GENETIC AND MOLECULAR DISSECTION OF HOMOCYSTEINEMIA IN MICE
by
SHEILA ERNEST
Submitted in partial fulfillment of the requirements
For the degree of Doctor of Philosophy
Dissertation Adviser: Dr. Joseph H. Nadeau
Department of Genetics
CASE WESTERN RESERVE UNIVERSITY
August, 2004 CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
______
candidate for the Ph.D. degree *.
(signed)______(chair of the committee)
______
______
______
______
______
(date) ______
*We also certify that written approval has been obtained for any proprietary material contained therein. À mes parents très chers 1
TABLE OF CONTENTS 2
TABLE OF CONTENTS ...... 1 LIST OF TABLES...... 4 LIST OF FIGURES ...... 7 ACKNOWLEDGEMENTS...... 11 ABSTRACT...... 12 CHAPTER I INTRODUCTION AND RESEARCH AIMS...... 14 HOMOCYSTEINEMIA ...... 15 Definitions of homocysteinemia ...... 15 Environmental determinants of homocysteinemia ...... 16 Genetic determinants of homocysteinemia...... 19 HOMOCYSTEINEMIA AND HUMAN DISEASES...... 22 Neural Tube Defects...... 23 Vascular disease ...... 26 Cancer ...... 28 Neurodegenerative disorders...... 30 Down Syndrome ...... 31 HOMOCYSTEINEMIA AND ANIMAL MODELS ...... 32 Environmental determinants of homocysteinemia ...... 32 Mouse models with deficiencies of genes involved in HCY-folate metabolism ...... 33 Mouse models with phenotypes associated with hyperhomocysteinemia in humans...... 34 MECHANISM OF HYPERHOMOCYSTEINEMIA IN DISEASE ...... 42 Neural tube defects...... 42 Vascular disease ...... 42 Cancer ...... 44 Neurodegenerative disorders...... 45 Down Syndrome ...... 46 HOMOCYSTEINE...... 47 Chemistry...... 47 Regulation of HCY metabolism...... 48 FOLATE...... 53 Chemistry...... 53 Regulation of folate metabolism...... 53 HOMOCYSTEINE AND FOLATE ...... 57 RESEARCH AIMS ...... 59 3
CHAPTER II GENETIC AND MOLECULAR CONTROL OF FOLATE-HOMOCYSTEINE METABOLISM IN MUTANT MICE ...... 63 ABSTRACT...... 64 INTRODUCTION ...... 65 RESULTS ...... 68 DISCUSSION ...... 79 MATERIAL and METHODS...... 87 CHAPTER III PARALLEL CHANGES IN METABOLITE LEVELS AND EXPRESSION PROFILES IN FOLATE-DEFICIENT AND CROOKED TAIL MUTANT MICE...96 ABSTRACT...... 97 INTRODUCTION ...... 98 RESULTS ...... 101 DISCUSSION ...... 113 MATERIAL AND METHODS ...... 115 CHAPTER IV NUTRIGENES, FUNCTIONAL GENOMICS AND SYSTEMS BIOLOGY ...... 119 ABSTRACT...... 120 INTRODUCTION ...... 121 RESULTS AND DISCUSSION ...... 122 CHAPTER V INHERITANCE OF HOMOCYSTEINEMIA IN A/J AND C57BL/6J MICE: DIET, GENDER, PARENTAL AND GENETIC EFFECTS...... 124 ABSTRACT...... 125 INTRODUCTION ...... 126 RESULTS ...... 133 DISCUSSION ...... 148 MATERIAL AND METHODS ...... 152 CHAPTER VI QTL ANALYSIS OF HOMOCYSTEINEMIA IN A/J AND C57BL/6J MICE USING C57BL/6J-Chr(i)A/J CHROMOSOME SUBSTITUTION STRAINS...... 157 ABSTRACT...... 158 INTRODUCTION ...... 159 RESULTS ...... 162 DISCUSSION ...... 179 MATERIAL AND METHODS ...... 184 CHAPTER VII SUMMARY AND FUTURE DIRECTIONS...... 190 APPENDIX 1 ...... 198 APPENDIX 2 ...... 200 BIBLIOGRAPHY ...... 202 4
LIST OF TABLES 5
CHAPTER ONE
I-1. Nutrition correction of NTDs in seven mouse models………………………..37
BOX 1: Homocysteine metabolism……………………………………………….....50
BOX 2: Folate metabolism……………………………………………………………54
CHAPTER TWO
II-1. List of genes used on the arrays……………………………………………….75
II-2. Summary of RNA abundance results………………………………………….77
II-3. Significant differences in RNA abundance in mutants ……………………...86
CHAPTER THREE
III-1. Effect of folic acid treatment on craniofacial and neural tube defects in Ski homozygous mutant mice…………………………………………………………...112
CHAPTER FIVE
V-1. Gender and diet effects on mean homocysteine levels……………………138
V-2. Gender and diet effects on homocysteine variance………………………..139
V-3. Variation in mean homocysteine levels and in variance for females that were raised on the ‘lower folate’ diet……………………………………………….144
V-4. Variation in mean HCY levels and in variance for males that were raised on the ‘lower folate’ diet……………………………………………….145 6
CHAPTER FIVE (continued)
V-5. Variation in mean HCY levels and in variance for females that were raised on the ‘higher folate’ diet……………………………………………...146
V-6. Variation in mean HCY levels and in variance for males that were raised on the ‘higher folate’ diet……………………………………………...147
V-7. Heritability of homocysteinemia……………………………………………….147
CHAPTER SIX
V1-1. Advantages and limitations of CSS versus F2 intercross mapping……...160
APPENDIX 1
Nutrients composition of Harland Teklad 7013 (‘lower’ folate) and Lab Diet 5010
(‘higher’ folate) diets………………………………………………………………….198
APPENDIX 2
List of missense mutations that differ between A/J and B6 strains on Chr 17……………………………………………………………………………...200 7
LIST OF FIGURES 8
CHAPTER ONE
I-1. Homocysteine and folate metabolism…………………………………………..18
I-2. Different forms of homocysteine…………………………………………………49
I-3. Chemical structure of folate……………………………………………………...55
CHAPTER TWO
II-1. Homocysteine levels for mutants congenic on the B6 background…………71
II-2. Homocysteine levels for mutants on different backgrounds…………………72
CHAPTER THREE
III-1. Serum homocysteine levels of Cd female mice……………………………102
III-2. Serum folate levels of Cd female mice………………………………………103
III-3. Hepatic folate levels of Cd female mice……………………………………..104
III-4. Serum glutathione levels of Cd female mice………………………………..105
III-5. Serum cysteine levels of Cd female mice…………………………………...107
III-6. Hierarchical clustering of metabolite and gene expression profiles of Cd mutant mice…………………………………………………………………………...108
III-7. Hierarchical clustering of metabolite and gene expression profiles of Cd and other previously analyzed mutant mice……………………………………………111
CHAPTER FIVE
V-1. Homocysteine levels among 14 different inbred strains……………………129 9
CHAPTER FIVE (continued)
V-2. Homocysteine levels in female mice that were raised on the ‘lower folate’ diet…………………………………………………………………140
V-3. Homocysteine levels in male mice that were raised on the ‘lower folate’ diet…………………………………………………………………141
V-4. Homocysteine levels in female mice that were raised on the ‘higher folate’ diet………………………………………………………………..142
V-5. Homocysteine levels in male mice that were raised on the ‘higher folate’ diet………………………………………………………………..143
CHAPTER SIX
VI-1. Engineering chromosome substitution strains……………………………...161
VI-2A: Mean homocysteine levels of CSS females that were raised on the ‘lower folate’ diet……………………………………………………………..164
VI-2B. Variance analysis of homocysteine of CSS females that were raised on the ‘lower folate’ diet……………………………………………………………..165
VI-3A: Mean homocysteine levels of CSS males that were raised on the ‘lower folate’ diet……………………………………………………………..166
VI-3B. Variance analysis of homocysteine levels of CSS males that were raised on the ‘lower folate’ diet……………………………………………………………..167
VI-4A: Mean homocysteine levels of CSS males that were raised on the ‘higher folate’ diet…………………………………………………………….169 10
CHAPTER SIX (continued)
VI-4B. Variance analysis of homocysteine levels of CSS males that were raised on the ‘higher folate’ diet…………………………………………………………….170
VI-5. Mean homocysteine levels of parental strains and F1 hybrid females…..172
VI-6. Congenic strains derived from B6-Chr17A/J to map the trait mean of normal homocysteine levels………………………………………………………………….173
VI-7. Fine-mapping of the trait mean of normal homocysteine levels
in females…………………………………………………………………………….175
VI-8. Candidate region on chromosome 17 that reduces mean homocysteine levels of B6 female mice…………………………………………………………….176
VI-9. Variance analysis of homocysteine levels of parental strains and F1 hybrid females………………………………………………………………………………..178
VI-10. Congenic strains derived from B6-Chr17A/J to map the trait variance of normal homocysteine levels………………………………………………………...180
VI-11. Fine-mapping of the trait variance of normal homocysteine levels……..181 11
ACKNOWLEDGEMENTS
I learned that a support system is a key component to navigate through your graduate career. I am thanking friends, mentor and family who helped me through my doctoral studies.
I would like to thank Jodi Jackson and Martha Susiarjo for brightening up my mornings, Gail Franklin for our great conversations at tea time, Michelle and
Malana Bey and my Caribbean brother, Sheldon Joseph aka ‘Trinidaddy’.
My sickle cell has been a great challenge and I was fortunate to be surrounded with generous friends who took care of me through the crises. I would like to thank Petrice Brown, Lesil Brihn, Toshi Kitami, Josephine Lam,
Kirsten Youngren (my favorite bay mate!), Pat Hunt and Terri Hassold who welcomed me into their home.
I would like to thank my mentor, Joe Nadeau, who provided valuable guidance and made me realize my potentials. In addition, I would like to thank his wife, Diane, and they both cared for me in this country far from home.
Most importantly, I would like to thank my parents, whom I most privileged to have, for always putting my needs ahead of theirs and for their continued support. 12
Genetic and Molecular Dissection of Homocysteinemia in Mice
Abstract
by
SHEILA ERNEST
Hyperhomocysteinemia (significantly elevated levels of blood homocysteine, a sulfur-containing amino acid) occurs at a frequency of 5% in the general population and adversely affects fundamental aspects of fetal development, adulthood and aging. However, the role of elevated homocysteine
(HCY) in these birth defects and adult diseases remains unclear. Several genetic and environmental factors affect HCY levels, making homocysteinemia a complex trait and its genetic and molecular control difficult to dissect in humans.
The laboratory mouse is an important model to investigate control of homocysteinemia because HCY levels in humans and mice are influenced by similar genetic and physiological factors. Single gene mutants, multigenic models and dietary perturbations were used to investigate the genetic and molecular control of homocysteinemia in disease and in health. Analysis of genetic and physiological perturbations of HCY metabolism led to identification of new single gene mouse models of homocysteinemia. The studies provided clues to new genes, pathways and functions that adversely affect homocysteinemia and 13 perhaps contribute to the pathogenesis of birth defects and adult diseases.
Characterization of the genetic control of normal HCY levels in multigenic mouse models revealed complex interplay among genes, gender, diet and parental effects on homocysteinemia. The laboratory mouse aided in simplifying the complexity of the homocysteinemia trait and provided important clues and new approaches for the understanding of human homocysteinemia. 14
1 4
CHAPTER I
INTRODUCTION AND RESEARCH AIMS 15
Various birth defects and adult diseases occur more frequently when blood levels of homocysteine (HCY), a sulfur-containing amino acid, exceed 12
µmol/L (KANG et al. 1992). Hyperhomocysteinemia (significantly elevated homocysteine levels) occurs at a frequency of 5% in the general population
(HANKEY and EIKELBOOM 2000). The variety of birth defects and adult diseases associated with elevated HCY levels is striking, but it remains unclear how 1 elevated levels of this single amino acid and its metabolites could adversely 5 affect so many different organs, tissues, and physiological processes. This chapter describes homocysteinemia and the factors that influence it, impact on human diseases, lessons from animal models, possible mechanisms in causing disease, and regulation.
HOMOCYSTEINEMIA
1) Definitions of homocysteinemia
In humans, normal HCY blood levels range from 5 to 12 µmol/L (PIETRZIK and
BRONSTRUP 1997) and hyperhomocysteinemia is subclassified into three categories, marginal (or mild; 12-30 µmol/L), moderate (30-100 µmol/L), and severe (>100 µmol/L) (PIETRZIK and BRONSTRUP 1997). In this thesis, homocysteinemia will be defined as HCY levels in the blood, 16
hyperhomocysteinemia, as significantly elevated HCY levels and normal
HCY homocysteinemia, as normal levels.
2) Environmental determinants of homocysteinemia
Several environmental factors affect homocysteinemia, such as: 1 6
a) Age and gender. In early childhood, HCY levels are similar between boys and girls (ranges 3 to 8 µmol/L; (BEILBY and ROSSI 2000)), but from puberty to old age, gender differences are recognizable and HCY levels increase (~3 to 5
µmol/L) with age in both genders (TONSTAD et al. 1996; TONSTAD et al. 1997). In adults, men usually have higher HCY levels than women and the average difference is 1-2 µmol/L (NYGARD et al. 1995). Hormonal effects account for these gender differences, as higher estrogen levels are associated with decreased
HCY levels (ZMUDA et al. 1997; GIRI et al. 1998; MORRIS et al. 2000).
b) Vitamin deficiencies. Deficiencies in B vitamin cofactors, such as folate,
vitamin B6 and B12, usually result in increased HCY levels (MINER et al. 1997).
Vitamin deficiencies are often caused by disorders in malabsorption (LINDENBAUM
1979; ROSENBLATT and FENTON 2001), malnutrition (ROSENBLATT and FENTON
2001) or loss of nutrients during food processing (SCHNEEDE et al. 2000). 17
c) Drug therapy. Medications influence HCY levels, often by affecting
absorption or metabolism of folate, vitamin B6 or B12 (UELAND and REFSUM 1989;
REFSUM and UELAND 1990). For instance, HCY levels increase during therapy with anticancer or antiepilectic drugs (antagonists of folate metabolism; (REFSUM et al. 1986; REFSUM et al. 1991)) or the anesthetic gas, nitrous oxide (an
antagonist of vitamin B12; (CHRISTENSEN et al. 1994; GUTTORMSEN et al. 1994)). 1 Also, cholesterol-lowering drugs, such as fibrates, increase HCY levels (DIERKES 7 et al. 1999).
d) Lifestyle. Smoking (BERGMARK et al. 1993; NYGARD et al. 1995), high consumption (more than 6 cups per day) of caffeinated coffee (NYGARD et al.
1997), chronic and high consumption of alcohol (CRAVO et al. 1996), and lack of physical activity (NYGARD et al. 1995; NYGARD et al. 1997) significantly increase
HCY levels. Smoking and caffeine probably increase HCY levels by interfering
with vitamin B6 synthesis (NYGARD et al. 1995) and alcohol by interfering with the activity of methionine synthase, an enzyme of HCY metabolism (MTR; Figure I-1,
{step 2} and see “HOMOCYSTEINE” section below; (KENYON et al. 1998)). 18 {steps 1 to 8}. Homocysteine is metabolized through metabolism
Homocysteine
and folate metabolism
and 6 . 12 Figure I-1. Homocysteine the remethylation or the trans- sulfuration pathways. Homocysteine metabolism is linked to glutathione metabolism {steps 20 to 23} through cysteine. Folate metabolism {steps 9 to 19}. Folate and homocysteine pathways are interrelated and are linked by 5,10-methylene- THF. Labels: vitamins B B 19
3) Genetic determinants of homocysteinemia
Several genetic factors are known to affect homocysteinemia. The ones that have been identified mostly involve mutations of genes involved in HCY and folate metabolism (Figure I-1 and see “HOMOCYSTEINE” and “FOLATE” sections below) and result in deficiencies of the following:
a) Cystathionine b-synthase (CBS; Figure I-1, {step 7}). CBS is the first step in the HCY trans-sulfuration and it irreversibly catabolizes HCY into cystathionine. CBS deficiency is inherited as an autosomal recessive trait (MUDD et al. 2001). Heterozygosity and homozygosity for CBS mutations lead to moderate and severe hyperhomocysteinemia, respectively (MUDD et al. 1964;
BOERS et al. 1985b). Heterozygous CBS mutations are present at a frequency of
0.5-1.5% in the general population (BOERS et al. 1985a) and more than 92 distinct mutations have been identified (KRAUS et al. 1999). CBS deficiency causes hyperhomocysteinemia because HCY is not efficiently removed from the cycle, accumulates intracellularly, and is then exported in the plasma (MUDD et al. 2001). In addition to increasing HCY levels, CBS deficiency increases the risk for vascular disease (MUDD et al. 1985; WILCKEN and WILCKEN 1998) and may cause dislocation of the optic lens, osteoporosis, thinning and lengthening of the long bones, and mental retardation (MUDD et al. 2001). 20
b) Methylenetetrahydrofolate reductase (MTHFR; Figure I-1, {step 1}).
MTHFR reduces 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate, which is the major folate species involved in methylation reactions of HCY metabolism (ROSENBLATT and FENTON 2001). MTHFR deficiency is inherited as an autosomal recessive trait (ROSENBLATT and FENTON 2001). Homozygosity and heterozygosity for MTHFR deficiency lead to mild hyperhomocysteinemia
(FROSST et al. 1995; KLUIJTMANS et al. 1996). Homozygosity and heterozygosity for a common MTHFR mutation (C677T) are present at frequency of ~10% and
~43%, respectively in the general population (BRATTSTROM et al. 1998; WALD et al. 2002). This variant renders the enzyme thermolabile by decreasing its activity by as much as 50% at 55oC. MTHFR deficiency causes hyperhomocysteinemia by reducing the synthesis of methylfolates or 5-methyltetrahydrofolate (FROSST et al. 1995; ROSENBLATT and FENTON 2001). In addition to increasing HCY levels,
MTHFR deficiency may cause developmental delay, motor and gait abnormalities or seizures (ROSENBLATT and FENTON 2001). Moreover, the common MTHFR variant increases the risk for neural tube defects (STEEGERS-THEUNISSEN et al.
1995; WHITEHEAD et al. 1995; CHRISTENSEN et al. 1999; BOTTO and YANG 2000), cardiovascular disease (WALD et al. 2002), perhaps Down syndrome (JAMES et al. 1999; HOBBS et al. 2000), and paradoxically provides protection from colon cancer (CHEN et al. 1996; MA et al. 1997). 21
c) Methionine synthase (MTR; Figure I-1, {step 2}). MTR deficiency is inherited as an autosomal recessive trait (WATKINS and ROSENBLATT 1989;
ROSENBLATT and FENTON 2001). Homozygosity for MTR deficiency lead to severe hyperhomocysteinemia (WATKINS and ROSENBLATT 1989). MTR deficiency reduces the synthesis of tetrahydrofolate and consequently causes hyperhomocysteinemia (ROSENBLATT and FENTON 2001). Homozygosity for a common MTR mutation (A2756G) is present at frequency of ~15-20% in the general population (LECLERC et al. 1996; CHEN et al. 1997). However, whether this common variant causes hyperhomocysteinemia is unclear (JACQUES et al.
2003) (HARMON et al. 1999; TSAI et al. 2000; DEKOU et al. 2001). In addition to increasing HCY levels, MTR deficiency may cause developmental delay and megaloblastic anemia (WATKINS and ROSENBLATT 1989; ROSENBLATT and FENTON
2001); it may also increase the risk for NTDs (MORRISON et al. 1997; VAN DER PUT et al. 1997; LUCOCK 2000; DOOLIN et al. 2002) and Down syndrome (BOSCO et al.
2003).
d) Methionine synthase reductase (MTRR). MTRR is an activator of MTR
(ROSENBLATT and FENTON 2001). MTRR deficiency is inherited as an autosomal recessive trait (WATKINS and ROSENBLATT 1989; ROSENBLATT and FENTON 2001).
Homozygosity for MTRR deficiency leads to severe hyperhomocysteinemia
(WATKINS and ROSENBLATT 1989). Heterozygosity for a common MTRR mutation
(A66G) is present at frequency of ~50% in the general population (HOBBS et al. 22
2000; O'LEARY et al. 2002). However, this common variant does not cause hyperhomocysteinemia (WILSON et al. 1999; BROWN et al. 2000; JACQUES et al.
2003). In addition to increasing HCY levels, MTRR deficiency may cause developmental delay and megaloblastic anemia (WATKINS and ROSENBLATT 1989;
ROSENBLATT and FENTON 2001), and increase the risk for NTDs (DOOLIN et al.
2002) and Down Syndrome (HOBBS et al. 2000; O'LEARY et al. 2002; BOSCO et al.
2003).
e) Glutamate carboxypeptidase II (GCPII). GCPII hydrolyzes dietary folates for absorption. Heterozygosity for GCPII mutations leads to mild hyperhomocysteinemia. Heterozygosity for a GCPII mutation (C1561T) is present at a frequency of ~8% in a Caucasian population. This deficiency alters absorption of folates, results in low levels of folate and consequently increases
HCY levels (DEVLIN et al. 2000).
HOMOCYSTEINEMIA AND HUMAN DISEASE
Hyperhomocysteinemia is an independent risk factor for neural tube defects (GROUP 1991; STEEGERS-THEUNISSEN et al. 1991; STEEGERS-THEUNISSEN et al. 1994; WENSTROM et al. 2000a; WENSTROM et al. 2000b), vascular disease 23
(MCCULLY 1969; BOUSHEY et al. 1995; MAYER et al. 1996; WELCH and LOSCALZO
1998; WALD et al. 2002) and Alzheimer disease (SESHADRI et al. 2002). It also has been suggested to be associated with certain cancers, such as colon (KATO et al. 1999) breast, ovarian and pancreatic carcinoma (CORONA et al. 1997; WU and WU 2002), Parkinson disease (KUHN et al. 1998a; KUHN et al. 1998b; YASUI et al. 2000; BLANDINI et al. 2001), and Down syndrome (JAMES et al. 1999; BOSCO et al. 2003).
1) Neural Tube Defects
Elevated HCY levels in the mother during pregnancy increases the risk of neural tube defects (NTDs; Wenstrom, 2000 #472;Wenstrom, 2000
#473;Steegers-Theunissen, 1991 #187;Steegers-Theunissen, 1994 #112]).
During normal pregnancy, HCY levels in the mother usually decrease and are lowest in the second trimester (WALKER et al. 1999). However, in NTD-affected pregnancies, HCY levels are often increased in the plasma of mothers (ESKES
1998), in the amniotic fluid (STEEGERS-THEUNISSEN et al. 1995; WENSTROM et al.
2000a; WENSTROM et al. 2000b), and perhaps in plasma of affected children
(ESKES 2001). Moreover, mothers heterozygous for CBS deficiencies (MUDD et al. 2001) or with low levels of plasma folate (GEORGE et al. 2002) may show increased risk of pregnancy losses. It is unknown whether increased HCY levels are a cause or consequence of NTDs. 24
a) Genetic determinants
Genetic factors, more precisely mutations of genes involved in the HCY- folate metabolism, affect the risk for NTDs. Mothers with the following mutations may show increased risk to give birth to a child with NTDs:
i) MTHFR, homozygosity for the C677T variant (VAN DER PUT et al.
1995; WHITEHEAD et al. 1995; CHRISTENSEN et al. 1999; BOTTO and YANG
2000). Maternal gene-nutrient interactions also influence the NTD risk:
mothers with low folate status combined with MTHFR mutation are at a
greater risk than either variable alone (MOLLOY and SCOTT 1997;
CHRISTENSEN et al. 1999).
ii) MTR, homozygosity and heterozygosity for the A2756G variant
(MORRISON et al. 1997; VAN DER PUT et al. 1997; LUCOCK 2000; DOOLIN et al.
2002). NTD risk increases with the number of mutant alleles (DOOLIN et al.
2002).
iii) MTRR, homozygosity and heterozygosity for the A66G variant. NTD
risk increases with the number of mutant alleles (DOOLIN et al. 2002).
iv) MTHFD1, homozygosity for the R653Q variant (Figure I-1, {steps 11,
12 and 13}). This variant does not result in hyperhomocysteinemia or folate
deficiency (BRODY et al. 2002). MTHFD1 (methylenetetrahydrofolate
dehydrogenase/ methenyltetraydrofolate cyclohydrolase/ 10-
formyltetrahydrofolate synthetase) is a trifunctional enzyme that provides 25
single carbon units for purine and pyrimidine synthesis (WEST et al. 1996;
ROSENBLATT and FENTON 2001).
v) RFC-1, homozygosity for the A80G variant. NTD risk is increased
only when this variant is combined with low folate status. Also, this variant
does not lead to significantly increased HCY levels or decreased folate levels
(MORIN et al. 2003). RFC-1 (reduced folate carrier 1) mediates the transport of
folate into the cell (SIROTNAK and TOLNER 1999).
b) Treatment
It is now widely accepted that folic acid taken before conception and during pregnancy reduces the NTD risk. Evidence was first presented in 1965
(HIBBARD et al. 1965) and numerous trials (GROUP 1991; CZEIZEL and DUDAS
1992; BERRY et al. 1999) have confirmed that recurrence of NTDs can be reduced by 50-70%. Folic acid fortification of the US food supply (e.g. grains and flour) is associated with a 19% reduction in the occurrence of NTDs (HONEIN et al. 2001). In other instances, methionine has also been associated with reduced risk for NTDs (SHAW et al. 1997). The mechanism for the protective effect of folic acid or methionine is unknown. 26
2) Vascular disease
Elevated plasma HCY levels were first linked to vascular disease in patients with deficiencies in CBS gene (Figure I-1 {step 7};(MCCULLY 1969)). Subsequent studies found associations between hyperhomocysteinemia and cardiovascular, peripheral vascular and cerebrovascular disease (COOK et al. 2002). In cardiovascular disease, hyperhomocysteinemia appears to act independently of other risk factors, such as cholesterol, smoking, age and high blood pressure
(MCCULLY 1969; BOUSHEY et al. 1995; MAYER et al. 1996; WELCH and LOSCALZO
1998; WALD et al. 2002) (REFSUM et al. 1998). The risk is increased when hyperhomocysteinemia is combined with the traditional risk factors, such as cholesterol, smoking and hypertension (GRAHAM et al. 1997). A 5 µmol/L HCY increment may elevate risks for coronary heart disease (BOUSHEY et al. 1995), ischaemic heart disease, deep vein thrombosis and stroke (WALD et al. 2002).
Meta-analyses show that hyperhomocysteinemia causes coronary heart disease
(KLERK et al. 2002), ischaemic heart disease, deep vein thrombosis and stroke
(WALD et al. 2002).
a) Genetic determinants
Both mild and severe hyperhomocysteinemia have been shown to be associated with vascular disease.
i) Deficiencies of the CBS protein lead to severe
hyperhomocysteinemia and have been associated with premature vascular 27
disease (BOERS et al. 1985b; MUDD et al. 1985; WILCKEN and WILCKEN 1998).
No association was detected between CBS mutations and coronary disease
(ROBINSON 2001), and interestingly, no CBS mutation was found in vascular
disease patients probably because of its rare frequency in the population
(ROBINSON 2001).
ii) The C677T mutation in MTHFR gene may lead to mild
hyperhomocysteinemia and is associated with cardiovascular disease (KLERK
et al. 2002; WALD et al. 2002). The homozygous TT mutation combined with
low folate status further increases the risk for coronary heart disease (KLERK
et al. 2002), ischaemic heart disease, deep vein thrombosis and stroke (WALD
et al. 2002).
b) Treatment
Several clinical trials are underway to test whether the HCY lowering
vitamin supplements, i.e. folic acid, vitamins B6 and B12, reduce the risk for vascular disease (CLARKE and ARMITAGE 2000). It was estimated that by HCY lowering vitamin cofactors in individuals aged 45 years and older, the potentially preventable deaths in the United States is ~10% for men and ~6% for women
(BOUSHEY et al. 1995). Furthermore, a 3 µmol/L HCY decrease, achievable by
0.8 mg of folic acid, would reduce the risk of ischaemic heart disease by ~16%, deep vein thrombosis by ~25% and stroke by ~24% (WALD et al. 2002).
Preliminary studies already showed that B vitamins attenuate thrombin 28 generation (UNDAS et al. 1999), improve vascular endothelial function in men with coronary heart disease and significantly lower the risk for ischaemic heart disease (SCHNYDER et al. 2001). The mechanism for the protective effect of B vitamins is unclear but the damaging effects of HCY may partly be reversed.
3) Cancer
Increased levels of serum HCY levels are often found in colon (KATO et al.
1999) breast, ovarian and pancreatic carcinoma (CORONA et al. 1997; WU and
WU 2002). HCY concentrations parallel the profile of tumor markers in serum of carcinoma patients (WU and WU 2002). And similarly to those markers, HCY levels reflect the rapid proliferation rate of tumor cells, suggesting that HCY itself is a tumor marker (WU and WU 2002). Additionally, folate deficiency is a risk factor for colon (GIOVANNUCCI et al. 1993; GIOVANNUCCI et al. 1995; CHEN et al.
1996; MA et al. 1997; CHEN et al. 1998) and breast cancers (GRAHAM et al. 1991;
FREUDENHEIM et al. 1996). In situations of both low folate status and high alcohol consumption (≥ 15 g/day), colon cancer risk is further increased (GIOVANNUCCI et al. 1993; GIOVANNUCCI et al. 1995) whereas the breast cancer risk is observed only in these conditions (ZHANG et al. 1999; ROHAN et al. 2000). Because increased HCY levels are often associated with folate deficiency, hyperhomocysteinemia could also be risk factors for breast cancer. 29
a) Genetic determinants
Interestingly, the homozygous C677T MTHFR mutation protects from colorectal cancer, however only with low consumption of alcohol and high dietary folate or methione (CHEN et al. 1996; MA et al. 1997; CHEN et al. 1998). A possible mechanism for this protection is that the mutated MTHFR spares 5,10- methylenetetrahydrofolate (Figure I-1 {step 1}) and consequently leads to a decreased rate of uracil missincorporation, which contributes to strand breaks, during DNA synthesis (CHEN et al. 1996; MA et al. 1997; CHEN et al. 1998;
FENECH 1998). Maintaining low levels of alcohol is essential because alcohol causes folate deficiency by disrupting absorption and accelerating breakdown of folate (HILLMAN and STEINBERG 1982).
b) Treatment
Supplementation with B vitamins can reduce the colon cancer risk from
~25% to as much as ~60%-75% (GIOVANNUCCI et al. 1995; GIOVANNUCCI et al.
1998; SU and ARAB 2001) and the breast cancer risk by ~25%, especially when consumption of alcohol is relatively high (ZHANG et al. 1999). B vitamin supplementation may protect from cancer by reversing the effects of oxidative stress, abnormal methylation and nucleotide imbalance cause by folate deficiency or increased HCY levels (see “MECHANISMS OF
HYPERHOMOCYSTEINEMIA IN DISEASE” section below). For instance, 30 supplementation with folic acid reverses DNA hypomethylation in individuals with colorectal cancer (CRAVO et al. 1994).
4) Neurodegenerative disorders
Hyperhomocysteinemia has recently been shown to be an independent risk factor for Alzheimer disease (AD) (SESHADRI et al. 2002). The levels of HCY may determine the rate of progression of AD (CLARKE et al. 1998) and hyperhomocysteinemia precedes the onset of AD (JOOSTEN et al. 1997; SESHADRI et al. 2002). A 5 µmol/L HCY increment may elevate the risk for AD by ~40%
(SESHADRI et al. 2002). Hyperhomocysteinemia is also found in patients with
Parkinson disease (PD (KUHN et al. 1998a; KUHN et al. 1998b)) but is secondary to treatment with the antiparkinsonian drug, levodopa (MULLER et al. 1999).
Levodopa acts as a potent methyl acceptor from S-adenosylmethionine and the resulting S-adenosylhomocysteine is rapidly metabolized to HCY (Figure I-1
{step 5}; (BOTTIGLIERI et al. 1994; KUHN et al. 1998a).
Genetic determinants
Hyperhomocysteinemia is more severe in PD patients with the homozygous
MTHFR C677T mutation than in patients without the mutation (YASUI et al. 2000;
KUHN et al. 2001). These observations suggest that levodopa treatment 31 combined with MTHFR mutation could accelerate the neurodegenerative process in PD patients (YASUI et al. 2000; KUHN et al. 2001).
5) Down Syndrome
Elevated HCY levels during pregnancy may increase the risk of having a child with Down syndrome (DS) or trisomy 21 (JAMES et al. 1999; BOSCO et al. 2003).
Genetic determinants
The following polymorphisms may increase the risk of having a child with DS:
i) MTHFR, homozygosity and heterozygosity for the C677T variant
(JAMES et al. 1999; HOBBS et al. 2000).
ii) MTR, homozygosity and heterozygosity for the A2756G variant
(BOSCO et al. 2003).
iii) MTRR, homozygosity and heterozygosity for the A66G variant
(HOBBS et al. 2000; O'LEARY et al. 2002; BOSCO et al. 2003).
Moreover, certain combinations of gene polymorphisms may further increase the risk: homozygosity for MTRR 66GG combined with homozygosity or heterozygosity for MTHFR C677T (HOBBS et al. 2000; O'LEARY et al. 2002); heterozygosity for MTRR 66AG combined with heterozygosity for MTR 2756AG
(BOSCO et al. 2003). 32
HOMOCYSTEINEMIA AND ANIMAL MODELS
Little is known about the range of normal HCY levels in animal models.
However, a survey, done in our laboratory, of fourteen different inbred strains of mice with no obvious phenotypes, showed that HCY levels ranged from 3.1 to
7.0 mmol/L (see Chapter Five, Figure V-1). These levels could represent reference levels of normal HCY levels in mice.
1) Environmental determinants of homocysteinemia
It is likely that HCY levels in animal models are influenced by similar factors affecting homocysteinemia in humans. For instance, rats fed diets deficient in
folate (MILLER et al. 1994b; HUANG et al. 2001; EBBESEN et al. 2003) or vitamin B6, as well as mice treated with the antiepilectic drug, valproate (HISHIDA and NAU
1998) (SMOLIN and BENEVENGA 1982; MILLER et al. 1992; MILLER et al. 1994a), show increased HCY levels,. Little study has evaluated the effect of age and gender on HCY level in animals. In Chapter Five of this thesis, HCY levels were evaluated in two inbred strains of mice, A/J and C57BL/6J, and, in contrast to findings in humans, levels were higher in females than in males. Therefore, estrogen may have contrasting effects on HCY levels in humans and mice. 33
2) Mouse models with deficiencies of genes involved in HCY-folate metabolism
Targeted disruptions in the following HCY-folate genes caused hyperhomocysteinemia in mice:
a) Methylenetetrahydrofolate reductase (MTHFR; Figure I-1 {step 1}). HCY levels are increased ~10-fold and ~1.6-fold respectively in homozygous and heterozygous Mthfr mutant mice compared to levels in wild-type controls.
MTHFR-deficient homozygotes show growth retardation and laminar structure defects in the cerebellum. In addition, significant lipid deposition in the aorta was observed in both heterozygous and homozygous mutants (CHEN et al. 2001).
These phenotypes are similar to those associated with MTHFR deficiency in humans (ROSENBLATT and FENTON 2001).
b) Methionine synthase (MTR; Figure I-1 {step 2}). Heterozygous Mtr mutant mice have a ~1.7-fold increase in HCY levels compared to their wild-type controls, whereas homozygous Mtr mutant mice are embryonic lethal. Otherwise, wild-type controls and MTR-deficient heterozygotes are phenotypically indistinguishable, despite a 40-50% decrease in MTR activity in the mutants
(SWANSON et al. 2001).
c) Cystathionine b-synthase (CBS; Figure I-1 {step 7}). Compared to their wild-type controls, homozygous and heterozygous CBS-deficient mutant mice 34 have a ~40-fold and ~2-fold increase in HCY levels respectively. Homozygous mutants suffer from severe growth retardation and the majority die within 5 weeks after birth (WATANABE et al. 1995). In contrast to humans, Cbs mutant mice do not show atherosclerotic lesions (WATANABE et al. 1995). However, the endothelium function (i.e. the ability of the endothelium to relax in response to a vasodilator) is disrupted in Cbs heterozygotes due to hyperhomocysteinemia
(EBERHARDT et al. 2000; DAYAL et al. 2001; WEISS et al. 2002). Endothelium dysfunction plays an important role in the pathogenesis of atherosclerosis (CHEN et al. 2000). Moreover, consistent with CBS deficiency in humans (MUDD et al.
2001), livers of homozygous Cbs mutants are enlarged and contain lipid droplets
(WATANABE et al. 1995).
