Cobalamin (vitamin B12) is an essential cofactor for several enzyme systems in mammals, and adequate concentrations of this cofactor are required for nucleic acid synthesis.1 Animals are unable to synthesize cobalamin and therefore depend entirely on dietary sources.1 Absorption of cobalamin from the gastrointestinal tract is a complex process. First, cobalamin is bound to haptocorrin, then it is bound to gastric or pancreatic intrinsic factor, and then it is transferred to receptors on the surface of enterocytes in the ileum.2
Hypocobalaminemia can develop in dogs for several reasons, including pancreatic and intestinal disease.3 In humans, cobalamin deficiency attributable to selective malabsorption is a rare autosomal-recessive hereditary disorder that can be detected in early childhood.4,5 Hereditary cobalamin deficiency has been detected in Giant Schnauzers, Australian Shepherd Dogs, and Chinese Shar-Peis.6–8 Moreover, other authors have diagnosed cobalamin deficiency in Border Collies9–11 and 1 Beagle.12 Cobalamin is a cofactor for the conversion of methylmalonyl-CoA to succinyl-CoA via methylmalonyl-CoA mutase and for remethylation of homocysteine via methionine synthase.1 Cobalamin deficiency causes a reduction in activity of both of these enzymes, resulting in increases in methylmalonic acid and total homocysteine concentrations.1 Measurement of these metabolites allows assessment of availability of cobalamin for cells and is the test of choice to detect early or mild cobalamin deficiency in humans.13
Correlation of urine methylmalonic acid and plasma total homocysteine concentrations with serum cobalamin concentrations in dogs has not been determined, to the authors' knowledge. In addition, reported reference ranges for serum cobalamin concentrations in dogs have not been compared to reference ranges for urine methylmalonic acid or plasma total homocysteine concentrations. Because we have diagnosed cobalamin deficiency in Border Collies that had nonspecific clinical signs, we believed that cobalamin deficiency might be more prevalent than currently recognized. Thus, the purpose of the study reported here was to establish reference ranges for serum cobalamin, urine methylmalonic acid, and plasma total homocysteine concentrations in healthy dogs and to determine concentrations of these markers of cobalamin metabolism in Border Collies.
Materials and Methods
Animals—From July 2009 through September 2010, 113 purebred Border Collies were tested for cobalamin deficiency at the Clinic for Small Animal Internal Medicine, Vetsuisse Faculty, University of Zurich. Border Collie owners were recruited for participation in the study via advertisements on the Swiss Border Collie club website homepage and in Swiss dog magazines and by referring veterinarians who were informed of the study. Assessment of each Border Collie included obtaining a detailed history and performance of a physical examination, CBC, serum biochemical analysis, and urinalysis.
Thirty-five healthy control dogs were recruited for the study; these dogs were owned by staff and students of the clinic. Inclusion criteria for these dogs were as follows: breeds other than Border Collie or Border Collie crossbred, no history of disease within 12 months prior to recruitment, judged to be healthy by the owner, and unremarkable results of physical examination, CBC, serum biochemical analyses, and urinalysis. This group of healthy control dogs comprised 19 mixed-breed dogs, 3 Labrador Retrievers, 2 Golden Retrievers, and 11 dogs of other pedigree breeds. Median age of these dogs was 5 years (range, 1 to 15 years), and median body weight was 12.6 kg (range, 5.1 to 43 kg). These dogs included 9 sexually intact females, 5 sexually intact males, 9 spayed females, and 12 neutered males.
An additional 12 healthy dogs that were fed bone and raw food exclusively were included in the study. These dogs included 2 Australian Shepherd Dogs, 1 Jack Russell Terrier, 1 Alaskan Malamute, 1 Tervueren, 1 Airedale Terrier, and 6 mixed-breed dogs. Median age of these dogs was 5.4 years (range, 1.9 to 13.3 years), and median body weight was 22.7 kg (range, 6.1 to 39.7 kg). These dogs included 1 sexually intact female, 1 sexually intact male, 7 spayed females, and 3 neutered males. The study was approved by the Committee for the Permission of Animal Experimentation, Canton of Zurich, Zurich, Switzerland. Owners provided informed consent for inclusion of their dogs in the study.
