Final Report Summary - INTERPREGGEN (Genetic studies of pre-eclampsia in Central Asian and European populations)
Executive Summary:
InterPregGen investigated the genetic predisposition to pregnancy-related hypertension, through a collaborative project involving 12 partners from Europe (UK, Finland, Iceland, Norway) and Central Asia (Uzbekistan, Kazakhstan).
PROJECT AIMS AND OUTCOMES
Objective 1: To establish collaborative links between Central Asian and European research groups.
InterPregGen has enjoyed excellent scientific collaboration between Central Asian and European researchers.
a. On-line meetings of all partners have taken place monthly, and annual Collaborators meetings at venues in London, Helsinki and Oslo have provided opportunities for scientific and social interaction.
b. Training in good research practice, advanced genetic laboratory techniques, and statistical genetics have been provided through training and monitoring visits to Uzbekistan and Kazakhstan by researchers from the UK; through training placements for Central Asian researchers in London and Iceland; and through the provision of extensive on-line material for self-tuition.
c. Following a highly effective recruitment effort, biobanks incorporating DNA samples from 9700 Kazakhs and 7000 Uzbeks have been established in Kazakhstan and Uzbekistan as a resource for future collaborative research.
Objective 2: To identify inherited factors in women and their babies which predispose to pre-eclampsia.
InterPregGen has conducted genetic studies on DNA from 16,000 Central Asian subjects recruited in the course of the study, and existing DNA collections from European cohorts, comprising over 300,000 cases and controls.
a. Genome-wide chip genotyping was performed on 2,600 mother-baby pairs and 2,600 controls from Central Asia, and 2,500 affected babies from the UK and Iceland. The data were supplemented with previously analysed maternal genome-wide data from 6,500 European pregnancies. This is the largest ever genome-wide study of pre-eclamptic women, and the first screen of the fetal genome in pre-eclampsia.
b. A Central Asian reference dataset was generated by whole genome sequencing of 100 Kazakh and 100 Uzbek individuals. This has not been undertaken previously in any Central Asian population.
c. Following genotype imputation over 7 million DNA sequence variants were tested for evidence of association with pre-eclampsia, firstly within individual cohorts, then in the combined results generated by meta-analysis.
d. A region of the fetal genome is associated with pre-eclampsia in both European and Central Asian populations, the first reported evidence implicating fetal DNA sequence variants in pre-eclampsia.
e. Multiple DNA variants in the maternal genome predispose to pre-eclampsia. Many of these variants are also associated with essential hypertension, Type 2 diabetes and coronary artery disease.
f. Clinical and genetic data have been assembled into a single harmonised database and used to model a pre-eclampsia risk-prediction tool. This requires field testing in a clinical setting.
MAJOR IMPACTS
a. Identification of DNA variants in mothers and their offspring which predispose to pre-eclampsia has potential to provide novel insights into disease mechanisms, and improve prevention and treatment of pre-eclampsia.
b. Genome-wide association screening data and whole genome sequencing data will be deposited on managed-access online repositories for future use by researchers.
c. Central Asian biobanks will be available for further collaborative research.
d. A pre-eclampsia risk-prediction tool has been modelled and is available for testing in a clinical setting.
Project Context and Objectives:
SUMMARY DESCRIPTION OF PROJECT CONTEXT AND OBJECTIVES
InterPregGen addresses the aims and objectives of HEALTH.2011.2.1.1.-4: Population genetics studies on cardio-metabolic disorders in EU/AC and EECA populations, specifically the genetic predisposition to pregnancy-related hypertension, through a collaborative project involving leading research groups in this field in EU (UK, Finland) AC (Iceland, Norway) and EECA (Uzbekistan, Kazakhstan) countries.
Pregnancy-related hypertension (high blood pressure) is recognised as one of the most serious disorders of pregnancy, endangering the lives of pregnant women and their babies. About 10% of women develop hypertension during pregnancy; in about 3% of women this is associated with the loss of protein in the urine (proteinuria), a sign that the kidneys are also involved. This is the most severe form of pregnancy-related hypertension, known as pre-eclampsia, which claims the lives of an estimated 50,000 women across the world each year. Medication may control the mother’s blood pressure, but there is only one cure for pre-eclampsia – delivery of the baby. The baby is often small, due to poor growth in the uterus, and in severe cases of pre-eclampsia may have to be delivered prematurely. The combined risks of poor intrauterine growth and prematurity due to pre-eclampsia kill an estimated 900,000 babies annually.
Pre-eclampsia usually develops after the 20th week of pregnancy, but in its early stages women may be unaware that all is not well. A priority for good antenatal care is to identify those women at highest risk of pre-eclampsia at an early stage so that the health of mother and baby can be carefully monitored and managed. Women at high risk of pre-eclampsia include those who had a previous affected pregnancy, and women with predisposing medical conditions, including essential hypertension, diabetes, kidney disease, and obesity.
Women with a family history of pre-eclampsia are also at increased risk, and this observation is an important motivation for the InterPregGen study. A woman whose mother had pre-eclampsia is two or three times more likely than other women to develop pre-eclampsia in her own pregnancy. This has led to a search for changes in DNA which may predispose to pre-eclampsia, and be handed on from mother to daughter. What has been recognised more recently is that men are more likely to father a pre-eclamptic pregnancy if they were themselves born of a pre-eclamptic pregnancy. This raises the possibility that DNA which the father passes on to his baby may also affect the risk of pre-eclampsia. Searching for changes in DNA in mothers, fathers and babies of families affected by pre-eclampsia may give some clues to the causes of this condition, which are currently incompletely understood.
In spite of over 2 decades of research, the results of genetic studies of pre-eclampsia have been inconsistent. Two important factors have contributed to the uncertainty surrounding the genetic basis of pre-eclampsia: most studies have included insufficient numbers of patients and controls to detect DNA sequence variants with individually small effects; and very few studies have investigated the DNA variants in offspring of pre-eclamptic pregnancies. The InterPregGen project has studied many thousands of women affected by pre-eclampsia from Western Europe and Central Asia, together with their partners and babies, using the technique of genome-wide association screening (GWAS) to detect inherited variations in their DNA. A total of 15,600 women and 8,600 babies from pregnancies affected by pre-eclampsia have been tested, and their DNA make-up compared with over 300,000 controls. This is the largest study of the genetic basis of pre-eclampsia ever undertaken, and provides a unique opportunity to extend our understanding of this condition, and also to develop long-lasting research partnerships between scientists and clinicians in Europe and Central Asia.
The central aims of the project are
1. To identify the inherited factors in women and their babies which predispose to pre-eclampsia in Central Asian and European populations.
2. To establish sustainable and productive collaborative links between Central Asian and European research groups through training programmes, exchange visits and the creation of DNA research biobanks in Central Asia.
3. To develop a predictive tool for pre-eclampsia and its complications
DESCRIPTION OF WORK PERFORMED AND MAIN RESULTS
➢ The Research Team
The InterPregGen consortium comprises researchers from Europe and Central Asia with a wide range of expertise, including clinicians, scientists, bioethicists, geneticists, data analysts and IT specialists. Research groups from 4 European countries are represented in InterPregGen: Finland, Iceland, Norway, and the United Kingdom. The ethnic diversity of the study is extended by collaboration with research groups from Kazakhstan and Uzbekistan. These populations bring a unique dimension to InterPregGen. They have been poorly represented in previous studies of pre-eclampsia, in spite of the higher incidence of the disorder in these countries. During the five years since the consortium was assembled the group has held annual face-to-face meetings, most recently in London in March 2016. Monthly on-line meetings have helped to bring a strong sense of collective energy to the search for inherited factors which predispose to pre-eclampsia, and an opportunity to share challenges and their solutions.
➢ Who is taking part in the study?
InterPregGen benefits from access to DNA samples collected in Europe from affected families for earlier studies of pre-eclampsia, and suitable control samples for comparison. The pioneering nature of this research in Central Asia called for the recruitment of volunteers to provide new samples. Kazakhstan and Uzbekistan each had recruitment targets of 2000 pre-eclamptic women, babies and partners and 2000 non-hypertensive pregnant control women. These targets have been achieved or exceeded in both countries, establishing the first pre-eclampsia DNA biobanks in Central Asia.
➢ Training research teams
Training has been a vital element in building a strong, well-informed team to carry out our research programme. The University of Leeds (UK) has extensive experience in conducting clinical and genetic research projects, and assembled a team of research clinicians who undertook training of researchers in Central Asia in the vital ethical aspects of clinical research, and other elements of good research practice. This has been a two-way learning process as research personnel gained familiarity with the health care systems in operation in other countries and their traditions. The Central Asian leaders have in turn shared this training with their research teams, and maintained regular audit of their activities, ensuring that recruitment, recording of clinical data, and collection of samples for DNA analysis have been performed to the highest research and ethical standards.
Training courses in genetic data analysis and advanced molecular laboratory techniques have been provided by the London School of Hygiene and Tropical Medicine, deCODE Genetics, and the Wellcome Trust Sanger Institute. The training has been cascaded by team leaders in Central Asia to their research teams, establishing a sound basis for future genetic research in their home countries.
➢ Analysing the DNA
InterPregGen has used GWAS to identify variations in DNA in mothers and offspring from pre-eclamptic pregnancies. This involved using a genotyping chip to establish the DNA sequence at up to 1 million locations in the genome, known as single nucleotide polymorphisms (SNPs). InterPregGen benefitted from the availability of previous GWAS results from pre-eclamptic mothers and controls in Iceland and the UK. Further GWAS analysis has been completed on over 1000 Icelandic babies, 1000 UK babies, and 2700 pre-eclamptic mothers-baby pairs and 2700 controls from Central Asia.
➢ Data processing
The value of the results of chip genotyping has been enhanced by the process of genotype imputation, a statistical technique which provides a dense map of the DNA variation in each individual. Imputation requires a reference set of complete DNA sequence information for individuals from the same population. International projects such as 1000Genomes have yielded publicly available sequence data from many populations, but they do not include any from Central Asia. InterPregGen has addressed this deficiency by sequencing the whole genome of 100 Uzbek and 100 Kazakh subjects, and using the results in association with data from Phase 3 of the 1000 Genomes project to assist in accurate genotype imputation in Central Asian samples. This provided the data analysts with millions of results summarising DNA variation within each individual, which have been compared in cases and controls. The results from each population have been combined by the process of meta-analysis, which increased the power of the study to detect differences in DNA sequence between cases and controls, and successfully identified DNA variations which occur at increased frequency in mothers and babies with pre-eclampsia.
A special feature of InterPregGen has been the search for rare DNA variants associated with pre-eclampsia. Such variation is not well detected using the GWAS approach, but we have benefitted from the rich source of DNA sequence information available from the Icelandic population. Although a number of rare DNA variants in Icelandic mothers and babies appeared to associate with pre-eclampsia, these were not confirmed in subsequent rounds of DNA testing.
➢ Replicating the results
Differences in DNA sequence between cases and controls detected by GWAS may simply be chance findings. We have therefore checked these results in completely independent sets of European samples from Norway, Finland and Denmark, and Central Asian samples from Uzbekistan and Kazakhstan. We have also examined the DNA in fathers of babies affected by pre-eclampsia to determine whether the effect of a gene variant in the baby is dependent on the parent from which the variant was inherited, a phenomenon known as genetic imprinting.
➢ Main results of genetic investigations
These investigations have revealed several DNA variants in women which are associated with hypertension in pregnancy. Some of these variations have been associated previously with blood pressure in non-pregnant individuals, Type 2 diabetes, and coronary vascular disease, all conditions which occur at increased frequency in later life in women with a history of pre-eclampsia. There is also persuasive evidence for DNA variants in the baby which predispose to pre-eclampsia. This is the first time the baby’s DNA has been extensively investigated in pre-eclampsia, and these results are entirely novel. There was no evidence that the association of fetal DNA variants with pre-eclampsia was influenced by genetic imprinting.
➢ Implications for clinical practice
A clear aim of the InterPregGen consortium is to use the findings of the study to improve the clinical approach to pre-eclampsia. The opportunities to use the results of this study in the clinical arena include the development of computer software tools which combine clinical and genetic risk factors from an individual woman to assess her risk of developing pre-eclampsia, or its severe complications, later in pregnancy. To achieve this we have incorporated clinical and genetic data from thousands of pregnancies into a single database, constructing a powerful resource for creating and testing risk prediction models.
Further benefits arising from the results of this study are the insights which these are providing to the mechanisms of pre-eclampsia and its short- and long-term complications. This is vital information in the search for more effective strategies for the prevention and treatment of pre-eclampsia.
➢ Dissemination of information
InterPregGen has a bilingual website, in English and Russian, at www.interpreggen.org This provides information on the aims and design of the study, together with news updates. Publications including information leaflets, and scientific papers and abstracts, can be downloaded from this site. A password-protected area of the website has been used throughout the study for sharing confidential documents between partners.
