Transcobalamin 766G Homozygosity, Raised Homocysteine, Raised Methylmalonic Acid and High Creatinine: A Dementia-predisposing Phenotype? Implications for Dementia and Alzheimer’s Disease

Research Article

Austin J Clin Neurol 2015;2(7): 1062.

Transcobalamin 766G Homozygosity, Raised Homocysteine, Raised Methylmalonic Acid and High Creatinine: A Dementia-predisposing Phenotype? Implications for Dementia and Alzheimer’s Disease

Percy ME¹*, Somerville MJ², Podemski L², Wright E¹, Colelli T¹, Cheung A¹, Leung A¹, Kitaygorodsky J¹ and Garcia A³*

¹Departments of Physiology and Obstetrics & Gynaecology, University of Toronto, Canada

²Department of Medical Genetics, University of Alberta, Canada

³Division of Geriatrics, Department of Medicine, Queen’s University, Canada

*Corresponding author: Percy ME, Departments of Physiology and Obstetrics & Gynaecology, University of Toronto, Surrey Place Centre, Room 216, 2 Surrey Place, Toronto, Ontario M5S 2C2, Canada

Garcia A, Division of Geriatrics, Department of Medicine,Queens University, St. Mary’s of the Lake Hospital,Kingston, PO Box 3600, 340 Union Street, Kingston,Ontario K7L 5A2, Canada

Received: April 25, 2015; Accepted: June 25, 2015; Published: June 30, 2015


Raised homocysteine may be a weak, nutritionally modifiable risk factor for Alzheimer’s disease, an etiologically complex disorder likely consisting of subtypes. To further understanding of homocysteine in dementia, we evaluated a set of variables known to reflect/affect the metabolism of vitamin B12 as risk factors for dementia development in a pilot, exploratory nested case/control study. These included raised homocysteine, raised methylmalonic acid, high creatinine and the C766G alleles of transcobalamin II (TCN2), the major blood transporter of B12. High creatinine reflects kidney dysfunction, and can affect levels of homocysteine and methylmalonic acid. In the absence of kidney dysfunction, abnormally high homocysteine is a marker of B12 and/or folate deficiency while raised methylmalonic acid reflects B12 deficiency. Transcobalamin produced from 766G may differ functionally from that produced from 766C. Participants were stringently-selected, community volunteers who at study entry were >age 65 and had high cognitive function. Among those who developed dementia within 6.39 years, all with raised serum homocysteine (>15μmol/L) at study entry also had co-occurring raised methylmalonic acid (>350nmol/L), high creatinine (>87μmol/L) and transcobalamin II 766G homozygosity. The odds ratios of TCN2 766G homozygosity plus raised homocysteine, raised methylmalonic acid, or high creatinine at study entry for dementia were high compared to factors singly. We have tentatively dubbed this four-factor cluster “the TCN2 dementiapredisposing phenotype, TDPP”. These original findings warrant investigation in larger samples. New questions to ask are if TDPP heralds a distinct subtype of impending Alzheimer’s and if TDPP might distinguish a group to target with nutritional intervention before onset of cognitive impairment.

Keywords: B12; Creatinine; Dementia; Folate; Homocysteine; Kidney; Methylmalonic acid; Transcobalamin


Genetic terms

A: Adenine; C: Cytosine; G; Guanine; T: Thymine; APOE: Apolipoprotein E Gene Symbol; e2,e3,e4: the three most common alleles of APOE; MTHFR: Methylenetetrahydrofolate Reductase Gene Symbol; MTHFR C677T: C-to-T Substitution at Position 677 of the Gene; 677C/C: MTHFR 677C Homozygote (having 2 C alleles); 677C/T: MTHFR 677C/T Heterozygote (having 1 C and 1 T allele); 677T/T: MTHFR 677T Homozygote (having 2 T alleles); MTHFR A1298C: A-to-C Substitution at Position 1298 of the Messenger Ribonucleic Acid (mRNA); 1298A/A: MTHFR 1298A Homozygote (having 2 A alleles); 1298A/C: MTHFR 1298A/C Heterozygote (having 1 A and 1 C allele); 1298C/C: MTHFR C Homozygote (having 2 C alleles); TCN2: Transcobalamin II Gene Symbol; TCN2 C766G: C-to-G Substitution at Position 766 in Relation to the first Nucleotide (+1) of the ATG Translation Initiation Codon of TCN2; 766C/C: TCN2 766C Homozygote (having 2 C alleles); 766C/G: TCN2 766G Heterozygote (having 1 C and 1 G allele); 766G/G: TCN2 766G Homozygote (i.e., having 2 G alleles)


