sign in sign up
<<
Home , Folate

Bull. Org. mond. Sante 1963, Bull. Wld Hith Org.j1 28, 517-529

Folates in Megaloblastic Anaemia

J. METZ, M.D.'

The importance of deficiency of the folic acid group of compounds () in the pathogenesis of nutritional anaemias is receiving increasing recognition. There is evidence that the megaloblastic anaemias, due to either B12 or deficiency, may be the cause of widespread morbidity in malnourished populations. It was therefore considered timely to review certain aspects of the role offolates in megaloblastic anaemia, with special reference to the dietary intake in relation to human requirements, and the recognition of in man.

The nutritional effects of folate compounds were The term " folic acid " has since been variously recognized before the chemical structure was known, applied to connote either a single chemical substance and a number of naturally occurring compounds (PGA) or a whole group of unconjugated, conju- were named whose function is now known to be due gated, or formylated compounds which owe their to folate activity. These naturally occurring com- activity to the PGA radical. For this reason it is pounds received the names of Wills' factor, vitamin better to use the term " folate " for the whole group M, vitamin Bc, Norite eluate factor, factor U, of substances, reserving the term " folic acid " for casei factor, L. casei factor, and PGA only. Early work on substances with folate acti- folic acid. In 1945, Angier et al. showed the structure vity has been comprehensively reviewed (Stokstad, of the liver L. casei factor to be N-(4-([(2-amino-4- 1954; Jukes, 1955; Hutner, Nathan & Baker, 1959). hydroxy-6-pteridyl)-methyl]amino)benzoyl}-, and suggested the name pteroylglutamic acid (PGA). It was subsequently demonstrated that the METHODS OF MEASURING FOLATE ACTIVITY naturally occurring forms all had a similar nucleus, 4-(2-amino-4-hydroxy-6-pteridyl)-methyl aminoben- Chemical zoic acid, and were conjugated with one or more Reduction of the pteroylglutamates at acid pH glutamic acid moieties. yields p-aminobenzoylglutamic acid, and this reac- " Folic acid" is a term first given by Mitchell, tion has been suggested as the basis for a chemical Snell & Williams (1941) to a substance obtained from method of measuring compounds with folate activity which stimulated the growth of Streptococcus (Hutchings et al., 1947). The sensitivity is limited. A faecalis andL. casei. The factor was found to be abun- fluorometric method (Allfrey et al., 1949) has been dant in leaves, and was thus named folic acid. This described by which PGA is converted to 2-amino-4- substance has the same biological properties as PGA. hydroxypteridine-6-carboxylic acid, which fluoresces In that a 1948, Sauberlich & Baumann reported strongly at 470 mpt when irradiated with light of folic-acid-like substance was necessary for the growth wave-length 365 The increment in intensity of to be m,t. of Pediococcus cerevisiae (formerly thought fluorescence is directly proportional to PGA con- Leuconostoc citrovorum). The substance thus received centration. If fluorescent pigments which are altered the name of citrovorum factor and is also referred by permanganate oxidation are absent, the reaction to as . The structure was established may be standardized to permit the determination of as 5-formyl-5,6,7,8-tetrahydropteroylglutamic acid PGA in concentrations between 0.01 and 10 acid can pug/ml. (Brockman et al., 1950). Like PGA, folinic These methods are too insensitive for the determina- exist in polyglutamate conjugates. tion of folate activity in small quantities of biological materials, and the results obtained are usually higher 1 Head, Department of Haematology and Radioisotope Laboratory, South African Institute for Medical Research, than those obtained by microbiological assay Johannesburg. methods.

1254 -517 518 J. METZ

Animal assays inactive for S. faecalis or L. casei. However, if the Folate is necessary for the of a number of polyglutamates are converted to monoglutamates by species of animals. The effects of deficiency have conjugase , the results of microbiological been extensively studied in monkeys (Wills & Stewart, assays of folate activity in various natural materials 1935; Day, 1944) and swine (Wintrobe, Samter & are in agreement with the chick assay (Bird et al., Lisco, 1939; Cartwright et al., 1948, 1949). The chick 1945). The assay of folate by micro-organisms must (Campbell et al., 1945; Hutchings, Oleson & therefore be interpreted as a function of their ability Stokstad, 1946; Frost, Dann & McIntire, 1946) and to utilize polyglutamates. the rat (Asenjo, 1948; Ransone & Elvehjem, 1954) Microbiological assays for folate activity are not are commonly used for the assay of folate activity, completely specific. Thymine in adequate amounts most work having been done with the chick. The can completely replace folic acid for S. faecalis, and basal diet in the chick and rat assay is inadequate for partially for L. casei, in assay media which contain maintenance of growth and results in the develop- purine bases. In samples containing significant ment of anaemia. The haemopoietic response to quantities of thymine, high assay values for folate added PGA is parallel to the growth response, but activity may be obtained. the sensitivity is greater when growth rather than haemopoiesis is used in evaluating the results. Animal assays measure PGA as well as its conjugates, DIETARY INTAKE as animals apparently can utilize conjugated forms as readily as the parent compounds. Animal assays Amount offolate in foodstuffs are thus of special nutritional significance, but the Folates in plant foods are present to a large extent method has limitations for use in routine analysis. as polyglutamates, and this may also be true of animal foods, but a variable amount of N-5-methyl- Microbiological THFA may be present. Both PGA and folinic acid Although many micro-organisms require folate have been isolated from natural sources as poly- activity for growth, only L. casei, S. faecalis (for- glutamyl conjugates. The activity of foods has been merly called S. lactis R.) and P. cerevisiae are general- determined by assay with L. casei or S. faecalis, both ly useful for the assay of folate activity. L. casei is before and after conjugase digestion. If the poly- more sensitive to PGA, requiring only one-sixth as glutamates have not been converted to monoglu- much as S. faecalis for growth. These organisms tamate forms, the L. casei value most nearly re- respond differently to various PGA conjugates and presents folate activity, and the S. faecalis value derivatives. S. faecalis and L. casei respond maxi- approximates most closely to the PGA content. mally to the non-reduced monoglutamates except Thus natural products generally give a higher result N-5-methyltetrahydrofolic acid (N-5-methyl-THFA), withL. casei than with S.faecalis; as these organisms which is active for L. casei alone. Increasing numbers are unable to utilize the higher conjugates, values of glutamic acid residues result in rapid reduction in determined by direct assay without conjugase treat- growth response, to which S. faecalis is more sen- ment cannot be applied to animals able to absorb sitive than L. casei. P. cerevisiae responds to all conjugated forms, nor will the post-conjugase treat- known reduced monoglutamate forms except N-5- ment results be strictly applicable until human con- methyl-THFA. Polyglutamates are active for man jugase activity can be accurately determined. and other animals, which probably possess the ability The validity of microbiological assays for folate in to convert the polyglutamates into monoglutamate food depends on a number of other factors, inclu- forms. The efficiency of this utilization is, however, ding the extent of conjugase digestion, and the com- unknown. There is therefore a discrepancy between pleteness of the extraction of the activity from the the utilization of polyglutamates by animals and by natural material. Because of the variable release of micro-organisms. This is of considerable importance folate activity by different sources of conjugase, because naturally occurring compounds with folate Snell (1950) observed that no single extraction activity are almost exclusively polyglutamates (Bird procedure would give maximum assay values for all et al., 1945). Thus the folate in yeast, which is samples. Further complications are the presence of effective in the treatment of megaloblastic anaemia conjugase inhibitors (Swendseid et al., 1947) in associated with (Wills, 1931), occurs natural materials, and the destruction of labile mainly in the heptaglutamic acid form, which is compounds by autoclaving during extraction. FOLATES IN MEGALOBLASTIC ANAEMIA 519

