University of Nebraska - Lincoln University of Nebraska - Lincoln
DigitalCommons@University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln
College of Law, Faculty Publications Law, College of
5-1995
A candidate genetic risk factor for vascular disease: A common A candidate genetic risk factor for vascular disease: A common
mutation in methylenetetrahydrofolate reductase mutation in methylenetetrahydrofolate reductase
P. Frosst
McGill University
H. J. Blom
University Hospital Nijmegen, Netherlands
R. Milos
McGill University
P. Goyette
McGill University
Christal A. Sheppard
University of Nebraska-Lincoln
, christalsheppard@unl.edu
See next page for additional authors
Follow this and additional works at: https://digitalcommons.unl.edu/lawfacpub
Part of the Legal Studies Commons
Frosst, P.; Blom, H. J.; Milos, R.; Goyette, P.; Sheppard, Christal A.; Matthews, R. G.; Boers, G. J.H.; den
Heijer, M.; Kluijtmans, L. A.J.; van den Heuvel, L. P.; and Rozen, Rima, "A candidate genetic risk factor for
vascular disease: A common mutation in methylenetetrahydrofolate reductase" (1995).
College of Law,
Faculty Publications
. 124.
https://digitalcommons.unl.edu/lawfacpub/124
This Article is brought to you for free and open access by the Law, College of at DigitalCommons@University of
Nebraska - Lincoln. It has been accepted for inclusion in College of Law, Faculty Publications by an authorized
administrator of DigitalCommons@University of Nebraska - Lincoln.
Authors Authors
P. Frosst, H. J. Blom, R. Milos, P. Goyette, Christal A. Sheppard, R. G. Matthews, G. J.H. Boers, M. den
Heijer, L. A.J. Kluijtmans, L. P. van den Heuvel, and Rima Rozen
This article is available at DigitalCommons@University of Nebraska - Lincoln: https://digitalcommons.unl.edu/
lawfacpub/124
Abstract
Hyperhomocysteinaemia has been identied as a risk factor for
cerebrovascular, peripheral vascular, and coronary heart dis-
ease.
1-4
Elevated levels of plasma homocysteine can result from
genetic or nutrient-related disturbances in the trans-sulphuration
or re-methylation pathways for homocysteine metabolism.
1,5–7
5,10-Methylenetetrahydrofolate reductase (MTHFR) catalyzes
the reduction of 5,10-methylenetetrahydrofolate to 5-methyltetra-
hydrofolate, the predominant circulatory form of folate and carbon
donor for the re-methylation of homocysteine to methionine. Re-
duced MTHFR activity with a thermolabile enzyme has been re-
ported in patients with coronary and peripheral artery diseases.
6
We have identied a common mutation in MTHFR which alters
a highly-conserved amino acid; the substitution occurs at a fre-
quency of approximately 38% of unselected chromosomes. The
mutation in the heterozygous or homozygous state correlates with
reduced enzyme activity and increased thermolability in lympho-
cyte extracts; in vitro expression of a mutagenized cDNA contain-
ing the mutation conrms its effect on thermolability of MTHFR. Fi-
nally, individuals homozygous for the mutation have signicantly
elevated plasma homocysteine levels. This mutation in MTHFR
may represent an important genetic risk factor in vascular disease.
Severe MTHFR deciency, the most common inborn error
of folate metabolism, results in hyperhomocysteinaemia, ho-
mocystinuria, and hypomethioninaemia. Patients with severe
MTHFR deciency (0-20% residual activity in cultured bro-
blasts) present in infancy or adolescence with developmental
delay, motor and gait dysfunction, seizures, psychiatric distur-
bances, and other neurological abnormalities; they are also at
risk for vascular complications.
8
Individuals with 50% resid-
ual activity, due to a thermolabile form of the reductase, were
rst reported in approximately 17% of 212 North American
patients with coronary artery disease.
