FIELD OF THE INVENTION

This invention relates to a method of diagnosing cobalamin deficiency and folic acid deficiency and distinguishing between the two, in warm¬ blooded animals, particularly humans, by measuring serum levels of cystathionine and 2-methylcitric acid.

BACKGROUND OF THE INVENTION Accurate and early diagnosis of cobalamin

(Vitamin B ₁₂ ) and folic acid deficiencies in warm¬ blooded animals is important because these deficiencies can lead to life-threatening hematologic abnormalities which are completely reversible by treatment with cobalamin or folic acid, respectively. Accurate and early diagnosis of cobalamin deficiency is especially important because it can also lead to incapacitating and life-threatening neuropsychiatric abnormalities; administration of exogenous cobalamin stops the progression of these abnormalities, almost always leads to significant improvement in symptoms, and frequently leads to their complete correction. The distinction between cobalamin and folic acid deficiency is often difficult because both deficiencies lead to indistinguishable hematologic abnormalities; the distinction is important because use of the proper vitamin results in the greatest improvement in hematologic abnormalities and, more importantly, only cobalamin will correct the neuropsychiatric abnormalities which are only seen in cobalamin deficiencies. The use of folic acid to treat cobalamin deficiency is extremely dangerous, since some or all of the hematologic abnormalities may improve, but neuropsychiatric abnormalities will not improve and may progress, or even be precipitated.

Assays for cobalamin and folate in serum or plasma have long been the most widely utilized and recommended tests for diagnosing and distinguishing cobalamin and folic acid deficiency. However, in 1978 it was discovered that cobalamin analogues are present in human plasma and that their presence could mask cobalamin deficiency because the radioisotope dilution assays for serum cobalamin then in use were not specific for true cobalamin. This problem could be corrected by using pure or purified intrinsic factor as the binding protein in the radioisotope dilution assay for cobalamin. This modification has almost totally replaced assays existing in 1978 that used a nonspecific cobalamin-binding protein. See, e.g., U.S. patent 4,188,189 (Allen), U.S. Patent 4,351,822

(Allen), U.S. Patent 4,451,571 (Allen), and Kolhouse, J.F., H. Kondo, N.C. Allen, E. Podell, and R.H. Allen, N. Enσ. J. Med. 299;785-792 (1978). These improved assays for serum cobalamin are now utilized in thousands of laboratories throughout the world and appear to give low values for about 90% of patients with cobalamin deficiency. R. H. Allen, S.P. Stabler, D.G. Savage and J. Lindenbaum, American Journal of Hematoloαv, 34:90-98 (1990); J. Lindenbaum, D.G. Savage, S.P. Stabler and R. H. Allen, American Journal of Hematolocry, 34:99-107 (1990); S.P. Stabler, R.H. Allen, D.G. Savage and J. Lindenbaum, Blood, 76(5): 871-81 (1990).

The improved assays avoid the problem of cobalamin analogues masking true cobalamin deficiency, but they have been severely criticized because they frequently give low values in patients who lack any evidence of actual cobalamin deficiency. This problem of false positive testing led experts in the field to take the position that cobalamin deficiency should be considered and serum cobalamin values should be obtained only in patients who have hematologic or

neurologic abnormalities that are typical of patients with cobalamin deficiency. Dr. Schilling and his coworkers, who are experts in the field of cobalamin deficiency and laboratory diagnosis, stated: "We conclude that the 'improved* vitamin B ₁₂ assay kits will yield an increased proportion of clinically unexplained low results for serum B ₁₂ . It seems prudent for scientific and economic reasons to measure serum vitamin B ₁₂ only in patients who have hematological or neurological findings that suggest a reasonable probability of vitamin B ₁₂ deficiency. Measuring serum

B ₁₂ as a screening test in the anemic or the geriatric population will result in a high proportion of low values that cannot be correlated with clinical disease."

Schilling, R.F., V.F. Fairbanks, R. Miller, K. Schmitt, and M.J. Smith, Clin. Chem. 29(3) :582-583 (1983).

Thus, the cobalamin assays referred to by Dr.

Schilling frequently provided low serum cobalamin levels in patients who were not truly cobalamin deficient. Such findings are confusing or misleading to the physician and may result in unnecessary and expensive further testing. To avoid that, it was generally taught in the art that the clinical spectrum of cobalamin deficiency is relatively narrow and well- defined and that the possibility of cobalamin deficiency should only be considered in those who have concurrent hematological or neurological symptoms. Routine screening of the general population or those with only moderate anemia, or moderate macrocytosis, or other neuropsychiatric abnormalities, would lead to high numbers of false positives.

