WO2008144940A1

WO2008144940A1 - Biomarker for hypertriglyceridemia

Info

Publication number: WO2008144940A1
Application number: PCT/CA2008/001057
Authority: WO
Inventors: Robert A. Hegele
Original assignee: University of Western Ontario
Current assignee: University of Western Ontario
Priority date: 2007-05-31
Filing date: 2008-05-30
Publication date: 2008-12-04
Anticipated expiration: 2009-11-30

Abstract

A method of assessing the risk of severe HTG in a mammal is provided comprising determining in a nucleic acid-containing sample from the mammal the presence of at least one polymorphism within the APOA5 gene, wherein the identification of an APOA5 polymorphism is indicative of a risk of severe HTG.

Description

BIOMARKER FOR HYPERTRIGLYCERIDEMIA

Field of the Invention

[0001] The present invention relates to biomarkers useful in the determination of risk of hypertrigylceridemia in a mammal.

Background of the Invention

[0002] Hypertriglyceridemia (HTG) is a commonly encountered phenotype that is a defining component of metabolic syndrome and is associated with numerous comorbidities, including coronary heart disease (CHD) and diabetes. Furthermore, plasma triglyceride (TG) concentrations of >10 mmol/L, a level that defines adult patients with Frederickson type 5 hyperlipoproteinemia, are associated with increased risk of acute pancreatitis. Plasma TG concentration >10 mmol/L is seen in ~1 in 600 adult North Americans.

[0003] Moderate HTG may generally be associated with plasma TG concentrations equal to or greater than 5 mmol/L and less than 10 mmol/L and may generally be associated with Frederickson type 4 hyperlipoproteinemia. Mild hypertriglyceridemia may generally be associated with plasma TG concentrations equal to or greater than 2 mmol/L and less than 5 mmol/L and may generally also be associated with Frederickson type 4 hyperlipoproteinemia. Plasma TG concentrations <2 mmol/L may qualify as representing the normal range of TG levels. Patients with TG levels falling within the normal, mild, or moderate range remain at risk of developing severe HTG.

[0004] While both genetic and lifestyle factors are determinants of plasma TG concentration, the genetic component remains incompletely defined. Since complex quantitative traits, such as plasma TG, do not conform to Mendelian inheritance patterns, their genetic basis may be the cumulative contribution of multiple DNA variants.

[0005] Given the prevalence of hypertriglyceridemia in numerous pathological conditions, it would be desirable to develop a means to identify risk of acquiring HTG. Summary of the Invention

[0006] It has now been found that the occurrence of an APOA5 polymorphism in a mammal, along with one or more low frequency polymorphisms, is associated with risk for hypertriglyceridemia, including sever HTG, in a mammal and, therefore, is also associated with risk of HTG-related disease.

[0007] Accordingly, in one aspect of the invention, there is provided a method of assessing the risk of HTG in a mammal comprising determining in a nucleic acid- containing sample from the mammal the presence of polymorphisms within the APOA5 gene, wherein the identification of one or more APOA5 polymorphisms is indicative of a risk of HTG.

[0008] In another aspect of the invention, a method of assessing the risk of HTG in a mammal is provided comprising determining in a nucleic acid-containing sample from the mammal the presence of at least one of the APOA5 polymorphisms, APOA5 S19W and APOA5 -1 131T>C, in combination with one or more secondary polymorphisms selected from the group consisting of the GCKR rs780094 A allele, TRIBl rsl7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and the APOE non-E3 allele, wherein the identification of at least one APOA5 polymorphism in combination with one or more secondary polymorphisms is indicative of risk of HTG.

[0009] In a further aspect, a method of assessing the risk of HTG in a mammal comprising:

1) identifying the occurrence of HTG-related polymorphisms in said mammal; and

2) determining the risk based thereon.

[0010] In another aspect, the invention provides a method of assessing the risk of disease associated with HTG in a mammal comprising the step of identifying the occurrence in said mammal of at least one APOA5 polymorphism, wherein the identification of one or more APOA5 polymorphisms is indicative of a risk of disease associated with HTG. [0011] In another aspect of the invention, a kit useful to assess risk of HTG in a mammal is provided comprising at least one reagent useful to identify the presence of at least one APOA5 polymorphism.

[0012] In another aspect of the invention, an array useful to assess the risk of HTG in a mammal is provided comprising reagents useful to detect at least one APOA5 polymorphism, and at least one secondary polymorphism selected from the group consisting of the GCKR rs780094 A allele, TRIBl rsl7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and the APOE non-E3 allele.

[0013] These and other aspects of the invention will become apparent from the detailed description and drawings which follow.

Brief Description of the Drawings

[0014] Figure 1 is a bar graph illustrating the plasma lipoprotein response in patients with severe HTG to fibrate monotherapy;

[0015] Figure 2 illustrates he relationship between plasma TG quartile and APOA5 variant frequencies;

[0016] Figure 3 is a line graph illustrating that the presence of APOA5 S19W, -

1131T>C or either allele was strongly associated with Fredrickson hyperlipoproteinemia (HLP) phenotypes as indicated by odds ratios for each allele and each HLP phenotype; and

[0017] Figure 4 illustrates the risk associated with select clinical and genetic variables for severe HTG.

Detailed Description of the Invention

[0018] A method of assessing the risk of HTG in a mammal is provided comprising identifying the occurrence in the mammal of one or more polymorphisms within the APOA5 gene. Detection of an APOA5 polymorphism is indicative of a risk of severe HTG.

[0019] The term "HTG" refers to a TG concentration that may fall within the normal, mild, or moderate range, e.g. from < 2 mmol/L to > 10 mmol/L, while the term "severe HTG" refers to a fasting plasma triglyceride (TG) concentration greater than 10 mmol/L documented on at least 2 distinct occasions, wherein "fasting" refers to no intake of anything other than water for at least 12 hours.

[0020] As used herein, the term "mammal" refers to human and non-human mammals such as domestic animals, livestock and wild animals.

[0021] The term "polymorphism" refers to one of two or more alternate forms

(alleles) of a nucleotide sequence that can exist at a particular site in a gene.

[0022] The term "APOA5" refers to the apolipoprotein A-V gene having the nucleotide sequence depicted by accession no. rs662799.

[0023] In order to assess the risk of HTG in a mammal, the occurrence of APOA5 gene polymorphisms in a nucleic acid-containing sample from the mammal is determined. The term "nucleic acid-containing sample" is used to refer to any biological sample that may be obtained from the mammal that contains nucleic acid. Appropriate DNA- containing biological samples for use in the present method include, but are not limited to, saliva, urine, semen and other bodily secretions, as well as hair, epithelial cells and the like. Although such non-invasively obtained biological samples are preferred for use in the present method, one of skill in the art will appreciate that invasively-obtained DNA- containing biological samples, may also be used in the present method, including for example, blood, serum, bone marrow, cerebrospinal fluid (CSF) and tissue biopsies such as lymph node samples. Techniques for the invasive process of obtaining such samples are known to those of skill in the art.