5
A recent study of the
Netherlands population identied the thermolabile variant in
different forms of premature vascular disease,
6
and estimated
its incidence to be 7% of vascular patients. The presence of a
thermolabile MTHFR is predictive of coronary artery stenosis,
independent of other risk factors, such as age, smoking, hyper-
cholesterolaemia, and hypertension.
9
Our recent isolation of a cDNA for human MTHFR
10
has
enabled us to identify nine mutations in this gene, in the se-
verely-decient group of patients, by SSCP analysis and direct
sequencing of PCR fragments.
10,11
Using the same procedures,
we identied a C to T substitution at nucleotide (nt) 677, which
converts an alanine to a valine residue (Figure 1a). This altera-
tion creates a HinfI site (Figure 1b), which was used to screen
114 unselected French Canadian chromosomes; the allele fre-
quency of the substitution was 0.38. The frequency of the three
genotypes were as follows: –/–, 37%; +/–, 51%; and +/+, 12%
(+ indicates the presence of the HinfI site and a valine residue).
As these individuals were not examined clinically or biochem-
ically, they cannot be considered as a control group.
We next performed genotypic analysis and measured en-
zyme activity and thermolability in a total of 40 lymphocyte
pellets from patients with premature vascular disease and
controls (Table I). We selected 13 vascular patients from our
previous study, among whom ve were considered to have
thermolabile MTHFR.
6
From a large reference group of 89
controls, we studied all seven individuals who had thermo-
labile MTHFR, and selected at random an additional 20 con-
trols with normal MTHFR from the same reference group.
The mean MTHFR activity for individuals homozygous for
the Ala to Val substitution (+/+) was approximately 30% of
the mean activity for (–/–) individuals. Heterozygotes had a
mean MTHFR activity of 65% compared to (–/–) individuals,
intermediate between values for (–/–) and (+/+) individuals.
The ranges of activities showed some overlap for the hetero-
zygous and (–/–) genotypes, but homozygous (+/+) individ-
uals showed virtually no overlap with the other two groups.
A one-way analysis of variance yielded a P value <.0001;
Published in Nature Genetics 10 (May 1995), pp. 111–113.
Copyright © 1995 Nature Publishing Group. Used by permission.
Submitted December 7, 1994; accepted March 1995.
A candidate genetic risk factor for vascular disease:
A common mutation in methylenetetrahydrofolate reductase
P. Frosst,
1
H.J. Blom,
2
R. Milos,
1
P. Goyette,
1
C. A. Sheppard,
3
R. G. Matthews,
3
G.J.H. Boers,
4
M. den Heijer,
2
L.A.J. Kluijtmans,
2
L. P. van den Heuvel,
2
and R. Rozen
1
1. Departments of Human Genetics, Pediatrics and Biology, McGill University, Montreal Children’s Hospital, Montreal, Canada H3H 1P3
2. Department of Pediatrics, University Hospital Nijmegen, 6500 HB Nijmegen, The Netherlands
3. Biophysics Research Division and Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan, 48109, USA
4. Department of Medicine, University Hospital Nijmegen, 6500 HB Nijmegen, The Netherlands
5. Department of Hematology, Municipal Hospital Leyenburg, 2545 CH The Hague, The Netherlands
Corresponding author — R. Rozen
111
Figure 1. Sequence change and restriction enzyme analysis for the al-
anine to valine substitution. a) Sequence of two individuals, a homo-
zygote for the alanine residue and a homozygote for the valine resi-
due. The antisense strands are depicted. The primers for analysis of
the A→V change are: 5’-TGAAGGAGAA GGTGTCTGCG GGA-3’ (ex-
onic) and 5’-AGGACGGTGC GGTGAGAGTG-3’ (intronic); these prim-
ers generate a fragment of 198 bp. b) The substitution creates a HinfI
recognition sequence which digests the 198 bp fragment into 175 and
23 bp fragments; the latter fragment has been run off the gel. All three
possible genotypes are shown.
