BrifishJournal ofHaematology, 1976,33, 415.
In Vivo Plasma and Urine Folate Binding after Ingestion of 'H-Folk Acid and '"C-Methyl-Folate F. P. RETIEF, A. DU P. HEYNS,MARIETHA OOSTHUIZEN AND 0. R. VAN REENEN
Faculty of Medicine, University of Qrunge Free State, Bloemfontein, Republic of South Africa (Received 27 August 1975; acceptedfor publication 25 September 1975) After simultaneous ingestion of equivalent amounts of ['H]folic acid ('H-PteGlu) and [' "C]N~-niet1iy1-tetrahydrofo1icacid ( I "C-CH,-H,PteGlu) we were able to demonstrate progressive macromolecular binding of radiofolate in plasma, which appeared to be near maximal at 6 h. Bound radiofolate was predominantly of' "CCH,H,PteGlu origin, and only at 24 h could 3H incorporation be demonstrated. The binder eluted with albumin from Sephadex DEAE-A5o columns. In urine a smaller bound radiofolate fraction, with approximately equal amounts of 'H and "C, appeared after 5.5 h. Plasma chromatography showed radio-PteGlu (peak I) to be rapidly converted to CH,-H,PteGlu (peak 2), with subsequent appearance of two further radiofolate peaks (peaks 3 and 4) the nature of which is as yet unclear. Urine showed similarly placed fractions but their magnitude differed, and urinary peak 3 in particular was much more prominent than its plasma counterpart.
'
Most available studies of plasma folate binding have concentrated on unsaturated binders, measured in vitro. Elsborg (1972) reported significant plasma binding of folic acid but other authors suggested that such binders are virtually absent from normal plasma (Metz & Herbert, 1967; Retief& Huskisson, 1970; Waxman & Schreiber, 1972; da Costa & Rothenberg, 1974), cccasionally present in uraeniia (Hines et al, 1973), pregnancy and women on oral contraceptives (da Costa & Rothenberg, 1974) and folate-deficiency (Waxman & Schreiber, 1972). Other workers, again, could not confirm significant unsaturated binders in pregnancy (Waxman & Schreiber, 1972; Retief et al, 1976a) or folate-deficiency (Retief et al, 1976a). Gel filtration (Markkanen, 1968) and exhaustive dialysis (Retief & Huskisson, 1969) of normal serum showed that endogenous folate was partially bound to macromolecules, a suggestion previously made by Herbert (1961). In the same year Johns et al(1961) reported that plasma radiofolate, obtained after intravenous injection of tritiated folic acid, was largely non-dialysable, resistant to ultra-filtration, and thus probably protein-bound. Good evidence therefore exists of an endogenous plasma folate binder-a macromolecular substance virtually saturated with folate under normal circumstances.W e previously reported preliminary attempts at labelling the endogenous binder in vivo with ['4C]folic acid (Retief et a!, 1975)~ and in the present study we pursued the project, making oral use of both unphysiological (['Hlfolic acid) and physiological (['4C]methyl-folate) radiofolates. Correspondence: Professor F. P. Retief, Department of Internal Medicine, University of Orange Free State, Bloemfontein, Republic of South Africa.
415
416
F. P. Retiejet a1 MATERIALS AND METHODS
Subjects A 52-year-old white male (J.W.) previously healthy, with normal folate status (Table I), and in his third week of uneventful convalescence after a myocardial infarction, gave informed consent for the following study to be performed on him. 1154pg of folate (20 pg/kg), consisting of 579 p g folic acid (100 pCi ['Hlfolic acid; 'H-PteGlu) and 575 p g NS-methyltetrahydrofolic acid (50 pCi ['4C]methylfolate; 4C-CH3-H,PteGlu) was taken by mouth in IOO ml Auid volume after an overnight fast. The methyl-folate solution was freshly prepared within 10 min of ingestion. Fasting was continued for the next 6 h. 20 ml heparinized venous blood samples in Vacutainer tubes were then taken at 15 min, 30 min, I h, 3 h, 6 h and 24 h, and plasma separated by centrifugation. A pre-test urine sample was taken under aseptic conditions, followed by total urine collections at I h, 1.5 h, 3 h, 5.5 h and 24 h. Volumes were recorded and aliquots stored. Plasma and urine samples were kept in the dark at - 20°C until further analysis. A 26-year-old white male (L.L.) with chronic brucellosis and subclinical folate-depletion (normal blood count and plasma folate, but decreased red cell folate, Table I) was also investigated: 1954 pg total folate (20 pg/kg) was ingested, containing 1379 pg folic acid (roo pCi 'H-PteGlu) and 575 pg CH3-H4PteGlu (50 pCi 14C-CH,-H4PteGlu). Blood samples were taken at 30 min, I h, 2 h, 3 h and 6 h. Urine was not examined. Radioactive F o h e s (Radiochemical Centre, Amersham, U.K.) (i) Folk acid-t-'H (potassium salt), specific activity 13 I pCi/mg, radiochemical purity 98%. (ii) 5[methyl-14C]tetrahydr~folic acid (barium salt), specific activity 87 pCi/mg, radiochemical purity 99%. Weir et a2 (1973) showed that the two diastereoisomers of 5-methylfolate have differential absorption characteristics. Liquid chromatography utilized in this study did not separate these