InterPregGen researchers have disseminated details of the study via their institutional websites, local press, radio and television, and have presented papers at local and international scientific meetings and published results in academic journals.
IMPACT
The primary aim of the InterPregGen study, the identification of genetic variations in mother and baby which predispose to pre-eclampsia, has been achieved, and results will be published in 2017. These results provide new insights into the mechanism of pre-eclampsia, and will stimulate novel areas of research, ultimately leading to new approaches to the prediction, prevention and treatment of pre-eclampsia.
We have developed a model for risk prediction in pre-eclampsia, using a combination of clinical and genetic information. The model will require field testing in an unselected population, but if successful would have considerable implications for health care, both in effective targeting of limited antenatal care resources to those at highest risk, and in guiding management choices in severely affected pregnancies.
A major aim of this project has been the development of sustainable research links between scientists and clinicians in Europe and Central Asia. This has been a two-way process, as European colleagues have become more aware of the Central Asian health care and research infrastructure, and Kazakh and Uzbek partners have become acquainted with research practices familiar to Europeans. Biobanks have been established in Kazakhstan and Uzbekistan, incorporating many thousands of DNA samples and associated clinical data. This places Central Asian researchers in a strong position to bid for future funding for collaborative research. The value of this resource is greatly enhanced by the availability of whole genome sequence data generated by the InterPregGen project from 100 Kazakh and 100 Uzbek individuals. Results will be made available through managed on-line databases to all bona fide researchers, and will prove an invaluable resource for future genetic studies in Central Asia.
ACKNOWLEDGMENTS
The InterPregGen consortium would like to thank all the women and their families who have generously participated in our preeclampsia research (past and present).
Project Results:
The overarching aim of the InterPregGen project has been to further understanding of the causes and mechanisms of the hypertensive disorders of pregnancy, with the main focus on the inherited predisposition to pre-eclampsia. Over the five years of the project, InterPregGen has:
1. Established collaborative links with an extensive network of researchers active in the field of pre-eclampsia research in Europe, Central Asia, and South America.
2. Enhanced research capacity in Uzbekistan and Kazakhstan through investment in infrastructure and training in good research practice.
3. Recruited over 16,000 volunteers from two Central Asian countries for research into pre-eclampsia.
4. Established biobanks in Uzbekistan and Kazakhstan for future genetic research.
5. Completed the first whole genome sequence of 100 Kazakh and 100 Uzbek individuals.
6. Conducted genome-wide association screening (GWAS) of over 9000 European and Central Asian women affected by pre-eclampsia; confirmed significant findings in 8 independent sample sets; and identified DNA sequence variants in mothers which confer susceptibility to pre-eclampsia.
7. Conducted the first genome-wide association screen of the fetal genome in pre-eclampsia in 5,000 offspring of pre-eclamptic pregnancies, and confirmed significant findings in independent sample sets. This has led to the identification of DNA sequence variants in the fetus which predispose to pre-eclamptic pregnancies.
8. Developed models of pre-eclampsia pathogenesis based on the results of genetic studies.
9. Developed a computer-based tool for prediction of pre-eclampsia risk from clinical and genetic data available in the first trimester of pregnancy.
10. Conducted the study in compliance with European Commission ethical standards, as advised by the Ethics Review Report and an independent ethics consultant.
1. Collaboration.
From its inception InterPregGen has been a collaborative project, involving research groups from the UK, Iceland, Norway, Finland, Kazakhstan and Uzbekistan. The European groups had worked together previously, and had access to substantial volumes of background research data and sample collections. Their ready willingness to share these valuable resources played a significant part in the successful outcome of the project, as it was possible to commence analysis of GWAS data from Icelandic and UK women from the outset.
By contrast the 3 Central Asian research groups (2 in Uzbekistan, 1 in Kazakhstan) had limited experience of working with European researchers, and were less familiar with European research practices. Their enthusiasm was evident during the 18 months of preparation for the start of the project, and a rapport between researchers was established at the kick-off meeting in London in January 2012. The relationship has been cemented through monthly online conferences of all partners, and annual consortium meetings. Training courses and placements have been particularly successful in establishing strong bonds between European and Central Asian teams. Language has provided some challenges which have been successfully addressed, partly through the use of interpreters, but largely because of the willingness of Central Asian colleagues to sharpen their English language skills.
InterPregGen has been widely recognised as a well-designed and well-conducted project, and has consequently been able to widen the network of collaborators providing valuable data and samples for the project. These include the UK ALSPAC study and the Danish Birth Cohort study. Parallel projects have generated valuable additional genetic data which has been generously shared with InterPregGen, notably GWAS data from the Norwegian MoBa study. InterPregGen’s partnership with researchers from the Wellcome Trust Case Control Consortium 2 (WTCCC2) has provided access to GWAS data from 2000 maternal cases and controls from Colombia. In consequence the study now covers the 3 continents of Europe, Central Asia and South America. The importance of this ethnic diversity cannot be overstated; whilst pre-eclampsia remains an important cause of maternal and perinatal death in Europe, its incidence is considerably higher in populations from Central Asia and South America. In spite of this disparity in the health and economic burdens imposed by pre-eclampsia, most previous genetic research into the condition has been conducted in populations of European origin.
2. Enhanced research capacity in Uzbekistan and Kazakhstan.
InterPregGen has invested in laboratory and computing infrastructure in Uzbekistan and Kazakhstan, and has prioritised training in good research practice, laboratory techniques, and advanced statistical analysis of genetic data. Laboratory equipment purchases have included -80C freezers and racking, automated pipettes, GeneQuant DNA spectrophotometers, and a Pyromark pyrosequencer and work station. Computing facilities have been enhanced through the purchase of laptop computers for field researchers, and desk top computers for advanced data analysis.
Training in good research practice, including ethical issues, recruitment of volunteers, and adherence to standard operating procedures (SOPs), has been provided by an experienced team from the University of Leeds, UK. This work started in January 2012 with a two-day course in Leeds for Central Asian researchers. This established the differences in health service provision and research experience and practice in Central Asian centres, and was followed by an intensive 9-month period of preparation and translation of research documentation and SOPs, and subsequent applications for Research Ethics Committee approval. Although this delayed the start of recruitment by several months the time was well-spent in establishing a firm basis for high-quality research. A team from Leeds made annual training and monitoring visits to Central Asia throughout the study, to audit recruitment and data collection, monitor close-down procedures at recruitment centres, and verify secure storage of hard-copy and electronic data. The training was well-received, and was cascaded to researchers at outlying centres. It was particularly pleasing that the visits were used as an opportunity for leading obstetricians from the UK to share their clinical experience with local clinicians, and to observe clinical practice in Uzbekistan and Kazakhstan. The team was accompanied by Dr Wai Kwong Lee from the University of Glasgow, UK who provided training and advice on laboratory procedures, specifically related to follow-up genotyping to be undertaken in the Central Asian laboratories.
3. Recruitment of volunteers in Uzbekistan and Kazakhstan
The majority of research into the genetics of pre-eclampsia has been conducted in populations of Western European origin; no such studies have been conducted in Central Asian populations, in spite of the higher incidence of pre-eclampsia in these countries. It was therefore an important aim of InterPregGen to recruit Central Asian women affected by pre-eclampsia, their babies and partners, and healthy pregnant control women for genetic studies. To achieve adequate statistical power to detect DNA variants with modest effects on the risk of pre-eclampsia a target of 2000 affected family trios and 2000 healthy pregnant women was set for Kazakhstan and Uzbekistan, a total of 4000 cases and 4000 controls. The target was achieved in Uzbekistan and exceeded in Kazakhstan, where 2600 cases and controls were recruited. This was a remarkable achievement by research teams with no previous experience of recruitment for genetic studies on this scale. Substantial time and effort were devoted during the first 9-12 months of the project to training research teams, developing standard operating procedures, and piloting recruitment.
Recruitment was established in 7 maternity centres in Kazakhstan: 3 in Almaty, 2 in Shymkent (724 km from Almaty), and one each in Taraz (547 km from Almaty) and Taldykorgan (270 km from Almaty). Volunteers were all ethnic Kazakhs by grandparental ethnicity. In Uzbekistan, recruitment of volunteers took place in 13 maternity units, located in Tashkent (7 units), Namangan, Jizzakh, Karshi, Chirchik, Samarkand and Navoi. Volunteers were ethic Uzbeks by grandparental ethnicity. All volunteers provided written informed consent to genetic studies. Each participant provided detailed clinical and obstetric information which was entered into an online clinical database. A venous blood sample for DNA extraction was taken from pre-eclamptic and control women, and partners of pre-eclamptic women. Fetal DNA was extracted from umbilical venous blood or a fragment of umbilical cord. Together the Uzbek and Kazakh teams have established the largest collection of samples and data for genetic studies of pre-eclampsia in Central Asia, and indeed is one of the largest collections in the world.
4. Biobanks in Uzbekistan and Kazakhstan.
A primary aim of the InterPregGen project was the creation of biobanks in Kazakhstan and Uzbekistan for future genetic studies of a range of complex disorders. Limited biobanked resources for genetic studies in complex human diseases were previously available in these countries. InterPregGen has therefore provided infrastructure requirements for biobank sustainability, including -800C freezers with uninterruptable power supplies, freezer racking and bar coding equipment for sample tracking. Biobanks are now established in Kazakhstan, comprising 9,700 DNA samples from ethnic Kazakhs, and in Uzbekistan, comprising 7,413 DNA samples from ethnic Uzbeks. The collections include samples from genetic studies of pre-eclampsia and samples collected for whole genome sequencing.
These are the largest biobank resources in Kazakhstan and Uzbekistan, and are supported by high-quality clinical databases, genome-wide genotyping data from a total of 3000 subjects in each country, and whole genome sequencing data from 100 Uzbeks and 100 Kazakhs. These features ensure the international standing of the biobanks, making them attractive to researchers seeking partners for collaborative ventures. A priority for InterPregGen has been the development of robust governance protocols for the biobanks, recognising the technical and ethical dimensions of good management of the resource. We have worked closely with biobanking experts within the consortium, and consulted widely with leading authorities in Europe, to develop biobank governance protocols which recognise the evolving nature of the field.
5. The first whole genome sequence of Kazakhs and Uzbeks.
Sequencing the entire genome of multiple individuals from different populations through projects such as 1000 Genomes has provided fascinating insights into the ancestry of ethnic groups, produced catalogues of rare genetic variants, and provided essential reference sequence for the full exploitation of genome-wide array genotyping. Amongst the main reference panels providing whole genome sequence data or dense genome-wide genotype data (HapMap, 1000Genomes, Human Genome Diversity panel), none includes populations from Central Asia. InterPregGen has addressed this deficiency by conducting whole genome sequencing of 100 Kazakh and 100 Uzbek volunteers, equally divided between males and females. This has achieved two main purposes. Firstly, the whole genome sequence data from Kazakhstan and Uzbekistan has provided reference sequences for each population, enabling InterPregGen to derive the maximum information from genome-wide association studies of pre-eclampsia. Secondly, the whole genome sequence data from Central Asia will provide a future resource useable by others conducting genetic studies in Central Asia, and will be of value to genetic researchers with a more general interest in population genetics.
Analysis of the whole genome sequencing data has revealed correlations in DNA variants seen in Central Asians and both Asian and European populations, reflecting their geographical location and their position on the historical trading route between China and Europe, the Silk Road. An added dimension to the sequencing project in Kazakhs is their descent from three historical hordes, or zhus. Among the Kazakh volunteers, the percentage of each Zhu was approximately 25%, 30% and 45% for the “younger”, “middle” and “older” Zhus, respectively. Exploring whether each Zhu is reflected in the DNA sequence has produced some interesting results which will be published in the scientific literature. In addition the whole genome sequencing data will be deposited at the European Genome-phenome Archive, and will be made available through managed access to bona fide researchers.
6. Genome-wide association screening identifies DNA variants in mothers which predispose to pre-eclampsia
It has long been recognised that there is a familial predisposition to pre-eclampsia. Women whose mother or sisters were affected by pre-eclampsia are at increased risk of developing the condition. Furthermore, men who were born of a pre-eclamptic pregnancy are more likely to father a pregnancy which is affected by the disorder. The overarching objective of InterPregGen was the identification of DNA variants which affect the risk of developing pre-eclampsia. This has been a target for researchers for over two decades, resulting in many hundreds of published research papers. Unfortunately none of the reported variants has been reproducibly associated with pre-eclampsia. This has been a familiar problem with many common disorders, such as coronary heart disease and diabetes, where multiple DNA variants with individually small effects influence the risk of disease. Reliable detection of such variants requires large studies, involving thousands of individuals. It has also become apparent that causal DNA variants do not necessarily occur within obvious candidate genes, or indeed within DNA regions coding for proteins. The answer has been to conduct a scan of the entire genome, through the technique of genome-wide association screening (GWAS). This involves genotyping hundreds of thousands of DNA variants, known as single nucleotide polymorphisms (SNPs), which together capture the majority of variation in the genome. InterPregGen conducted GWAS of over 9000 European and Central Asian women affected by pre-eclampsia, and confirmed significant findings in 8 independent sample sets, comprising over 18,000 samples.