112 P. Fr o ss t et a l . i n Nat u r e G e N e t i c s 10 (1995)
a pairwise Bonferroni t test showed that all three genotypes
were signicantly different with P < 0.01 for the three possi-
ble combinations.
The three genotypes were all signicantly different (P < .01)
with respect to enzyme thermolability. The mean residual ac-
tivity after heat inactivation (5 min at 46°C) was 67% (–/–),
56% (+/–) and 22% (+/+). While the degree of thermolability
overlaps somewhat for (–/–) individuals and heterozygotes,
individuals with two mutant alleles had a distinctly lower
range. Every individual with the (+/+) genotype had residual
activity <35% after heating, and specic activity <50% of that
of the (–/–) genotype.
Fasting homocysteine levels in (+/+) individuals were al-
most twice the value for (+/–) and (–/–) individuals. The dif-
ferences among genotypes for plasma homocysteine were
maintained when homocysteine was measured following six
hours of methionine loading. A one-way anova yielded a P <
.01 for the fasting and post-methionine homocysteine levels.
A pairwise Bonferroni t test showed that homozygous mutant
individuals had signicantly elevated homocysteine levels (P
< .05),compared to either (+/–) or (–/–) individuals.
We have used the original MTHFR cDNA (1.3 kb) to isolate
a 2.2 kb cDNA, which contained an additional 900 bp at the 3’
end; the latter contained a termination codon, 100 bp of 3’ UTR
and a poly A tail (GenBank# UO9806). The open reading frame
of 1980 bp predicts a protein of 74.6 kDa. The puried porcine
liver enzyme has been shown to have subunits of 77 kDa.
12
Western analysis (Figure 2a) of several human tissues and of
porcine liver reveals a polypeptide of 77 kDa, as well as an ad-
ditional polypeptide of approximately 70 kDa in human fetal
liver and in porcine liver, suggesting the presence of isozymes.
The wild-type cDNA and a mutagenized cDNA, containing
the Ala to Val substitution, were expressed in E. coli to yield
a protein of approximately 70 kDa (Figure 2a), which co-mi-
grates with the smaller polypeptide mentioned above. Treat-
ment of extracts at 46°C for ve minutes revealed that the
enzyme containing the substitution was signicantly more
thermolabile than the wild -type enzyme (P < .001; Figure 2b).
The expression experiments were not designed to measure
differences in specic activity before heating, since variation
in efciencies of expression could contribute to difculties in
interpretation. Curiously though, the specic activity for the
mutant construct was higher in both experiments. It is possi-
ble that the mutant protein has increased stability in E. coli, or
that inclusion bodies in our extracts contributed to differences
in recovery of properly-assembled enzyme.
The alanine residue is conserved in porcine MTHFR and
in the corresponding bacterial metF genes.
10
We have also ob-
served a region of homology in the human dihydrofolate re-
ductase (DHFR) gene,
11
although the alanine residue itself is
not conserved; this region of amino acids 130-149 of DHFR
contains Thr136, which has been implicated in folate binding
of human DHFR.
13
This region in MTHFR might also be in-
volved in folate binding, and the enzyme may be stabilized in
the presence of folate. This hypothesis is compatible with the
well-documented inuence of folate on homocysteine levels
7,14
and with the reported correction of mild hyperhomocysteinae-
mia by folic acid in individuals with premature vascular dis-
ease
14
and thermolabile MTHFR.
15
Table 1. Correlation between MTHFR genotype and enzyme activity,
thermolability, and plasma homocysteine level
Genotype
–/– +/– +/+
n=19 n=9 n=12
Specic activity
a,b
22.9±1.7 15.0±0.8 6.9±0.6
(nmol CH
2
O/mg protein/hr) (11.8–33.8) (10.2–18.8) (2.6–10.2)
Residual activity 66.8±1.5 56.2±2.8 21.8±2.8
after heatings
a,b
(%) (55–76) (41–67) (10–35)
Plasma homocysteine
a,c
12.6±1.1 13.8±1.0 22.4±2.9
(μM)(after fasting) (7–21) (9.6–20) (9.6–42)
Plasma homocysteine
a,c
41.3±5.0
d
41±2.8 72.6±11.7
e
(μM)(post-methionine load) (20.9-110) (29.1-54) (24.4-159)
Enzyme activity and plasma homocysteine were determined as pre-
viously reported.