isomers. The salts were dissolved in sodium hydroxide and ethanol as in preparing the standard folic acid curve for L. casei bioassay (Herbert, 1966), dried ascorbic acid (I mg/ml) added to methylfolate, and stored in the dark at - 20°C until used. Solutions were analysed for bioactivity (Herbert, 1966) to ensure that biofolate content corresponded to calculated radiofolate concentration. Purity of 14C-CH3-H4 PteGlu in solution was confirmed by ultraviolet absorption spectroscopy (Chanarin, 1969). Radioactivity was measured by liquid scintillation counting on an Intertechnique SL 30 connected to a Multi-8 computer, as previously described (Retief et ~ 2 1976a). , Quantitative radiofolate calculations were deduced from the known specific activity of oral radiofolate doses. Other Methods Urine and plasma samples were further analysed by means of: (i) Liquid chromatography with Sephadex-DEAE-50 (Retief et al, 1976a): 20x 0.8 cm columns were used and 0.5 ml samples eluted stepwise with 7x 5 ml buffer-NaC1 solutions at pH 7.0. Forty-fivex I ml eluates were collected over approximately 6 h and assessed for total protein (ultra-violet extinction at 280 nm) and radioactivity, with differentiation between 14C and 3H. The
In Vivo Folate Binding
417
addition of ascorbic acid to the eluting fluid did not materially change folate elution patterns in this relatively brief chromatography procedure and was thus not routinely used. Samples applied to the column were, however, treated with ascorbic acid (I mglml). (ii) Haemoglobin-coated charcoal adsorption (HCCA), using 50 mg haemoglobin-coated charcoal pellets/o.j ml test sample in 1.5 ml 0.9% NaCl (Retief et al, 1976a). (iii) Dialysis for 36 h in Visking casing against 0.9% NaCl (Retief & Huskisson, 1970). (iv) L. cusei biofolate determinations (Herbert, 1966) were performed on plasma and urine of J.W. RESULTS
Postabsorption Plasma and Urinary Folate Levels The results of biofolate and radiofolate determinations on plasma and urine samples are summarized in Tables I and 11. I h after J.W. ingested the folate, maximum total plasma folate (biofolate) peak was TABLE I. Plasma folates after radiofolate ingestion, with indication of bound fractions (HCCA) 3H-PteGfu (pg/l.) Samples Biofolate (postabsorption) (pg/l.)
Total Bound
~~
%
14C-CH3-H4PteGlu (pgfl.) Total
Patient L.L.(plasma folate 4.0 pg/l., red cell folate 121 fig/].) 30 min 0.1 6.0 5.2 1.6 Ih 1.1 0.1 9.0 2.8
-
-
6h
Bound
%
Total
Bound
%
0.4 0.4
1.6 1.9 7.6 40.7 64.6 72.6
24.9 24.8
0.6
2.0
0.6
2.0
0.0
0.4
0.4
13.2
0-7 0.8
33.3 38.7 59.6
6.8 3.9 3.5 3.5 3.3
~~
Patient J.W. (plasma folate 5.0 pg/l., red cell folate 168.0&I.) 15 min 20 1.5 0.4 12.4 23.4 30 min 29 3.2 0.2 6.5 21.6 Ih 32 4.9 0.2 3.7 15.4 4.5 8.6 3h IS 2.6 0.1 6h I0 2.6 0.2 2.4 5.5 24 h 8 1.6 1.0 63.2 4.5
2h 3h
Total radiofalate (pgll.)
1.3
0.2
1.3
0.2
15.0 15.0
1.1
0.2
18.0
2.2
2.2 2.2
1.2
3.5 3.6 3.2
1.3
20.3 11.2
8.1 6.1
1.4 3.6 3.8 4.2
0.1
6.0 32.0
46.0 68.0
1.5
0.5
12.8
0.9
25.7 28.6 45.4
1.0
1.5
TABLE 11. Urinary folate after radiofolate ingestion, with indication of bound fractions (HCCA) H-PteGlu (pg/l.) Specimens (postabsorption)
14GCH3-H4PteGlu(pgll.)
Total radiofolate (pg/l.)
Biofolate (pg/l.)
Total Bound
Patient J.W. (urine folate 5.0 pg/l.) 30 min 80 0.4 1.5 h 48 10.4 3h I0 5.1 8 2.4 5.5 h 24 h 6 3.4
0.1
0.4 0.4 0.4 1.3
%
Total
Bound
%
Total
Bound
%
14.1 4.2 7.1 15.6
2.5 30.9 16.4 9.8 5.9
0.1
2.5 3.1
2.9 41.3
0.2
1.0
6.8 3.3
0.7
4.2
5.1
5.0
21.5 12.2
1.1
0.5
0.9
16.2
9.3
2.3
7.3 24.7
55.1
1.0
1.4
F. P. Retiefet a1
418
reached, and at 24 h this had returned to approximately normal (pre-absorption) levels. Divergence between biofolate and radiofolate was greatest at I h, after which the two fractions approximated closely. l4C-CH,H4PteGlu reached peak levels at 15 min (23.4 pg/l.) whereas 3H (originating from ,H-PteGlu) reached a much lower maximum at I h (4.9 pg/l.), in spite of comparable initial oral intake. Conversion of PteGlu to CH,-H4PteGlu occurred during the first half hour (Fig I), so that plasma ,H-activity soon represented both PteGlu and CH,-H,PteGlu. Total urinary folate was maximal in the 3 0 min collection. Radiofolate reached a peak at 1.5 h and wide divergence between total and radiofolate values was evident in the first two samples only, with 4C the predominant isotope.
'
Di@erentiation between Bound and Unbound Folate Using HCCA to differentiate between 'free' and macromolecularly-bound radiofolate, a progressive increase in the bound plasma fraction was evident (Table I). At 6 h 49.9% (J.W.) and 45.4% (L.L) total radiofolate was bound, and at 24 h 68.8% (J.W.). The bound fraction contained predominantly ,C,indicating a radio-CH-,H,PteGlu origin, while the contribution of ,H (radio-PteGlu origin) became significant only at 24 h. In the urine 0fJ.W. a bound total radiofolate fraction was convincingly present (> 10%) in the 5+ h sample only, and then as 24.7% of the total radiofolate (Table 11). 14C and contributed equally to this bound fraction.