It was thought that the hematological or neurological symptoms that were necessary to justify an assay for serum cobalamin and the resulting risk of a false positive, were fairly severe. Those symptoms included significant anemia, displayed for example in decreased hematocrit or hemoglobin, with macrocytic red

blood cells (i.e., mean cell volume (MCV) generally greater than 100 fl), or neurologic symptoms of peripheral neuropathy and/or ataxia. Anemia associated with cobalamin deficiency was described as typically severe with hemoglobin < 8 g% or hematocrit < 25% and the size of the red blood cells was described as greatly increased to levels > 110 fl. See, for example, Babior and Bunn (1983) in Harrison's Principles of Internal Medicine (Petersdorf et al. , eds.) McGraw-Hill Book Co., New York; Lee and Gardner (1984) in Textbook of Family Practice, 3rd Ed. (Rakel, ed. ) Saunders & Co, Philadelphia). While it was well recognized that individuals with cobalamin deficiency could display neurologic disorders in the absence of anemia, such situations were believed to be exceptional and rare. See Beck (1985) in Cecil Textbook of Medicine, 17th Ed. (Wyngaarden and Smith, eds.) W.B. Saunders, Philadelphia, p. 893-900; Babior and Bunn (1987) in Harrison's Principles of Internal Medicine, 11th Ed. (Braunwald et al. , eds.) McGraw-Hill, New York pp. 1498-1504; Walton (1985) Brain's Diseases of the Nervous System, 9th Ed. Oxford University Press, Oxford, UK. The neurologic symptoms of cobalamin deficiency were considered to be late manifestations of the disease most typically occurring after the onset of anemia or, if they occurred first, were soon to be followed by the onset of anemia. See Woltmann (1919) Am. J. Med. Sci. 157:400-409; Victor and Lear (1956) Am. J. Med. 20:896-911. It was later discovered that the clinical spectrum of cobalamin deficiency is much broader than previously recognized and that many cobalamin-deficient patients are not anemic, or only moderately anemic; that in many cases their red blood cells are not macrocytic, or only moderately macrocytic; that in many cases a variety of neurologic abnormalities other than peripheral neuropathy and ataxia are present; and that

in many cases the serum cobalamin level is only slightly decreased and may actually be normal, even with the improved assays using purified intrinsic factor. See U.S. Patent no. 4,940,658 by Allen et al; S.P. Stabler, R.H. Allen, D.G. Savage and J.

Lindenbaum, Blood, 76(5): 871-81 (1990). Accordingly, there was a need for an improved assay for cobalamin deficiency, preferably one in which cobalamin deficiency could be distinguished from folic acid deficiency.

An improved assay is disclosed in U.S. Patent No. 4,940,658, the contents of which are hereby incorporated by reference, by Allen et a_L. , who are the inventors of the present invention. The Allen et al. patent teaches a method of assaying total homocysteine serum concentrations to predict cobalamin or folic acid deficiency, and assaying serum methylmalonic acid concentrations to distinguish between the two deficiencies. Briefly, it was determined that there were elevated levels of total homocysteine in about 85- 95% of patients with cobalamin deficiency and in about 90-95% of patients with folic acid deficiency. Therefore, elevated homocysteine levels pointed toward one of those two deficiencies, but did not distinguish between the two. However, distinguishing between the two was assisted by the discovery that there were elevated levels of methylmalonic acid in about 85-95% of patients with cobalamin deficiency but methylmalonic acid levels were normal in patients with folic acid deficiency. The two assays are relatively inexpensive and accurate, and can be performed concurrently by taking a single serum sample.

The assay described in the Allen patent is a very important test procedure but may not be infallible in detecting either kind of deficiency or in distinguishing between the two. About 5-15% of the cobalamin deficiencies will not be detected because

these patients will not show elevated homocysteine levels. About 5-10% of the folic acid deficiencies will not be detected because these patients will not show elevated homocysteine levels. When elevated homocysteine levels are in fact detected, thereby indicating either a cobalamin deficiency or folic acid deficiency, about 5-10% of the cobalamin deficiencies can not be distinguished from the folic acid deficiencies because the patients will not show elevated methylmalonic acid levels. Therefore the assays may produce a small but not insignificant number of false negative results.

The existence of this small potential of false negative results led the inventors of the present invention to develop another assay procedure to detect cobalamin and folic acid deficiencies and to distinguish between the two. It has been discovered that elevated levels of cystathionine are present in the serum and urine of about 80-90% of patients with cobalamin deficiency or folic acid deficiency. It has also been discovered that elevated levels of 2- methylcitric acid are present in the serum, urine and cerebral spinal fluid of 80-90% of patients with cobalamin deficiency but not in patients with folic acid deficiency. The present invention uses an assay for cystathionine as a test for the existence of either of the two deficiencies or as a check on an assay for total homocysteine when testing for the two deficiencies. It also uses an assay for 2-methylcitric acid as a test for distinguishing between the two deficiencies or as a check on an assay for methylmalonic acid in distinguishing between the two deficiencies.

It was previously known that increased amounts of cystathionine are present in the urine and serum of children and adults with inherited defects of cystathionase who cannot convert cystathionine to

cysteine and alpha-ketobutyrate. The serum levels of cystathionine range from undetectable to as high as 80,000 nanomoles per liter. Serum levels of methionine, homocysteine and cysteine are usually not elevated in these patients. See, e.g., Mudd, S.H., H.L. Levy and F. Skovby, in The Metabolic Basis of Inherited Disease , 6th Ed. (CR. Scriver, A.L. Beaudet, W.S. Sly and D. Valle, eds.) (McGraw-Hill, Inc., New York, 1989) pp. 693-734. Elevated levels of urine cystathionine have also been seen in a variety of other conditions including pyridoxine (Vitamin B ₆ ) deficiency, hyperthyroidism, liver disease, various tumors of the central nervous system and liver, and inherited defects of cystathionine transport in the kidney. See, e.g., Mudd, S.H., in The Metabolic Basis of Inherited Disease , supra.