3) Mouse models with phenotypes associated with hyperhomocysteinemia in humans
Several mouse models of human diseases associated with hyperhomocysteinemia have been studied and this section summarizes the data for each mouse model.
a) Neural tube defects
We found increased HCY levels in certain mouse models of NTDs, i.e.
Pax3, Gli3 and Apob heterozygous female mutants, compared to their wild-type 35 controls (see Chapter Two, Figure II-1; (ERNEST et al. 2002)). Heterozygotes were tested because homozygotes for these mutations show anencephaly or spina bifida and die during embryogenesis (Mouse Locus Catalog; www.informatics.jax.org). And female mice were tested because in humans maternal hyperhomocysteinemia is a risk factor for fetal neural tube defects
(MALINOW et al. 1998), although anomalies in fetal metabolism also appear to be important (FLEMING and COPP 1998). Folate also plays an essential role in normal development of the early mouse embryo for DNA (O'NEILL 1998) and its deficiency leads to increased fetal deaths, decreased fetal weight and delayed craniofacial and heart development (BURGOON et al. 2002).
In mouse models for NTDs, several compounds suppress NTDs (Table I-
1):
i) Folic acid reduces the severity of NTDs in Cart1 (cartilage
homeoprotein 1; (ZHAO et al. 1996)), Pax3Sp (pax3 transcription factor;
(FLEMING and COPP 1998)), Cd (crooked-tail; (CARTER et al. 1999)), Cited2
(CBP/p300 interacting transactivators with glutamic acid (E)/(D)-rich C-
terminal domain; (BARBERA et al. 2002)), Rfc1 (reduced folate carrier 1; (ZHAO
et al. 2001)) and Folbp1 (folic acid-binding protein 1; (PIEDRAHITA et al. 1999)
mice.
ii) Inositol, but not folate, suppresses NTDs in ct (curly-tail) mice
(GREENE and COPP 1997; TRAN et al. 2002). 36
iii) Methionine, but not folate, suppresses NTDs in Axd (axial defects)
mice (ESSIEN 1992).
The mechanism by which folic acid, methionine and inositol correct the birth defects in these mutant mice is unknown. It has been proposed that folic acid suppresses the defect in Pax3Sp by correcting disruptions in pyrimidine biosynthesis (FLEMING and COPP 1998), in Cited2 by compensating for cell death
(BARBERA et al. 2002), and inositol may suppress the defect in ct by enhancing cell proliferation through up-regulating the retinoic acid receptor b (GREENE and
COPP 1997). Few mouse models have been analyzed for defect correction in response to nutrients (~8 of out a total of ~65 models for NTDs (JURILOFF and
HARRIS 2000; ZHAO et al. 2001; BARBERA et al. 2002)) and a pattern may emerge with the characterization of several mutants with similar type of defects, for example mice with spina bifida versus mice with excencephaly. 37
Table I-1. Nutrition correction of NTDs in seven mouse models. The type of NTDs, the nutrient delivery, its amount and time vary by experiment. Mutant NTD Nutritional Delivery Amount Time of NTD supplement method delivered nutrient correction delivery 1Axd Spina bifida Methionine IP injection 180 mg/kg bw dpc 8 and 9 Yes Folinic acid IP injection 33 mg/kg bw dpc 8, 9 and 10 No Vitamin B12 IP injection 330 mg/kg bw dpc 8 and 9 No 2Cart1 Acrania, Folic acid IP injection 2.5-3 mg/kg E0.5-E9.5 Yes anencephaly bw 3Cited2 Exencephaly Folic acid IP injection 3 mg/kg bw E0.5-gestation Yes
Emryos in 50 mM dpc 8.5 Yes culture 4Cd Exencephaly Folic acid Chow 4,7 and 10 Before Yes mg/kg chow conception- gestation 5ct Spina bifida aMethionine IP injection 200, 400, 800 dpc 9 No and 1600 mg/kg bw Water ad 10 g/L Before No libitum conception- gestation bInositol IP injection 4 and 2000 E9.5 and/or Yes mg/kg bw E10.5 cFolic acid Chow 100 mg/kg Before No chow conception- gestation 6Folbp1 Spina bifida Folinic acid Oral 25 mg/kg bw Before Yes intubation conception- gestation 7Sp2H Spina bifida, Folic acid IP injection 10 mg/kg bw E8.5; 9.5 Yes anencephaly Emryos in 200 ug/ml E8.5-E10.5 Yes culture Thymidine IP injection 10 mg/kg bw E8.5; E9.5 Yes Emryos in 250 ug/ml E8.5-E10.5 Yes culture Methionine Embyos in 1.5 mg/ml E8.5-E10.5 No culture Symbols: IP: intraperitoneal, bw: body weight, dpc: day post coitum 1Axial defects(ESSIEN 1992) 2Cartilage homeoprotein 1(ZHAO et al. 1996) 3CBP/p300 interacting transactivators with glutamic acid (E)/(D)-rich C-terminal domain(BARBERA et al. 2002) 4Crooked-tail(CARTER et al. 1999) 5curly-taila(VAN STRAATEN et al. 1995),b(GREENE and COPP 1997),c(TRAN et al. 2002) 6Folic acid-binding protein 1(PIEDRAHITA et al. 1999) 7PAX3 transcription factor(FLEMING and COPP 1998) 38
b) Vascular disease
Diets enriched with methionine or low levels of B vitamins induce hyperhomocysteinemia, and cause atherosclerosis in rats (MATTHIAS et al. 1996;
MORITA et al. 2001) and accelerate atherosclerosis in apoE-null mouse, a model of hypercholesterolemia and atherosclerosis (HOFMANN et al. 2001; ZHOU et al.
2001; TROEN et al. 2003). A recent study showed that atherosclerotic lesions in
ApoE-null mice were caused by excess plasma methionine rather than plasma
HCY: a methionine-enriched diet leads to both hyperhomocysteinemia and lesions, whereas a B vitamin-deficient diet leads to hyperhomocysteinemia and no lesions (TROEN et al. 2003). This is in agreement with the observation of Cbs knockout mouse does not develop atherosclerosis (WATANABE et al. 1995).
However, Cbs mutant mice (EBERHARDT et al. 2000; DAYAL et al. 2001; WEISS et al. 2002), as well as cultured human and pig HCY-treated endothelial cells,
(CHEN et al. 2000) do show endothelial dysfunction . Therefore, the role of increased HCY levels on atherosclerosis remains to be determined.
B vitamin treatment ameliorates atherosclerosis and endothelial dysfunction in hyperhomocysteinemic mice. Atherosclerotic lesions, as well as
HCY levels, are significantly reduced in apoE-null mice fed a diet-supplemented
with folic acid or vitamins B6 and B12 (HOFMANN et al. 2001; TROEN et al. 2003).
Also, the severity of endothelial dysfunction in Cbs heterozygous mice correlates with folate content in the diet (LENTZ et al. 2000). 39
c) Cancer
We found increased HCY levels in a colon cancer model, Apc heterozygous mutants, compared to their wild-type controls (see Chapter Two,
Figure II-1; (ERNEST et al. 2002)). Depending on the stage of tumorigenesis, dietary folate deficiency has divergent effects on tumor formation:
i) Dietary deficiencies in folate or methyl (methionine, choline or vitamin
B12) groups induce tumor formation in rats, primarily in the liver (MIKOL et al.
1983; GHOSHAL and FARBER 1984; LOCKER et al. 1986; HENNING et al. 1997a).
Moreover, folate deficiency exacerbates the chemically induced-colorectal
tumorigenesis in rats (CRAVO et al. 1992; KIM et al. 1996b).
ii) Conversely, folate deficiency reduces the number of chemically
induced-aberrant crypt foci, probably the precursors of colorectal cancer in
rodents and humans (PRETLOW et al. 1992; BIRD 1995; RONCUCCI et al. 1998;
LE LEU et al. 2000). In addition, folate deficiency reduces the number of small
intestine polyps, after the establishment of neoplastic foci, in mouse genetic
models of colon cancer, single Apc mutants (SONG et al. 2000a) and double
Apc +/- Msh2-/- mutants (SONG et al. 2000b). The inhibitory effect of folate
deficiency on tumor formation is in agreement with the role of folate
metabolism in DNA and purine synthesis and the use of antifolate drugs for
cancer therapy (e.g. methotrexate, which inhibits dihydrofolate reductase
(DHFR, Figure I-1 {step 19}), (ROSENBLATT and FENTON 2001)). Tissues with
active cell proliferation, such as colonic cells, are more vulnerable to folate 40
deficiency and, consequently, DNA synthesis is disrupted and tumor growth is
inhibited (LE LEU et al. 2000; SONG et al. 2000a; SONG et al. 2000b). For
example, growth of a transplanted cancer is inhibited in folate-deficient rats
(ROSEN and NICHOL 1962) and folate deprivation delays onset of leukemia in
Friend Virus-Infected mice (KOURY et al. 1997). These results suggest that, in
advanced tumorigenesis, folate deficiency has an inhibitory effect on tumor
progression or may even cause tumor regression (SONG et al. 2000a; SONG et
al. 2000b).
The effect of dietary folate supplementation on cancer incidence also depends on the stage of tumorigenesis at which folate is introduced:
i) When provided before the onset of colorectal cancer, dietary folate
supplementation suppresses the development of small intestine polyps in Apc
and Apc +/- Msh2-/- mutant mice.
ii) In contrast, dietary folate supplementation, provided after the onset
of the disease, promotes intestinal tumorigenesis in these mutant mice (SONG
et al. 2000a; SONG et al. 2000b). Furthermore, in folate-deficient rats, folate
supplementation enhances tumor development in the liver (CHANDAR and
LOMBARDI 1988) and colon (LE LEU et al. 2000). In humans, accelerated
progression of acute leukemia is observed in children treated with folate
supplementation (SONG et al. 2000a; SONG et al. 2000b).
In summary, dietary folate supplementation may provide protection against the initiation of carcinogenesis in normal cells, whereas folate deficiency 41 predisposes to cancer. By contrast, in advanced tumorigenesis or in preexisting folate-deficient cells, folate supplementation would promote the growth and progression of cancer lesions (MELNYK et al. 1999; LE LEU et al. 2000; SONG et al.
2000a; SONG et al. 2000b; JAMES et al. 2003; LAMPRECHT and LIPKIN 2003).
b) Neurodegenerative disorders
In Alzheimer disease (AD), the death of neurons is caused by abnormal processing of the amyloid precursor protein (APP) leading to increased production and accumulation of amyloid b-peptide (Ab) (YANKNER 1996; MATTSON
1997). Hyperhomocysteinemia, induced by diet enriched with HCY and devoid of folate, accelerates neuronal death and increases cellular DNA damage of the hippocampus in a mouse model of AD (KRUMAN et al. 2002), created by overexpressing APP with the “Swedish” mutation under the control of a prion promoter (BORCHELT et al. 1996). However, folate deficiency does not alter Ab levels in the brains of APP transgenic mice, suggesting that damaging effects of folate deficiency on neurons are independent of Ab deposition (KRUMAN et al.
2002).
Experimental models of sporadic Parkinson disease (PD) are created through systemic injection of the toxin 1-methyl-4-phenyl-1,2,3,6- tetrahydropyridine (MPTP) in mice (BEZARD et al. 1997). In folate-deficient mice with subsequent MPTP-treatment, direct injection of HCY into the substantia nigra of the brain exacerbates dopamine depletion, neuronal degeneration and 42 motor dysfunction (DUAN et al. 2002). These results suggest that folate deficiency and elevated HCY levels endanger dopaminergic neurons.
MECHANISM OF HYPERHOMOCYSTEINEMIA IN DISEASE
1) Neural tube defects
The mechanism by which hyperhomocysteinemia causes NTDs is unknown.
However, animal studies suggest that that HCY itself is teratogenic where injection of non-physiologic doses of HCY in avian embryos in culture induced
NTDs (ROSENQUIST et al. 1996), possibly through inhibition of N-methyl-D- aspartate (NMDA) receptors (ROSENQUIST et al. 1996).
2) Vascular disease
Mechanisms of action of HCY on vascular disease are unclear but two main hypotheses have been proposed:
a) Oxidative damage. During the intracellular or extracellular auto-oxidation of HCY, potent reactive oxygen species, such as superoxide and hydrogen 43 peroxide, are produced (DE VRIESE et al. 1998; WELCH and LOSCALZO 1998).
These reactive species expose the underlying matrix and smooth-muscle cells, which in turn proliferate and promote the activation of platelets (STAMLER and
SLIVKA 1996; BELLAMY and MCDOWELL 1997; DE VRIESE et al. 1998; TANG et al.
1998). Moreover, increased HCY levels is associated with increased levels of oxidized proteins and decreased levels of plasma antioxidants, suggesting that
HCY can alter the antioxidant capacity of the cell (UPCHURCH et al. 1997;
VENTURA et al. 2000).
b) Molecular target. Hyperhomocysteinemia partly contributes to enhanced vascular inflammation and hypercoagulability, factors implicated in the development of atherosclerosis (HOFMANN et al. 2001). For instance, increased
HCY levels are associated with increased aggregation of platelets (DURAND et al.
1997; EIKELBOOM et al. 1999) and activation of coagulation factors (RODGERS and
KANE 1986; RODGERS and CONN 1990; LENTZ and SADLER 1991). HCY can also interact with specific targets: it inhibits the binding of plasminogen to annexin and consequently synthesis of plasmin is reduced leading to blood vessel occlusion
(HAJJAR et al. 1998). 44
3) Cancer
Hypotheses have been proposed to explain the effect of increased HCY levels on cancer risk:
a) Oxidative damage. During the auto-oxidation of HCY, potent reactive oxygen species, such as superoxide and hydrogen peroxide, are produced (DE
VRIESE et al. 1998; WELCH and LOSCALZO 1998). These reactive species can lead to oxidative DNA damage (CROTEAU and BOHR 1997; STOLZENBERG-SOLOMON et al. 1999), which can accumulate and be detected in cancerous tissue (CROTEAU and BOHR 1997).
b) Abnormal DNA methylation. The trans-methylation pathway of HCY metabolism is involved in several methylation reactions of nucleic acids, proteins, lipids and small molecules (Figure I-1 {step 5}; (CLARKE and BANFIELD 2001).
Increased HCY levels or decreased folate levels could disrupt DNA methylation patterns, which are common in human tumors (LAIRD and JAENISCH 1994; LAIRD et al. 1995; FENECH et al. 1998), and both hypomethylation and hypermethylation have been shown to be important in carcinogenesis (LAIRD and JAENISCH 1994;
LAIRD et al. 1995; FENECH et al. 1998). For instance, low folate status and global
DNA hypomethylation are associated with human colorectal neoplasia (PUFULETE et al. 2003). Moreover, in a rat model of chemically induced-colon cancer, folate 45 deficiency leads to exon-specific DNA hypomethylation of the p53 tumor suppressor gene (KIM et al. 1996a).
c) Nucleotide imbalance. Hyperhomocysteinemia is often linked with disruptions in folate metabolism, which is involved in DNA and pyrimidine synthesis (Figure I-1, {step 18}). Folate deficiency leads to accumulation of deoxyuridine triphosphate (dUTP), and consequently uracil is incorporated in place of thymine during DNA synthesis (BLOUNT et al. 1997; DUTHIE and HAWDON
1998). Uracil misincorporation contributes to DNA strand breaks (CHEN et al.
1996; MA et al. 1997; CHEN et al. 1998; FENECH et al. 1998) and chromosome breaks are important in nearly all human cancers and are especially common in colorectal cancer (MA et al. 1997).
4) Neurodegenerative disorders
HCY-folate metabolism is present in the brain but may not be complete; for instance, enzymes such as cystathionine g-lyase (CTH, Figure I-1 {step 8}) may be absent (FINKELSTEIN 1990; MUDD et al. 2001). Moreover, folate transporters are expressed in neurons and glial cells suggesting an important role for HCY- folate metabolism in the nervous system (SIROTNAK and TOLNER 1999). 46
Mechanisms have been proposed to explain the effect of elevated HCY levels or folate deficiency on neurons:
a) Oxidative damage. In cultured neurons, increased HCY and folate deficiency both induce cytosolic Ca2+ influx (HO et al. 2002). In addition, HCY promotes glutamate neurotoxicity through NMDA receptors (KRUMAN et al. 2000;
KRUMAN et al. 2002). Increased Ca2+ and glutamate excitoxicity may lead to neuronal oxidative damage (KRUMAN et al. 2000; KRUMAN et al. 2002).
b) Neuronal cell death. Increased HCY and folate deficiency can kill and damage neurons in culture by inducing DNA damage through nucleotide imbalance, induced by uracil misincorporation (KRUMAN et al. 2000).
5) Down Syndrome
Down Syndrome results from an extra chromosome 21, which is caused by failure of normal chromosome segregation, primarily during maternal meiosis
(HASSOLD and SHERMAN 2000). Folate deficiency has been associated with altered chromosome recombination and segregation (HOBBS et al. 2000).
Because folate deficiency or increased HCY levels may induce abnormal DNA methylation, DNA hypomethylation may increase the risk of chromosome non- disjunction (JAMES et al. 1999). 47
HOMOCYSTEINE
HCY is formed from methionine (Figure I-1), an essential amino acid found in animal and plant proteins (FINKELSTEIN 1998). It contributes directly to methylation of DNA, proteins, lipids and small molecules (Figure I-1, {step 5};
(CLARKE and BANFIELD 2001)) and contributes indirectly to glucose metabolism
(Figure I-1, through conversion of cysteine {step 8} into pyruvate), citric acid cycle (Figure I-1, through a-ketoglutarate {step 8}; (MUDD et al. 2001)) and glutathione metabolism (Figure I-1, {steps 20 to 23}; (MOSHAROV et al. 2000)).
1) Chemistry
HCY is an amino acid whose sulfur group (Figure I-2) is highly reactive
(UELAND 1995; JACOBSEN 2001). Therefore, small amounts (<1%) of HCY itself or reduced HCY are detected in plasma (MANSOOR et al. 1992; UELAND 1995). Most
HCY exists in its oxidized form and comprises 98-99% of total HCY in the plasma
(MANSOOR et al. 1992; UELAND 1995). Both reduced and oxidized HCY are 70-
80% protein bound (Figure I-2), mainly to albumin through disulfide bonds with cysteine. The remaining 20-30% combines with other thiols (Figure I-2), including cysteine (5-15% of HCY-cysteine mixed disulfides) or with another HCY molecule
(5-15% of HCY-HCY molecules to form homocystine; (MANSOOR et al. 1992)).
Total serum or plasma HCY refers to the sum of all forms, free and bound, of 48
HCY (MANSOOR et al. 1992). During the auto-oxidation of HCY (HCY-HCY molecules) in presence of molecular oxygen, hydrogen peroxide and other reactive oxygen species are produced (DE VRIESE et al. 1998; WELCH and
LOSCALZO 1998). Therefore, HCY may cause disease via these reactive molecules (STAMLER and SLIVKA 1996; BELLAMY and MCDOWELL 1997; DE VRIESE et al. 1998; TANG et al. 1998).
2) Regulation of HCY metabolism
HCY has three possible fates: it can be remethylated through the trans- methylation pathway (Figure I-1, {steps 1 to 6}; (MUDD et al. 2001)), irreversibly catabolized through the trans-sulfuration pathway (Figure I-1, {steps 7 and 8};
(MUDD et al. 2001)), or exported to the extracellular space (STEAD et al. 2000).
Box 1 describes HCY metabolism in detail.
a) Trans-methylation and trans-sulfuration pathways
Liver is a key organ for HCY metabolism (STEAD et al. 2000). In mammalian liver, about half of HCY is remethylated and half is irreversibly catabolized through the trans-sulfuration pathway (FOWLER 1997). The trans-methylation pathway is present in all mammalian cells whereas the tissue distribution of the trans-sulfuration metabolism is limited. According to animal 49
Figure I-2. Different forms of homocysteine (from (BEILBY and ROSSI 2000)) HCY is an amino acid whose sulfur group is highly reactive. HCY can bind to proteins, to another HCY molecule or to other thiols such as cysteine. 50
BOX 1: Homocysteine Metabolism
Homocysteine (HCY) is metabolized through the trans-methylation or trans-sulfuration pathways. Trans-sulfuration of HCY is linked to glucose metabolism and glutathione metabolism {steps 20 to 23} through cysteine, citric acid cycle through a-ketobutyrate.
Trans-methylation pathway {steps 1 to 6} 1. 5,10-methylenetetrahydrofolate reductase {MTHFR, step 1} oxidizes 5,10- methylenetetrahydrofolate (5,10-methylene-THF) into 5-methyltetrahydrofolate (5-methyl-THF, a predominant form of folate in serum and many tissues). 2. HCY is converted into methionine by methionine synthase {MTR, step 2}, which uses vitamin B12 as a cofactor and 5-methyl-THF as methyl donor, or by betaine-homocysteine methyltransferase {BHMT, step 3} which uses betaine as methyl donors. 3. S-adenosylmethionine synthase {AMS, step 4} converts methionine into S- adenosylmethionine (SAM) in a magnesium and potassium dependant reaction. Methionine is the only metabolite of HCY that mammals are unable to synthesize therefore it must come from the diet. 4. SAM is converted into S-adenosylhomocysteine (SAH) (glycine N-methyltransferase {GNMT step 5} by donating its methyl group through different methyltransferases in a wide variety of methylation reactions (such as methylation of proteines, DNA and lipids), and is also a strong inhibitor of those methyltransferases. 5. Adenosylhomocysteine hydrolase {AHCY, step 6} hydrolyzes SAH into HCY and adenosine.
Trans-sulfuration pathway {steps 7 and 8} This reaction is irreversible and consists of two major steps: 1. HCY is condensed with serine to form cystathionine through cystationine b-synthase {CBS, step 7} using vitamin B6 as cofactor. 2. cystathionine is converted into cysteine and a-ketobutyrate through cystathionine g-lyase {CTH, step 8} using vitamin B6 as cofactor. Decarboxylation of a-ketobutyrate leads to propionyl- CoA, which enters the citric acid cycle as succinyl-CoA.
Glutathione metabolism {steps 20 to 23} 1. Cysteine is converted into glutathione (GSH) by sequential actions of g-glutamyl cysteine synthetase {GCLC, step 20} and glutathione synthetase {GSS, step 21}. Cysteine can also be converted into pyruvate, which is a metabolite of glucose metabolism. 2. Glutathione peroxidase {GPx, step 22} uses GSH to reduce H2O2 and form oxidized GSH, GSSG. 3. Glutathione reductase {GSR, step 23} converts the oxidized form, GSSG, back to the reduced form, GSH. In mammalian cells, most glutathione occurs in the form of GSH. 51 studies, trans-sulfuration pathway is complete in only four tissues: liver, kidney, small intestine and pancreas (FINKELSTEIN 1990).
The trans-methylation and trans-sulfuration pathways are tightly regulated by nutrients, such as methionine (SELHUB 1999) and by two classes of enzymes.
The “methionine conserving” enzymes (Figure I-1), i.e. methylenetetrahydrofolate reductase (MTHFR, {step 1}), methionine synthase (MTR, {step 2}), betaine-HCY methyltransferase (BHMT, {step 3}), S-adenosylmethionine synthase (AMS, {step
4}), and S-adenosylHCY hydrolase (AHCY, {step 6}), are inhibited by their products, impaired by S-adenosylmethionine, and their hepatic content decreases with increased dietary methionine. Conversely, the “methionine catabolizing” enzymes, i.e. glycine N-methyltransferase (GNMT, {step 5}), cystathionine b-synthase (CBS, {step 7}), and cystathionine g-lyase (CTH, {step
8}), are not inhibited by their products, are activated by S-adenosylmethionine, and their hepatic content increases in response to increased dietary methionine
(FINKELSTEIN 1990). According to these regulations, two predictions can be considered:
i) With high dietary methionine, S-adenosylmethionine levels increase
and inhibit MTHFR and BHMT resulting in suppressed 5-
methyltetrahydrofolate (5-methyl-THF), and therefore in activation of GNMT
(whose inhibition by 5-methyl-THF is lifted). Also, increased S-
adenosylmethionine levels activate CBS. In short, with high levels of 52
methionine, the trans-sulfuration pathway is favored over the trans-
methylation of HCY (FINKELSTEIN 1990; SELHUB 1999).
ii) In contrast, with low dietary methonine, S-adenosylmethionine
levels decrease and are insufficient for MTHFR and BHMT inhibition, resulting
in increased 5-methyl-THF which in turn inhibits GNMT. Also, decreased S-
adenosylmethionine levels activate CBS. In summary, with low levels of
methionine, the trans-methylation, over the trans-sulfuration, metabolism is
favored (FINKELSTEIN 1990; SELHUB 1999).
b) Extracellular export of HCY
Liver largely contributes to plasma HCY (STEAD et al. 2000). Cellular export of HCY may be critical in sustaining vital cellular functions by maintaining low intracellular levels of HCY (UELAND et al. 1986; CHRISTENSEN et al. 1991). In agreement with HCY export, animal studies showed that HCY is present at lower levels in tissue than in plasma (FOWLER 2001). Then, extracellular or plasma
HCY represents most likely the balance between HCY production and catabolism
(FOWLER 2001). When the equilibrium is disrupted, HCY accumulates in the cells and is then released into the plasma and other body fluids leading to hyperhomocysteinemia (PIETRZIK and BRONSTRUP 1997). 53
FOLATE
Folate, also referred as tetrahydrofolate (THF), participates in purine and pyrimidine biosynthesis and in the management of the major pool of single carbon units, where serine serves as the major source of carbon units (Figure I-
1, {step 9}; (ROSENBLATT and FENTON 2001)). Folate coenzymes (Figure I-1,
{steps 9-19}) are present in virtually every known organism and cell type
(APPLING 1991). Box 2 describes folate metabolism in detail.
1) Chemistry
Figure I-3 illustrates the structure of folate. Extracellular folate contains a single glutamate residue whereas intracellular folate contains a polyglutamate chain usually consisting of 5-8 glutamate residues (SUH et al. 2001). The polyglutamate chain is essential in retaining folate within the cell and increases the affinity of folate for folate-dependent enzymes (SCHIRCH and STRONG 1989;
SUH et al. 2001). Serum folate represents circulating folate, which is greatly influenced by diet, whereas red blood cell folate reflects intracellular folate and strongly correlates with hepatic levels (SUH et al. 2001).
2) Regulation of folate metabolism
As opposed to most bacteria and yeast, mammals cannot synthesize folates de novo, and thus depend on dietary sources or intestinal bacteria. Dietary folates include liver, leafy green vegetables, citrus fruits (LUCOCK 2000), flour and 54
BOX 2: Folate metabolism Because folate (THF) metabolism has a higher degree of interrelated reactions than HCY, it will be described in the following steps: conversion of THF, de novo purine synthesis, DNA and pyrimidine synthesis,. Conversion of THF into 5,10-methylene-THF and 5,10-methenyl-THF {steps 9-10} 1. Serinehydroxymethyltransferase {SHMT, step 9} converts THF into 5,10-methylene-THF, and catalyzes the conversion of serine, the major source of 1-carbon units to glycine. 2. A bifunctional enzyme, with glutamate formiminotransferase and formiminotetrahydrofolate cyclodeaminase acitivities {FTCD, step 10}, converts THF into 5,10-methenyl-THF, and catalyzes the transfer of a formimino group from formiminoglutamate {FIGLU} onto THF to form glutamate {GLU}. de novo purine synthesis {steps 11-17} 1. The interconversion of 5,10-methylene-THF, 5,10-methenyl-THF and 10-formyl-THF is mediated by a trifunctional enzyme in the cytoplasm (methylene-THF dehydrogenase/methenyl- THF cyclohydrolase/10-formyl-THF synthase; {MTHFD1 steps 11, 12, 13}) and bifunctional enzyme in the mitochondria (methylene-THF dehydrogenase/methenyl-THF cyclohydrolase; {MTHFD2 steps 11 and 12}) that bear activity for: a. 5,10-methylenetetrahydrofolate dehydrogenase, which reversibly converts 5,10- methylene-THF into 5,10-methenyl-THF {step 11} b. 5,10-methenyltetrahydrofolate cyclohydrolase, which reversibly converts 5,10-methenyl- THF into 10-formyl-THF {step 12} c. 10-formyltetrahydrofolate synthetase {step 13}, which reversibly converts 10-formyl-THF into THF and formic acid {HCOOH}. This reaction permits either the release of single carbon from folate as formic acid or the scavenging of potentially toxic formic acid 2. 10-formyl-THF is used as formyl donor to form THF and participates in de novo purine nucleotide biosynthesis by: a. 5-phosphoribosylglycineamide transformylase {GART, step 14} transfers the formyl group onto glycineamide ribonucleotide {GAR} to form formylglycineamide ribonucleotide {FGAR} b. 5-phosphoribosyl-5-aminoimidazole-4-carboxamide transformylase {ATIC, step 15} transfers the formyl group onto 5-phosphoribosyl-5-aminoimidazole-4-carboxamide ribonucleotide {AICAR} to form FAICAR. 3. 5,10-methenyltetrahydrofolate synthetase {MTHFS, step 16} irreversibly catalyzes 5-formyl- THF into 5,10-methenyl-THF. 4. Formyltetrahydrofolate dehydrogenase {FTHFD, step 17} irreversibly converts 10-formyl-THF into THF and carbon dioxide {CO2}. This reaction serves to maintain sufficient THF to permit acceptance of single carbons in folate-dependant reaction. DNA and pyrimidine synthesis {steps 18-19} 1. Thymidylate synthase {TYMS, step 18} oxidizes 5,10-methylene-THF into DHF and also catalyzes the reductive methylation of uridine monophosphate {dUMP} into tyrosine monophosphate {dTMP}. 2. Dihydrofolate reductase {DHFR, step 19} reduces DHF and folic acid into THF 55
Figure I-3. Chemical structure of folate (from (SUH et al. 2001)) Structure of tetrahydrofolate (THF) with three molecules of glutamate. 56 cereal grains sources (ROHAN et al. 2000). Folic acid is a dietary supplement and a synthetic form of folate, and has to be modified into THF by dihydrofolate reductase (DHFR; Figure I-1, {step 19}) (ROSENBLATT and FENTON 2001). Dietary folates and supplements are then absorbed, transported, stored and utilized by the cells.
a) Absorption
Dietary folates are transported across the brush border membrane of the intestine and absorbed by intestinal cells (LUCOCK 2000). Before absorption, polyglutamate folates must be hydrolyzed into monoglutamates, which are then metabolized into 5-methyl-THF by MTHFR (Figure I-1, {step 1}). 5-methyl-THF is the primary form supplied to tissues (LUCOCK 2000; SUH et al. 2001).
b) Transport
In the circulation, 5-methyl-THF is associated with low affinity binding proteins, such as albumin, a2-macroglobulin and transferrin (LUCOCK 2000).
There are two principal modes of 5-methyl-THF transport into the cell:
i) Folate receptors. They are membrane-bound receptors and mediate
the unidirectional transport of folate into the cell probably via endocytosis
(SUH et al. 2001).
ii) Reduced folate carriers. They are mobile, mediate bidirectional
transport of folate, may act as anion exchangers to concentrate intracellular 57
folate (SIROTNAK and TOLNER 1999) and are more efficient in transporting
folate than folate receptors (SUH et al. 2001).
c) Intracellular storage
Most circulating 5-methyl-THF is taken up by the liver (LUCOCK 2000).
Because 5-methyl-THF is a poor substrate for folate polyglutamylation, it must be converted into THF by MTR (Figure I-1, {step 2}) for synthesis of polyglutamate forms, which are retained by the cells (LUCOCK 2000; COOK 2001). The liver is the main organ that stores folates (STEINBERG 1984). Negligible free folate is present and most intracellular folate is protein-bound, such as to 10-formyl-THF dehydrogenase (FTHFD; Figure I-1, {step 17}) and to GNMT (Figure I-1, {step 5};
(SUH et al. 2001)). In agreement with this concept, in rats, intracellular folate levels do not proportionally increase with folate intake suggesting that intracellular folates are independent of excess of exogenous folate supply and are dependent on folate-binding capacity of the cell (SUH et al. 2001).
HOMOCYSTEINE AND FOLATE
Homocysteine and folate metabolism are two inter-related pathways that are tightly linked through 5,10-methylene-THF, which is reduced to 5-methyl-THF for 58
HCY metabolism, oxidized to 10-formyl-THF for purine synthesis or unchanged for DNA synthesis (ROSENBLATT and FENTON 2001). There is often an inverse relationship between HCY and folate levels: with high serum folate, HCY is low and vice versa (ROSENBLATT and FENTON 2001). When HCY and folate metabolisms are disrupted, these two pathways compete for folate cofactors
(SCOTT et al. 1981; GREEN et al. 1988; HERBIG et al. 2002). With low levels of methionine or folate, folate coenzymes are preferentially directed towards HCY metabolism to protect methylation reactions (by maintaining adequate levels of
S-adenosylmethionine, the major intracellular methyl donor) and suppress DNA synthesis Scott, 1981 #310;Green, 1988 #577]. Whereas, with increased glycine levels, folate coenzymes are shuttled towards DNA synthesis at the expense of
HCY metabolism (HERBIG et al. 2002).
Folate deficiency or hyperhomocysteinemia are often associated with various birth defects and adult diseases (see “HOMOCYSTEINEMIA AND
HUMAN DISEASES” section). Folate, vitamins B6 and B12 are often used to treat hyperhomocysteinemia, although folate treatment alone seems sufficient to treat the mild and moderate forms of hyperhomocysteinemia (COLLABORATION 1998;
VAN DER GRIEND et al. 2000). Moreover, supplementation with B vitamins reduces the risk for neural tube defects (GROUP 1991; HONEIN et al. 2001), ischaemic heart disease (SCHNYDER et al. 2001), colon (GIOVANNUCCI et al. 1995;
GIOVANNUCCI et al. 1998; SU and ARAB 2001) and breast (ZHANG et al. 1999) cancers. Therefore, maintenance of adequate extracellular (or levels in plasma at 59
≥ 6.8 nmol/L; (PRINZ-LANGENOHL et al. 2001)) and intracellular (or levels in red blood cells at ≥ 317 nmol/L; (PRINZ-LANGENOHL et al. 2001)) levels of folate combined with reduced HCY levels (or plasma levels at ≤ 12 mmol/L; (PIETRZIK and BRONSTRUP 1997)) would be favored in the prevention of disease. Daily
supplementation with both 500-5000 mg folic acid and 500 mg of vitamin B12
would reduce plasma HCY levels by 25-30%, whereas vitamin B6 would not have any additional effect (COLLABORATION 1998). The current reference daily intake is
400 mg for folic acid, 6 mg for vitamin B12, and 2000 mg for vitamin B6 (FAIRFIELD and FLETCHER 2002). To meet the folate intake requirement, all flour and cereal grains have been supplemented with 140 mg of folate per 100 g of flour since
1996 in the United States (FAIRFIELD and FLETCHER 2002). This fortification increased plasma folate levels by ~54% (JACQUES et al. 1999) and reduced the occurrence of NTDs by 19% (HONEIN et al. 2001).
RESEARCH AIMS
Several genetic and environmental factors affect HCY levels, making homocysteinemia a complex trait and its genetic and molecular control difficult to dissect in humans. The laboratory mouse is an important model to investigate control of homocysteinemia because HCY levels in humans and mice are 60 influenced by similar genetic and physiological factors. I used single gene mutants, multigenic models and dietary perturbations to investigate the control of homocysteinemia in disease and in health.
Chapters Two, Three and Four address the following question: Do mutations in genes other than CBS, MTHFR and MTR affect HCY-folate metabolism? Mutations in CBS, MTHFR and MTR alone do not explain the frequency or variety of birth defects and adult diseases in humans. Other sources of variation include mutant alleles that have more subtle effects in these genes, mutations in other genes in HCY-folate metabolism, or perhaps mutations in genes in other developmental and physiological pathways. Chapters Two and
Three explore genetic or physiological perturbation to investigate causes and consequences of homocysteinemia. Chapter Four examines the significance of perturbations to study complex biological systems.
Chapter Two describes studies that tested whether HCY-folate metabolism was disrupted in single gene mouse models of NTDs and colon cancer, which are phenotypes associated with hyperhomocysteinemia in humans. Increased serum HCY levels and altered expression profiles for genes involved in HCY-folate metabolism were detected in mice with partial protein deficiencies of GLI3, PAX3, APC and APOB. These results suggested that WNT and hedgehog signal transduction as well as lipid transport adversely affect HCY- folate metabolism. 61
Chapter Three describes studies that tested the status of HCY-folate metabolism in Crooked tail mutant mice, whose developmental defects are folate-responsive, under various dietary folate supplementations. In humans, elevated HCY levels are a risk factor for NTDs and folic acid taken before conception and during pregnancy reduces the NTD recurrence risk. Moreover,
HCY and folate levels are usually inversely related. Therefore, maintenance of adequate folate levels combined with reduced HCY levels would be favored in the prevention of NTDs. Surprisingly, the Cd mutation did not significantly affect
HCY levels but altered expression profiles of genes involved in HCY-folate metabolism. These results provide clues to gene-diet interactions that modulate
HCY-folate metabolism.