'
Chromatography with Sephadex D E A E - A 5 o The results for plasma and urine samples from patient J.W. are represented in Figs I and 2. The normal elution pattern of 4C-CH,-H,PteGlu and ,H-PteGlu is evident from standards in Fig 3 . In plasma (Fig I) from ingested radio-PteGlu, appeared at 15 min predominantly as PteGlu (peak I ) , but by 3 h had moved to the CH,-H,PteGlu position (peak 2). At 6 h a new radiofolate peak, consisting predominantly of H, had appeared more proximally in eluates 7-13 (peak 4, and at 24 h this peak contained only ,H. 14C(from ingested radio CH,-H,PteGlu) migrated in the position of CH,-H,PteGlu, as was expected. The significance of a small 14C-fraction in eluates 18-24, noticed in the 15 min to 3 h samples (also evident at 6 h in L.L., Fig 4) was unclear but interesting in view of a corresponding major ',C-peak in urine (peak 3). In urine (Fig z) radioactivity present at 3 0 min appeared as I4C in eluates 19-26, i.e. in the position ofpeak 3. At 1.5 h it was present in peaks z (CH,-H,PteGlu), 3 and 4, and at 5.5 and 24 h, in peaks 3 and 4 only. appeared at 1.5 h in all four peaks, but gradually disappeared from peaks 1-3. At 24 h it was present only in peak 4. In order to ascertain whether folate breakdown in urine may have been partially responsible for the various radiofolate peaks noticed in Fig 2, tracer amounts of 14C-CH,-H4PteGlu and ,H-PteGlu were incubated for 6 h at 37°C with the patient's pre-test urine. The resulting chromatogram did not significantly differ from that of radiofolate added to urine or plasma without prolonged incubation (Fig 3). In order to further identify the bound radiofolate fraction(s) a 6 h plasma sample from L.L. was dialysed in Visking casing against 0.9% NaCl for 36 h before chromatography.
In Vivo Folate Binding
419
In Fig 4 the pre- and postdialysis plasma chromatograms are compared. Non-dialysable activity, consisting only of 3C-CH,-H,PteGlu, migrated under the albumin peak. A 24 h urine sample of patient J.W. was similarly treated, but postdialysis radioactivity proved insufficient for adequate characterization in the present chromatography system.
*t
Ih
3h
6h
'1-
24 h
I
0
1
0
I
10
20 Eluotes
-.---\-
__* I
30
I
40
FIG I. The results of liquid chromatography (Sephadex DEAE-ASO)of plasma taken at various intervals after ingestion of 3H-PteGlu and 14C-CH3-H4PteGlu, showing radiofolate elution. Standard protein elution with this technique is indicated at the top ofthe graph: (a) gamma-globulin, (b) fibrinogen, (c) beta-globulin with transferrin, (d) alphaglobulins, (e) albumin, and (f) pre-albumin (Retief et d,1976a). Radiofolate fractions referred to in the text are peak I (PteGlu, eluates 33-40), peak z (CH,-H,PteGlu, eluates 26-33), peak 3 (eluates 18-24) and peak 4 (eluates 7-13).
DISCUSSION 4C-CH,-H4PtcGlu present in plasma after being absorbed from the intestine undergoes progressive macromolecular binding, when tested with either Visking dialysis or HCCA (Retief
F. P. Retiefet a1
420
et a!, 1976b). The two subjects investigated in the present study showed that 45.4% (1.5 pg/l.) (L.L.) and 46.9% (3.8 pg/l.) (J.W.) of their total plasma radiofolate was bound at 6 h; in J.W. 68.8% (4.2pgll.) was bound at 24 h. These findings suggest that maximal in vivo niacromolecular plasma binding occurs within 6 h of folate absorption, but that additional binding may occur subsequently. After 6 h the bound fraction was comparable with the dialysis-resistant
30 min
1.5 h
3h
't
5.5 h
2l I
,f'
10
.-'..
20 Eluates
30
24 h
I
40
FIG2. The results of liquid chromatography (SephadexDEAE-A~O)of total urines collected at various intervals after ingestion of 3H-PteGlu and 14C-CH,-H4PteGlu, showing radiofolate elution with peaks 1-4 as in Fig I .
biofolate values of approximately I .5-6 pg/l. previously reported for normal plasma (Retief & Huskisson, 1970; Retief et al, 1976a). Delayed 'H-incorporation at 24 h (Table I) may be
explained on the basis that 'H-PteGlu must first be converted to CH,-H,PteGlu as only the latter folatc can be attached to the binder. Markkanen et al (1972), doing bioassays on eluates from a comparable chromatography system, suggested that three protein carriers of endogenous folate exist in plasma : albumin,
FIG 3 . The chromatogram (Sephadex DEAE-Aso) of pre-test urine to which tracer amounts of 3HPteGlu and 14C-CH4-3H-PteGlu had been added in vitro, is shown before and after 6 h incubation a t 36"C, and compared with chromatograms of the patient's plasma to which the same radiofolates had been added in uitro. (a) Plasma; (b) urine; (c) urine after 6 h incubation.
I50 100
50
Eluates
FIG4. The chromatograms (Sephadex DEAE-Aso) of plasma from patient L.L. taken 6 h after ingestion of 3H-PteGlu and l4C-CH3-H,PteGlu, is shown before and after 36 h dialysis in Visking casing. Protein elution (u.v. extinction at 280 nm) is indicated as a dotted curve. (a) Plasma; (b) plasma after dialysis.