6
Each value represents mean ± standard error. The
range is given in parentheses below the mean. a. one-way anova P <
.01. b. paired t test for all combinations P < .01. c. paired t test P < .05
for +/+ group versus +/– group or –/– group; P > .05 for +/– versus –/–
group. d. n = 18. e. n = 11.
Figure 2. Expression analysis of MTHFR in E. coli. a) Western blot of bacterial extracts and tissues. Two μg of bacterial extract protein was used
for lanes 1-3. The tissues (lanes 4-6) were prepared by homogenization in 0.25 M sucrose with aprotinin and leupeptin, followed by sonication
on ice. The extracts were spun for 15 min in a microcentrifuge at 14,000g and 100 μg of supernatant protein was used for western analysis. h,
human; p, porcine. b) Thermolability assay of bacterial extracts. Two separate experiments (with 3-4 replicates construct experiment) were per-
formed to measure thermostable activity of the wild-type and mutagenized MTHFR cDNAs. The values shown represent mean ± standard error for
each experiment, as % of residual activity after heating. The means of the specic activities before heating (expressed as nmol formaldehyde/hr/
mg protein) were as follows: 3.8 and 5.3 (Exp. 1) and 6.2 and 7.5 (Exp. 2) for MTHFR and MTHFR A→V, respectively.
Mu tat i on in MtHFr a s c an d i d at e g e n et i c r i s k Fa c to r F o r va s cu l ar d i se a se 113
Our data have identied a common genetic change in
MTHFR which results in thermolability; our experiments do
not directly address the relationship between this change and
vascular disease. Nonetheless, this mutation represents a di-
agnostic test for evaluation of MTHFR thermolability in hy-
perhomocysteinaemia. Large case-control studies are required
to evaluate the frequency of this genetic change in various
forms of occlusive arterial disease and to examine the interac-
tion between this genetic marker and dietary factors, such as
folate intake. Well-dened populations need to be examined,
as the limited data set thus far suggests that population-spe-
cic allele frequencies may exist. More importantly, however,
the identication of a candidate genetic risk factor for vascu-
lar disease, which may be inuenced by nutrient intake, rep-
resents a critical step in the design of appropriate therapies for
the homocysteinaemic form of arteriosclerosis.
Methodology
Mutation identication. Primers were designed from the cDNA se-
quence to generate 250-300 bp fragments which overlapped 50-75 bp
at each end. When PCR amplication of human genomic DNA yielded
larger fragments than expected for the coding region alone, these frag-
ments were presumed to contain introns and were sequenced directly
(Cycle Sequencing kit, GIBCO). Intronic primer sequences were ob-
tained with this strategy. PCR products were analyzed by a non-ra-
dioactive SSCP protocol as described.
10
Fragments showing a shift
on SSCP gels were subcloned into Bluescript and sequenced (Seque-
nase kit, USB). To conrm the sequence changes, a new PCR was per-
formed with genomic DNA; the PCR product was digested with HinfI
and analyzed by polycrylamide gel electrophoresis.
Clinical material. To determine the frequency of the A→V mutation,
DNA from 57 individuals from Quebec was analyzed by PCR and re-
striction digestion. The individuals, all French Canadian, were not ex-
amined clinically or biochemically. The 40 individuals analyzed in Ta-
ble 1 have been described.
6
Of the 13 cardiovascular patients, eight
had cerebrovascular arteriosclerosis, and ve had peripheral arterio-
sclerosis. Five had thermolabile MTHFR, while eight had thermosta-
ble MTHFR (>33% residual activity after heating). Controls and patients
were all Dutch-Caucasian, between 20–60 years of age. None of these in-
dividuals used vitamins that could alter homocysteine levels. Enzyme
assays and homocysteine determinations have also been reported.