Chapters Five and Six address the following question: Do genes contribute to variation in normal HCY levels? In humans, genetic variants are involved in hyperhomocysteinemia whereas their role in the variation of HCY levels in the healthy population is unclear. Chapters Five and Six investigate the genetic control of normal homocysteinemia in two inbred strains, A/J and
C57BL/6J (B6).
Chapter Five investigates the inheritance of normal homocysteinemia in
A/J and B6 mice. Normal homocysteinemia was a complex trait that depended on gender, diet and parental effects. Moreover, physiological factors strongly influenced heritability of normal homocysteinemia. The results provided important clues about the complexity and inheritance of normal homocysteinemia. 62
Chapter Six consists of detecting genetic factors affecting normal homocysteinemia using the mode of inheritance provided by the results of
Chapter Five. Genetic mapping was used to detect quantitative trait loci (QTL) that control normal homocysteinemia in A/J and B6 mice. According to studies of
Chapter Five, chromosome substitution strains (CSSs) rather than intercross F2 mice were the best resource for genetic mapping of normal homocysteinemia in
A/J and B6 mice. Several QTLs were detected suggesting that genetic factors contribute to normal homocysteinemia. 63
CHAPTER II GENETIC AND MOLECULAR CONTROL OF FOLATE-HOMOCYSTEINE METABOLISM IN MUTANT MICE 64
AUTHORS
Sheila Ernest, Benedicte Christensen, Brian M. Gilfix, Orval A. Mamer, Angela
Hosack, Mitchell Rodier, Clemencia Colmenares, James McGrath, Allen Bale,
Rudi Balling, David Sankoff, David S. Rosenblatt and Joseph H. Nadeau
REFERENCE
Mammalian Genome 13, 259-267 (2002)
ABSTRACT
Hyperhomocysteinemia adversely affects fundamental aspects of fetal development, adulthood and aging, but the role of elevated homocysteine levels in these birth defects and adult diseases remains unclear. Mouse models are valuable for investigating the causes and consequences of hyperhomocysteinemia. We used a phenotype-based approach to identify mouse mutants for studying the relation between single gene mutations, homocysteine levels as a measure of the status of homocysteine metabolism, and gene expression profiles as a way to assess the impact of protein deficiency in mutant mice on steady-state transcription levels of genes in the folate-homocysteine pathways. These mutants were selected based on their propensity to produce 65 phenotypes that are reminiscent of those associated with anomalies in folate- homocysteine metabolism in humans. We report identification of new single gene mouse models of homocysteinemia and characterization of their molecular and physiological impact on folate-homocysteine metabolism. Mutations in several genes involved in the hedgehog and WNT signal transduction pathways as well as a gene involved in lipid metabolism resulted in elevated homocysteine levels and altered expression profiles of folate-homocysteine metabolism genes. These results begin to unravel the complex relations between elevation of a single amino acid in the blood and the diverse birth defects and adult diseases associated with hyperhomocysteinemia.
INTRODUCTION
Various birth defects and adult diseases occur more frequently when blood levels of homocysteine (HCY), a sulfur-containing amino acid, exceed
12 µmol/L (KANG et al. 1992). Normal human HCY blood levels usually range from 4.9 to 11.7 µmol/L (PIETRZIK and BRONSTRUP 1997). Hyperhomocysteinemia
(significantly elevated HCY levels) occurs at a frequency of 5% in the general population (HANKEY and EIKELBOOM 2000). It is an independent risk factor for neural tube defects (NTDs; (GROUP 1991), (STEEGERS-THEUNISSEN et al. 1991) 66 and is also associated with colon, breast, ovarian and pancreatic cancer (MAYER et al. 1996), Alzheimer disease (JOOSTEN et al. 1997; CLARKE et al. 1998; MILLER
1999) and with certain treatments for Parkinson disease (MULLER et al. 1999).
The variety of birth defects and adult diseases associated with elevated HCY levels is striking. It is not obvious how elevated levels of a single amino acid and its metabolites could adversely affect so many organs, tissues, and physiological processes. Moreover, it is unclear whether elevated HCY is itself pathogenic or simply a disease marker.
Homocysteine is remethylated through the trans-methylation pathway or irreversibly catabolized through the trans-sulfuration pathway (MUDD et al. 2001);
Figure I-1). It contributes directly to methylation of DNA, proteins and lipids; it contributes indirectly to glucose metabolism and citric acid cycle. Homocysteine metabolism is also tightly linked to folate metabolism (Figure I-1), which is the major pool of single carbon units. Both pathways participate in purine and pyrimidine biosynthesis and amino acid metabolism (alanine, glycine and serine;
(ROSENBLATT and FENTON 2001). Regulation of the folate and HCY pathways is complex and, because of their inter-relatedness, it is difficult to predict the biochemical and phenotypic consequences of inhibiting particular steps (KISLIUK
1999). 67
Mutations in MTHFR (methylenetetrahydrofolate reductase; (KANWAR et al.
1976; KANG et al. 1991; CHRISTENSEN et al. 1999), CBS (cystathionine b- synthase; (MUDD et al. 1964; BOERS et al. 1985b), MTR (HCY:methionine methyltransferase, also known as methionine synthase; (GULATI et al. 1996;
LECLERC et al. 1996), and MTRR (methionine synthase reductase) are also associated with hyperhomocysteinemia (LECLERC et al. 1998). More importantly, the cumulative effect of these various mutations does not readily account for the frequency or diversity of these birth defects and adult diseases. Other sources of variation include mutant alleles that have more subtle effects in these genes
(GOYETTE et al. 1996; VAN DER PUT et al. 1998), mutations in other genes in folate-HCY metabolism, or perhaps mutations in genes in other developmental and physiological pathways.
As a first step towards understanding the genetic and molecular controls of HCY levels, a panel of mutant mice was characterized for serum HCY levels and for hepatic expression profiles of most genes in the folate-HCY pathways as well as a comparable number of other genes as controls. An important part of the experimental design was to characterize HCY levels and gene expression in mice with single gene mutations that have phenotypes similar to birth defects and adult diseases associated with hyperhomocysteinemia in humans, such as neural tube defects and cancer. By testing whether HCY levels and expression profiles are altered in these single gene mutant mice, we can begin to identify 68 genes and processes that adversely affect HCY metabolism. This phenotypic- driven approach does not make assumptions about the identity of genes or biochemical processes that control HCY levels. As a result, we may discover novel and unexpected influences on HCY metabolism.
This study revealed mutant mice in which HCY levels were significantly elevated as a consequence of the mutation as compared to their wild-type controls. We also found significant differences in gene expression patterns, often involving genes in the folate-HCY pathways in these mutant versus control comparisons. These differences provide clues to new genes, pathways and functions that adversely affect folate-HCY metabolism.
RESULTS
Serum HCY levels in mutant mice
Serum HCY levels were measured in mutant mice with single gene mutations (Figures II-1 and II-2). The criteria for selecting mutant mice for this study was their propensity to cause disorders, such as neural tube defects and cancer, that are similar to anomalies associated with hyperhomocysteinemia in humans (MUDD et al. 2001; ROSENBLATT and FENTON 2001). Female mice were 69 used because in humans maternal hyperhomocysteinemia is a risk factor for fetal neural tube defects (MALINOW et al. 1998), although anomalies in fetal metabolism also appear to be important (FLEMING and COPP 1998). Virgin female mice were used to control for possible effects of pregnancy on HCY levels (KANG et al. 1986; STEEGERS-THEUNISSEN et al. 1997). Levels were measured in heterozygous ApobtmlUnc, Pax3Sp, Apcmin, Ptch1+/-, Ski+/- and Gli3XtJ compared to their respective wild-type (+/+) controls (Figures II-1 and II-2). Because homozygotes for these mutations show anencephaly or spina bifida and die during embryogenesis (Mouse Locus Catalog; www.informatics.jax.org), it was not possible to collect sufficient blood from these fetuses to measure HCY levels.
Because protein content can vary between plasma and serum, and because HCY is mainly protein-bound, levels were compared in serum and plasma. Plasma and serum samples were obtained by collecting blood from the retro-orbital sinus of virgin female (A/J x C57BL/6J)F1 hybrids that were 6-8 weeks old. Homocysteine levels were not significantly different between plasma and serum (mean HCY level in serum: 5.5±1.1 µmol/L; mean HCY in plasma:
4.9±0.7 µmol/L; paired two-tailed Student's t test, p=0.18). HCY levels were therefore measured in serum samples in the following studies.
When compared to their respective controls, Apcmin, ApobtmlUnc, Gli3XtJ and
Pax3Sp mice had significantly higher HCY levels than the C57BL/6J (B6) 70 background control strain (Figure II-1). The average level increased by 18% in
Pax3Sp, 24% in Gli3XtJ, 25% in ApobtmlUnc, and 31% in Apcmin mice. Gli3XtJ (on the
C3HeB/FeJ background), Ptch1+/-, and Ski+/- showed substantial overlap with their respective Ptch1+/+ and Ski+/+ wild-type control (Figure II-2). These results show that several single gene mutations adversely affect folate and HCY metabolism and that the effect can depend on genetic background. Pax3 and
Gli3 are part of the hedgehog signal transduction pathway (GORLIN 1987;
DOMINGUEZ et al. 1996; GOODRICH et al. 1996), Apc contributes to WNT signal transduction (BEHRENS et al. 1998), and Apob participates in lipid metabolism
(FARESE and HERZ 1998). This is the first evidence that mutations in these pathways adversely affect folate-HCY metabolism.
To test whether vitamin B12 (cobalamin) deficiency might be responsible for elevated HCY levels, methylmalonic acid (MMA) levels were measured
(UBBINK et al. 1991; GILFIX et al. 1997) in selected inbred strains and mutant
mice. Vitamin B12 deficiency is associated with increased MMA levels (STABLER et
al. 1986). Methylmalonyl-CoA mutase requires vitamin B12 as a cofactor, whereas
methionine synthase (MTR) requires both folic acid and vitamin B12 as cofactors
(KIRKE et al. 1993). If vitamin B12 was deficient, both MMA and HCY levels should be elevated. However, none of the mice with elevated HCY had high MMA levels
(data not shown), arguing that these cases of homocysteinemia did not result
from vitamin B12 deficiency. 71 1.2; ± , 7.4 XtJ Gli3 1.6; ± , 7.8 min Apc 1.6; ± , 6.8 Sp Pax3 0.6; ± 0.9. To assess differences in mean HCY levels, Dunnett’s test was used. Values are expressed as in mean HCY levels, Dunnett’s test was 0.9. To assess differences ± , 7.5 standard deviation. *P<0.05, **P<0.01 standard deviation. *P<0.05, ± tmlUnc FIGURE II-1. Homocysteine levels for mutants congenic on the B6 background for mutants congenic Homocysteine levels FIGURE II-1. (µmol/L) for: B6, 5.6 Mean homocysteine levels Apob mean 72 0.8 and ± , 4.6 +/+ Ptch1 0.3; ± consisted of their wild-type sibs from consisted of their wild-type t test was used. Values are t test was used. Values , 4.7 +/- Xt Ski and Student’s +/- Ptch1 0.4; C3HeB/FeJLe- ± 1.6. Controls for ± , 7.7 +/- Ski 1.0 and ± , 8.3 +/+ standard deviation. ns, non significant. standard deviation. ns, ± Ski 1.2; ± , 4.7 +/- FIGURE II-2. Homocysteine levels for mutants on different backgrounds for mutants on different Homocysteine levels FIGURE II-2. (µmol/L) for: C3HeB/FeJLe, 5.0 Mean homocysteine levels Ptch1 levels, assess differences in mean homocysteine a segregating cross. To expressed as mean 73
Expression studies on mutant mice
Expression profiles (BROWN and BOTSTEIN 1999; DUGGAN et al. 1999;
TAMAYO et al. 1999) are a powerful method to identify genes whose RNA abundance is modulated in mutant mice. Arrays of gene-specific PCR products
(GRESS et al. 1992; BERTUCCI et al. 1999) were used to compare profiles as a way of testing whether the RNA abundance of genes in folate-HCY metabolism was altered in mutant versus control mice. These arrays were composed of most
(25 out of 27 genes) of the genes involved in folate-HCY metabolism (46.7%) with the remaining 28 genes (53.3%) being functionally unrelated to this pathway
(Table II-1; MTHFS {step 16} of Figure I-1, and FTHFD {step 17} of Figure I-1, were not included because no sequence was available). Liver was used as a source of RNA because folate-HCY metabolism occurs primarily in this organ
(MUDD et al. 2001; ROSENBLATT and FENTON 2001). Arrays were hybridized with four replicate pools of radiolabelled total liver cDNA from the same mutants and controls that were used for the HCY survey.
Two methods were used to identify outliers in each mutant versus wild- type control comparison: 1) two standard deviations from the mean expression level for all genes, and 2) greater than 3-fold difference in RNA level (see
Materials and Methods for a detailed description of the data analysis). Because only heterozygotes could be tested, any changes in RNA abundance were expected to be subtle. Interestingly, the observed changes ranged between two- 74 and eight-fold. Twelve genes, eight of which were directly involved in folate-HCY metabolism, showed a difference of at least 2 standard deviations and two additional genes differed at least 3-fold in expression levels (Table II-2). For instance, FBP2 levels were significantly increased by 4.9-fold and GNMT levels were decreased by 3-fold in heterozygous B6-Pax3Sp mutants compared to B6 wild-type controls. In general, the increases ranged from 1.8-fold for AHCY in
C3HeB- Gli3XtJ to 5.6-fold for ATIC in B6- Apcmin mice. The decreases ranged from 1.8-fold for SHMT in B6- ApobtmlUnc to 7.9-fold for TYMS in B6-Pax3Sp.
Therefore, 36% of HCY-folate genes used in the array (9/25) have altered expression levels.
Results for the GLI3 are noteworthy in two respects. First, the increase in
GLI3 RNA abundance in mice that were partially deficient for PTCH1 (Table II-2) and its decrease in mice partially deficient for PAX3 are consistent with the inhibitory effect of PTCH1 on GLI3 function (MURONE et al. 1999) and the activating influences of GLI3 on PAX3 function (DAHL et al. 1997). Second, the
GLI3 RNA level decreased in both B6-Pax3 and Apc mutant mice, both of which showed elevated HCY levels as compared to their controls, whereas the GLI3 levels increased in Ptch mutant mice, which showed similar HCY levels as their control. The association between changes in HCY levels and GLI3 RNA levels suggest that a correlation, and perhaps a causal relation, might be involved. 75
TABLE II-1. List of genes used on the arrays
The name and symbol of the genes as well as to which pathway the genes belong to are indicated. The DNA sequences (Genbank accession number) and the portion of the genes (bases) used for the design of PCR products are also indicated. Eight genes (AHCY, AMS, DHFR, LEPR, NRAMP, PLI, RFC and
TYMS) were represented by two PCR products, one for the 5’ end of the gene and the other for the 3’ end of the gene.
GENES Genbank Bases acc# Name Symbol Pathway Adenosylhomocysteine hydrolase 3' end AHCY 3' L32836 1369-1874 HCY-Folate Adenosylhomocysteine hydrolase 5' end AHCY 5' L32836 171-619 HCY-Folate Plasmin inhibitor alpha 2 3' end PLI 3' Z36774 951-1458 Inflammatory response Plasmin inhibitor alpha 2 5' end PLI 5' Z36774 207-626 Inflammatory response Apolipoprotein E APOE M12414 352-843 Lipid transport Beta actin ACTB X03672 294-893 WNT signal transduction/ Cystoskeleton Betaine homocysteine methyltransferase BHMT W16199 89-315 HCY-Folate Breast Cancer 1 BRCA1 U36475 5997-6398 Carbohydrate metabolism, cell growth and/or maintenance Cartilage homeoprotein 1 CART1 X92346 1052-1455 DNA-dependent transcription regulation Cystathionine beta-synthase CBS AA096780 25-236 HCY-Folate Ceruloplasmin CP U49430 2278-2714 Inflammatory response UDP galactose ceramide-galactosyl transferase CGT X92122 441-850 Inflammatory response Cystathionine gamma lyase CTH W11479 114-316 HCY-Folate Dihydrofolate reductase 3' end DHFR 3' L26316 795-1253 HCY-Folate Dihydrofolate reductase 5' end DHFR 5' L26316 62-478 HCY-Folate 5-Phosphribosyl-5-aminoimidazole-4-carboxamide transformylase ATIC W08334 116-318 HCY-Folate Folate binding protein 1 FOLBP1 M64782 260-699 HCY-Folate Folate binding protein 2 FOLBP2 M64817 402-896 HCY-Folate Fibrinogen A alpha polypeptide FGA D43759 72-571 Inflammatory response Glyceraldehyde-3-phosphate dehydrogenase GAPD M32599 322-910 Glycolysis Glutamate formiminotransferase FTCD AA051787 175-466 HCY-Folate GLI-Kruppel family member GLI3 GLI3 X95255 317-884 Hedegehog signal transduction Glucose transporter 1 GLUT1 D10230 26-544 Glucose transport Glycine N-methyltransferase GNMT W83078 202-420 HCY-Folate Haptoglobin HP M96827 165-672 Inflammatory response Heme Oxygenase 1 HMOX1 X13356 181-587 Inflammatory response Inositol triphosphate type 2 receptor ITPR2 Z71173 2436-2837 Phosphoinositide signaling pathway Inositol triphosphate type 3 receptor ITPR3 Z71174 169-571 Phosphoinositide signaling pathway Inositol polyphosphate-1 phosphatase INPP1 U27295 912-1314 Phosphoinositide signaling pathway Inositol polyphosphate-5 phosphatase IPOLY5 W17555 130-358 Phosphoinositide signaling pathway Leptin Receptor 3' end LEPR 3' U42467 3779-4352 Signal transduction Leptin Receptor 5' end LEPR 5' U42467 24-516 Signal transduction Methenyl cyclohydrolase, mitochondrial MTHFD1 (formerD21754 MTCH) 139-357 HCY-Folate 5-methyltetrahydrofolate-homocysteine methyltransferase or methionine synthaseMTR Shane B, pers. communiHCY-Folate 5,10-Methylenetetrahydrofolate reductase MTHFR AA183742 114-359 HCY-Folate Neuropeptide Y receptor Y1 NPY1R Z1438069 851-1363 G-protein coupled-receptor signaling pathway Neuropeptide Y receptor Y6 NPY6R U58367 1518-1918 G-protein coupled-receptor signaling pathway Natural resistance-associated macrophage protein 1 NRAMP1 L13732 1014-1422 Inflammatory response Natural resistance-associated macrophage protein 2 3' end NRAMP2 3' L33415 1083-1492 Inflammatory response Natural resistance-associated macrophage protein 2 5' end NRAMP2 5' L33415 167-588 Inflammatory response Oncostatin M OSM D31942 339-800 Inflammatory response Plasminogen activator inhibitor type 1 PAI1 M33960 180-741 Inflammatory response Transthyretin (Prealbumin) TTR X03351 57-376 Inflammatory response 5-Phosphoribosylglycineamide transformylase GART U01024 2481-2896 HCY-Folate Reduced folate carrier 3' end RFC 3' U32469 1663-2233 HCY-Folate Reduced folate carrier 5' end RFC 5' U32469 314-731 HCY-Folate Serum amyloid A5 SAA5 U02554 162-509 Inflammatory response S-adenosylmethionine synthetase 3' end AMS 3' L13622 2575-3058 HCY-Folate S-adenosylmethionine synthetase 5' end AMS 5' L13622 302-706 HCY-Folate Serum amyloid P-component SAP X14079 195-634 Inflammatory response Serinehydroxymethyltransferase, cytoplasmic SHMT1 AA028497 250-477 HCY-Folate NADP+ dependent methylenetetrahydrofolate dehydrogenase MTHFD2 J04627 1377-1907 HCY-Folate cyclohydrolase synthetase Thymidylate synthase 3' end TYMS 3' M13019 432-863 HCY-Folate Thymidylate synthase 5' end TYMS 5' M13019 32-458 HCY-Folate 76
Results for TYMS in Pax3Sp mutant mice are also noteworthy. These mutant mice are thymidine-deficient (FLEMING and COPP 1998) and this may result from TYMS down-regulation (Table II-2; thymidine is represented as dTMP in Figure I-1 {step 18}). Therefore, these results are consistent with known physiological processes.
The outlier genes AHCY, LEPR, RFC and TYMS were each represented by two PCR products, one for the 5’ end and the other for the 3’ end of the gene.
There was no correlation in expression level between the 5' and 3' probes for each of these genes, i.e. only one of the products was altered in the mutant versus control comparison: TYMS 3’, AHCY 5’, RFC 3’ and LEPR 5’ in B6-Pax3Sp mice; LEPR 3’ in Apcmin ; AHCY 5’ in C3HeB-Gli3XtJ; and RFC 3’ in Ski mice.
These results raise the possibility that these genes are alternatively spliced.
Review of the literature reveals evidence that both LEPR (MERCER et al. 1996) and RFC (TOLNER et al. 1997) produce alternative transcripts. Our LEPR 3’ PCR product detected the ob-r transcript, whereas the LEPR 5' PCR product detected the a, b, d and e transcripts. Moreover, Mercer and coworkers also found expression differences between LEPR 3' and 5' sequences in various mouse brain regions (MERCER et al. 1996). By contrast, our 5’ and 3’ PCR products did not distinguish the alternative RFC transcripts. To test whether the other two genes, AHCY and TYMS, produce alternative transcripts, we performed a 77 Decreased SHMT, 1.8X TYMS, 7.9X SAP, 4.9X GNMT, 3.0X LEPR, 3.7X GLI3, 3.4X GLI3, 3.8X AHCY, 3.5X RFC, 2.0X RNA abundance FBP2, 4.9X ATIC, 3.4X INPP1, 3.2X ATIC, 5.6X LEPR, 3.4X AHCY, 1.83X MTHFR, 2X GART, 2.5X GLI3, 2.4X SAA5, 2.5X FBP2, 3.0X Increased 3 3 1 1 4 2 1 defect Neural tube Exencephaly and spina bifida No obvious defect Exencephaly Exencephaly Spina bifida Exencephaly Exencephaly showing consistent changes in gene expression levels unique to expression levels changes in gene showing consistent levels Effect on HCY Increased Increased Increased Unchanged Unchanged Unchanged Increased ), whereas their expression is often elevated in strains with normal HCY is often elevated in strains with normal ), whereas their expression XtJ ). Gli3 Ski Control Ptch1 +/+ Ski +/+ B6 B6 B6 B6 C3HeB/FeJLe and and B6-
min Ptch1 , , Apc XtJ Sp 2000. 2000. Gli3
Northern analysis of liver mRNA from B6 and B6-Pax3Sp for TYMS, and from B6 and B6-Gli3XtJ for AHCY. Expression differences between the 5' and 3' end sequences were confirmed for AHCY but not for TYMS (data not shown). Finally, the sequence of the human genome reveals an average of 3-4 alternative transcripts per gene (CONSORTIUM 2001; VENTER et al. 2001) , and it is likely that many genes in mice also have alternative transcripts.
Because of the inherent nature of the data analysis for expression profiles, some outlier genes are expected to be ‘false positives’, at least with respect to their association with changes in HCY levels. Two examples are suggested in this survey. FBP2 levels increased in both Apob and Ski mutant mice (Table II-2).
Apob mutant mice showed elevated HCY levels, whereas Ski mutant mice showed unchanged HCY levels. Similarly, LEPR levels decreased in B6-Pax3 mutant mice and increased in B6-Apc mutant mice. Both Apob and Ski mutants showed elevated HCY levels as compared to their controls. Thus, for both FBP2 and LEPR, there does not seem to be a simple relationship between changes in
HCY and RNA levels. Whether results for FBP2 and LEPR are ‘false positives’ or instead reflect more complex relations between HCY levels and RNA abundance remains to be determined. 79
Validation studies
Gene expression levels in liver samples were independently measured with Northern analysis for TYMS (probes for the 5' and 3' portion of the gene) in
B6 and B6-Pax3Sp; FBP2 in B6, B6-Pax3Sp, and in Ski+/+ and Ski+/-; AHCY 5' (the
3' end is not detected on the array and, therefore, could not be included in the validation) in B6 and B6-Gli3XtJ. Then, these results were compared to those for the corresponding genes on the array between the mutants and their respective controls. Using beta actin and GAPDH as controls for normalization, the correlation between expression levels for the Northerns and arrays was 0.95
(actin control: N=5, P<0.05) and 0.89 (GAPDH control: N=5, P<0.05). Thus, expression levels for Northerns and arrays were highly concordant in all 10 comparisons.
DISCUSSION
Overview
Given the complex relationships between the folate and HCY pathways on various birth defects and adult diseases, we studied HCY levels in mutant mice with phenotypes (NTDs and colon cancer) often associated with hyperhomocysteinemia in humans. Mice with mutations in the Apob, Gli3, Pax3, 80
Ptch and Ski genes were used as models for NTDs and mice with a mutation in the Apc gene as a model for colon cancer. Serum HCY levels were measured and hepatic gene expression profiles were monitored in heterozygous mutants, because homozygotes are embryonic lethal. An important attribute of surveying heterozygous mutants is the ability to test for subtle and specific consequences of partial protein deficiencies as compared to the severe phenotypic and metabolic abnormalities that occur in homozygous mutants. Partial deficiencies of APC, APOB, PAX3 and GLI3 (on the B6 background) increased HCY levels, whereas PTCH, SKI and GLI3 (on the C3H background) did not affect HCY levels. We also found significant differences in gene expression patterns, often affecting folate-HCY metabolism, in these mutants. These results provide clues to new pathways, i.e. WNT and hedgehog signal transduction (APC, GLI3 and
PAX3) and lipid transport (APOB) that adversely affect HCY metabolism and perhaps contribute to the pathogenesis of birth defects and adult diseases in these mice.
HCY levels in humans and mice
The magnitude of changes in HCY levels in mutant mice is comparable to that found in humans who are heterozygous for certain genetic mutations. In humans for example, plasma HCY levels are elevated 50% - 75% in heterozygotes for mutations in the CBS gene (WELCH and LOSCALZO 1998) whereas heterozygotes for the Ala / Val (C677T) mutation in the MTHFR gene 81 have similar HCY levels as Ala/Ala (C/C) ‘wild-type’ homozygotes (BRATTSTROM et al. 1998). Gene targeting was used to create mice with targeted deficiencies of the CBS (WATANABE et al. 1995), MTR (SWANSON et al. 2001) and MTHFR (CHEN et al. 2001) proteins. Compared to their respective wild-type controls, heterozygous CBS mutant mice show a 50% increase in the HCY level
(WATANABE et al. 1995), heterozygous MTR mutant mice show a 49% increase in
HCY level in males and a 70% increase in females (SWANSON et al. 2001), and heterozygous MTHFR mutant mice a 38% increase (CHEN et al. 2001). In our survey, mutant mice that are prone to NTDs or colon cancer have HCY levels increased between 18% and 31% when compared to their wild-type control
(Figures II-1 and II-2). The differences between these results is that CBS, MTR and MTHFR in humans and mice are directly involved in folate-HCY metabolism, whereas the mutant genes included in our survey are involved primarily in other developmental and physiological pathways and their metabolic effects may be modulated as their effects are transmitted through various pathways. Differences in diet may also contribute to the modest differences between HCY levels in humans and mice. Dietary folate levels vary considerably in humans, whereas the mutant and control mice were maintained on defined diets. In the present study, mice were maintained on a relatively high folate diet (7.5 mg/kg).
Nevertheless, the effects on HCY levels in these mutant mice are in the same direction and roughly of the same magnitude as those found in humans. 82
HCY metabolism and neural tube defects
Pregnant women with elevated HCY levels have an increased risk of giving birth to a child with a NTD (MILLS et al. 1995). When taken before and during pregnancy, folic acid prevents most (70%) NTDs (MRC Vitamin Study
Research Group, 1991), demonstrating the critical role that folate-HCY metabolism plays in neural tube development.
Distinct sites of NTDs in mutant mice reflect regional differences in mechanisms of normal neural fold elevation and closure (HARRIS and JURILOFF
1999). For instance, exencephaly is a failure of elevation of the neural folds in the region between the caudal border of the forebrain and rostral border of the hindbrain, whereas spina bifida is a failure to close the posterior neuropore
(HARRIS and JURILOFF 1999). Surveyed mice with exencephaly have both normal
(C3HeB-Gli3XtJ and Ski) and increased HCY levels (B6-ApobtmlUnc and B6-Gli3XtJ).
B6-Pax3Sp mice, which have both exencephaly and spina bifida showed the smallest increase in HCY level (Table II-2, Figure II-1). Thus, a relation between the type or severity of neural tube defect and HCY levels in these mutant mice was not evident; a pattern might become apparent with a larger survey of mutants. 83
Folate-HCY metabolism and cancer
Mutations in the APC gene, a key component of the WNT transduction pathway, contribute to colorectal cancer in humans (COTTRELL et al. 1992) and mice (SU et al. 1992) possibly by inducing chromosome instability (FODDE et al.
2001; KAPLAN et al. 2001). The folate-HCY pathway is involved in important methylation reactions, DNA synthesis and repair, and chromosome stability
(FENECH 2001). Disruption of folate and HCY pathways, for instance with low vitamin B12, methionine and folate intake, is an important risk factor for many cancers (FENECH 2001). Moreover, dietary folate supplementation suppresses colonic polyps in B6-ApcMin mice in an age-dependent manner (SONG et al.
2000a). Finally, folate is used as a therapeutic against colon cancer (JANNE and
MAYER 2000). The elevated HCY level (Figure II-1) as well as the changes in
RNA abundance levels (Table II-2) in Apc mutant mice provide a new model for studying the relations between WNT signal transduction, folate-HCY metabolism, and susceptibility to colon cancer.
HCY levels and expression profiles
HCY levels and expression profiles were monitored in heterozygous B6-
Apcmin, B6-ApobtmlUnc, B6-Gli3XtJ, B6-Pax3Sp, Ptch1+/-, Ski+/- and C3HeB-Gli3XtJ mutant mice to study the physiological and molecular responses of folate-HCY metabolism to specific genetic perturbations. The biochemical functions of the 84 outlier genes in the mutant versus control comparisons were evaluated to gain clues to the basis for elevated versus unchanged HCY levels in mice with partial deficiencies of APOB, PAX3, APC, GLI3, PTCH1 and SKI proteins (Table II-2). If these mutations adversely affect folate-HCY metabolism, the changes in RNA abundance levels should be consistent with particular physiological consequences, assuming that changes in RNA abundance result in altered functional activities.
One of the several provocative differences among the mutant mice is that expression of folate-HCY genes is often reduced in strains with elevated HCY levels (ApobtmlUnc, Pax3Sp, Apcmin and B6-Gli3XtJ), whereas their expression is often elevated in strains with normal HCY levels (C3HeB/FeJ-Gli3XtJ, Ptch1 and Ski)
(Table II-3). Perhaps C3HeB/FeJ-Gli3XtJ, Ptch1 and Ski mice are able to maintain normal levels by increasing the levels of transcription of these genes, despite their sensitizing mutation (Figure II-2, Table II-3).
The decreased GNMT transcript levels in Pax3Sp mice are associated with increased HCY levels, perhaps because of glycine hypomethylation (Table II-2).
GNMT (glycine N-methyltransferase) transfers the methyl group of S- adenosylmethionine to glycine (OGAWA et al. 1998). A population study showed that DNA hypomethylation is associated with elevated HCY levels (YI et al.
2000). Whether deficiency of other methyltransferases results in elevated HCY 85 levels in Pax3Sp mice as well as the relation between PAX3 and GNMT remain to be determined.
Genetic background may account for the contrasting effects on AHCY and
HCY levels. Under normal physiological conditions, the condensation of adenosine and HCY to form S-adenosylhomocysteine is favored, a reaction that decreases HCY levels (Figure I-1 {step 6}; (FINKELSTEIN 1990)). Therefore, the reduced level of AHCY in Gli3XtJ mutant mice on the C57BL/6J background may account for their increased HCY (Table II-2). By contrast, the Gli3XtJ mutation on the C3HeB/FeJ background shows increased AHCY levels and normal HCY
(Table II-2). The nature of this genetic background effect remains to be determined.
An inverse relationship between folate and HCY levels is often found
(ROSENBLATT and FENTON 2001). Increasing the availability of folate, which might result from increased in ATIC, FBP2, MTHFR and GART observed in Pax3Sp,
ApcMin, C3HeB/FeJ-Gli3XtJ; Ptch1 and Ski mice, should reduce HCY levels.
However, Pax3Sp and ApcMin mice seem incapable of adequately regulating their
HCY levels by this mechanism. Consistent with interpretation is the observation that prenatal treatment with folate corrects NTD in Pax3Sp mice (FLEMING and
COPP 1998) suggesting that these mice are folate-deficient. 86 Genes whose expression are decreased RFC compared to the B6 wild-type tmlUnc Apob Mutants with unchanged HCY levels Genes whose expression are increased MTHFR AHCY GART GLI3 SAA5 2 standard deviation) in RNA abundance in mutants compared to their in mutants deviation) in RNA abundance 2 standard ≥ SHMT TYMS GNMT AHCY GLI3 SAP Genes whose expression are decreased Mutants with elevated HCY levels (2X) P. Genes whose expression are increased ATIC INPP1 folate binding protein-2; GNMT, glycine N-methyltransferase; GLI3, transcription factor gli3; LEPR, leptin receptor; SAP, transcription factor gli3; LEPR, leptin receptor; GNMT, glycine N-methyltransferase; GLI3, folate binding protein-2; 5-phosphoribosyl-5-aminoimidazole-4-carboxamide TYMS, thymidylate synthase; ATIC, serum amyloid P-component; GART, 5- 5,10-methylenetetrahydrofolate reductase; adenosylhomocysteine hydrolase; MTHFR, transformylase; AHCY, SAA5, serum amyloid A5; RFC, reduced folate carrier; SAP, serum amyloid phosphoribosylglycineamide transformylase; Table II-3. Significant differences ( Significant differences Table II-3. controls respective wild-type in folate- are listed. RNAs that are directly involved and the type of neural tube defect phenotype The effect on HCY levels threshold that were below the two standard-deviation (Figure I-1) are shown in bold. The genes homocysteine metabolism are and are underlined. For instance, HCY levels or higher than 3-fold were also evaluated but showed changes equal C57BL/6J- expression levels are decreased in heterozygous increased and SHMT FBP2, inositol polyphosphate-1 phosphatase; SHMT, serine hydroxymethyltransferase; INPP1, control. Abbreviations: 87
Summary
Our phenotype-based approach based on measurement of HCY levels and expression profiles in mice with single gene mutations has provided the clues to the identity of novel genes and pathways that adversely affect folate and
HCY metabolism. Although the relations between reactions in folate-HCY metabolism are complex (KISLIUK 1999), making it difficult to predict the metabolic consequences of many perturbations, the consistency between elevated HCY levels, altered RNA abundance levels and metabolic expectations is provocative. The changes in gene expression are consistent with the literature but will have to be assessed in more detail by measuring additional biological markers, such as purine and thymidine levels and global methylation.
MATERIAL and METHODS
Mice
Mice were purchased from the Jackson Laboratory, except for the Ptch1 and Ski mutant and their wild-type controls. Mice were either bled upon their arrival or were maintained under SPF conditions. All mice shared the same animal room with controlled temperature, humidity, and 12 h light-dark cycle.
Mice were provided food and water ad libitum. Number of animals used per mutants and controls: C57BL/6J, n=30; C57BL/6J-Pax3Sp, n=14; C57BL/6J- 88
Apcmin, n=12; C57BL/6J-Gli3XtJ, n=7; C57BL/6J-ApobtmlUnc, n=5; C3HeB/FeJLe, n=4; C3HeB/FeJLe-Xt, n=10; Ptch1+/+, n=7; Ptch1+/-, n=5; Ski+/+, n=5; Ski+/-, n=5.
Controls for Ptch1+/- and Ski+/- consisted of their wild-type sibs Ptch1+/+ and Ski+/+ respectively, from a segregating cross.
Diets
All mice were maintained on the Harlan-Teklad LM-485 diet, except for
PTCH1 mice that were maintained on the Prolab RMH 3000 diet.