Study protocol—Food was withheld from all dogs for 8 to 12 hours prior to obtaining blood samples. Blood (8 mL) was obtained from each dog via jugular venipuncture and placed into blood collection tubes.a Urine samples were collected by owners the evening or morning before dogs were brought to the clinic. Owners were instructed to store urine samples at 4° to 6°C. Paired urine samples were collected by owners from 6 of the dogs immediately before and 8 hours after food consumption for determination of the effect of food intake on urinary methylmalonic acid excretion.
Determination of serum cobalamin, plasma total homocysteine, and urine methylmalonic acid concentrations—For determination of serum cobalamin concentrations, blood samples were allowed to clot at room temperature (approx 22°C) for 20 minutes and then centrifuged at 1,500 × g for 10 minutes. Serum was harvested and stored at −80°C until measurement. Serum cobalamin concentrations were measured by use of an automated chemiluminescence assayb as previously described.8 The upper limit of detection of this assay was 1,000 ng of cobalamin/L; serum samples were diluted ≥ 1:2 for analysis if the cobalamin concentration was greater than the upper detection limit. The intra-assay and interassay CVs were 2.1% and 3.4%, respectively. The lower detection limit of the assay was 150 ng of cobalamin/L.
Plasma total homocysteine concentrations were determined by use of high-performance liquid chromatography and fluorometric detection.14 Blood samples collected into chilled tubes containing sodium citratea were centrifuged immediately after collection at 1,570 × g at 4°C for 10 minutes. Plasma was obtained and stored at −80°C until performance of assays. Homocysteine was added to a pooled sample of citrated canine plasmac (final homocysteine concentration, 100 μmol/L). This pooled plasma sample was sequentially diluted to prepare standards with homocysteine concentrations of 50, 25, 12.5, 5.0, and 2.5 μmol/L. Aliquots of these diluted standards were stored at −80°C until performance of assays; these standards were used to generate standard curves for each assay. Recoveries were assessed by including 3 standards (25, 12.5, and 5.0 μmol homocysteine/L) as samples in 5 assays during the 3-week testing period. Recoveries were > 96% for each standard tested. Because no quality-control samples were commercially available for plasma total homocysteine concentration assays, a pooled sample of canine plasma was tested for each assay (mean homocysteine concentration, 16.8 μmol/L). The between-assay CV for homocysteine concentration in this plasma sample was < 6%. The within-assay CV was < 3% and < 6% for the 50 and 5 μmol of homocysteine/L standards, respectively. The lower limit of detection of this assay was 2.5 μmol of homocysteine/L.
Urine samples were stored at −80°C until measurement. Urine methylmalonic acid concentrations were determined by use of gas chromatography and mass spectrometryd with a lower limit of detection of 0.15 mmol.15 Results were expressed as millimoles of urine methylmalonic acid per mole of urine creatinine. Creatinine concentrations were determined via the Jaffe method with an analyzer.e This method had been previously validated for determination of methylmalonic acid concentrations in canine urine samples at the University School of Veterinary Medicine, Giessen, Germany with the same model instrument that was used in the present study.
Statistical analysis—Data were analyzed by use of software.f Each data set was evaluated for normality with the Kolmogorov-Smirnov test. Within the 2 groups (Border Collies and healthy control dogs), serum cobalamin, urine methylmalonic acid, and plasma total homocysteine concentrations and results of a CBC and serum biochemical analyses were compared by use of the Mann-Whitney U test. The Spearman rank correlation coefficient was used to determine the relationship between urine methylmalonic acid, serum cobalamin, and plasma total homocysteine concentrations for both of those groups of dogs. Values of P < 0.05 were considered significant. Reference ranges were established by use of the nonparametric percentile method. The 2.5 and 97.5 percentiles were determined to achieve the 95% 2-sided reference interval for serum cobalamin and plasma total homocysteine concentrations. For urine methylmalonic acid concentration, the 95th percentile was used to determine a 1-sided reference range (ie, the lower limit of the reference range was 0 mmol/mol of creatinine; the upper limit of the reference range was the 95th percentile value). Serum cobalamin and urine methylmalonic acid concentrations outside the working range of the assays were assumed to be 149 ng/L and 1.9 mmol/mol of creatinine, respectively. Only values for samples obtained from healthy control dogs were used to calculate the reference limit and ranges.
Results
Healthy control dogs—Cobalamin concentrations in serum samples obtained from healthy control dogs ranged from 261 to 1,001 ng/L (median, 441 ng/L; mean ± SD, 540.5 ± 235.5 ng/L; Figure 1). The calculated reference range for serum cobalamin concentration was 261 to 1,001 ng/L.