[0024] It may be necessary, or preferable, to extract the DNA from the biological sample prior to polymorphism determination. Methods of DNA extraction are well-known to those of skill in the art and include chemical extraction techniques utilizing phenol- chloroform (Sambrook et al., 1989), guanidine-containing solutions, or CTAB-containing buffers. As well, as a matter of convenience, commercial DNA extraction kits are also widely available from laboratory reagent supply companies, including for example, the QIAamp DNA Blood Minikit available from QIAGEN (Chatsworth, CA), or the Extract- N-Amp blood kit available from Sigma (St. Louis, MO). [0025] Once an appropriate nucleic-acid-containing sample is obtained, polymorphisms in target genes can be determined using well-established techniques, for example, genotyping assays such as TaqMan assays or restriction endonuclease analysis.

[0026] In accordance with the present methods, APOA5 polymorphisms in both the coding and non-coding regions of the gene have been identified as biomarkers associated with risk of HTG. For example, the polymorphisms, APOA5 S19W in the coding region of the gene, and APOA5 -1131T>C in the non-coding region of the gene, are each distinct biomarkers associated with risk of HTG in a mammal. By "distinct" it is meant that each of these polymorphisms is itself a biomarker of HTG.

[0027] The identification of APOA5 polymorphisms in combination with one or more secondary polymorphisms may provide a more definitive risk assessment, i.e. a risk assessment having an odds ratio of at least about 2.00, preferably at least about 4.00, and more preferably at least about 10. The greater the odds ratio, the greater the degree of risk in a mammal of developing HTG. The term "secondary polymorphism" is used herein to refer to a polymorphism which, when it occurs in conjunction with an APOA5 polymorphism indicative of severe HTG risk e.g. one or the other of , it will generally increase the odds ratio associated with HTG risk. Such secondary polymorphisms include, but are not necessarily limited to, GCKR rs780094 A allele, TRIBl rsl 7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and APOE non-E3 allele, e.g. E2 (C112(rs429358) + C158(rs7412)) and E4 (R112(s429358) + R158(rs7412)) alleles. For clarity, the term "GCKR " refers to glucokinase regulatory protein; "TRIBl " is also known as the C8FW gene; "GALNT2 " refers to UDP-N-acetyl-alpha-D- galactosamine:polypeptide N-acetylgalactosaminyltransferase 2; "TBL2" refers to transducin-(beta)-like 2 gene; and "APOE" refers to apolipoprotein E gene.

[0028] Additionally, the identification of a combination of genetic factors, such as

APOA5 and secondary polymorphisms as set out above, in combination with one or more clinical factors may also provide a more definitive risk assessment, i.e. a risk assessment having an odds ratio of at least about 2.00, and preferably at least about 4.00. The term "clinical factors" is used herein to refer to factors such as diabetes, obesity (defined as having a BMI>33kg/m²), age, sex, diet, alcohol intake, ethnicity (e.g. asian, aboriginal), total fat content and psychological state (e.g. stress) of the candidate mammal to be assessed. [0029] In a further aspect of the invention, a method of assessing the risk of disease associated with HTG in a mammal is provided. The method comprises the step of identifying the occurrence in said mammal of at least one APOA5 polymorphism. Identification of one or more APOA5 polymorphisms is indicative of a risk of disease associated with HTG, including but not limited to metabolic syndrome, disorders in which there is an increased risk of developing cardiovascular disease, diabetes, hypertension, polycystic ovarian syndrome, non-alcoholic fatty liver disease, as well as pancreatitis.

[0030] Assessment of risk may be calculated manually, or automatically using computer technology. For example, the relevant genetic and clinical factors (or characteristics) of a mammal may be input into a computer adapted to calculate the risk based on these factors, e.g. a system comprising the appropriate software to conduct the risk assessment, and the computer will provide an output of the risk for the mammal of developing severe HTG, or a disease associated with severe HTG.

[0031] In another aspect of the present invention, a kit is provided that is useful in the determination of risk in a mammal of developing HTG. The kit comprises one or more reagents useful to determine the presence of an APOA5 polymorphism in a nucleic acid- containing sample from the mammal. The kit may optionally include reagents useful to identify the presence of one or more secondary polymorphisms such as the GCKR rs780094 A allele, TRIBl rsl7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl 7145738 T allele and the APOE non-E3 allele. The reagent useful to identify the presence of an APOA5 polymorphism, or a secondary polymorphism, may comprise one or more probes, primers or other nucleic acid capable of specific binding to a selected polymorphism, or any other component useful to detect a selected polymorphism. The reagents may be detectably labeled and may optionally be fixed to a solid support. In addition, the kit may include controls, buffers, and instructions for use.

[0032] In a further aspect of the invention, an array is provided that is useful to assess the risk of HTG in a mammal. Generally, the array will include a support to which a series of reagents, as set out above, is bound that is useful to identify the APOA5 and secondary polymorphisms identified herein. The reagents, typically polynucleotide probes obtained by, e.g., polymerase chain reaction (PCR) amplification of specific gene segments that target the polymorphism of interest, are affixed to the array support using methods well established in the art. Detection of target polymorphisms within a nucleic acid-containing sample from a mammal is conducted using the array in conjunction with a labeling technique, as one of skill in the art will appreciate, which enables detection of binding of the polymorphism-containing nucleic acid.

[0033] Other features and advantages of the present invention will become apparent from the following examples. It should be understood, however, that the specific examples while indicating preferred embodiments of the invention are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

Example 1

[0034] The study included 167 patients of European geographical ancestry with severe HTG, defined as having fasting plasma TG >10 mmol/L documented on >2 occasions, from a single tertiary referral lipid clinic. Patients underwent a complete medical history and examination; basic clinical, biochemical and demographic variables were collected. Normolipidemic adult controls were taken from the European subgroup of the Study of Health Assessment and Risk in Ethnic groups (SHARE), a survey of cardiovascular risk factors in Canadian sub-populations as described in Anand SS et al. (2000) Lancet 356: 279-84.OnIy individuals with TG <2 mmol/L were included. All patients provided informed consent for DNA analysis.

[0035] Baseline attributes of the study sample are shown in Table 1 below.