Blood and tissue samples
Blood samples were obtained from the sub-orbital sinus of virgin female mice that were 6-8 weeks old, and were collected in non-heparinized tubes. After centrifugation, serum samples were stored at -80°C. At autopsy, liver samples were placed immediately on dry ice.
Plasma versus serum test
Blood samples were obtained from the sub-orbital sinus of virgin (A/J x
C57BL/6J)F1 hybrids female mice that were 6-8 weeks old. Each sample was collected in non-heparinized tubes (serum) and heparinized tubes (plasma; 89 potassium-EDTA microvette CB300 tubes, Sarstedt). After centrifugation, serum samples were stored at -80°C.
Homocysteine and methylmalonic acid measurements
Homocysteine and methylmalonic acid levels were measured with either
GC/MS or HPLC methods. Homocysteine levels were measured with either
MS/GC or HPLC methods.
MS/GC. Mouse serum samples (100 µl - 200 µl) were diluted to 0.5 ml with PBS.
Then 1 ml of dH20 was added along with octadeuterohomocysteine and trideutero-MMA internal standards (20 µl each of 165 and 25.8 µg/ml saline solutions, respectively). The samples were reduced by the addition of 51 µl dithiothreitol (10 mg/ml in 1 mol/L NaOH) and heating at 37°C for 30 min, and then applied onto a column of AG MP-1 resin (400 mg). The column was washed successively with 9 ml of dH20, 3 ml of methanol, 1.2 ml of 0.3 mol/L acetic acid in methanol (to elute HCY), and finally, 1.2 ml of 3.6 mol/L acetic acid/0.1 mol/L HCl in 1:9 dH20:methanol (to elute MMA). The eluates were evaporated to dryness under vacuum. The HCY and MMA fractions were derivatized with 50 µl of BSTFA and 50 µl MTBSTFA, respectively (50°C for 30 min). The derivatization mixtures were then mixed in autoinjector vials and then analyzed with a HP5988 in positive ion El mode. Chromatography was on a 30 m x 0.25 mm capillary column coated with a 0.25 µm DB-1 film. Injection was in 90 splitless mode with a 1 min delay at 100°C and then ramped to 180°C at
40°C/min, then to 210°C at 5°C/min followed by a 3 min bake-out at 280°C.
Under these conditions, MMA-2TBDMS and homocysteine-3TMS elute in that order 7 min after injection and about 8 sec apart. Ions monitored were m/z 289
(Do-MMA-2TBDMS), 292 (D3-MMA-2TBDMS), 234 (Do-homocysteine-3TMS), and 238 (D4-homocysteine-3TMS).
HPLC. The HPLC method of Ubbink and Vermaak (1991) was used to measure total serum HCY levels.
Reagents for the array
Gene and EST databases were surveyed to identify sequences for genes involved in folate-HCY metabolism or in other pathways as controls. The sequence of each candidate PCR product was evaluated for motifs, repeats and closely related gene family members that might confound its specificity in expression profile experiments. Eight genes (AHCY, AMS, DHFR, LEPR,
NRAMP, PLI, RFC and TYMS) were represented by two PCR products, one for the 5’ end of the gene and the other for the 3’ end of the gene. Oligonucleotides were synthesized by Research Genetics (Huntsville, Alabama).
PCR products for expression arrays
Trizol Reagent kits (Gibco BRL) were used to prepare total RNA from fresh or frozen tissues (usually liver) from female C57BL/6J mice. cDNAs made 91 from these RNAs (Superscript II RNase H- Reverse Transcriptase; Gibco BRL) were used as template in PCR reactions. The resulting PCR products were purified and quantitated. To make the arrays, 160 ng of purified and denatured
PCR products were aliquoted in duplicate on Hybond N+ nylon membranes
(Amersham).
Hybridization probes
To make the hybridization probe, 20 µg of total RNA was radiolabelled with 32P-deoxycytosine (dCTP) in first strand cDNA synthesis reactions
(Superscript II RNase H- Reverse Transcriptase; Gibco BRL). Hybridizations were done at 65°C for 20-24 h in Church buffer after pre-hybridization at 65°C for
2-3 h. Four replicate pools from each strain or mutant were prepared by combining equal portions (by weight) of liver samples from (the same) 4-5 genetically and phenotypically identical mice. The radiolabelled cDNA pools were independently hybridized to new replicate membrane arrays. A phosphorimager
(Molecular Dynamics) was used to quantitate hybridization signals.
Data analysis
Each gene was represented twice on each array to control for the variation in background noise on the nylon membrane. In summary, the correlation 92 coefficient for the duplicate spots was calculated. If the correlation coefficient fell below 0.99, inconsistent duplicate data points were discarded. The expression values were log transformed and normalized by center mean, similar to the procedure used in the Cluster program (EISEN et al. 1998) to account for the variation in probe intensity between replicate membranes. The normalized values for the four replicate membranes were used to calculate the mean and the standard deviation. If the signal for a gene in one of the four replicates exceeded
2 standard deviations from the mean, the remaining three data points were used for the analysis. Thus, for the majority of genes, the estimate of the expression level for each gene was based on eight signals.
In more detail, the null hypothesis is that Xijk = MiGjRk(1 + Eijk), where Xijk,
Mi, Gj, and Rk are the expression level, strain or mutant effect, gene effect and replicate effect, respectively, such that the replicates are indexed separately by k
within each strain. Eijk is an error term that is identically and independently distributed for each reading. Taking the natural logarithm of both sides, we obtain
yijk = mi + gj + rk + eijk, where yijk, mi, gj and rk are the loge of Xijk, Mi, Gj and Rk,
respectively, and eijk ~ Eijk are independently and identically distributed normal
errors. To accurately estimate the gene effects, we set rk ~ dj xijk / n, where n is
the number of genes in the replicate, and letting yijk = xijk - rk, we used the model
xijk = mi + gj + eijk, where k now indexes the several identically distributed
measures for each combination of strain and gene. For each cell Cij we can
calculate the mean yij and the standard deviation s. Those s that exceeded the 93 critical value s* = 1.96 were considered to result from at least one aberrant measure. Where possible, these values are removed and the entire analysis repeated on the edited data. For the 2520 signals in the 120 = (15 x 16)/ 2 pairwise comparisons, only 4.1% had to be eliminated.
We next examined the RNA abundance data to identify outliers in each pairwise mutant versus control comparisons. To do this, the control normalized values was subtracted from the mutant normalized values across all the genes.
Then, genes falling outside of the 95% confidence interval were considered outliers.
In more detail, we inspected the cell means y1jk and y2jk for each gene j and noted those where the variation exceeds d= 1.96, as indicating significantly heterogeneous levels of gene expression in the two strains. Although we formally included strain effects in our model, we were really only interested in the interactions of gene and strain, and any main effect due to strain was removed
along with the replicate effect in defining yijk, because of the nesting structure of the experimental design. The critical values s* and d are determined from the mean squared error remaining after all gene effects and (negligible) strain effects have been removed.
Northern analysis
Gene expression levels were validated for TYMS 3' and 5' end in B6 and
B6-Pax3Sp; FBP2 in B6, B6-Pax3Sp, Ski+/+ and Ski+/-; AHCY 3' and 5' end in B6 94 and B6-Gli3XtJ. Beta actin and glyceraldehyde-3-dehydrogenase were used as controls to normalize differences in mRNA loading. Total RNA from liver was extracted using a Trizol Reagent kit (Gibco BRL) and messenger RNA was isolated with the Oligotex mRNA Midi Kit (Qiagen). The mRNA was transferred onto Hybond N+ nylon membranes (Amersham) and the hybridization was performed with the NorthernMax kit (Ambion). Probes were radiolabelled with
32P-deoxycytosine (dCTP) by PCR. A phosphorimager (445SI, Molecular
Dynamics) was used to quantitate hybridization signals.
Acknowledgements
We thank Robert E. MacKenzie and Shawn McCandless for many valuable discussions about the biochemistry of folate and HCY metabolism and
Barry Shane for the methionine synthase sequence prior to publication. This work was supported by grants from the March of Dimes Foundation and from the
National Heart, Lung and Blood Institute (HL58982) to JHN, the National Institute of Child Health and Development (HD30728) to CC, the Medical Research
Council to DSR, the Natural Science and Engineering Research Council of
Canada and the Canadian Genome Analysis and Technology Program to DS, by a grant from the Charles B. Wang Foundation to the Center for Computational
Genomics, by a grant from the Keck Foundation to the Department of Genetics, and by a Howard Hughes Medical Institute grant to Case Western Reserve 95
University School of Medicine. BC was a recipient of a fellowship from the Royal
Victoria Hospital Research Institute. DS is a fellow of the Canadian Institute for
Advanced Research. 96
CHAPTER III
PARALLEL CHANGES IN METABOLITE LEVELS AND EXPRESSION
PROFILES IN FOLATE-DEFICIENT AND CROOKED TAIL MUTANT MICE 97
AUTHORS
Sheila Ernest, Michelle Carter, Haifeng Shao, Angela Hosack, Natalia Lerner,
Clemencia Colmenares, David S. Rosenblatt, Yoh-Han Pao, M. Elizabeth Ross, and Joseph H. Nadeau
REFERENCE
Will be submitted for publication
ABSTRACT
Anomalies in homocysteine (HCY) and folate metabolism are associated with common birth defects and adult diseases, several of which can be suppressed with dietary folate supplementation (see Chapter One). Crooked-tail
(Cd) mutant mice are an important model of folate-responsive neural tube defects (NTDs). Because dietary folate supplementation suppresses NTDs in Cd mutant mice, we investigated modulation of these pathways in Cd mice.
Metabolic assays and expression profiles were used to survey the status of the
HCY, folate and other pathways in +/+, Cd/+ and Cd/Cd mice on control, folate- supplemented and -deficient diets. Folate suppressed NTDs through a mechanism that did not directly involve modulating HCY levels. Instead, parallel changes in metabolite and expression profiles in folate-supplemented Cd/Cd mice and folate-deficient +/+ and Cd/+ mice suggest that homozygosity for the 98
Cd mutation leaded to a defect in the utilization of intracellular folate. Expression profiles for a panel of mutant mice with phenotypes associated with anomalies in
HCY and folate metabolism revealed two major clusters, one that included all of the folate-responsive mutants and another with mutants whose responsiveness to folate has not been tested. Based on these profiles, we predicted and verified that NTDs in Ski mutants are folate-resistant.
INTRODUCTION
Birth defects such as NTDs and adult diseases such as vascular disease, and certain cancers are often associated with anomalies in HCY and folate metabolism(CARMEL 2001; MUDD et al. 2001; ROSENBLATT and FENTON 2001).
Disease pathogenesis is thought to result directly from the adverse effects of elevated HCY levels and perhaps indirectly from redox reactions that are mediated through homocyteine, cysteine and glutathione metabolism (HEINECKE
2001). HCY and folate levels are usually inversely related (ROSENBLATT and
FENTON 2001) and dietary supplementation with folate typically restores normal folate levels, lowers HCY levels, and dramatically reduces the risk for NTDs
(GROUP 1991), colon cancer (GIOVANNUCCI et al. 1998) and ischaemic heart disease (SCHNYDER et al. 2001). In many cases however, folate supplementation 99 fails to suppress NTDs. By defining the functional relations between mutant genes that cause NTDs and the metabolism of HCY and folate and by characterizing the genetic, molecular and physiological basis for responsiveness to folate supplementation, new ways to treat folate-resistant cases may be identified.
The laboratory mouse is an important model to understand the basis of folate-responsiveness. Locations and frequencies of the various neural tube closure defects are similar in humans and mice, suggesting that the pathogenesis of NTDs may be similar in these two organisms (VAN ALLEN et al.
1993). Crooked-tail (Cd) mice are an important model of the human folate- responsive NTDs. Homozygosity for the Cd mutation leads to early lethality, exencephaly, or to small body weight in the rare survivors; a crooked tail is found in Cd/+ heterozygotes (MORGAN 1954; CARTER et al. 1999). Penetrance of the crooked tail phenotype is incomplete and depends on genetic background
(CARTER et al. 1999). A diet supplemented with folic acid reduces the severity and recurrence risk of exencephaly in homozygous Cd/Cd mice (CARTER et al.
1999). However, the genetic, cellular, molecular and developmental basis for the birth defects in Cd mutant mice are not known (CARTER et al. 1999).
Because disruption of HCY and folate metabolism has been associated with NTDs(STEEGERS-THEUNISSEN et al. 1995) and Cd mice are folate- responsive(CARTER et al. 1999), we investigated modulation of these pathways in 100
Cd mice. Serum metabolite levels and hepatic expression profiles of genes in the
HCY and folate pathways were monitored in wild-type (+/+), heterozygous (Cd/+) and homozygous (Cd/Cd) females fed a control, folate-deficient or
-supplemented diet. Surprisingly, the Cd mutation did not significantly affect HCY levels, suggesting that folate might correct NTDs through another mechanism.
Cluster analysis showed that metabolite and gene expression profiles of folate- supplemented Cd/Cd mice were similar to those of folate-deficient +/+ and Cd/+ mice, suggesting that homozygosity for the Cd mutation causes a defect in the utilization of intracellular folate that does not affect HCY levels. Finally, cluster analysis showed that folate-responsive mouse mutants have distinct expression profiles from those whose responsiveness to folate has not been tested. Based on these clustered expression profiles, we predicted and subsequently verified that one of the untested mutants in the second cluster had phenotypes that are resistant to folate treatments, demonstrating that expression levels of genes in these profiles can predict folate responsiveness. 101
RESULTS
Metabolites
We surveyed serum levels of HCY and folate as key metabolic indicators of the status of the HCY and folate pathways in wild-type (+/+), heterozygous
(Cd/+) and homozygous (Cd/Cd) females fed a control, folate-deficient or folate- supplemented diet. As expected, HCY levels were elevated only in mice that were raised on the folate-deficient diet (Figure III-1). Contrary to expectations however, Cd mice had normal HCY levels as well as normal levels of extracellular and intracellular folate, regardless of diet (Figure III-2 and III-3), suggesting that the beneficial effects of folate treatments were not mediated through a reduction in HCY levels.
We also measured cysteine and glutathione levels as indicators of the status of the trans-sulfuration pathway, through which HCY is catabolized
(MOSHAROV et al. 2000; MUDD et al. 2001). GSH is a major cellular antioxidant and maternal glutathione may play an important role in protecting the embryo during early development by scavenging endogenous reactive oxygen species
(LEGGE and SELLENS 1991). GSH levels were consistently increased in Cd/Cd mice and tended to increase with decreasing amounts of dietary folate (Figure III-
4). Elevated serum levels in Cd/Cd mice suggest that GSH stores in the liver, which contribute largely to GSH in the blood (MEISTER 1988), may be depleted. A 102 10 mg, 8.1 Cd/+ female mice on control, female mice 4.0, n=15; ± 2.9, n=14; wild-type 4 mg, 2.9, n=14; wild-type 4 ± Cd standard deviation. ± 0 mg, 13.2 Cd/+ 16.5, n=15; ± 1.2, n=13; wild-type 10 mg, 9.3 1.2, n=13; wild-type 10 ± 4mg, 4.7 Cd/Cd levels of wild-type, heterozygous and homozygous and homozygous heterozygous levels of wild-type, 0.8,n=13; 3.3, n=6. To assess differences in mean HCY levels, the Newman-Keuls multiple in mean HCY levels, the Newman-Keuls 3.3, n=6. To assess differences ± ± 4mg, 4.5 homocysteine 10 mg, 8.1 Cd/+ Cd/Cd 0.9, n=18; ± 3.9, n=15; Figure III-1. Serum Figure III-1. and folate-deficient diets folate-supplemented (µmol/L) for: wild-type 0 mg, 26.5 Mean homocysteine levels 5.1 ± as mean post-test was used. Values are expressed comparison one-way ANOVA 103 ± 16.4, ± 10 mg, 58.3 Cd/+ 4.1, n=5; ± female mice on control, folate- female mice standard deviation. Cd ± 2.4, n=16; wild-type 4 mg, 72.2 2.4, n=16; wild-type 4 ± 0 mg, 5.3 Cd/+ 4.1, n=15; 14.4, n=14; wild-type 10 mg, 54.5 14.4, n=14; wild-type ± ± 4mg, 57.8 Cd/Cd 18.4, n=6. To assess differences in mean folate levels, the Newman-Keuls multiple differences in mean folate levels, the Newman-Keuls 18.4, n=6. To assess ± levels of wild-type, heterozygous and homozygous and homozygous heterozygous levels of wild-type, 12.5, n=4; ± folate 10 mg, 81.3 Cd/Cd 4mg, 60.9 Cd/+ Figure III-2. Serum Figure III-2. diets supplemented and folate-deficient for: wild-type 0 mg, 7.0 Mean folate levels (pmol/ml) n=8; 21.5, n=5; as mean post-test was used. Values are expressed comparison one-way ANOVA 104 ± 0 mg, 0.7 Cd/+ 0.7, n=16; ± 2.0, n=14; wild-type 10 mg, 5.5 2.0, n=14; wild-type 10 ± female mice on control, folate- female mice Cd 4mg, 5.1 Cd/Cd 4.2, n=7. To assess differences in mean tissue folate 4.2, n=7. To assess differences 2.5, n=10; ± ± 4mg, 6.8 10 mg, 6.0 Cd/+ Cd/Cd 5.0, n=10; ± 2.4, n=10; ± levels of wild-type, heterozygous and homozygous and homozygous heterozygous levels of wild-type, 10 mg, 4.1 Hepatic folate Cd/+ 0.4, n=17; wild-type 4 mg, 8.0 0.4, n=17; wild-type 4 2.9, n=10; Figure III-3. Figure III-3. diets supplemented and folate-deficient for: wild-type 0 mg, 1.0 levels. Mean folate levels (pmol/mg of protein) Mean liver tissue folate ± ± mean was used. Values are expressed as multiple comparison one-way ANOVA post-test levels, the Newman-Keuls standard deviation. 105 ± ± control, on 10 mg, 3.4 mice Cd/+ female
Cd 0.5, n=14; ± 0.8, n=16; wild-type 4 mg, 4.0 0.8, n=16; wild-type 4 ± standard deviation. ± homozygous homozygous 0 mg, 4.1 and Cd/+ 0.7, n=13; ± heterozygous heterozygous 1.4, n=13; wild-type 10 mg, 3.2 1.4, n=13; wild-type 10 ± wild-type, 4mg, 4.7 of Cd/Cd levels
1.2, n=7. To assess differences in mean glutathione levels, the Newman-Keuls multiple in mean glutathione levels, the Newman-Keuls 1.2, n=7. To assess differences ± 0.7, n=13; ± glutathione 10 mg, 4.6 4mg, 3.4 Serum Cd/+ Cd/Cd III-4. Figure and folate-deficient diets folate-supplemented (µmol/L) for: wild-type 0 mg, 4.5 Mean glutathione levels 0.9, n=18; 0.7, n=15; as mean post-test was used. Values are expressed comparison one-way ANOVA 106 folate-deficient diet reduces hepatic antioxidant concentrations (HENNING et al.
1997b) and promotes hepatic oxidative stress (HUANG et al. 2001). Perhaps the
Cd mutation adversely affects antioxidant functions in the liver in a similar manner. CYS plays numerous roles such as substrate for protein synthesis and probably as an extracellular antioxidant (UELAND et al. 1996). However, no obvious patterns were observed for CYS levels among genotypes and diets
(Figure III-5).
Expression studies
Hepatic expression profiles were used to identify genotypes and diets (as perturbations) that clustered together and to identify the patterns of metabolites and gene expression (as consequences of perturbations) that distinguished each cluster. Two main clusters of strains (A and B) and two main clusters of genes and metabolites (1 and 2) were found (Figure III-6). Cluster A included wild-type
+/+ and Cd/+ mice on the control and folate-supplemented diets (4 and 10 mg
FA/kg, respectively), whereas Cluster B included mutant Cd/Cd homozygotes on the control and folate-supplemented diets and wild-type +/+ and Cd/+ mice on the folate-deficient diet (0 mg FA/kg). Cluster 1 included genes in various metabolic pathways (HCY and folate metabolism, inflammatory response, lipid transport, phosphoinositide signaling, and glucose transport and metabolism), whereas Cluster 2 included the remainder of the genes except those involved in 107 0.3, folate- ± 0.6, n=5; ± control, on 10 mg, 5.7 mice Cd/+ female
Cd 0.7, n=17; wild-type 4 mg, 4.9 0.7, n=17; wild-type 4 0.6, n=5; ± ± 0 mg, 4.4 homozygous homozygous Cd/+ and standard deviation. ± 0.5, n=13; ± heterozygous heterozygous 1.1, n=13; wild-type 10 mg, 4.5 1.1, n=13; wild-type 10 ± wild-type, 4mg, 4.4 of Cd/Cd levels
0.8, n=4; cysteine ± 0.4, n=7. To assess differences in mean cysteine levels, the Newman-Keuls multiple comparison one- in mean cysteine levels, the Newman-Keuls 0.4, n=7. To assess differences ± Serum 4mg, 4.5 III-5. 10 mg, 3.5 Cd/+ Figure diets supplemented and folate-deficient for: wild-type 0 mg, 4.5 Mean cysteine levels (µmol/L) n=7; Cd/Cd was used. Values are expressed as mean way ANOVA post-test 108
Figure III-6. Hierarchical clustering of metabolite and gene expression profiles The expression data were analyzed with hierarchical clustering (EISEN et al. 1998) using the Multiple Viewer Experiment (MeV) software package developed by TIGR. Expression levels were normalized and mean-centered across strains and genes/metabolites. The hierarchical analysis was statistically supported by jack-knife resampling with 1000 iterations, with resampling conducted on both strains and genes separately. Branches of resulting trees are color-coded to denote the percentage of times a given node was supported over the resampling trials. Two main clusters of strains (A and B) and two main clusters of genes and metabolites (1 and 2) were found. Legend: red, expression greater than the mean; green, expression lower than the mean; black, expression unchanged; grey, no available data. 109 phosphoinositide signaling and in glucose transport and metabolism. Jack-knifing was used to test whether particular pathways were critical for defining the clusters. Sequential elimination of each pathway did not affect the fundamental structure or composition of the various clusters, suggesting that clustering did not depend on results for particular pathways (data no shown).
Folate-deficient wild-type mice and folate-supplemented Cd homozygous mutant mice share molecular features that are characteristic of a ‘methyl trap’.
Both showed significantly increased MTHFR and SHMT RNA levels and significantly reduced MTR RNA levels. Assuming that RNA and protein levels are correlated, these changes in gene expression profiles typically result in a ‘methyl trap’, which is usually caused by deficiency of MTR and perhaps by increases in
SHMT enzyme activity (HERBIG et al. 2002). The methyl trap results in accumulation of 5-methyl-THF, an increase in HCY levels and MTHFR enzyme activity, and a decrease in S-adenosylmethionine levels (HERBIG et al. 2002). The methyl trap changes the production of individual folate species that compose the folate pools and may not affect levels of total folate (HORNE et al. 1989).
Functional folate deficiency combined with normal plasma folate levels is not unusual; children with Down Syndrome have normal plasma folate levels and their intracellular folate deficiency is caused by a methyl trap and decreased HCY plasma levels (POGRIBNA et al. 2001). 110
To test whether similarities among expression profiles that might be diagnostic for particular genetic or metabolic conditions, we compared expression profiles of the various Cd mice with those of previously studied
(ERNEST et al. 2002) mouse mutants, Apob, Gli3, Pax3, Ptch and Ski (Figure III-
7). Interestingly, mice whose phenotypic defects are folate-responsive grouped together: one cluster contained Cd (CARTER et al. 1999), Apc (SONG et al. 2000a) and Pax3 (FLEMING and COPP 1998) mutant mice, whereas the other cluster contained Apob, Ptch and Ski, whose responsiveness to folate is unknown.
These results predict that Apob, Ptch and Ski mutant mice would be folate- resistant, or perhaps respond to folate treatments in a manner that is functionally distinct from the response of mutants in the first group.
To test this hypothesis, we treated Ski mutant mice with folic acid during the first half of gestation. Ski mutants were tested in both the C57BL/6J (N9-N10 backcross generations) and 129S6 (N4 backcross generation) backgrounds to determine whether folic acid treatment reduced the incidence of either the clefting or exencephaly phenotypes observed in these strains (COLMENARES et al.
2002). As predicted, folic acid had no effect on the incidence of either strain- dependent phenotype in Ski mutant homozygotes (Table III-1). A broader survey of NTDs and other mutants is needed to test the generality of these results. A key application of gene expression surveys is developing profiles that are diagnostic for particular disease conditions or treatment responses. Results with 111
Figure III-7. Hierarchical clustering of metabolite and gene expression profiles of Cd and other previously analyzed mutant mice (ERNEST et al. 2002)or see Chapter Two. The expression data were analyzed with hierarchical clustering using the same software and methods as described in figure III-6. Cluster A contains mutants whose defects are folate-responsive. Legend: red, expression greater than the mean; green, expression lower than the mean; black, expression unchanged; grey, no data available. 112 Facial cleft (%) 13 (87) 7 (78) 0 0 homozygous mutant mice. homozygous Ski Exencephaly (%) 1 (11) 13 (92) 8 (100) 1 (6) -/- (%) Ski 9 (22) 14 (24) 8 (21) 15 (24) 41 58 39 Total Offspring 63 -null mutants (two-sided Fisher’s exact test: B6, P = 0.8, and 129, P = 1.0) Fisher’s exact test: B6, P = 0.8, and 129, -null mutants (two-sided Ski 129 129 + Folic acid C57BL/6J C57BL/6J + Folic acid Table III-1. Effect of folic acid treatment on craniofacial and neural tube defects in and neural tube on craniofacial Effect of folic acid treatment Table III-1. Folic acid treatment had no effect on the treated with folic acid as described in Methods. Pregnant females were phenotypes of 113
Cd and Ski and the other mutants suggest that development of these tests for folate responsiveness may be feasible.
DISCUSSION
We studied the physiological and genetic modulation of folate and HCY metabolism in Crooked-tail mutant mice, an important model of folate-responsive
NTDs. The expectation from various epidemiological and clinical studies in humans was that elevated HCY levels is an important risk factor for NTDs
(STEEGERS-THEUNISSEN et al. 1995). The ability of folate to suppress NTDs in Cd homozygotes (CARTER et al. 1999) predicted that these mice would have elevated HCY and reduced folate levels. Contrary to expectations however, Cd mice had normal HCY levels, regardless of diet, as well as normal levels of extracellular and intracellular folate, suggesting that the beneficial effects of folate treatments were not mediated through a reduction in HCY levels, at least in
Cd mutant mice. The molecular and physiological basis for folate-protection in Cd mice is therefore perplexing.
The strikingly similarities among the metabolite and gene expression profiles in folate-supplemented Cd/Cd mice and in folate-deficient +/+ and Cd/+ 114 mice suggest that Cd homozygotes have a defect in folate utilization despite high folate levels in diet, serum and liver. Cd/Cd cells either do not have adequate access to folate or are not able to utilize the available folate effectively. Perhaps by increasing folate levels with dietary supplementation, a threshold is reached that restores normal neural tube development. Alternatively, perhaps folate utilization differs between fetal metabolism, when correction occurs, and adults, when measurements are typically made.
Expression profiles correctly predicted that the Ski mutant mice NTDs would either be resistant to folate treatment or at least respond to folate treatments in a different manner from the mutants in the folate-responsive expression profile cluster, i.e. Cd, Pax3 and Apc. A broader survey of NTDs and other mutants is needed to test the generality of these results. A key application of gene expression surveys is developing profiles that are diagnostic for particular disease conditions or treatment responses. Results with Ski and the other mutants suggest that development of these tests for folate responsiveness may be feasible. 115
MATERIAL AND METHODS
Crooked tail mutants: Study population, diets and metabolite measures
Female mice were used because in humans maternal HCY as well as fetal metabolism are risk factors for fetal neural tube defects (FLEMING and COPP
1998). Virgin females were used to control for possible effects of pregnancy on
HCY levels (KANG et al. 1986). An additional reason for focusing on females is that among human anencephalies and certain mouse exencephalies, such as
Cd, females are more frequently affected than males (CARTER et al. 1999). Mice were raised from conception on control (4 mg folic acid (FA)/kg chow), folate- supplemented (10 mg FA/kg chow), or folate-deficient (0 mg FA/kg chow) defined diets (P.J. Noyes, NH). Homozygous Cd/Cd mice on the folate-deficient diet were not analyzed because of their severely reduced survival.
Cd genotypings
The polymorphism detection protocol was previously described (CARTER et al. 1999). The D6Mit135, D6Mit111 (Invitrogen, CA; formerly Research Genetics,
AL) and Ms6hm3 markers were used for genotyping. 116
Blood and tissue samples
After deep anesthetization of 6-8 week old virgin female mice, blood samples were obtained via cardiac puncture and were collected in non- heparinized tubes. After centrifugation, serum samples were stored at -80°C. At autopsy, liver samples were placed immediately on dry ice.
Metabolite measurements
The HPLC method of Ubbink and Vermaak (UBBINK et al. 1991) was used to measure total serum homocysteine, glutathione and cysteine levels. Total folate serum levels were measured by microbiological enzymatic assay (HORNE
1997; MOLLOY and SCOTT 1997). For hepatic folate levels, tissues were boiled in an extraction buffer (2% sodium ascorbate, 0.2M b-mercaptoethanol, 0.05M
HEPES and 0.05M CHES, pH 7.85), homogenized, centrifuged for 10 minutes and treated with 0.5 mg/ml of carboxypeptidase G (from Pseudomonas species;
Sigma, MO) for 6 hours (sample to enzyme ratio, 4:1)(HUANG et al. 2001). The samples were then boiled and centrifuged 5 minutes each and folate levels were measured by microbiological assay (HORNE 1997; MOLLOY and SCOTT 1997).
Liver protein content was measured by BCA (bicinchoninic acid) protein assay reagent kit (Pierce, IL) 117
Ski mutants: Study population and folate treatment
Pregnant females were treated daily with folic acid (5 mg/kg body weight, by i.p. injection) from E0.5 (day of plug) to E 9.5. Fetuses were collected at
E16.5-E18.5, examined for craniofacial phenotypes, and genotyped by Southern blotting (BERK et al. 1997).
Expression arrays
Arrays of gene-specific PCR products were used to evaluate the impact of the mutation and diet on the RNA abundance of genes involved in homocysteine- folate metabolism as well as a panel of genetically and biochemically independent genes as controls. The methods that we used to make the arrays, gene-specific reagents, hybridization conditions and data analysis were described previously21 (Refer also to Chapter II). Genes that were included on the array are listed in Table II-1 of Chapter Two.
Data analysis
We used to hierarchical clustering (EISEN et al. 1998) to analyze the expression data with the Multiple Viewer Experiment (MeV) software package developed by The Institute of Genome Research was used. The hierarchical 118 analysis was statistically supported by jack-knife resampling with 1000 iterations, with resampling conducted on both strains and genes separately.
ACKNOWLEDGMENTS
This work was supported by grants from the NIH (HL58982 to JHN,
HD30728 to CC, and NS34998 to MER), and by gift from the Charles B. Wang
Foundation to YHP and JHN). 119
CHAPTER IV
NUTRIGENES, FUNCTIONAL GENOMICS AND SYSTEMS BIOLOGY 120
AUTHORS
Sheila Ernest, Michelle Carter, Angela Hosack, David Rosenblatt, Elizabeth Ross and Joseph H. Nadeau
REFERENCE
Journal of Nutrition 133, 4267-4268 (2003)
ABSTRACT
Traditionally, the classic reductionist approach attributes functions to individual genes. For instance, this has involved the analysis of motifs or the amino acid sequences of single gene products. It is unclear how the products of particular collections genes act together to provide higher order functionality in health and disease. To address this higher order problem, the function of collections of genes, as opposed to “one gene at a time” has to be studied. 121
INTRODUCTION
A model system is needed to test systems biology. We used homocysteine- folate (HCY-folate) metabolism as an ideal model system because it contains many useful attributes: 1) it is relevant to common and important human diseases, such as neural tube defects (NTD; (GROUP 1991; STEEGERS-
THEUNISSEN et al. 1991), vascular disease (VERHOEF and STAMPFER 2001) and cancer (MAYER et al. 1996); 2) its core pathway is known (Figure I-1); 3) it contains intermediate phenotypes that are measurable, such as metabolite levels and enzyme activity; 4) it is influenced by environmental factors, such as diet
(CARMEL 2001; ROSENBLATT and FENTON 2001); 5) it contains targets for pharmacologic interventions, such as anticancer drugs (ROSENBLATT and FENTON
2001). A key point in testing systems biology is to monitor consequences or responses of specific causes or perturbations. We have been developing a platform based on collections of NTD mutant mice and dietary supplementation in combination with HCY-folate metabolism to begin studying complex biological systems. 122
RESULTS AND DISCUSSION
We first assessed the effect of genetic perturbations on HCY-folate metabolism using NTD mice with a series of single gene mutations (PAX3, GLI3,
APOB, PTCH and SKI; see Chapter Two). We monitored the response of HCY- folate metabolism to these genetic perturbations by measuring serum metabolite levels and gene expression. Increased serum homocysteine levels and altered expression profiles of HCY-folate metabolism genes were detected in mice with partial protein deficiencies in the WNT and hedgehog signal transduction (GLI3 and PAX3) and lipid transport (APOB) (ERNEST et al. 2002). This systems approach revealed functional relations between pathways which otherwise were not evident and may contribute to the pathogenesis of birth defects.
We then studied the effect of genetic and physiological perturbations on HCY- folate metabolism using a single gene mouse mutant model, Crooked tail (Cd), whose developmental defects are suppressed with a folate-supplemented diet
(CARTER et al. 1999); see Chapter Three). Crooked tail mice model the human folate-responsive NTD, permitting an assessment of the ways in which dietary supplementations suppress the adverse developmental outcomes of single gene mutations. A combination of metabolic assays and expression profiles were used to survey wild-type, heterozygous Cd/+ and homozygous Cd/Cd mice on control, folate-supplemented and folate-deficient diets. The mutation and the diet were 123 used as genetic and physiological perturbations, respectively. Surprisingly, the
Cd mutation had no significant effect on HCY levels, suggesting that folate might correct the defect through a mechanism that may not involve HCY per se.
Interestingly, cluster analyses revealed that gene expression and metabolite profiles of homozygous Cd/Cd mice parallel that of folate-deficient mice, suggesting that homozygosity for the Cd mutation might be a functional folate deficiency without alterations in HCY levels. These studies illustrate the ways in which gene expression and metabolic assays can be combined with perturbations of physiological and developmental systems to define the complex networks of functional interactions that relate genetic mutations with disease outcomes.
In summary, the HCY-folate metabolism has many useful attributes for studies on systems biology in mouse models of human disease and provides a unique opportunity to integrate genetics, expression and metabolite profiles and complex phenotypes. 124
CHAPTER V
INHERITANCE OF HOMOCYSTEINEMIA IN A/J AND C57BL/6J MICE:
DIET, GENDER, PARENTAL AND GENETIC EFFECTS 125
AUTHORS
Sheila Ernest, Angela Hosack, William E. O’Brien, David S. Rosenblatt and
Joseph H. Nadeau
REFERENCE
Will be submitted for publication
ABSTRACT
Hyperhomocysteinemia is associated with various birth defects and adult diseases and affects 5% of the general population. Hyperhomocysteinemia is a complex trait that depends on several genetic and environmental factors. The extent to which genetic factors control homocysteine (HCY) levels in healthy individuals is unclear. Laboratory mice are valuable models for dissecting the genetic and environmental controls of homocysteinemia. We assessed the inheritance of normal homocysteine (HCY) levels in two inbred strains, A/J and
C57BL/6J (B6), under controlled physiological conditions and assessed the relative importance of diet, gender, parental and genetic effects. Diet affected mean HCY levels whereas gender affected both mean and variance of HCY levels. Moreover, gender of the parents influenced homocysteinemia in reciprocal
F1 hybrids suggesting that maternal effects modulate HCY levels. Finally, gene- diet interactions affected heritability of normal HCY levels. These studies showed 126 that each of these factors contribute to normal homocysteinemia and provided important clues in the understanding of human homocysteinemia.
INTRODUCTION
Homocysteinemia and human diseases
Five percent of the general population shows significantly elevated HCY levels in the blood (KANG and WONG 1996). Hyperhomocysteinemia is an independent risk factor for neural tube defects (NTDs) (GROUP 1991; STEEGERS-
THEUNISSEN et al. 1991), vascular disease (BOUSHEY et al. 1995; WELCH and
LOSCALZO 1998; WALD et al. 2002) and Alzheimer disease (SESHADRI et al. 2002), and is often associated with diseases, such as colon cancer (KATO et al. 1999) and Down Syndrome (JAMES et al. 1999; BOSCO et al. 2003). Although it is unclear whether elevated HCY levels are a cause or consequence of disease, a meta-analysis demonstrated that hyperhomocysteinemia causes cardiovascular disease (WALD et al. 2002).