Elevated levels of cystathionine are present in the urine and serum of some children with inherited defects involving methionine synthase, who cannot convert homocysteine and N ⁵ -methyltetrahydrofolate to methionine and tetrahydrofolate, respectively. These patients usually have low levels of methionine, high levels of total homocysteine, and normal levels of cysteine in their serum. In these patients, the inherited defects were due to: 1) failure to synthesize N ⁵ -methyltetrahydrofolate; 2) failure to synthesize methyl-cobalamin, which is a required co-factor for methionine synthase; and 3) a lack of the plasma transport protein transcobalamin II, which is required to deliver cobalamin to cells. See, e.g., Levy, H.L., S.H. Mudd, J.D. Schulman, P.M. Dryfus, and R.H. Abeles, Am. J. Med. 48:390-397, 197; Baumgartner, E.R., H. Wick, R. Mauere, N. Egli, and B. Steinmann, Helv. Paediat. Acta 34:465-482, 1979; Ribes, A., M.A. Vilaseca, P. Briones, A. Maya, J. Sabater, P. Pascual, L. Alvarez, J. Ros and E.G. Pascual, J. Inher. Metabol. Pis. 7(Suppl. 2):129-130, 1984; Baumgartner, R. , H.

Wick, J.C. Linnell, E. Gaull, C. Bachmann, and B. Steinmann, HeIv. Paedia . Acta. 34:483-496, 1979; Mudd, S.H., in The Metabolic Basis of Inherited Disease, supra; Carmel, R. , A.A. Bedros, J.W. Mace, and S.I. Goodman, Blood 55:570-579, 1980.

On the other hand, other children with similar defects did not have elevated levels of cystathionine, which tended to cast doubt on the reliability of such an indicator. See, e.g., Goodman, S.I., P.G. Moe, K.B. Hammond, S.H. Mudd, and B.W. Uhlendorf, Biochem. Med 4:500-515, 1970; Barshop, B.A., J. Wolff, W.L. Nyhan, A. Yu, C. Pordanos, G. Jones, L. Sweetman, J. Leslie, J. Holm, R. Green, D.W. Jacobsen, B.A. Cooper, and D. Rosenblatt, Am. J. Med. Genet. 35:222-228, 1990; Baumgartner, R. , H. Wick, H.C. Linnell, E. Gaull, C. Bachmann, and B. Steinmann, Helv. Paediat. Acta. 34:483-496, 1979; Mudd, S.H., in The Metabolic Basis of Inherited Disease, supra.

Elevated levels of cystathionine have been reported in the serum of pigs with severe experimental vitamin B ₁₂ deficiency. See Levy, H.L. and G.J. Cardinale, Fed. Proc. 29:634, 1970: Mudd, S.H., in Heritable Disorders of Amino Acid Metabolism: Patterns of Clinical Expression and Genetic Variation, (W.O. Nyhan, ed. ) John Wiley & Sons, New York, 1974, pp. 429-451.

Increased amounts of cystathionine have been observed in the urine of several children with life-threatening cobalamin deficiency. See Higgenbottom, M.C., L. Sweetman, and W.L. Nyhan, N. Enσl. J. Med. 299:317-323, 1978; Davis, Jr., J.R., J. Goldenring, and B.H. Lugin, Am. J. Pis. Child. 135:566-567, 1981. In other infants with life-threatening cobalamin deficiency, amino acids in urine were found to be normal or cystathionine was present in undetectable or normal amounts. See

Grasbeck, R. , R. Gordin, I. Kantero, and B. Kuhlback, Acta Medica Scandinavica. 167:289-296, 1960; Lambert,

H.P., T.A.J. Prankerd, and J.M. Smellie, Q.J. Med♦ 30:71-92, 1961; Lampkin, B.C. and A.M. Mauer, Blood 30:495-502, 1967; Hollowell, Jr., J.G., W.K. Hall, M.E. Coryell, J. McPherson, Jr., and D.A. Hahn, Lancet 2:1428, 1969; Frader, J., B. Reibman, and D. Turkewitz, N. Enσl. J. Med. 299:1319-1320, 1978. Cystathionine was not detected in the serum or urine of a child with life-threatening folic acid deficiency although it may have been present in this patient's cerebral spinal fluid. See Corbeel, L., G. Van den Berghe, J. Jaeken, J. Van Tornout, and R. Eeckels, Eur. J. Pediatr. 143:284-290, 1985. It is believed that elevated cystathionine has not been reported in the serum or urine of children with moderate or mild cobalamin or folic acid deficiencies, or in adults with any degree of cobalamin or folic acid deficiencies, or in the serum of normal children or adults.