Table 1: Baseline attributes of study subjects severe HTG controls P-value number 167 277 percent female 32.9% 53.7% O.0001 age (years) 50.9±13.0 50.1±14.8 NS treatment for diabetes mellitus 35.3% 1.4% <0.0001 body mass index (kg/m²) 30.3±4.8 27.0±4.6 O.0001 body mass index >33 kg/m² 28.0% 8.6% O.0001 total cholesterol (mmol/L) 11.9±5.9 5.0±0.9 <0.0001 triglycerides (mmol/L) 31.3±25.0 1.15±0.41 O.0001 HDL-cholesterol (mmol/L) 0.77±0.36 1.25±0.36 <0.0001 [0036] Severe HTG cases and controls were matched for age. There was a significant excess of males and of patients receiving medical treatment for type 2 diabetes among severe HTG patients. Body mass index (BMI) was significantly higher in severe HTG patients. By definition, severe HTG patients had markedly higher plasma TG and total cholesterol and significantly lower plasma HDL cholesterol. Plasma TG concentration in severe HTG patients ranged from 10.1 to 180 mmol/L. In addition, 47/167 severe HTG patients (28.1%) had been hospitalized on >1 occasion with pancreatitis. Corresponding data were not available for the control group.

[0037] DNA was extracted as described in Al-Shali K et al. (Clin Biochem 2000.

35: 125-30, the contents of which are incorporated herein by reference). Coding regions and intron-exon boundaries of APOA5 (4 exons) were amplified, purified and then directly sequenced in 5'- and 3'- directions in an ABI 3730 DNA Analyzer using reagents shown in Table 2.

Table 2: Amplification primers for coding regions of APOA5

exon primer sequence Tanneal (⁰C) size (bp)

1,2 F: 5'-CAGGTAATGGCAAGCATGG 60 353

R: 5'-CAGAGGCAGGTCATCATGG

3 (half) F: 5'-GCTTCGTGCGTGAGTTGTTA 60 541

R: 5'-CTTCCGGGAGAGCACCTG

3 (half) F: 5'-GAGCTGCACCGCAGTGTG 60 570

R: 5'-AGGCCACTTTCAAGGACTGA

[0038] Genomic DNA sequences were analyzed using Sequence Navigator software. Sequence variants were confirmed on an independent sample on a second day. Screening of normolipidemic controls for sequence variants found in severe HTG patients was performed using allele-specific methods such as restriction endonuclease analysis or a method called SNaPshot, as summarized in Table 3.

Table 3: Genotyping methodology for sequence variants in ΛPOA5 variant amplification primers Tanneal(°C) size (bp)Comments p S19W

F 5'-CAGGTAATGGCAAGCATGG 60 353 RE endonuclease Eael, fragment sizes

R 5'-CAGAGGCAGGTCATCATGG S19 177,101,75 , W19 252, 101

p N66S

F 5'-GCTTCGTGCGTGAGTTGTTA 60 1200 ASOpπmer 5'GAGCAAGACCTCAACA

R 5'-AGGCCACTTTCAAGGACTGA p V153M

F 5'-GCTTCGTGCGTGAGTTGTTATTGTTA 60 1200 ASOpimer 5'CTGTGGCCTGGTGGAGGTGGC

R 5'-AGGCCACTTTCAAGGACTGA

RE, restriction enzyme, ASO, allele-specific oligonucleotide method (SNaPshot)Blinded between-day replicated genotypes of a random 5% of samples showed >99.9% concordance.

[0039] The PANTHER database was used to impute dysfunction of sequence variants. The output of PANTHER is the subPSEC score, which represents the negative logarithm of the probability ratio of dysfunction of the wild-type and mutant amino acids at a particular position of the gene product; scores range between 0 (neutral) and -10 (most likely to be deleterious). PANTHER also calculates a probability of the mutation having a deleterious effect between 0 (neutral) and 1 (certainly deleterious). Predictions of dysfunction are very highly correlated with in vitro functional assessment. Most biochemically-proven functional mutations have a subPSEC score between -3 and -10 with P_dei_eteπous between 0.5 and 1.0, while most neutral non-dysfunctional polymorphisms have a subPSEC score between 0 and -2, with Pdeieteπou_s between 0.01 and 0.2.

[0040] All analyses were performed using SAS version 9.1. Between-group differences in discrete and quantitative traits were determined using chi-square or Fisher exact analysis and unpaired Student t-tests, respectively. Odds ratios were determined from contingency analyses. Nominal levels for significance were taken to be PO.05 for all comparisons.

Candidate gene polymorphisms

[0041] Other polymorphisms representing DNA sequence variants with frequency

<1% in controls and/or that are known functional disease-causing changes, also commonly referred to as mutations that were found in the genomic DNA of severe HTG patients are summarized in Table 4. Table 4: Rare DNA sequence polymorphisms in APOA5 in patients with severe hypertriglyceridemia number of carriers variant name new or known PANTHER published evidence severe HTG/controls subPSEC Pdeleterious (N= 167) /(N=277)

Gene: APOA5 p.N66S new -2.38 0.35 none 1 0

P.A315V known -2.51 0.38 none 1 0

[0042] In the severe HTG patients, APOA5 p.N66S and APOA5 p.A315 V each had modest scores with uncertain potential dysfunction.

Common DNA sequence variants

[0043] By re-sequencing, two reported candidate gene single nucleotide polymorphisms (SNPs) were found,: APOA5, namely p.S19W and p.A315V (Table 5). PANTHER predicted substantial dysfunction for APOA5 p.S19W. PANTHER predicted that APOA5 p.V315M was neutral. APOA5 p.S19W was significantly more prevalent, respectively, in severe HTG cases compared with controls (Table 5).

Table 5: DNA sequence polymorphisms in APOA5 in patients with severe hypertriglyceridemia allele frequency variant name mutation PANTHER published evidence severe HTG/controls or SNP subPSEC Pdeleterious (N=167) /(N=277)

Gene: APOA5

P.S19W known SNP -4.02 0.73 associated with 0.162 /0.049f dyslipidemia (ref 31-34); 50% reduced secretion from HepG2 cells

P.V153M known SNP -1.01 0.12 none 0.015 / 0.015

*P<0.005; fP<0.0001

[0044] In contrast to other dyslipidemias, a very strong association was found between severe HTG and the dysfunctional APOA5 p.S19W variant such that the variant was present in 31.7% of severe HTG patients but only 9.8% of normotriglyceridemic controls (OR 4.30 [95% CI 2.58 to 7.19]). Among severe HTG subjects, 111, 46 and 7 individuals had the S 19/Sl 9, W 19/Sl 9 and W 19/Wl 9 genotypes, respectively and among normolipidemic subjects 250, 23 and 4 individuals had the S 19/Sl 9, W 19/Sl 9 and W 19/Wl 9 genotypes, respectively (PO.0001). Among severe HTG patients, the mean plasma TG in APOA5 p.S19W carriers (29.0±27.4 mmol/L) was not significantly different from that in non-carriers (32.3±23.8 mmol/L).