Treatment of hyperhomocysteinemia and related diseases
Normal HCY blood levels range from 5 to 12 µmol/L (PIETRZIK and
BRONSTRUP 1997) and hyperhomocysteinemia is classified in three categories, 127 mild (12-30 µmol/L), moderate (30-100 µmol/L), and severe (>100 µmol/L) (KANG et al. 1992; PIETRZIK and BRONSTRUP 1997). In this paper, homocysteinemia is defined as the level of HCY in the blood, hyperhomocysteinemia, as significantly elevated HCY levels and normal HCY homocysteinemia, as normal levels. Folate
and vitamins B6 and B12 are often used to treat hyperhomocysteinemia, although folate treatment alone is often sufficient (COLLABORATION 1998; VAN DER GRIEND et al. 2000). Moreover, folate reduces the risk for neural tube defects (GROUP
1991; HONEIN et al. 2001) and colon cancer (GIOVANNUCCI et al. 1995;
GIOVANNUCCI et al. 1998) and several clinical trials are underway to assess the effect of folate on vascular disease (CLARKE and ARMITAGE 2000). Folate and
vitamins B6 and B12 are cofactors in HCY-folate metabolism, whose disruption often leads to elevated HCY levels and increased risk for various birth defects and adult diseases (ROSENBLATT and FENTON 2001).
Factors affecting homocysteinemia
Several physiological and genetic factors affect homocysteinemia.
Physiological factors include age, where HCY increases with increasing age, gender (women tend to have lower levels than men), and diet (vitamin cofactors
such as folate and vitamin B6 and B12 intake usually reduces HCY levels) (MINER et al. 1997). Deficiencies in genes involved in HCY and folate metabolism result in hyperhomocysteinemia: these genes include cystathionine b-synthase (CBS 128
(MUDD et al. 1964; BOERS et al. 1985b)), methionine synthase (MTR (WATKINS and ROSENBLATT 1989)), methionine synthase reductase (MTRR (WATKINS and
ROSENBLATT 1989)), methylenetetrahydrofolate reductase (MTHFR; (KANG et al.
1988; ENGBERSEN et al. 1995; FROSST et al. 1995; KLUIJTMANS et al. 1996)), and glutamate carboxypeptidase II (GCPII; (DEVLIN et al. 2000)). Heterozygosity and homozygosity for MTHFR deficiencies (FROSST et al. 1995; KLUIJTMANS et al.
1996), as well as heterozygosity for GCPII (DEVLIN et al. 2000)) lead to mild hyperhomocysteinemia whereas heterozygosity and homozygosity for CBS deficiencies lead to moderate and severe hyperhomocysteinemia, respectively
(MUDD et al. 1964; BOERS et al. 1985b). In addition to increasing HCY levels,
CBS deficiencies increase the risk for vascular disease (MUDD et al. 1985;
WILCKEN and WILCKEN 1998) and a common MTHFR mutation (C677T) increases the risk for NTDs (CHRISTENSEN et al. 1999), cardiovascular disease (WALD et al.
2002), and paradoxically provides protection from colon cancer (CHEN et al.
1996; MA et al. 1997).
Homocysteinemia and mouse models
Little is known about normal HCY levels in mice. A survey of fourteen different inbred strains with no obvious phenotypes showed that HCY levels ranged from 3.1 to 7.0 mmol/L (Figure V-1). Several single gene mutant mice with 129 0.1; 0.2; ± ± standard ± 0.2; SWR/J, 3.8 ± 0.3; BALB/byJ, 6.0 ± 0.1; DDY/JeL, 3.5 ± 0.1; SWXL-4, 5.6 0.6. Values are expressed as mean 0.6. Values are expressed ± ± 0.4; NOD/LtJ, 3.3 ± 0.4; ABP/Le, 7.0 ± 0.1; CBA/CaJ, 5.6 ± 0.1; C3H/HeJ, 6.6 ± 0.1; C57BL/6J, 5.6 ± mol/L) among 14 different inbred strains 14 different inbred mol/L) among m 0.3; A/J, 5.4 ± 0.1; DBA/2J, 6.3 ± Figure V-1. HCY levels ( Figure V-1. (µmol/L) for: EL/SuZ, 3.2 Mean homocysteine levels NOR/LtJ, 4.2 CBA/J, 5.9 ANOVA Newman-Keuls multiple comparison one-way assess differences in mean HCY levels, the error of the mean. To size, n, and P values are indicated. post-test was used. Sample 130 phenotypes similar to those associated with hyperhomocysteinemia in humans also show disruptions of HCY-folate metabolism (ERNEST et al. 2002). Increased
HCY levels are found in certain mouse models of NTDs, i.e. Pax3, Gli3 and Apob heterozygous mutants, and the Apc colon cancer model heterozygous mutants, compared to their wild-type controls (ERNEST et al. 2002). As in humans, folicacid as well as other compounds suppress birth defects in several mouse models of
NTDs. Folic acid suppresses defects in Cart1 (cartilage homeoprotein 1; (ZHAO et al. 1996)), Cd (Crooked tail; (CARTER et al. 1999)), Cited2 (CBP/p300 interacting transactivators with glutamic acid (E)/(D)-rich C-terminal domain; (BARBERA et al.
2002)), Folbp1 and Folbp2 (folic acid-binding protein 1 and 2; (PIEDRAHITA et al.
1999)), folic acid and thymidine suppress defects in Pax3Sp (PAX3 transcription factor; (FLEMING and COPP 1998)) mutants, inositol (GREENE and COPP 1997) but not folate (TRAN et al. 2002) suppresses defects in ct (curly-tail) mutants, and methionine, but not folate suppresses NTDs in Axd (axial defects) mice (ESSIEN
1992). Folic acid supplementation also suppresses the number of colonic polyps in Apc mice (SONG et al. 2000a) and reduces the number of atherosclerotic lesions in Apoe null mice, a mouse model of atherosclerosis, (HOFMANN et al.
2001).
Mouse models of hyperhomocysteinemia were created by targeted disruptions of the CBS (WATANABE et al. 1995), MTR (SWANSON et al. 2001) and
MTHFR (CHEN et al. 2001) proteins. As in humans, hyperhomocysteinemia 131 results from CBS, MTR and MTHFR deficiencies. Compared to wild-type control mice, homozygous and heterozygous Cbs mutant mice have a ~40-fold and 2- fold increase in HCY levels respectively (WATANABE et al. 1995), a ~10-fold and
~1.5-fold increase respectively in homozygous and heterozygous Mthfr mutant mice (CHEN et al. 2001), and a 1.7-fold increase heterozygous Mtr mutant mice whereas homozygous Mtr mutant mice are embryonic lethal (SWANSON et al.
2001).
Homocysteine and folate metabolism
HCY and folate pathways are metabolically linked and their regulation is complex. However, HCY and folate levels are usually inversely related
(ROSENBLATT and FENTON 2001). HCY, a sulfur-containing amino acid, is remethylated through the trans-methylation pathway or irreversibly catabolized through the trans-sulfuration pathway (MUDD et al. 2001). It contributes directly to methylation of DNA, proteins and lipids and indirectly to glucose metabolism and citric acid cycle. Folate metabolism manages the major pool of single carbon units. Both pathways participate in purine and pyrimidine biosynthesis and amino acid metabolism (ROSENBLATT and FENTON 2001).
Homocysteinemia is a complex trait that is affected by several genetic and environmental factors and consequently its genetic control in humans is difficult 132 to dissect. Genetic factors contribute to mild, moderate (WALD et al. 2002) and severe (MUDD et al. 2001) hyperhomocysteinemia, whereas their role in the control of normal homocysteinemia is unclear (REED et al. 1991; BERG et al.
1992; CESARI et al. 2000). The laboratory mouse is an important and useful animal model to study the genetic control of normal homocysteinemia, independent of disease. We investigated HCY metabolism using two inbred strains of mice, A/J and C57BL/6J (B6), on two diets with different amounts of folate. These two inbred strains are analogous two humans with distinct genotypes in contrast to average effects in families and populations (HOIT and
NADEAU 2001). A unique aspect of our study was the analysis of inheritance not only of mean HCY levels but also of variance itself. Gender and diet affected mean and variance of homocysteinemia to different extents in A/J and B6 mice.
In addition, gender of the parents affected HCY mean and variance differently in reciprocal F1 hybrids suggesting a possible role for maternal effects in homocysteinemia. Physiological factors affected heritability with genetic contributions being greatest in males on the ‘higher folate’ diet and lowest in both females and males on the ‘lower folate’ diet. These studies revealed important clues about the inheritance of normal homocysteinemia and provided the first evidence of genetic control of HCY variance. 133
RESULTS
The levels of HCY in serum were measured in A/J and C57BL/6J (B6) inbred strains as well as in their reciprocal F1 hybrids and F2 progeny. HCY levels were analyzed in males and females on two different diets, ‘lower folate’
(Harlan Teklad 7013) and ‘higher folate’ (Lab Diet 5010). For simplicity, folate content was used to distinguish the two diets, although they differ in the amounts of many other ingredients (See Appendix 1). We analyzed patterns of variation in both mean and variance of HCY levels.
Variation in mean HCY levels
Gender effects. HCY levels differed significantly between females and males in both A/J and B6 strains that were raised on either diet (Table V-1). However, contrary to the general trend in humans (NYGARD et al. 1995), females had significantly higher levels than males in both strains (Table V-1).
Diet effects. Only in B6 female and male mice that HCY significantly differed between the 2 diets. HCY levels were more elevated on the ‘lower’ versus ‘higher folate’ diets in both genders (Table V-1).
Strain effects. HCY levels were compared between A/J and B6 mice to study the effects of genetic background. A/J females and males both had significantly 134 lower HCY levels than B6 mice of the respective gender. These differences were found regardless of the diet on which the mice were raised (Figures V-2, V-3, V-
4, V-5; Tables V-3, V-4, V-5 and V-6).
Parental effects. To test the effects of the gender of the parents, HCY levels were compared between reciprocal F1 hybrid mice, i.e. ABF1 (where A refers to the
A/J mother) versus BAF1 (where B refers to the B6 mother). Gender of the parents did not affect HCY levels in F1 females raised on either diet, i.e. HCY levels in ABF1 = BAF1 females. However, HCY levels differed in males, i.e. HCY levels in ABF1 ≠ BAF1 (Tables V-3 and V-5). HCY levels were significantly lower in ABF1 than BAF1 males regardless of diet (Tables V-4 and V-6).
Dominance effects. HCY levels for the parental strains and F1 hybrids were compared to determine dominance of homocysteinemia; comparisons are presented according to gender and diet effects.
Females on the ‘lower folate’ diet Because ABF1 and BAF1 females had similar HCY levels (see above), their data were pooled. F1 hybrids had significantly higher HCY levels than A/J mice and significantly lower levels than
B6 females demonstrating that homocysteinemia was inherited as a semi- dominant trait (Table V-3). 135
Males on the ‘lower folate’ diet Dominance of HCY levels in hybrid males was parent-dependent. ABF1 hybrid males had similar levels to A/J males and significantly lower levels than B6 males, demonstrating that the low HCY trait, which was characteristic of the A/J parent, was inherited in a dominant manner.
In contrast, BAF1 hybrid males had significantly higher and lower levels than A/J and B6 males respectively, showing that homocysteinemia in these mice was inherited in a semi-dominant manner (Table V-4).
Females on ‘higher folate’ diet Because ABF1 and BAF1 females had similar levels, data were pooled. Interestingly, F1 hybrid females had lower levels than both B6 and A/J males, indicating that homocysteinemia was under-dominant
(Table V-5).
Males on ‘higher folate’ diet Dominance of HCY levels in hybrid males was parent-dependent. ABF1 hybrid males had lower levels than both B6 and A/J males, indicating that homocysteinemia was under-dominant. By contrast, BAF1 hybrids had similar levels to B6 males and significantly higher levels than A/J, indicating that the high HCY trait in BAF1 males was inherited in a dominant manner from the B6 parent (Table V-6).
Heritability. Heritability is the proportion of genetic variance that affects a trait and uses data from the two parental strains, F1 hybrids and F2 population for its estimation. When parental gender influenced HCY levels, which was the case for males on both diets (Tables V-4 and V-6), heritability was estimated separately. 136
Genetic factors contributed significantly to homocysteinemia in males on the
‘higher folate’ diet but more modestly in both females and males raised on the
‘lower folate’ diet (Table V-7).
Variation in HCY variance
We then studied variance in HCY levels as a measure of the ability of distinct genetic systems to modulate perturbations of HCY metabolism (GIBSON and WAGNER 2000; RUTHERFORD 2000).
Gender effects. HCY variance differed significantly between females and males only in mice that were raised on the ‘lower folate’ diet. (Table V-2). In general, females had significantly more variable HCY levels than males.
Diet effects. Diet had no effect on HCY variance in either A/J or B6 mice (Table
V-2).
Strain effects. To study the effects of genetic background, HCY variance was compared between A/J and B6 mice. When raised on the ‘lower folate’ diet, A/J females had less variable HCY levels than B6 females (Figure V-2 and Table V-
3). Genetic background did not affect HCY variance in males raised on the ‘lower 137 folate’ diet (Figure V-3, Table V-4) or in both females and males raised on the
‘higher folate’ diet (Figures V-4 and V-5, Tables V-5 and V-6).
Parental effects. To test the effects of parental gender on variance in HCY levels in hybrid mice, reciprocal ABF1 versus BAF1 mice were compared. In F1 females on the ‘lower folate’ diet, parental effects affected variance with HCY levels being more variable in ABF1 than in BAF1 females (Figure V-2, Table V-
3). Not surprisingly, gender of the parents did not affect HCY variance in males raised on the ‘lower folate’ diet or in females and males on the ‘higher folate’ diet
(Figures V-3, V-4 and V-5; Tables V-4, V-5 and V-6).
Dominance effects. To measure dominance effects on variance, HCY variances for the parental strains and F1 hybrids were compared. Dominance of HCY variance was parent-dependent for F1 females raised on the ‘lower folate’ diet. In
ABF1 females, variance was under-dominant because hybrids had significantly less variable levels than B6 and A/J females (Table V-3). In BAF1 females, the low variance trait was inherited in a dominant manner from the A/J parent with 138 0.1) 0.2) 0.1) 0.2) ± ± ± ± = 0.05/4 = 0.0125. a M (2.4 M (2.8 H (2.4 H (2.8 > > = > 0.2) 0.3) 0.1) 0.2) ± ± ± ± ‘Higher folate’ diet F (3.9 P = 2E-21 (43 F; 45 M) F (4.4 P = 7E-13 (46 F, 49 M) Males L (2.5 ns (41 L; 45 H) L (4.0 P = 1E-12 (55 L; 49 H) test was used and to account for multiple test was used and to t 0.1) 0.2) 0.2) 0.3) ± ± ± ± M (2.5 M (4.0 H (3.9 H (4.4 > > = > 0.2) 0.4) 0.2) 0.4) ± ± ± ± mol/L) ± 1.96 * standard error of the mean and values are indicated in parentheses. error of the mean and values are indicated mol/L) ± 1.96 * standard ‘Lower folate’ diet F (3.6 P = 5E-10 (44 F; 41 M) F (5.9 P = 1E-14 (60 F; 55 M) Females L (3.6 ns (44 L; 43 H) L (5.9 P = 1E-08 (60 L; 46 H) m : HCY levels were compared between females and males. For instance, when raised on the ‘lower folate’ between females and males. For instance, : HCY levels were compared To assess differences in mean HCY levels, Student’s To assess differences : HCY levels were compared between mice raised on the ‘lower folate’ and those raised on the ‘higher folate’ between mice raised on the ‘lower folate’ : HCY levels were compared Gender effects(Females vs Males) A/J B6 Diet effects (‘Lower’ vs ‘Higher’ folate) A/J B6 HCY levels are expressed as mean ( HCY levels are expressed gender or diet comparisons. P, and sample size, n, are indicated below the The test significance, ns, non-significant. males; L, ‘lower folate’; H, ‘higher folate’; Symbols: F, females; M, Table V-1. Gender and diet effects on mean HCY levels Gender and diet effects Table V-1. Gender effects higher HCY levels than A/J males. diet, A/J females had significantly Diet effects folate’ diet. lower HCY than females on the ‘higher females on the ‘lower folate’ diet had significantly diet. For instance, B6 Data analysis: to a Bonferroni correction, 4 comparisons, significance levels were subjected hypothesis testing with 139 = 0.05/4 = a M (0.2) M (0.5) H (0.2) H (0.5) = = = = ‘Higher folate’ diet F (0.3) ns (43 F; 45 M) F (1.2) ns (46 F, 49 M) Males L (0.2) ns (41 L; 45 H) L (0.7) ns (55 L; 49 H) M (0.2) M (0.7) H (0.3) H (1.2) > > = = ‘Lower folate’ diet F (0.7) P = 4E-04 (44 F; 41 M) F (1.9) P = 2E-04 (60 F; 55 M) Females L (0.7) ns (44 L; 43 H) L (1.9) ns (60 L; 46 H) : HCY variance was compared between females and males and variance values are indicated in between females and males and variance : HCY variance was compared To assess differences in variance of HCY levels, Levene’s test was used and to account for multiple in variance of HCY levels, Levene’s test was To assess differences : HCY variance was compared between mice raised on the ‘lower folate’ and those raised on the ‘higher folate’ between mice raised on the ‘lower : HCY variance was compared B6 Gender effects(Females vs Males) A/J B6 Diet effects (‘Lower’ vs ‘Higher’ folate) A/J Data analysis: to a Bonferroni correction, 4 comparisons, significance levels were subjected hypothesis testing with gender or diet comparisons. P, and sample size, n, are indicated below 0.0125.Test significance, H, ‘higher folate’. Symbols: F, females; M, males; L, ‘lower folate’; Table V-2. Gender and diet effects on HCY variance Gender and diet effects Table V-2. Gender effectsmore levels in A/J females were significantly when raised on the ‘lower folate’ diet, HCY parentheses. For instance, size, n, are indicated below the gender comparison. A/J males. Test significance, P, and sample variable than those of Diet effects diet had A/J females that were raised on ‘lower folate’ are indicated in parentheses. For instance, diet and variance values raised on the ‘higher folate’ diet. similar variance to those 140 1.96 * standard error of the mean. 1.96 * standard error ± mice that were raised on the ‘lower folate’ were raised on the ‘lower mice that female mol/L) in parental strains, F1 and F2 strains, F1 and mol/L) in parental m Figure V-2. HCY levels ( Figure V-2. diet mean sample size (n) and lower numbers represent Upper numbers represent 141 1.96 * standard error of the mean 1.96 * standard error ± mice that were raised on the ‘lower folate’ diet were raised on the ‘lower mice that male mol/L) in parental strains, F1 and F2 strains, F1 and mol/L) in parental m Figure V-3. HCY levels ( Figure V-3. mean sample size (n) and lower numbers represent Upper numbers represent 142 1.96 * standard error of the mean 1.96 * standard error ± mice that were raised on the ‘higher folate’ were raised on the ‘higher mice that female mol/L) in parental strains, F1 and F2 strains, F1 and mol/L) in parental m Figure V-4. HCY levels ( Figure V-4. diet mean sample size (n) and lower numbers represent Upper numbers represent 143 1.96 * standard error of the mean 1.96 * standard error ± mice that were raised on the ‘higher folate’ were raised on the ‘higher mice that male mol/L) in parental strains, F1 and F2 strains, F1 and mol/L) in parental m Figure V-5. HCY levels ( Figure V-5. diet mean sample size (n) and lower numbers represent Upper numbers represent 144 B6 A/J = = ns BAF1 ns Not applicable BAF1 BAF1 Test significance, P, and
BAF1 (0.9) < B6 (1.9) < B6 A/J < < Variance A/J (0.7) P = 0.002 (44 A/J; 60 B6) ABF1 (0.2) P= 6E-04 (29 ABF1; 30 BAF1) ABF1 ABF1 P = 0.002 ABF1 P = 5E-06 Under-dominant that were raised on the ‘lower folate’ diet: raised on the ‘lower that were females 0.3) ± 1.96 * standard error of the mean. ± 0.4) ± mol/L) m BAF1 (4.2 = B6 (5.9 < 0.1) ± 0.2) ± = 0.05/4 = 0.0125. 0.2, n=59) a A/J B6 ± > < (pooled data) Mean A/J (3.6 P = 2E-18 (44 A/J; 60 B6) ABF1 (4.4. ns (29 ABF1; 30 BAF1) F1 (4.3 F1 P = 1E-05 (59 F1; 44 A/J) F1 P = 5E-12 (59 F1; 60 B6) Semi-dominant : mean HCY levels or variance were compared between the parental strains and F1 progeny. Values variance were compared between the parental : mean HCY levels or : mean HCY levels and variance were compared between ABF1 and BAF1 (reciprocal) F1 hybrid mice. variance were compared between ABF1 and : mean HCY levels and : mean HCY levels and variance were compared between A/J and B6 and the values are indicated in variance were compared between A/J and : mean HCY levels and = 0.05/2 = 0.025 or a test was used; differences in variance of HCY levels was assessed by Levene’s test To account for multiple test was used; differences in variance of HCY t Trait (F1 vs parental) A/J B6 Student’s or 4 comparisons, significance levels were subjected to a Bonferroni hypothesis testing with 2 (when F1 were pooled) correction, ( Average HCY levels are expressed as mean sample size, n, are indicated below comparisons. Strain effect (A/J vs B6) Parental effect (ABF1 vs BAF1) Dominance effect similar for ABF1 and BAF1 mice. similar for ABF1 and BAF1 Dominance effects levels intermediate between A/J and B6 females For instance, F1 hybrid females had are indicated in parentheses. in a semi-dominant manner. To assess differences in mean HCY levels, indicating the mean HCY trait was inherited Table V-3. Variation in mean HCY levels and in variance for levels and in variance Variation in mean HCY Table V-3. effects. strain, parental and dominance Strain effects and less variable HCY levels than B6 females. A/J females had significantly lower mean parentheses. For instance, Parental effects progeny had similar mean HCY levels, whereas parentheses. For instance, the reciprocal F1 Values are indicated in were F1 data were pooled when HCY levels were more variable than those of BAF1 females. levels of ABF1 females 145 BAF1 (0.4) = B6 (0.7) = A/J B6 = = (pooled data) Variance A/J (0.2) ns (41 A/J; 55 B6) ABF1 (0.3) ns (30 ABF1; 29 BAF1) F1 (variance = 0.4, n=59) F1 ns F1 ns Not applicable A/J B6 > < 0.2) ± 0.2) BAF1 BAF1 P = 0.0001 BAF1 P = 7E-07 Semi-dominant ± BAF1 (3.1 < B6 (4.0 < 0.2) ± 0.1) ± A/J B6 = < P = 3E-13 A/J or low mean HCY trait is dominant Mean A/J (2.5 B6) P = 1E-18 (41 A/J; 55 ABF1 (2.7 P = 0.01 (30 ABF1; 29 BAF1) ABF1 ABF1 ns ABF1 (ABF1 vs BAF1) Dominance (F1 vs parental) A/J B6 Trait Strain Effect (A/J vs B6) Parental effect Table V-4. Variation in mean HCY levels and in variance for males that were raised on the ‘lower folate’ diet: raised on the ‘lower for males that were levels and in variance Variation in mean HCY Table V-4. effects. strain, parental and dominance See Table V-3. 146 BAF1 (0.5) = B6 (1.2) = A/J B6 = = (pooled data) Variance A/J (0.3) ns (43 A/J; 46 B6) ABF1 (0.3) ns (29 ABF1; 30 BAF1) F1 (variance = 0.4, n=59) F1 ns F1 ns Not applicable 0.3) ± 0.3) ± BAF1 (3.6 = B6 (4.4 See Table V-3. < 0.2) ± 0.2) ± 0.1, n=59) A/J B6 ± < < (pooled data) Under-dominant Mean A/J (3.9 P= 0.007 (43 A/J; 46 B6) ABF1 (3.5 ns (29 ABF1; 30 BAF1) F1 (3.5 F1 P= 3E-03 F1 P = 7E-6 Dominance (F1 vs parental) A/J B6 Trait Strain Effect (A/J vs B6) Parental effect (ABF1 vs BAF1) Table V-5. Variation in mean HCY levels and in variance for females that were raised on the ‘higher folate’ diet: were raised on the ‘higher for females that levels and in variance Variation in mean HCY Table V-5. effects. strain, parental and dominance 147 BAF1 (0.2) = B6 (0.5) = BAF1-BAF2 0.58 A/J B6 = = (pooled data) Variance A/J (0.2) ns (45 A/J; 49 B6) ABF1 (0.1) ns (30 ABF1; 30 BAF1) F1 (variance = 0.4, n=60) F1 ns F1 ns Not applicable ‘higher folate’ diet Not done ABF1-ABF2 0.61 0.2) ± A/J B6 > = 0.2) ± BAF1-BAF2 0.10 BAF1 P = 6E-05 BAF1 ns B6 or high HCY mean trait is dominant BAF1 See Table V-3. BAF1 (2.9 < B6 (2.8 < 0.1) ± 0.1) ± A/J B6 < < Mean A/J (2.4 B6) P = 0.007 (45 A/J; 49 ABF1 (2.0 BAF1) P = 3E-11 (30 ABF1; 30 ABF1 ABF1 P = 1E-04 ABF1 P = 4E-09 Over-dominant ‘lower folate’ diet 0.40 ABF1-ABF2 0.49 Females Males B6 Dominance (F1 vs parental) Table V-7. Heritability of homocysteinemia in males, heritability was determined separately for the ABF1-ABF2 and Because parental effects influenced HCY levels BAF1-BAF2 populations. (ABF1 vs BAF1) A/J Table V-6. Variation in mean HCY levels and in variance for males that were raised on the ‘higher folate’ diet: raised on the ‘higher for males that were levels and in variance Variation in mean HCY Table V-6. effects. strain, parental and dominance Strain Effect (A/J vs B6) Parental effect 148 hybrids having similar variance as A/J females and significantly less variable levels than B6 females (Table V-3).
DISCUSSION
Both genetic and environmental factors control hyperhomocysteinemia.
However, little is known about the relative impact of these factors on normal HCY levels among individuals who do not have hyperhomocysteinemia. Therefore, we studied the inheritance of homocysteinemia in mouse models to gain insight into the genetic and environmental controls of variation in normal HCY levels. Using
A/J and B6 inbred mouse strains we conducted systematic studies to characterize influences on patterns of HCY variation. In our mouse models, normal homocysteinemia was a complex trait that depended on gender, diet and parental effects and both the mean and variance in HCY levels were affected.
These results provided crucial information that will guide the design of future experiments in mice and humans. 149
Variation in mean HCY levels
Females tended to have greater HCY levels than males
In humans, hormonal effects account for gender differences, where higher estrogen levels are often associated with decreased HCY levels (ZMUDA et al.
1997; GIRI et al. 1998; MORRIS et al. 2000). In A/J and B6 mice, however, HCY levels were usually higher in females than males (Table V-1). This trend was also found in other inbred strains that were analyzed in our laboratory (unpublished data). Therefore, estrogen may have contrasting effects on mean HCY levels in humans and mice.
Diet affected homocysteinemia
In humans, dietary supplementation with folic acid usually lowers HCY levels (COLLABORATION 1998; VAN DER GRIEND et al. 2000), but some individuals do not respond to vitamin supplementation (VAN DER GRIEND et al. 2000). We also observed differences in responses to folate in our mouse models (Table V-1).
The ‘higher folate’ diet reduced HCY levels to a greater extent in female and male B6 mice, whereas it did not affect HCY levels in A/J females and males
(Table V-1). The A/J mouse might correspond to humans whose HCY levels are resistant to the beneficial effects of folate supplementation whereas the B6 mouse to humans whose HCY levels are folate-responsive. 150
Parental effects influenced homocysteinemia only in hybrid male progeny
In humans, the HCY phenotype of the mother is often implicated in homocysteinemia of the fetus; mothers with elevated HCY levels are more likely to give birth to a child who also has elevated HCY levels (MALINOW et al. 1998).
In humans, however, it is often difficult to establish effects of parental gender on the phenotype of the progeny in humans. Our mouse models provide valuable information about parental effects on normal homocysteinemia. Parental effects strongly influenced HCY levels in F1 males but not in F1 females (Tables V-3, V-
4, V-5 and V-6), suggesting that genomic imprinting, as well as many other maternal effects, may account for normal homocysteinemia in the F1 hybrid males. Imprinting is known to be sensitive to folate levels (WOLFF et al. 1998;
WATERLAND and JIRTLE 2003), which is the source of CH3 for DNA methylation
(ROSENBLATT and FENTON 2001), but direct tests of imprinting on homocysteinemia in humans and mice remain to be done.
Heritability of homocysteinemia
Twin studies have been used to estimate heritability of homocysteinemia in populations of healthy humans but results are conflicting. Two studies found that normal HCY levels are heritable (~0.72) (REED et al. 1991; BERG et al. 1992) and another study suggests that genetic factors do not contribute significantly to variation in HCY levels (CESARI et al. 2000). In our studies, gene-diet interaction strongly influenced heritability estimates; heritability was greater for mice that 151 were raised on the ‘higher folate’ than on the ‘lower folate’ diet (Table V-7).
Interestingly, mice raised on the ‘lower folate’ diet showed a greater environmental variance than those raised on the ‘higher folate’ diet, suggesting that metabolic pathways that are stressed with low levels of essential factors, such as folate, are more vulnerable to stochastic influences. Similar factors, such as the diet, might complicate estimation of heritability in humans.
Variation in HCY variance
Perhaps our most striking observation was the evidence for genetic control of variance of HCY levels in genetically identical inbred mice, which were expected to have comparable variances. Females had more variable HCY levels than males (Table V-2) and hormonal changes associated with the estrus cycle may explain these differences. In humans, estrogen affects mean HCY levels
(ZMUDA et al. 1997; GIRI et al. 1998; MORRIS et al. 2000), however it is unknown whether it can also have an effect on HCY variance.
HCY levels were more variable in B6 than A/J females raised on the ‘lower folate’ diet (Table V-3). Perhaps B6 females had difficulty maintaining stable HCY levels on the ‘lower folate’ diet and, consequently they may be at greater risk for
HCY-related disease than A/J females on the ‘lower folate’ diet. B6 females may represent individuals whose HCY levels vary greatly on a ‘lower folate’ diet and 152 may be at greater risk to homocysteinemia-related diseases than other individuals whose HCY levels do not vary as much. This type of variability is not unusual in humans and genetic factors, which are generally silent, may increase certain disease risks under specific environmental conditions (RUTHERFORD
2000). The same situation could apply to homocysteinemia.
Gender of the parents affected variance of F1 females raised on the ‘lower folate’ but not the ‘higher folate’ diet. The low HCY variance trait was inherited in a dominant manner from the A/J parent in both reciprocal F1 female progeny suggesting that greater variance under ‘low folate’ and high HCY levels conditions may expose some fetuses to risk for birth defects because of stochastic factors.
MATERIAL AND METHODS
Mice
Inbred mice were purchased from the Jackson Laboratory (Bar Harbor,
Maine) and maintained under SPF conditions. All mice shared the same animal room with controlled temperature, humidity, and 12 h light-dark cycle. Mice were provided food and water ad libitum. 153
Diets
Inbred strain survey: mice were maintained on the 5010 LabDiet
(Indiana). Inheritance of homocysteinemia in A/J and B6 strains: mice were maintained either on the 5010 LabDiet (6.0 mg folic acid/Kg chow; Laboratory autoclavable rodent diet 5010; Indiana) or 7013 Harlan Teklad (1.83 mg folic acid/Kg chow; NIH-31 modified mouse/rat sterilizable diet; Wisconsin). The progeny were raised from conception on either diet.
Mouse breeding
Parental strains
Separate mating pairs of A/J and B6 mice were bred on each diet to produce progeny that were tested for gender, diet and strain effects.
F1 progeny
Crosses between A/J and B6 mice produced F1 hybrids. Crosses between
A/J females and B6 males produced ABF1 mice and crosses between B6 females and A/J males produced reciprocal BAF1 mice.
F2 progeny
Brother-sister matings of F1 hybrid mice produced the F2 progeny.
Crosses between ABF1 produced ABF2 mice and crosses between BAF1 mice produced reciprocal BAF2 progeny. 154
Blood samples
Blood samples were obtained from the retro-orbital sinus of male and virgin female mice that were 6-8 weeks old, and were collected in non- heparinized tubes. After centrifugation, serum samples were stored at -80°C.
Virgin females were used to control for possible effects of pregnancy on HCY levels (KANG et al. 1986; STEEGERS-THEUNISSEN et al. 1997).
Homocysteine measurements
Inbred strain survey: total serum HCY levels were measured by HPLC method. Inheritance of homocysteinemia in A/J and B6 mice: total serum HCY levels were measured by HPLC and tandem mass spectroscopy methods. Both methods consist of reducing protein-HCY disulfide bonds prior analysis. The concentration of HCY was determined by HPLC or tandem mass spectroscopy.
HPLC
HPLC method of Ubbink and Vermaak (UBBINK et al. 1991) was used.
Tandem mass spectroscopy
Tandem mass spectroscopy method of McCann and co-authors was used
(MCCANN et al. 2003).
Tandem mass spectroscopy vs HPLC: correction of HCY levels
Preliminary data showed a consistent bias in which samples measured by
HPLC gave higher HCY levels than those measured by tandem mass 155 spectroscopy (TMS). The following method was used to adjust differences in techniques to pool the data. Inbred mouse strains, whose HCY levels were measured by both methods, were selected to control for genetic and environmental factors because HCY levels should be similar within each group.
HCY levels from 490 samples that were collected from four different strains under various conditions were used. Of the 490 samples, 282 were measured by
HPLC and the remainder by TMS. The mean HCY levels were calculated for each group and the correction factor was applied as followed:
mHPLC (5.849 mmol/L) - mTMS (4.238 mmol/L) = 1.611 mmol/L
mHPLC: mean HCY levels across all samples measured by HPLC
mTMS: mean HCY levels across all samples measured by TMS
HCY levels differed by 1.611 mmol/L with a correlation factor (r2) of 0.85 between the two methods after correction. Then, each sample measured by HPLC had its
HCY level subtracted by 1.611 mmol/L and pooled with samples measured by
TMS.
Data analysis
Statistical analysis
Inbred strain survey: the Newman-Keuls multiple comparison one-way
ANOVA post-test was used to assess differences in mean HCY levels.
Inheritance of homocysteinemia in A/J and B6 mice: to assess differences in 156 mean HCY levels, Student’s t test was used and significance levels were subjected with the appropriate Bonferroni correction factor to account for multiple testing. To assess differences in variance, Levene's test was used (MANLY 1997).
Heritability estimates (WRIGHT 1968; DIETRICH et al. 1993)
Environmental variance (VE) = (VA/J+VB6+VF1)/3
Total variance (VT) = VF2
Genetic variance (VG) = VT-VE
Heritability = VG/VE
Abbreviations: V = variance, F1 = F1 population, F2 = F2 population 157
CHAPTER VI
QTL ANALYSIS OF HOMOCYSTEINEMIA IN A/J AND C57BL/6J MICE USING C57BL/6J-Chr(i)A/J CHROMOSOME SUBSTITUTION STRAINS
157 158
AUTHORS
Sheila Ernest, Annie E. Hill, Jonathan B. Singer, Angela Hosack, William O’Brien,
David S. Rosenblatt, Eric S. Lander, Joseph H. Nadeau
ABSTRACT
In A/J and C57BL/6J (B6) mice, normal homocysteinemia is a complex genetic trait that depends on gender, diet and parental effects. To characterize the genetic control of normal homocysteinemia, chromosome substitution strains
(CSSs) derived from A/J and B6 mice were used to detect quantitative trait loci
(QTL). Because this phenotypic analysis is not based on prior expectations of gene function, genes from HCY-folate metabolism as well as other pathways might be detected. Survey of normal homocysteinemia of the B6-Chr(i)A/J CSS panel, where each mouse carries a single substituted chromosome from the A/J donor strain onto a uniform background of the B6 host strain, identified a total of
30 QTLs that influenced mean and variance of HCY levels in females and males raised on a ‘lower folate’ diet, and males on a ‘higher folate’ diet. Moreover, novel genetic interactions between A/J and B6 mice were found. These results reveal the complex interplay among genes, gender, diet and parental effects on
158 159 homocysteinemia and may explain why studies of homocysteinemia are so difficult in humans.