Cobalamin concentrations in serum samples obtained from healthy Border Collies (white circles; n = 109), Border Collies with clinical signs and clinicopathologic evidence of cobalamin deficiency (asterisks; 4), and healthy control dogs (black diamonds; 35). Median values for Border Collies and control dogs are indicated by horizontal lines. The reference range (gray shading; 261 to 1,001 ng/L) was calculated as the middle 95% of values for the group of healthy control dogs.
Citation: American Journal of Veterinary Research 73, 8; 10.2460/ajvr.73.8.1194
Methylmalonic acid concentrations in urine samples obtained from healthy control dogs ranged from < 2 to 6.6 mmol/mol of creatinine (median, 1.9 mmol/mol of creatinine; mean ± SD, 2.1 ± 0.8 mmol/mol of creatinine; Figure 2); 32 of the dogs had a urine methylmalonic acid concentration < 2 mmol/mol of creatinine, and 3 dogs had urine methylmalonic acid concentrations of 2.5, 3.4, and 3.6 mmol/mol of creatinine. The upper reference limit calculated for urine methylmalonic acid concentration was 4.2 mmol/mol of creatinine. Food intake had no effect on methylmalonic acid concentrations in urine samples obtained from the 6 dogs for which paired samples were available; concentration of methylmalonic acid in each of these urine samples was < 2 mmol/mol of creatinine.
Methylmalonic acid concentrations in urine samples obtained from the dogs in Figure 1. The upper reference limit (4.2 mmol/mol of creatinine) of the reference range (gray shading) was calculated as the 95th percentile value for the group of healthy control dogs. Notice that the y-axis is not continuous and not linear. †Urine methylmalonic acid concentrations for the 109 healthy Border Collies that did not have signs of cobalamin deficiency were significantly (P < 0.001) higher than they were for the 35 healthy control dogs. MMA = Methylmalonic acid. See Figure 1 for remainder of key.
Citation: American Journal of Veterinary Research 73, 8; 10.2460/ajvr.73.8.1194
Total homocysteine concentrations in plasma samples obtained from healthy control dogs ranged from 4.3 to 18.4 μmol/L (median, 9.1 μmol/L; mean ± SD, 10.4 ± 4.5 μmol/L; Figure 3). The calculated reference range for plasma total homocysteine concentration was 4.3 to 18.4 μmol/L. No correlations were detected (Spearman rank correlation coefficient) between serum cobalamin and plasma total homocysteine concentrations, serum cobalamin and urine methylmalonic acid concentrations, or urine methylmalonic acid and plasma total homocysteine concentrations. Results for the 12 dogs that had been fed bone and raw food exclusively were similar to results for the 35 healthy control dogs.
Total homocysteine concentrations in plasma samples obtained from the dogs in Figure 1. The reference range (gray shading; 4.3 to 18.4 μmol/L) was calculated as the middle 95% of values for the group of healthy control dogs. See Figure 1 for remainder of key.
Citation: American Journal of Veterinary Research 73, 8; 10.2460/ajvr.73.8.1194
Healthy Border Collies—One hundred nine Border Collies were determined to be healthy. None of these dogs had received cobalamin during the study from any source other than their typical diet (ie, supplemental cobalamin had not been administered). All of these dogs were in athletic physical fitness, and no abnormalities were detected during physical examination. Results of a CBC, serum biochemical analyses, and urinalysis were unremarkable for each of these dogs. Median age of these dogs was 4 years (range, 0.2 to 14 years), and there were 32 sexually intact males, 30 sexually intact females, 28 spayed females, and 19 neutered males. Median body weight of these dogs was 17.3 kg (range, 2.7 to 29 kg). Median cobalamin concentration in serum samples obtained from these dogs was 592 ng/L (range, 150 to 1,855 ng/L; mean ± SD, 641.4 ± 304.5 ng/L; Figure 1), which was not significantly different from that in serum samples obtained from healthy control dogs.
Methylmalonic acid concentrations in urine samples obtained from healthy Border Collies ranged from < 2 to 360 mmol/mol of creatinine (median, 1.9 mmol/mol of creatinine; mean ± SD, 23.7 ± 60.1 mmol/mol of creatinine; Figure 2); 47 of the 109 (43.1%) healthy Border Collies had a urine methylmalonic acid concentration > 2 mmol/mol of creatinine (range, 3.2 to 360 mmol/mol of creatinine), and 41 (37.6%) dogs had a urine methylmalonic acid concentration higher than the upper reference limit (4.2 mmol/mol of creatinine). Methylmalonic acid concentrations in urine samples obtained from healthy Border Collies were significantly (P < 0.001) higher than they were in urine samples obtained from healthy control dogs.