It was previously known that large amounts of 2-methylcitric acid are present in the urine of children with inherited defects in propionyl-CoA carboxylase, who cannot convert propionyl-CoA to D- methylmalonyl-CoA, and with inherited defects in L- methylmalonyl-CoA mutase, who cannot convert L- methylmalonyl-CoA to succinyl-CoA. See, e.g., Ando, T., K. Rasmussen, J.M. Wright, and W.L. Nyhan, J. Biol. Chem. 247:2200-2204, 1972: Sweetman, L., W. Weyler, T. Shafai, P.E. Young, and W.L. Nyhan, JAMA 242:1048-1052, 1979; Weidman, S.W. and G.R. Drysdale, Biochem. J. 177:169-174, 1979. It has also been reported that 2- methylcitric acid is present in increased amounts in the amniotic fluid and urine of pregnant women with fetuses that were shown after birth to have inherited defects in either propionyl-CoA carboxylase or L- methylmalonyl-CoA mutase. See Naylor, G., L. Sweetman, W.L. Nyhan, C. Hornbeck, J. Griffiths, L. Morch, and S. Brandange, Clinica Chimica Acta 107:175-183, 1980; Sweetman, L. , G. Naylor, T. Ladner, J. Holm, W.L.

Nyhan, C. Hornbeck, J. Griffiths, L. Morch, S. Brandange, L. Gruenke, and J.C. Craig, in Stable Isotopes (H.L. Schmidt, H. Forstel, and K. Heinzinger, eds.) Elsevier Scientific Publishing Company, Amsterdam, 1982, pp. 287-293; Aramaki, S., P. Lehotay, W.L. Nyahn, P.M. MacLeod and L. Sweetman, J. Inher. Metab. Pis. 12:86-88, 1989; Coude, M., B. Chadefaux, P. Rabier, and P. Kamound, Clinica Chimica Acta. 187:329- 332, 1990. Elevated levels of 2-methylcitric acid have also been observed in the urine of some (Barshop, B.A., J. Wolff, W.L. Nyhan, A. Yu, C. Prodanos, G. Jones, L. Sweetman, J. Leslie, J. Holm, R. Green, P.W. Jacobsen, B.A. Cooper, and P. Rosenblatt, Am. J. Med. Genet. 35:222-228, 1990; Baumgartner, R. , H. Wick, J.C. Linnell, E. Gaull, C.Bachmann, B. Steimann, Helv. Paediat. Acta. 34:483-496, 1979; Higgenbottom, M.C., L. Sweetman, and W. L. Nyhan, N. England J. Med. 299:317- 323, 1978) but not all (Levy, H.L., S.H. Mudd, J.P. Schulman, P.M. Preyfus, and R.H. Abelese, A . J. Med.

48:390-397, 1970; Baumgartner, E.R. H. Wick, R. Maurer, N. Egli, and B. Steinmann, Helv. Paediat. Acta. 34:465- 482, 1979; Ribes, A., M.A. Vilaseca, P. Briones, A. Maya and J. Sabater, J. Inherit. Metabol. Pis. 7(Suppl. 2):129-130, 1984; Mudd, S.H., and B.W. Uhlendorf,

Biochem. Med. 4:500-515, 1970; Carmel, R., A.A. Bedros, J.W. Mace, and S.I. Goodman, Blood 55:570-579, 1980) children with inherited defects involving: 1) the inability to convert cobalamin to adenosyl-cobalamin, which is a required co-factor for L-methylmalonyl-CoA mutase; and 2) transcobalamin II deficiency which results in the inability to transport cobalamin from plasma to cells.

Large amounts of 2-methylcitric acid have been found in the urine of one child with life-threatening cobalamin deficiency. See Higgenbottom, M.C.L. Sweetman, and W.L. Nyhan, N. Enql. J. Med. 299:317-323,

1978. In other infants with life-threatening cobalamin deficiency, the presence or elevated levels of 2- methylcitric acid were not observed. See Pavis, Jr., J.R., J. Goldenring, and B.H. Lubin, Am. J. Pis. Child. 135:566-567, 1981; Hollowell, Jr. J.G., W.K. Hall, M.E. Coryell, J. McPherson, Jr., and P.A. Hahn, Lancet 2:1428, 1969; Frader, J. , Reibman, and P. Turkewitz, N. Enσl. J. Med. 299:1319-1320, 1978). It is believed that the presence of 2-methylcitric acid in detectable or elevated amounts has not been reported in the urine of children with moderate or mild cobalamin deficiency or adults with severe, moderate or mild cobalamin deficiency. The presence of 2-methylcitric acid in detectable or elevated amounts has not been reported in the serum of any subject with any of the inherited disorders involving propionyl-CoA carboxylase, L- methylmalonyl-CoA mutase, the synthesis of adenosyl cobalamin or the plasma transport of cobalamin nor has it been reported in the serum of children or adults with any degree of cobalamin deficiency nor in the serum of normal children or adults.

SUMMARY OF THE INVENTION The present invention measures cystathionine levels in the serum or urine and measures 2- methylcitric acid levels in the serum, urine or cerebral spinal fluid for the purpose of predicting cobalamin and folic acid deficiencies and for distinguishing between the two deficiencies. It has been determined that elevated cystathionine levels tend to correspond with the existence of either such deficiency, while elevated 2-methylcitric acid levels tend to correspond with the existence of cobalamin deficiency but not folic acid deficiency and can thereby be used to distinguish between the two deficiencies.