INTRODUCTION
Homocysteinemia is a complex trait in A/J and B6 mice
In humans, homocysteinemia is a complex trait that is affected by several genetic and environmental factors. Genetic factors contribute to mild, moderate
(WALD et al. 2002) and severe (MUDD et al. 2001) hyperhomocysteinemia, whereas their role in the control of normal homocysteinemia is unclear (REED et al. 1991; BERG et al. 1992; CESARI et al. 2000), perhaps because of gene-diet interactions. The inheritance of homocysteinemia was assessed in mouse models to gain insight into the genetic and environmental controls of variation in normal HCY levels (see Chapter Five). Using A/J and B6 inbred mouse strains, systematic studies were conducted to characterize influences of various genetic and non-genetic on patterns of HCY variation. These two genetically defined mouse strains are equivalent to two randomly selected individuals from human populations as opposed to average effects across populations (HOIT and NADEAU
2001). Gender and diet affected both mean and variance of HCY levels to different extents in A/J and B6 mice (Chapter Five, Tables V-1 and V-2). In addition, gender of the parents affected HCY means and variances (Chapter
Five, Tables V-3, V-4, V-5 and V-6) differently in F1 hybrids suggesting a
159 160 possible role for maternal effects in homocysteinemia. As in humans, normal homocysteinemia is a complex trait that depends on gender, diet and parental effects in this mouse model.
QTL analysis of complex traits
QTL analysis is an approach for detecting genes, or quantitative trait loci
(QTL), that contribute to complex traits. Mouse chromosome substitution strains
(CSSs) is a powerful mapping paradigm to analyze both Mendelian and multigenic traits. The CSS approach consists of substituting a whole chromosome of a host strain with the corresponding one of a donor strain.
Because there are 19 autosomes, 2 sex and 1 mitochondrial chromosomes in the mouse genome, the complete panel consists of 22 strains ((NADEAU et al. 2000);
Figure VI-I). The genome scan consists of phenotyping each CSS and deviation from the host strain indicates presence of QTL on that chromosome (NADEAU et al. 2000; BELKNAP 2003). CSSs are more powerful than the traditional F2 intercrosses to analyze complex traits (Table VI-1). However, a combination of different mapping panels is most effective in detecting QTLs (NADEAU et al. 2000;
BELKNAP 2003).
160 161 2004) et al. INGER 2000; S et al. ADEAU (N
of strain B is i (where chromosome A ) i Chr( substituted with that of strain A). the substituted chromosome. Because the substituted chromosome. 2 sex and 1 there are 19 autosomes, in the mouse mitochondrial chromosomes genome, 22 CSS lines can be made: B- Figure VI-1. Engineering chromosome substitution strains (CSSs) substitution Engineering chromosome Figure VI-1. are formed substituted with that of a donor strain. They a whole chromosome of a host strain is CSSs are strains in which two different from an initial cross between by 10 backcross inbred strains followed onto the generations from the donor strain is recipient strain. The recipient except for inbred throughout its genome 162
Table V1-1. Advantages and limitations of CSS versus F2 intercross mapping (NADEAU et al. 2000; BELKNAP 2003; SINGER et al. 2004)
CSS F2 Advantages Limitations Genome Scan - Phenotyping - Phenotyping - No genotyping - Genotyping of ~100 loci for at least 300 animals Mouse background Inbred Genetically heterogenous QTL detection Detects many strong and Detect primarily strong weak QTLS QTLs Number of generations 3-4 10 for fine resolution map with congenic strains Limitations1 Advantages Genome scan - Synteny - 20-50 cM resolution - Does not distinguish - Distinguishes multiple multiple QTLs QTLs
1 Limitations can quickly be resolved with fine-resolution mapping based on crosses between a CSS and the host strain 163
QTL mapping of homocysteinemia
Because homocysteinemia is a complex genetic trait where both HCY mean and variance differ between A/J and B6 mice (see Chapter Five), CSSs rather than intercross mapping are the more effective method for dissecting genetic factors controlling normal HCY levels. Moreover, because F2 mice are genetically and phenotypically heterogeneous, QTL analysis of the trait variance of homocysteinemia would not have been possible with the intercross method.
Two traits of normal HCY levels, mean and variance, were evaluated. The genome scan was performed for females and males on the ‘lower folate’ diet, and for males on the ‘higher folate’ diet using the B6-Chr(i)A/J CSSs (SINGER et al.
2004). This study revealed a total of 30 QTLs influencing mean and variance of serum HCY levels of B6 mice. Fine-mapping was initiated for females on the
‘lower folate’ diet to localize the QTL on chromosome 17 that was responsible for the reduction in both mean and variance of HCY levels. Using congenic strains, the QTL affecting mean HCY levels was mapped to a ~14 cM interval that contained two genes involved in HCY-folate metabolism, cystathionine b- synthase and glycine N-methyltransferase. The QTL affecting variance HCY levels has to date reached inconclusive results due to limited sample size. This phenotype-based approach revealed that, in addition to hyperhomocysteinemia, normal HCY levels can be controlled by genetic factors, which possibly involve genes from HCY-folate metabolism or other biological pathways. 164
RESULTS
B6 mice had significantly higher (Chapter Five, Tables V-3, V-4, V-5 and
V-6) and usually more variable (Chapter Five, Table V-3) levels than A/J mice.
The QTL analysis with B6-Chr(i)A/J CSSs was used to find QTLs that control HCY levels in B6 and A/J mice.
CSS genome scan
The genome scan consisted of measuring serum levels of HCY for each
B6-Chr(i)A/J CSS. A phenotypic difference between a CSS and the B6 host strain indicated the presence of at least one QTL on the substituted chromosome.
Genome scans were done for females and males that were raised on the ‘lower folate’ diet (Harlan Teklad 7013), and males that were raised on the ‘higher folate’ diet (Lab Diet 5010). Patterns of variation in mean and variance of HCY levels were analyzed.
Females on the ‘lower folate’ diet
HCY levels were surveyed for 16 CSSs (Figures VI-2A and VI-2B); work on the remaining five strains is in progress. Four QTLs were detected on three different chromosomes (Chrs): QTLs on Chrs 9 and 19 affected the trait mean 165
(Figure VI-2A), whereas QTLs on Chr 17 affected both the trait mean and variance (Figures VI-2A and VI-2B). The QTL on Chr 19 significantly reduced mean HCY levels in B6 mice, without however reaching the low levels of A/J mice.
Interestingly, the QTL on Chr 9 increased mean HCY levels above that in
B6, implying that interactive effects between at least one gene on Chr 9 and genes on other chromosomes influence homocysteinemia. QTLs on Chr 17 reduced not only the mean but also variance of B6 HCY levels. There are no known HCY-folate genes on Chr 19 raising the possibility that genes from other biological pathways affect normal HCY levels in females on the ‘lower folate’ diet.
Males on the ‘lower folate’ diet
HCY levels were surveyed for 17 CSSs (Figures VI-3A and 3B); work on the remaining five strains is in progress. Seven QTLs were detected on six different chromosomes: QTLs on Chrs 3, 10, 18 and X affected the trait mean
(Figure VI-3A), the QTL on Chr 4 affected the trait variance (Figure VI-3B), and
QTLs on Chr 8 affected both the traits mean and variance (Figures VI-3A and
3B). Surprisingly, these significant QTLs increased, instead of decreased, mean and variance of HCY levels of B6 mice (Figure VI-3A and 3B). Because there are no known HCY-folate genes on Chr 18 and X, the genome scan revealed novel genes that affect normal HCY levels in males on the ‘lower folate’ diet. 166 that were raised on the ‘lower folate’ diet raised on the ‘lower that were females mol/L) of CSS mol/L) of CSS m HCY levels ( HCY levels Mean = 0.05/21 = 0024. test was used. To differences. mean. Sample size, n, and P values are indicated for significant HCY levels, Student’s t account for multiple hypothesis testing with 21 CSSs, significance levels were subjected to a Bonferroni correction, a Values are expressed in mean ± 1.96 * standard error of the CSS that showed a significant difference from B6. To assess differences in mean Figure VI-2A: Figure VI-2A: Mean HCY levels of each CSS were compared to those of the B6 host strain. Underlined strain numbers indicate the 167 = 0.05/21 = 0024. P values are = 0.05/21 = 0024. P values a that were raised on the ‘lower folate’ diet that were raised on females mol/L) of CSS m are shown under a different representation. Variance of HCY levels of each CSS was representation. Variance of HCY levels are shown under a different analysis of HCY levels ( analysis of HCY levels Variance Figure VI-2B. Same data asdifference indicate the CSS that showed significant the B6 host strain. Underlined strain numbers compared with that of Figure VI-2A test was used. To account for multiple hypothesis in variance of HCY levels, Levene’s from B6. To assess differences correction, significance levels were subjected to a Bonferroni testing with 21 CSSs, differences. indicated for significant 168 that were raised on the ‘lower folate’ diet raised on the ‘lower that were males mol/L) of CSS mol/L) of CSS m HCY levels ( HCY levels Mean = 0.05/22 = 0023. test was used. To significance levels were subjected to a Bonferroni correction, a Values are expressed in mean ± 1.96 * standard error of the mean. Sample size, n, and P values are indicated for significant differences. t account for multiple hypothesis testing with 22 CSSs, Figure VI-3A: Figure VI-3A: the host strain. Underlined strain numbers indicate CSS were compared with those of the B6 Mean HCY levels of each CSS that showed significant difference from B6. To assess differences in mean HCY levels, Student’s 169 = 0.05/22 = 0023. P values are = 0.05/22 = 0023. P values a that were raised on the ‘lower folate’ diet raised on the ‘lower that were males mol/L) of CSS mol/L) of CSS m are shown under a different representation. Variance of HCY levels of each CSS was representation. Variance of HCY levels are shown under a different analysis of HCY levels ( analysis of Variance Figure VI-3B. Figure VI-3B. Same data asdifference indicate the CSS that showed significant the B6 host strain. Underlined strain numbers compared with that of Figure VI-3A test was used. To account for multiple hypothesis in variance of HCY levels, Levene’s from B6. To assess differences correction, significance levels were subjected to a Bonferroni testing with 22 CSSs, differences. indicated for significant 170
Males on the ‘higher folate’ diet
HCY survey of 15 CSSs showed unexpected results: every line had significantly higher mean HCY levels than the B6 host strain (Figure VI-4A).
Additionally, QTLs on Chrs 10, 15, 16 and 17 significantly increased the variance of B6 (Figure VI-4B), although the progenitor strains, A/J and B6, had similar variances (Chapter Five, Figure V-5; Table V-6). All mouse records were reviewed for anomalies and showed that all males were kept under similar physiological conditions, i.e. similar diet and husbandry, and in addition sample collection and testing of B6 control and various CSSs were done at different occasions. Because there are no known HCY-folate genes on Chrs 15, 19, X and
Y, these data suggest that novel genes affected normal homocysteinemia in B6 males that were raised on the ‘higher folate’ diet. Work on the remaining seven strains is in progress.
Fine-mapping
Fine-mapping was undertaken to determine the location of QTL(s) on Chr
17 that significantly decreased mean and variance of HCY levels of B6 females that were raised on the ‘lower folate’ diet (Figures VI-2A and 2B). F1 hybrids were produced by crossing the B6 host strain with the B6-Chr17A/J CSS and were tested to determine the dominance of the trait. Then, congenic strains were generated by crossing F1 hybrids back to the recessive parental strain, i.e. either 171 that were raised on the ‘higher folate’ diet raised on the ‘higher that were males mol/L) of CSS mol/L) of CSS m HCY levels ( HCY levels Mean test was t = 0.05/22 = 0023. P values are indicated for significant differences. hypothesis testing with 22 CSSs, significance levels were subjected to a Bonferroni correction, a Values are expressed in mean ± 1.96 * standard error of the mean. Sample size, n, and HCY levels, Student’s used. To account for multiple Figure VI-4A: Figure VI-4A: the host strain. Underlined strain numbers indicate CSS were compared with those of the B6 Mean HCY levels of each CSS that showed significant difference from B6. To assess differences in mean 172 = 0.05/22 = 0023. P values are = 0.05/22 = 0023. P values a that were raised on the ‘higher folate’ diet raised on the ‘higher that were males mol/L) of CSS mol/L) of CSS m are shown under a different representation. Variance of HCY levels of each CSS was representation. Variance of HCY levels are shown under a different analysis of HCY levels ( analysis of Figure VI-4A Variance
Figure VI-4B. Figure VI-4B. Same data asdifference indicate the CSS that showed significant the B6 host strain. Underlined strain numbers compared with that of test was used. To account for multiple hypothesis in variance of HCY levels, Levene’s from B6. To assess differences correction, significance levels were subjected to a Bonferroni testing with 22 CSSs, differences. indicated for significant 173
B6 or B6-Chr17A/J mice. The Chr 17 fine-mapping for the mean and variance traits of homocysteinemia are separately described below.
Mean
Parental and dominance effects
To test the effects of the gender of the parents, mean HCY levels were compared between reciprocal F1 hybrid mice, i.e. B6 x B6-Chr17A/J (where B6 is the mother) versus B6-Chr17A/J x B6 (where B6-Chr17A/J is the mother). Gender of the parents did not affect mean HCY levels in F1 females, i.e. HCY levels in
B6 x B6-Chr17A/J = B6-Chr17A/J x B6 (Figure VI-5), and the F1 data were consequently pooled.
HCY levels of the parental strains were compared with those of F1 hybrids to test dominance of homocysteinemia. Pooled F1 hybrids had similar levels to
B6 females and significantly higher levels than B6-Chr17A/J females (Figure VI-5), indicating that the high HCY trait in F1 females was inherited in a dominant manner from the B6 parent.
Congenic strains
To create the congenic strains, (B6-Chr17A/J x B6)F1 hybrid females were backcrossed to the recessive parental strain, B6-Chr17A/J males. From this cross
(N2F1 generation), recombinant female progeny were selected and used to establish nine different congenic lines (Figure VI-6). From subsequent 174 mol/L) of parental strains and F1 hybrid females strains and F1 mol/L) of parental m )F1 A/J = a HCY levels ( HCY levels test was t = B6- mean HCY x B6)F1, that A/J Dominance A/J Mean vs F1), significance x B6. : mean HCY levels A/J A/J and P values are indicated. ns: non significant. 0.05/2 = 0.025. Values are expressed in mean ± 1.96 * standard error of the mean. Sample size, n, effects were compared between the parental strains and pooled F1 progeny. To assess differences in mean HCY levels, Student’s used. To account for multiple hypothesis testing with 2 comparisons (B6 vs F1; Chr17 levels were subjected to a Bonferroni correction, folate’ diet. F1 data were levels pooled because HCY B6 x B6-Chr17 Chr17 Figure VI-5. Figure VI-5. Parental effects:levels were compared hybrid between reciprocal F1 females, (B6 x B6-Chr17 vs (B6-Chr17 were raised on the ‘lower 175 to map the trait mean of normal HCY levels trait mean of normal to map the A/J x B6)F1 (AB genotype) hybrid females were backcrossed to the recessive x B6)F1 (AB genotype) A/J (AA genotype) males. From this cross (N2F1 generation), recombinant female progeny were From this cross (N2F1 generation), recombinant (AA genotype) males. A/J Figure VI-6. Congenic strains derived from B6-Chr17 Congenic strains derived Figure VI-6. strains, (B6-Chr17 To create the congenic parental strain, B6-Chr17 confer B6 trait is dominant, AB = BB are alleles that 9 different congenic lines. Because selected and used to establish HCY trait in a recessive manner. manner; AA is allele that confers lower higher HCY trait in a dominant 176 backcrosses (from N3F1 to N6F1 generations), female progeny that shared similar genotype with their mother was phenotyped for HCY levels.
To localize the QTL, mean HCY levels of females from each congenic line were compared to those of the B6 host strain (Figure VI-7). The QTL that significantly reduced B6 mean HCY levels was located in a ~ 14 cM region, or
~28 Mb, that is distal to the D17Mit133 marker and proximal to the D17Mit178 marker (Figure VI-8). More female mice from each congenic line, especially M9 and M20 lines, (Figure VI-7) are needed to confirm and further reduce the size of this interval.
Candidate genes
Polymorphisms between A/J and B6 mice were examined using the
Celera database to determine which gene(s) in the 14 cM region may contribute to differences in mean HCY levels. There are two genes from the HCY-folate metabolism in this region:
1. Cystathionine b-synthase (Cbs, Figure I-1 {step 7}). This gene is located at 17 cM. One polymorphism is located within an exon and changes the amino acid at position 407 (histidine for A/J and arginine for B6 mice) 177 = 0.05/9 = a test was used. t values are indicated. 0.0056. Values are expressed in mean ± 1.96 * standard error of the mean. Sample size, n, and P reduces mean HCY levels of B6 female mice that were raised on the ‘lower folate’ diet. To assess differences in mean HCY levels, Student’s To account for multiple hypothesis testing with 9 congenic strains, significance levels were subjected to a Bonferroni correction, confers lower HCY trait in confers lower HCY trait black recessive manner. The rectangle represents the candidate region that Figure VI-7. Fine resolution mapping of the trait mean of normal HCY levels in females of normal HCY levels of the trait mean Fine resolution mapping Figure VI-7. B6 host line were compared with those of the congenic (backcross generations N3F1-N6F1) Mean HCY levels of each trait is differences from B6. Because B6 numbers indicate congenic line showing significant strain. Underlined strain dominant, AB = BB are alleles that confer higher HCY trait in a dominant manner; AA is allele that 178 Figure VI-8. Candidate region on chromosome 17 that reduces mean HCY levels of B6 female mice levels of B6 female reduces mean HCY chromosome 17 that Candidate region on Figure VI-8. to the to the D17Mit133 marker and proximal to a region of ~ 14 cM, or ~ 28 Mb, that is distal The QTL was mapped and involved in HCY-folate metabolism, CBS chromosome 17. This interval contains 2 genes D17Mit178 marker on genetic distances are indicated in cM. end of the chromosome is on the left and GNMT. The centromeric 179 and 27 polymorphisms are located within intronic regions. The exonic mutation is not positioned in any known functional domain of the Cbs gene, but could contribute, as well as intronic mutations, to differences in enzyme activity between A/J and B6.
2. Glycine N-methyltransferase (Gnmt, Figure I-1 {step 5}). This gene is located at 24 cM. Three polymorphisms are located in intronic but not exonic regions and could contribute to differences in enzyme activity between A/J and
B6.
In addition to polymorphisms in Cbs and Gnmt genes, several other missense mutations were found (see Appendix 2) as well as 25 intronic poymorphisms and one mutation affecting the donor splice site of the chloride intracellular channel 1 (Clic1) gene. Clic1 allows the passage of ions across lipid bilayer membranes. Thus far, Cbs and Gnmt genes are the best candidates.
Variance
Parental and dominance effects
Variance of HCY levels was compared between reciprocal F1 hybrid mice and were not affected by gender of the parents (Figure VI-9). Therefore, HCY levels of reciprocal F1 hybrids were pooled. Pooled F1 hybrids had similar variance to B6-Chr17A/J females and significantly lower variance than B6 (Figure
VI-9), demonstrating that the low HCY trait variance, which was characteristic 180 Dominance = 0.05/2 = 0.025. x B6. a A/J = B6-Chr17 A/J variance of HCY levels were variance of HCY levels Parental effects: x B6)F1 reciprocal hybrid females that were raised on the x B6)F1 reciprocal hybrid A/J )F1 and (B6-Chr17 A/J vs F1), significance levels were subjected to a Bonferroni correction, were subjected to a Bonferroni correction, vs F1), significance levels A/J mol/L) of parental strains and F1 hybrid females strains and F1 mol/L) of parental are shown under a different representation. are shown under a different m Figure VI-5
: variance of HCY levels were compared between the parental strains and pooled F1 progeny. To assess were compared between the parental strains : variance of HCY levels Values are indicated. ns: non-significant. Values are indicated. ns: Figure VI-9. HCY levels ( Figure VI-9. Same data as x B6-Chr17 compared between (B6 levels B6 x B6-Chr17 were pooled because variance of HCY ‘lower folate’ diet. F1 data effects 2 account for multiple hypothesis testing with of HCY levels, Levene’s test was used. To differences in variance Chr17 comparisons (B6 vs F1; 181 of the A/J parent, was inherited in a dominant manner.
Congenic strains
(B6-Chr17A/J x B6)F1 hybrid females were backcrossed to the recessive parental strain, B6 males. From this cross (N2F1 generation), recombinant female progeny were selected and used to establish nine different congenic lines
(Figure VI-10). Female progeny of subsequent backcross generations (from
N3F1 to N6F1 generations) were phenotyped for HCY levels.
HCY variance of females from each congenic line was compared to that of the B6 female littermates (Figure VI-11). Unfortunately, the QTL was not mapped due to insufficient sample size.
DISCUSSION
The present studies demonstrated that several QTLs were detected to control both mean and variance of normal HCY levels of A/J and B6 mice and that genetic factors can contribute to homocysteinemia in health, at least in these two strains of inbred mice. Moreover, normal homocysteinemia may be controlled 182 trait is dominant, AB = AA are alleles that confer trait is dominant, AB = A/J to map the trait variance of normal HCY levels trait variance of normal to map the A/J x B6)F1 (AB genotype) hybrid females were backcrossed to the recessive x B6)F1 (AB genotype) A/J Figure VI-10. Congenic strains derived from B6-Chr17 Congenic strains Figure VI-10. strains, (B6-Chr17 To create the congenic selected recombinant female progeny were genotype) males. From this cross (N2F1 generation), parental strain, B6 (BB 9 different congenic lines. Because B6-Chr17 and used to establish higher trait variance in a recessive manner. a dominant manner; BB is allele that confers lower trait variance in 183 trait is dominant, AB = AA are alleles that confer lower trait variance in a dominant manner; AA are alleles that confer lower trait variance trait is dominant, AB = A/J = a sample size. Bonferroni correction, 0.05/9 = 0.0056. Values are expressed in mean ± 1.96 * standard error of the mean. Sample size, n, is indicated. Comparisons did not reach significance level due to limited testing with 9 congenic strains, significance levels were subjected to a Figure VI-11. Fine resolution mapping of the trait variance of normal HCY levels of normal HCY of the trait variance Fine resolution mapping Figure VI-11. the B6 host line were compared with those of each congenic (backcross generations N3F1-N6F1) Variance HCY levels of strain. Because B6-Chr17 levels, To assess differences in variance of HCY higher trait variance in a recessive manner. BB is allele that confers Levene’s test was used. To account for multiple hypothesis 184 by gene interactions that involved genes from HCY-folate metabolism as well as from other pathways.
Linakge analysis with CSS is a powerful method that was recently used to survey of 53 complex traits and easily detected 150 QTLs (SINGER et al. 2004).
Using the same panel in the present study, a total of 30 QTLs were found to affect mean and variance of serum HCY levels in CSS females and males on the
‘lower folate’ diet, and in males on the ‘higher folate’ diet. Moreover, interactions between the two genomes, A/J and B6, were found. These results suggest that normal homocysteinemia in A/J and B6 mice is controlled by genetic factors.
B6 mice have significantly higher and more variable HCY levels than A/J mice. This linkage analysis was therefore expected to identify QTLs that significantly reduced mean and variance of HCY levels of B6 mice. Surprisingly, most of the QTLs that were found (27 out of 30) further increased mean and variance of HCY levels of B6 mice, even in cases where progenitor strains did not show differences (Figure VI-4B). These results are perplexing because when both genomes were equally represented in ABF1 or BAF1 hybrid mice (see
Chapter Five), HCY levels did not exceed that of B6. Thus, replacement of a single B6 chromosome with an A/J chromosome in homozygous state revealed interactions that were otherwise unknown and seemingly require specific allelic combinations. 185
Fine-mapping was started for the QTL that reduced both mean and variance of HCY levels in females that were raised on the ‘lower folate’ diet. The
QTL that reduced mean HCY levels was mapped to a 14 cM region on Chr 17.
With the Celera database, several missense (Appendix 2) and splice site mutations between genes of A/J and B6 mice were found. The best candidates are two genes involved in HCY-folate metabolism, Cbs and Gnmt. The association of Cbs with hyperhomocysteinemia is known (MUDD et al. 2001;
ROSENBLATT and FENTON 2001), whereas its potential role in normal homocysteinemia is novel, suggesting that Cbs may have also have important role in maintaining normal HCY levels.
Genetic dissection of a complex trait under healthy conditions is a challenging task due to small but reproducible differences in trait as opposed to disease, where larger variation is more evident. It is not surprising that in humans, the role of genetic factors in normal homocysteinemia is difficult to determine (REED et al. 1991; BERG et al. 1992; CESARI et al. 2000), possibly due to the complexity of the trait. Perhaps, particular individuals may carry unidentified genetic variants that may affect susceptibility to disease. 186
MATERIAL AND METHODS
Mouse colonies
B6-Chr(i)A/J CSS mice were provided by Annie Hill from our laboratory and maintained under SPF conditions. Sixteen CSSs (Chrs 3, 4, 6, 7, 8, 9, 10, 11, 12,
14, 15, 16, 17, 18, 19 and X) were analyzed for females; 17 CSSs (Chrs 3, 4, 6,
7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19, X and Y) were analyzed for males on the ‘lower folate’ diet; 15 CSS lines (Chrs 2, 3, 4, 6, 7, 9, 10, 11, 14, 15, 16, 17,
19, X and Y) were analyzed for males on the ‘higher folate’ diet. All mice shared the same animal room with controlled temperature, humidity, and 12 h light-dark cycle. Mice were provided food and water ad libitum.
Diets
Breeding mice were maintained either on the ‘higher folate’ diet, 5010
LabDiet (6.0 mg folic acid/Kg chow; Laboratory autoclavable rodent diet 5010;
Richmond, Indiana) or the ‘lower folate’ diet, 7013 Harlan Teklad (1.83 mg folic acid/Kg chow; NIH-31 modified mouse/rat sterilizable diet; Madison, Wisconsin).
The progeny were raised from conception on either diet. 187
Mouse breeding
CSS
Separate mating pairs of each B6-Chr(i)A/J CSS line were bred on the
‘lower folate’ diet for females and each diet for males to produce progeny that were tested for QTL analysis.
Congenic strains for female B6-Chr17A/J mice on the ‘lower folate’ diet
Crosses between B6-Chr17A/J and B6 mice produced F1 hybrids. Crosses between B6-Chr17A/J females and B6 males produced (B6-Chr17A/J x B6)F1 mice and crosses between B6 females and B6-Chr17A/J males produced reciprocal B6 x B6-Chr17A/J)F1 mice.
Congenic strains to map mean HCY levels
(B6-Chr17A/J x B6)F1 hybrid females were backcrossed to the recessive parental strain, B6-Chr17A/J males to generate N2F1 mice. Recombinant N2F1 females were genotyped and used to establish nine different congenic lines
(Figure VI-6). Only females from subsequent generations (from N3F1 to N6F1 generations) and that shared similar genotype with their mother were phenotyped for HCY levels.
Congenic strains to map variance of HCY levels
(B6-Chr17A/J x B6)F1 hybrid females were backcrossed to the recessive parental strain, B6 males to generate N2F1 mice. Recombinant N2F1 females were genotyped and used to establish nine different congenic lines 188
(Figure VI-10). Only females from subsequent generations (from N3F1 to N6F1 generations) and that shared similar genotype with their mother were phenotyped for HCY levels.
Polymorphism detection
Genomic DNA was prepared by digesting tail samples with proteinase K at
55oC overnight in 1X buffer (50 mM KCl, 10 mM Tris-HCl, 0.1% Triton X-100, pH
9.0 at 25oC). The samples were then placed at 100oC for 1 hour to inactivate the enzyme. Recombination events in the N2F1-N6F1 generations were detected by
PCR using polymorphic markers (Invitrogen, CA) on chromosome 17. Each PCR reaction contained a final concentration of 1x of PCR buffer (Invitrogen, CA),
2mM of MgCl2 (Invitrogen, CA), 0.2mm of each forward and reverse primers
(Invitrogen, CA), 1.2 mM dNTPs (Invitrogen, CA), 0.03 U/ml of Taq DNA polymerase (Invitrogen, CA), and 1.5 ml of 1:5 dilution of genomic DNA. PCR reactions were amplified at 94oC for 2 mins, then 35 cycles of 94oC for 1 min,
50oC for 1 min, 72oC for 1 min, with a final extension at 72oC for 10 mins.
Products were visualized by ethidium bromide after separation on 6% polyacrilamide gels. 189
The following markers on Chr 17 were used to genotype females from N2F1 to
N6F1 generations:
Markers position in cM
D17Mit113 6.5 D17Mit133 10.4 D17Mit198 16 D17Mit176 22.5 D17Mit178 24.5 D17Mit139 30.2 D17Mit20 34.3 D17Mit193 41.5 D17Mit39 45.3 D17Mit2 49.7 D17Mit155 55.7
Blood samples
Blood samples were obtained from the sub-orbital sinus of male and virgin female mice that were 6-8 weeks old, and were collected in non-heparinized tubes. After centrifugation, serum samples were stored at -80°C. Virgin females were used to control for possible effects of pregnancy on HCY levels (KANG et al.
1986; STEEGERS-THEUNISSEN et al. 1997).
Homocysteine measurements
Total serum HCY levels were measured by HPLC and tandem mass spectroscopy methods. Both methods consist of reducing protein-HCY disulfide 190 bonds prior analysis. The concentration of HCY was determined by HPLC or tandem mass spectroscopy.
HPLC
HPLC method of Ubbink and Vermaak (UBBINK et al. 1991) was used.
Tandem mass spectroscopy
Tandem mass spectroscopy method of McCann and co-authors was used
(MCCANN et al. 2003).
Tandem mass spectroscopy vs HPLC: correction of HCY levels
Preliminary data showed a consistent bias in which samples measured by
HPLC gave higher HCY levels than those measured by tandem mass spectroscopy (TMS). The following method was used to adjust differences in techniques to pool the data. Inbred mouse strains, whose HCY levels were measured by both methods, were selected to control for genetic and environmental factors because HCY levels should be similar within each group.
HCY levels from 490 samples that were collected from four different strains under various conditions were used. Of the 490 samples, 282 were measured by
HPLC and the remainder by TMS. The mean HCY levels were calculated for each group and the correction factor was applied as followed:
mHPLC (5.849 mmol/L) - mTMS (4.238 mmol/L) = 1.611 mmol/L 191
mHPLC: mean HCY levels across all samples measured by HPLC
mTMS: mean HCY levels across all samples measured by TMS
HCY levels differed by 1.611 mmol/L with a correlation factor (r2) of 0.85 between the two methods after correction. Then, each sample measured by HPLC had its
HCY level subtracted by 1.611 mmol/L and pooled with samples measured by
TMS.
Statistical analysis
To assess differences in mean HCY levels, the Student’s t test was used.
To assess differences in variance, the Levene's test was used (MANLY 1997).
Significance levels were subjected to a Bonferroni correction to account for multiple testing. 192
CHAPTER VII
SUMMARY AND FUTURE DIRECTIONS 193
The goal of this project was to dissect the genetic and molecular control of homocysteinemia in disease and in health using the laboratory mouse as a model, as well as phenotypic-driven approaches, which did not make assumptions about the identity of genes or biochemical processes that control
HCY levels. As a result, novel and unexpected influences on HCY metabolism were discovered.
Chapters Two, Three and Four addressed the following question: Do mutations in genes other than CBS, MTHFR and MTR affect HCY-folate metabolism? To approach this subject, single gene mouse models with phenotypes associated with hyperhomocysteinemia in humans were used in combination with metabolite and expression profiles.
Chapter Two described studies that tested whether HCY-folate metabolism was disrupted in mouse models of neural tube defects (NTDs) and colon cancer. Mice with mutations in the Apob, Gli3, Pax3, Ptch1 and Ski genes were used as models for NTDs and mice with a mutation in the Apc gene as a model for colon cancer. Serum HCY levels were measured and hepatic gene expression profiles were monitored in heterozygous mutants, because homozygotes are embryonic lethal. Partial deficiencies of APC, APOB, PAX3 and
GLI3 (on the B6 background) increased HCY levels, whereas PTCH, SKI and
GLI3 (on the C3H background) did not affect HCY levels. Significant differences in gene expression patterns, often affecting folate-HCY metabolism, in these mutants were also found. These results provide clues to new pathways, i.e. WNT 194 and hedgehog signal transduction (APC, GLI3 and PAX3) and lipid transport
(APOB) that adversely affect HCY metabolism and perhaps contribute to the pathogenesis of birth defects and adult diseases.
This approach revealed important functional relations between a metabolic and signal transduction pathways, that otherwise were not evident. Interestingly, a recent study showed that alteration in folate transport adversely affects hedgehog signal transduction (TANG and FINNELL 2003), clearly highlighting the interrelatedness among these biological pathways. Additional characterization of causes and consequence of genetic perturbations is needed to further understand these complex relationships, i.e. analysis of HCY-folate metabolism in mutants of WNT and hedgehog signaling and reciprocal analysis of WNT and hedgehog signaling in mutants of HCY-folate metabolism. Investigation of several mouse mutants can also test whether relationships among genes in the same pathway are reproduced in another pathway. For instance, because
PTCH1 and GLI3 have contrasting effects on hedgehog signaling (MURONE et al.
1999), they would be expected to also have antagonizing effects on HCY-folate metabolism, which was true for HCY levels (Chapter Two, Figures II-1 and II-2).
Do HCY-folate metabolism, WNT and hedgehog signaling pathways, and lipid transport/metabolism contribute together to disease phenotypes? To test this hypothesis, phenotypic effects of pair of gene mutations could be evaluated in double-mutant mice. For instance, do mutations in both HCY-folate metabolism and WNT signaling worsen, improve or even affect 195 homocysteinemia, or disease phenotypes of mice compared to either mutation alone? Are these effects additive or interactive? To control for possible genetic background effects, mutations ought to be congenic on the same inbred background. These studies will provide clues about the degree of relatedness among diverse biological pathways and mechanisms contributing to the pathogenesis of disease in these mice.
A related topic to the abovementioned question is the significance of genetic background: Could genetic background affect relationships among these biological pathways, i.e. HCY-folate metabolism, WNT and hedgehog signaling?
In Chapter Two, several single gene mutations adversely affected HCY-folate metabolism but the effect depended on genetic background: partial deficiency of
GLI3 showed contrasting effects on HCY levels on C57BL/6J vs C3HeB/FeJ background (Chapter Two, Figures II-1 and II-2). One approach to test the background is to compare phenotypes and other traits, such as expression and metabolite profiles, of single- and double-mutant mice on different genetic backgrounds. Obvious genetic background effects would suggest that several mechanisms contribute to the pathogenesis of disease.
Chapter Three described studies that monitored the status of HCY-folate metabolism in Crooked-tail mutant mice, whose developmental defects are folate-responsive, under various dietary folate supplementations. Serum metabolite levels suggest that folate suppressed NTDs through a mechanism that did not directly involve modulating HCY levels. Hepatic expression profiles 196 suggest that homozygosity for the Cd mutation led to a defect in the utilization of intracellular folate that does not affect HCY levels. Expression profiles for mutants with phenotypes associated with anomalies in HCY and folate metabolism (mice with mutations in the Apob, Gli3, Pax3, Ptch1, Ski and Apc genes as well as Cd mutants) revealed that folate-responsive mouse mutants have distinct expression profiles. Based on these profiles, folate-resistance of
NTDs in Ski mutant mice was predicted and verified, demonstrating that expression levels of genes in these profiles can predict folate responsiveness
An important aspect of HCY-folate metabolism is the ability of folic acid supplementation to reduce the risk of certain birth defects and adult diseases.
Prediction of a defect-responsiveness to particular treatment is significant in development of diagnostic tools. According to cluster results of metabolite and expression profiles (Chapter Three, Figure III-7), the defects in Ski, Ptch1, Apob and Gli3 (on C57BL/6J background) mutants may not respond to folate supplementation, or may respond in a manner functionally distinct from the response of Cd, Apc, Pax3 and Gli3 (on C3HeB/FeJ background) mutants. A broader survey of mutants is needed to test the generality of these observations by analyzing: 1) additional mutants whose defects are folate-responsive; 2) mutants whose defects are folate-resistant; 3) mutants whose response to folate is unknown. Assays on pertinent genes, i.e. genes that consistently predict response to folate, could be developed and applied to humans. 197
Chapters Five and Six addressed the following question: Do genes contribute to variation in normal HCY levels? Both genetic and environmental factors control hyperhomocysteinemia. However, little is known about the relative impact of these factors on HCY levels among individuals who do not have hyperhomocysteinemia. Inheritance of normal homocysteinemia as well as linkage analysis were investigated in two inbred strains, A/J and C57BL/6J (B6), that had normal but significantly different levels in serum HCY.