Toepfer et al. (1951) summarized the available curve. They then administered natural information on the folate content of foods, and folate by mouth, measured the urinary excretion of measured the folate activity of a large number of folate, and by reference to the previously constructed foods. Enzymatic digestion was carried out with injected dose/urinary excretion curve, calculated the acetone-dried . Total folate activity amount of folate absorbed. About 25 % of naturally was assayed with L. casei in the -treated occurring folate in yeast, 60 % of the natural activity foods, and PGA activity in untreated foods was after deconjugation with chicken pancreas, and over assayed with S. faecalis. Brewer's yeast, broadleaf 95 % of synthetic PGA were absorbed from the , calabrese , liver concentrate, chicken gastro-intestinal tract in two normal men. Anderson liver, , spinach and leaf were found et al. (1960) administered by mouth 0.2 mg or to be the best sources of folate, while , , corn 40 ,ug/kg body-weight of tritiated PGA, and found and other were poor in folate. The results that control subjects usually excreted less than 25 % presented in tables of food folate activity have been of the administered radioactivity in the faeces. justifiably questioned, as it is uncertain whether or The estimation of folate assimilated from the diet is not they represent the activity available from the further complicated by the fact that intestinal diet. Thus the megaloblastic anaemia of tropical bacterial synthesis may constitute a source of the sprue can be corrected by as little as 25 ,ug a day of vitamin. PGA orally, yet the disease develops in patients receiving a diet containing, according to microbio- Human requirements for folate logical assay, 500-1500 ,ug of folate (Perez-Santiago, Whether a particular dietary intake can be re- Sheehy & Rubini, 1960). The data given in tables garded as adequate depends on the human daily represent sources of natural folate activity, but can- requirement for folate. Human requirements have not be extrapolated to the amount of activity avail- been deduced from the smallest dose of PGA able to man, which can probably be determined only therapeutically active in a patient with folate- by feeding experiments in folate-deficient patients. deficient megaloblastic anaemia. Jandl & Lear (1956) The folate activity in the diet can be greatly obtained a haematological response in two patients altered by the way in which the food is prepared, following the administration of 500 and 250 ,ug of particularly as the vitamin is water-soluble. PGA, respectively, and suggested that 200-300 ,tg of and curing meat may result in losses of up to 95 % PGA daily was adequate therapy for an adult with of the activity (Cheldelin, Woods & Williams, 1943; clinical folate deficiency. Basing their calculations Schweigert, Pollard & Elevehjem, 1946). Losses may on the daily urinary excretion, they suggested that be especially great in canned foods (Ives et al., 1946). the requirements do not exceed 200-300 ,ug per day. The effects of preparing foods in various ways can be Zalusky & Herbert (1961) suggested 50 ,ug a day, deduced from the data compiled by Toepfer et al. and the human requirement is probably of this order. (1951). While daily requirements have been estimated in terms of PGA, probably only a small fraction of Amount offolate in diets natural folate exists in this form. More information Estimates of the dietary intake of folate vary concerning the ability of man to absorb natural widely. Denko et al. (1946) estimated the L. casei folate is necessary before the dietary needs can be active material in a normal diet to average 67 ,ug finally assessed. daily. Jandl & Lear (1956) found the total folate activity of a general hospital diet to range from 1000 LEVELS OF FOLATE ACTIVITY IN to 1500 ,tg daily, while Jukes (1961) suggested that a Whole bloodfolate activity good diet could supply about 400 tsg daily. Pace et Schweigert & Pearson (1947) measured folate al. (1960) estimated the average daily intake of folate activity with S. faecalis in whole blood or plasma in 36 healthy girls to range from 52 to 97 tsg. As diluted with water, and found the level of activity these estimates depend on the microbiological assay to be much greater in whole blood than in plasma. of activity in foods, they do not necessarily reflect After blood samples were digested with takadiastase, the activity available to man. Jandl & Lear (op. cit.) a marked increase in folate activity was noted. This injected varying doses of PGA into normal subjects increase was subsequently shown to be due to con- and measured the urinary folate activity excreted, jugases in the blood releasing folate activity from and in this way constructed an injected dose/urinary the takadiastase preparation (Simpson & Schweigert, 520 J. METZ