Creatinine concentrations in urine samples obtained from the 41 Border Collies with an elevated urine methylmalonic acid concentration were not significantly different from those in urine samples obtained from the 68 Border Collies with a urine methylmalonic acid concentration within the reference range. Total homocysteine concentrations in plasma samples obtained from the 109 healthy Border Collies ranged from 2.8 to 22.4 μmol/L (median, 8.5 μmol/L; mean ± SD, 9.5 ± 4.0 μmol/L; Figure 3) and were not significantly different from those in plasma samples obtained from healthy control dogs. Five of the healthy Border Collies had a serum cobalamin concentration lower than the reference range (261 to 1,001 ng/L); serum cobalamin concentrations ranged from 150 to 259 ng/L (median, 251 ng/L) in these dogs. These 5 Border Collies had urine methylmalonic acid and total homocysteine concentrations within the reference ranges.
Serum cobalamin and plasma total homocysteine concentrations of the 47 healthy Border Collies with a urine methylmalonic acid concentration > 2 mmol/mol of creatinine were not significantly different from those of healthy control dogs. Spearman rank correlation coefficient analysis did not reveal significant correlations among serum cobalamin, plasma total homocysteine, or urine methylmalonic acid concentrations in healthy Border Collies or in Border Collies with a urine methylmalonic acid concentration higher than the upper reference limit.
Border Collies with cobalamin deficiency—Cobalamin deficiency was diagnosed in 4 of the 113 Border Collies. Median age of these 4 Border Collies was 11.5 months (range, 8 to 42 months), median body weight was 11.6 kg (range, 11 to 12.1 kg), and all of these dogs were sexually intact females. All of these dogs had serum cobalamin concentrations < 150 ng/L (Figure 1); median urine methylmalonic acid concentration for these dogs was 4,064 mmol/mol of creatinine (range, 1,800 to 6,665 mmol/mol of creatinine; Figure 2), and median plasma total homocysteine concentration was 51.5 μmol/L (range, 40 to 81.6 μmol/L; Figure 3). Each of these 4 dogs had been fed a different commercial dog food.
Abnormalities detected in cobalamin-deficient Border Collies included stunted growth (4/4), lethargy (4/4), glossitis (2/4), fever (1/4), mild nonregenerative anemia (3/4), neutropenia (1/4), elevated isolated aspartate aminotransferase activity (3/4), and mild proteinuria (4/4). Parenteral (SC) administration of cobalamin to these dogs caused resolution of all abnormalities except proteinuria and increased isolated aspartate aminotransferase activity.
Discussion
To the authors' knowledge, serum cobalamin concentrations have not been correlated with concentrations of the biomarkers methylmalonic acid and total homocysteine in healthy pet dogs, and methods used to determine reference ranges for these factors have not been reported. Reference ranges determined in the present study were similar to previously reported16 reference ranges for these factors. Although, to the authors' knowledge, no criterion-referenced method exists for determination of cobalamin status in humans or other animals, a plasma methylmalonic acid concentration within the reference range is typically considered to be supportive of a clinically normal cobalamin status in humans (even when serum cobalamin concentration is low).17
Little information has been published regarding methylmalonic acid concentrations in dogs. Cats with elevated serum methylmalonic acid concentrations have low serum cobalamin concentrations; serum methylmalonic acid concentrations decrease after administration of cobalamin to these cats.18 Similarly, other authors15 recently reported that there is a negative correlation between serum cobalamin and methylmalonic acid concentrations in dogs. Results of that study15 also indicate that measurement of serum methylmalonic acid concentration may be a better diagnostic test for detection of cobalamin deficiency than is measurement of serum cobalamin concentration. Measurement of urine methylmalonic acid concentration has only sporadically been reported, and no reference ranges have been established.9,10,12,19 Measurement of urine methylmalonic acid concentration may have several advantages for determination of cobalamin status. Methylmalonic acid concentrations in urine are up to 40-fold as high as they are in serum in humans and are therefore easy to determine.20 In addition, use of the ratio of urine methylmalonic acid to creatinine concentration minimizes influences of hemoconcentration and kidney disease on values.20,21 Additionally, methylmalonic acid is very stable in urine,22 whereas no data have been reported regarding stability of methylmalonic acid in serum, to the authors' knowledge. As well, free-catch urine samples can be obtained less invasively, compared with collection of blood samples, and they can be easily obtained by owners.