[0045] To adjust for possible confounding due to obesity, a post hoc analysis was performed among subjects with BMI < 30 kg/m². This subgroup was comprised of 81 severe HTG patients (50.2±13.5 years, 33% female, BMI 26.7±2.3 kg/m²) and 186 controls (50.5±13.5 years, 53% female, BMI 25.1±2.5 kg/m²). Carrier OR for severe HTG for subjects with >1 copy of APOA5 S19W was 4.7 (95% CI 2.4 to 9.5; PO.0001).

Response toflbrate therapy according to genotype

[0046] In order to determine a possible between-genotype difference in plasma TG response to oral fibrate treatment, a 53 non-diabetic HTG patient sub-group whose treatment consisted only of dietary counseling and usual doses of one of fenofibrate, gemfibrozil or bezafibrate were examined. In this subgroup (50.2±12.8 years, 32% female), the maximal percent change from baseline of plasma lipoproteins within 12 months of initiating treatment was examined. The subgroup was comprised of 7, 18 and 28 HTG patients who had >1 copy of either the heterozygous rare mutations, >1 copy of the APOA5 p.S19W allele and neither, respectively. A significant difference in plasma TG response between genotypes was noted as follows: patients who had >1 copy of rare mutations had a blunted maximal decrease in plasma TG compared with other subjects (Figure 1). Significant between-group differences were also observed in increased plasma HDL cholesterol and decreased total cholesterol on treatment (Figure 1).

Example 2

[0047] This study included 678 consecutive unrelated Caucasian subjects from a tertiary referral lipid clinic. Patients underwent a complete medical history and examination; basic clinical, biochemical and demographic variables were collected. Normolipidemic adult controls were taken from the European subgroup of the Study of Health Assessment and Risk in Ethnic groups (SHARE), a survey of cardiovascular risk factors in Canadian sub-populations together with healthy population-based controls from Ontario. All patients provided informed consent for DNA analysis (University of Western Ontario Institutional Review Board protocol #07920E).

Biochemical determinations and classification by hyperlipoproteinemia phenotype

[0048] Plasma lipoprotein profiles were determined as described for lipid clinic patients (Hegele RA et al. (2003) Arterioscler Thromb Vase Biol 23: 111-6, the relevant contents of which are incorporated herein by reference) and for normal controls according to Anand et al. 2000, the relevant contents of which are incorporated herein by reference. Subjects were classified as having familial hypercholesterolemia (FH; HLP type 2A) based on the presence of definite diagnostic criteria as set out in Yuan G et al. (2006) CMAJ 174: 1124-9, the relevant contents of which are incorporated herein by reference, which in all cases included demonstration of heterozygosity for a disease-causing mutation (as described in Wang J et al (2005) . J Lipid Res 46: 366-72 and Wang J et al. (2001. Hum Mutat 18: 359, the relevant contents of which are incorporated herein by reference). Subjects were classified as having combined hyperlipidemia (CHL) as described (Eichenbaum-Voline S et al (2004) Arterioscler Thromb Vase Biol 24: 167-74, the relevant contents of which are incorporated herein by reference). Briefly, based on previous studies, subjects were identified as having CHL on the basis of both cholesterol and TG higher than age- and sex-specific 95th and 90th percentile values, respectively, together with the presence of cholesterol or TG higher than age- and sex-specific 90th percentile values in a blood relative. Subjects were classified as having dysbetalipoproteinemia (DBL; HLP type 3) essentially as described in Evans D et al. (2005) Clin Genet 68: 369-72, the relevant contents of which are incorporated herein by reference. Briefly, DBL was diagnosed based on the presence of an APOE E2/E2 homozygous genotype, TG exceeding age- and sex-specific 90th percentile values and/or documentation of a ratio of very-low density lipoprotein cholesterol to TG >0.30, determined as described in Whitman SC et al. (1997) . Arterioscler Thromb Vase Biol 17: 1707-15, the relevant contents of which are incorporated herein by reference.

[0049] Subjects were classified as having primary hypertriglyceridemia (HTG;

HLP type 4) based on TG concentrations exceeding age- and sex-specific 90th percentile values, but not exceeding 10 mmol/L, with no documented chylomicronemia and absence of other lipoprotein phenotypes. Subjects were classified has having severe HTG, sometimes also called 'mixed hyperlipidemia' (MHL; HLP type 5) based on fasting plasma TG >10 mmol/L documented on >2 occasions with documented chylomicronemia. Children with fasting plasma TG >10 mmol/L with documented chylomicronemia and homozygous or compound heterozygous mutations in LPL were excluded.

DNA analysis

[0050] DNA was extracted as described above. APOA5 S19W (dbSNP rs3135506) was genotyped using a validated TaqMan genotyping assay (Assay ID C 25638153 10; TaqMan® SNP Genotyping Assays, Applied Biosystems, Foster City, CA). APOA5 T[- 1131C] (dbSNP rsl 729411) was genotyped using a custom designed TaqMan genotyping assay (TaqMan® SNP Custom Genotyping Assays, Applied Biosystems, Foster City, CA). A 600 nucleotide sequence (300 upstream and 300 downstream) from NT 033899.7 was submitted to RepeatMasker (www.repeatmasker.org) to detect repetitive sequences and then the sequence was submitted to BLASTN2.2.17 to confirm unique alignment to the human Build 36 genome database. After passing these criteria, the 600 nucleotide sequence was edited to place an "N" where any other SNPs or indels were present to allow for Applied Biosystems to design the custom probe. The custom probe uses primers as follows: 5'- CCC TGC GAG TGG AGT TCA -3' and 5'- CTC TGA GCC CCA GGA ACT G. SNP genotyping was performed using an allelic discrimination assay using the 7900HT Fast Real-Time PCR System and genotypes were read using automated software (SDS 2.3, Applied Biosystems, Foster City, CA). Reactions were run in 5 μL volumes using an amplification protocol of 95⁰C for 10 minutes, followed by 50 cycles of 95⁰C for 15 seconds, then 6O⁰C for 1.5 minutes.

Statistical analysis

[0051] Analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC).

Between-group differences in discrete and quantitative traits were determined using chi- square analysis and unpaired Student t-tests, respectively. Odds ratios (OR) were calculated using the "case-control" method in the FREQ procedure in SAS. Log transformed TG were used for parametric analyses, but untransformed values are shown in Tables and Figures. Maximal likelihood linkage disequilibrium was estimated using PHASE (v2.0) (as described in Stephens M and Donnelly P (2003). Am J Hum Genet 73: 1162-9, the relevant contents of which are incorporated herein by reference). To assess the relationship of both variants with HLP and high TG concurrently, subjects who had >1 copy or 0 copies of either APOA5 S19W or -1131T>C were further classified as having "either" or "neither", respectively. Due to the relatively large number of comparisons, the nominal level for significance was adjusted to PO.01 to be conservative and minimize type I errors.