1949; Wolff, Drouet & Karlin, 1949). Tonnies, values were obtained with L. casei, which in the Usdin and co-workers in a series of papers (Toennies, maternal blood ranged from 24 to 70.8 to 163 m,ug/ml. Frank & Gallant, 1953; Toennies, Usdin & Phillips, The S. faecalis values were lower and showed a much 1956; Toennies & Phillips, 1959) demonstrated that wider variation than did the L. casei values, ranging dialysates prepared from blood haemolysates showed in maternal blood from 1.85 to 11.0 to 48.0 m,tg/ml. high folate activity resulting from the interaction The folinic acid values in the maternal blood were during and after haemolysis of cell and plasma 0.4-21.2 m,ug/ml. This group of workers sub- factors. The active material was not present in blood sequently studied the folate activity in the whole as such, and it was suggested that a precursor in the blood and serum of patients with anaemia of red cells, when acted upon by an enzyme factor in pregnancy (Izak et al., 1961). In normal persons the the plasma, resulted in the liberation of material with L. casei activity in whole blood averaged 89 m,ug/ml, microbiological folate activity. The active material the S.faecalis activity 12 mpg/ml and the P. cerevisiae was shown to represent a number of compounds activity 6.34 mptg/ml. The activity in serum alone, with high L. casei, low S. faecalis and even lower as compared with that in whole blood, was less P. cerevisiae activity (Usdin, Phillips & Toennies, than 2 % with L. casei and about 5 % with S. faecalis 1956; Usdin, 1959). Neither PGA, pteroyltriglu- and P. cerevisiae. In anaemic pregnant women, tamic acid nor folinic acid seemed to be the major the L. casei levels were significantly lower (mean component. The precursor material and the plasma 62.4 mpg/ml for whole blood and 4.16 m,ug/ml enzymatic factor likewise consisted of several active for serum), but this difference was not found in the subfractions. It was suggested that the greater folate S. faecalis or P. cerevisiae levels. Significantly low activity when measured by L. casei than when values with L. casei were found in about 50% of measured by S. faecalis was probably due to the anaemic patients. Grossowicz et al. (1962) later presence of folate derivatives at the triglutamyl or reported that folate metabolites in the whole blood of higher level. However, only slight increases in patients with nutritional megaloblastic anaemia were S. faecalis activity follow incubation of blood folate remarkably reduced, the most striking reductions concentrates with active conjugase preparations being in L. casei and P. cerevisiae activity. (Izak et al., 1961). Hansen & Nystrom (1961) reported that whole Toennies, Frank & Gallant (1956) measured the blood S. faecalis activity was subnormal in only folate activity released by incubation of blood 6 out of 10 patients with megaloblastic anaemia haemolysates in normal man and in rats. Finger-tip caused by folate deficiency. blood was diluted in phosphate buffer, incubated, That the assay cannot be specific appears likely, dialysed, and then assayed with L. casei. A range of and Herbert et al. (1960) have suggested that the 12 to 150 m,ug/ml was found in 100 normal adults. formed elements of the blood contain sufficient The folate active material of rats dropped to 5 % of folate by-passing compounds to vitiate the assay the normal value on folate-deficient diets. of folate activity as a measure of clinical deficiency. Niewig et al. (1954) measured whole blood folate Although there has been much work on the nature with S. faecalis, but did not give details of the of the folate activity of whole blood, relatively little method used. In 43 normal persons, values were has been published on the measurement of whole 2.9-8.05-24.1 m,ug/ml. In four patients with sprue blood folate levels in folate-deficient humans. Whole and one patient with megaloblastic anaemia of blood folate measured with S. faecalis does not pregnancy the levels were low, but some patients show consistent differences between normal and with nutritional megaloblastic anaemia and megalo- folate-deficient subjects. Some studies suggest blastic anaemia of pregnancy had normal levels. Of significant differences with L. casei and possibly 16 patients with pernicious anaemia, half showed also with S. faecalis and P. cerevisiae. The whole low values. Cox et al. (1960) found the S. faecalis blood assay for folate activity requires more study activity in whole blood samples from 6 patients with in patients with megaloblastic anaemia before its megaloblastic anaemia of pregnancy to be within the place in the detection of folate deficiency can be normal range. evaluated. Grossowicz et al. (1960) measured S. faecalis, L. casei and P. cerevisiae activity in foetal and Serum folate activity maternal whole blood and serum by the method Many attempts have been made to measure folate of Toennies, Frank & Gallant (1952). The highest activity in serum with S. faecalis or L. casei. There FOLATES IN MEGALOBLASTIC ANAEMIA 521 is little or no activity in serum for S. faecalis They subsequently substituted enzymatic casein (Schweigert & Pearson, 1947; Chanarin, Anderson hydrolysate for the -free acid-hydrolysed casein & Mollin, 1958; Pitney, Joske & MacKinnon, 1960) in the medium, increased the amount of ascorbic and measurement of serum folate with this organism acid and reduced the final volume in each assay yields little or no information of value. Consistent tube. Ascorbic acid was added to serum which differences in serum S. faecalis activity between was to be stored. With these modifications, serum normal and folate-deficient subjects have not been folate activity was measured in normal subjects, demonstrated (Cox et al., 1960). in patients with pernicious anaemia, with idiopathic There is, however, activity for L. casei in normal steatorrhoea, and with nutritional folate deficiency, serum. Herbert, Larrabee & Buchanan (1962) have in women with megaloblastic anaemia of pregnancy, presented evidence that this activity, which is not and in normal pregnant women. In normal subjects available to S. faecalis or P. cerevisiae, is due largely the levels were 5.9-21.0 m,tg/ml, while in patients to a monoglutamate form of folate, N-5-methyl- with megaloblastic anaemia, with idiopathic stea- THFA. Narayanan, Shenoy & Ramasarma (1956) torrhoea, and with nutritional folate deficiency the measured the serum L. casei levels of 13 patients levels were less than 3.5 mHtg/ml. There was no with nutritional megaloblastic anaemia and found overlap between the levels in cases of megaloblastic them to be 0.28-0.94-2.55 m,ug/ml; the results in anaemia due to folate deficiency and in those due to normal persons were not given, but these figures deficiency. While the mean level in are considerably lower than those usually reported normal pregnant women was significantly higher in normal subjects. Baker et al. (1959) introduced than that in normal control subjects, in women a modified L. casei serum folate assay. Serum was with severe megaloblastic anaemia of pregnancy diluted in phosphate buffer with added ascorbic the levels were as low as those in patients with acid, and incubated at 37°C for 90 minutes. After idiopathic steatorrhoea and nutritional folic acid autoclaving, the coagulated protein is removed by deficiency; some patients with mild megaloblastic centrifugation, and the supernatant assayed for changes had levels between 4 and 7 mptg/ml. folate activity. In 10 normal persons, the levels Cooper & Lowenstein (1961) found serum values ranged from 7.5 to 24 mjtg/ml, while in 8 patients of less than 3.0 m,ug/ml in all patients with megalo- with alcoholic cirrhosis and megaloblastic anaemia blastic anaemia due to folate deficiency. Stevens et the levels were 1-3.3 m,ug/ml. In 9 patients with al. (1962) reported subnormal serum L. casei levels pernicious anaemia, the levels were 9-23 m,tg/ml. in patients with folate-deficient megaloblastic an- In all cases investigated the levels of folate agreed aemia, and normal or high levels in those with with the clinical findings. In a subsequent report megaloblastic anaemia due to . (Herbert et al., 1960) the levels in 13 patients with The measurement of serum folate with L. casei folate-deficient megaloblastic anaemias associated thus yields valuable information. There is usually with , pregnancy, infancy, and the no overlap between the levels obtained in megalo- syndrome ranged from 0.5 to blastic anaemia cases due to folate deficiency and in 4.8 m,tg/ml. those due to vitamin B12 deficiency. The L. casei Cooperman, Luhby & Avery (1960) modified the assay appears to be a valid index of folate nutrition method, claiming a marked increase in sensitivity except where there is associated vitamin B22 defi- and accuracy. L. casei levels were measured in ciency, which may result in elevated levels. It is normal adults, folate-deficient patients, normal possible that in pregnancy the serum L. casei level pregnant women, folate-deficient pregnant women, may not reflect the folate nutriture of the mother. and normal infants and children. The levels in A different approach to the measurement of body folate-deficient subjects overlapped those of the folate nutrition was introduced by Chanarin, Moilin normal persons, and it was suggested that the assay & Anderson (1958b), who measured the rate of had no clinical usefulness as a diagnostic test. clearance from the plasma of PGA injected intra- Waters & Mollin (1961) reported that, with the venously with S. faecalis. Schweigert (1948) had method of Baker et al. (1959), they found consider- earlier injected a large dose (12 mg) of PGA or an able overlap between the results in normal subjects equivalent amount of pteroyltriglutamic acid intra- and those in patients with folate deficiency. There venously into normal subjects. Whole blood samples was poor reproducibility in repeat assays, and the were drawn 2, 4, 8 and 24 hours after the injection, diagnostic value of the method was therefore limited. and the folate activity was assayed with S. faecalis 522 J. METZ and L. casei. Chanarin, Mollin & Anderson (1958a) subject to subject. The bulk of the folinic acid injected PGA intravenously, in a dose of 15 ,ug/kg, activity in urine is in an oxygen-labile form which is into normal persons and patients with megaloblastic destroyed rapidly in air (Silverman, Ebaugh & anaemia and measured the folate activity with Gardiner, 1956). Jandl & Lear (1956) found the S. faecalis 3, 15 and 30 minutes after the injection. urinary L. casei activity in 8 normal men to range The rate of clearance was abnormally high in all the from 1.58 to 3.93 to 5.94 ,ug/day, while in 4 patients patients with folate-deficient megaloblastic anaemia, with megaloblastic anaemia associated with cirrhosis, and this increased rate was considered to indicate the levels ranged from 0.51 to 1.38 to 2.52 ,ug/day. folate deficiency. Of 23 patients with vitamin B12 Chatterjea (1960) reported that no significant dif- deficiency, however, the clearance was abnormally ference could be demonstrated in S. faecalis activity rapid in 12. between patients with nutritional megaloblastic Burgen (1961) injected tritiated PGA and found anaemia and normal subjects. Levels of folate the plasma clearance more rapid than normal in activity in the urine are generally too low to provide patients with pernicious anaemia and non-tropical any information of value in the diagnosis of folate sprue. deficiency, and measurement of urinary folate acti- Chanarin et al. (1959) reported that the rate of vity will not differentiate normal subjects from those clearance of injected PGA was increased in two- with folate deficiency. thirds of women in late pregnancy, and in all women Attempts have been made to demonstrate dif- with twin pregnancy. Girdwood & Delamore (1961) ferences in the urinary excretion of folate in normal likewise found abnormally rapid clearance in and folate-deficient subjects following administra- pregnant women. tion of a test dose of PGA. Steinkamp et al. (1946) Metz et al. (1961), using the same technique, measured urinary S. faecalis activity following an found the clearance to be abnormally rapid in all oral dose of PGA. The mean recovery in normal patients with megaloblastic anaemia due to either subjects was 28.5%. Hospitalized patients excreted folate or vitamin B12 deficiency. Abnormally rapid less of the oral dose, and it was suggested that this clearance was found also in non-anaemic mal- might indicate a low degree of saturation. Bethell nourished females. et al. (1947) found the urinary L. casei activity fol- Herbert & Zalusky (1962) measured serum folate lowing oral administration of the hexaglutamyl con- levels after intravenous injection of PGA with both jugate of PGA to be lower in patients with macro- S. faecalis and L. casei. While the S. faecalis activity cytic anaemia than in normal subjects. disappeared rapidly in both folate- and vitamin-B12- Spray, Fourman & Witts (1951) studied the ex- deficient megaloblastic anaemias, the L. casei activity cretion of small doses of PGA given intravenously to disappeared rapidly in folate deficiency, but slowly normal subjects, patients with steatorrhoea, and in vitamin B12 deficiency. treated and untreated patients with pernicious The rapidity of clearance from the plasma of PGA anaemia. Folate activity was assayed with L. casei injected intravenously is a sensitive index of folate in a 24-hour specimen of urine following the injec- deficiency, but may be abnormal also in vitamin B12 tion of PGA. Three patients with steatorrhoea and deficiency. The technique has the disadvantage of three patients with untreated pernicious anaemia requiring the administration of a therapeutically excreted less of the injected dose than did normal active dose of PGA. subjects. A number of workers (Swendseid et al., 1952; IN URINE Girdwood, 1953; Tasker, 1955; Cox et al., 1958; FOLATE ACTIVITY Chatterjea, 1960) have measured with S. faecalis the Small amounts of folate activity are excreted in folate activity excreted in a 24-hour urine specimen the urine. With S. faecalis, the levels in 24-hour following the injection of a 5-mg dose of PGA. The samples range from 0 to 5.5 ,tg (Steinkamp et al., mean excretion in normal subjects has been reported 1946; Girdwood, 1953; O'Brien, 1960); withL. casei as 1.5 mg (Swendseid et al.), 2.02 mg (Tasker), from 1.0 to 10.0 ,ug (Register & Sarett, 1951; Spray, 2.62 mg (Cox et al., 1958), and 2.58 mg (Chatterjea). Fourman & Witts, 1951; Jandl & Lear, 1956); and Patients with leukaemia retained a larger proportion with P. cerevisiae from 0.19 to 1.0 ,ug (Register & of the dose, and it was suggested that this was Sarett, 1951; Jandl & Lear, 1956). The urinary folate probably due to folate deficiency (Swendseid et al., activity varies widely from day to day and from 1952). FOLATES IN MEGALOBLASTIC ANAEMIA 523