Unexpectedly, urine methylmalonic acid concentrations in healthy Border Collies were significantly higher than those in healthy control dogs in the present study. One cause of elevated urine methylmalonic acid concentration in humans is food intake prior to collection of urine samples, although postprandial urine methylmalonic acid concentrations have only been reported to reach as high as 3 mmol/mol of creatinine.23 It is unlikely that the diet of the dogs influenced results of the present study because conditions under which samples were collected were similar for Border Collies and control dogs. Furthermore, methylmalonic acid concentrations in paired urine samples collected from 6 dogs before and after feeding a standard meal were similar (data not shown). Even if diet of the dogs had a minor impact on elevation of urine methylmalonic acid concentrations in healthy Border Collies, values determined in the present study were much higher than those reported for humans that had not fasted.23
Theoretically, bacterial overgrowth in the small intestine may increase urine methylmalonic acid concentration. An overgrowth of bacteria producing propionic acid, a precursor of methylmalonyl-CoA, could cause an increase in urine methylmalonic acid concentration.24 We cannot exclude this possibility, but consider it unlikely because none of the healthy Border Collies in the present study had a history of digestive problems. Notably, diets did not differ between control dogs and Border Collies.
Extremely high methylmalonic acid concentrations (237, 264, and 360 mmol/mol of creatinine) were detected in urine samples obtained from 3 healthy Border Collies in the present study; these dogs were unrelated to each other and were living in the same household. Each of these dogs were fed a diet of bone and raw food. Because feeding bone and raw food usually requires that food undergo a freeze-thaw cycle, loss of water-soluble B vitamins was suspected. To investigate this possibility, serum cobalamin, plasma total homocysteine, and urine methylmalonic acid concentrations were determined for 12 healthy dogs that were fed bone and raw food exclusively. Results for these dogs were similar to those for control dogs.
It is possible that the healthy eucobalaminemic Border Collies with methylmalonic aciduria in the present study were subclinical carriers of hereditary selective cobalamin malabsorption. Genetic testing would have been required to verify this. However, the lack of difference in serum cobalamin and plasma total homocysteine concentrations between healthy control dogs and the healthy Border Collies indicates it was unlikely that these Border Collies were carriers.
In humans, inborn errors of cellular cobalamin metabolism can cause methylmalonic aciduria.25,26 Intracellular cobalamin metabolism involves multiple steps between release of cobalamin from lysosomes and synthesis of adenosylcobalamin in mitochondria (which is required for function of the mitochondrial enzyme methylmalonyl-CoA mutase) and methylcobalamin in cytosol (which is required for function of the cytoplasmic enzyme methionine synthase). Nine defects of this pathway in humans have been identified; these defects cause methylmalonic aciduria, homocysteinemia, or both, depending on the metabolic step that is affected.25,26 In affected humans, serum cobalamin concentrations are typically within the reference range, as was found for healthy Border Collies in the present study. However, in humans, most defects in that cobalamin metabolism pathway cause overt clinical signs and life-threatening disease; humans that have methylmalonic aciduria and are asymptomatic for disease are rarely identified.26
The finding that urine methylmalonic acid concentration was increased in 37.6% of healthy Border Collies in the present study might indicate these dogs had benign methylmalonic aciduria, which, to the authors' knowledge, has not been identified in dogs. Benign methylmalonic aciduria has been identified in human children with no evidence of cobalamin deficiency; methylmalonic aciduria does not resolve after administration of cobalamin to these children.27 Two siblings in that study27 were found to have a defect in the methylmalonyl-CoA mutase enzyme. Authors of another report28 described benign methylmalonic aciduria in a Turkish family, 3 members of which had serum cobalamin and plasma and urine total homocysteine concentrations within the respective reference limits. Results of extensive biochemical screening for other known causes of methylmalonic aciduria, including function of the methylmalonyl-CoA mutase system, were unremarkable for those family members.28
All Border Collies with cobalamin deficiency in the present study had elevated plasma total homocysteine concentrations; these concentrations were higher than those in any of the healthy control dogs. Homocysteine is the intermediate product of methionine metabolism, and subsequent steps in the metabolic pathway are cobalamin-dependent. Homocysteine concentration is a sensitive indicator of cobalamin status in humans; homocysteine concentrations increase early in the course of cobalamin deficiency and often precede development of clinical signs. Renal disease, hemoconcentration, thyroid disease, folate deficiency, and use of certain medications (eg, omeprazole, cyclosporine, and antiepileptic drugs) are known causes of hyperhomocysteinemia in humans.29 Similarly, increased total homocysteine concentrations are associated with renal and cardiac diseases in dogs.30 None of these causes of elevated homocysteine concentration were detected in the Border Collies with cobalamin deficiency in the present study.