InterPregGen benefitted at the outset from access to existing maternal GWAS data from 3 European cohorts: the Icelandic deCODE dataset, UK GOPEC, and UK ALSPAC data. Data were analysed by the data analysis team at the Wellcome Trust Sanger Institute in the UK, and deCODE Genetics in Iceland. The technique of genotype imputation was used to extend the available information, using data from directly genotyped SNPs to infer the genotype at hundreds of thousands of SNPs which had not been directly genotyped. The results of GWAS in these three datasets were combined through meta-analysis. This identified several SNPs with suggestive evidence for association with pregnancy-induced hypertension. Interestingly, a number of these SNPs had been associated previously with hypertension unrelated to pregnancy. To test for genuine association with disease, over 100 SNPs showing the strongest association with pregnancy-induced hypertension were genotyped in independent sample sets of cases and controls from Finland (the FINNPEC collection, made available by the University of Helsinki), Norway (HUNT and MoBa biobanks, made available by the Norwegian University of Science and Technology and the National Institute of Public Health, Norway, respectively) and Denmark (Danish Birth Cohort). The replication testing confirmed the GWAS association signals in 3 regions of the maternal genome. Individually each variant increases a woman’s risk of developing pre-eclampsia by approximately 10%.
During the analysis of European maternal data, recruitment of volunteers in Uzbekistan and Kazakhstan was on-going. When 1000 women and offspring affected by pre-eclampsia and 1000 healthy pregnant controls had been recruited, DNA samples were shipped to deCODE Genetics and the Wellcome Trust Sanger Institute for genome-wide chip genotyping, using the Illumina Omni 2.5-8 chip array, which genotypes approximately 2.5 million SNPs. After elimination of genotyped samples that failed QC criteria, 920 Uzbek maternal cases and 924 controls and 864 Kazakh maternal cases and 923 controls were included in further data analysis. Genotypes were imputed with IMPUTE2 software using reference data from the 1000 Genomes project and the newly-available InterPregGen whole genome sequencing data for Central Asia, yielding approximately 8 million DNA variants that were subjected to GWAS analysis using SNPTEST software. Both the Kazakh and the Uzbek GWAS results showed a number of association signals of suggestive significance (P < 1 x 10-5) for which evidence of genuine disease association with pre-eclampsia would be strengthened if there were a corresponding association signal in the alternative Central Asian dataset and/or European datasets.
Meta-analysis of the 2 Central Asian GWAS results with the results of the 3 European GWAS datasets was performed. Following re-imputation of European datasets using the latest version of whole genome sequence data from the 1000 genomes project, 7.5 million SNPs and 8 million indels genotyped or imputed in all 5 datasets were available for meta-analysis. This revealed a number of suggestive association signals (P < 1 x 10-6) which were prioritised for more testing by genotyping implicated DNA variants in independent sample collections from Europe and Central Asia. These have not so far yielded any new variants with confirmed association with pre-eclampsia in addition to those discovered earlier in the European datasets. However our data analysts have recently been given access to substantial additional genome-wide data generated by 4 other pre-eclampsia research projects, increasing the statistical power for the discovery of associated DNA variants, and they will continue to explore the maternal genome beyond the end of the InterPregGen project.
As a complementary approach to understanding the role of maternal genetic variants, InterPregGen conducted a search for risk variants which pre-eclampsia shares with other common diseases. The principle is to identify the large mass of DNA variants with provisional but inconclusive evidence for association in GWAS of pre-eclampsia and then test whether the same mass of alleles also exhibits association with a second disease. This is done by applying a single statistical test that evaluates the mass of variants in aggregate and which can yield highly statistically significant evidence that shared variants predispose to both diseases, even though the method does not identify individual variants among the large mass of tested variants. Three such aggregate-tests (polygenic score analysis (PGSA); genome-wide complex trait analysis (GCTA); and linkage disequilibrium score regression (LDSR)) were applied to the InterPregGen meta-analysis results from maternal European cases and controls in order to test for overlap of susceptibility variants for pre-eclampsia and other disease phenotypes. The results of these tests provided strong evidence that pre-eclampsia susceptibility alleles are shared with essential hypertension, cardiovascular disease and type II diabetes.
7. First genome-wide association screen of offspring of pre-eclamptic pregnancies identifies fetal genetic variants which predispose to maternal disease
Relatively limited genome-wide data from the offspring of pre-eclamptic pregnancies were available to InterPregGen at the start of the study. To study the fetal genome in Europeans, an additional 605 Icelandic subjects were recruited for GWAS, providing a total of 980 offspring of pre-eclamptic pregnancies. Genome-wide chip genotyping was performed on Icelandic and 1063 UK subjects using on the Illumina Omni-Express chip which comprises a total of 730,525 SNPs. Following QC, genotype imputation was performed using IMPUTE2 software, yielding 8 million genotyped and imputed SNP genotypes for GWAS. Association analysis was performed on each dataset individually, then combined through meta-analysis, augmented by available GWAS data from the UK ALSPAC dataset. These data were subsequently combined with the results of GWAS of 2000 newly-collected Central Asian pre-eclamptic offspring and controls.
The results revealed a region of chromosome 13 which is strongly associated with pre-eclampsia. This region was tested in 2 independent European sample sets, from Norway and Finland, and independent collections of a further 1000 cases and controls from each of the two participating Central Asian countries. The replication testing results demonstrate a DNA sequence variant close to the gene Fms-like tyrosine kinase 1 (FLT1) which is strongly associated with pre-eclampsia (P = 5.38x10-11). The position of the SNP near FLT1 provides important support for the biological significance of the result, since considerable evidence implicates the protein sFlt-1, which is encoded by FLT1, in the aetiology of pre-eclampsia. This is the first report of genetic variation in the fetus leading to disease of the mother in pregnancy.
8. New models of pre-eclampsia pathogenesis.
In spite of the heavy burden which pre-eclampsia imposes on maternal and fetal health, its causes are incompletely defined. Pre-eclampsia is understood to have its origins in early pregnancy, at the time of placental implantation and growth, long before the onset of clinical disease. This is a technically and ethically difficult period in which to obtain tissues for basic scientific research. Studies conducted later in pregnancy, when the disease process is established, suffer from uncertainty about whether observed pathophysiological changes are causal, or secondary phenomena. Genetics offers a powerful tool in the search for the underlying disease mechanisms in disorders of pregnancy, as the DNA of mother and baby is present at the outset of pregnancy, is unchanged by the disease process, and is available for study when the woman and her baby are restored to health.
An entirely novel finding from the genetic studies of pre-eclampsia conducted by InterPregGen is the presence of a strong association signal in the fetal genome within an enhancer region of the FLT1 gene. FLT1 encodes Fms-like tyrosine kinase 1 (Flt-1), a receptor for the angiogenic factors vascular endothelial growth factor (VEGF) and placental growth factor (PlGF). The soluble form of this receptor, sFlt-1, is believed to be of placental origin, and circulates in maternal plasma during pregnancy. sFlt-1 levels increase as pregnancy progresses, and are markedly elevated in women with established pre-eclampsia. The finding that the fetal genotype at the FLT1 locus is associated with pre-eclampsia expands the hypotheses developing around the role of sFlt-1. A widely held view is that excess sFlt-1 is released from a damaged placenta and binds the angiogenic factors VEGF and PlGF in the maternal circulation, resulting in vascular endothelial dysfunction. The genetic results from InterPregGen strongly suggest that factors encoded by FLT1 play a primary, aetiological role in pre-eclampsia, and are not released secondary to placental damage only. Data analysis has shown that the association at this locus with pre-eclampsia is limited to the fetal genotype; there is no evidence for association with the maternal FLT1 genotype, supporting a placental origin for the high circulating sFlt-1 seen in pre-eclampsia. The fetal FLT1 gene has been indirectly implicated previously through the increased incidence of PE, and high maternal plasma levels of sFlt-1, in women carrying a fetus with 3 copies of this gene due to trisomy 13. The location of the SNP showing the strongest association with pre-eclampsia within a gene enhancer region raises the intriguing possibility that fetal (and therefore placental) FLT1 gene expression is affected from very early in pregnancy, possibly with effects on placental angiogenesis. An unexpected finding is that DNA variants at the site of FLT1 are associated with late-onset pre-eclampsia (after 34 weeks gestation) but not with early-onset pre-eclampsia. This is contrary to the clinical observation that the highest levels of sFlt-1 are seen in women with early-onset pre-eclampsia, and suggests that variants near FLT1 may have a novel or modified mechanism of action on pre-eclampsia risk. This opens up avenues for further research.
The extensive collections of samples and data available to InterPregGen have enabled the identification of several DNA sequence variants in the maternal genome which are associated with pre-eclampsia, with modest individual effect sizes in the region of 10% (i.e. odds ratio ~ 1.10). Most of these variants have been described previously in association with (non-gestational) essential hypertension. InterPregGen has benefitted from access to unpublished GWAS data from ~3000 UK and Icelandic women with (non-proteinuric) gestational hypertension. Although both gestational hypertension and pre-eclampsia are hypertensive disorders of pregnancy, their clinical course differs, with gestational hypertension usually having a better outcome than pre-eclampsia. For this reason they have often been considered as aetiologically distinct conditions. The evidence from InterPregGen challenges this view, as pre-eclampsia and gestational hypertension are associated with the same genetic variants.
We conclude that pre-eclampsia, gestational hypertension and essential hypertension are overlapping conditions, not only in their clinical manifestations, but also in their aetiology. This accounts at least in part for the increased susceptibility of women to develop pre-eclampsia if they have a history of essential hypertension. It also suggests a genetic basis for the epidemiological observation that women with a history of pre-eclampsia are at increased risk of essential hypertension later in life.
Through the technique of LD score analysis, InterPregGen has studied the extent of sharing of sequence variants with other genetically complex disorders. This method of analysis does not identify specific shared variants, but examines the entire genome for evidence of genetic correlation between phenotypes. In addition to the sharing of susceptibility genes with essential hypertension, as described above, there is significant genetic correlation between pre-eclampsia and coronary artery disease and type II diabetes. Like hypertension, type 2 diabetes is a risk factor for pre-eclampsia. Furthermore, women with a history of pre-eclampsia have an increased risk of developing type II diabetes and coronary artery disease in later life. The clinical overlap between these disorders appears to be due at least in part to shared genetic determinants.
9. A tool for pre-eclampsia risk prediction.
Effective management of pre-eclampsia depends on prompt diagnosis. This in turn relies on good antenatal care, with regular monitoring of blood pressure. The identification at an early stage in pregnancy of those women at highest risk would enable efficient and cost-effective targeting of antenatal care resources. Currently risk prediction is based on clinical data, including personal and family history; trials of biomarkers are on-going. InterPregGen has explored whether incorporation of genetic data available in the first trimester would improve predictive models for pre-eclampsia. As an initial step the University of Helsinki coordinated the assembly of a database of clinical and genetic data from the participating cohorts: GOPEC (UK), FINNPEC (Finland), MoBa (Norway), HUNT (Norway), Uzbekistan, Kazakhstan and data from a large case-control cohort from Colombia (GenPE). Initial preparatory work included extensive quality control and harmonisation of clinical data from disparate clinical databases. The final database incorporates details of 13,294 control pregnancies and 13,511 pregnancies affected by pre-eclampsia, including maternal, fetal and paternal data where available.
Using statistical and exploratory data analysis, genetic and clinical variables with evidence of association with preeclampsia and available in the first trimester were selected for development of a predictive tool. The analyses were planned using variables with data available for at least 50% of each cohort. Mothers who, prior to 20 weeks of gestation, were affected by essential hypertension, renal disease, type 1 or 2 diabetes mellitus, or excreting protein in the urine were excluded.
Using computational bootstrapping analysis, selected variables were ranked in order of the strength of their association with pre-eclampsia. Available data were randomly divided into 70% ‘training’ and 30% ‘testing’ data. Predictive models were developed from training data using the high-ranked variables in a cumulative manner, and the best predictive model(s) were identified using the model quality assessment criteria: adjusted R2 and Akaike’s Information Criterion (AIC). These models were tested on the testing data. Two cohorts, Icelandic and Colombian, were reserved for external validation of the models. Using prediction accuracy analysis, the performance of the models was assessed by calculating the area under the curve (AUC) of the receiver operating characteristic (ROC).
The evidence from all the models tested was that clinical data alone has stronger predictive power than genetic data alone, but a combination of clinical and genetic data provides the best predictive model. A model based on 10 clinical and 2 genetic variables yielded the best predictive accuracy (AUC: 0.9). The predictive tool requires field testing in a clinical setting, and will need to evolve as additional biomarkers are identified by other research groups.
