| Human Molecular Genetics | Pages |
©1999 Oxford University Press |
A genome-wide scan reveals a maternal susceptibility locus for pre-eclampsia on chromosome 2p13
Introduction
Results
Discussion
Materials And Methods
Subjects and family collection
Genealogy database and cluster function
Genotyping
Statistical analysis
Acknowledgements
References
A genome-wide scan reveals a maternal susceptibility locus for pre-eclampsia on chromosome 2p13
Received May 4, 1999; Revised and Accepted June 29, 1999
Pre-eclampsia is a common and serious disease and a major cause of maternal and infant mortality. Antenatal care systems world-wide screen for signs of the disease such as hypertension and proteinuria. Unlike most other human disorders it impacts two individuals, the mother and the child, both of whom can be severely affected. The pathophysiology of the disorder is incompletely understood, but familial clustering of the disease is apparent. Here we report the results of a genome-wide screen of Icelandic families representing 343 affected women. Including those patients with non-proteinuric pre-eclampsia (gestational hypertension), proteinuric pre-eclampsia and eclampsia, we detected a significant locus on 2p13 with a lod score of 4.70 (single point P < 3.49 × 10-6). This is the first reported locus for pre-eclampsia meeting the criteria for genome-wide significance.
INTRODUCTION
Pre-eclampsia is a complex disorder with a wide clinical spectrum but its initial hallmark is new onset hypertension in the latter half of pregnancy, resolving post-partum (1-4). More severe cases also have significant proteinuria. The most worrisome complication of pre-eclampsia is its unpredictable progression to eclampsia, represented by neurological involvement leading to seizures and morbidity or death in the mother and/or child. The best treatment is early delivery by caesarian section if the child is sufficiently mature (4). The rate of eclampsia has decreased to <0.1% of all deliveries due to screening for pre-eclampsia and earlier deliveries in pre-eclamptic women, but in developing countries mortality among these women remains high (5-7). It is evident from higher perinatal mortality and the more frequent small for gestational age births in pre-eclampsia that the child is also affected (8).
The aetiology and pathogenesis of pre-eclampsia are unclear but may involve abnormal placentation and its interaction with the maternal microvasculature (9,10). It is also associated with other vascular diseases (11,12). Although most cases are sporadic, there is substantial evidence for a familial component (13-16). For example, daughters of women with a history of pre-eclampsia are much more likely to develop pre-eclampsia than the daughters-in-law (15,16).
RESULTS
To study the effect of maternal susceptibility genes in the development of pre-eclampsia, we sent questionnaires to a population-based cohort of 2585 Icelandic women who were diagnosed with gestational hypertension, pre-eclampsia or eclampsia during the years 1984-1993. Approximately 18% of those who responded claimed that they had a first or second degree relative with pre-eclampsia. The maternity records of these familial cases of pre-eclampsia were re-examined and the patients sub-classified into gestational hypertension (non-proteinuric pre-eclampsia), proteinuric pre-eclampsia or eclampsia. These initial multiplex nuclear families were later run through an extensive computerized genealogy database, yielding, in some cases, much larger pedigrees than the families found using the questionnaire approach (17). The study included 124 pedigrees with 343 women who had any one of the three diagnostic classifications above (termed general criteria), of which 186 women in 72 pedigrees had either proteinuric pre-eclampsia or eclampsia (termed strict criteria) (Table 1).
Table 1. Composition of familial material according to the general and strict criteria
| No. of affecteds | General | Strict |
| 2 | 74 | 48 |
| 3 | 30 (1) | 16 (1) |
| 4 | 10 | 3 (1) |
| 5 | 2 (1) | 2 (1) |
| 6 | 4 (2) | 1 |
| 7 | 2 | 2a,b |
| 8 | 1b | 0 |
| 9 | 1a | 0 |
aIncludes family A.
bIncludes family B.
A genome-wide scan was performed using a framework map of 440 markers randomly distributed throughout the genome with an average spacing of ~9 cM. The data were analysed using affecteds-only non-parametric/allele sharing methods (18-20). Individuals who were not known to be affected, including males, were classified as `unknown' for the purposes of this analysis. Figure 1 presents the allele sharing lod scores for the genome scan using the markers in the framework map. For both the general and strict criteria, no region achieved a lod score >2 with the exception of chromosome 2. Under the general criteria, the lod score curve displayed two peaks with a lod of 2.42 at 93.68 cM and a lod of 3.14 at 110 cM, separated by a dip in between. An additional 22 markers were added to the region between 50 and 130 cM (Fig. 2). Under the general criteria, the right peak decreased, but the left peak increased or a lod score of 4.77 (NPL score 5.46, P < 3.49 × 10-6) at the marker D2S286 location at 94.05 cM. This satisfied the criterion for genome-wide significance (because it has P < 2 × 10-5) (21). Under the strict criteria, the lod score peaked at an almost identical location between the markers D2S2111 and D2S1394 (lod 3.23, NPL score 4.00, P < 7.7 × 10-5).