422
F. P. Retiefet a1
an ct ,-macroglobulin and a 8-globulin, probably transferrin. Although Waxman & Schreibcr (1973) also favoured transfcrrin as a folate binder, this could not be confirmed by Jacob & Herbert (1974).On labclling the macromolecular folate binder in vivo wewere able to demonstrate only albumin-associated non-dialysablc (bound) radiofolate at 6 h (Fig 4). O u r previous inability to demonstrate radioactivity in the plasma binder 6 h after oral radio-PteGlu (Retief ct al, 1976a),is in accordance with the absence of 3 H (from 'H-PteGlu) from the nondialysable fraction in Fig 4, and the delayed appearance (at 24 11) o f 3 H in the HCCA resistant fraction in Table I. Radiofolate fractionated into four peaks in plasma (Fig I), and these fractions showed some rcseinblance to those described by Markkanen et al(197.2). Comparable radioactive peaks also appeared in urine. Protein-associated fractions in plasma may, however, represent physiological folate moieties with dissimilar elution propcrties rather than folate bound to different proteins. At present there is no conclusive evidence that the three folate carricrs described by Markkanen et al(1972)could not partly rcpresent protein association rather than protein binding. These workers did not attempt to separate bound from unbound plasma folatc prior to chromatography, although they did previously state that approximately 60% of plasma folate is not protein-bound (Markkanen & Peltola, 1971).The altcration of normal folate elution pattern described in various disease states (Markkanen et al, 1973) including pernicious anaemia (Markkanen ct a!, 1974), may then be due to abnormal folate metabolism and deranged physiological folates in plasma rather than a primary defect of binding proteins, unless it can be shown that alterations specifically occurred in the bound folate fraction. Waxinan (1975) has suggested in a review that the unsaturated serum folic acid binder (FABP) may have a functional role as withholder of folate from tissues, and relates this binder to intracellular and membrane binders, as well as to the well-known original milk binder (Ghitis, 1967). However, the amounts of FABP demonstrated in normal and even folatedeficient sera (Waxman & Schreiber, 1973, Retief et al, 1976a) are so small that the possibility of non-physiological in vitro artefacts must be seriously considered, especially in view of the fact that unphysiological folates are bound in preference to physiological folate (Waxman & Schreibcr, 1973; da Costa & Rothenberg, 1974). It is quite uncertain whether unsaturated and saturated binders are structurally or functionally related. Our findings suggest that plasma folate is attached to a binder during its passage through tissues (possibly the liver), but that this binder is virtually saturated. The binder may be of intracellular origin, but it is interesting to note that red cell biofolate (considered to be a valid indicator of folate stores) is extensively dialysable in Visking casing and thus not significantly attached to macromolecular binders (Retief& Huskisson, 1970). The minor urinary folate binder demonstrated in Table I1 could not be further characterized in this study. Although the present study was primarily designed to investigate in vivo plasma folate binding, interesting additional findings regarding folate metabolism may be commented on. Displacement of storage folate by absorbed folate (Whitehead & Cooper, 1967; Melikian et a!, 1971) was demonstrated by the initial divergence between total biofolate and radiofolate values (Table I). Plasma 3H-PteGlu (peak I) was totally converted to CH,-H,PteGh (peak 2) within 3 h after ingestion of 3H-PteGlu and then at 6 h 3 H appeared in the more proximal fraction, pcak 4 (Fig I). At 6 h peak 4 consisted predominantly of 3 H and at 24 h almost totally of 3H. This unexpected finding may perhaps be explained by postulating that the
In Vivo Folate Binding 423 unidentified folate in peak 4 is gradually derived from CH,-H,PteGlu in peak 2, and that peak 2 incorporates folate from ingested PteGlu (,H) much more slowly than from readily available CH,-H,PteGh ('"C). Once formed peak 4 seems fairly stable. At 24 h radioactivity in this peak was comparable to that in peak 2 (Fig I). Another possible explanation is that PteGlu is less rapidly absorbed than CH3-H,PteGlu, and gradually metabolized in the in peak 4 may represent a nonphysiological metabolite af folic acid. intestine, so that the However, the presence of some I4C in peak 4 (in plasma and urine) would then have to indicate similar intestinal degradation of CH,-H,PteGlu. It is unlikely that peaks 3 and 4 represent in vitro folate change, due to oxidation in plasma or urine, as this would not explain in peak 4. PteGlu is a stable folate. the differential presence of In urine four peaks of radioactivity were evident: PteGlu (peak I), CH,-H,PteGlu (peak 2), a prominent fraction in which radiofolate made its first appearance at 3 0 min (peak 3), and a lesser fraction (peak 4) (Fig 2). A small peak 3 was also present in plasma (Fig I), but this was quite minor compared with the prominent urinary equivalent. McLean & Chanarin (1966) described three folate peaks in urine using DEAE-Sephadex-ASo after i.v. H-PteGlu, and suggested that this represented 10-forniyl-tetrahydrofolate (with p-amino-benzoyl fragments), CH,-H,PteGlu and PteGlu. The presence of lo-forniyl-folate and 5-10-methenyltetrahydrofolate in urine after oral PteGlu were first recognized by Silverman et al(1965) and Albreclit & Broquist (1956). The nature of peaks 3 and 4, demonstrated in the present study, is being further investigated, and may partly represent formyl- and methenyl-folates. Osborne-White & Smith (1973) showed 10-formyl-tetrahydrofolate to elute proximal to CH,-H,PteGlu on DEAE-A~s,in positions which could correspond to the peaks 4 (even 3) and 2 respectively. However, forniyl-folate is not a recognized constituent of plasma (Herbert et al, 1962), and the possibility exists that fractions 3 and 4 may even represent di- or triglutamates of CH,-H,PteGlu, although Osborne-White & Smith (1973) claimed that polyglutamates elute distal to monoglutamates in this system. One may also speculate that peak 3 (and possibly peak 4) are prominent in urine because they represent less essential physiological folates produced by the experimental conditions, and lost in urine because the kidney's folate receptors are primarily geared for retention of CH,-H,PteGlu (peak 2 ) . ACKNOWLEDGMENT
This study was supported by an ad hoc grant from the South African Medical Research Council. REFERENCES ALBRECHT, A.M. & BROQUIST, H.P. (1956) Evidence for occurrence of Io-formyletrahydrofolic acid in human urine. Proceedings of the Society for Experimental Biology and Medicine, 92, 158. CHANARIN, I. (1969) The Megaloblastic Anaemias, p 23 3 . Blackwell Scientific Publications, Oxford. DACOSTA, M. & ROTHENBERG, S.P. (1974)Appearance of a folate binder in leukocytes and serum of women who are pregnant or taking oral contraceptives. Journal of Laboratory and Clinical Medicine, 83, 207.