In a preferred embodiment, samples of serum or urine in the assay for cystathionine, and samples of

serum, urine or cerebral spinal fluid in the assay for 2-methylcitric acid, are mixed by vortexing with water and known amounts of stable isotope forms of cystathionine or 2-methylcitric acid, respectively. The mixtures are washed, incubated, eluted and dried in the manner described herein, and then are analyzed by gas chromatography/mass spectrometry.

The assays for cystathionine and 2-methylcitric acid may be combined with assays for homocysteine or methylmalonic acid or both. Further, while the assays are described for use with serum or urine in the case of cystathionine and with serum, urine or cerebral spinal fluid in the case of 2-methylcitric acid, it should be appreciated that "serum" may include plasma and that it may be feasible to apply the assays to other body fluids as well.

BRIEF PESCRIPTION OF THE PRAWINGS FIG. 1 shows a diagram of the metabolic pathways of homocysteine, methionine and cystathionine in humans.

FIG. 2 illustrates the mass spectra of cystathionine.

FIG. 3 illustrates the chromatograms of cystathionine.

FIG. 4 shows a diagram of the metabolic pathways of 2-methylcitric acid and methylmalonic acid.

FIG. 5 illustrates the mass spectra of 2- methylcitric acid. FIG. 6 illustrates the chromatograms of 2- methylcitric acid.

FIG. 7 shows clinical data showing the serum cystathionine levels in nanomoles per liter in normal patients and patients with cobalamin or folic acid deficiency.

FIG. 8 shows clinical data showing the 2- methylcitric acid levels in nanomoles per liter in

normal patients and patients with cobalamin or folic acid deficiencies.

PETAILEP PESCRIPTION OF THE INVENTION The Metabolism

The metabolism of homocysteine, methionine and cystathionine in humans is generally known and is shown diagrammatically in FIG. 1. Methionine is converted to S-adenosylmethionine by the transfer of the adenosyl moiety of ATP to methionine. S-adenosylmethionine is a high energy compound, and each sulfonium atom is capable of participating in one or more transfer reactions to produce the sulfur-containing compound S- adenosylhomocysteine. Hydrolysis cleaves S- adenosylhomocysteine into homocysteine and adenosine. Homocysteine may be converted to cystathionine by the transsulfuration pathway or may be remethylated to form methionine. In methylation, the reaction utilizes N ⁵ -methyltetrahydrofolate (N ⁵ -MTHF) as the methyl donor and utilizes methylcobalamin as a cofactor. The remethylation of homocysteine results in methionine and tetrahydrofolate, the later of which is converted back to N ⁵ -MTHF by PNA and RNA pathways. A deficiency in either folic acid or cobalamin will block the methylation of homocysteine into methionine. Such a block may lead to increases in serum or urine levels of homocysteine and cystathionine, although this outcome is not necessarily apparent in view of the complexity of the metabolic pathways and the possibility of other reactions.

In the synthesis of cystathionine from homocysteine in the transsulfuration pathway, the homocysteine is condensed with serine in a reaction catalyzed by cystathionine 3 synthase. Cystathionine cleaves into cysteine and -ketobutyrate in a reaction catalyzed by -cystathionase to complete the transsulfuration sequence.

In the pathway for the formation of methylmalonic acid, shown in FIG. 4, propionyl-CoA is converted into methylmalonyl-CoA. The methylmalonyl is converted to succinyl-CoA in a reaction requiring adenosylcobala in as a cofactor. A deficiency in cobalamin may block the conversion of methylmalonyl-CoA to succinyl-CoA, thereby resulting in methylmalonic aciduria or propionic aciduria, although again such an outcome is not necessarily apparent from the metabolic pathway because a variety of other reactions could occur.

In the synthesis of 2-methylcitric acid, propionyl-CoA and oxaloacetic acid and water react to form 2-methylcitric acid, using the enzyme citrate synthase as a catalyst. It has been found that the blocking of the pathway from propionyl-CoA to succinyl- CoA by a deficiency of adenosylcobalamin may increase the conversion of propionyl-CoA to 2-raethylcitric acid, thereby increasing 2-methylcitric acid levels. Thus it can be seen that cobalamin (in the form of methylcobalamin) and folic acid are vital to the methylation of homocysteine into methionine, and that cobalamin (in the form of adenosylcobalamin) but not folic acid is vital to the conversion of methylmalonyl- CoA to succinyl-CoA. A deficiency in either cobalamin or folic acid may lead to increased levels of homocysteine or cystathionine. A deficiency in cobalamin but not folic acid may also result in increased levels of methylmalonic acid or 2- methylcitric acid.

Assay for Cystathionine A preferred embodiment of the invention utilizes an assay for cystathionine levels in the manner described below. The following components are added in sequence to 12 x 75 mm glass test tubes:

(1) 51 ul of H ₂ 0 containing 40 picomoles of P, L [2,2,3,3-P ₄ ] cystathionine, available from MSD Isotopes, Montreal, Canada, as a custom synthesis; (2) 40 ul of serum or urine; and

(3) 100 ul of H ₂ 0.