Figure 1. Genome-wide scan for pre-eclampsia genes. A framework map of 440 polymorphic markers was used. Each box represents a chromosome and the y-axis depicts the multipoint non-parametric lod score. The solid line represents the lod score for the general criteria; the dashed line represents the lod score for the strict criteria. The x-axis represents the cM distance of the markers used.
Figure 2. Dense mapping of the pre-eclampsia locus on chromosome 2p13. There are a total of 107 markers genotyped on the chromosome. Twenty-two additional markers around the peak location were used in the analysis. The x-axis represents the cM distance of the markers used. The solid line represents the lod score for the general criteria; the dashed line represents the lod score for the strict criteria.
Close inspection of the data revealed that support for the chromosome 2 locus came mostly from the two largest families, families A and B, displayed in Figure 3. For family A, all of the patients shared a haplotype spanning 12 markers, between 85.48 (D2S2368) and 99.41 cM (D2S1777). For family B, all the affecteds shared a haplotype between 89.76 (D2S291) and 111.21 cM (D2S2216). Therefore, the susceptibility gene is most likely located between 88.15 (D2S292) and 101.02 cM (D2S329) (Table 2). The two haplotypes of these two families were not identical over this wide region. Further work is needed to determine whether they share a small region identical by descent (IBD). Figure 4 shows the results of the genome scan with families A and B removed from the data. For the region between 88.15 and 99.41 cM on chromosome 2 the lod score ranged from 0.16 to 0.37 under the general criteria and from 0.26 to 0.38 under the strict criteria. However, under the general criteria the lod score reaches 2.08 at a different locus on chromosome 2, at 157.5 cM (D2S321). It should be noted that we found no evidence at all for linkage at previously reported candidate regions or genes, including 4q, angiotensinogen on 1q42-43 or endothelial nitrous oxide synthase on 7q36 (22-25).
Figure 3. Two families with a large contribution to the lod at the 2p13 locus. Half-filled circles indicate cases with non-proteinuric pre-eclampsia, filled circles indicate cases with proteinuric pre-eclampsia and cross-hatched circles indicate cases with eclampsia. All other individuals are not known to be affected and are treated as unknown in the analysis. Blood samples were available from all living members of these pedigrees. Some sexes in the two upper generations of these pedigrees have been altered and unaffected, untyped siblings of affecteds are not displayed, to protect the confidentiality of these families.
Figure 4. Genome-wide scan for pre-eclampsia genes using all the families except families A and B. The figure is constructed in the same way as Figure 1.
Table 2. Putative mutation-carrying haplotype in the affected members of families A and B
| D | D | D | D | D | D | D | D | D | D | D | D | D | D | D | D | D | D | |
| 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | |
| S | S | S | S | S | S | S | S | S | S | S | S | S | S | S | S | S | S | |
| 3 | 2 | 2 | 4 | 3 | 2 | 2 | 2 | 1 | 2 | 2 | 1 | 1 | 3 | 1 | 2 | 3 | 2 | |
| 3 | 3 | 8 | 4 | 5 | 9 | 9 | 1 | 3 | 1 | 8 | 6 | 7 | 2 | 7 | 3 | 8 | 2 | |
| 7 | 6 | 5 | 1 | 8 | 2 | 1 | 1 | 9 | 1 | 6 | 9 | 7 | 9 | 9 | 3 | 8 | 1 | |
| 8 | 1 | 4 | 0 | 7 | 0 | 3 | 6 | |||||||||||
| A III:4 | 9 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 3 | 4 | 4 | 6 | 1 |
| A III:5 | 9 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 3 | 4 | 4 | 6 | 1 |
| A III:6 | 9 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 3 | 4 | 4 | 6 | 1 |
| A III:7 | 9 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 3 | 4 | 4 | 6 | 1 |
| A IV:1 | 8 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 3 | 4 | 4 | 6 | 1 |
| A IV:2 | 8 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 3 | 4 | 4 | 6 | 1 |
| A IV:3 | 9 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 4 | 6 | 10 | 3 | 3 |
| A IV:4 | 9 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 4 | 6 | 10 | 3 | 3 |
| A IV:5 | 9 | 5 | 9 | 8 | 3 | 5 | 4 | 4 | 4 | 4 | 8 | 4 | 2 | 3 | 4 | 4 | 6 | 1 |
| B III:1 | 10 | 3 | 8 | 9 | 9 | 11 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
| B III:2 | 10 | 3 | 8 | 9 | 9 | 11 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
| B III:3 | 10 | 3 | 8 | 9 | 9 | 11 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
| B III:4 | 10 | 3 | 8 | 9 | 9 | 11 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
| B III:6 | 9 | 6 | 8 | 9 | 9 | 11 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
| B IV:2 | 7 | 6 | 9 | 8 | 7 | 5 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
| B IV:3 | 9 | 6 | 8 | 9 | 9 | 11 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
| B V:1 | 7 | 6 | 9 | 8 | 7 | 5 | 1 | 4 | 3 | 5 | 6 | 8 | 2 | 3 | 4 | 6 | 4 | 5 |
DISCUSSION
Our study design, while it used a computerized genealogy database, did not focus on large family clusters with many patients. Indeed, half of the families contained only two patients and most of these were sib pairs. However, the evidence supporting the chromosome 2 locus came mostly from the two largest family clusters. This supports the notion that, even with complex traits and the use of non-parametric/allele sharing methods, large families which are created by linking up affecteds who are not closely related can sometimes be very useful.