The test has been reported as abnormal in 6 out That folate may play an additional role in of 18 patients with megaloblastic anaemia associated degradation is indicated by the report that liver with pregnancy (Girdwood, 1953; Layrisse et al., urocanase is lowered in folate-deficient rats (Bald- 1960), and in 8 out of 29 patients with nutritional ridge, 1958) suggesting that folate may be involved megaloblastic anaemia (Tasker, 1955). as a co-factor in an intermediate reaction in the con- Cox et al. (1958) found normal subjects excreted version of urocanic acid to a formamidinoglutaric 1.39-2.62-3.21 mg, while 20 patients with idiopathic acid. Folate may also be involved in tryptophane steatorrhoea excreted 0.25-2.35-3.13 mg, and con- in that formic acid, possibly derived from cluded that the test was unsatisfactory for determi- tryptophane metabolism, appears in the urine in high ning folate deficiency. Similarly Chatterjea (1960) concentration in folate-deficient animals (Rabinowitz reported that in patients with nutritional megalo- & Tabor, 1958). Of these metabolic disturbances, blastic anaemia the mean excretion was lower than only that of histidine degradation has been exten- in controls, but there was considerable overlap when sively applied to the study of folate deficiency. the values in nutritional megaloblastic anaemia were compared with those obtained in deficiency and Methods for the detection of FIGLU in urine tropical eosinophilia, and the value of the test as an A number of methods have been described for the indicator of folate deficiency was therefore limited. detection of FIGLU in the urine. Microbiological Thus the measurement of urinary folate activity, assay with L. arabinosus (Silverman, Gardiner & either with or without administration of a test dose Bakerman, 1952) utilizes the method of Henderson of PGA, will not consistently differentiate normal & Snell (1948) for glutamic acid estimation, the from folate-deficient subjects. Most of these studies FIGLU being converted to glutamic acid by auto- were, however, carried out without taking adequate claving. The sensitivity is about 10 ,tg/ml. The precautions to prevent destruction of folate activity assay is complicated by the glutamic acid and glu- in the urine. tamine normally present in the urine. If interfering glutamic acid and are first removed by allowing the organism to grow in unheated urine EVIDENCE OF DISTURBANCE IN METABOLISM AS A (Davis & Onesti, 1960) the specificity of the assay is MEASURE OF FOLATE DEFICIENCY improved. While the sensitivity has been increased by repeated subculture of the organism in progres- Role offolate in metabolism sively decreasing concentrations of glutamic acid, it Folate has been shown to be active as a 1-carbon remains relatively poor and the method is very time- unit donor in a number of important metabolic pro- consuming. Urinary FIGLU can also be detected by cesses, including the closure of the purine ring, the paper chromatography (Broquist, 1956) but the sen- of pyrimidine, the inter-conversion of sitivity is poor, being in the range of 20-50 ,ug/ml. and , and the formation of Silverman, Gardiner & Condit (1958) introduced from homocystine. The 1-carbon units are trans- a combined enzymatic and microbiological method. ferred as hydroxymethyl (-CH2OH), formyl (CHO) Liver extract was used to reduce PGA to tetrahydro- or formimino (-CH-NH) groups. An example of this folic acid (THFA), which was formylated by the function of folate which has received considerable FIGLU in the test urine. The N-10-formyl-THFA attention is in the transfer of the 1-carbon unit as a was converted to N-5-formyl-THFA, which was formyl group from assayed microbiologically with P. cerevisiae. This (FIGLU), an intermediate in the metabolism of method is more sensitive than microbiological assay, histidine. In 1951, Bakerman, Silverman & Daft but is also time-consuming. It measures urocanic detected a precursor of glutamic acid in the urine acid and perhaps other intermediates of histidine of folate-deficient rats. This material was subse- metabolism as well, and to determine the amount of quently identified as FIGLU (Silverman, Gardiner FIGLU only, urine samples are assayed before and & Bakerman, 1952), and shown to be an intermediate after autoclaving. product of histidine metabolism (Tabor et al., 1953). Tabor & Wyngarden (1958) described an entirely For the further metabolism of FIGLU to glutamic enzymatic assay method. Urine is incubated in acid, folate in the form of is phosphate buffer with a mixture of transferase, required (Tabor & Rabinowitz, 1956; Miller & cyclodeaminase and THFA. During the reaction Waelsch, 1956). 5-10-methenyl-THFA is converted to 10-formyl- 524 J. METZ