ELSBORG, L. (1972) Binding of folk acid to human plasma proteins. Acta Haematologica, 48, 207. GHITIS, J. (1967) The folate binding in milk. American Journal of Clinical Nutrition, 20, I. HERBERT, V. (1961) The assay and nature of folic acid activity in human serum. Journal of Clinical Investigation, 40. 81. HERBERT, V. (1966) Aseptic addition method for Lactobacillus casei assay of folate activity in human serum.Journal of Clinical Pathology, 19. 12.
424
F. P. Retiefet a1
HERBERT, V., LARRABEE, A.R. & BUCHANAN, J.M. (1962) Studies on the identification of a folate compound of human serum. journal oj’ Clinical Investigation, 41, 1134. HINES, J.D., KAMEN, B. & CASTON, D. (1973) Abnormal folate binding protein(s) in azotemic patients. (Abstract). Blood, 42, 997. JACOB,E. 81 HERBERT, V. (1974) Evidence against transferrin as a binder of either vitamin Blz or folic acid. Blood, 43, 767. JOHNS, D.G.,SPERTI, S. & BURGEN, A.S.V. (1961) The metabolism of tritiated folic acid in man. Jortrnal of Clinical Investigation, 40, 1684. MCLEAN, A. & CHANARIN, I. (1966) Urinary excretion of 5-methyl-tetrahydrofolate in man. Blood, 27, 386. MARKKANEN, T. (1968) Pteroylglutamic acid (PGA) activity of serum in gel filtration. Life Sciences, 7. 11, 887. MARKKANEN, T. & PELTOLA, 0.(1971) Carrier proteins of folic acid activity in human serum. Acta ffaematologica, 45, 106. MARKKANEN, T., VIRTANEN, S., HIMANEN, P.& PAJULA, R.-L. (1972) Transferrin, the third carrier protein of folic acid activity in human serum. Acta HaematolOgb.3, 48, 213. MARKANNEN, T., PAJULA,R.-L., HIMANEN, P. & VIRTANEN, S. (1973) Serum folk and activity (L. casei) in Sephadex gel chromatography. Journal qf Clinical Pathology, 26, 486. MARKKANEN, T., HIMANEN, P. & PAJULA, R.-L. (1974) Binding of folic acid to serum proteins. 111. The effect of pernicious anaemia. Acta Haematologica, 51, 193. MELIKIAN, V., PATON, A., LEEMING, R.J. & PORTMANGRAHAM, H. (1971) Site of reduction and methylation of folk acid in man. Lancet, ii, 955. METZ,J. & HERBERT, V. (1967) Folate binders in milk and human serum; their use in coated charcoal assay of folic acid; their possible physiologic role. (Abstract). Journal of Clinical Investigation, 46, 1 ~ 9 6 .
OSBORNE-WHITE, W.S. & SMITH,R.M. (1973) Identification and measurement of the folates in sheep liver. Biochemical Journal 136. 265. RETIBF,F.P. & HUSKISSON, Y.J. (1969) Serum and urinary folate in liver disease. British Medical Journal, ii, 150. RETIEF, F.P. & HUSKISSON, Y.J. (1970) Folate binders in body fluids.Journal ofClinica1 Pathology, 23, 703. RETEIEF,F.P., HEYNS,A. DU P., OOSTHUIZEN, M. & BADENHORST, C.J. (1975) Plasma folate binding. (Abstract). South African Medical Journal, 49, 1060. A. DU P., OOSTHUIZEN, M., VAN RETIEF, F.P., HEYNS, REENEN, O.R. & BADENHORST, C.J. (1976a) In vitro binding of folates by body fluids. British Journal of’ Haematology, 32, I I 3. RETIBF, F.P., HEWS,A. DU P., OOSTHUIZEN, M. & VAN REENEN, O.R. (1976b) Aspects of radiofolate absorption, metabolism and plasma binding. South African Medical Journal, 50, 212. SILVERMAN, M., EBAUGH, F.G., JR, & GARDINER, R.C. (1956) The nature of labile citrovorum factor in human urine. Journal of Biological Chemistry, 223, 259. WAXMAN, S. (1975) Folate binding proteins. British Journal ofHaematology, a9, 23. WAXMAN, S. & SCHREIBER, C. (1972) Further studies of serum folk acid binding protein (FABP). (Abstract). American Journal of Clinical Nutrition, 25, 450. WAXMAN, S. & SCHREIBER, C. (1973) Characteristics of folk acid-binding protein in folate-deficient serum. Blood, 42, 291. WEIR,D.G., BROWN, J.P., FREEDMAN, D.S. & SCOTT, J.M. (1973) The absorption of the diastereoisomers of 5-methyltetrahydropteroylglutamate in man: a carrier-mediated process. Clinical Science and Molecular Medicine, 45, 625. WHITEHEAD, V.M. & COOPER, B.A. (1967) Absorption of unaltered folic acid from the gastro-intestinal tract in man. British Journal offfaematology, 13, 679.