The samples are mixed well by vortexing and 51 ul of 0.083 H ₃ B0 ₃ -NaOH, pH 10.0, containing 16.6 mg/ml of D, L-dithiothreitol is followed by mixing. After incubation at 40°C for 30 minutes, 51 ul of H ₂ 0 containing 83 mg/ml of iodacctamide is added followed by mixing. After incubating for 30 minutes at 40°C, 1 ml of 0.03N HC1 is added followed by mixing and the samples are then applied to 300 ul columns of a cation exchange resin (AG MP-50, 100-200 mesh, hydrogen form (Bio-Rad Laboratories, Richmond, CA) ) that has previously been washed with 1 ml of MeOH and 3.3 ml of H ₂ 0, to remove any negatively charged salts. After the sample is applied, each column is washed twice with 3 ml of H ₂ 0 and once with 3 ml of MeOH. Each column is then eluted with 1.1 ml of 4N NH ₄ OH in MeOH. The eluates are taken to dryness by vacuum centrifugation in a Savant vacuum centrifuge. The dried eluates are then derivatized by adding 30 ul of a solution containing 10 ul of N-methyl-N(tert-butyl dimethylsylyl) trifluoracetamide and 20 ul of acetonitrile. After incubation at 40°C for 60 minutes in sealed autosampler vials, 1 ul is analyzed by gas chromatography/mass spectrometry using a 10 meter SPB-1 capillary column (Supelco, Inc., Bellefonte, PA) that has an internal diameter of 0.25 mm and a 0.25 um film thickness.

Gas chromatography/mass spectrometry is performed using a Hewlett Packard 5890 Gas Chromatograph and a Hewlett Packard 5970 or 5971A Mass Selective Detector. The initial column temperature is 140°C which is held for approximately 0.6 minutes after

sample injection and is then increased to 300°C at a rate of 30°C/minute. The mass selective detector is operated in the selected ion monitoring mode in which ions 362.2 are monitored for endogenous cystathionine and 366.2 for P,L[2,2,3,3-P ₄ ] cystathionine.

Cystathionine is quantitated by dividing the integrated area of the M/Z 362.2 peak that elutes at approximately 5.7 minutes (the exact times are determined daily with controls) by the integrated area of the M/Z 366.2 peak that elutes at the same time and then multiplying by 1,000 nanomoles/liter, which is the equivalent amount of P,L[2,2,3,3-D ₄ ] cystathionine that was added to each sample.

In some experiments, the method may be altered in the following manner: 1) 51 ul of H ₂ 0 containing 400 picomoles of D,L[2,2,3,3-D ₄ ] cystathionine, 400 ul of serum (or 40 ul of urine with 360 ul of 0.15M NaCl), and 1 ml of H ₂ 0 are added in the initial sequence; 2) 51 ul of IN NaOH containing 10 mg/ml of dithiothreitol are added in place of the H ₃ B0 ₃ dithiothreitol prior to the first 30 minute incubation; 3) the addition of iodoacctamide and the second 30 minute incubation are omitted; 4) the 0.03N HC1 is omitted and the samples are applied to 300 ul columns of an anion exchange resin (AG MP-1, 100-200 mesh, chloride form); and 5) samples are eluted with 0.4N acetic acid (range 0.04 to 0.1N) in methanol. Comparable results are obtained with both methods.

The same processes may be used to assay for total homocysteine if a suitable internal standard for it is added to the samples.

FIG 2. shows the molecular weights of cystathionine and a diagram of the cystathionine molecule. Peaks representing the entire derivative, i.e. [M] ⁺ , were not observed. FIG 3 shows the gas chromotogram of cystathionine, the top graph being for

endogenous cystathionine and the bottom graph being for D,L[2,2,3,3-D ₄ ] cystathionine.

Assay for 2-Methylcitric Acid A preferred embodiment of the invention utilizes an assay for 2-methylcitric acid levels in the manner described below. The following components are each added in sequence to 12 x 75 mm glass test tubes:

1) 51 ul of H ₂ 0 containing 200.4 picomoles of [methyl-D ₃ ]2-methylcitric acid II and 172.8 picomoles of [methyl-D ₃ ]2-methylcitric acid I available from MSP Isotopes, Montreal, Canada, as a custom synthesis;

2) 400 ul of serum, cerebral spinal fluid, or urine (in the case of urine, 40 ul are used together with 360 ul of 0.15M NaCl); and

3) 1 ml of H ₂ 0.

The samples are mixed by vortexing and then applied to 300 ul columns of an anion exchange resin (AG MP-1, 100-200 mesh, chloride form (Bio-Rad Laboratories, Richmond, CA) that has previously been washed with 1 ml of MeOH and 3.3 ml of H ₂ 0. After the sample is applied, each column is washed with 3 ml of H ₂ 0 and 3 times with 3 ml of .06N acetic acid (range 0.01 to 0.1) in MeOH. Each column is then eluted with 1.1 ml of 3.6M acetic acid/0.IN HC1 in MeOH. The eluates are taken to dryness by vacuum centrifugation in a Savant vacuum centrifuge. The dried eluates are then derivatized by adding 30 ul of a solution containing 10 ul of N-methyl-N(tert-butyl dimethylsylyl)tri luoracetamide and 20 ul of acetonitrile. After incubation at 90°C for 30 minutes in sealed autosampler vials, 1 ul is analyzed by gas chromatography/mass spectrometry using a 20 meter SPB-1 capillary column (Supelco, Inc.) that has an internal diameter of 0.25 mm and a 0.25 um film thickness.