THFA, and the latter is completely reconverted to reported the presence of FIGLU in the urine of 5-10-methenyl-THFA upon acidification, and is read patients with megaloblastic anaemia of pregnancy. in a spectrophotometer at 350 m,u. The enzyme As differences in FIGLU excretion by folate-defi- preparation has traces of urocanase. While this cient individuals may vary with the dietary intake of enzymatic assay is the most sensitive and specific histidine, a histidine metabolic loading procedure to technique (0.1 ,tg of FIGLU per ml can be detected), equalize the quantity of histidine presented for it requires the use of highly unstable THFA. metabolism was devised (Luhby et al., 1958). Fol- Knowles, Prankerd & Westall (1960) have re- lowing an 8-12 g loading-dose of histidine, urinary ported a method for detecting urinary FIGLU by FIGLU was measured microbiologically. Normal high-voltage paper electrophoresis. After staining subjects did not excrete detectable amounts, but in with ninhydrin, 10 ,ug of FIGLU per ml could be macrocytic anaemia of pregnancy large amounts detected. The procedure can be made semi-quantita- were excreted. The excess FIGLU was abolished by tive by serial dilution of the urine. Kohn, Mollin & PGA therapy. FIGLU was not detected in patien's Rosenbach (1961) introduced conventional voltage with pernicious anaemia. Luhby, Cooperman & electrophoresis for FIGLU determination. FIGLU Teller (1959) subsequently reported the results in a is separated together with glutamic acid by cellulose series of normal subjects and anaemic patients. A acetate electrophoresis. One half of the strip is 15-g loading-dose of l-histidine monohydrochloride exposed to ammonia vapour, which converts any was administered daily for 2 to 3 days, all the urine FIGLU present to glutamic acid, which will then was collected, and the FIGLU was measured by the stain with ninhydrin. A semi-quantitative estimation enzyme technique of Tabor & Wyngarden (1958). of the amount of FIGLU present can be obtained by Urinary FIGLU in all patients with non-pernicious comparing the intensity of the colour of a marker megaloblastic anaemias was excessive, while normal spot with the patient's FIGLU spot. Zalusky & amounts were excreted by patients with pernicious Herbert (1962) eliminate exposure to ammonia anaemia. vapour by the use of strong alkali before electro- Spray & Witts (1959) measured urinary FIGLU by phoresis. The sheets are sprayed with ninhydrin, the method of Silverman, Gardiner & Condit (1958) and the colour quantitated after elution. after administering a 2-g histidine load to normal Chanarin & Bennett (1962) describe a simplified subjects, patients with megaloblastic anaemia due quantitative method based on that of Silverman, to folate deficiency, and patients with megaloblastic Gardiner & Condit (1958). A crude liver enzyme pre- anaemia due to deficiency of vitamin B12. Negligible paration is employed to reduce PGA to THFA and amounts of FIGLU were detected in the urine of the then to transfer the formimino group of FIGLU to normal subjects. In the folate-deficient patients, the THFA. The end-product of the reaction is con- FIGLU was excessive. In patients with vitamin B12 verted to 5-10-methenyl-THFA with hydrochloric deficiency, there was a slight excess of FIGLU but acid and read spectrophotometrically. The crude the levels were considerably lower than those in the liver contains urocanase, so that urocanic acid is folate-deficient patients. converted to FIGLU, and the method measures Knowles, Prankerd & Westall (1960) administered combined FIGLU and urocanic acid. This may be of a 20-g histidine load, collected urine for 12 hours, importance as the authors claim that urocanic acid and measured the FIGLU electrophoretically. In may comprise more than 80 % of the histidine deriva- normal subjects, FIGLU could not be detected, while tives appearing in the urine of patients with folate in 5 patients with steatorrhoea there was a marked deficiency. The urocanic acid and FIGLU may be excess. Kohn, Mollin & Rosenbach (1961) collected estimated separately after destroying the FIGLU by urine between the third and the eighth hour after autoclaving. the administration of 15 g of histidine. All patients with clinical and microbiological evidence of folate Urinary FIGLU in normal and in folate-deficient deficiency showed an increase in urinary FIGLU. persons Of 14 patients with vitamin B12 deficiency, 10 showed Increased urinary FIGLU in man was first demon- no increase in FIGLU. Zalusky & Herbert (1962) strated in leukaemic patients receiving treatment administered a 20-g histidine load and measured with folate antagonists (Broquist, 1956; Hiatt, the FIGLU electrophoretically in a 12-hour urine Goldstein & Tabor, 1958) and was interpreted as specimen. In 6 normal subjects the amount of indicative of induced folate deficiency. Luhby (1957) FIGLU excreted ranged from 0 to 19 to 54 mg, FOLATES IN MEGALOBLASTIC ANAEMIA 525 while 8 subjects with folate deficiency excreted 253- This was confirmed by chromatography of urine, 561-705 mg and 7 subjects with vitamin B12 deficiency using pure urocanic acid for comparison. After excreted 12-288-989 mg. administration of the histidine load, urocanic acid Griisbeck, Bjorksten & Nyberg (1961) measured and FIGLU were present in the urine of normal urinary FIGLU by the method of Tabor & Wyn- subjects and of patients with pernicious anaemia, garden (1958) after administering 3 daily doses of post-gastrectomy megaloblastic anaemia, idiopathic 15 g ofhistidine. All patients with idiopathic steator- steatorrhoea, iron-deficiency anaemia, haemolytic rhoea and studied excreted an excess anaemia, carcinomatosis, and acute leukaemia. of FIGLU. In 3 out of 4 patients with severe vitamin Urocanic acid comprised 0-97% of the histidine B12 deficiency there was a mild excess of FIGLU, but derivatives excreted in the urine. in no patient were the levels as high as the levels in folate deficiency. The authors conclude that there is no evidence that vitamin B12 deficiency alone DIAGNOSIS OF FOLATE DEFICIENCY causes increased urinary excretion of FIGLU. It is possible to advance some comments on the Metz, Stevens & Brandt (1962), using a modified available tests of folate deficiency. Three procedures microbiological assay, reported that after a 2-g histi- warrant full consideration at this stage-namely, the dine load 6 out of 11 patients with megaloblastic an- plasma clearance of injected doses of PGA; the aemia due to folate deficiency excreted FIGLU in the serum L. casei level; and the urinary excretion of urine, while no FIGLU was detected in 14 patients FIGLU after a histidine load. It should be appreci- with vitamin B22 deficiency. When the histidine load ated that these three tests probably measure dif- was increased to 5 g all the folate-deficient patients ferent and distinct parameters. The clearance of excreted FIGLU; with this load, 3 out of 9 patients injected PGA is a measure mainly of the rate at with vitamin B12 deficiency also excreted FIGLU, but which the tissue receptors (and thus probably folate in smaller amounts than the folate-deficient patients. stores) remove the monoglutamate entering the Chanarin, Bennett & Berry (1962) measured plasma. Serum L. casei levels may reflect the amount urinary FIGLU and urocanic acid by their modified of folate fed into the plasma from tissue stores, enzymatic method after a 1 5-g histidine load. possibly representing the folate available to the sites Abnormal amounts of histidine derivatives were of utilization. Urinary FIGLU is a measure of a excreted in 9 out of 10 patients with megaloblastic function of folate, the transfer of a 1-carbon unit anaemia due to folate deficiency. Abnormal amounts which takes place in the liver, and is therefore were also excreted by 10 out of 15 patients with probably a measure of the liver stores of folate. In vitamin B12 deficiency, but the amounts excreted the folate-depleted patient, all three procedures will were smaller than those excreted by the folate- demonstrate the defect; when, however, there is deficient patients. deficiency of vitamin B12, the differences in the para- Considerable data are thus available on the meters become manifest. The three tests will be urinary excretion of FIGLU as an index of folate considered briefly with reference to sensitivity, deficiency. Consistently increased excretion has specificity and practicability. been demonstrated in patients with folate deficiency. In the patient whose folate deficiency is severe A proportion of patients with primary vitamin B12 enough to cause megaloblastic anaemia, all three deficiency will also show increased urinary FIGLU, tests are usually abnormal. In the diagnosis of defi- but in lesser amounts than patients with folate ciency prior to the development of anaemia, all three deficiency. tests have been reported as abnormal (Chanarin, Mollin & Anderson, 1958a; Knowles, Prankerd Excretion of urocanic acid in folate deficiency & Westall, 1960; Cooper & Lowenstein, 1961; Metz Bennett & Chanarin (1961), in studying the urinary et al., 1961), but there is limited information avail- excretion of histidine derivatives after a 1 5-g loading- able on the relative sensitivity of the various pro- dose of histidine, found that a high proportion of cedures when applied to the same patient. Extensive the material in urine which donated a single carbon studies (Chanarin et al., 1959; Girdwood & Dela- unit to THFA was both heat- and alkali-stable. This more, 1961) on the plasma clearance of injected doses suggested that the material was not FIGLU, which is of PGA in pregnancy leave no doubt as to the rapidly destroyed by autoclaving at alkaline pH, sensitivity of this procedure. Latent folate deficiency and that the substance concerned was urocanic acid. is most likely to become manifest during pregnancy, 526 J. METZ and application of the three tests to the same patient of the test as an index of folate deficiency depends during pregnancy should provide an assessment of on the attitude to secondary changes in folate meta- their relative sensitivity. bolism resulting from vitamin B12 deficiency. All three tests have been reported as abnormal in In the plasma clearance test PGA is assayed with conditions where folate deficiency is well recognized. S. faecalis, a relatively simple procedure in a labora- However, the clearance of injected doses of PGA is tory carrying out microbiological assays. An im- often rapid in patients with vitamin B12 deficiency, portant drawback of the test, however, is that it indicating increased avidity of body tissues for PGA; involves the administration of a dose of PGA which similarly, the urinary excretion of PGA after a test may be therapeutically effective. The L. casei assay dose is diminished in vitamin B12 deficiency. Whether is more difficult to perform in that stringent precau- this rapid clearance in the case of vitamin B12 defi- tions to protect the labile folate activity to be as- ciency can be regarded as detracting from the sayed are necessary. The FIGLU test was initially specificity of the test is debatable; if the finding is of limited application, owing to the formidable enzy- construed as the result of depletion of folate stores matic and microbiological procedures involved. conditioned by the vitamin B12 deficiency, the test Electrophoretic methods have greatly simplified the remains specific for tissue depletion of PGA. The detection of FIGLU, but are relatively insensitive, L. casei serum folate level will differentiate vitamin and not strictly quantitative. The simplified enzy- B12 from folate deficiency. Excess urinary FIGLU matic procedure of Chanarin & Bennett (1962) is has been reported in patients ?pparently suffering both sensitive and quantitative, and is suitable for from primary vitamin B12 deficiency, but in lesser routine analysis. amounts than in patients with folate deficiency. The In our present state of knowledge of the tests of differences in FIGLU excretion between folate- and folate nutrition, it is desirable that as many para- B12-deficient subjects may be related to the size of meters as possible be examined. Furthermore, in the histidine load. As only a small proportion of the view of the secondary changes in folate metabolism reported cases of vitamin B12 deficiency have demon- which may be associated with vitamin B12 deficiency, strated excess urinary FIGLU, it can be said that knowledge of the serum vitamin B12 concentration is vitamin B12 deficiency per se does not result in excess of considerable value in the interpretation of the urinary FIGLU. Again, therefore, the interpretation results of tests of folate nutrition.