In Vivo Plasma and Urine Folate Binding after Ingestion of 'H-Folk Acid and '"C-Methyl-Folate F. P. RETIEF, A. DU P. HEYNS,MARIETHA OOSTHUIZEN AND 0. R. VAN REENEN
Faculty of Medicine, University of Qrunge Free State, Bloemfontein, Republic of South Africa (Received 27 August 1975; acceptedfor publication 25 September 1975) After simultaneous ingestion of equivalent amounts of ['H]folic acid ('H-PteGlu) and [' "C]N~-niet1iy1-tetrahydrofo1icacid ( I "C-CH,-H,PteGlu) we were able to demonstrate progressive macromolecular binding of radiofolate in plasma, which appeared to be near maximal at 6 h. Bound radiofolate was predominantly of' "CCH,H,PteGlu origin, and only at 24 h could 3H incorporation be demonstrated. The binder eluted with albumin from Sephadex DEAE-A5o columns. In urine a smaller bound radiofolate fraction, with approximately equal amounts of 'H and "C, appeared after 5.5 h. Plasma chromatography showed radio-PteGlu (peak I) to be rapidly converted to CH,-H,PteGlu (peak 2), with subsequent appearance of two further radiofolate peaks (peaks 3 and 4) the nature of which is as yet unclear. Urine showed similarly placed fractions but their magnitude differed, and urinary peak 3 in particular was much more prominent than its plasma counterpart.
'
Most available studies of plasma folate binding have concentrated on unsaturated binders, measured in vitro. Elsborg (1972) reported significant plasma binding of folic acid but other authors suggested that such binders are virtually absent from normal plasma (Metz & Herbert, 1967; Retief& Huskisson, 1970; Waxman & Schreiber, 1972; da Costa & Rothenberg, 1974), cccasionally present in uraeniia (Hines et al, 1973), pregnancy and women on oral contraceptives (da Costa & Rothenberg, 1974) and folate-deficiency (Waxman & Schreiber, 1972). Other workers, again, could not confirm significant unsaturated binders in pregnancy (Waxman & Schreiber, 1972; Retief et al, 1976a) or folate-deficiency (Retief et al, 1976a). Gel filtration (Markkanen, 1968) and exhaustive dialysis (Retief & Huskisson, 1969) of normal serum showed that endogenous folate was partially bound to macromolecules, a suggestion previously made by Herbert (1961). In the same year Johns et al(1961) reported that plasma radiofolate, obtained after intravenous injection of tritiated folic acid, was largely non-dialysable, resistant to ultra-filtration, and thus probably protein-bound. Good evidence therefore exists of an endogenous plasma folate binder-a macromolecular substance virtually saturated with folate under normal circumstances.W e previously reported preliminary attempts at labelling the endogenous binder in vivo with ['4C]folic acid (Retief et a!, 1975)~ and in the present study we pursued the project, making oral use of both unphysiological (['Hlfolic acid) and physiological (['4C]methyl-folate) radiofolates. Correspondence: Professor F. P. Retief, Department of Internal Medicine, University of Orange Free State, Bloemfontein, Republic of South Africa.
415
416
F. P. Retiejet a1 MATERIALS AND METHODS
Subjects A 52-year-old white male (J.W.) previously healthy, with normal folate status (Table I), and in his third week of uneventful convalescence after a myocardial infarction, gave informed consent for the following study to be performed on him. 1154pg of folate (20 pg/kg), consisting of 579 p g folic acid (100 pCi ['Hlfolic acid; 'H-PteGlu) and 575 p g NS-methyltetrahydrofolic acid (50 pCi ['4C]methylfolate; 4C-CH3-H,PteGlu) was taken by mouth in IOO ml Auid volume after an overnight fast. The methyl-folate solution was freshly prepared within 10 min of ingestion. Fasting was continued for the next 6 h. 20 ml heparinized venous blood samples in Vacutainer tubes were then taken at 15 min, 30 min, I h, 3 h, 6 h and 24 h, and plasma separated by centrifugation. A pre-test urine sample was taken under aseptic conditions, followed by total urine collections at I h, 1.5 h, 3 h, 5.5 h and 24 h. Volumes were recorded and aliquots stored. Plasma and urine samples were kept in the dark at - 20°C until further analysis. A 26-year-old white male (L.L.) with chronic brucellosis and subclinical folate-depletion (normal blood count and plasma folate, but decreased red cell folate, Table I) was also investigated: 1954 pg total folate (20 pg/kg) was ingested, containing 1379 pg folic acid (roo pCi 'H-PteGlu) and 575 pg CH3-H4PteGlu (50 pCi 14C-CH,-H4PteGlu). Blood samples were taken at 30 min, I h, 2 h, 3 h and 6 h. Urine was not examined. Radioactive F o h e s (Radiochemical Centre, Amersham, U.K.) (i) Folk acid-t-'H (potassium salt), specific activity 13 I