10. Conducted the study in compliance European Commission ethical standards.
The study has been conducted in compliance with European Commission standards of research ethics, following the requirements of the Ethics Review Report, and advised by Dr. Nicola Stingelin, our independent ethics consultant. A number of ethical issues have arisen as a result of differences in clinical and research practice between Europe and Uzbekistan and Kazakhstan. Mechanisms for obtaining informed consent for participation in genetic research were carefully reviewed to ensure that they met the standards which apply in the European Union. Both Central Asian countries have existing research ethics structures and guidelines, including established procedures for applying for Research Ethics Committee (REC) approval. Local REC approval was obtained by all partners, European and Central Asian, at the start of the study to cover their individual contributions to the project. After some initial difficulty in identifying a REC which was willing to review the project as a whole, over-arching approval was obtained from the University of Nottingham REC. The co-ordinating centre has maintained the ethics database, including updates of Research Ethics Committee approval as required. Emphasis has been placed throughout on protection of the privacy of participants, by coding or anonymising clinical and genetic data, using secure sites for the transfer of electronic data, and ensuring that electronic and hard copy data are stored securely. European and Central Asian partners have developed publication policies, data sharing policies, and biobank governance procedures, which will be made publicly available via the InterPregGen website. These may prove of value to other researchers embarking on research projects involving collaborative working between Central Asian and European countries.
Potential Impact:
Potential Impact
• InterPregGen has completed the largest genome-wide association study of pre-eclampsia conducted to date, and achieved its primary aim of identification of DNA sequence variants in mothers and their offspring which predispose to disease. The findings are novel and convincingly reproducible in independent sample collections. The past two decades of research into the molecular genetics of pre-eclampsia have seen considerable investment of energy and resources without producing definitive results. InterPregGen has shown that the DNA of many thousands of affected individuals and controls must be examined in order to detect the sequence variants which predispose to disease, as each variant individually has a small effect on disease risk. This should provide a strong message to researchers that effort should be invested not in conducting small inconclusive case-control studies, but in building large, well-phenotyped collections of samples which can provide definitive evidence for genetic associations.
The ability to merge genetic data from multiple large sample collections through meta-analysis has been key to InterPregGen’s successful identification of variants associated with pre-eclampsia. The limit of detection of genetic variants of small effect is determined mainly by the size of the study, and ever-increasing sample sizes from meta-analyses will likely identify further variants associated with pre-eclampsia and/or its complications. To facilitate future use of InterPregGen data in meta-analysis GWAS data will be deposited in the European Genome-phenome Archive (www.ebi.ac.uk/ega/) with managed access for bona fide researchers.
The motivation for InterPregGen has been and remains the application of genetic discoveries to improved pregnancy outcomes, based on a better understanding of pre-eclampsia and its complications. The genetic discoveries have stimulated the articulation of novel hypotheses of the molecular basis of pre-eclampsia and its subtypes, opening prospects for further research and modifications to clinical practice. One key discovery has been the evidence that pre-eclampsia shares genetic association with gestational hypertension, essential hypertension, coronary artery disease, and type 2 diabetes. This provides a firm molecular basis for the observed clinical overlap between these conditions. The challenge now is to tease out the shared pathological mechanisms through which the genetic variants are acting, and develop a molecular classification of these conditions which can guide disease prevention and therapy.
InterPregGen is the first study to conduct a thorough investigation of fetal DNA variants in pre-eclampsia. The finding of DNA variants close to the FLT1 gene, which encodes Flt-1 and its soluble isoform sFlt-1, is of considerable interest in view of the established association of high sFlt-1 levels in women with pre-eclampsia. The sFlt-1 in the maternal circulation is believed to be of placental origin, which is consistent with involvement of the fetal gene. What is unexplained is the observation that fetal FLT1 genotype is associated only with late-onset pre-eclampsia, as maternal sFlt-1 is regarded as a marker for early-onset disease. The future challenge for researchers is to explore how modest alterations in Flt-1 or sFlt-1 expression from the start of pregnancy could cause changes, perhaps in the formation of the placenta, which are manifested as pre-eclampsia in late pregnancy.
The prospect of predicting in early pregnancy which women are at highest risk of pre-eclampsia is attractive, as this would enable the effective targeting of limited ante-natal care resources. In common with most other genetically complex disorders, genetic data alone has inadequate predictive power for pre-eclampsia. Nevertheless there is evidence that a combination of clinical and genetic data has improved predictive value over clinical data alone, and this merits further testing in a clinical setting. InterPregGen has shown that the DNA variant with the strongest effect on pre-eclampsia risk is in the fetal, not the maternal, genome. The fetal genotype is currently not readily available antenatally in the majority of pregnancies. However, techniques which reliably isolate fetal DNA from maternal venous blood are developing rapidly, and this holds out the prospect that fetal genotype could be included in a model with better predictive value for pre-eclampsia.
• InterPregGen was designed to establish sustainable and productive collaborative links between Central Asian and European research groups. Close collaboration has been a strong feature of the project, and will yield benefits for future research. Partners in Uzbekistan and Kazakhstan had not previously been involved in genetic research on the scale of InterPregGen. By sharing expertise in good research practice, advanced genetic laboratory techniques, and genetic data analysis, InterPregGen has ensured that Central Asian colleagues are conversant with research practices in Europe, and are trained to a high level in managing complex research procedures and analysing high level genetic data. Equally important has been the insights which European partners have obtained into health care systems in Uzbekistan and Kazakhstan. Together these factors have established mutual understanding and a vital infrastructure for further collaborative research. This has been underpinned by investment in infrastructure hardware in Uzbekistan and Kazakhstan, including -80C freezers and racking for long-term sample storage, sample bar-coding facilities, DNA spectrophotometers, genotyping equipment, and computers.
Sample biobanks have been established in Kazakhstan and Uzbekistan containing approximately 13,000 DNA samples for genetic research from ethnic Kazakhs and Uzbeks, supported by a high-quality clinical database, and genetic data including genome-wide association screening data. The availability of the biobanks for collaborative research will be advertised through the InterPregGen website and in academic publications arising from the project. These are amongst the largest biobanked resources for genetic research into pre-eclampsia in the world, and will attract international interest from researchers in the field.
InterPregGen has generated the first whole genome sequence of 100 ethnic Kazakhs and 100 Uzbeks, which will be deposited at the European Genome-phenome Archive for managed access to researchers. The importance of the whole genome data lies in the lack of any such datasets from Central Asians on public databases such as 1000Genomes. The deposited whole genome data will be of interest to scientists working in the field of population genetics, and to researchers into genetic disorders in Central Asian populations.
Dissemination activities
• Throughout the 5 years of the project InterPregGen partners have promoted the study at presitigious international scientific meetings, including the European Society for Human Genetics, the American Society for Human Genetics, and the International Society for the Study of Hypertension in Pregnancy. Abstracts at scientific meetings and academic papers published to date are:
1. L. Morgan, H. Laivuori; on behalf of the InterPregGen Consortium. Genetic studies of pre-eclampsia in Central Asian and European population-Information on the InterPregGen Study funded by the Seventh Framework Programme of the European Community. XVIII International Society for the Study of Hypertension in Pregnancy (ISSHP) World Congress, July 2012, Geneva
2. Sally Chappell, Wai Kwong Lee, Scott Shooter, Ralph McGinnis, Linda Morgan on behalf of the InterPregGen Consortium. Evaluation of GWAS chip-genotyping of fetal DNA extracted from umbilical cord tissue and WGA-DNA extracted from filter paper bloodspots. European Society for Human Genetics conference, June 2013, Paris
3. Morgan L on behalf of the InterPregGen consortium. Genetic studies of pre-eclampsia in Central Asian and European populations: a progress report on the InterPregGen study. International Society for the Study of Hypertension in Pregnancy (ISSHP) European Congress, June 2013, Tromsø
4. Svyatova G; Zakhidova N; Morgan L; Walker JJ; on behalf of the InterPregGen Consortium. Clinical features of pre-eclampsia in 2613 Central Asian women and babies recruited for genetic studies. XIX ISSHP World Congress, October 2014, New Orleans
5. Morgan L, McGinnis R, Steinthorsdottir V, Svyatova G, Zakhidova N, Kwong Lee W, Iversen AC, Magnus P, Walker JJ, Casas JP, Sultanov S, Laivuori H, on behalf of the InterPregGen consortium. InterPregGen: genetic studies of pre-eclampsia in three continents. Norsk Epidemiologi 2014; 24 (1-2): 141-146
6. N. O. Williams, F. Dudbridge, S. Chappell, C. Franklin, R. McGinnis, On behalf of the InterPregGen Consortium. Investigation of the polygenic genetics of Pre-eclampsia and its relationship with other phenotypes. European Society for Human Genetics conference, June 2015, Glasgow
7. Ralph McGinnis, Frank Dudbridge, Debbie Lawlor, Chris Franklin, Nicholas Williams. On behalf of the InterPregGen Consortium. Investigation of DNA variants responsible for Pre-eclampsia. American Society for Human Genetics annual meeting, October 2015, Baltimore
8. Svyatova G, Berezina G, Salimbaeva D, Choroshilova I, Uspanova M, Dolbakova S. First international Intergen DNA bank in the Republic of Kazakhstan. 4th International Conference, May 2016, Nazarbayev University, Astana
9. McGinnis R, Steinthorsdottir V, Williams N et al. First genome-wide significant locus for pre-eclampsia susceptibility discovered on fetal chromosome 13. American Society of Human Genetics annual meeting, October 2016, Vancouver
10. McGinnis R, Steinthorsdottir V, Williams N et al. First sequence variants associated with pre-eclampsia identified in fetal genome. Nature Genetics 2016. Manuscript under consideration
• The InterPregGen team is continuing to prepare manuscripts for submission to prestigious academic journals. Manuscripts in preparation include:
11. Genome-wide association study identifies multiple loci associated with pre-eclampsia in affected mothers
12. Whole genome sequencing of Central Asian populations from Kazakhstan and Uzbekistan
13. Clinical features of pre-eclampsia in Uzbekistan and Kazakhstan
14. The challenges and rewards of research collaboration between European and Central Asian countries
15. Genome-wide scanning in Central Asian mothers and babies affected by pre-eclampsia
16. Development of risk prediction models for pre-eclampsia incorporating genotyping data
• Bilingual (English/Russian) website.
The bilingual InterPregGen website www.interpreggen.org is maintained by the University of Nottingham. Confidential material has been deposited on the password-protected researchers’ area of the website throughout the study. With the end of the project, the password-protected area will be removed, and selected non-confidential study documentation will be moved to a publicly accessible researchers’ page. This will include standard operating procedures, information and consent forms, clinical record booklets and protocols developed in the course of InterPregGen for use by researchers planning similar studies in future. The public-facing website will remain active for a minimum of 2 years after the end of the project.
• Wider public dissemination
Partners have provided invited talks to public interest groups describing and promoting the study, and disseminated information leaflets in clinics in Kazakhstan and Uzbekistan. Examples can be downloaded from the InterPregGen website www.interpreggen.org. Training and monitoring visits by the University of Leeds team to Uzbekistan and Kazakhstan have provided opportunities for open lectures and discussions with local obstetricians covering clinical guidelines for pre-eclampsia. InterPregGen is promoted on individual institutional websites, and partners have provided interviews for local radio and television. Major InterPregGen publications will be accompanied by co-ordinated releases to local and national media by institutional press offices.
List of Websites:
WEBSITE: www.interpreggen.org
CONTACT DETAILS
University of Nottingham, United Kingdom Dr Linda Morgan linda.morgan@nottingham.ac.uk
Scientific Center for Gynecology, Obstetrics and Perinatology, Kazakhstan Prof.Gulnara Svyatova respmgk@mail.ru
Republic Specialized ScientificPractical Medical Center of Obstetrics and Gynecology, Uzbekistan Dr Nodira Zakhidova odira_z@yahoo.com
deCODE Genetics, Iceland . Dr Valgerdur Steinthorsdottir Valgerdur.steinthorsdottir@decode.is
University of Helsinki, Finland. Prof. Hannele Laivuori Hannele.laivuori@helsinki.fi
Norwegian University of Science and Technology, Norway. Dr Ann-Charlotte Iversen ann-charlotte.iversen@ntnu.no
National Institute of Public Health, Norway. Prof. Per Magnus Per.magnus@fhi.no
London School of Hygiene and Tropical Medicine, United Kingdom. Dr Frank Dudbridge Frank.dudbridge@lshtm.ac.uk
Wellcome Trust Sanger Institute, United Kingdom. Dr Ralph McGinnis rm2@sanger.ac.uk
University of Glasgow, United Kingdom. Prof. Anna Dominiczak a.dominiczak@clinmed.gla.ac.uk
University of Leeds, United Kingdom. Prof. James Walker j.j.walker@leeds.ac.uk
Institute of Immunology, Uzbek Academy of Science, Uzbekistan. Dr Nodira Zakhidova odira_z@yahoo.com
InterPregGen investigated the genetic predisposition to pregnancy-related hypertension, through a collaborative project involving 12 partners from Europe (UK, Finland, Iceland, Norway) and Central Asia (Uzbekistan, Kazakhstan).