RESULTS

Clinical and biochemical features

[0052] Baseline attributes of the 678 Lipid Clinic patients, the subgroups with defined HLP phenotypes and normolipidemic controls are shown in Table 6. Lipid clinic patients were significantly older, with significantly higher body mass index, plasma total cholesterol and triglyceride concentration, but not HDL cholesterol concentration.

Table 6. Clinical, biochemical and genetic attributes of study subjects according to lipoprotein phenotypes

FH CHL DBL HTG Mixed Lipid clinic Normolipidemic

(HLP2A) (HLP2B) (HLP3) (HLP4) (HLP5) patients controls number 88 92 48 38 151 678 373 percent female 454% 46 7% 396% 10 5% 31 8% 43 1% 40 1% age (years) 57 7+13 6 56 6±11 7 51 6±12 0 59 5±13 4 50 8+12 7 54 7+14 7 472+15 2 body mass index (kg/m²) 24 1+3 4 29 1+4 3 28 9+3 1 31 2+8 1 30 5+4 8 28 7±4 7 27 2±42 plasma cholesterol (mmol/L)

- total 11 2+1 6 8 2+1 4 9 3+1 8 4 9+0 8 12 0+6 0 6 3+2 1 5 10+0 84

- high density lipoprotein 0 7+0 01 1 2+0 3 1 1+0 3 08±02 0 8+04 1 2+0 4 1 3+0 3 plasma triglyceride (mmol/L) 2 2 9+4 9 4 7±1 3 6 7±2 8 49±1 5 30 9±25 2 3 8±9 7 I 18±0 41

APOA5 S 19W

-genotype frequency

S/S 87 5% (77) 804% (74) 72 9% (35) 81 6% (31) 67 6% (102) 82 0% (550) 92 8% (346)

SAV 11 4% (10) 18 5% (17) 25 0% (12) 15 8% (6) 27 8% (42) 17 2% (122) 62% (23)

W/W 1 1% (1) 1 1% (1) 2 1% (1) 2 6% (1) 4 6% (7) 0 8% (6) 1 0% (4)

- allele frequency 7 3% 103%*** 14 6%*** 9 2%*** 18 5%*** 94%*** 42%

I - carrier frequency 12 5% 9 6%** 27 1%*** 18 4%* 32 5%*** 18 1%*** 7 2% M

Cn

Λ/^>OΛ5 -1131T>C

-genotype frequency

T/T 83 0% (73) 793% (73) 729% (35) 684%(26) 66 9% (101) 80 2% (549) 89 5% (334)

CAT 16 0% (14) 19 6% (18) 25 0% (12) 29 0%(l l) 27 1% (41) 18 6% (120) 10 2% (38)

C/C 1 0% (1) 1 1% (1) 2 1% (1) 2 6%(1) 6 0% (9) 1 2% (9) 0 3% (1)

- allele frequency 9 1% 109%** 14 6%** 17 1%** 19 5%*** 10 5%*** 5 4%

- carrier frequency 17 1% 20 7%** 27 1%** 31 6%** 33 1%*** 19 8%*** 10 4%

APOAS S19W or -1131T>C neither 71 5% (63 63 1% (58) 54 2% (26) 55 3%(21) 42 3%(64) 64 8% (439) 82 6% (308) either 28 5% (25) 369% (34) *** 45 8% (22) *** 44 7 (17) *** 57 6% (87) *** 35 2% (239) *** 174 (65)

**P<0 01, ***P<0 0001 compared to normolipidemic controls Abbreviations as in Table 1

Minimal linkage disequilibrium between ΛPOA5 variants

[0053] Pairwise linkage disequilibrium between APOA5 S19W and -1 131T>C, as estimated by the correlation coefficient r from PHASE, was 0.037 (P=0.33) and 0.017 (P=0.83) in Lipid Clinic patients and in normolipidemic controls, respectively. Estimates of pairwise linkage disequilibrium between APOA5 S19W and -1 131T>C in HLP subgroups was similarly non-significant, with r values ranging from 0.004 to 0.165 and P-values ranging from 0.23 to 0.98. Thus, in these samples, the alleles of the two SNPs were not significantly associated with each other and SNP genotypes could be considered as being independent of each other.

Higher prevalence ofAPOAS variants in lipid clinic patients

[0054] APOA5 S19W and -1131T>C genotype frequencies did not deviate significantly from expectations of Hardy- Weinberg equilibrium. Table 2 shows that APOA5 S19W was found at significantly higher allele frequency compared to controls (9.4% vs. 4.2%, respectively, PO.0001) and at significantly higher carrier frequency compared to controls (18.1% vs. 7.2%, respectively, P<0.0001). APOA5 -1 131T>C was found at significantly higher allele frequency compared to controls (10.5% vs. 5.4%, respectively, PO.0001) and at significantly higher carrier frequency compared to controls (19.8% vs. 10.4%, respectively, PO.0001). Carriers of either APOA5 S19W or -1131T>C were significantly more frequent among lipid clinic patients compared to controls (carrier frequency of 35.2% vs. 17.4%, respectively, PO.0001). The overall odds ratio for carriers of APOA5 S19W, -1131T>C or either one among lipid clinic patients was 2.98 (95% confidence interval [CI] 1.93 to 4.60), 2.01 (95% CI 1.38 to 2.95) and 2.58 (95% CI 1.89 to 3.52), respectively.

Association ofAPOAS variants with plasma TG

[0055] To further explore the association of the APOA5 variants with dyslipidemia, the 678 lipid clinic patients were stratified according to quartile of fasting plasma TG concentration and tested for between-quartile differences in the APOA5 allele and carrier frequencies. The relationship between plasma TG quartile and APOA5 variant frequencies is shown in Figure 2. A significant increasing trend of AP OA 5 S19W allele frequency across TG quartiles: 6.2%, 7.0%, 10.3% and 16.2% in quartiles 1, 2, 3 and 4, respectively (P for trend<0.0001). A stepwise relationship between APOA5 S19W carrier frequency and TG quartile: 11.8%, 13.4%, 20.0% and 30.5% in quartiles 1, 2, 3 and 4, respectively (P for trendO.OOOl) was similarly observed. A significant increasing trend of APOA5 -1131T>C allele frequency across TG quartiles: 6.5%, 7.0%, 11.5% and 15.9% in quartiles 1, 2, 3 and 4, respectively (P for trend<0.0001) was observed, as well as a stepwise relationship between APOA5 -1131T>C carrier frequency and TG quartile: 12.4%, 14.0%, 21.2% and 28.8% in quartiles 1, 2, 3 and 4, respectively (P for trendO.OOOl). When the presence of at least one copy of S19W or -1 131T>C was considered, a stepwise relationship between carrier frequency and TG quartile: 23.7%, 26.7%, 38.8% and 52.1% in quartiles 1, 2, 3 and 4, respectively (P for trendO.OOOl) was similarily observed.