Gas chromatography/mass spectrometry is performed using a Hewlett Packard 5890 Gas Chro atograph and a Hewlett Packard 5971A Mass Selective Petector. The initial column temperature is 80°C which is held for approximately 0.6 minutes after sample injection and is then increased to 300°C at a rate of 30°C/minute. The mass selective detector is operated in the selected ion monitoring mode in which ions 605.4 are monitored for endogenous 2-methylcitric acid II and I and 608.4 are monitored for [methyl-D ₃ ]2- methylcitric acid II and I. 2-methylcitric acid II is quantitated by dividing the integrated area of the M/Z 605.4 peak that elutes at approximately 8.4 minutes (the exact times are determined daily with controls) by the integrated area of the M/Z 608.4 peak that elutes at the same time and then multiplying by 501 nanomoles/liter, which is the equivalent amount of [methyl-P ₃ ]2-methylcitric acid II that was added to each sample. 2-methylcitric acid I is quantitated in the same manner utilizing the M/Z 605.4 ^' and M/Z 608.4 peaks that elute at approximately 8.5 minutes and then multiplying their ratio by 432 nanomoles/liter which is the equivalent amount of [methyl-P ₃ ]2-methylcitric acid I added to each sample. The integrated areas for the two internal standard peaks, i.e. the M/Z 608.4 peaks eluting at about 8.4 and 8.5 minutes, are corrected for the amounts contributed to them by endogenous 2- methylcitric acid II and I, as a result of naturally occurring isotope abundance. These corrections, which are determined using samples containing only unenriched 2-methylcitric acid II and I, are approximately 6.6% of the areas of the 605.4 peaks at 8.4 and 8.5 minutes. It has been found that some serum, urine and cerebral spinal fluid samples contain an endogenous peak of M/Z 608.4 that elutes at the same time, i.e. approximately 8.5 minutes, as the M/Z 608.4 peak for 2-methylcitric acid I and that this endogenous peak interferes with

the quantitation of endogenous 2-methylcitric acid in these samples. This problem can be solved by using the M/Z 608.4 peak for 2-methylcitric acid II as the internal standard for quantitation of 2-methylcitric acid I and 2-methylcitric acid II.

The same process may be used to assay for methylmalonic acid if a suitable internal standard for it is added to the samples.

FIG. 5 shows the molecular weights of 2- methylcitric acid and a diagram of the 2-methylcitric acid molecule. Peaks representing the entire molecule, i.e. [M] ⁺ were not observed. FIG. 6 shows the gas chromatogram of 2-methylcitric acid I and II, the top graph being for endogenous 2-methylcitric acid I and II and the bottom graph being for [methyl-P ₃ ] 2- methylcitric acid I and II.

Although the assay described above is for both 2-methylcitric acid I and 2-methylcitric acid II, it should be appreciated that the assay can actually be used for either compound or both in combination.

Combined Assays In other experiments of a preferred embodiment of the invention, the second method for assaying cystathionine was combined with the method for assaying 2-methylcitric acid in the following manner: 1) 51 ul of H ₂ 0 containing the internal standards for cystathionine, 2-methylcitric acid, homocysteine, and methylmalonic acid was added to 400 ul of serum or cerebral spinal fluid or urine (40 ul urine plus 360 ul of 0.15M NaCl) and 1 ml of H ₂ 0 in the initial sequence; 2) 51 ul of H ₂ 0 containing 20 mg/ml of dithiothreitol was then added followed by mixing and a 30 minute incubation at 40°C; 3) the entire sample was then treated and followed up exactly as described for the 2- methylcitrate assay except that the run-through from the AG MP-1 was saved instead of being discarded; 4) 51

ul of IN NaOH containing 10 mg/ml of dithiothreitol was added to the run-through followed by mixing and a second 30 minute incubation at 40°C; and 5) the treated run-through samples from the preceding step were then applied directly to AG MP-1 columns and eluted with 0.4N acetic acid (range 0.04 to 0.1N) and methanol exactly as described in the second cystathionine method. In this combined assay, two autosampler vials were obtained for each sample, with the first autosampler vial containing endogenous and internal standards for 2-methylcitric acid and methylmalonic acid and the second autosampler vial containing endogenous and internal standards for cystathionine and homocysteine. Puring the GC/MS analysis, the peaks for endogenous and the internal standard for methylmalonic acid eluted at approximately 5.0 minutes which were well separated and ahead of the corresponding peaks for 2-methylcitric acid which eluted at approximately 8.5 minutes (see above). Puring the GC/MS analysis, the peaks for endogenous and the internal standard for homocysteine eluted at approximately 3.8 minutes which were well separated and ahead of the corresponding peaks for cystathionine which eluted at approximately 5.7 minutes (see above).

Clinical Results

FIG. 7 shows on a logarithmic scale the serum cystathionine levels in nanomoles per liter in 50 patients having no manifestations of any cobalamin or folic acid deficiency, in 30 patients having a known cobalamin deficiency and in 20 patients having a known folic acid deficiency. As the Figure suggests, high serum levels of cystathionine generally but not always correspond to either cobalamin or folic acid deficiency.