PROJECT AIMS AND OUTCOMES
Objective 1: To establish collaborative links between Central Asian and European research groups.
InterPregGen has enjoyed excellent scientific collaboration between Central Asian and European researchers.
a. On-line meetings of all partners have taken place monthly, and annual Collaborators meetings at venues in London, Helsinki and Oslo have provided opportunities for scientific and social interaction.
b. Training in good research practice, advanced genetic laboratory techniques, and statistical genetics have been provided through training and monitoring visits to Uzbekistan and Kazakhstan by researchers from the UK; through training placements for Central Asian researchers in London and Iceland; and through the provision of extensive on-line material for self-tuition.
c. Following a highly effective recruitment effort, biobanks incorporating DNA samples from 9700 Kazakhs and 7000 Uzbeks have been established in Kazakhstan and Uzbekistan as a resource for future collaborative research.
Objective 2: To identify inherited factors in women and their babies which predispose to pre-eclampsia.
InterPregGen has conducted genetic studies on DNA from 16,000 Central Asian subjects recruited in the course of the study, and existing DNA collections from European cohorts, comprising over 300,000 cases and controls.
a. Genome-wide chip genotyping was performed on 2,600 mother-baby pairs and 2,600 controls from Central Asia, and 2,500 affected babies from the UK and Iceland. The data were supplemented with previously analysed maternal genome-wide data from 6,500 European pregnancies. This is the largest ever genome-wide study of pre-eclamptic women, and the first screen of the fetal genome in pre-eclampsia.
b. A Central Asian reference dataset was generated by whole genome sequencing of 100 Kazakh and 100 Uzbek individuals. This has not been undertaken previously in any Central Asian population.
c. Following genotype imputation over 7 million DNA sequence variants were tested for evidence of association with pre-eclampsia, firstly within individual cohorts, then in the combined results generated by meta-analysis.
d. A region of the fetal genome is associated with pre-eclampsia in both European and Central Asian populations, the first reported evidence implicating fetal DNA sequence variants in pre-eclampsia.
e. Multiple DNA variants in the maternal genome predispose to pre-eclampsia. Many of these variants are also associated with essential hypertension, Type 2 diabetes and coronary artery disease.
f. Clinical and genetic data have been assembled into a single harmonised database and used to model a pre-eclampsia risk-prediction tool. This requires field testing in a clinical setting.
MAJOR IMPACTS
a. Identification of DNA variants in mothers and their offspring which predispose to pre-eclampsia has potential to provide novel insights into disease mechanisms, and improve prevention and treatment of pre-eclampsia.
b. Genome-wide association screening data and whole genome sequencing data will be deposited on managed-access online repositories for future use by researchers.
c. Central Asian biobanks will be available for further collaborative research.
d. A pre-eclampsia risk-prediction tool has been modelled and is available for testing in a clinical setting.
Project Context and Objectives:
SUMMARY DESCRIPTION OF PROJECT CONTEXT AND OBJECTIVES
InterPregGen addresses the aims and objectives of HEALTH.2011.2.1.1.-4: Population genetics studies on cardio-metabolic disorders in EU/AC and EECA populations, specifically the genetic predisposition to pregnancy-related hypertension, through a collaborative project involving leading research groups in this field in EU (UK, Finland) AC (Iceland, Norway) and EECA (Uzbekistan, Kazakhstan) countries.
Pregnancy-related hypertension (high blood pressure) is recognised as one of the most serious disorders of pregnancy, endangering the lives of pregnant women and their babies. About 10% of women develop hypertension during pregnancy; in about 3% of women this is associated with the loss of protein in the urine (proteinuria), a sign that the kidneys are also involved. This is the most severe form of pregnancy-related hypertension, known as pre-eclampsia, which claims the lives of an estimated 50,000 women across the world each year. Medication may control the mother’s blood pressure, but there is only one cure for pre-eclampsia – delivery of the baby. The baby is often small, due to poor growth in the uterus, and in severe cases of pre-eclampsia may have to be delivered prematurely. The combined risks of poor intrauterine growth and prematurity due to pre-eclampsia kill an estimated 900,000 babies annually.
Pre-eclampsia usually develops after the 20th week of pregnancy, but in its early stages women may be unaware that all is not well. A priority for good antenatal care is to identify those women at highest risk of pre-eclampsia at an early stage so that the health of mother and baby can be carefully monitored and managed. Women at high risk of pre-eclampsia include those who had a previous affected pregnancy, and women with predisposing medical conditions, including essential hypertension, diabetes, kidney disease, and obesity.
Women with a family history of pre-eclampsia are also at increased risk, and this observation is an important motivation for the InterPregGen study. A woman whose mother had pre-eclampsia is two or three times more likely than other women to develop pre-eclampsia in her own pregnancy. This has led to a search for changes in DNA which may predispose to pre-eclampsia, and be handed on from mother to daughter. What has been recognised more recently is that men are more likely to father a pre-eclamptic pregnancy if they were themselves born of a pre-eclamptic pregnancy. This raises the possibility that DNA which the father passes on to his baby may also affect the risk of pre-eclampsia. Searching for changes in DNA in mothers, fathers and babies of families affected by pre-eclampsia may give some clues to the causes of this condition, which are currently incompletely understood.
In spite of over 2 decades of research, the results of genetic studies of pre-eclampsia have been inconsistent. Two important factors have contributed to the uncertainty surrounding the genetic basis of pre-eclampsia: most studies have included insufficient numbers of patients and controls to detect DNA sequence variants with individually small effects; and very few studies have investigated the DNA variants in offspring of pre-eclamptic pregnancies. The InterPregGen project has studied many thousands of women affected by pre-eclampsia from Western Europe and Central Asia, together with their partners and babies, using the technique of genome-wide association screening (GWAS) to detect inherited variations in their DNA. A total of 15,600 women and 8,600 babies from pregnancies affected by pre-eclampsia have been tested, and their DNA make-up compared with over 300,000 controls. This is the largest study of the genetic basis of pre-eclampsia ever undertaken, and provides a unique opportunity to extend our understanding of this condition, and also to develop long-lasting research partnerships between scientists and clinicians in Europe and Central Asia.
The central aims of the project are
1. To identify the inherited factors in women and their babies which predispose to pre-eclampsia in Central Asian and European populations.
2. To establish sustainable and productive collaborative links between Central Asian and European research groups through training programmes, exchange visits and the creation of DNA research biobanks in Central Asia.
3. To develop a predictive tool for pre-eclampsia and its complications
DESCRIPTION OF WORK PERFORMED AND MAIN RESULTS
➢ The Research Team
The InterPregGen consortium comprises researchers from Europe and Central Asia with a wide range of expertise, including clinicians, scientists, bioethicists, geneticists, data analysts and IT specialists. Research groups from 4 European countries are represented in InterPregGen: Finland, Iceland, Norway, and the United Kingdom. The ethnic diversity of the study is extended by collaboration with research groups from Kazakhstan and Uzbekistan. These populations bring a unique dimension to InterPregGen. They have been poorly represented in previous studies of pre-eclampsia, in spite of the higher incidence of the disorder in these countries. During the five years since the consortium was assembled the group has held annual face-to-face meetings, most recently in London in March 2016. Monthly on-line meetings have helped to bring a strong sense of collective energy to the search for inherited factors which predispose to pre-eclampsia, and an opportunity to share challenges and their solutions.
➢ Who is taking part in the study?
InterPregGen benefits from access to DNA samples collected in Europe from affected families for earlier studies of pre-eclampsia, and suitable control samples for comparison. The pioneering nature of this research in Central Asia called for the recruitment of volunteers to provide new samples. Kazakhstan and Uzbekistan each had recruitment targets of 2000 pre-eclamptic women, babies and partners and 2000 non-hypertensive pregnant control women. These targets have been achieved or exceeded in both countries, establishing the first pre-eclampsia DNA biobanks in Central Asia.
➢ Training research teams
Training has been a vital element in building a strong, well-informed team to carry out our research programme. The University of Leeds (UK) has extensive experience in conducting clinical and genetic research projects, and assembled a team of research clinicians who undertook training of researchers in Central Asia in the vital ethical aspects of clinical research, and other elements of good research practice. This has been a two-way learning process as research personnel gained familiarity with the health care systems in operation in other countries and their traditions. The Central Asian leaders have in turn shared this training with their research teams, and maintained regular audit of their activities, ensuring that recruitment, recording of clinical data, and collection of samples for DNA analysis have been performed to the highest research and ethical standards.
Training courses in genetic data analysis and advanced molecular laboratory techniques have been provided by the London School of Hygiene and Tropical Medicine, deCODE Genetics, and the Wellcome Trust Sanger Institute. The training has been cascaded by team leaders in Central Asia to their research teams, establishing a sound basis for future genetic research in their home countries.
➢ Analysing the DNA
InterPregGen has used GWAS to identify variations in DNA in mothers and offspring from pre-eclamptic pregnancies. This involved using a genotyping chip to establish the DNA sequence at up to 1 million locations in the genome, known as single nucleotide polymorphisms (SNPs). InterPregGen benefitted from the availability of previous GWAS results from pre-eclamptic mothers and controls in Iceland and the UK. Further GWAS analysis has been completed on over 1000 Icelandic babies, 1000 UK babies, and 2700 pre-eclamptic mothers-baby pairs and 2700 controls from Central Asia.
➢ Data processing
The value of the results of chip genotyping has been enhanced by the process of genotype imputation, a statistical technique which provides a dense map of the DNA variation in each individual. Imputation requires a reference set of complete DNA sequence information for individuals from the same population. International projects such as 1000Genomes have yielded publicly available sequence data from many populations, but they do not include any from Central Asia. InterPregGen has addressed this deficiency by sequencing the whole genome of 100 Uzbek and 100 Kazakh subjects, and using the results in association with data from Phase 3 of the 1000 Genomes project to assist in accurate genotype imputation in Central Asian samples. This provided the data analysts with millions of results summarising DNA variation within each individual, which have been compared in cases and controls. The results from each population have been combined by the process of meta-analysis, which increased the power of the study to detect differences in DNA sequence between cases and controls, and successfully identified DNA variations which occur at increased frequency in mothers and babies with pre-eclampsia.
A special feature of InterPregGen has been the search for rare DNA variants associated with pre-eclampsia. Such variation is not well detected using the GWAS approach, but we have benefitted from the rich source of DNA sequence information available from the Icelandic population. Although a number of rare DNA variants in Icelandic mothers and babies appeared to associate with pre-eclampsia, these were not confirmed in subsequent rounds of DNA testing.
➢ Replicating the results
Differences in DNA sequence between cases and controls detected by GWAS may simply be chance findings. We have therefore checked these results in completely independent sets of European samples from Norway, Finland and Denmark, and Central Asian samples from Uzbekistan and Kazakhstan. We have also examined the DNA in fathers of babies affected by pre-eclampsia to determine whether the effect of a gene variant in the baby is dependent on the parent from which the variant was inherited, a phenomenon known as genetic imprinting.
➢ Main results of genetic investigations
These investigations have revealed several DNA variants in women which are associated with hypertension in pregnancy. Some of these variations have been associated previously with blood pressure in non-pregnant individuals, Type 2 diabetes, and coronary vascular disease, all conditions which occur at increased frequency in later life in women with a history of pre-eclampsia. There is also persuasive evidence for DNA variants in the baby which predispose to pre-eclampsia. This is the first time the baby’s DNA has been extensively investigated in pre-eclampsia, and these results are entirely novel. There was no evidence that the association of fetal DNA variants with pre-eclampsia was influenced by genetic imprinting.
➢ Implications for clinical practice
A clear aim of the InterPregGen consortium is to use the findings of the study to improve the clinical approach to pre-eclampsia. The opportunities to use the results of this study in the clinical arena include the development of computer software tools which combine clinical and genetic risk factors from an individual woman to assess her risk of developing pre-eclampsia, or its severe complications, later in pregnancy. To achieve this we have incorporated clinical and genetic data from thousands of pregnancies into a single database, constructing a powerful resource for creating and testing risk prediction models.
Further benefits arising from the results of this study are the insights which these are providing to the mechanisms of pre-eclampsia and its short- and long-term complications. This is vital information in the search for more effective strategies for the prevention and treatment of pre-eclampsia.
➢ Dissemination of information
InterPregGen has a bilingual website, in English and Russian, at www.interpreggen.org This provides information on the aims and design of the study, together with news updates. Publications including information leaflets, and scientific papers and abstracts, can be downloaded from this site. A password-protected area of the website has been used throughout the study for sharing confidential documents between partners.