Association ofAPOA5 S19Wwith classical HLP phenotypes

[0056] Using the specific diagnostic criteria outlined above, 88, 92, 48, 38 and 151 subjects were unequivocally classified with FH, CHL, DBL, HTG and MHL (HLP types 2A, 2B, 3, 4 and 5, respectively). There were no patients with HLP type 1, defined as children or adolescents with LPL deficiency due to absent post-heparin LPL activity and/or mutated LPL or APOC2 alleles. Clinical, biochemical and genetic features of study subjects are shown in Table 2. For all HLP phenotypes, APOA5 S19W and -1131T>C genotype frequencies did not deviate significantly from expectations of the Hardy- Weinberg equilibrium. All phenotype classes, except for FH, had markedly elevated plasma TG concentrations. Furthermore, APOA5 allele and carrier frequencies were significantly higher than in control subjects for all HLP phenotypes, except for FH as illustrated in Figure 3. Specifically, AP O AS S19W carrier odds ratios for HLP types 2B, 3, 4 and 5 were 3.11 (95% confidence interval [CI] 1.63 to 5.95), 4.76 (95% CI 2.25 to 10.1), 2.89 (95% CI 1.17 to 7.18) and 6.16 (95% CI 3.66 to 10.3), respectively. APOA5 -1 131T>C carrier odds ratios for HLP types 2B, 3, 4 and 5 were 2.23 (1.21 to 4.08), 3.18 (95% CI 1.55 to 6.52), 3.95 (95% CI 1.85 to 8.45) and 4.24 (95% CI 2.64 to 6.81), respectively. The presence of either allele was similarly strongly associated with HLP types: the overall odds ratio for the presence of either allele in lipid clinic patients was 2.58 (95% CI 1.89 to 3.52). The ORs for the TG-containing HLP phenotypes are shown in

Figure 2. DISCUSSION

[0057] The findings of the foregoing study are as follows:

1) a higher frequency of carriers of APO ^]A5 variants in lipid clinic patients, with overall odds ratio for carriers of S19W, -1 131T>C or either one of 2.98 (95% CI 1.93 to 4.60), 2.01 (95% CI 1.38 to 2.95) and 2.58 (95% CI 1.89 to 3.52), respectively;

2) significant stepwise relationships between APOA5 minor allele carrier frequencies and plasma triglyceride (TG) quartiles;

3) APOA5 S19W carrier odds ratios for HLP types 2B, 3, 4 and 5 of 3.11 (95% CI 1.63 to 5.95), 4.76 (95% CI 2.25 to 10.1), 2.89 (95% CI 1.17 to 7.18) and 6.16 (95% CI 3.66 to 10.3), respectively; and

4) APOA5 -1 131T>C carrier odds ratios for HLP types 2B, 3, 4 and 5 of 2.23 (95% CI 1.21 to 4.08), 3.18 (95% CI 1.55 to 6.52), 3.95 (95% CI 1.85 to 8.45) and 4.24 (95% CI 2.64 to 6.81), respectively.

[0058] Thus, APOA5 variants S19W and -1131T>C are strongly and specifically associated with HTG in lipid clinic patients and with several HLP phenotypes defined by elevated plasma TG concentration. The findings confirm the importance of these APOA5 variants in the study of patient pathogenesis and response to intervention, and also for diagnosis of HTG and TG-containing HLP phenotypes.

[0059] The associations with APOA5 S19W were consistent across the 'polygenic'

HLP type 2B, 3, 4 and 5 phenotypes. Between 30 and 60% of subjects with these four classical HLP phenotypes were carriers of either the APOA5 S 19W or - 1131T>C alleles, with significant odds ratios between 2 and 7. Thus, the results indicate that both variants individually have comparably strong associations with HTG and HLP 2B, 3, 4 and 5. Furthermore, given the absence of significant linkage disequilibrium between APOA5 S19W and -1131T>C, the genotypes function effectively as independent determinants of HTG and the essentially independent information that each provides can be combined. Perhaps the most impressive example of this is that more than half (57.6%) of HLP type 5 patients in this study carried either APOA5 variant, compared to only one-in-six normolipidemic controls (17.4%). Consistent associations so strong involving such prevalent polymorphic alleles are unusual in lipoprotein metabolism and indeed in most complex metabolic traits.

[0060] The associations with several seemingly unrelated HLP phenotypes indicate that the APOA5 variants are common determinants linking them.

[0061] In summary, APOA5 S19W and -1131T>C are herein shown to be clinical genetic markers for HTG. APOA5 S19W and -1131T>C are consistent and important genetic determinants of complex traits defined by elevated TG, specifically classical HLP phenotypes, as well as general HTG.

Example 3

[0062] Patients (148) of European geographic ancestry with severe HTG, defined as having fasting (>12 hour) plasma TG >10 mmol/L documented on 2 occasions, from a single tertiary referral lipid clinic were studied. Patients underwent complete medical history and examination; basic clinical, biochemical, and demographic variables were collected. Normolipidemic adult controls were taken from the European subgroup of the Study of Health Assessment and Risk in Ethnic groups (SHARE), and combined with healthy population-based controls from the same region of Canada. No control had ischemic heart disease and there was no use of medications among these healthy control subjects. All patients provided informed consent for DNA analysis.

DNA Analysis

[0063] DNA was extracted as described previously. For SNP genotyping, markers that were replicably associated with plasma TG and showed relatively strong association were selected. The selected genes and dbSNP identification numbers were: GALNT2 rs4846914, TBL2 rsl7145738, TRIBl rsl7321515, ANGPTL3 rsl2130333, GCKR rs780094, APOA5 rs3135506 (S 19W) and LPL rs328 (S447X). These genes were genotyped using validated genotyping assays (TaqMan® SNP Genotyping Assays, Applied Biosystems, Foster City, CA). APOA5 -1131T>C (dbSNP rs662799) was genotyped using a custom designed genotyping assay (TaqMan® SNP Custom Genotyping Assays, Applied Biosystems, Foster City, CA). The custom probe uses primers as follows: 5'- CCC TGC GAG TGG AGT TCA -3' and 5'- CTC TGA GCC CCA GGA ACT G. SNP genotyping was performed using an allelic discrimination assay using the 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA) and genotypes were read using automated software (SDS 2.3, Applied Biosystems, Foster City, CA). Reactions were run in 5 μL volumes using an amplification protocol of 95⁰C for 10 minutes, followed by 42 cycles of 95⁰C for 15 seconds, then 6O⁰C for 1.5 minutes. An established method was used to genotype APOE isoforms (described in Hixson et al. (1990). J Lipid Res 31, 545-548. the relevant contents of which are incorporated herein by reference).