6
Constructs for expression analysis. A human colon carcinoma cDNA
library (gift of Nicole Beauchemin, McGill University) was screened by
plaque hybridization with the original 1.3 kb cDNA
10
to obtain addi-
tional coding sequences. A cDNA of 2.2 kb was isolated, which contained
900 additional bp at the 3’ end (Genbank accession number UO9806). Se-
quencing was performed on both strands for the entire cDNA. Addi-
tional 5’ sequences (800 bp) were obtained from a human kidney cDNA
library (Clontech), but these sequences did not contain additional coding
sequences and were therefore used for the PCR-based mutagenesis only
(see below) and not for the expression analysis. The two cDNAs (2.2 kb
and 800 bp) were ligated using the EcoRI site at nt 199 and inserted into
Bluescript (Stratagene). The 2.2 kb cDNA was subcloned into the expres-
sion vector pTrc99A (Pharmacia) using the NcoI site at nt 11 and the XbaI
site in the polylinker region of both vectors. Sequencing was performed
across the cloning sites to verify the wild-type construct.
PCR-based mutagenesis, using the cDNA-containing Bluescript
vector as template, was used to create the A to V mutation identied.
16
Vent polymerase (NEB) was used to reduce PCR errors. The following
primers were used: primer 1, bp -200 to -178, sense; primer 2, bp 667
to 687, antisense, containing a mismatch, A, at nt 677; primer 3, 667 to
687, sense, containing a mismatch, T, at nt 677; primer 4, bp 1092 to
1114, antisense. PCR was performed using primers 1 and 2 to generate
a product of 887 bp, and using primers 3 and 4 to generate a product
of 447 bp. The two PCR fragments were isolated from a 1.2% agarose
gel by Geneclean (BIO 101). A nal PCR reaction, using primers 1 and
4 and the rst two PCR fragments as template, was performed to gen-
erate a 1.3 kb band containing the mutation. The 1.3 kb fragment was
digested with NcoI and MscI, and inserted into the wild-type cDNA-
containing expression vector by replacing the sequences between the
NcoI site at bp 11 and the MscI site at bp 943. The entire replacement
fragment and the cloning sites were sequenced to verify that no addi-
tional changes were introduced by PCR.
Expression analysis. Overnight cultures of JM 105 containing vector,
wild-type, or mutagenized MTHFR cDNA were grown at 37 °C in 2
× YT media with .05 mg ml
–1
ampicillin. Fresh 10 ml cultures of each
were inoculated with approximately 50 μl of overnight cultures and
grown at 37 °C to an O.D. of 1 at 420 nM. Cultures were then induced
for 2 h with 1 mM IPTG and pelleted. The cells were resuspended in
TE buffer with 2 μg ml
–1
aprotinin and leupeptin (3.5 × wet weight
of cells). Cell suspensions were sonicated on ice for 3 × 15 s and cen-
trifuged for 30 min at 4 °C to pellet cell debris and unlysed cells. The
supernatant was removed and assayed for protein concentration with
the Bio-Rad protein assay. Western analysis was performed using the
Amersham ECL kit with antiserum generated against puried porcine
liver MTHFR.
12
Enzymatic assays were performed by established pro-
cedures.
17
Thermolability was assessed by pre-treating the extracts at
46 °C for 5 min before determining activity. Specic activities (nmol
formaldehyde/h/mg protein) were calculated for the 2 cDNA-con-
taining constructs after subtraction of the values obtained with vector
alone (to subtract background E. coli MTHFR activity).
Acknowledgments — We thank E. Stevens, H. van Lith-Zanders,
C Mandel, and N. Beauchemin for their contribution to this work.