InterPregGen researchers have disseminated details of the study via their institutional websites, local press, radio and television, and have presented papers at local and international scientific meetings and published results in academic journals.
IMPACT
The primary aim of the InterPregGen study, the identification of genetic variations in mother and baby which predispose to pre-eclampsia, has been achieved, and results will be published in 2017. These results provide new insights into the mechanism of pre-eclampsia, and will stimulate novel areas of research, ultimately leading to new approaches to the prediction, prevention and treatment of pre-eclampsia.
We have developed a model for risk prediction in pre-eclampsia, using a combination of clinical and genetic information. The model will require field testing in an unselected population, but if successful would have considerable implications for health care, both in effective targeting of limited antenatal care resources to those at highest risk, and in guiding management choices in severely affected pregnancies.
A major aim of this project has been the development of sustainable research links between scientists and clinicians in Europe and Central Asia. This has been a two-way process, as European colleagues have become more aware of the Central Asian health care and research infrastructure, and Kazakh and Uzbek partners have become acquainted with research practices familiar to Europeans. Biobanks have been established in Kazakhstan and Uzbekistan, incorporating many thousands of DNA samples and associated clinical data. This places Central Asian researchers in a strong position to bid for future funding for collaborative research. The value of this resource is greatly enhanced by the availability of whole genome sequence data generated by the InterPregGen project from 100 Kazakh and 100 Uzbek individuals. Results will be made available through managed on-line databases to all bona fide researchers, and will prove an invaluable resource for future genetic studies in Central Asia.
ACKNOWLEDGMENTS
The InterPregGen consortium would like to thank all the women and their families who have generously participated in our preeclampsia research (past and present).
Project Results:
The overarching aim of the InterPregGen project has been to further understanding of the causes and mechanisms of the hypertensive disorders of pregnancy, with the main focus on the inherited predisposition to pre-eclampsia. Over the five years of the project, InterPregGen has:
1. Established collaborative links with an extensive network of researchers active in the field of pre-eclampsia research in Europe, Central Asia, and South America.
2. Enhanced research capacity in Uzbekistan and Kazakhstan through investment in infrastructure and training in good research practice.
3. Recruited over 16,000 volunteers from two Central Asian countries for research into pre-eclampsia.
4. Established biobanks in Uzbekistan and Kazakhstan for future genetic research.
5. Completed the first whole genome sequence of 100 Kazakh and 100 Uzbek individuals.
6. Conducted genome-wide association screening (GWAS) of over 9000 European and Central Asian women affected by pre-eclampsia; confirmed significant findings in 8 independent sample sets; and identified DNA sequence variants in mothers which confer susceptibility to pre-eclampsia.
7. Conducted the first genome-wide association screen of the fetal genome in pre-eclampsia in 5,000 offspring of pre-eclamptic pregnancies, and confirmed significant findings in independent sample sets. This has led to the identification of DNA sequence variants in the fetus which predispose to pre-eclamptic pregnancies.
8. Developed models of pre-eclampsia pathogenesis based on the results of genetic studies.
9. Developed a computer-based tool for prediction of pre-eclampsia risk from clinical and genetic data available in the first trimester of pregnancy.
10. Conducted the study in compliance with European Commission ethical standards, as advised by the Ethics Review Report and an independent ethics consultant.
1. Collaboration.
From its inception InterPregGen has been a collaborative project, involving research groups from the UK, Iceland, Norway, Finland, Kazakhstan and Uzbekistan. The European groups had worked together previously, and had access to substantial volumes of background research data and sample collections. Their ready willingness to share these valuable resources played a significant part in the successful outcome of the project, as it was possible to commence analysis of GWAS data from Icelandic and UK women from the outset.
By contrast the 3 Central Asian research groups (2 in Uzbekistan, 1 in Kazakhstan) had limited experience of working with European researchers, and were less familiar with European research practices. Their enthusiasm was evident during the 18 months of preparation for the start of the project, and a rapport between researchers was established at the kick-off meeting in London in January 2012. The relationship has been cemented through monthly online conferences of all partners, and annual consortium meetings. Training courses and placements have been particularly successful in establishing strong bonds between European and Central Asian teams. Language has provided some challenges which have been successfully addressed, partly through the use of interpreters, but largely because of the willingness of Central Asian colleagues to sharpen their English language skills.
InterPregGen has been widely recognised as a well-designed and well-conducted project, and has consequently been able to widen the network of collaborators providing valuable data and samples for the project. These include the UK ALSPAC study and the Danish Birth Cohort study. Parallel projects have generated valuable additional genetic data which has been generously shared with InterPregGen, notably GWAS data from the Norwegian MoBa study. InterPregGen’s partnership with researchers from the Wellcome Trust Case Control Consortium 2 (WTCCC2) has provided access to GWAS data from 2000 maternal cases and controls from Colombia. In consequence the study now covers the 3 continents of Europe, Central Asia and South America. The importance of this ethnic diversity cannot be overstated; whilst pre-eclampsia remains an important cause of maternal and perinatal death in Europe, its incidence is considerably higher in populations from Central Asia and South America. In spite of this disparity in the health and economic burdens imposed by pre-eclampsia, most previous genetic research into the condition has been conducted in populations of European origin.
2. Enhanced research capacity in Uzbekistan and Kazakhstan.
InterPregGen has invested in laboratory and computing infrastructure in Uzbekistan and Kazakhstan, and has prioritised training in good research practice, laboratory techniques, and advanced statistical analysis of genetic data. Laboratory equipment purchases have included -80C freezers and racking, automated pipettes, GeneQuant DNA spectrophotometers, and a Pyromark pyrosequencer and work station. Computing facilities have been enhanced through the purchase of laptop computers for field researchers, and desk top computers for advanced data analysis.
Training in good research practice, including ethical issues, recruitment of volunteers, and adherence to standard operating procedures (SOPs), has been provided by an experienced team from the University of Leeds, UK. This work started in January 2012 with a two-day course in Leeds for Central Asian researchers. This established the differences in health service provision and research experience and practice in Central Asian centres, and was followed by an intensive 9-month period of preparation and translation of research documentation and SOPs, and subsequent applications for Research Ethics Committee approval. Although this delayed the start of recruitment by several months the time was well-spent in establishing a firm basis for high-quality research. A team from Leeds made annual training and monitoring visits to Central Asia throughout the study, to audit recruitment and data collection, monitor close-down procedures at recruitment centres, and verify secure storage of hard-copy and electronic data. The training was well-received, and was cascaded to researchers at outlying centres. It was particularly pleasing that the visits were used as an opportunity for leading obstetricians from the UK to share their clinical experience with local clinicians, and to observe clinical practice in Uzbekistan and Kazakhstan. The team was accompanied by Dr Wai Kwong Lee from the University of Glasgow, UK who provided training and advice on laboratory procedures, specifically related to follow-up genotyping to be undertaken in the Central Asian laboratories.
3. Recruitment of volunteers in Uzbekistan and Kazakhstan
The majority of research into the genetics of pre-eclampsia has been conducted in populations of Western European origin; no such studies have been conducted in Central Asian populations, in spite of the higher incidence of pre-eclampsia in these countries. It was therefore an important aim of InterPregGen to recruit Central Asian women affected by pre-eclampsia, their babies and partners, and healthy pregnant control women for genetic studies. To achieve adequate statistical power to detect DNA variants with modest effects on the risk of pre-eclampsia a target of 2000 affected family trios and 2000 healthy pregnant women was set for Kazakhstan and Uzbekistan, a total of 4000 cases and 4000 controls. The target was achieved in Uzbekistan and exceeded in Kazakhstan, where 2600 cases and controls were recruited. This was a remarkable achievement by research teams with no previous experience of recruitment for genetic studies on this scale. Substantial time and effort were devoted during the first 9-12 months of the project to training research teams, developing standard operating procedures, and piloting recruitment.
Recruitment was established in 7 maternity centres in Kazakhstan: 3 in Almaty, 2 in Shymkent (724 km from Almaty), and one each in Taraz (547 km from Almaty) and Taldykorgan (270 km from Almaty). Volunteers were all ethnic Kazakhs by grandparental ethnicity. In Uzbekistan, recruitment of volunteers took place in 13 maternity units, located in Tashkent (7 units), Namangan, Jizzakh, Karshi, Chirchik, Samarkand and Navoi. Volunteers were ethic Uzbeks by grandparental ethnicity. All volunteers provided written informed consent to genetic studies. Each participant provided detailed clinical and obstetric information which was entered into an online clinical database. A venous blood sample for DNA extraction was taken from pre-eclamptic and control women, and partners of pre-eclamptic women. Fetal DNA was extracted from umbilical venous blood or a fragment of umbilical cord. Together the Uzbek and Kazakh teams have established the largest collection of samples and data for genetic studies of pre-eclampsia in Central Asia, and indeed is one of the largest collections in the world.
4. Biobanks in Uzbekistan and Kazakhstan.
A primary aim of the InterPregGen project was the creation of biobanks in Kazakhstan and Uzbekistan for future genetic studies of a range of complex disorders. Limited biobanked resources for genetic studies in complex human diseases were previously available in these countries. InterPregGen has therefore provided infrastructure requirements for biobank sustainability, including -800C freezers with uninterruptable power supplies, freezer racking and bar coding equipment for sample tracking. Biobanks are now established in Kazakhstan, comprising 9,700 DNA samples from ethnic Kazakhs, and in Uzbekistan, comprising 7,413 DNA samples from ethnic Uzbeks. The collections include samples from genetic studies of pre-eclampsia and samples collected for whole genome sequencing.
These are the largest biobank resources in Kazakhstan and Uzbekistan, and are supported by high-quality clinical databases, genome-wide genotyping data from a total of 3000 subjects in each country, and whole genome sequencing data from 100 Uzbeks and 100 Kazakhs. These features ensure the international standing of the biobanks, making them attractive to researchers seeking partners for collaborative ventures. A priority for InterPregGen has been the development of robust governance protocols for the biobanks, recognising the technical and ethical dimensions of good management of the resource. We have worked closely with biobanking experts within the consortium, and consulted widely with leading authorities in Europe, to develop biobank governance protocols which recognise the evolving nature of the field.
5. The first whole genome sequence of Kazakhs and Uzbeks.
Sequencing the entire genome of multiple individuals from different populations through projects such as 1000 Genomes has provided fascinating insights into the ancestry of ethnic groups, produced catalogues of rare genetic variants, and provided essential reference sequence for the full exploitation of genome-wide array genotyping. Amongst the main reference panels providing whole genome sequence data or dense genome-wide genotype data (HapMap, 1000Genomes, Human Genome Diversity panel), none includes populations from Central Asia. InterPregGen has addressed this deficiency by conducting whole genome sequencing of 100 Kazakh and 100 Uzbek volunteers, equally divided between males and females. This has achieved two main purposes. Firstly, the whole genome sequence data from Kazakhstan and Uzbekistan has provided reference sequences for each population, enabling InterPregGen to derive the maximum information from genome-wide association studies of pre-eclampsia. Secondly, the whole genome sequence data from Central Asia will provide a future resource useable by others conducting genetic studies in Central Asia, and will be of value to genetic researchers with a more general interest in population genetics.
Analysis of the whole genome sequencing data has revealed correlations in DNA variants seen in Central Asians and both Asian and European populations, reflecting their geographical location and their position on the historical trading route between China and Europe, the Silk Road. An added dimension to the sequencing project in Kazakhs is their descent from three historical hordes, or zhus. Among the Kazakh volunteers, the percentage of each Zhu was approximately 25%, 30% and 45% for the “younger”, “middle” and “older” Zhus, respectively. Exploring whether each Zhu is reflected in the DNA sequence has produced some interesting results which will be published in the scientific literature. In addition the whole genome sequencing data will be deposited at the European Genome-phenome Archive, and will be made available through managed access to bona fide researchers.
6. Genome-wide association screening identifies DNA variants in mothers which predispose to pre-eclampsia
It has long been recognised that there is a familial predisposition to pre-eclampsia. Women whose mother or sisters were affected by pre-eclampsia are at increased risk of developing the condition. Furthermore, men who were born of a pre-eclamptic pregnancy are more likely to father a pregnancy which is affected by the disorder. The overarching objective of InterPregGen was the identification of DNA variants which affect the risk of developing pre-eclampsia. This has been a target for researchers for over two decades, resulting in many hundreds of published research papers. Unfortunately none of the reported variants has been reproducibly associated with pre-eclampsia. This has been a familiar problem with many common disorders, such as coronary heart disease and diabetes, where multiple DNA variants with individually small effects influence the risk of disease. Reliable detection of such variants requires large studies, involving thousands of individuals. It has also become apparent that causal DNA variants do not necessarily occur within obvious candidate genes, or indeed within DNA regions coding for proteins. The answer has been to conduct a scan of the entire genome, through the technique of genome-wide association screening (GWAS). This involves genotyping hundreds of thousands of DNA variants, known as single nucleotide polymorphisms (SNPs), which together capture the majority of variation in the genome. InterPregGen conducted GWAS of over 9000 European and Central Asian women affected by pre-eclampsia, and confirmed significant findings in 8 independent sample sets, comprising over 18,000 samples.