Table I set forth below shows urine

cystathionine levels ("UCYSTAT") from 50 normal subjects in nanomoles per liter.

TABLE I (normal urine cystathionine)

SAMPLE IP UCYSTAT

7426 467

10021 6077 3887

13259 2343 2472 4817 3736 8945 4144 3515

38834 8329 5723

20761 3594 3603

19063 4515 7361 6665 680 2339 6017 3592 547

21236 6811 3245 3474 2632 58

11953 691 5783 3844 1875 3639 877 1599 5318 1607 2079

936

10747

12060

6175 1644

Table II set forth below shows serum 2- methylcitric acid levels in nanomoles per liter for 50 patients with no cobalamin deficiency. The table is broken into columns for total 2-methylcitric acid

("TOTMC"), the ratio of the two isόmers of 2- methylcitric acid ("MCI/MCII"), the second isomer

("MCII") and the first isomer ("MCI").

TABLE II (normal serum 2-methylcitric acid) C MCI/MCII MCII MCI

100 47 100 35 94 33 93 36 93 31 92 30 92 33 91 45 91 35 87 31 83 31 80 29 79 26 78 27 70 29 69 29 66 22 66 28

Table III set forth below shows serum 2- methylcitric acid levels for 50 patients with clinically confirmed cobalamin deficiencies, using the same units and column labels as in Table II.

TABLE III (cobalamin deficient serum 2-methylcitric acid)

SAMPLE IP MCTOT MCI/MCII MCII MCI

485

612

468

322

303

276

283

289

245

231

191

206

162

160

148

130

120

109

106

92

88

69

91

71 38

Table IV set forth below shows serum 2- methylcitric acid levels for 25 patients with clinically confirmed folic acid deficiencies, using the same units and column labels as in Table II.

TABLE IV (folic acid deficient 2-methylcitric acid)

SAMPLE ID MCTOT MCI/MCII MCII MCI

48 15 41 14 40 11 34 11 26 10 20 7 10 2

The data tabulated in Tables II-IV is presented in graphic form in FIG. 8 for the two 2-methylcitric acid isomers totalled. As evident from the tables and graph, high levels of 2-methylcitric acid tend to correspond to a cobalamin deficiency but not to a folic acid deficiency.

Table V set forth below shows urine 2- methylcitric acid levels from the same normal subjects as presented in Table II for serum 2-methylcitric acid levels. The table is broken into columns for total urine 2-methylcitric acid ("UMCTOT"), the ratio of the two isomers of urine 2-methylcitric acid

("UMCI/UMCII"), the second isomer ("UMCII") and the first isomer ("UMCI").

TABLE V (normal urine 2-methylcitric acid)

UMCI/UMCII UMCII UMCI

13395 .60

12630 .60

12339 .75

11740 .51

11527 .57

11094 .67

10893 .53

9883 .76

9330 .56

9200 .84

9165 .55

9162 .76

5910 .53

8246 .45

7733 .54

7358 .54

7159 .95

7145 .54

6265 .53

6008 .53

5570 .69

3492 .60

3416 .55

3062 .52

2535 .62

2483 .09

1631 .96

1617 .51 859 .57

Table VI set forth below shows cerebral spinal fluid 2-methylcitric acid levels for 5 patients with clinically confirmed cobalamin deficiencies, using the same units and column labels as in Table II. Samples

1-5 are from 5 different patients, while samples 6 and

7 are from the same patient as sample 5 during (sample

6) and after (sample 7) cobalamin therapy.

TABLE VI (cobalamin deficient csf 2-methylcitric acid)

SAMPLE IP TOTMC MCI/MCII MCII MCI

Table VII set forth below shows the declining levels of both serum and urine cystathionine and 2- methylcitric acid levels in a cobalamin deficient patient that is periodically administered 1000 ug doses of cobalamin over a 13 day period. The cobalamin administrations took place on days 0, 2, 6 and 13.

TABLE VII

Table VIII set forth below shows the serum cystathionine and 2-methylcitric acid levels in nanomoles per liter in a cobalamin patient that was mistakenly treated with oral folic acid at the rate of one mg per day from days 0-11, and then was treated with weekly cobalamin injections of 1000 ug starting on day 35. As the table shows, the folic acid treatments had no significant effect in reducing cystathionine or 2-methylcitric acid levels, but the cobalamin treatment did decrease both cystathionine and 2-methylcitric acid to approximately normal levels.

TABLE VIII

PAY CYSTA MCTOTAL MCII MCI

0 552 1242 709 533 23 565 995 529 466 26 513 1915 1128 787 56 180 211 107 104

From this clinical data, it can be concluded that elevated levels of serum or urine cystathionine levels suggest either a cobalamin deficiency or a folic acid deficiency. Further, elevated levels of serum, urine or cerebral spinal fluid 2-methylcitric acid suggest a cobalamin deficiency but not a folic acid deficiency. Once a cobalamin or folic acid deficiency is detected and distinguished, it can be effectively treated with administrations of the deficient compound. A significant decrease or normalization indicates that the deficiency was due to the vitamin used for treatment. The method described herein for detecting and distinguishing between cobalamin and folic acid deficiencies can be used alone or can be used in combination with or as a backup to other methods such as those that rely on measuring homocysteine or methylmalonic acid.