Interestingly, none of the 41 healthy Border Collies in the present study that had an elevated urine methylmalonic acid concentration had an elevated plasma total homocysteine concentration, which indicates these dogs may have had a subclinical defect in methylmalonic-CoA mutase. Hypocobalaminemia (range, 150 to 259 ng/L; median, 251 ng/L [reference range, 261 to 1,001 ng/L]) was detected in 5 of the healthy Border Collies; 4 of these dogs had a serum cobalamin concentration (230, 251, 254, and 259 ng/L) that was near the reference range. Unlike the 4 Border Collies with cobalamin deficiency that had signs of disease, these hypocobalaminemic healthy Border Collies had urine methylmalonic acid and plasma total homocysteine concentrations that were within the reference ranges. In contrast to results for the Border Collies that had clinical signs of disease, these healthy Border Collies were in excellent physical condition. The possibility that enzyme-bound cobalamin in tissues may prevent cellular cobalamin deficiency indicates the importance of measuring markers of cellular cobalamin metabolism.12
The finding in the present study of methylmalonic aciduria in healthy Border Collies with serum cobalamin concentrations within the reference limits and in Border Collies with clinical signs of cobalamin deficiency that had low serum cobalamin concentrations is intriguing and should be investigated further. These results may indicate different disease processes. The healthy Border Collies with methylmalonic aciduria and serum cobalamin concentrations within the reference limits may have had a defect in the mitochondrial metabolic cobalamin pathway (ie, methylmalonyl-CoA mutase). The Border Collies with clinical signs of cobalamin deficiency may have had selective intestinal malabsorption of cobalamin. Future studies should focus on genetic testing, measurement of intestinal cobalamin absorption, and determination of methylmalonyl-CoA mutase function of dogs.
ABBREVIATIONS
| CoA | Coenzyme A |
| CV | Coefficient of variation |
Sarstedt, Sevelen, Switzerland.
Immulite 2000, Vitamin B12, Siemens Healthcare Diagnostics Inc, Newark, Del.
EQCAPC-0100, Europa Bioproducts, Cambridge, England.
QP5050A, Shimadzu, Kyoto, Japan.
ABXPentra 400 analyzer, AxonLab, Stuttgart, Germany.
GraphPad Prism, version 5.0, GraphPad Software Inc, San Diego, Calif.
References
1 Markle HV. Cobalamin. Crit Rev Clin Lab Sci 1996; 33: 247–356.
2 Marcoullis G, Rothenberg SP. Intrinsic factor-mediated intestinal absorption of cobalamin in the dog. Am J Physiol 1981; 241: G294–G299.
3 Batt RM, Morgan JO. Role of serum folate and vitamin B12 concentrations in the differentiation of small intestinal abnormalities in the dog. Res Vet Sci 1982; 32: 17–22.
4 Grasbeck R, Gordin R, Kantero I, et al. Selective vitamin B12 malabsorption and proteinuria in young people. A syndrome. Acta Med Scand 1960; 167: 289–296.
5 Imerslund O. Idiopathic chronic megaloblastic anemia in children. Acta Paediatr Suppl 1960; 49(suppl 119):1–115.
6 Fyfe JC, Hall CA, Jezyk PF, et al. Inherited selective intestinal cobalamin malabsorption and cobalamin deficiency in dogs. Pediatr Res 1991; 29: 24–31.
7 He Q, Madsen M, Kilkenney A, et al. Amnionless function is required for cubilin brush-border expression and intrinsic factor-cobalamin (vitamin B12) absorption in vivo. Blood 2005; 106: 1447–1453.
8 Grützner N, Bishop MA, Suchodolski JS, et al. Association study of cobalamin deficiency in the Chinese Shar Pei. J Hered 2010; 101: 211–217.