MTHFR: Methylenetetrahydrofolate Reductase [NAD(P) H]; MTRR: [Methionine Synthase] Reductase (alternative name, Methionine Synthase); MS: Methionine Synthase (see MTRR); MUT: Methylmalonyl-CoA Mutase

Enzyme substrates

methyleneTHF: 5,10-methylenetetrahydrofolate; methylTHF: 5-methyltetrahydrofolate; THF: Tetrahydrofolate

Vitamins and co-factors

B2: Vitamin B2 (Riboflavin); B6: Vitamin B6 (Pyridoxine); B12: Vitamin B12 (Cobalamin); CoA: co-enzyme A; FAD: Flavin Adenine Dinucleotide, a derivative of B2; NADPH: Reduced form Nicotinamide Adenine Dinucleotide Phosphate

Indicators of homocysteine metabolism

Cr: Creatinine; HCY: Homocysteine; MMA: Methylmalonic Acid


α: Alpha - the probability of rejecting the null hypothesis given that it is true; AD: Alzheimer’s Disease; C: Centigrade; CI: Confidence Interval; O: Degree; ID#: Participant Identification Number; L: Liter; μmol: Micromole; nmol: Nanomol; pmol: Picomol; MCI: Mild Cognitive Impairment; MMSE: Mini-Mental State Examination; L: n.d: Not Determined; n.s: Not Significant; OR: Odds Ratio; P: Probability of observing an effect given that the null hypothesis is true; PCR: Polymerase Chain Reaction; RBC: Red Blood Cell; RFLP; Restriction Fragment Length Polymorphism; SAS: Statistical Analysis System; sec: Seconds


Overview of dementia, Alzheimer’s disease and homocysteinemia

Dementia is the most serious form of memory disorder affecting the elderly. Late-onset Alzheimer’s disease (AD) occurring after age 65 is the most common form of dementia. As of 2015, diagnostic and intervention efforts still focus on “prototype AD”, though variability in AD clinical presentation, progression and pathology point to the existence of AD subtypes [1]. Understanding the genetic nature of this variability is expected to be important in the application of “personalized medicine” to AD and other forms of dementia [1]. In this approach, an individual’s genetic profile along with other information will be used to guide decisions made in regard to dementia prevention, diagnosis, and treatment.

Homocysteinemia (raised levels of homocysteine in the serum or plasma) is considered to be one nutritionally modifiable risk factor that if treated may delay the onset of AD. However, intervention efforts in AD patients with B vitamin supplementation have been disappointing, and prospective studies of factors associated with homocysteinemia in the dementia field are warranted. In order to further our understanding of the role of homocysteinemia in dementia, we undertook a pilot study to evaluate a set of candidate “risk factors” known to reflect or affect the metabolism of vitamin B12 (cobalamin) and homocysteine metabolism. A brief review of information relevant to this study is presented next. Following this is a description of the study design, hypotheses, and reasons for choosing particular variables as candidate risk factors. [Note: Risk factors are variables associated with increased risk of disease; statistically significant associations should not be equated with causality.]

The enigmatic relation of homocysteine with dementia

Homocysteine (HCY) is a non-protein, sulfur-containing amino acid that is the major metabolite of the essential amino acid methionine. It is found in small quantities in all cells and plays an important role in human health [2-4]. Serum or plasma concentrations of HCY are determined in large part by the body’s status of folate, B12 and B6, because these B vitamins are cofactors or substrates for particular enzymes involved in HCY metabolism [5,6]. However, particular polymorphisms of various enzymes [7- 10], kidney dysfunction [11,12], and other factors [13-15] also can affect HCY concentrations. High dose supplements of folic acid and vitamins B12 and B6 can reduce homocysteine levels even in the absence of vitamin deficiency [6,16].

High levels of circulating HCY (denoted as homocysteinemia or hyperhomocysteinemia) precede and/or co-occur not only with heart disease and vascular disease [17,18], but also with cognitive deterioration in aging, elderly adults [5], dementia [19], AD (the most common form of dementia) [20,21], and vascular dementia [22], as well as with other neurological disorders [23,24]. However, such findings have not been found consistently [25]. Because elevated HCY is associated with deficiency of B12, B6 and/or folate [5,6], and even normal levels of HCY can be lowered with B vitamin supplementation [6,16], deficiencies of B vitamins are suspected of being causal or contributing factors in these disorders.