ACKNOWLEDGEMENTS

I wish to thank Dr C. A. Finch, Dr V. Herbert, Dr and Dr V. N. Patwardhan for their help in preparing this D. L. Mollin, Dr. M. Rachmilewitz, Dr M. M. Wintrobe review.

RItSUMIl

L'auteur pense que le terme <( derives foliques > ou des folates exprime en fait la capacite des micro-orga- folates # est preferable a celui d'(( acide folique )) que l'on nismes 'a utiliser des polyglutamates. doit reserver a l'acide pteroyl-glutamique (PGA). La quantite de folates d'origine alimentaire est etudiee. L'auteur passe en revue differentes methodes de mesure Les donnees fournies par les tables du contenu en sub- de l'activite des folates. Les methodes chimiques sont peu stance foliques des produits alimentaires representent applicables lorsqu'il s'agit de determiner l'activite folique bien les sources de l'activite folique naturelle, mais l'on de petites quantites de produits naturels. Les essais sur ne peut en extrapoler la quantite d'activite utile a l'orga- l'animal ont un interet tout particulier sur le plan nutri- nisme humain. De meme, lorsqu'il s'agit de la quantite tionnel du fait qu'ils mesurent aussi bien l'acide pteroyl- de folates contenus dans certains regimes, des estimations glutamique que ses derives; cependant, ils sont difficile- basees sur les resultats des essais microbiologiques des ment utilisables pour l'analyse courante. Les micro- aliments ne renseignent pas necessairement sur l'activite organismes , Streptococcus faecalis et folique fournie a l'organisme. Pediococcus cerevisiae au moyen desquels l'on teste I'acti- Les besoins de l'homme en substances foliques ont et vite folique repondent de fa9on differente aux divers deduits de la plus petite quantite d'acide pteroyl-gluta- conjugues et derives foliques, et un essai microbiologique mique active chez un sujet atteint d'anemie megaloblas- FOLATES IN MEGALOBLASTIC ANAEMIA 527 tique par carence folique; ils sont probablement de 50g convenient d'exiger l'administration d'une dose thera- par jour. Du fait que ce besoin journalier est exprime en peutique d'acide pteroyl-glutamique. quantites d'acide pteroyl-glutamique et que seule une La mesure des folates dans l'urine avec ou sans admi- petite partie des substances foliques naturelles existe sous nistration d'une dose standard d'acide pteroyl-glutamique cette forme, l'on a besoin d'en savoir davantage sur ne donne pas chez le sujet carence des chiffres tres diff& l'absorption digestive des folates naturels avant d'etablir rents de ceux obtenus chez l'homme normal. de fagon certaine les besoins alimentaires. L'acide a-forminino-L-glutamique (FIGLU), produit L'auteur passe en revue les methodes actuelles de intermediaire du metabolisme de l'histidine, s'accumule diagnostic de carence folique chez l'homme. Le taux de dans l'urine des sujets carences en folates. Le FIGLU peut folates dans le sang total mesure grace a S. faecalis n'est etre mis en evidence dans l'urine par des techniques pas tres different chez le sujet normal et chez le malade microbiologiques, chromatographiques, electrophor& atteint de carence folique. Cependant, certaines etudes tiques ou enzymatiques, ces dernieres etant les plus incitent a penser que des differences significatives peuvent sensibles et les plus precises. L'existence d'une excretion apparaitre lorsqu'on etudie la mesure a l'aide de L. casei; urinaire marquee de FIGLU a et demontree chez des peut-8tre meme aussi avec S. faecalis et P. cerevisiae. Le malades atteints de carence folique. Certains sujets test de l'activite folique du sang total doit faire l'objet de atteints de carence primitive en vitamine B12 peuvent nouvelles etudes chez des malades atteints d'anemie mega- egalement avoir un taux urinaire exagere de FIGLU, mais loblastique avant que son r6le puisse etre evalue pour le habituellement cette augmentation est moins marquee. depistage des carences foliques. Les trois methodes les plus utilisees pour le diagnostic L'on n'a pu jusqu'ici mettre en evidence par l'inter- des carences foliques chez l'homme sont donc la clearance mediaire de S. faecalis des differences notables entre la plasmatique de l'acide pteroyl-glutamique injecte par teneur en folates seriques de sujets normaux et de sujets voie veineuse, le taux d'acide folique du serum mesure carences. L'utilisation de L. casei pour determiner le taux grace a L. casei et l'excretion urinaire de l'acide a-formi- de substance folique dans le serum fournit d'excellents nino-L-glutamique (FIGLU) apres surcharge en histidine. renseignements et habituellement les chiffres obtenus dans I1 est probable que ces tests mesurent des parametres les cas d'anemie megaloblastique par carence folique et differents les uns des autres. Lorsque la carence folique dans ceux d'anemie megaloblastique par hypovitaminose est assez importante pour entrainer une anemie megalo- B12 sont bien distincts. La methode utilisant L. casei blastique, les trois tests donneront des reponses positives. parait donner une indication valable de la richesse en Lorsqu'il s'agit de detecter des carences foliques au stade folates, excepte lorsqu'il y a carence combinee en acide pre-anemique, l'on a rapporte des resultats anormaux folique et en vitamine B12, peut-etre aussi au cours de avec les trois tests. Cependant, l'on ne sait que peu de la grossesse. choses sur leur sensibilite respective lorsqu'ils sont tous La clearance plasmatique de l'acide pteroyl-glutamique trois effectues chez le meme sujet. Dans l'etat actuel de injecte par voie veineuse est un indice sensible de la nos connaissances sur les tests d'absorption des folates, carence folique, mais peut etre egalement anormale en il est souhaitable de s'adresser au plus grand nombre cas de carence vitaminique B12. Cette methode a l'in- possible de parametres.