pCi/mg, radiochemical purity 98%. (ii) 5[methyl-14C]tetrahydr~folic acid (barium salt), specific activity 87 pCi/mg, radiochemical purity 99%. Weir et a2 (1973) showed that the two diastereoisomers of 5-methylfolate have differential absorption characteristics. Liquid chromatography utilized in this study did not separate these isomers. The salts were dissolved in sodium hydroxide and ethanol as in preparing the standard folic acid curve for L. casei bioassay (Herbert, 1966), dried ascorbic acid (I mg/ml) added to methylfolate, and stored in the dark at - 20°C until used. Solutions were analysed for bioactivity (Herbert, 1966) to ensure that biofolate content corresponded to calculated radiofolate concentration. Purity of 14C-CH3-H4 PteGlu in solution was confirmed by ultraviolet absorption spectroscopy (Chanarin, 1969). Radioactivity was measured by liquid scintillation counting on an Intertechnique SL 30 connected to a Multi-8 computer, as previously described (Retief et ~ 2 1976a). , Quantitative radiofolate calculations were deduced from the known specific activity of oral radiofolate doses. Other Methods Urine and plasma samples were further analysed by means of: (i) Liquid chromatography with Sephadex-DEAE-50 (Retief et al, 1976a): 20x 0.8 cm columns were used and 0.5 ml samples eluted stepwise with 7x 5 ml buffer-NaC1 solutions at pH 7.0. Forty-fivex I ml eluates were collected over approximately 6 h and assessed for total protein (ultra-violet extinction at 280 nm) and radioactivity, with differentiation between 14C and 3H. The
In Vivo Folate Binding
417
addition of ascorbic acid to the eluting fluid did not materially change folate elution patterns in this relatively brief chromatography procedure and was thus not routinely used. Samples applied to the column were, however, treated with ascorbic acid (I mglml). (ii) Haemoglobin-coated charcoal adsorption (HCCA), using 50 mg haemoglobin-coated charcoal pellets/o.j ml test sample in 1.5 ml 0.9% NaCl (Retief et al, 1976a). (iii) Dialysis for 36 h in Visking casing against 0.9% NaCl (Retief & Huskisson, 1970). (iv) L. cusei biofolate determinations (Herbert, 1966) were performed on plasma and urine of J.W. RESULTS
Postabsorption Plasma and Urinary Folate Levels The results of biofolate and radiofolate determinations on plasma and urine samples are summarized in Tables I and 11. I h after J.W. ingested the folate, maximum total plasma folate (biofolate) peak was TABLE I. Plasma folates after radiofolate ingestion, with indication of bound fractions (HCCA) 3H-PteGfu (pg/l.) Samples Biofolate (postabsorption) (pg/l.)
Total Bound
~~
%
14C-CH3-H4PteGlu (pgfl.) Total
Patient L.L.(plasma folate 4.0 pg/l., red cell folate 121 fig/].) 30 min 0.1 6.0 5.2 1.6 Ih 1.1 0.1 9.0 2.8
-
-
6h
Bound
%
Total
Bound
%
0.4 0.4
1.6 1.9 7.6 40.7 64.6 72.6
24.9 24.8
0.6
2.0
0.6
2.0
0.0
0.4
0.4
13.2
0-7 0.8
33.3 38.7 59.6
6.8 3.9 3.5 3.5 3.3
~~
Patient J.W. (plasma folate 5.0 pg/l., red cell folate 168.0&I.) 15 min 20 1.5 0.4 12.4 23.4 30 min 29 3.2 0.2 6.5 21.6 Ih 32 4.9 0.2 3.7 15.4 4.5 8.6 3h IS 2.6 0.1 6h I0 2.6 0.2 2.4 5.5 24 h 8 1.6 1.0 63.2 4.5
2h 3h
Total radiofalate (pgll.)
1.3
0.2
1.3
0.2
15.0 15.0
1.1
0.2
18.0
2.2
2.2 2.2
1.2
3.5 3.6 3.2
1.3
20.3 11.2
8.1 6.1
1.4 3.6 3.8 4.2
0.1
6.0 32.0
46.0 68.0
1.5
0.5
12.8
0.9
25.7 28.6 45.4
1.0
1.5
TABLE 11. Urinary folate after radiofolate ingestion, with indication of bound fractions (HCCA) H-PteGlu (pg/l.) Specimens (postabsorption)
14GCH3-H4PteGlu(pgll.)
Total radiofolate (pg/l.)
Biofolate (pg/l.)
Total Bound
Patient J.W. (urine folate 5.0 pg/l.) 30 min 80 0.4 1.5 h 48 10.4 3h I0 5.1 8 2.4 5.5 h 24 h 6 3.4
0.1
0.4 0.4 0.4 1.3
%
Total
Bound
%
Total
Bound
%
14.1 4.2 7.1 15.6
2.5 30.9 16.4 9.8 5.9
0.1
2.5 3.1
2.9 41.3
0.2
1.0
6.8 3.3
0.7
4.2
5.1
5.0
21.5 12.2
1.1
0.5
0.9
16.2
9.3
2.3
7.3 24.7
55.1
1.0
1.4
F. P. Retiefet a1
418
reached, and at 24 h this had returned to approximately normal (pre-absorption) levels. Divergence between biofolate and radiofolate was greatest at I h, after which the two fractions approximated closely. l4C-CH,H4PteGlu reached peak levels at 15 min (23.4 pg/l.) whereas 3H (originating from ,H-PteGlu) reached a much lower maximum at I h (4.9 pg/l.), in spite of comparable initial oral intake. Conversion of PteGlu to CH,-H4PteGlu occurred during the first half hour (Fig I), so that plasma ,H-activity soon represented both PteGlu and CH,-H,PteGlu. Total urinary folate was maximal in the 3 0 min collection. Radiofolate reached a peak at 1.5 h and wide divergence between total and radiofolate values was evident in the first two samples only, with 4C the predominant isotope.