With respect to cardiovascular and neurological diseases, it still is not clear if homocysteinemia is a causal or contributing factor, a marker of underlying vitamin deficiency, or the result of some other factor or factors [20,21,26-27]. One reason for controversy is that clinical trials with B vitamin supplementation to alleviate homocysteinemia have had no or few clinical effects aside from HCY lowering [16,28]. However, one clinical trial involving participants with mild cognitive impairment (MCI), a precursor to dementia, B vitamin therapy was shown to reduce the rate of brain shrinkage which occurs during normal aging and also in persons with MCI and AD [29]. In this particular trial, B vitamins also appeared to slow cognitive and clinical decline in the MCI participants, particularly in those with elevated HCY [30,31]. The success of the latter study may be related to the fact that the trial involved participants with predementia (i.e., MCI) rather than relatively established disease.

Relationship between B12 deficiency, raised homocysteine and raised methylmalonic acid

B12 deficiency results in raised HCY and raised methylmalonic acid (MMA). Only two enzymatic reactions are dependent upon B12. The first involves the methylation of HCY to methionine by methionine synthase [MTRR] which uses methylcobalamin as a cofactor [32]. The reaction catalyzed by MTRR converts 5-methyltetrahydrofolate (methylTHF) into tetrahydrofolate (THF) while transferring a methyl group to HCY to form methionine. (MethylTHF is produced from 5,10-methylenetetrahydrofolate via the enzyme, methylenetetrahydrofolate reductase (MTHFR) which contains a bound flavin cofactor and uses NADP(H) as the reducing agent.) B12 is thus crucial for the cycling of folate in the folate cycle as well as for generating the methionine needed for synthesis of S-adenosylmethionine, the source of active methyl groups for methylation reactions [5,6]. Interference with the MTRR reaction from deficiency of methylcobalamin or methylTHF will result in buildup of HCY [5,6]. The second B12 dependent reaction catalyzes the conversion of L-methylmalonyl-CoA to succinyl-CoA by methylmalonyl-CoA mutase (MUT); this requires 5-deoxyadenosylcobalamin [33]. (Succinate is an intermediate of the energy generating Krebs cycle.) When B12 deficiency interferes with MUT, methylmalonyl-CoA accumulates, leading to buildup of methylmalonic acid (MMA). Buildup of methylmalonyl-CoA results in buildup of propionyl-CoA, the precursor of methylmalonyl-CoA. Through the catalytic action of citrate synthase, excessive propionyl- CoA results in the buildup of 2-methylcitric acid via condensation with oxaloacetate. The MUT reaction mediates the catabolism of propionyl-CoA derived from odd chain fatty acids and some branched chain amino acids [5,6].

Design of the present study

The study described in this paper used a nested case/control design. It took advantage of a previously characterized cohort of stringentlyselected, community volunteers who at study start had high cognitive function and were >age 65 years. Over an observation window of 6.39 years, it was possible to distinguish those who developed dementia by the study end (the cases) and those who still had normal cognitive function (the controls) from others who became afflicted only with MCI [34,35].

Hypotheses of the present study

Because previous studies indicated that mild homocysteinemia is a relatively weak risk factor for dementia [19], we hypothesized that the association strength of homocysteinemia for dementia might be increased by considering raised HCY in combination with a genetic polymorphism thought to affect B12 and HCY metabolism as a risk factor. Additionally, we assumed that if B vitamin and/or folate abnormalities were contributing to homocysteinemia, then levels of other variables reflecting or affecting HCY status also should be evaluated. The rationales for selecting particular candidate genetic and biochemical risk factors for development of dementia are discussed next.

Genetic factors selected for evaluation in the development of dementia

Genetic factors selected as potential risk factors for development of dementia in the present study included the C766C and 766Gpolymorphisms of transcobalamin II (TCN2) the major blood transporter of vitamin B12 to tissues and cells. The TCN2 C766G polymorphisms were selected for evaluation for two main reasons. First, compared to common polymorphisms of MTHFR [36,37], TCN2 C766G polymorphisms have been understudied in the dementia field. Zetterberg and colleagues reported that 766G was strongly associated with lower levels of holo-transcobalamin (the B12-transcobalamin complex) compared to other 766 genotypes in the cerebrospinal fluid of patients with AD, in spite of normal vitamin levels in the peripheral blood [38]. McCaddon et al. had observed that “proportionately fewer 776C homozygote’s appear to develop AD at any given age [than other TCN2 genotypes], but this will require confirmation in a longitudinal study” [39]. Second, as detailed below, other evidence is suggestive that the 766C and 766G alleles may produce transcobalamins that do not function identically, making them plausible candidates for evaluation.