At this point it is difficult to assess the contribution of the chromosome 2 locus to pre-eclampsia cases in general. The data are consistent with the chromosome 2 locus harbouring a highly penetrant variant which is not very common in the population and is responsible more for cases of pre-eclampsia which have a very strong familial component. For example, under the general criteria, by fitting a dominant model with no sporadic cases (but still using an affecteds-only analysis) and allowing for locus heterogeneity (26), a parametric lod score of 3.91 was obtained with the mutation frequency set to 0.01 and 17% of the families were estimated to be affected by this mutation. It should be noted that the lod score was lower under the strict criteria primarily because the number of affecteds was lower and not because of decreased IBD sharing at the chromosome 2 locus. In fact, applying the same dominant modelling approach to the strict material, we achieved a parametric lod score of 3.13 and estimated that 22% of the families were affected by this mutation. However, given the complexity of the disease phenotype, a reliable assessment of the effect of the chromosome 2 locus will probably have to wait until after the gene is isolated.
MATERIALS AND METHODS
Subjects and family collection
In 1993 the collection of information about women with a family history of pre-eclampsia was initiated. Computerized records containing diagnosis codes for the years 1984-1993 at the National University Hospital in Reykjavik were searched for women with the international classification of diagnosis codes for transient hypertension of pregnancy (gestational hypertension), pre-eclampsia and eclampsia (e.g. ICD-9 codes 642.0-642.9). A total of 2585 women were found to have these diagnoses. Of these, 113 were living abroad, seven were deceased and one could not be located. A structured questionnaire was sent to 2467 women seeking information about their pregnancy history, dates and places of delivery, family history of pre-eclampsia, medications, hospital admissions, cardiovascular disorders, diabetes mellitus, renal disorders and lifestyle, such as smoking habits and contraceptives. The study was approved by the Icelandic Data Protection Commission and the Ethics Committee. There was a 70% response rate and 310 women (index cases) claimed to have first or second degree relatives with a history of pre-eclampsia. Subsequently, the maternity records for the index women and their relatives were reviewed and the diagnoses verified. Women with a confirmed diagnosis according to the criteria below were then invited to participate in the study. They were classified into: (i) gestational hypertension (mild pre-eclampsia without proteinuria); (ii) proteinuric pre-eclampsia (gestational hypertension with proteinuria or severe gestational hypertension); and (iii) eclampsia (pre-eclampsia with seizures). Gestational hypertension was defined as a diastolic blood pressure of [ge]90 mmHg (required in all cases) plus a systolic blood pressure of [ge]140 mmHg or a systolic blood pressure increase [ge]30 mmHg or a diastolic blood pressure increase [ge]15 mmHg over pre-pregnancy or early pregnancy values and which resolved by 6 months post-partum (2,3). Two readings 6 h apart were required and the fourth Korotkoff sound was used to determine the diastolic value. Women with pre-existing hypertension or other medical disease known to predispose to pre-eclampsia were excluded. Proteinuria was defined as >300 mg/24 h or at least two urinary dipstick measurements >1+ (separated by at least 6 h) (2,3). No further sub-division on clinical grounds was done.
Blood from members of these family clusters was collected and stored at -70°C. Personal information about the patients and family members were subsequently coded according to the rules of the Data Protection Commission. The blood and DNA samples were also coded in the same way.