[0064] For SNP analysis, patients with a known sequence-proven loss-of-function mutation in LPL, APOC2 or APOA5, encoding lipoprotein lipase, apo C-II and apo A-V, respectively, were excluded. Blinded between-day replicated genotypes of a random 3% of samples showed >99.9% concordance across all markers.

Statistical Analysis

[0065] The two-sample t-test was used to compare the difference between case and control groups for quantitative traits, while Pearson's chi-square test was used to compare discrete traits with exact P-values obtained whenever cell sizes <5. Deviations of genotype frequency from the Hardy-Weinberg assumption were assessed using a chi-square test. Maximal likelihood linkage disequilibrium was estimated using PHASE v2.0. To assess the relationship of SNPs with severe HTG, dominant and recessive models of minor allele genotypes were tested for each gene. A simple logistic regression model was used to assess univariate association between each SNP and severe HTG. A multiple logistic regression model with backward elimination procedure was adopted to assess the joint effects of genes and clinical variables such as presence of diabetes and marked obesity, i.e. body mass index (BMI) >33 kg/m². The adequacy of the final models was assessed using the Hosmer- Lemeshow goodness-of-fit test. Relative importance of genetic and clinical variables was quantified using the R-square computed with logistic regression raw residuals as described in Heinze et al. (2003). Comput Methods Programs Biomed 71, 155-163, the relevant contents of which are incorporated herein by reference. Statistical significance was taken at nominal P-value <0.05 for all comparisons. All analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC), with the exception the exact tests which were performed using StatXactδ (Cytel Inc, Cambridge, MA). [0066] Baseline clinical and biochemical attributes of the study sample are shown in Table 7.

Table 7. Clinical, biochemical and genetic attributes of study subjects

[0067] After excluding 16 patients with heterozygous loss-of-function mutation in

LPL, APOC2 or APOA5, 132 patients or cases with severe HTG remained for analysis. These were each matched with up to 4 normolipidemic controls based on age within 5 years and sex. By definition, severe HTG patients had markedly higher plasma TG and total cholesterol and significantly lower HDL cholesterol (Table 7). Plasma TG concentration in severe HTG patients ranged from 10.1 to 180 mmol/L. In addition, 37/132 severe HTG patients (28.0%) had been hospitalized on >1 occasion with pancreatitis.

Results

Differences in Distribution of DNA Variants between Severe HTG Cases and Controls

[0068] Genotype counts and frequencies in severe HTG patients and controls are shown in Table 8.

Table 8. Genotype counts and frequencies of candidate genes evaluated

Abbreviations APOA5, gene encoding apolipoprotein A-V, GCKR, gene encoding glucokinase receptor, TRIBl, gene encoding homologue ofDrosophila Tnbbles 1, GALNT2, gene encoding UDP-N-acetyl-alpha-D-galactosamine polypeptide N-acetylgalactosaminyltransferase, TBL2 gene encoding transducin-beta-like-2, ANGPTL3 gene encoding angiopoietin-hke 3, APOE, gene encoding apolipoprotein E, LPL, gene encoding lipoprotein lipase [0069] Minor allele frequencies (MAFs) for each genotype in severe HTG cases and controls are shown in Table 9.

Table 9. Candidate Gene Minor Allele Frequencies

[0070] Frequencies of each genotype did not deviate from Hardy-Weinberg equilibrium. The significance of the differences in genotype frequencies between severe HTG cases and controls are shown in Table 9: In univariate chi-square analysis, genotype frequencies of each evaluated marker had a significantly different distribution in HTG cases compared with controls (range of P-values 0.024 to 1.5 X 10^'12). The significance of the differences in MAFs between severe HTG cases and controls are shown in Table 9: The MAF of each marker studied was significantly different between groups.

Genetic Risk of Severe Hypertriglyceridemia: Univariate Odds Ratios

[0071] Univariate ORs for severe HTG were determined for two clinical variables - namely diabetes and marked obesity (defined as BMI >33 kg/m2) and for the HTG-risk genotype for each genomic variant. Both dominant and recessive models for each genotype were evaluated and the model that provided the strongest and most significant OR was chosen to serve as the nominal genotype variable for subsequent multivariate analyses. There was no significant linkage disequilibrium between the two APOA5 variants (P=0.23), so these were treated as independent variables for the purpose of subsequent analyses. For APOE, presence or absence of the common E3 allele was evaluated.

[0072] Univariate ORs and 95% CIs are shown in Table 10 for the most significantly associated genetic model for each genotype: only the ANGPTL3 and LPL genotypes were not significant for either dominant or recessive model. However, both APOA5 variants, APOE non-E3/3 genotype, GCKR AA recessive genotype; TRIBl AA recessive genotype and TBL2 CC recessive genotype each had significant ORs for severe HTG.

Table 10: Univariate Odds Ratios for Severe HTG

[0073] Variables entered into model are defined as follows: APOA5 Wl 9 dominant had the test genotypes SW and WW and the reference genotype SS; APOA5 -1 131C dominant had the test genotypes CC and TC and the reference genotype TT; GALNT2 G recessive had the test genotype GG and the reference genotypes AA and AG; GCKR A recessive had the test genotype AA and the reference genotypes GA and GG; TBL2 C recessive had the test genotype CC and the reference genotypes CT and TT; APOE non-E3 allele had the test genotypes 2/2, 4/2, 4/4 and the reference genotypes 3/2, 3/3 and 4/3; TRIBl A recessive had the test genotype AA and the reference genotypes AG and GG; LPL S447 recessive had the test genotype SS and the reference genotypes SX (there were no XX individuals); ANGPTL3 C recessive had the test genotype CC and the reference genotypes CT and TT. The last 6 rows show odds ratios for individuals with combinations of at-risk genotypes, starting with either APOA5 genotype and then adding non-APOA5 at-risk genotypes. [0074] Univariate ORs for severe HTG were also determined for combinations of genetic variables. Since both APOA5 variants were very strongly associated with HTG, the presence of either served as the primary genetic predictor: the OR was 6.93 (95% CI 4.44 to 10.8). Adding any one or two of the other genetic variables did not substantially change this OR. However, adding 3, 4 or 5 additional genetic markers to the presence of either APOA5 marker sequentially increased the OR from 7.58 to 8.92 to 25.0, so that when an individual had 7 genetic risk markers (i.e. APOA5 plus any other five), the resulting OR for severe HTG was very high indeed.