Chapter Five investigated the inheritance of normal homocysteinemia in
A/J and B6 mice. Results showed that normal homocysteinemia was a complex trait that depended on gender, diet and parental effects and, most interestingly, both mean and variance in HCY levels were affected.
Does variability in HCY levels correlate with increased risk for HCY-related disease? HCY levels were more variable in B6 than A/J females raised on the
‘lower folate’ diet (Chapter Five, Table V-3 and Figure V-2). Perhaps B6 females had difficulty maintaining stable HCY levels on the ‘lower folate’ diet and, consequently may be at greater risk for HCY-related disease than A/J females on the ‘lower folate’ diet. This type of variability is not unusual in humans, and genetic factors that are generally silent may increase disease risk under specific environmental conditions (RUTHERFORD 2000). One approach to assess the significance of HCY variability is to evaluate onset of disorders between B6 and
A/J female mice fed a folate-deficient diet, assuming that differences in HCY variance between the two inbred strains are maintained. Disorders such as 198 hepatocarcinogenesis could be monitored because dietary deficiencies in folate induce tumor formation in rats, primarily in the liver (MIKOL et al. 1983; GHOSHAL and FARBER 1984; LOCKER et al. 1986; HENNING et al. 1997a). If HCY variability correlates with increased risk for disease, onset of tumor formation should occur more rapidly in B6 than A/J females. These experiments may provide clues about the significance of variability in disease.
Chapter Six consisted of detecting genetic factors that affected normal homocysteinemia. Linkage analysis was used to detect quantitative trait loci
(QTL) that control normal homocysteinemia in A/J and B6 mice. Several QTLs were found to affect mean and variance of serum HCY levels. These two HCY traits may be controlled by related or different QTLs, which may involve genes from HCY-folate metabolism as well as genes from other biological pathways.
These studies demonstrate there are multiple genetic ways to maintain normal
HCY levels, as there are several genetic mutations that lead to hyperhomocysteinemia.
The linkage analysis revealed novel interactions where replacement of a single B6 chromosome with an A/J chromosome in homozygous state further increased HCY levels of B6 mice. However, this linkage analysis was expected to identify QTLs that significantly reduced mean and variance of HCY levels of
B6 mice. Perhaps a progressive increase in number of A/J chromosomes onto the B6 host strain may attenuate these effects. To test this hypothesis, crosses between two different CSSs will increase in the number of A/J chromosomes 199 onto the B6 host strains. The resulting ‘double CSS’ would be expected to have lower mean and variance of HCY levels than either ‘single CSS’ alone. This would enable fine mapping of interacting loci.
Several QTLs were found to affect homocysteinemia in A/J and B6 mice and they may also affect mean and variance levels of HCY in humans. These results may provide important insights into numerous mechanisms utilized by an organism to maintain healthy conditions or clues to genetic variants that are tolerable in certain circumstances but increase certain risks in specific environmental conditions. 200
APPENDIX 1 201
Nutrients composition of Harland Teklad 7013 (‘lower’ folate) and Lab Diet 5010 (‘higher’ folate) diets. Nutrients in bold differ in quantity (at least 2 fold) between ‘higher’ and ‘lower’ folate diets. Nutrients both in bold and italic are known to influence HCY levels. Nutrients Harland Teklad 7013 Lab diet 5010 (‘lower’ folate diet) (‘higher’ folate diet) Crude protein 18% 23% Crude fat 6% 4.5% Crude fiber 5% 6% Arginine 1.27% 1.42% Methionine 0.35% 0.43% Cystine 0.32% 0.35% Histidine 0.41% 0.58% Isoleucine 0.90% 1.22% Leucine 1.51% 1.85% Lysine 0.96% 1.36% Phenylalanine+Tyrosine 1.54% 1.73% Threonine 0.71% 0.89% Tryptophane 0.22% 0.27% Valine 0.91% 1.17% Calcium 1.19% 1.01% Phosphorus 1.07% 0.74% Sodium 0.33% 0.28% Chlorine 0.35% 0.43% Potassium 0.61% 1.08% Magnesium 0.24% 0.21% Iron 271.12 mg/Kg 250 mg/Kg Manganese 152.80 mg/Kg 69 mg/Kg Zinc 56.70 mg/Kg 71 mg/Kg Copper 12.53 mg/Kg 12 mg/Kg Iodine 2.01 mg/Kg 1.17 mg/Kg Cobalt 0.56 mg/Kg 1.5 mg/Kg Selenium 0.24 mg/Kg 0.23 mg/Kg Vitamin A 26.51 IU/g 12 IU/g Vitamin D3 4.19 IU/g 3.3 IU/g Vitamin E 43.93 IU/g 32 IU/g Niacin 95.76 mg/Kg 92.2 mg/Kg Pantothenic acid 41.49 mg/Kg 12.5 mg/Kg Pyrodoxine 10.16 mg/Kg 6.5 mg/Kg Riboflavin 8.10 mg/Kg 4.5 mg/Kg Menadione 66.68 mg/Kg 0 Folic Acid 1.83 mg/Kg 6.0 mg/Kg Biotin 0.33 mg/Kg 0.20 ppm Vitamin B12 81.62 mcg/Kg 19.8 mcg/Kg 202
APPENDIX 2 203 and D17Mit133 Genes, distal to Genes, distal with missense mutations are represented. with missense mutations D17Mit178, List of missense mutations that differ between A/J and B6 strains on Chr 17. and B6 strains on Chr differ between A/J mutations that List of missense proximal to 204
BIBLIOGRAPHY 205
APPLING, D. R., 1991 Compartmentation of folate-mediated one-carbon metabolism in eukaryotes. Faseb J 5: 2645-2651.
BARBERA, J. P., T. A. RODRIGUEZ, N. D. GREENE, W. J. WENINGER, A. SIMEONE et al., 2002 Folic acid prevents exencephaly in Cited2 deficient mice. Hum Mol Genet 11: 283-293.
BEHRENS, J., B. A. JERCHOW, M. WURTELE, J. GRIMM, C. ASBRAND et al., 1998 Functional interaction of an axin homolog, conductin, with beta- catenin, APC, and GSK3beta. Science 280: 596-599.
BEILBY, J., and E. ROSSI, 2000 Number 58: homocysteine and disease. Pathology 32: 262-273.
BELKNAP, J. K., 2003 Chromosome substitution strains: some quantitative considerations for genome scans and fine mapping, pp. 723-732 in Mamm Genome.
BELLAMY, M. F., and I. F. MCDOWELL, 1997 Putative mechanisms for vascular damage by homocysteine. J Inherit Metab Dis 20: 307-315.
BERG, K., M. R. MALINOW, P. KIERULF and B. UPSON, 1992 Population variation and genetics of plasma homocyst(e)ine level. Clin Genet 41: 315-321.
BERGMARK, C., M. A. MANSOOR, J. SWEDENBORG, U. DE FAIRE, A. M. SVARDAL et al., 1993 Hyperhomocysteinemia in patients operated for lower extremity ischaemia below the age of 50--effect of smoking and extent of disease. Eur J Vasc Surg 7: 391-396.
BERK, M., S. Y. DESAI, H. C. HEYMAN and C. COLMENARES, 1997 Mice lacking the ski proto-oncogene have defects in neurulation, craniofacial, patterning, and skeletal muscle development. Genes Dev 11: 2029-2039.
BERRY, R. J., Z. LI, J. D. ERICKSON, S. LI, C. A. MOORE et al., 1999 Prevention of neural-tube defects with folic acid in China. China-U.S. Collaborative Project for Neural Tube Defect Prevention [corrected; erratum to be published]. N Engl J Med 341: 1485-1490. 206
BERTUCCI, F., K. BERNARD, B. LORIOD, Y. C. CHANG, S. GRANJEAUD et al., 1999 Sensitivity issues in DNA array-based expression measurements and performance of nylon microarrays for small samples. Hum Mol Genet 8: 1715-1722.
BEZARD, E., S. DOVERO, B. BIOULAC and C. GROSS, 1997 Effects of different schedules of MPTP administration on dopaminergic neurodegeneration in mice. Exp Neurol 148: 288-292.
BIRD, R. P., 1995 Role of aberrant crypt foci in understanding the pathogenesis of colon cancer. Cancer Lett 93: 55-71.
BLANDINI, F., R. FANCELLU, E. MARTIGNONI, A. MANGIAGALLI, C. PACCHETTI et al., 2001 Plasma homocysteine and l-dopa metabolism in patients with Parkinson disease. Clin Chem 47: 1102-1104.
BLOUNT, B. C., M. M. MACK, C. M. WEHR, J. T. MACGREGOR, R. A. HIATT et al., 1997 Folate deficiency causes uracil misincorporation into human DNA and chromosome breakage: implications for cancer and neuronal damage. Proc Natl Acad Sci U S A 94: 3290-3295.
BOERS, G. H., B. FOWLER, A. G. SMALS, F. J. TRIJBELS, A. I. LEERMAKERS et al., 1985a Improved identification of heterozygotes for homocystinuria due to cystathionine synthase deficiency by the combination of methionine loading and enzyme determination in cultured fibroblasts. Hum Genet 69: 164-169.
BOERS, G. H., A. G. SMALS, F. J. TRIJBELS, B. FOWLER, J. A. BAKKEREN et al., 1985b Heterozygosity for homocystinuria in premature peripheral and cerebral occlusive arterial disease. N Engl J Med 313: 709-715.
BORCHELT, D. R., G. THINAKARAN, C. B. ECKMAN, M. K. LEE, F. DAVENPORT et al., 1996 Familial Alzheimer's disease-linked presenilin 1 variants elevate Abeta1-42/1-40 ratio in vitro and in vivo. Neuron 17: 1005-1013.
BOSCO, P., R. M. GUEANT-RODRIGUEZ, G. ANELLO, C. BARONE, F. NAMOUR et al., 2003 Methionine synthase (MTR) 2756 (A --> G) polymorphism, double heterozygosity methionine synthase 2756 AG/methionine synthase reductase (MTRR) 66 AG, and elevated homocysteinemia are three risk 207
factors for having a child with Down syndrome. Am J Med Genet 121A: 219-224.
BOTTIGLIERI, T., K. HYLAND and E. H. REYNOLDS, 1994 The clinical potential of ademetionine (S-adenosylmethionine) in neurological disorders. Drugs 48: 137-152.
BOTTO, L. D., and Q. YANG, 2000 5,10-Methylenetetrahydrofolate reductase gene variants and congenital anomalies: a HuGE review. Am J Epidemiol 151: 862-877.
BOUSHEY, C. J., S. A. BERESFORD, G. S. OMENN and A. G. MOTULSKY, 1995 A quantitative assessment of plasma homocysteine as a risk factor for vascular disease. Probable benefits of increasing folic acid intakes. Jama 274: 1049-1057.
BRATTSTROM, L., D. E. WILCKEN, J. OHRVIK and L. BRUDIN, 1998 Common methylenetetrahydrofolate reductase gene mutation leads to hyperhomocysteinemia but not to vascular disease: the result of a meta- analysis. Circulation 98: 2520-2526.
BRODY, L. C., M. CONLEY, C. COX, P. N. KIRKE, M. P. MCKEEVER et al., 2002 A polymorphism, R653Q, in the trifunctional enzyme methylenetetrahydrofolate dehydrogenase/methenyltetrahydrofolate cyclohydrolase/formyltetrahydrofolate synthetase is a maternal genetic risk factor for neural tube defects: report of the Birth Defects Research Group. Am J Hum Genet 71: 1207-1215.
BROWN, C. A., K. Q. MCKINNEY, J. S. KAUFMAN, R. A. GRAVEL and R. ROZEN, 2000 A common polymorphism in methionine synthase reductase increases risk of premature coronary artery disease. J Cardiovasc Risk 7: 197-200.
BROWN, P. O., and D. BOTSTEIN, 1999 Exploring the new world of the genome with DNA microarrays. Nature Genetics 21: 33-37.
BURGOON, J. M., J. SELHUB, M. NADEAU and T. W. SADLER, 2002 Investigation of the effects of folate deficiency on embryonic development through the establishment of a folate deficient mouse model. Teratology 65: 219-227. 208
CARMEL, R., 2001 Folate deficiency, pp. 271-288 in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
CARTER, M., S. ULRICH, Y. OOFUJI, D. A. WILLIAMS and M. E. ROSS, 1999 Crooked tail (Cd) models human folate-responsive neural tube defects. Hum Mol Genet 8: 2199-2204.
CESARI, M., A. B. BURLINA, K. NARKIEWICZ, M. T. SARTORI, A. SACCHETTO et al., 2000 Are fasting plasma homocyst(e)ine levels heritable? A study of normotensive twins. J Investig Med 48: 351-358.
CHANDAR, N., and B. LOMBARDI, 1988 Liver cell proliferation and incidence of hepatocellular carcinomas in rats fed consecutively a choline-devoid and a choline-supplemented diet. Carcinogenesis 9: 259-263.
CHEN, C., M. E. HALKOS, S. M. SUROWIEC, B. S. CONKLIN, P. H. LIN et al., 2000 Effects of homocysteine on smooth muscle cell proliferation in both cell culture and artery perfusion culture models. J Surg Res 88: 26-33.
CHEN, J., E. GIOVANNUCCI, S. E. HANKINSON, J. MA, W. C. WILLETT et al., 1998 A prospective study of methylenetetrahydrofolate reductase and methionine synthase gene polymorphisms, and risk of colorectal adenoma. Carcinogenesis 19: 2129-2132.
CHEN, J., E. GIOVANNUCCI, K. KELSEY, E. B. RIMM, M. J. STAMPFER et al., 1996 A methylenetetrahydrofolate reductase polymorphism and the risk of colorectal cancer. Cancer Res 56: 4862-4864.
CHEN, L. H., M. L. LIU, H. Y. HWANG, L. S. CHEN, J. KORENBERG et al., 1997 Human methionine synthase. cDNA cloning, gene localization, and expression. J Biol Chem 272: 3628-3634.
CHEN, Z., A. C. KARAPLIS, S. L. ACKERMAN, I. P. POGRIBNY, S. MELNYK et al., 2001 Mice deficient in methylenetetrahydrofolate reductase exhibit hyperhomocysteinemia and decreased methylation capacity, with neuropathology and aortic lipid deposition. Hum Mol Genet 10: 433-443. 209
CHRISTENSEN, B., L. ARBOUR, P. TRAN, D. LECLERC, N. SABBAGHIAN et al., 1999 Genetic polymorphisms in methylenetetrahydrofolate reductase and methionine synthase, folate levels in red blood cells, and risk of neural tube defects. Am J Med Genet 84: 151-157.
CHRISTENSEN, B., A. B. GUTTORMSEN, J. SCHNEEDE, B. RIEDEL, H. REFSUM et al., 1994 Preoperative methionine loading enhances restoration of the cobalamin-dependent enzyme methionine synthase after nitrous oxide anesthesia. Anesthesiology 80: 1046-1056.
CHRISTENSEN, B., H. REFSUM, O. VINTERMYR and P. M. UELAND, 1991 Homocysteine export from cells cultured in the presence of physiological or superfluous levels of methionine: methionine loading of non- transformed, transformed, proliferating, and quiescent cells in culture. J Cell Physiol 146: 52-62.
CLARKE, R., and J. ARMITAGE, 2000 Vitamin supplements and cardiovascular risk: review of the randomized trials of homocysteine-lowering vitamin supplements. Semin Thromb Hemost 26: 341-348.
CLARKE, R., A. D. SMITH, K. A. JOBST, H. REFSUM, L. SUTTON et al., 1998 Folate, vitamin B12, and serum total homocysteine levels in confirmed Alzheimer disease. Arch Neurol 55: 1449-1455.
CLARKE, S., and K. BANFIELD, 2001 S-adenosylmethionine-dependent methyltransferase, pp. 63-78 in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
COLLABORATION, H. L. T., 1998 Lowering blood homocysteine with folic acid based supplements: meta-analysis of randomised trials. Homocysteine Lowering Trialists' Collaboration. Bmj 316: 894-898.
COLMENARES, C., H. A. HEILSTEDT, L. G. SHAFFER, S. SCHWARTZ, M. BERK et al., 2002 Loss of the SKI proto-oncogene in individuals affected with 1p36 deletion syndrome is predicted by strain-dependent defects in Ski-/- mice. Nat Genet 30: 106-109. 210
CONSORTIUM, I. G. S., 2001 Initial sequencing and analysis of the human genome. Nature 409: 860-921.
COOK, J. W., L. M. TAYLOR, S. L. ORLOFF, G. J. LANDRY, G. L. MONETA et al., 2002 Homocysteine and arterial disease. Experimental mechanisms. Vascul Pharmacol 38: 293-300.
COOK, R. J., 2001 Folate metabolism, pp. 113-134 in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
CORONA, G., G. TOFFOLI, M. FABRIS, A. VIEL, A. ZARRELLI et al., 1997 Homocysteine accumulation in human ovarian carcinoma ascitic/cystic fluids possibly caused by metabolic alteration of the methionine cycle in ovarian carcinoma cells. Eur J Cancer 33: 1284-1290.
COTTRELL, S., D. BICKNELL, L. KAKLAMANIS and W. F. BODMER, 1992 Molecular analysis of APC mutations in familial adenomatous polyposis and sporadic colon carcinomas. Lancet 340: 626-630.
CRAVO, M., P. FIDALGO, A. D. PEREIRA, A. GOUVEIA-OLIVEIRA, P. CHAVES et al., 1994 DNA methylation as an intermediate biomarker in colorectal cancer: modulation by folic acid supplementation. Eur J Cancer Prev 3: 473-479.
CRAVO, M. L., L. M. GLORIA, J. SELHUB, M. R. NADEAU, M. E. CAMILO et al., 1996 Hyperhomocysteinemia in chronic alcoholism: correlation with folate, vitamin B-12, and vitamin B-6 status. Am J Clin Nutr 63: 220-224.
CRAVO, M. L., J. B. MASON, Y. DAYAL, M. HUTCHINSON, D. SMITH et al., 1992 Folate deficiency enhances the development of colonic neoplasia in dimethylhydrazine-treated rats. Cancer Res 52: 5002-5006.
CROTEAU, D. L., and V. A. BOHR, 1997 Repair of oxidative damage to nuclear and mitochondrial DNA in mammalian cells. J Biol Chem 272: 25409-25412.
CZEIZEL, A. E., and I. DUDAS, 1992 Prevention of the first occurrence of neural- tube defects by periconceptional vitamin supplementation. N Engl J Med 327: 1832-1835. 211
DAHL, E., H. KOSEKI and R. BALLING, 1997 Pax genes and organogenesis. Bioessays 19: 755-765.
DAYAL, S., T. BOTTIGLIERI, E. ARNING, N. MAEDA, M. R. MALINOW et al., 2001 Endothelial dysfunction and elevation of S-adenosylhomocysteine in cystathionine beta-synthase-deficient mice. Circ Res 88: 1203-1209.
DE VRIESE, A. S., J. H. DE SUTTER, M. DE BUYZERE and D. DUPREZ, 1998 Mild to moderate hyperhomocysteinaemia in cardiovascular disease. Acta Cardiol 53: 337-344.
DEKOU, V., V. GUDNASON, E. HAWE, G. J. MILLER, D. STANSBIE et al., 2001 Gene- environment and gene-gene interaction in the determination of plasma homocysteine levels in healthy middle-aged men. Thromb Haemost 85: 67-74.
DEVLIN, A. M., E. H. LING, J. M. PEERSON, S. FERNANDO, R. CLARKE et al., 2000 Glutamate carboxypeptidase II: a polymorphism associated with lower levels of serum folate and hyperhomocysteinemia. Hum Mol Genet 9: 2837-2844.
DIERKES, J., S. WESTPHAL and C. LULEY, 1999 Serum homocysteine increases after therapy with fenofibrate or bezafibrate. Lancet 354: 219-220.
DIETRICH, W. F., E. S. LANDER, J. S. SMITH, A. R. MOSER, K. A. GOULD et al., 1993 Genetic identification of Mom-1, a major modifier locus affecting Min- induced intestinal neoplasia in the mouse. Cell 75: 631-639.
DOMINGUEZ, M., M. BRUNNER, E. HAFEN and K. BASLER, 1996 Sending and receiving the hedgehog signal: control by the Drosophila Gli protein Cubitus interruptus. Science 272: 1621-1625.
DOOLIN, M. T., S. BARBAUX, M. MCDONNELL, K. HOESS, A. S. WHITEHEAD et al., 2002 Maternal genetic effects, exerted by genes involved in homocysteine remethylation, influence the risk of spina bifida. Am J Hum Genet 71: 1222-1226. 212
DUAN, W., B. LADENHEIM, R. G. CUTLER, KRUMAN, II, J. L. CADET et al., 2002 Dietary folate deficiency and elevated homocysteine levels endanger dopaminergic neurons in models of Parkinson's disease. J Neurochem 80: 101-110.
DUGGAN, D. J., M. BITTNER, Y. CHEN, P. MELTZER and J. M. TRENT, 1999 Expression profiling using cDNA microarrays. Nature Genetics 21: 10-14.
DURAND, P., S. LUSSIER-CACAN and D. BLACHE, 1997 Acute methionine load- induced hyperhomocysteinemia enhances platelet aggregation, thromboxane biosynthesis, and macrophage-derived tissue factor activity in rats. Faseb J 11: 1157-1168.
DUTHIE, S. J., and A. HAWDON, 1998 DNA instability (strand breakage, uracil misincorporation, and defective repair) is increased by folic acid depletion in human lymphocytes in vitro. Faseb J 12: 1491-1497.
EBBESEN, L. S., K. CHRISTIANSEN and J. INGERSLEV, 2003 Hyperhomocysteinemia due to folate deficiency is thrombogenic in rats. J Nutr 133: 2250-2255.
EBERHARDT, R. T., M. A. FORGIONE, A. CAP, J. A. LEOPOLD, M. A. RUDD et al., 2000 Endothelial dysfunction in a murine model of mild hyperhomocyst(e)inemia. J Clin Invest 106: 483-491.
EIKELBOOM, J. W., E. LONN, J. GENEST, JR., G. HANKEY and S. YUSUF, 1999 Homocyst(e)ine and cardiovascular disease: a critical review of the epidemiologic evidence. Ann Intern Med 131: 363-375.
EISEN, M. B., P. T. SPELLMAN, P. O. BROWN and D. BOTSTEIN, 1998 Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 95: 14863-14868.
ENGBERSEN, A. M., D. G. FRANKEN, G. H. BOERS, E. M. STEVENS, F. J. TRIJBELS et al., 1995 Thermolabile 5,10-methylenetetrahydrofolate reductase as a cause of mild hyperhomocysteinemia. Am J Hum Genet 56: 142-150. 213
ERNEST, S., B. CHRISTENSEN, B. M. GILFIX, O. A. MAMER, A. HOSACK et al., 2002 Genetic and molecular control of folate-homocysteine metabolism in mutant mice. Mamm Genome 13: 259-267.
ESKES, T. K., 1998 Open or closed? A world of difference: a history of homocysteine research. Nutr Rev 56: 236-244.
ESKES, T. K., 2001 Homocysteine and human reproduction in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
ESSIEN, F. B., 1992 Maternal methionine supplementation promotes the remediation of axial defects in Axd mouse neural tube mutants. Teratology 45: 205-212.
FAIRFIELD, K. M., and R. H. FLETCHER, 2002 Vitamins for chronic disease prevention in adults: scientific review. Jama 287: 3116-3126.
FARESE, R. V., JR., and J. HERZ, 1998 Cholesterol metabolism and embryogenesis. Trends Genet 14: 115-120.
FENECH, M., 1998 Chromosomal damage rate, aging, and diet. Ann N Y Acad Sci 854: 23-36.
FENECH, M., 2001 The role of folic acid and Vitamin B12 in genomic stability of human cells. Mutat Res 475: 57-67.
FENECH, M., C. AITKEN and J. RINALDI, 1998 Folate, vitamin B12, homocysteine status and DNA damage in young Australian adults. Carcinogenesis 19: 1163-1171.
FINKELSTEIN, J. D., 1990 Methionine metabolism in mammals. Journal of Nutrition Biochemistry 1: 228-237.
FINKELSTEIN, J. D., 1998 The metabolism of homocysteine: pathways and regulation. Eur J Pediatr 157 Suppl 2: S40-44. 214
FLEMING, A., and A. J. COPP, 1998 Embryonic folate metabolism and mouse neural tube defects. Science 280: 2107-2109.
FODDE, R., J. KUIPERS, C. ROSENBERG, R. SMITS, M. KIELMAN et al., 2001 Mutations in the APC tumour suppressor gene cause chromosomal instability. Nat Cell Biol 3: 433-438.
FOWLER, B., 1997 Disorders of homocysteine metabolism. J Inherit Metab Dis 20: 270-285.
FOWLER, B., 2001 Transport and tissue distribution of homocysteine and related S-adenosyl compounds, pp. 163-175 in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
FREUDENHEIM, J. L., J. R. MARSHALL, J. E. VENA, R. LAUGHLIN, J. R. BRASURE et al., 1996 Premenopausal breast cancer risk and intake of vegetables, fruits, and related nutrients. J Natl Cancer Inst 88: 340-348.
FROSST, P., H. J. BLOM, R. MILOS, P. GOYETTE, C. A. SHEPPARD et al., 1995 A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat Genet 10: 111-113.
GEORGE, L., J. L. MILLS, A. L. JOHANSSON, A. NORDMARK, B. OLANDER et al., 2002 Plasma folate levels and risk of spontaneous abortion. Jama 288: 1867- 1873.
GHOSHAL, A. K., and E. FARBER, 1984 The induction of liver cancer by dietary deficiency of choline and methionine without added carcinogens. Carcinogenesis 5: 1367-1370.
GIBSON, G., and G. WAGNER, 2000 Canalization in evolutionary genetics: a stabilizing theory? Bioessays 22: 372-380.
GILFIX, B. M., D. W. BLANK and D. S. ROSENBLATT, 1997 Novel reductant for determination of total plasma homocysteine. Clin Chem 43: 687-688. 215
GIOVANNUCCI, E., E. B. RIMM, A. ASCHERIO, M. J. STAMPFER, G. A. COLDITZ et al., 1995 Alcohol, low-methionine--low-folate diets, and risk of colon cancer in men. J Natl Cancer Inst 87: 265-273.
GIOVANNUCCI, E., M. J. STAMPFER, G. A. COLDITZ, D. J. HUNTER, C. FUCHS et al., 1998 Multivitamin use, folate, and colon cancer in women in the Nurses' Health Study. Ann Intern Med 129: 517-524.
GIOVANNUCCI, E., M. J. STAMPFER, G. A. COLDITZ, E. B. RIMM, D. TRICHOPOULOS et al., 1993 Folate, methionine, and alcohol intake and risk of colorectal adenoma. J Natl Cancer Inst 85: 875-884.
GIRI, S., P. D. THOMPSON, P. TAXEL, J. H. CONTOIS, J. OTVOS et al., 1998 Oral estrogen improves serum lipids, homocysteine and fibrinolysis in elderly men. Atherosclerosis 137: 359-366.
GOODRICH, L. V., R. L. JOHNSON, L. MILENKOVIC, J. A. MCMAHON and M. P. SCOTT, 1996 Conservation of the hedgehog/patched signaling pathway from flies to mice: induction of a mouse patched gene by Hedgehog. Genes Dev 10: 301-312.
GORLIN, R. J., 1987 Nevoid basal-cell carcinoma syndrome. Medicine (Baltimore) 66: 98-113.
GOYETTE, P., B. CHRISTENSEN, D. S. ROSENBLATT and R. ROZEN, 1996 Severe and mild mutations in cis for the methylenetetrahydrofolate reductase (MTHFR) gene, and description of five novel mutations in MTHFR. Am J Hum Genet 59: 1268-1275.
GRAHAM, I. M., L. E. DALY, H. M. REFSUM, K. ROBINSON, L. E. BRATTSTROM et al., 1997 Plasma homocysteine as a risk factor for vascular disease. The European Concerted Action Project. Jama 277: 1775-1781.
GRAHAM, S., R. HELLMANN, J. MARSHALL, J. FREUDENHEIM, J. VENA et al., 1991 Nutritional epidemiology of postmenopausal breast cancer in western New York. Am J Epidemiol 134: 552-566. 216
GREEN, J. M., R. E. MACKENZIE and R. G. MATTHEWS, 1988 Substrate flux through methylenetetrahydrofolate dehydrogenase: predicted effects of the concentration of methylenetetrahydrofolate on its partitioning into pathways leading to nucleotide biosynthesis or methionine regeneration. Biochemistry 27: 8014-8022.
GREENE, N. D., and A. J. COPP, 1997 Inositol prevents folate-resistant neural tube defects in the mouse. Nat Med 3: 60-66.
GRESS, T. M., J. D. HOHEISEL, G. G. LENNON, G. ZEHETNER and H. LEHRACH, 1992 Hybridization fingerprinting of high-density cDNA-library arrays with cDNA pools derived from whole tissues. Mamm Genome 3: 609-619.
GROUP, M. V. S. R., 1991 Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. MRC Vitamin Study Research Group. Lancet 338: 131-137.
GULATI, S., P. BAKER, Y. N. LI, B. FOWLER, W. KRUGER et al., 1996 Defects in human methionine synthase in cblG patients. Hum Mol Genet 5: 1859- 1865.
GUTTORMSEN, A. B., H. REFSUM and P. M. UELAND, 1994 The interaction between nitrous oxide and cobalamin. Biochemical effects and clinical consequences. Acta Anaesthesiol Scand 38: 753-756.
HAJJAR, K. A., L. MAURI, A. T. JACOVINA, F. ZHONG, U. A. MIRZA et al., 1998 Tissue plasminogen activator binding to the annexin II tail domain. Direct modulation by homocysteine. J Biol Chem 273: 9987-9993.
HANKEY, G. J., and J. W. EIKELBOOM, 2000 Homocysteine and vascular disease. Indian Heart J 52: S18-26.
HARMON, D. L., D. C. SHIELDS, J. V. WOODSIDE, D. MCMASTER, J. W. YARNELL et al., 1999 Methionine synthase D919G polymorphism is a significant but modest determinant of circulating homocysteine concentrations. Genet Epidemiol 17: 298-309. 217
HARRIS, M. J., and D. M. JURILOFF, 1999 Mini-review: toward understanding mechanisms of genetic neural tube defects in mice. Teratology 60: 292- 305.
HASSOLD, T., and S. SHERMAN, 2000 Down syndrome: genetic recombination and the origin of the extra chromosome 21. Clin Genet 57: 95-100.
HEINECKE, J. W., 2001 Unique aspects of sulfur chemistry: homocysteine and lipid oxidation, pp. 32-38 in Homocysteine in health and disease, edited by D. W. JACOBSEN. Cambridge University Press, Cambridge.
HENNING, S. M., M. E. SWENDSEID and W. F. COULSON, 1997a Male rats fed methyl- and folate-deficient diets with or without niacin develop hepatic carcinomas associated with decreased tissue NAD concentrations and altered poly(ADP-ribose) polymerase activity. J Nutr 127: 30-36.
HENNING, S. M., M. E. SWENDSEID, B. T. IVANDIC and F. LIAO, 1997b Vitamins C, E and A and heme oxygenase in rats fed methyl/folate-deficient diets. Free Radic Biol Med 23: 936-942.
HERBIG, K., E. P. CHIANG, L. R. LEE, J. HILLS, B. SHANE et al., 2002 Cytoplasmic serine hydroxymethyltransferase mediates competition between folate- dependent deoxyribonucleotide and S-adenosylmethionine biosyntheses. J Biol Chem 277: 38381-38389.
HIBBARD, B. M., E. D. HIBBARD and T. N. JEFFCOATE, 1965 Folic acid and reproduction. Acta Obstet Gynecol Scand 44: 375-400.
HILLMAN, R. S., and S. E. STEINBERG, 1982 The effects of alcohol on folate metabolism. Annu Rev Med 33: 345-354.
HISHIDA, R., and H. NAU, 1998 VPA-induced neural tube defects in mice. I. Altered metabolism of sulfur amino acids and glutathione. Teratog Carcinog Mutagen 18: 49-61.
HO, P. I., D. ORTIZ, E. ROGERS and T. B. SHEA, 2002 Multiple aspects of homocysteine neurotoxicity: glutamate excitotoxicity, kinase hyperactivation and DNA damage. J Neurosci Res 70: 694-702. 218
HOBBS, C. A., S. L. SHERMAN, P. YI, S. E. HOPKINS, C. P. TORFS et al., 2000 Polymorphisms in genes involved in folate metabolism as maternal risk factors for Down syndrome. Am J Hum Genet 67: 623-630.
HOFMANN, M. A., E. LALLA, Y. LU, M. R. GLEASON, B. M. WOLF et al., 2001 Hyperhomocysteinemia enhances vascular inflammation and accelerates atherosclerosis in a murine model. J Clin Invest 107: 675-683.
HOIT, B. D., and J. H. NADEAU, 2001 Phenotype-driven genetic approaches in mice: high-throughput phenotyping for discovering new models of cardiovascular disease. Trends Cardiovasc Med 11: 82-89.
HONEIN, M. A., L. J. PAULOZZI, T. J. MATHEWS, J. D. ERICKSON and L. Y. WONG, 2001 Impact of folic acid fortification of the US food supply on the occurrence of neural tube defects. Jama 285: 2981-2986.
HORNE, D. W., 1997 Microbiological assay of folates in 96-well microtiter plates. Methods Enzymol 281: 38-43.
HORNE, D. W., D. PATTERSON and R. J. COOK, 1989 Effect of nitrous oxide inactivation of vitamin B12-dependent methionine synthetase on the subcellular distribution of folate coenzymes in rat liver. Arch Biochem Biophys 270: 729-733.
HUANG, R. F., Y. C. HSU, H. L. LIN and F. L. YANG, 2001 Folate depletion and elevated plasma homocysteine promote oxidative stress in rat livers. J Nutr 131: 33-38.
JACOBSEN, D. W., 2001 Practical chemistry of homocysteine and other thiols, pp. 9-20 in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
JACQUES, P. F., A. G. BOSTOM, J. SELHUB, S. RICH, R. C. ELLISON et al., 2003 Effects of polymorphisms of methionine synthase and methionine synthase reductase on total plasma homocysteine in the NHLBI Family Heart Study. Atherosclerosis 166: 49-55. 219
JACQUES, P. F., J. SELHUB, A. G. BOSTOM, P. W. WILSON and I. H. ROSENBERG, 1999 The effect of folic acid fortification on plasma folate and total homocysteine concentrations. N Engl J Med 340: 1449-1454.
JAMES, S. J., M. POGRIBNA, I. P. POGRIBNY, S. MELNYK, R. J. HINE et al., 1999 Abnormal folate metabolism and mutation in the methylenetetrahydrofolate reductase gene may be maternal risk factors for Down syndrome. Am J Clin Nutr 70: 495-501.
JAMES, S. J., I. P. POGRIBNY, M. POGRIBNA, B. J. MILLER, S. JERNIGAN et al., 2003 Mechanisms of DNA damage, DNA hypomethylation, and tumor progression in the folate/methyl-deficient rat model of hepatocarcinogenesis. J Nutr 133: 3740S-3747S.
JANNE, P. A., and R. J. MAYER, 2000 Chemoprevention of colorectal cancer. N Engl J Med 342: 1960-1968.
JOOSTEN, E., E. LESAFFRE, R. RIEZLER, V. GHEKIERE, L. DEREYMAEKER et al., 1997 Is metabolic evidence for vitamin B-12 and folate deficiency more frequent in elderly patients with Alzheimer's disease? J Gerontol A Biol Sci Med Sci 52: M76-79.
JURILOFF, D. M., and M. J. HARRIS, 2000 Mouse models for neural tube closure defects. Hum Mol Genet 9: 993-1000.
KANG, S. S., and P. W. WONG, 1996 Genetic and nongenetic factors for moderate hyperhomocyst(e)inemia. Atherosclerosis 119: 135-138.
KANG, S. S., P. W. WONG, H. G. BOCK, A. HORWITZ and A. GRIX, 1991 Intermediate hyperhomocysteinemia resulting from compound heterozygosity of methylenetetrahydrofolate reductase mutations. Am J Hum Genet 48: 546-551.
KANG, S. S., P. W. WONG and M. R. MALINOW, 1992 Hyperhomocyst(e)inemia as a risk factor for occlusive vascular disease. Annu Rev Nutr 12: 279-298. 220
KANG, S. S., P. W. WONG, J. M. ZHOU and H. Y. COOK, 1986 Total homocyst(e)ine in plasma and amniotic fluid of pregnant women. Metabolism 35: 889-891.