Genealogy database and cluster function
deCODE Genetics has an ongoing genealogy project which involves electronic registration of all available genealogies for the last 10 centuries in Iceland, including genealogical manuscripts, censuses, church books and phone books (17). The genealogy database is stored and maintained within a relational database. Each record in the database consists of a personal identifier, identifier to parents, sex and dates of birth and death. All personal identifiers in the genealogy database are reversibly coded by the Data Protection Commission outside our laboratory. All algorithms that work on family relations were implemented outside the database, in memory-based programs, since relational databases are not well suited for implementing dynamic programming algorithms. We have developed recursive pedigree algorithms that find all ancestors in the database who are related to each member of the input list within a given number of generations back. Using these groups, the cluster function then searches for ancestors who are common to any two or more members of the input lists. These pedigrees are then drawn by the Cyrillic pedigree drawing program (Cherwell Scientific Publishing). The Data Protection Commission coded lists of patients participating in the study in the same manner as the genealogy database and the cluster function created, in some cases, much larger pedigrees than the nuclear families found using the questionnaire approach. (In the families generated, any affected member was at most five meiotic events away from an affected relative, e.g. five meiotic events separate a proband from his first cousin once removed.) The study in the end included 124 pedigrees with 343 women who had any one of the three diagnostic classifications above, of which 186 women had either proteinuric pre-eclampsia or eclampsia.
Genotyping
We extracted DNA from the peripheral blood of a total of 343 affected and 239 unaffected relatives and genotyped them using 440 fluorescently labelled primers (a combination of 330 microsatellite markers from the Research Genetics v8 linkage screening set and 110 additional dinucleotide repeat markers from published genetic maps) with an average spacing of 8.7 cM (27,28). PCRs were set up, run and pooled on Perkin Elmer/Applied Biosystems 877 Integrated Catalyst Thermocyclers. The reactions were 5 µl in volume and used 10 ng of template genomic DNA and 0.25 U of Amplitaq Gold (Applied Biosystems), 2 pmol of each primer, 0.2 mM deoxynucleotides and 2.5 mM MgCl2; the buffer was supplied by the manufacturer. The PCR conditions used were 95°C for 10 min to activate the Amplitaq Gold, then 34 cycles of denaturation at 94°C for 15 s, annealing at 55°C for 30 s and elongation at 72°C for 1 min. The pooled products were supplemented with the internal size standards and were electrophoresed and detected on an Applied Biosystems model 377 Sequencer using Genescan v3.0 peak calling software. The genotypes were defined and edited in the Applied Biosystems Genotyper v.2.0 program. Forty-six Icelandic controls were genotyped to derive allelic frequencies. Additional polymorphic markers were used to increase the information content of suggestive loci on chromosome 2 and were also used to examine the eNOS locus on 7q36 (20). The marker orders and genetic distances used were obtained from publicly available genetic maps at the Marshfield Medical Clinic web site (www.marshmed.org ) (27).
Statistical analysis
Mendelian inheritance was verified in part by using the PEDCHECK program (29). All linkage results reported were produced by the GENEHUNTER-PLUS modification of the GENEHUNTER program (18,19). The GENEHUNTER program implements a non-parametric method of linkage analysis (19). It produces NPL scores which indicate the amount of excess IBD sharing among related affecteds as measured by a chosen scoring function. The scoring function used for this manuscript was Sall, which is found to be quite powerful for a wide range of inheritance models and is particularly suitable for dominant inheritance with or without locus heterogeneity (19,20). All calculations were fully multipoint, i.e. all markers from the same chromosome were used simultaneously. Apart from the standard output of GENEHUNTER, the GENEHUNTER-PLUS program gives a lod score which is closely related to the NPL score (18). The allele sharing lod scores reported here are all computed with respect to the exponential model (18). This lod score gives the same statistical significance as the NPL score when the IBD information is complete. However, when the IBD information is incomplete, the P-values calculated based on the NPL score are, in general, conservative, sometimes extremely so when the information is far from complete (30). The lod score does not have this weakness because it is a true log likelihood ratio and, therefore, is much more appropriate for a genome-wide scan with the framework map markers. However, the P-values reported for chromosome 2 in the final analysis are the conservative `exact' P-values reported by GENEHUNTER and this explains the use of the `<' sign. Because with the additional markers the information at the chromosome 2 locus is close to complete, the P-values reported are expected to be only slightly conservative.