REFERENCES

Allfrey, V., Teply, L. J., Geffen, C. & King, C. G. (1949) Baldridge, R. C. (1958) J. biol. Chem., 231, 207 J. biol. Chem., 178, 465 Bennett, M. C. & Chanarin, I. (1961) Lancet, 2, Anderson, B., Belcher, E. H., Chanarin, I. & Mollin, 1095 D. L. (1960) Brit. J. Haemat., 6, 439 Bethel, F. H., Meyers, M. C., Andrews, G. A., Arbor, A., Angier, R. B., Boothe, J. H., Hutchings, B. L., Mowat, Swendseid, M. E., Bird, 0. D. & Brown, R. A. (1947) J. H., Semb, J., Stokstad, E. L. R., SubbaRow, Y., J. Lab. clin. Med., 32, 3 Waller, C. W., Cosulich, D. B., Fahrenbach, M. J., Bird, 0. D., Bressler, B., Brown, R. A., Campbell, C. J. Hultquist, M. E., Kuh, E., Northey, E. A., Seeger, & Emmett, A. D. (1945) J. biol. Chem., 159, 631 D. R., Sickels, J. P. & Smith, J. M., jr (1945) Science, Brockman, J. A., jr, Roth, B., Broquist, H. P., Hultquist, 102, 227 M. E., Smith, J. M., jr, Fahrenbach, M. J., Cosulich, Asenjo, C. F. (1948) J. Nutr., 36, 601 D. B., Parker, R. P., Stokstad, E. L. R. & Jukes, T. H. Baker, H., Herbert, V., Frank, O., Pasher, I., Hutner, (1950) J. Amer. chem. Soc., 72, 4325 S. H., Wasserman, L. R. & Sobotka, H. A. (1959) Broquist, H. P. (1956) J. Amer. chem. Soc., 78, 6205 Clin. Chem., 5, 275 Burgen, A. S. V. (1961) Blood, 17, 367 Bakerman, H. A., Silverman, M. & Daft, F. S. (1951) Campbell, C. J., McCabe, M. M., Brown, R. A. & J. biol. Chem., 188, 117 Emmett, A. D. (1945) Amer. J. Physiol., 144, 348 528 J. METZ