InterPregGen benefitted at the outset from access to existing maternal GWAS data from 3 European cohorts: the Icelandic deCODE dataset, UK GOPEC, and UK ALSPAC data. Data were analysed by the data analysis team at the Wellcome Trust Sanger Institute in the UK, and deCODE Genetics in Iceland. The technique of genotype imputation was used to extend the available information, using data from directly genotyped SNPs to infer the genotype at hundreds of thousands of SNPs which had not been directly genotyped. The results of GWAS in these three datasets were combined through meta-analysis. This identified several SNPs with suggestive evidence for association with pregnancy-induced hypertension. Interestingly, a number of these SNPs had been associated previously with hypertension unrelated to pregnancy. To test for genuine association with disease, over 100 SNPs showing the strongest association with pregnancy-induced hypertension were genotyped in independent sample sets of cases and controls from Finland (the FINNPEC collection, made available by the University of Helsinki), Norway (HUNT and MoBa biobanks, made available by the Norwegian University of Science and Technology and the National Institute of Public Health, Norway, respectively) and Denmark (Danish Birth Cohort). The replication testing confirmed the GWAS association signals in 3 regions of the maternal genome. Individually each variant increases a woman’s risk of developing pre-eclampsia by approximately 10%.
During the analysis of European maternal data, recruitment of volunteers in Uzbekistan and Kazakhstan was on-going. When 1000 women and offspring affected by pre-eclampsia and 1000 healthy pregnant controls had been recruited, DNA samples were shipped to deCODE Genetics and the Wellcome Trust Sanger Institute for genome-wide chip genotyping, using the Illumina Omni 2.5-8 chip array, which genotypes approximately 2.5 million SNPs. After elimination of genotyped samples that failed QC criteria, 920 Uzbek maternal cases and 924 controls and 864 Kazakh maternal cases and 923 controls were included in further data analysis. Genotypes were imputed with IMPUTE2 software using reference data from the 1000 Genomes project and the newly-available InterPregGen whole genome sequencing data for Central Asia, yielding approximately 8 million DNA variants that were subjected to GWAS analysis using SNPTEST software. Both the Kazakh and the Uzbek GWAS results showed a number of association signals of suggestive significance (P < 1 x 10-5) for which evidence of genuine disease association with pre-eclampsia would be strengthened if there were a corresponding association signal in the alternative Central Asian dataset and/or European datasets.
Meta-analysis of the 2 Central Asian GWAS results with the results of the 3 European GWAS datasets was performed. Following re-imputation of European datasets using the latest version of whole genome sequence data from the 1000 genomes project, 7.5 million SNPs and 8 million indels genotyped or imputed in all 5 datasets were available for meta-analysis. This revealed a number of suggestive association signals (P < 1 x 10-6) which were prioritised for more testing by genotyping implicated DNA variants in independent sample collections from Europe and Central Asia. These have not so far yielded any new variants with confirmed association with pre-eclampsia in addition to those discovered earlier in the European datasets. However our data analysts have recently been given access to substantial additional genome-wide data generated by 4 other pre-eclampsia research projects, increasing the statistical power for the discovery of associated DNA variants, and they will continue to explore the maternal genome beyond the end of the InterPregGen project.
As a complementary approach to understanding the role of maternal genetic variants, InterPregGen conducted a search for risk variants which pre-eclampsia shares with other common diseases. The principle is to identify the large mass of DNA variants with provisional but inconclusive evidence for association in GWAS of pre-eclampsia and then test whether the same mass of alleles also exhibits association with a second disease. This is done by applying a single statistical test that evaluates the mass of variants in aggregate and which can yield highly statistically significant evidence that shared variants predispose to both diseases, even though the method does not identify individual variants among the large mass of tested variants. Three such aggregate-tests (polygenic score analysis (PGSA); genome-wide complex trait analysis (GCTA); and linkage disequilibrium score regression (LDSR)) were applied to the InterPregGen meta-analysis results from maternal European cases and controls in order to test for overlap of susceptibility variants for pre-eclampsia and other disease phenotypes. The results of these tests provided strong evidence that pre-eclampsia susceptibility alleles are shared with essential hypertension, cardiovascular disease and type II diabetes.
7. First genome-wide association screen of offspring of pre-eclamptic pregnancies identifies fetal genetic variants which predispose to maternal disease
Relatively limited genome-wide data from the offspring of pre-eclamptic pregnancies were available to InterPregGen at the start of the study. To study the fetal genome in Europeans, an additional 605 Icelandic subjects were recruited for GWAS, providing a total of 980 offspring of pre-eclamptic pregnancies. Genome-wide chip genotyping was performed on Icelandic and 1063 UK subjects using on the Illumina Omni-Express chip which comprises a total of 730,525 SNPs. Following QC, genotype imputation was performed using IMPUTE2 software, yielding 8 million genotyped and imputed SNP genotypes for GWAS. Association analysis was performed on each dataset individually, then combined through meta-analysis, augmented by available GWAS data from the UK ALSPAC dataset. These data were subsequently combined with the results of GWAS of 2000 newly-collected Central Asian pre-eclamptic offspring and controls.
The results revealed a region of chromosome 13 which is strongly associated with pre-eclampsia. This region was tested in 2 independent European sample sets, from Norway and Finland, and independent collections of a further 1000 cases and controls from each of the two participating Central Asian countries. The replication testing results demonstrate a DNA sequence variant close to the gene Fms-like tyrosine kinase 1 (FLT1) which is strongly associated with pre-eclampsia (P = 5.38x10-11). The position of the SNP near FLT1 provides important support for the biological significance of the result, since considerable evidence implicates the protein sFlt-1, which is encoded by FLT1, in the aetiology of pre-eclampsia. This is the first report of genetic variation in the fetus leading to disease of the mother in pregnancy.
8. New models of pre-eclampsia pathogenesis.
In spite of the heavy burden which pre-eclampsia imposes on maternal and fetal health, its causes are incompletely defined. Pre-eclampsia is understood to have its origins in early pregnancy, at the time of placental implantation and growth, long before the onset of clinical disease. This is a technically and ethically difficult period in which to obtain tissues for basic scientific research. Studies conducted later in pregnancy, when the disease process is established, suffer from uncertainty about whether observed pathophysiological changes are causal, or secondary phenomena. Genetics offers a powerful tool in the search for the underlying disease mechanisms in disorders of pregnancy, as the DNA of mother and baby is present at the outset of pregnancy, is unchanged by the disease process, and is available for study when the woman and her baby are restored to health.
An entirely novel finding from the genetic studies of pre-eclampsia conducted by InterPregGen is the presence of a strong association signal in the fetal genome within an enhancer region of the FLT1 gene. FLT1 encodes Fms-like tyrosine kinase 1 (Flt-1), a receptor for the angiogenic factors vascular endothelial growth factor (VEGF) and placental growth factor (PlGF). The soluble form of this receptor, sFlt-1, is believed to be of placental origin, and circulates in maternal plasma during pregnancy. sFlt-1 levels increase as pregnancy progresses, and are markedly elevated in women with established pre-eclampsia. The finding that the fetal genotype at the FLT1 locus is associated with pre-eclampsia expands the hypotheses developing around the role of sFlt-1. A widely held view is that excess sFlt-1 is released from a damaged placenta and binds the angiogenic factors VEGF and PlGF in the maternal circulation, resulting in vascular endothelial dysfunction. The genetic results from InterPregGen strongly suggest that factors encoded by FLT1 play a primary, aetiological role in pre-eclampsia, and are not released secondary to placental damage only. Data analysis has shown that the association at this locus with pre-eclampsia is limited to the fetal genotype; there is no evidence for association with the maternal FLT1 genotype, supporting a placental origin for the high circulating sFlt-1 seen in pre-eclampsia. The fetal FLT1 gene has been indirectly implicated previously through the increased incidence of PE, and high maternal plasma levels of sFlt-1, in women carrying a fetus with 3 copies of this gene due to trisomy 13. The location of the SNP showing the strongest association with pre-eclampsia within a gene enhancer region raises the intriguing possibility that fetal (and therefore placental) FLT1 gene expression is affected from very early in pregnancy, possibly with effects on placental angiogenesis. An unexpected finding is that DNA variants at the site of FLT1 are associated with late-onset pre-eclampsia (after 34 weeks gestation) but not with early-onset pre-eclampsia. This is contrary to the clinical observation that the highest levels of sFlt-1 are seen in women with early-onset pre-eclampsia, and suggests that variants near FLT1 may have a novel or modified mechanism of action on pre-eclampsia risk. This opens up avenues for further research.
The extensive collections of samples and data available to InterPregGen have enabled the identification of several DNA sequence variants in the maternal genome which are associated with pre-eclampsia, with modest individual effect sizes in the region of 10% (i.e. odds ratio ~ 1.10). Most of these variants have been described previously in association with (non-gestational) essential hypertension. InterPregGen has benefitted from access to unpublished GWAS data from ~3000 UK and Icelandic women with (non-proteinuric) gestational hypertension. Although both gestational hypertension and pre-eclampsia are hypertensive disorders of pregnancy, their clinical course differs, with gestational hypertension usually having a better outcome than pre-eclampsia. For this reason they have often been considered as aetiologically distinct conditions. The evidence from InterPregGen challenges this view, as pre-eclampsia and gestational hypertension are associated with the same genetic variants.
We conclude that pre-eclampsia, gestational hypertension and essential hypertension are overlapping conditions, not only in their clinical manifestations, but also in their aetiology. This accounts at least in part for the increased susceptibility of women to develop pre-eclampsia if they have a history of essential hypertension. It also suggests a genetic basis for the epidemiological observation that women with a history of pre-eclampsia are at increased risk of essential hypertension later in life.
Through the technique of LD score analysis, InterPregGen has studied the extent of sharing of sequence variants with other genetically complex disorders. This method of analysis does not identify specific shared variants, but examines the entire genome for evidence of genetic correlation between phenotypes. In addition to the sharing of susceptibility genes with essential hypertension, as described above, there is significant genetic correlation between pre-eclampsia and coronary artery disease and type II diabetes. Like hypertension, type 2 diabetes is a risk factor for pre-eclampsia. Furthermore, women with a history of pre-eclampsia have an increased risk of developing type II diabetes and coronary artery disease in later life. The clinical overlap between these disorders appears to be due at least in part to shared genetic determinants.
9. A tool for pre-eclampsia risk prediction.
Effective management of pre-eclampsia depends on prompt diagnosis. This in turn relies on good antenatal care, with regular monitoring of blood pressure. The identification at an early stage in pregnancy of those women at highest risk would enable efficient and cost-effective targeting of antenatal care resources. Currently risk prediction is based on clinical data, including personal and family history; trials of biomarkers are on-going. InterPregGen has explored whether incorporation of genetic data available in the first trimester would improve predictive models for pre-eclampsia. As an initial step the University of Helsinki coordinated the assembly of a database of clinical and genetic data from the participating cohorts: GOPEC (UK), FINNPEC (Finland), MoBa (Norway), HUNT (Norway), Uzbekistan, Kazakhstan and data from a large case-control cohort from Colombia (GenPE). Initial preparatory work included extensive quality control and harmonisation of clinical data from disparate clinical databases. The final database incorporates details of 13,294 control pregnancies and 13,511 pregnancies affected by pre-eclampsia, including maternal, fetal and paternal data where available.
Using statistical and exploratory data analysis, genetic and clinical variables with evidence of association with preeclampsia and available in the first trimester were selected for development of a predictive tool. The analyses were planned using variables with data available for at least 50% of each cohort. Mothers who, prior to 20 weeks of gestation, were affected by essential hypertension, renal disease, type 1 or 2 diabetes mellitus, or excreting protein in the urine were excluded.
Using computational bootstrapping analysis, selected variables were ranked in order of the strength of their association with pre-eclampsia. Available data were randomly divided into 70% ‘training’ and 30% ‘testing’ data. Predictive models were developed from training data using the high-ranked variables in a cumulative manner, and the best predictive model(s) were identified using the model quality assessment criteria: adjusted R2 and Akaike’s Information Criterion (AIC). These models were tested on the testing data. Two cohorts, Icelandic and Colombian, were reserved for external validation of the models. Using prediction accuracy analysis, the performance of the models was assessed by calculating the area under the curve (AUC) of the receiver operating characteristic (ROC).
The evidence from all the models tested was that clinical data alone has stronger predictive power than genetic data alone, but a combination of clinical and genetic data provides the best predictive model. A model based on 10 clinical and 2 genetic variables yielded the best predictive accuracy (AUC: 0.9). The predictive tool requires field testing in a clinical setting, and will need to evolve as additional biomarkers are identified by other research groups.