KANG, S. S., J. ZHOU, P. W. WONG, J. KOWALISYN and G. STROKOSCH, 1988 Intermediate homocysteinemia: a thermolabile variant of methylenetetrahydrofolate reductase. Am J Hum Genet 43: 414-421.
KANWAR, Y. S., J. R. MANALIGOD and P. W. WONG, 1976 Morphologic studies in a patient with homocystinuria due to 5, 10- methylenetetrahydrofolate reductase deficiency. Pediatr Res 10: 598-609.
KAPLAN, K. B., A. A. BURDS, J. R. SWEDLOW, S. S. BEKIR, P. K. SORGER et al., 2001 A role for the Adenomatous Polyposis Coli protein in chromosome segregation. Nat Cell Biol 3: 429-432.
KATO, I., A. M. DNISTRIAN, M. SCHWARTZ, P. TONIOLO, K. KOENIG et al., 1999 Serum folate, homocysteine and colorectal cancer risk in women: a nested case-control study. Br J Cancer 79: 1917-1922.
KENYON, S. H., A. NICOLAOU and W. A. GIBBONS, 1998 The effect of ethanol and its metabolites upon methionine synthase activity in vitro. Alcohol 15: 305- 309.
KIM, Y. I., I. P. POGRIBNY, R. N. SALOMON, S. W. CHOI, D. E. SMITH et al., 1996a Exon-specific DNA hypomethylation of the p53 gene of rat colon induced by dimethylhydrazine. Modulation by dietary folate. Am J Pathol 149: 1129-1137.
KIM, Y. I., R. N. SALOMON, F. GRAEME-COOK, S. W. CHOI, D. E. SMITH et al., 1996b Dietary folate protects against the development of macroscopic colonic neoplasia in a dose responsive manner in rats. Gut 39: 732-740.
KIRKE, P. N., A. M. MOLLOY, L. E. DALY, H. BURKE, D. G. WEIR et al., 1993 Maternal plasma folate and vitamin B12 are independent risk factors for neural tube defects. Q J Med 86: 703-708. 221
KISLIUK, R., 1999 Antifolate drugs in cancer therapy, pp. 13-36 in Anticancer drug development guide. Humana Press, New Jersey.
KLERK, M., P. VERHOEF, R. CLARKE, H. J. BLOM, F. J. KOK et al., 2002 MTHFR 677C-->T polymorphism and risk of coronary heart disease: a meta- analysis. Jama 288: 2023-2031.
KLUIJTMANS, L. A., L. P. VAN DEN HEUVEL, G. H. BOERS, P. FROSST, E. M. STEVENS et al., 1996 Molecular genetic analysis in mild hyperhomocysteinemia: a common mutation in the methylenetetrahydrofolate reductase gene is a genetic risk factor for cardiovascular disease. Am J Hum Genet 58: 35-41.
KOURY, M. J., D. J. PARK, D. MARTINCIC, D. W. HORNE, V. KRAVTSOV et al., 1997 Folate deficiency delays the onset but increases the incidence of leukemia in Friend virus-infected mice. Blood 90: 4054-4061.
KRAUS, J. P., M. JANOSIK, V. KOZICH, R. MANDELL, V. SHIH et al., 1999 Cystathionine beta-synthase mutations in homocystinuria. Hum Mutat 13: 362-375.
KRUMAN, II, C. CULMSEE, S. L. CHAN, Y. KRUMAN, Z. GUO et al., 2000 Homocysteine elicits a DNA damage response in neurons that promotes apoptosis and hypersensitivity to excitotoxicity. J Neurosci 20: 6920-6926.
KRUMAN, II, T. S. KUMARAVEL, A. LOHANI, W. A. PEDERSEN, R. G. CUTLER et al., 2002 Folic acid deficiency and homocysteine impair DNA repair in hippocampal neurons and sensitize them to amyloid toxicity in experimental models of Alzheimer's disease. J Neurosci 22: 1752-1762.
KUHN, W., T. HUMMEL, D. WOITALLA and T. MULLER, 2001 Plasma homocysteine and MTHFR C677T genotype in levodopa-treated patients with PD. Neurology 56: 281; author reply 281-282.
KUHN, W., R. ROEBROEK, H. BLOM, D. VAN OPPENRAAIJ and T. MULLER, 1998a Hyperhomocysteinaemia in Parkinson's disease. J Neurol 245: 811-812. 222
KUHN, W., R. ROEBROEK, H. BLOM, D. VAN OPPENRAAIJ, H. PRZUNTEK et al., 1998b Elevated plasma levels of homocysteine in Parkinson's disease. Eur Neurol 40: 225-227.
LAIRD, P. W., L. JACKSON-GRUSBY, A. FAZELI, S. L. DICKINSON, W. E. JUNG et al., 1995 Suppression of intestinal neoplasia by DNA hypomethylation. Cell 81: 197-205.
LAIRD, P. W., and R. JAENISCH, 1994 DNA methylation and cancer. Hum Mol Genet 3 Spec No: 1487-1495.
LAMPRECHT, S. A., and M. LIPKIN, 2003 Chemoprevention of colon cancer by calcium, vitamin D and folate: molecular mechanisms. Nat Rev Cancer 3: 601-614.
LE LEU, R. K., G. P. YOUNG and G. H. MCINTOSH, 2000 Folate deficiency diminishes the occurrence of aberrant crypt foci in the rat colon but does not alter global DNA methylation status. J Gastroenterol Hepatol 15: 1158- 1164.
LECLERC, D., E. CAMPEAU, P. GOYETTE, C. E. ADJALLA, B. CHRISTENSEN et al., 1996 Human methionine synthase: cDNA cloning and identification of mutations in patients of the cblG complementation group of folate/cobalamin disorders. Hum Mol Genet 5: 1867-1874.
LECLERC, D., A. WILSON, R. DUMAS, C. GAFUIK, D. SONG et al., 1998 Cloning and mapping of a cDNA for methionine synthase reductase, a flavoprotein defective in patients with homocystinuria. Proc Natl Acad Sci U S A 95: 3059-3064.
LEGGE, M., and M. H. SELLENS, 1991 Free radical scavengers ameliorate the 2- cell block in mouse embryo culture. Hum Reprod 6: 867-871.
LENTZ, S. R., R. A. ERGER, S. DAYAL, N. MAEDA, M. R. MALINOW et al., 2000 Folate dependence of hyperhomocysteinemia and vascular dysfunction in cystathionine beta-synthase-deficient mice. Am J Physiol Heart Circ Physiol 279: H970-975. 223
LENTZ, S. R., and J. E. SADLER, 1991 Inhibition of thrombomodulin surface expression and protein C activation by the thrombogenic agent homocysteine. J Clin Invest 88: 1906-1914.
LINDENBAUM, J., 1979 Aspects of vitamin B12 and folate metabolism in malabsorption syndromes. Am J Med 67: 1037-1048.
LOCKER, J., T. V. REDDY and B. LOMBARDI, 1986 DNA methylation and hepatocarcinogenesis in rats fed a choline-devoid diet. Carcinogenesis 7: 1309-1312.
LUCOCK, M., 2000 Folic acid: nutritional biochemistry, molecular biology, and role in disease processes. Mol Genet Metab 71: 121-138.
MA, J., M. J. STAMPFER, E. GIOVANNUCCI, C. ARTIGAS, D. J. HUNTER et al., 1997 Methylenetetrahydrofolate reductase polymorphism, dietary interactions, and risk of colorectal cancer. Cancer Res 57: 1098-1102.
MALINOW, M. R., A. RAJKOVIC, P. B. DUELL, D. L. HESS and B. M. UPSON, 1998 The relationship between maternal and neonatal umbilical cord plasma homocyst(e)ine suggests a potential role for maternal homocyst(e)ine in fetal metabolism. Am J Obstet Gynecol 178: 228-233.
MANLY, B. F., 1997 Randomization, bootstrap and Monte Carlo methods in biology. Chapman and Hall/CRC, Boca Raton.
MANSOOR, M. A., A. M. SVARDAL and P. M. UELAND, 1992 Determination of the in vivo redox status of cysteine, cysteinylglycine, homocysteine, and glutathione in human plasma. Anal Biochem 200: 218-229.
MATTHIAS, D., C. H. BECKER, R. RIEZLER and P. H. KINDLING, 1996 Homocysteine induced arteriosclerosis-like alterations of the aorta in normotensive and hypertensive rats following application of high doses of methionine. Atherosclerosis 122: 201-216.
MATTSON, M. P., 1997 Cellular actions of beta-amyloid precursor protein and its soluble and fibrillogenic derivatives. Physiol Rev 77: 1081-1132. 224
MAYER, E. L., D. W. JACOBSEN and K. ROBINSON, 1996 Homocysteine and coronary atherosclerosis. J Am Coll Cardiol 27: 517-527.
MCCANN, S. J., S. GILLINGWATER, B. G. KEEVIL, D. P. COOPER and M. R. MORRIS, 2003 Measurement of total homocysteine in plasma and blood spots using liquid chromatography-tandem mass spectrometry: comparison with the plasma Abbott IMx method. Ann Clin Biochem 40: 161-165.
MCCULLY, K. S., 1969 Vascular pathology of homocysteinemia: implications for the pathogenesis of arteriosclerosis. Am J Pathol 56: 111-128.
MEISTER, A., 1988 Glutathione metabolism and its selective modification. J Biol Chem 263: 17205-17208.
MELNYK, S., M. POGRIBNA, B. J. MILLER, A. G. BASNAKIAN, I. P. POGRIBNY et al., 1999 Uracil misincorporation, DNA strand breaks, and gene amplification are associated with tumorigenic cell transformation in folate deficient/repleted Chinese hamster ovary cells. Cancer Lett 146: 35-44.
MERCER, J. G., N. HOGGARD, L. M. WILLIAMS, C. B. LAWRENCE, L. T. HANNAH et al., 1996 Localization of leptin receptor mRNA and the long form splice variant (Ob-Rb) in mouse hypothalamus and adjacent brain regions by in situ hybridization. FEBS Lett 387: 113-116.
MIKOL, Y. B., K. L. HOOVER, D. CREASIA and L. A. POIRIER, 1983 Hepatocarcinogenesis in rats fed methyl-deficient, amino acid-defined diets. Carcinogenesis 4: 1619-1629.
MILLER, J. W., 1999 Homocysteine and Alzheimer's disease. Nutr Rev 57: 126- 129.
MILLER, J. W., M. R. NADEAU, D. SMITH and J. SELHUB, 1994a Vitamin B-6 deficiency vs folate deficiency: comparison of responses to methionine loading in rats. Am J Clin Nutr 59: 1033-1039.
MILLER, J. W., M. R. NADEAU, J. SMITH, D. SMITH and J. SELHUB, 1994b Folate- deficiency-induced homocysteinaemia in rats: disruption of S- 225
adenosylmethionine's co-ordinate regulation of homocysteine metabolism. Biochem J 298 ( Pt 2): 415-419.
MILLER, J. W., J. D. RIBAYA-MERCADO, R. M. RUSSELL, D. C. SHEPARD, F. D. MORROW et al., 1992 Effect of vitamin B-6 deficiency on fasting plasma homocysteine concentrations. Am J Clin Nutr 55: 1154-1160.
MILLS, J. L., J. M. MCPARTLIN, P. N. KIRKE, Y. J. LEE, M. R. CONLEY et al., 1995 Homocysteine metabolism in pregnancies complicated by neural-tube defects. Lancet 345: 149-151.
MINER, S. E., J. EVROVSKI and D. E. COLE, 1997 Clinical chemistry and molecular biology of homocysteine metabolism: an update. Clin Biochem 30: 189- 201.
MOLLOY, A. M., and J. M. SCOTT, 1997 Microbiological assay for serum, plasma, and red cell folate using cryopreserved, microtiter plate method. Methods Enzymol 281: 43-53.
MORGAN, W., 1954 A new crooked tail mutation involving distinctive pleiotropism. J Genet 52: 354-373.
MORIN, I., A. M. DEVLIN, D. LECLERC, N. SABBAGHIAN, C. H. HALSTED et al., 2003 Evaluation of genetic variants in the reduced folate carrier and in glutamate carboxypeptidase II for spina bifida risk. Mol Genet Metab 79: 197-200.
MORITA, H., H. KURIHARA, S. YOSHIDA, Y. SAITO, T. SHINDO et al., 2001 Diet- induced hyperhomocysteinemia exacerbates neointima formation in rat carotid arteries after balloon injury. Circulation 103: 133-139.
MORRIS, M. S., P. F. JACQUES, J. SELHUB and I. H. ROSENBERG, 2000 Total homocysteine and estrogen status indicators in the Third National Health and Nutrition Examination Survey. Am J Epidemiol 152: 140-148.
MORRISON, K., Y. H. EDWARDS, S. A. LYNCH, J. BURN, F. HOL et al., 1997 Methionine synthase and neural tube defects. J Med Genet 34: 958. 226
MOSHAROV, E., M. R. CRANFORD and R. BANERJEE, 2000 The quantitatively important relationship between homocysteine metabolism and glutathione synthesis by the transsulfuration pathway and its regulation by redox changes. Biochemistry 39: 13005-13011.
MUDD, S. H., J. D. FINKELSTEIN, F. IRREVERRE and L. LASTER, 1964 Homocystinuria: an enzymatic defect. Science 143: 1443-1445.
MUDD, S. H., H. L. LEVY and J. P. SKRAUS, 2001 Disorders of transsulfuration, pp. 2007-2056 in The metabolic and molecular bases of inherited disease, edited by C. R. SCRIVER, A. L. BEAUDET, W. S. SLY and D. VALLE. McGraw Hill, New York.
MUDD, S. H., F. SKOVBY, H. L. LEVY, K. D. PETTIGREW, B. WILCKEN et al., 1985 The natural history of homocystinuria due to cystathionine beta- synthase deficiency. Am J Hum Genet 37: 1-31.
MULLER, T., B. WERNE, B. FOWLER and W. KUHN, 1999 Nigral endothelial dysfunction, homocysteine, and Parkinson's disease. Lancet 354: 126- 127.
MURONE, M., A. ROSENTHAL and F. J. DE SAUVAGE, 1999 Sonic hedgehog signaling by the patched-smoothened receptor complex. Curr Biol 9: 76- 84.
NADEAU, J. H., J. B. SINGER, A. MATIN and E. S. LANDER, 2000 Analysing complex genetic traits with chromosome substitution strains. Nat Genet 24: 221- 225.
NYGARD, O., H. REFSUM, P. M. UELAND, I. STENSVOLD, J. E. NORDREHAUG et al., 1997 Coffee consumption and plasma total homocysteine: The Hordaland Homocysteine Study. Am J Clin Nutr 65: 136-143.
NYGARD, O., S. E. VOLLSET, H. REFSUM, I. STENSVOLD, A. TVERDAL et al., 1995 Total plasma homocysteine and cardiovascular risk profile. The Hordaland Homocysteine Study. Jama 274: 1526-1533. 227
O'LEARY, V. B., A. PARLE-MCDERMOTT, A. M. MOLLOY, P. N. KIRKE, Z. JOHNSON et al., 2002 MTRR and MTHFR polymorphism: link to Down syndrome? Am J Med Genet 107: 151-155.
O'NEILL, C., 1998 Endogenous folic acid is essential for normal development of preimplantation embryos. Hum Reprod 13: 1312-1316.
OGAWA, H., T. GOMI, F. TAKUSAGAWA and M. FUJIOKA, 1998 Structure, function and physiological role of glycine N- methyltransferase. Int J Biochem Cell Biol 30: 13-26.
PIEDRAHITA, J. A., B. OETAMA, G. D. BENNETT, J. VAN WAES, B. A. KAMEN et al., 1999 Mice lacking the folic acid-binding protein Folbp1 are defective in early embryonic development. Nat Genet 23: 228-232.
PIETRZIK, K., and A. BRONSTRUP, 1997 Causes and consequences of hyperhomocyst(e)inemia. Int J Vitam Nutr Res 67: 389-395.
POGRIBNA, M., S. MELNYK, I. POGRIBNY, A. CHANGO, P. YI et al., 2001 Homocysteine metabolism in children with Down syndrome: in vitro modulation. Am J Hum Genet 69: 88-95.
PRETLOW, T. P., M. A. O'RIORDAN, T. G. PRETLOW and T. A. STELLATO, 1992 Aberrant crypts in human colonic mucosa: putative preneoplastic lesions. J Cell Biochem Suppl 16G: 55-62.
PRINZ-LANGENOHL, R., I. FOHR and K. PIETRZIK, 2001 Beneficial role for folate in the prevention of colorectal and breast cancer. Eur J Nutr 40: 98-105.
PUFULETE, M., R. AL-GHNANIEM, A. J. LEATHER, P. APPLEBY, S. GOUT et al., 2003 Folate status, genomic DNA hypomethylation, and risk of colorectal adenoma and cancer: a case control study. Gastroenterology 124: 1240- 1248.
REED, T., M. R. MALINOW, J. C. CHRISTIAN and B. UPSON, 1991 Estimates of heritability of plasma homocyst(e)ine levels in aging adult male twins. Clin Genet 39: 425-428. 228
REFSUM, H., and P. M. UELAND, 1990 Clinical significance of pharmacological modulation of homocysteine metabolism. Trends Pharmacol Sci 11: 411- 416.
REFSUM, H., P. M. UELAND and S. KVINNSLAND, 1986 Acute and long-term effects of high-dose methotrexate treatment on homocysteine in plasma and urine. Cancer Res 46: 5385-5391.
REFSUM, H., P. M. UELAND, O. NYGARD and S. E. VOLLSET, 1998 Homocysteine and cardiovascular disease. Annu Rev Med 49: 31-62.
REFSUM, H., F. WESENBERG and P. M. UELAND, 1991 Plasma homocysteine in children with acute lymphoblastic leukemia: changes during a chemotherapeutic regimen including methotrexate. Cancer Res 51: 828- 835.
ROBINSON, K., 2001 Homocysteine and coronary artery disease, pp. 371-383 in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
RODGERS, G. M., and M. T. CONN, 1990 Homocysteine, an atherogenic stimulus, reduces protein C activation by arterial and venous endothelial cells. Blood 75: 895-901.
RODGERS, G. M., and W. H. KANE, 1986 Activation of endogenous factor V by a homocysteine-induced vascular endothelial cell activator. J Clin Invest 77: 1909-1916.
ROHAN, T. E., M. G. JAIN, G. R. HOWE and A. B. MILLER, 2000 Dietary folate consumption and breast cancer risk. J Natl Cancer Inst 92: 266-269.
RONCUCCI, L., S. MODICA, M. PEDRONI, M. G. TAMASSIA, M. GHIDONI et al., 1998 Aberrant crypt foci in patients with colorectal cancer. Br J Cancer 77: 2343-2348.
ROSEN, F., and C. A. NICHOL, 1962 Inhibition of the growth of an ame-thopterin- refractory tumor by dietary restriction of folic acid. Cancer Res 22: 495- 500. 229
ROSENBLATT, D. S., and W. A. FENTON, 2001 Inherited disorders of folate and cobalamin transport and metabolism, pp. 3897-3933 in The metabolic and molecular bases of inherited disease, edited by C. R. SCRIVER, A. L. BEAUDET, W. S. SLY and D. VALLE. McGraw Hill, New York.
ROSENQUIST, T. H., S. A. RATASHAK and J. SELHUB, 1996 Homocysteine induces congenital defects of the heart and neural tube: effect of folic acid. Proc Natl Acad Sci U S A 93: 15227-15232.
RUTHERFORD, S. L., 2000 From genotype to phenotype: buffering mechanisms and the storage of genetic information. Bioessays 22: 1095-1105.
SCHIRCH, V., and W. B. STRONG, 1989 Interaction of folylpolyglutamates with enzymes in one-carbon metabolism. Arch Biochem Biophys 269: 371-380.
SCHNEEDE, J., H. REFSUM and P. M. UELAND, 2000 Biological and environmental determinants of plasma homocysteine. Semin Thromb Hemost 26: 263- 279.
SCHNYDER, G., M. ROFFI, R. PIN, Y. FLAMMER, H. LANGE et al., 2001 Decreased rate of coronary restenosis after lowering of plasma homocysteine levels. N Engl J Med 345: 1593-1600.
SCOTT, J. M., J. J. DINN, P. WILSON and D. G. WEIR, 1981 Pathogenesis of subacute combined degeneration: a result of methyl group deficiency. Lancet 2: 334-337.
SELHUB, J., 1999 Homocysteine metabolism. Annu Rev Nutr 19: 217-246.
SESHADRI, S., A. BEISER, J. SELHUB, P. F. JACQUES, I. H. ROSENBERG et al., 2002 Plasma homocysteine as a risk factor for dementia and Alzheimer's disease. N Engl J Med 346: 476-483.
SHAW, G. M., E. M. VELIE and D. M. SCHAFFER, 1997 Is dietary intake of methionine associated with a reduction in risk for neural tube defect- affected pregnancies? Teratology 56: 295-299. 230
SINGER, J. B., A. E. HILL, L. C. BURRAGE, K. R. OLSZENS, J. SONG et al., 2004 Genetic dissection of complex traits with chromosome substitution strains of mice. Science 304: 445-448.
SIROTNAK, F. M., and B. TOLNER, 1999 Carrier-mediated membrane transport of folates in mammalian cells. Annu Rev Nutr 19: 91-122.
SMOLIN, L. A., and N. J. BENEVENGA, 1982 Accumulation of homocyst(e)ine in vitamin B-6 deficiency: a model for the study of cystathionine beta- synthase deficiency. J Nutr 112: 1264-1272.
SONG, J., A. MEDLINE, J. B. MASON, S. GALLINGER and Y. I. KIM, 2000a Effects of dietary folate on intestinal tumorigenesis in the apcMin mouse. Cancer Res 60: 5434-5440.
SONG, J., K. J. SOHN, A. MEDLINE, C. ASH, S. GALLINGER et al., 2000b Chemopreventive effects of dietary folate on intestinal polyps in Apc+/- Msh2-/- mice. Cancer Res 60: 3191-3199.
STABLER, S. P., P. D. MARCELL, E. R. PODELL, R. H. ALLEN and J. LINDENBAUM, 1986 Assay of methylmalonic acid in the serum of patients with cobalamin deficiency using capillary gas chromatography-mass spectrometry. J Clin Invest 77: 1606-1612.
STAMLER, J. S., and A. SLIVKA, 1996 Biological chemistry of thiols in the vasculature and in vascular-related disease. Nutr Rev 54: 1-30.
STEAD, L. M., M. E. BROSNAN and J. T. BROSNAN, 2000 Characterization of homocysteine metabolism in the rat liver. Biochem J 350 Pt 3: 685-692.
STEEGERS-THEUNISSEN, R. P., G. H. BOERS, H. J. BLOM, J. G. NIJHUIS, C. M. THOMAS et al., 1995 Neural tube defects and elevated homocysteine levels in amniotic fluid. Am J Obstet Gynecol 172: 1436-1441.
STEEGERS-THEUNISSEN, R. P., G. H. BOERS, F. J. TRIJBELS and T. K. ESKES, 1991 Neural-tube defects and derangement of homocysteine metabolism. N Engl J Med 324: 199-200. 231
STEEGERS-THEUNISSEN, R. P., G. H. BOERS, F. J. TRIJBELS, J. D. FINKELSTEIN, H. J. BLOM et al., 1994 Maternal hyperhomocysteinemia: a risk factor for neural-tube defects? Metabolism 43: 1475-1480.
STEEGERS-THEUNISSEN, R. P., N. C. WATHEN, T. K. ESKES, B. VAN RAAIJ-SELTEN and T. CHARD, 1997 Maternal and fetal levels of methionine and homocysteine in early human pregnancy. Br J Obstet Gynaecol 104: 20- 24.
STEINBERG, S. E., 1984 Mechanisms of folate homeostasis. Am J Physiol 246: G319-324.
STOLZENBERG-SOLOMON, R. Z., D. ALBANES, F. J. NIETO, T. J. HARTMAN, J. A. TANGREA et al., 1999 Pancreatic cancer risk and nutrition-related methyl- group availability indicators in male smokers. J Natl Cancer Inst 91: 535- 541.
SU, L. J., and L. ARAB, 2001 Nutritional status of folate and colon cancer risk: evidence from NHANES I epidemiologic follow-up study. Ann Epidemiol 11: 65-72.
SU, L. K., K. W. KINZLER, B. VOGELSTEIN, A. C. PREISINGER, A. R. MOSER et al., 1992 Multiple intestinal neoplasia caused by a mutation in the murine homolog of the APC gene. Science 256: 668-670.
SUH, J. R., A. K. HERBIG and P. J. STOVER, 2001 New perspectives on folate catabolism. Annu Rev Nutr 21: 255-282.
SWANSON, D. A., M. L. LIU, P. J. BAKER, L. GARRETT, M. STITZEL et al., 2001 Targeted disruption of the methionine synthase gene in mice. Mol Cell Biol 21: 1058-1065.
TAMAYO, P., D. SLONIM, J. MESIROV, Q. ZHU, S. KITAREEWAN et al., 1999 Interpreting patterns of gene expression with self-organizing maps: methods and application to hematopoietic differentiation. Proc Natl Acad Sci U S A 96: 2907-2912. 232
TANG, L., C. D. MAMOTTE, F. M. VAN BOCKXMEER and R. R. TAYLOR, 1998 The effect of homocysteine on DNA synthesis in cultured human vascular smooth muscle. Atherosclerosis 136: 169-173.
TANG, L. S., and R. H. FINNELL, 2003 Neural and orofacial defects in Folp1 knockout mice [corrected]. Birth Defects Res Part A Clin Mol Teratol 67: 209-218.
TOLNER, B., K. ROY and F. M. SIROTNAK, 1997 Organization, structure and alternate splicing of the murine RFC-1 gene encoding a folate transporter. Gene 189: 1-7.
TONSTAD, S., H. REFSUM, M. SIVERTSEN, B. CHRISTOPHERSEN, L. OSE et al., 1996 Relation of total homocysteine and lipid levels in children to premature cardiovascular death in male relatives. Pediatr Res 40: 47-52.
TONSTAD, S., H. REFSUM and P. M. UELAND, 1997 Association between plasma total homocysteine and parental history of cardiovascular disease in children with familial hypercholesterolemia. Circulation 96: 1803-1808.
TRAN, P., F. HIOU-TIM, P. FROSST, S. LUSSIER-CACAN, P. BAGLEY et al., 2002 The curly-tail (ct) mouse, an animal model of neural tube defects, displays altered homocysteine metabolism without folate responsiveness or a defect in Mthfr. Mol Genet Metab 76: 297-304.
TROEN, A. M., E. LUTGENS, D. E. SMITH, I. H. ROSENBERG and J. SELHUB, 2003 The atherogenic effect of excess methionine intake. Proc Natl Acad Sci U S A 100: 15089-15094.
TSAI, M. Y., M. BIGNELL, F. YANG, B. G. WELGE, K. J. GRAHAM et al., 2000 Polygenic influence on plasma homocysteine: association of two prevalent mutations, the 844ins68 of cystathionine beta-synthase and A(2756)G of methionine synthase, with lowered plasma homocysteine levels. Atherosclerosis 149: 131-137.
UBBINK, J. B., W. J. HAYWARD VERMAAK and S. BISSBORT, 1991 Rapid high- performance liquid chromatographic assay for total homocysteine levels in human serum. J Chromatogr 565: 441-446. 233
UELAND, P. M., 1995 Homocysteine species as components of plasma redox thiol status. Clin Chem 41: 340-342.
UELAND, P. M., M. A. MANSOOR, A. B. GUTTORMSEN, F. MULLER, P. AUKRUST et al., 1996 Reduced, oxidized and protein-bound forms of homocysteine and other aminothiols in plasma comprise the redox thiol status--a possible element of the extracellular antioxidant defense system. J Nutr 126: 1281S-1284S.
UELAND, P. M., and H. REFSUM, 1989 Plasma homocysteine, a risk factor for vascular disease: plasma levels in health, disease, and drug therapy. J Lab Clin Med 114: 473-501.
UELAND, P. M., H. REFSUM, R. MALE and J. R. LILLEHAUG, 1986 Disposition of endogenous homocysteine by mouse fibroblast C3H/10T1/2 Cl 8 and the chemically transformed C3H/10T1/2 MCA Cl 16 cells following methotrexate exposure. J Natl Cancer Inst 77: 283-289.
UNDAS, A., T. B. DOMAGALA, M. JANKOWSKI and A. SZCZEKLIK, 1999 Treatment of hyperhomocysteinemia with folic acid and vitamins B12 and B6 attenuates thrombin generation. Thromb Res 95: 281-288.
UPCHURCH, G. R., JR., G. N. WELCH, A. J. FABIAN, J. E. FREEDMAN, J. L. JOHNSON et al., 1997 Homocyst(e)ine decreases bioavailable nitric oxide by a mechanism involving glutathione peroxidase. J Biol Chem 272: 17012- 17017.
VAN ALLEN, M. I., D. K. KALOUSEK, G. F. CHERNOFF, D. JURILOFF, M. HARRIS et al., 1993 Evidence for multi-site closure of the neural tube in humans. Am J Med Genet 47: 723-743.
VAN DER GRIEND, R., D. H. BIESMA, F. J. HAAS, J. A. FABER, M. DURAN et al., 2000 The effect of different treatment regimens in reducing fasting and postmethionine-load homocysteine concentrations. J Intern Med 248: 223- 229.
VAN DER PUT, N. M., F. GABREELS, E. M. STEVENS, J. A. SMEITINK, F. J. TRIJBELS et al., 1998 A second common mutation in the methylenetetrahydrofolate 234
reductase gene: an additional risk factor for neural-tube defects? Am J Hum Genet 62: 1044-1051.
VAN DER PUT, N. M., R. P. STEEGERS-THEUNISSEN, P. FROSST, F. J. TRIJBELS, T. K. ESKES et al., 1995 Mutated methylenetetrahydrofolate reductase as a risk factor for spina bifida. Lancet 346: 1070-1071.
VAN DER PUT, N. M., E. F. VAN DER MOLEN, L. A. KLUIJTMANS, S. G. HEIL, J. M. TRIJBELS et al., 1997 Sequence analysis of the coding region of human methionine synthase: relevance to hyperhomocysteinaemia in neural-tube defects and vascular disease. Qjm 90: 511-517.
VAN STRAATEN, H. W., H. BLOM, M. C. PEETERS, A. M. ROUSSEAU, K. J. COLE et al., 1995 Dietary methionine does not reduce penetrance in curly tail mice but causes a phenotype-specific decrease in embryonic growth. J Nutr 125: 2733-2740.
VENTER, J. C., M. D. ADAMS, E. W. MYERS, P. W. LI, R. J. MURAL et al., 2001 The sequence of the human genome. Science 291: 1304-1351.
VENTURA, P., R. PANINI, C. VERLATO, G. SCARPETTA and G. SALVIOLI, 2000 Peroxidation indices and total antioxidant capacity in plasma during hyperhomocysteinemia induced by methionine oral loading. Metabolism 49: 225-228.
VERHOEF, P., and M. STAMPFER, 2001 Epidemiology of vascular and thrombbotic associations, pp. 357-370 in Homocysteine in health and disease, edited by R. CARMEL and D. W. JACOBSEN. Cambridge University Press, Cambridge.
WALD, D. S., M. LAW and J. K. MORRIS, 2002 Homocysteine and cardiovascular disease: evidence on causality from a meta-analysis. Bmj 325: 1202.
WALKER, M. C., G. N. SMITH, S. L. PERKINS, E. J. KEELY and P. R. GARNER, 1999 Changes in homocysteine levels during normal pregnancy. Am J Obstet Gynecol 180: 660-664. 235
WATANABE, M., J. OSADA, Y. ARATANI, K. KLUCKMAN, R. REDDICK et al., 1995 Mice deficient in cystathionine beta-synthase: animal models for mild and severe homocyst(e)inemia. Proc Natl Acad Sci U S A 92: 1585-1589.
WATERLAND, R. A., and R. L. JIRTLE, 2003 Transposable elements: targets for early nutritional effects on epigenetic gene regulation. Mol Cell Biol 23: 5293-5300.
WATKINS, D., and D. S. ROSENBLATT, 1989 Functional methionine synthase deficiency (cblE and cblG): clinical and biochemical heterogeneity. Am J Med Genet 34: 427-434.
WEISS, N., S. HEYDRICK, Y. Y. ZHANG, C. BIERL, A. CAP et al., 2002 Cellular redox state and endothelial dysfunction in mildly hyperhomocysteinemic cystathionine beta-synthase-deficient mice. Arterioscler Thromb Vasc Biol 22: 34-41.
WELCH, G. N., and J. LOSCALZO, 1998 Homocysteine and atherothrombosis. N Engl J Med 338: 1042-1050.
WENSTROM, K. D., G. L. JOHANNING, J. OWEN, K. E. JOHNSTON, S. ACTON et al., 2000a Amniotic fluid homocysteine levels, 5,10-methylenetetrahydrafolate reductase genotypes, and neural tube closure sites. Am J Med Genet 90: 6-11.
WENSTROM, K. D., G. L. JOHANNING, J. OWEN, K. E. JOHNSTON, S. ACTON et al., 2000b Role of amniotic fluid homocysteine level and of fetal 5, 10- methylenetetrahydrafolate reductase genotype in the etiology of neural tube defects. Am J Med Genet 90: 12-16.
WEST, M. G., D. W. HORNE and D. R. APPLING, 1996 Metabolic role of cytoplasmic isozymes of 5,10-methylenetetrahydrofolate dehydrogenase in Saccharomyces cerevisiae. Biochemistry 35: 3122-3132.
WHITEHEAD, A. S., P. GALLAGHER, J. L. MILLS, P. N. KIRKE, H. BURKE et al., 1995 A genetic defect in 5,10 methylenetetrahydrofolate reductase in neural tube defects. Qjm 88: 763-766. 236
WILCKEN, D. E., and B. WILCKEN, 1998 B vitamins and homocysteine in cardiovascular disease and aging. Ann N Y Acad Sci 854: 361-370.
WILSON, A., R. PLATT, Q. WU, D. LECLERC, B. CHRISTENSEN et al., 1999 A common variant in methionine synthase reductase combined with low cobalamin (vitamin B12) increases risk for spina bifida. Mol Genet Metab 67: 317-323.
WOLFF, G. L., R. L. KODELL, S. R. MOORE and C. A. COONEY, 1998 Maternal epigenetics and methyl supplements affect agouti gene expression in Avy/a mice. Faseb J 12: 949-957.
WRIGHT, S., 1968 Genetics and biometric foundations. University of Chicago Press, Chicago.
WU, L. L., and J. T. WU, 2002 Hyperhomocysteinemia is a risk factor for cancer and a new potential tumor marker. Clin Chim Acta 322: 21-28.
YANKNER, B. A., 1996 Mechanisms of neuronal degeneration in Alzheimer's disease. Neuron 16: 921-932.
YASUI, K., H. KOWA, K. NAKASO, T. TAKESHIMA and K. NAKASHIMA, 2000 Plasma homocysteine and MTHFR C677T genotype in levodopa-treated patients with PD. Neurology 55: 437-440.
YI, P., S. MELNYK, M. POGRIBNA, I. P. POGRIBNY, R. J. HINE et al., 2000 Increase in plasma homocysteine associated with parallel increases in plasma S- adenosylhomocysteine and lymphocyte DNA hypomethylation. J Biol Chem 275: 29318-29323.
ZHANG, S., D. J. HUNTER, S. E. HANKINSON, E. L. GIOVANNUCCI, B. A. ROSNER et al., 1999 A prospective study of folate intake and the risk of breast cancer. Jama 281: 1632-1637.
ZHAO, Q., R. R. BEHRINGER and B. DE CROMBRUGGHE, 1996 Prenatal folic acid treatment suppresses acrania and meroanencephaly in mice mutant for the Cart1 homeobox gene. Nat Genet 13: 275-283. 237
ZHAO, R., R. G. RUSSELL, Y. WANG, L. LIU, F. GAO et al., 2001 Rescue of embryonic lethality in reduced folate carrier-deficient mice by maternal folic acid supplementation reveals early neonatal failure of hematopoietic organs. J Biol Chem 17: 17.
ZHOU, J., J. MOLLER, C. C. DANIELSEN, J. BENTZON, H. B. RAVN et al., 2001 Dietary supplementation with methionine and homocysteine promotes early atherosclerosis but not plaque rupture in ApoE-deficient mice. Arterioscler Thromb Vasc Biol 21: 1470-1476.
ZMUDA, J. M., L. L. BAUSSERMAN, D. MACERONI and P. D. THOMPSON, 1997 The effect of supraphysiologic doses of testosterone on fasting total homocysteine levels in normal men. Atherosclerosis 130: 199-202.