Cartwright, G. E., Fay, J., Tatting, B. & Wintrobe, M. M. Hutchings, B. L., Stokstad, E. L. R., Boothe, J. H., (1948) J. Lab. clin. Med., 33, 397 Mowat, J. H., Waller, C. W., Angier, R. B., Semb, J. Cartwright, G. E., Tatting, B., Ashenbrucker, H. & & SubbaRow, Y. (1947) J. biol. Chem., 168, 705 Wintrobe, M. M. (1949) Blood, 4, 301 Hutner, S. H., Nathan, H. A. & Baker, H. (1959) Vitam. Chanarin, I., Anderson, B. B. & Mollin, D. L. (1958) and Horm., 17, 1 Brit. J. Haemat., 4, 156 Ives, M., Pollard, A. E., Elvehjem, C. A. & Strong, F. M. Chanarin, I. & Bennett, M. C. (1962) Brit. med. J., 1, 27 (1946) J. Nutr., 31, 347 Chanarin, I., Bennett, M. C. & Berry, V. (1962) J. clin. Izak, G., Rachmilewitz, M., Sadovsky, A., Bercovici, B., Path., 15, 269 Aronovitch, J. & Grossowicz, N. (1961) Amer. J. Chanarin, I., MacGibbon, B. M., O'Sullivan, W. & clin. Nutr., 9, 473 Mollin, D. L. (1959) Lancet, 2, 634 Jandl, J. & Lear, A. (1956) Ann. intern. Med., 45, 1027 Chanarin, I., Mollin, D. L. & Anderson, B. B. (1958a) Jukes, T. H. (1955) Meth. biochem. Anal., 2, 121 Brit. J. Haemat., 4, 435 Jukes, T. H. (1961) J. chron. Dis., 14, 283 Chanarin, I., Mollin, D. L. & Anderson, B. B. (1958b) Knowles, J. P., Prankerd, T. A. J. & Westall, R. G. (1960) Proc. roy. Soc. Med., 51, 757 Lancet, 2, 347 Chatterjea, J. B. (1960) Bull. Calcutta Sch. trop. Med., Kohn, J., Mollin, D. L. & Rosenbach, L. M. (1961) 8, 9 J. clin. Path., 14, 345 Cheldelin, V. H., Woods, A. M. & Williams, R. J. (1943) Layrisse, M., Aguero, O., Blumenfeld, N., Wallis, H., J. Nutr., 26, 477 Dugarte, I. & Ojeda, A. (1960) Blood, 15, 724 Cooper, B. A. & Lowenstein, L. (1961) Canad. med. Ass. Luhby, A. L. (1957) Clin. Res. Proc., 5, 8 J., 85, 987 Luhby, A. L., Cooperman, J. M. & Teller, D. N. (1959) Cooperman, J. M., Luhby, A. L. & Avery, C. M. (1960) Proc. Soc. exp. Biol. (N.Y.), 101, 350 Proc. Soc. exp. Biol. (N.Y.), 104, 536 Luhby, A. L., Cooperman, J. M., Teller, D. N. & Cox, E. V., Matthews, D., Meynell, M. J., Cooke, W. T. Donnenfeld, A. M. (1958) J. clin. Invest., 37, 915 & Gaddie, R. (1960) Blood, 15, 376 Metz, J., Stevens, K. & Brandt, V. (1962) Brit. med. J., Cox, E. V., Meynell, M. J., Cooke, W. T. & Gaddie, R. 2, 1440 (1958) Gastroenterology, 35, 390 Metz, J., Stevens, K., Krawitz, S. & Brandt, V. (1961) Davis, R. E. & Onesti, P. (1960) Aust. J. exp. Biol. med. J. clin. Path., 14, 622 Sci., 38, 239 Miller, A. & Waelsch, H. (1956) Arch. Biochem., 63, Day, P. L. (1944) Vitam. and Horm., 2, 71 263 Denko, C. W., Grundy, W. E., Wheeler, N. C., Hender- Mitchell, H. K., Snell, E. E. & Williams, R. J. (1941) son, C. R., Berryman, C. H., Friedmann, T. E. & J. Amer. chem. Soc., 63, 2284 Youmans, J. B. (1946) Arch. Biochem., 11, 109 Narayanan, M. S., Shenoy, K. G. & Ramasarma, G. B. Frost, D. V., Dann, F. P. & McIntire, F. C. (1946) Proc. (1956) Nature (Lond.), 178, 1347 Soc. exp. Biol. (N. Y.), 61, 65 Nieweg, H. O., Faber, J. G., Vries, J. A. de & Kroese, Girdwood, R. H. (1953) Lancet, 2, 53 W. F. S. (19541 J. Lab. clin. Med., 44, 118 Girdwood, R. [H.] & Delamore, I. W. (1961) Scot. med. J., O'Brien, J. S. (1960) Proc. Soc. exp. Biol. (N. Y.), 104, 354 6, 44 Pace, J. K., Stier, L. B., Taylor, D. D. & Goodman, P. A. Grasbeck, R., Bjorksten, F. & Nyberg, W. (1961) (1960) Fed. Proc., 19, 415 Nord. Med., 66, 1343 Perez-Santiago, E., Sheehy, T. W. & Rubini, M. E. (1960) Grossowicz, N., Aronovitch, J., Rachmilewitz, M., Clin. Res., 8, 215 Izak, G., Sadovsky, A. & Bercovici, B. (1960) Brit. J. Pitney, W. R., Joske, R. A. & MacKinnon, N. L. (1960) Haemat., 6, 296 J. clin. Path., 13, 440 Grossowicz, N., Rachmilewitz, G., Izak, G. & Zan, S. Rabinowitz, J. C. & Tabor, H. (1958) J. biol. Chem., (1962) Proc. Soc. exp. Biol. (N. Y.), 109, 770 233, 252 Hansen, H. A. &Nystr6m, B. (1961) Geront. clin. (Basel), Ransone, B. & Elvehjem, C. A. (1954) J. biol. Chem., 3, 173 151, 109 Henderson, L. M. & Snell, E. E. (1948) J. biol. Chem., Register, U. D. & Sarett, H. P. (1951) Proc. Soc. exp. 172, 15 Biol. (N.Y.), 77, 837 Herbert, V., Baker, H., Frank, O., Pasher, I., Sobotka Sauberlich, H. E. & Baumann, C. A. (1948) J. biol. Chem., H. & Wasserman, L. R. (1960) Blood, 16, 228 176, 165 Herbert, V., Larrabee, A. R. & Buchanan, J. M. (1962) Schweigert, B. S. (1948) J. Lab. clin. Med., 33, 1271 J. clin. Invest., 41, 1134 Schweigert, B. S. & Pearson, P. B. (1947) Amer. J. Herbert, V. & Zalusky, R. (1962) J. clin. Invest., 41, 1263 Physiol., 148, 319 Hiatt, H. H., Goldstein, M. & Tabor, H. (1958) J. clin. Schweigert, B. S., Pollard, A. E. & Elvehjem, C. A. (1946) Invest., 37, 829 Arch. Biochem., 10, 107 Hutchings, B. L., Oleson, J. J. & Stokstad, E. L. R. (1946) Silverman, M., Ebaugh, F. G., jr & Gardiner, R. C. (1956) J. biol. Chem., 163, 447 J. biol. chem., 223, 259 FOLATES IN MEGALOBLASTIC ANAEMIA 529

Silverman, M., Gardiner, R. C. & Bakerman, H. A. (1952) Toennies, G., Frank, H. G. & Gallant, D. L. (1952) J. biol. Chem., 194, 815 Growth, 16, 287 Silverman, M., Gardiner, R. C. & Condit, P. T. (1958) Toennies, G., Frank, H. G. & Gallant, D. L. (1953) J. nat. Inst., 20, 71 J. biol. Chem., 200, 23 Simpson, R. E. & Schweigert, B. S. (1949) Arch. Biochem., Toennies, G., Frank, H. G. & Gallant, D. L. (1956) 20, 32 Cancer (Philad.), 9, 1053 Snell, E. E. (1950) In: Vitamin methods, New York, Toennies, G. & Phillips, P. M. (1959) J. biol. Chem., 234, Academic Press, vol. 1, p. 395 2369 Spray, G. H., Fourman, P. & Witts, L. J. (1951) Brit. Toennies, G., Usdin, E. & Phillips, P. M. (1956) J. biol. med. J., 2, 202 Chem., 221, 855 Spray, G. H. & Witts, L. J. (1959) Lancet, 2, 702 Toepfer, E. W., Zook, E. G., Orr, M. L. & Richardson, Steinkamp, R. C., Shukers, C. F., Totter, J. R. & Day, L. R. (1951) Folic acid content of foods. Microbio- P. L. (1946) Proc. Soc. exp. Biol. (N. Y.), 63, 556 logical assay by standardized methods and compilation Stevens, K., Metz, J., Brandt, V. & Broekhuisen, L. van of data from the literature (US Department of Agri- (1962) E. Afr. med. J., 39, 221 culture Handbook No. 29) Stokstad, E. L. R. (1954) In: The , New York, Usdin, E. (1959) J. biol. Chem., 234, 2373 Academic Press, vol. 3, p. 89 Usdin, E., Phillips, P. M. & Toennies, G. (1956) J. biol. Swendseid, M. E., Bird, 0. O., Brown, R. A. & Bethell, Chem., 221, 865 F. H. (1947) J. Lab. clin. Med., 32, 23 Waters, A. H. & Mollin, D. L. (1961) J. clin. Path., 14, 335 Swendseid, M. E., Swanson, A. L., Meyers, M. C. & Wills, L. (1931) Brit. med. J., 1, 1059 Bethell, F. H. (1952) Blood, 7, 307 Wills, L. & Stewart, A. (1935) Brit. J. exp. Path., 16, 444 Tabor, H. & Rabinowitz, J. C. (1956) J. Amer. chem. Soc., Wintrobe, M. M., Samter, M. & Lisco, H. (1939) Bull. 78, 5705 Johns Hopk. Hosp., 64, 399 Tabor, H., Silverman, M., Mehler, A. H., Daft, F. S. & Wolff, R., Drouet, P. L. & Karlin, R. (1949) Bull. Soc. Bauer, H. (1953) J. Amer. chem. Soc., 75, 756 Chim. biol. (Paris), 31, 265 Tabor, H. & Wyngarden, L. (1958) J. clin. Invest., 37, Zalusky, R. & Herbert, V. (1961) New Engl. J. Med., 824 265, 1033 Tasker, P. W. G. (1955) Lancet, 2, 61 Zalusky, R. & Herbert, V. (1962) Lancet, 1, 108