ACKNOWLEDGEMENTS
We wish to thank the Icelandic patients, controls and their family members for their generous support for this work. This research was supported in part by the Wellcome Trust, UK, the Icelandic Research Council, the Research Fund of the University of Iceland and deCODE Genetics Inc. The authors would like to thank Professor J.M. Connor (University of Glasgow, Glasgow, UK) for his support in initiating this study. We would like to thank órgunnur Hjaltadóttir and Sigrún Valdimarsdottir for their help in data collection and the Heart Preventive Clinic, Icelandic Heart Association, Reykjavík, for facilities during data collection.
REFERENCES
+To whom correspondence should be addressed. R.A.- Tel: +354 525 3600; Fax: +354 525 3608; Email: reynirar{at}hi.is K.S.-Tel: +354 570 1900; Fax: +354 570 1901; Email: kstefans{at}decode.is
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Fitzpatrick, M.P. Johnson, T.D. Dyer, S. Forrest, K. Elliott, J. Blangero, S.P. Brennecke, and E.K. Moses Genetic association of the activin A receptor gene (ACVR2A) and pre-eclampsia Mol. Hum. Reprod., March 1, 2009; 15(3): 195 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.P. Johnson, E. Fitzpatrick, T.D. Dyer, J.B.M. Jowett, S.P. Brennecke, J. Blangero, and E.K. Moses Identification of two novel quantitative trait loci for pre-eclampsia susceptibility on chromosomes 5q and 13q using a variance components-based linkage approach Mol. Hum. Reprod., January 1, 2007; 13(1): 61 - 67. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. T. Chelbi, F. Mondon, H. Jammes, C. Buffat, T.-M. Mignot, J. Tost, F. Busato, I. Gut, R. Rebourcet, P. Laissue, et al. Expressional and Epigenetic Alterations of Placental Serine Protease Inhibitors: SERPINA3 Is a Potential Marker of Preeclampsia Hypertension, January 1, 2007; 49(1): 76 - 83. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Kalmyrzaev, A. Aldashev, M. Khalmatov, A. Polupanov, A. Jumagulova, L. Mamanova, M. R. Wilkins, and M. Town Genome-Wide Scan for Premature Hypertension Supports Linkage to Chromosome 2 in a Large Kyrgyz Family Hypertension, November 1, 2006; 48(5): 908 - 913. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E.K. Moses, E. Fitzpatrick, K.A. Freed, T.D. Dyer, S. Forrest, K. Elliott, M.P. Johnson, J. Blangero, and S.P. Brennecke Objective prioritization of positional candidate genes at a quantitative trait locus for pre-eclampsia on 2q22 Mol. Hum. Reprod., August 1, 2006; 12(8): 505 - 512. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Roberts and H. S. Gammill Preeclampsia: Recent Insights Hypertension, December 1, 2005; 46(6): 1243 - 1249. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Yamada, A. R. Carson, I. Caniggia, K. Umebayashi, T. Yoshimori, K. Nakabayashi, and S. W. Scherer Endothelial Nitric-oxide Synthase Antisense (NOS3AS) Gene Encodes an Autophagy-related Protein (APG9-like2) Highly Expressed in Trophoblast J. Biol. Chem., May 6, 2005; 280(18): 18283 - 18290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. B.M. Oudejans, J. Mulders, A. M.A. Lachmeijer, M. van Dijk, A. A.M. Konst, B. A. Westerman, I. J. van Wijk, P. A.J. Leegwater, H. D. Kato, T. Matsuda, et al. The parent-of-origin effect of 10q22 in pre-eclamptic females coincides with two regions clustered for genes with down-regulated expression in androgenetic placentas Mol. Hum. Reprod., August 1, 2004; 10(8): 589 - 598. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. F. Birkisson, E. Halapi, U. S. Bjornsdottir, D. L. Shkolny, E. Adalsteinsdottir, T. Arnason, D. Gislason, T. Gislason, J. Gulcher, K. Stefansson, et al. Genetic Approaches to Assessing Evidence for a T Helper Type 1 Cytokine Defect in Adult Asthma Am. J. Respir. Crit. Care Med., May 1, 2004; 169(9): 1007 - 1013. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hylenius, A.-M. Nybo Andersen, M. Melbye, and T. V. F. Hviid Association between HLA-G genotype and risk of pre-eclampsia: a case-control study using family triads Mol. Hum. Reprod., April 1, 2004; 10(4): 237 - 246. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Mein, M. J. Caulfield, R. J. Dobson, and P. B. Munroe Genetics of essential hypertension Hum. Mol. Genet., April 1, 2004; 13(90001): R169 - 175. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Levesque, J.-M. Moutquin, C. Lindsay, M.-C. Roy, and F. Rousseau Implication of an AGT Haplotype in a Multigene Association Study With Pregnancy Hypertension Hypertension, January 1, 2004; 43(1): 71 - 78. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||