9 Morgan LW, McConnell J. Cobalamin deficiency associated with erythroblastic anemia and methylmalonic aciduria in a Border Collie. J Am Anim Hosp Assoc 1999; 35: 392–395.
10 Battersby IA, Giger U, Hall EJ. Hyperammonaemic encephalopathy secondary to selective cobalamin deficiency in a juvenile Border Collie. J Small Anim Pract 2005; 46: 339–344.
11 Outerbridge CA, Myers SL, Giger U. Hereditary cobalamin deficiency in Border Collie dogs. J Vet Intern Med 1996; 10: 169–171.
12 Fordyce HH, Callan MB, Giger U. Persistent cobalamin deficiency causing failure to thrive in a juvenile Beagle. J Small Anim Pract 2000; 41: 407–410.
13 Savage DG, Lindenbaum J, Stabler SP, et al. Sensitivity of serum methylmalonic acid and total homocysteine determinations for diagnosing cobalamin and folate deficiencies. Am J Med 1994; 96: 239–246.
14 Vester B, Rasmussen K. High performance liquid chromatography method for rapid and accurate determination of homocysteine in plasma and serum. Eur J Clin Chem Clin Biochem 1991; 29: 549–554.
15 Sewell AC, Böhles HJ. 4-hydroxycyclohexanecarboxylic acid: a rare compound in urinary organic acid analysis. Clin Chem 1991; 37: 1301–1302.
16 Berghoff N, Suchodolski JS, Steiner JM. Association between serum cobalamin and methylmalonic acid concentrations in dogs. Vet J 2012; 191: 306–311.
17 Vogiatzoglou A, Oulhaj A, Smith AD, et al. Determinants of plasma methylmalonic acid in a large population: implications for assessment of vitamin B12 status. Clin Chem 2009; 55: 2198–2206.
18 Ruaux CG, Steiner JM, Williams DA. Early biochemical and clinical responses to Cbl supplementation in cats with signs of gastrointestinal disease and severe hypocobalaminemia. J Vet Intern Med 2005; 19: 155–160.
19 Fyfe JC, Jezyk PF, Giger U, et al. Inherited selective malabsorption of vitamin B12 in Giant Schnauzers. J Am Anim Hosp Assoc 1989; 25: 533–539.
20 Norman EJ, Cronin C. Cobalamin deficiency. Neurology 1996; 47: 310–311.
21 Norman EJ. Urinary methylmalonic acid/creatinine concentration ratio: a gold standard test for tissue vitamin B12 deficiency. J Am Geriatr Soc 1999; 47: 1158–1159.
22 Matchar DB, Feussner JR, Millington DS, et al. Isotope-dilution assay for urinary methylmalonic acid in the diagnosis of vitamin B12 deficiency. A prospective clinical evaluation. Ann Intern Med 1987; 106: 707–710.
23 Rasmussen K. Studies on methylmalonic acid in humans. I. Concentrations in serum and urinary excretion in normal subjects after feeding and during fasting, and after loading with protein, fat, sugar, isoleucine, and valine. Clin Chem 1989; 35: 2271–2276.
24 Bain MD, Jones M, Borriello SP, et al. Contribution of gut bacterial metabolism to human metabolic disease. Lancet 1988; 1: 1078–1079.
25 Whitehead VM. Acquired and inherited disorders of cobalamin and folate in children. Br J Haematol 2006; 134: 125–136.
26 Watkins D, Rosenblatt DS. Inborn errors of cobalamin absorption and metabolism. Am J Med Genet C Semin Med Genet 2011; 157: 33–44.
27 Ledley FD, Levy HL, Shih VE, et al. Benign methylmalonic aciduria. N Engl J Med 1984; 311: 1015–1018.
28 Sewell AC, Herwig J, Böhles H. A case of familial benign methylmalonic aciduria? J Inherit Metab Dis 1996; 19: 696–697.
29 Stanger O, Herrmann W, Pietrzik K, et al. DACH-LIGA homocystein (German, Austrian and Swiss Homocysteine Society): consensus paper on the rational clinical use of homocysteine, folic acid and B-vitamins in cardiovascular and thrombotic diseases: guidelines and recommendations. Clin Chem Lab Med 2003; 41: 1392–1403.
30 Rossi S, Rossi G, Giordano A, et al. Homocysteine measurement by an enzymatic method and potential role of homocysteine as a biomarker in dogs. J Vet Diagn Invest 2008; 20: 644–649.