Although a number of TCN2 alleles have been described, the C766G polymorphisms occur most commonly in the general population, though their frequencies and relative proportion vary markedly from one geographical region to another and among different ethnic groups [40]. The 766G allele contains a cytosine (C) to guanine (G) substitution at nucleotide position 766 (in relation to the first nucleotide (+1) of the ATG translation initiation codon of TCN2. This nucleotide substitution converts a proline to an arginine residue at position 259 of the transcobalamin protein [41,42]. While no basic research studies have been published to indicate that the proline and arginine variant transcobalamins are functionally different, this is implied from various clinical studies distinct from those described above. For example, Miller and colleagues concluded that the TCN2 genotype may influence susceptibility to B12 deficiency [43]. Subjects homozygous for the proline form of transcobalamin [i.e., who were TCN2 766C homozygote’s] had a significantly higher mean concentration of bound B12 than persons homozygous or heterozygous for the arginine form. As well, persons homozygous or heterozygous for the proline form of transcobalamin [i.e., who were TCN2 766C homozygote’s or 766C/G heterozygote’s] had a significantly lower mean level of methylmalonic acid (MMA, a metabolite that becomes elevated in B12 deficiency), than those homozygous for the arginine form [43]. Conversely, TCN2 776Ghomozygosity has been associated with higher MMA than other 776 genotypes [44]. Although the majority of studies have failed to find 766G genotype effects on serum or plasma HCY [44], HCY concentrations were higher in TCN2 776G homozygote’s with evidence of B12 deficiency than in individuals with other genotypes [9]. One working explanation for such genotype influences is that B12 may have a higher affinity for the proline form of transcobalamin than the arginine form [43]. Another is that the proline form may be more efficient in delivery of B12 to cells and tissues [9].

Biochemical factors at study entry selected for evaluation in development of dementia

Because TCN2 766G influences may manifest in combination with vitamin B12 deficiency (see above), we evaluated biochemical measures commonly used to probe B12 and folate status as risk factors for the development of dementia in our participants. Of the battery of standard and specialized tests already completed for the cohort study [34], we selected measures of serum HCY, B12, MMA and red cell folates for evaluation. High HCY and MMA are used as surrogate markers of B12 deficiency [45]. Folate deficiency [7] and high folate in the presence of B12 deficiency [46] are associated with high HCY. Because high creatinine, an indicator of kidney dysfunction, is often associated with high HCY and MMA [11,47], complicating the use of high MMA or HCY as B12 deficiency indicators [12], serum creatinine also was selected as a candidate risk factor.



This study involved: inspection of data and graphical analysis to search for trends; use of contingency table analysis and odds ratio (OR) analysis to determine association strengths of putative risk factors for dementia singly and in combination; and other analyses such as comparison of means. Results of statistical analyses were interpreted at the 95% level of confidence (α = 0.05). Bonferroni corrections to reduce the chance of obtaining spurious results through multiple comparisons were not applied since the study was pilot and exploratory in nature. As noted, the candidate risk factors selected for evaluation are related in that they all affect the metabolism of HCY. Hence the obtaining of multiple, positive results should not be surprising.


Study participants and ethics approval for involvement of human subjects have been described previously [34,35]. In brief, participants (N=281) were Caucasian, community volunteers. At study entry they were: >65 years old (average, 72.6); free of cerebrovascular, neurological and chronic kidney disease (participants with serum creatinine >146μmol/L were excluded); had a Mini-Mental State Examination (MMSE) score >25; and were not taking excessively high doseB12 supplement or getting B12 vitamin injections. After study entry, they had had up to three follow-up visits for assessment of cognitive function and health status over a period of 6.39 years. At the study end, 26 (15 females, 11 males) were classified as: having developed dementia (the cases), 192 (147 females, 45 males) as being still cognitively normal (the controls) while 22 had developed only MCI. Of this sample, a random subgroup of 15/26 cases (7 females, 8 males) and 171/192 controls (110 females, 36 males) underwent genetic analysis (see Genotyping).

The study began at approximately the same time that folic acid fortification of grain products was mandated in Canada [34,35] January 1999, and continued until mid-2005.For information, B2 (riboflavin) fortification of certain foods has been mandatory in North America for more than half a century [48].