This work was supported by the Medical Research Council of Can-
ada (R.R.), the Canadian Heart and Stroke Foundation (R.R.), in part
by the Netherlands Heart Foundation (H.B.), and by NIH Grant R37
GM24908 (CA.S. and R.G.M.). R.R. is a Principal Investigator of the
MRC Group in Medical Genetics.
References
1. Boers. G.H.J. et al. Heterozygosity for homocystinuria in premature pe-
ripheral and cerebral occlusive arterial disease. New Engl. J. Med.
313, 700-715 (1985).
2. Genest, J.J. Jr. et al. Plasma homocysteine levels In men with premature
coronary artery disease. J. Am. Coll. Cardiol. 18, 1114-1119 (1990).
3. Clarke, R. et aI. Hyperhomocysteinemia: An independent risk factor for
vascular disease. New Engl. J. Med. 324, 1149-1155 (1991)
4. Stampfer, M. J. et aI. A prospective study of plasma homocysteine and
risk of myocardial infarction in US physicians. J. Am. Med. Assoc. 288,
877–881 (1992).
5. Kang, S.-S. et aI. Thermolabile methylenetetrahydrofolate reductase: An
inherited risk factor for coronary artery disease. Am. J. Hum. Genet.
48, 536–545 (1991).
6. Engbersen, A.M.T. et al. Thermolabile 5,10-methylenetetrahydrofolate
reductase as a cause of mild hyperhomocysteinaemia. Am. J. Hum.
Genet. 56, 142-150 (1995).
7. Selhub, J., Jacques, P. F., Wilson, P.W.F., Rush, D., & Rosenberg, I. H.
Vitamin status and intake as primary determinants of homocysteinemia
in an elderly population, J. Am. Med. Assoc. 270, 2693-2698 (1993).
8. Rosenblatt, D. S. in The Metabolic Basis of Inherited Disease (ed.
Scriver, C. R., Beaudet, A. L, Sly, W. S., & Valle, D.) 2049-2064 (Mc-
Graw-Hill, New York, 1989).
9. Kang, S.-S., Passen, E. L., Ruggie, N., Wong, P.W.K., & Sora, H. Ther-
molabile defect of methylenetetrahydrofolate reductase in coronary ar-
tery disease. Circulation 88,1463-1469 (1993).
10. Goyette, P. et al. Human methylenetetrahydrofolate reductase: Isola-
tion of cDNA, mapping and mutation identication. Nature Genet. 7,
195-200 (1994).
11. Goyette, P., Frosst, P., Rosenblatt, D. S. ,& Rozen, R. Seven novel mu-
tations in the methylenetetrahydrofolate reductase gene and genotype/
phenotype correlations in severe methylenetetrahydrofolate reductase
deciency. Am. J. Hum. Genet. 56, 1052-1059 (1995).
12. Daubner, S. C. & Matthews, R. G. Purication and properties of methy-
lenetetrahydrofolate reductase from pig liver. J. Biol. Chem. 251. 140-
145 (1982).
13. Davies, J. F., et al. Crystal structures of recombinant human dihydrofo-
late reductase complexed with folate and 5-deazafolate, J. Biochem.
29, 9467-9479 (1990).
14. Franken, D. G., Boers, G.H.J., Blom, H. J, Trijbels, J.M.F., & Kloppen-
borg, P.W.C. Treatment of mild hyperhomocysteinemiain vascular dis-
ease patients. ArterioscIer. Thromb. 14, 465-470 (1994).
15. Kang, S.-S., Zhou, J., Wong, P.W.K., Kowalisyn, J., & Strokosch, G. In-
termediate homocysteinemla: A thermolabile variant of methylenetetra-
hydrofolate reductase. Am. J. Hum. Genet. 43, 414-421 (1988).
16. Horton, R. M., et al. Gene splicing by overlap extension. Math. Enzy-
mol. 217, 270-279 (1993).
17. Rosenblatt, D. S. & Erbe, R. W. Methylenetetrahydrofolate reductase
in cultured human cells. I. Growth and metabolic studies. Pediatr. Res.
11, 1137-1141 (1977).