10. Conducted the study in compliance European Commission ethical standards.
The study has been conducted in compliance with European Commission standards of research ethics, following the requirements of the Ethics Review Report, and advised by Dr. Nicola Stingelin, our independent ethics consultant. A number of ethical issues have arisen as a result of differences in clinical and research practice between Europe and Uzbekistan and Kazakhstan. Mechanisms for obtaining informed consent for participation in genetic research were carefully reviewed to ensure that they met the standards which apply in the European Union. Both Central Asian countries have existing research ethics structures and guidelines, including established procedures for applying for Research Ethics Committee (REC) approval. Local REC approval was obtained by all partners, European and Central Asian, at the start of the study to cover their individual contributions to the project. After some initial difficulty in identifying a REC which was willing to review the project as a whole, over-arching approval was obtained from the University of Nottingham REC. The co-ordinating centre has maintained the ethics database, including updates of Research Ethics Committee approval as required. Emphasis has been placed throughout on protection of the privacy of participants, by coding or anonymising clinical and genetic data, using secure sites for the transfer of electronic data, and ensuring that electronic and hard copy data are stored securely. European and Central Asian partners have developed publication policies, data sharing policies, and biobank governance procedures, which will be made publicly available via the InterPregGen website. These may prove of value to other researchers embarking on research projects involving collaborative working between Central Asian and European countries.
Potential Impact:
Potential Impact
• InterPregGen has completed the largest genome-wide association study of pre-eclampsia conducted to date, and achieved its primary aim of identification of DNA sequence variants in mothers and their offspring which predispose to disease. The findings are novel and convincingly reproducible in independent sample collections. The past two decades of research into the molecular genetics of pre-eclampsia have seen considerable investment of energy and resources without producing definitive results. InterPregGen has shown that the DNA of many thousands of affected individuals and controls must be examined in order to detect the sequence variants which predispose to disease, as each variant individually has a small effect on disease risk. This should provide a strong message to researchers that effort should be invested not in conducting small inconclusive case-control studies, but in building large, well-phenotyped collections of samples which can provide definitive evidence for genetic associations.
The ability to merge genetic data from multiple large sample collections through meta-analysis has been key to InterPregGen’s successful identification of variants associated with pre-eclampsia. The limit of detection of genetic variants of small effect is determined mainly by the size of the study, and ever-increasing sample sizes from meta-analyses will likely identify further variants associated with pre-eclampsia and/or its complications. To facilitate future use of InterPregGen data in meta-analysis GWAS data will be deposited in the European Genome-phenome Archive (www.ebi.ac.uk/ega/) with managed access for bona fide researchers.
The motivation for InterPregGen has been and remains the application of genetic discoveries to improved pregnancy outcomes, based on a better understanding of pre-eclampsia and its complications. The genetic discoveries have stimulated the articulation of novel hypotheses of the molecular basis of pre-eclampsia and its subtypes, opening prospects for further research and modifications to clinical practice. One key discovery has been the evidence that pre-eclampsia shares genetic association with gestational hypertension, essential hypertension, coronary artery disease, and type 2 diabetes. This provides a firm molecular basis for the observed clinical overlap between these conditions. The challenge now is to tease out the shared pathological mechanisms through which the genetic variants are acting, and develop a molecular classification of these conditions which can guide disease prevention and therapy.
InterPregGen is the first study to conduct a thorough investigation of fetal DNA variants in pre-eclampsia. The finding of DNA variants close to the FLT1 gene, which encodes Flt-1 and its soluble isoform sFlt-1, is of considerable interest in view of the established association of high sFlt-1 levels in women with pre-eclampsia. The sFlt-1 in the maternal circulation is believed to be of placental origin, which is consistent with involvement of the fetal gene. What is unexplained is the observation that fetal FLT1 genotype is associated only with late-onset pre-eclampsia, as maternal sFlt-1 is regarded as a marker for early-onset disease. The future challenge for researchers is to explore how modest alterations in Flt-1 or sFlt-1 expression from the start of pregnancy could cause changes, perhaps in the formation of the placenta, which are manifested as pre-eclampsia in late pregnancy.
The prospect of predicting in early pregnancy which women are at highest risk of pre-eclampsia is attractive, as this would enable the effective targeting of limited ante-natal care resources. In common with most other genetically complex disorders, genetic data alone has inadequate predictive power for pre-eclampsia. Nevertheless there is evidence that a combination of clinical and genetic data has improved predictive value over clinical data alone, and this merits further testing in a clinical setting. InterPregGen has shown that the DNA variant with the strongest effect on pre-eclampsia risk is in the fetal, not the maternal, genome. The fetal genotype is currently not readily available antenatally in the majority of pregnancies. However, techniques which reliably isolate fetal DNA from maternal venous blood are developing rapidly, and this holds out the prospect that fetal genotype could be included in a model with better predictive value for pre-eclampsia.
• InterPregGen was designed to establish sustainable and productive collaborative links between Central Asian and European research groups. Close collaboration has been a strong feature of the project, and will yield benefits for future research. Partners in Uzbekistan and Kazakhstan had not previously been involved in genetic research on the scale of InterPregGen. By sharing expertise in good research practice, advanced genetic laboratory techniques, and genetic data analysis, InterPregGen has ensured that Central Asian colleagues are conversant with research practices in Europe, and are trained to a high level in managing complex research procedures and analysing high level genetic data. Equally important has been the insights which European partners have obtained into health care systems in Uzbekistan and Kazakhstan. Together these factors have established mutual understanding and a vital infrastructure for further collaborative research. This has been underpinned by investment in infrastructure hardware in Uzbekistan and Kazakhstan, including -80C freezers and racking for long-term sample storage, sample bar-coding facilities, DNA spectrophotometers, genotyping equipment, and computers.
Sample biobanks have been established in Kazakhstan and Uzbekistan containing approximately 13,000 DNA samples for genetic research from ethnic Kazakhs and Uzbeks, supported by a high-quality clinical database, and genetic data including genome-wide association screening data. The availability of the biobanks for collaborative research will be advertised through the InterPregGen website and in academic publications arising from the project. These are amongst the largest biobanked resources for genetic research into pre-eclampsia in the world, and will attract international interest from researchers in the field.
InterPregGen has generated the first whole genome sequence of 100 ethnic Kazakhs and 100 Uzbeks, which will be deposited at the European Genome-phenome Archive for managed access to researchers. The importance of the whole genome data lies in the lack of any such datasets from Central Asians on public databases such as 1000Genomes. The deposited whole genome data will be of interest to scientists working in the field of population genetics, and to researchers into genetic disorders in Central Asian populations.
Dissemination activities
• Throughout the 5 years of the project InterPregGen partners have promoted the study at presitigious international scientific meetings, including the European Society for Human Genetics, the American Society for Human Genetics, and the International Society for the Study of Hypertension in Pregnancy. Abstracts at scientific meetings and academic papers published to date are:
1. L. Morgan, H. Laivuori; on behalf of the InterPregGen Consortium. Genetic studies of pre-eclampsia in Central Asian and European population-Information on the InterPregGen Study funded by the Seventh Framework Programme of the European Community. XVIII International Society for the Study of Hypertension in Pregnancy (ISSHP) World Congress, July 2012, Geneva
2. Sally Chappell, Wai Kwong Lee, Scott Shooter, Ralph McGinnis, Linda Morgan on behalf of the InterPregGen Consortium. Evaluation of GWAS chip-genotyping of fetal DNA extracted from umbilical cord tissue and WGA-DNA extracted from filter paper bloodspots. European Society for Human Genetics conference, June 2013, Paris
3. Morgan L on behalf of the InterPregGen consortium. Genetic studies of pre-eclampsia in Central Asian and European populations: a progress report on the InterPregGen study. International Society for the Study of Hypertension in Pregnancy (ISSHP) European Congress, June 2013, Tromsø
4. Svyatova G; Zakhidova N; Morgan L; Walker JJ; on behalf of the InterPregGen Consortium. Clinical features of pre-eclampsia in 2613 Central Asian women and babies recruited for genetic studies. XIX ISSHP World Congress, October 2014, New Orleans
5. Morgan L, McGinnis R, Steinthorsdottir V, Svyatova G, Zakhidova N, Kwong Lee W, Iversen AC, Magnus P, Walker JJ, Casas JP, Sultanov S, Laivuori H, on behalf of the InterPregGen consortium. InterPregGen: genetic studies of pre-eclampsia in three continents. Norsk Epidemiologi 2014; 24 (1-2): 141-146
6. N. O. Williams, F. Dudbridge, S. Chappell, C. Franklin, R. McGinnis, On behalf of the InterPregGen Consortium. Investigation of the polygenic genetics of Pre-eclampsia and its relationship with other phenotypes. European Society for Human Genetics conference, June 2015, Glasgow
7. Ralph McGinnis, Frank Dudbridge, Debbie Lawlor, Chris Franklin, Nicholas Williams. On behalf of the InterPregGen Consortium. Investigation of DNA variants responsible for Pre-eclampsia. American Society for Human Genetics annual meeting, October 2015, Baltimore
8. Svyatova G, Berezina G, Salimbaeva D, Choroshilova I, Uspanova M, Dolbakova S. First international Intergen DNA bank in the Republic of Kazakhstan. 4th International Conference, May 2016, Nazarbayev University, Astana
9. McGinnis R, Steinthorsdottir V, Williams N et al. First genome-wide significant locus for pre-eclampsia susceptibility discovered on fetal chromosome 13. American Society of Human Genetics annual meeting, October 2016, Vancouver
10. McGinnis R, Steinthorsdottir V, Williams N et al. First sequence variants associated with pre-eclampsia identified in fetal genome. Nature Genetics 2016. Manuscript under consideration
• The InterPregGen team is continuing to prepare manuscripts for submission to prestigious academic journals. Manuscripts in preparation include:
11. Genome-wide association study identifies multiple loci associated with pre-eclampsia in affected mothers
12. Whole genome sequencing of Central Asian populations from Kazakhstan and Uzbekistan
13. Clinical features of pre-eclampsia in Uzbekistan and Kazakhstan
14. The challenges and rewards of research collaboration between European and Central Asian countries
15. Genome-wide scanning in Central Asian mothers and babies affected by pre-eclampsia
16. Development of risk prediction models for pre-eclampsia incorporating genotyping data
• Bilingual (English/Russian) website.
The bilingual InterPregGen website www.interpreggen.org is maintained by the University of Nottingham. Confidential material has been deposited on the password-protected researchers’ area of the website throughout the study. With the end of the project, the password-protected area will be removed, and selected non-confidential study documentation will be moved to a publicly accessible researchers’ page. This will include standard operating procedures, information and consent forms, clinical record booklets and protocols developed in the course of InterPregGen for use by researchers planning similar studies in future. The public-facing website will remain active for a minimum of 2 years after the end of the project.
• Wider public dissemination
Partners have provided invited talks to public interest groups describing and promoting the study, and disseminated information leaflets in clinics in Kazakhstan and Uzbekistan. Examples can be downloaded from the InterPregGen website www.interpreggen.org. Training and monitoring visits by the University of Leeds team to Uzbekistan and Kazakhstan have provided opportunities for open lectures and discussions with local obstetricians covering clinical guidelines for pre-eclampsia. InterPregGen is promoted on individual institutional websites, and partners have provided interviews for local radio and television. Major InterPregGen publications will be accompanied by co-ordinated releases to local and national media by institutional press offices.
List of Websites:
WEBSITE: www.interpreggen.org
CONTACT DETAILS
University of Nottingham, United Kingdom Dr Linda Morgan linda.morgan@nottingham.ac.uk
Scientific Center for Gynecology, Obstetrics and Perinatology, Kazakhstan Prof.Gulnara Svyatova respmgk@mail.ru
Republic Specialized ScientificPractical Medical Center of Obstetrics and Gynecology, Uzbekistan Dr Nodira Zakhidova odira_z@yahoo.com
deCODE Genetics, Iceland . Dr Valgerdur Steinthorsdottir Valgerdur.steinthorsdottir@decode.is
University of Helsinki, Finland. Prof. Hannele Laivuori Hannele.laivuori@helsinki.fi
Norwegian University of Science and Technology, Norway. Dr Ann-Charlotte Iversen ann-charlotte.iversen@ntnu.no
National Institute of Public Health, Norway. Prof. Per Magnus Per.magnus@fhi.no
London School of Hygiene and Tropical Medicine, United Kingdom. Dr Frank Dudbridge Frank.dudbridge@lshtm.ac.uk
Wellcome Trust Sanger Institute, United Kingdom. Dr Ralph McGinnis rm2@sanger.ac.uk
University of Glasgow, United Kingdom. Prof. Anna Dominiczak a.dominiczak@clinmed.gla.ac.uk
University of Leeds, United Kingdom. Prof. James Walker j.j.walker@leeds.ac.uk
Institute of Immunology, Uzbek Academy of Science, Uzbekistan. Dr Nodira Zakhidova odira_z@yahoo.com