Preparation of genomic DNA from peripheral blood was described previously [35]. The focus was on TCN2C766G polymorphisms but for comparison MTHFR typing also was done. Genotypes resulting from the TCN2 766C and G alleles and the MTHR 677C and T alleles were identified using the RFLP-PCR procedures of Miller et al. [43], and Bottiglierri et al. [49], respectively. Genotypes resulting from the MTHFR 1298A and C alleles were identified using a TaqMan allelic discrimination assay. Briefly, labeled allele-specific probes (1298C_FAM: ACACTTGCTTCACTGGT, 1298A_VIC: AGACACTTTCTTCACTGGT) flanked by common amplification primers (F: AGAGCAAGTCCCCCAAGGA, R: CTTTGTGACCATTCCGGTTTG), were used in an amplification reaction on a GeneAmp 9700 (Life Technologies) with an initial incubation at 95OC for 105 sec, followed by 35 cycles of 95°C for 15 sec and 60°C for 60 sec, and endpoint quantitative analysis on a 7900HT Fast Genetic Analyzer (Life Technologies). TCN2 677C-G and MTHFR 677C-T typing was done in the Neurogenetics Laboratory at Surrey Place Centre, Toronto. MTHFR 1298A-C typing was done by M.J.S. and L.P. in the University of Alberta department of Medical Genetics.

Biochemical tests

Blood samples for the standard and specialized tests were nonfasting and taken by a single venipuncture at study entry. Total serum cobalamin (B12), red blood cell (RBC) folates, and serum creatinine were determined using standard procedures by the Kingston General Hospital Laboratory (Kingston, ON) [34,50]. Levels of HCY and MMA were measured in once-frozen serum samples using specialized procedures by Metabolite Laboratories, Inc. (Denver, CO) [5,34,50]. Serum preparation involved clotting at room temperature for 30 minutes prior to centrifugation [50]. Reference ranges for the study period were: cobalamin (B12), 165-740pmol/L; RBC folates, 750-1800nmol/L; total HCY, 5.1-13.9μmol/L; MMA, 73-271nmol/L; creatinine, 50-110μmol/L [34]. HCY levels are to some extent methodology dependent and measures in serum and plasma are not directly comparable [51]. We used serum samples prepared in the same way [50] to control for possible influences of environmental factors (e.g., ambient temperature, diurnal variation, centrifugation parameters) on biochemical variables within individuals [52].

Statistical analyses

We asked if particular genetic polymorphisms or genotypes or abnormal levels of biochemical markers of HCY metabolism at study entry were risk factors for dementia that developed by the study end (i.e., over the observation window of 6.39 years). The statistical methods that were used have mostly been described [35]. These include: comparison of the frequencies of candidate risk factors in the case and control groups using Fisher’s exact test (two-tailed); comparison of means using the independent samples t-test (twotailed, unequal variance); determination of odds ratios (ORs) and their 95% confidence intervals (CIs) to denote the strength of association between potential risk factors and outcomes; and post-hoc power analysis. MedCalc software was used for OR determinations [53]. (This software generates ORs identical to those produced by SAS.) For comparison of means, continuous variables were logarithmically transformed to reduce skewness. Analyses were conducted on up to 218 individuals whose blood work had been completed and on up to 186 individuals who also underwent genotyping (see Participants). Levels of red cell folates at study entry correlated strongly and positively with duration of exposure to mandatory folic acid fortification prior to study entry in persons who developed dementia and who retained normal cognitive function at study end, but levels of HCY, MMA, B12 or creatinine did not (data not shown).

Cut-off values for biochemical variables

Measures of HCY were denoted as raised if they exceeded 15μmol/L(the upper limit for “healthy” adults and in the range for “mild homocysteinemia” [51,54]); MMA levels were denoted as raised if they exceeded 350nmol/L (considered “elevated” in some previous studies [55]) Measures of creatinine were denoted as raised if they exceeded the upper lab limit of 110μmol/L. Measures of B12 were denoted as deficient if they were less than the lower lab limit of 165pmol/L. We also considered: creatinine measures within the top 15% of values for participants (>87μmol/L, denoted as “high”); and B12 levels within the lower 25% for participants (<210pmol/L, denoted as “low”).

Results and Discussion

Results are summarized in Tables 1-5.

Among those who developed dementia during the study, raised homocysteine, raised methylmalonic acid and high creatinine at study entry are associated with TCN2 766G homozygosity.

Inspection of Table 1 revealed substantial representation of TCN2 766G homozygosity among the 15 persons who developed dementia: 7 of the 15 (47%) had this marker. Of the TCN2 766G homozygote’s, 5of 6 whose blood results were complete had MMA >350nmol/L at study entry. Four of the 5with raised MMA had HCY >15μmol/L, and these same 4 had serum creatinine >87μmol/L) at study entry, possibly indicative of mild kidney dysfunction. Of those with raised HCY, raised MMA and high creatinine, one person had B12 <165pmol/L, two had B12 <210pmol/L plus low folate <750nmol.L) and one had raised folate >1800nmol/L.