Polygenic Determinants of Severe Hypertriglyceridemia: multivariate regression analysis

[0075] The multivariate ORs for severe HTG were calculated using the WaId statistic in multivariate logistic regression analysis with stepwise addition of variables and P<0.05 for each step (Table 1 1). The first model, which included two clinical variables in addition to nine genetic variables, found that diabetes, obesity, two APOA5 markers, APOE non-E3 genotype and GCKR, TRIBl and TBL2 genotypes were significantly associated with severe HTG. The C-statistic, which corresponds to the area under the receiver-operator curve for a diagnostic test, was 0.869 for this particular combination of clinical and genetic markers (Table 11). Hosmer and Lomeshow goodness of fit test showed that the models explained the observed data (χ₈ ² = 10.2; P=0.25 and χ₈ ² = 7.47; P=0.38). The second model assessed only genetic variables: the same genotypes from the first model remained significantly associated in the second model with one additional significantly associated genotype - namely GALNT2, assuming a recessive effect for the G allele. The C-statistic was 0.800 for this combination of genetic markers (Table 11). Table 11: Multivariate Odds Ratios for Severe HTG

The model used backward elimination and had a nominal P-value of 0 05 for each variable [0076] The proportion of contribution of specific variables to severe HTG was calculated using partial i^-values in multivariate linear regression analysis with stepwise addition of variables and P<0.05 for each step (Table 11). The first model, which included two clinical variables in addition to nine genetic variables, found that diabetes, APOA5 markers, obesity, TBL2 genotype, APOE genotype, TRIBl genotype and GCKR genotype were significantly associated with severe HTG. The model explained -43% of total variation in case versus control status, and of the explained variation, the total contribution of the genetic variables was -40% (range -1 to 25%). The second model assessed only genetic variables: the same genotypes from the first model remained significantly associated in the second model with one additional significantly associated genotype - namely GALNT2. The model accounted for -25% of total variation in case versus control status. Of explained variation, genetic markers accounted for at least -1% each as shown in Figure 4.

Discussion

[0077] In this study, the following was identified in patients with severe HTG:

\) APOA5 S19W, APOAS -1 131OT, APOE, GCKR rs780094, TRIBl rsl7321515, GALNT2 rs4846914 and TBL2 rsl7145738, were significantly associated with severe HTG;

2) ORs for these genetic variables were significant in both univariate and multivariate regression analyses, irrespective of the presence or absence of diabetes or obesity; and

3) a significant fraction - about one-quarter - of the attributable variation in disease status was associated with these genotypes.

[0078] In the current study, in which subjects with rare loss-of-function mutations were excluded, it was found that common SNP alleles, including those in APOA5 and those in TRIBl, GCKR, TBL2, GALNT2 and ANGPTU are found in a substantial proportion - almost two-thirds - of individuals with severe HTG. Thus, the genetic component of this complex metabolic trait is comprised of both rare and common variants.

Claims

CLAIMSWe Claim:

1. A method of assessing the risk of HTG in a mammal comprising determining in a nucleic acid-containing sample from the mammal the presence of at least one polymorphism within the APOA5 gene, wherein the identification of an APOA5 polymorphism is indicative of a risk of HTG.

2. A method as defined in claim 1, wherein the APOA5 polymorphism is selected from the group consisting of APOA5 S19W and APOA5 -1131OT.

3. A method as defined in claim 1, additionally comprising the identification of one or more secondary polymorphisms selected from the group consisting of GCKR rs780094 A allele, TPJBl rsl7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and APOE non-E3 allele.

4. A method as defined in claim 3, comprising the identification of at least two of the secondary polymorphisms.

5. A method as defined in claim 4, comprising the identification of each of the secondary polymorphisms.

6. A method of assessing the risk of HTG in a mammal comprising determining in a nucleic acid-containing sample from the mammal the presence of at least one of the AP0A5 polymorphisms, AP O AS S19W and APOA5 -1131OT, in combination with one or more secondary polymorphisms selected from the group consisting of GCKR rs780094 A allele, TRIBl rsl7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and APOE non-E3 allele.

7. A method as defined in claim 6, additionally comprising the identification of clinical risk factors.

8. A method as defined in claim 7, wherein the clinical risk factors are selected from the group consisting of diabetes, obesity, age, sex, diet, alcohol intake, ethnicity, total fat content and psychological state.

9. A method of assessing the risk of HTG in a mammal comprising:

1) identifying the occurrence of HTG-related polymorphisms in said mammal; and

2) determining the risk based thereon.

10. A method as defined in claim 9, wherein said HTG-related polymorphisms are selected from the group consisting of AP OA 5 polymorphisms, GCKR rs780094 A allele, TRIBl rsl7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and APOE non-E3 allele.

1 1. A method as defined in claim 9, wherein step (1) additionally comprises identifying HTG-related clinical factors for said mammal.

12. A method as defined in claim 1 1, wherein said clinical factors are selected from the group consisting of diabetes, obesity, age, sex, diet, alcohol intake, ethnicity, total fat content and psychological state.

13. A method of assessing the risk of disease associated with HTG in a mammal comprising the step of identifying the occurrence in said mammal of at least one APOA5 polymorphism, wherein the identification of one or more APOA5 polymorphisms is indicative of a risk of disease associated with HTG.

14. A method as defined in claim 13, wherein the disease is selected from the group consisting of metabolic syndrome, coronary heart disease and pancreatitis.

15. A method as defined in claim 13, wherein the APOA5 polymorphism is selected from the group consisting of APOA5 S19W and APOA5 -1131OT.

16. A method as defined in claim 13, additionally comprising the identification of the occurrence of one or more secondary polymorphisms selected from the group consisting of GCKR rs780094 A allele, TRIBl rs 17321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and APOE non-E3 allele, in said mammal.

17. A kit useful to assess risk of HTG in a mammal comprising at least one reagent useful to identify the presence of at least one APOA5 polymorphism.

18. A kit as defined in claim 17, wherein the polymorphism is selected from the group consisting of APOA5 Sl 9 W and APOA5 -1131OT.

19. A kit as defined in claim 17, additionally comprising one or more reagents useful to identify the presence of one or more secondary polymorphisms selected from the group consisting of GCKR rs780094 A allele, TRIBl rsl7321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl 7145738 T allele and APOE non-E3 allele.

20. A kit as defined in claim 17, comprising an indication of the risk of severe HTG associated with each polymorphism.

21. A kit as defined in claim 17, comprising an indication of the risk of severe HTG associated with combinations of said polymorphisms.

22. An array useful to assess the risk HTG in a mammal comprising reagents useful to detect at least one APOA5 polymorphism, and at least one secondary polymorphism selected from the group consisting of GCKR rs780094 A allele, TRIBl rs 17321515 G allele, GALNT2 rs4846914 G allele, TBL2 rsl7145738 T allele and APOE non-E3 allele.