'
Di@erentiation between Bound and Unbound Folate Using HCCA to differentiate between 'free' and macromolecularly-bound radiofolate, a progressive increase in the bound plasma fraction was evident (Table I). At 6 h 49.9% (J.W.) and 45.4% (L.L) total radiofolate was bound, and at 24 h 68.8% (J.W.). The bound fraction contained predominantly ,C,indicating a radio-CH-,H,PteGlu origin, while the contribution of ,H (radio-PteGlu origin) became significant only at 24 h. In the urine 0fJ.W. a bound total radiofolate fraction was convincingly present (> 10%) in the 5+ h sample only, and then as 24.7% of the total radiofolate (Table 11). 14C and contributed equally to this bound fraction.
'
Chromatography with Sephadex D E A E - A 5 o The results for plasma and urine samples from patient J.W. are represented in Figs I and 2. The normal elution pattern of 4C-CH,-H,PteGlu and ,H-PteGlu is evident from standards in Fig 3 . In plasma (Fig I) from ingested radio-PteGlu, appeared at 15 min predominantly as PteGlu (peak I ) , but by 3 h had moved to the CH,-H,PteGlu position (peak 2). At 6 h a new radiofolate peak, consisting predominantly of H, had appeared more proximally in eluates 7-13 (peak 4, and at 24 h this peak contained only ,H. 14C(from ingested radio CH,-H,PteGlu) migrated in the position of CH,-H,PteGlu, as was expected. The significance of a small 14C-fraction in eluates 18-24, noticed in the 15 min to 3 h samples (also evident at 6 h in L.L., Fig 4) was unclear but interesting in view of a corresponding major ',C-peak in urine (peak 3). In urine (Fig z) radioactivity present at 3 0 min appeared as I4C in eluates 19-26, i.e. in the position ofpeak 3. At 1.5 h it was present in peaks z (CH,-H,PteGlu), 3 and 4, and at 5.5 and 24 h, in peaks 3 and 4 only. appeared at 1.5 h in all four peaks, but gradually disappeared from peaks 1-3. At 24 h it was present only in peak 4. In order to ascertain whether folate breakdown in urine may have been partially responsible for the various radiofolate peaks noticed in Fig 2, tracer amounts of 14C-CH,-H4PteGlu and ,H-PteGlu were incubated for 6 h at 37°C with the patient's pre-test urine. The resulting chromatogram did not significantly differ from that of radiofolate added to urine or plasma without prolonged incubation (Fig 3). In order to further identify the bound radiofolate fraction(s) a 6 h plasma sample from L.L. was dialysed in Visking casing against 0.9% NaCl for 36 h before chromatography.
In Vivo Folate Binding
419
In Fig 4 the pre- and postdialysis plasma chromatograms are compared. Non-dialysable activity, consisting only of 3C-CH,-H,PteGlu, migrated under the albumin peak. A 24 h urine sample of patient J.W. was similarly treated, but postdialysis radioactivity proved insufficient for adequate characterization in the present chromatography system.
*t
Ih
3h
6h
'1-
24 h
I
0
1
0
I
10
20 Eluotes
-.---\-
__* I
30
I
40
FIG I. The results of liquid chromatography (Sephadex DEAE-ASO)of plasma taken at various intervals after ingestion of 3H-PteGlu and 14C-CH3-H4PteGlu, showing radiofolate elution. Standard protein elution with this technique is indicated at the top ofthe graph: (a) gamma-globulin, (b) fibrinogen, (c) beta-globulin with transferrin, (d) alphaglobulins, (e) albumin, and (f) pre-albumin (Retief et d,1976a). Radiofolate fractions referred to in the text are peak I (PteGlu, eluates 33-40), peak z (CH,-H,PteGlu, eluates 26-33), peak 3 (eluates 18-24) and peak 4 (eluates 7-13).
DISCUSSION 4C-CH,-H4PtcGlu present in plasma after being absorbed from the intestine undergoes progressive macromolecular binding, when tested with either Visking dialysis or HCCA (Retief
F. P. Retiefet a1
420
et a!, 1976b). The two subjects investigated in the present study showed that 45.4% (1.5 pg/l.) (L.L.) and 46.9% (3.8 pg/l.) (J.W.) of their total plasma radiofolate was bound at 6 h; in J.W. 68.8% (4.2pgll.) was bound at 24 h. These findings suggest that maximal in vivo niacromolecular plasma binding occurs within 6 h of folate absorption, but that additional binding may occur subsequently. After 6 h the bound fraction was comparable with the dialysis-resistant
30 min
1.5 h
3h
't
5.5 h
2l I
,f'
10
.-'..
20 Eluates
30
24 h
I
40
FIG2. The results of liquid chromatography (SephadexDEAE-A~O)of total urines collected at various intervals after ingestion of 3H-PteGlu and 14C-CH,-H4PteGlu, showing radiofolate elution with peaks 1-4 as in Fig I .
biofolate values of approximately I .5-6 pg/l. previously reported for normal plasma (Retief & Huskisson, 1970; Retief et al, 1976a). Delayed 'H-incorporation at 24 h (Table I) may be
explained on the basis that 'H-PteGlu must first be converted to CH,-H,PteGlu as only the latter folatc can be attached to the binder. Markkanen et al (1972), doing bioassays on eluates from a comparable chromatography system, suggested that three protein carriers of endogenous folate exist in plasma : albumin,
FIG 3 . The chromatogram (Sephadex DEAE-Aso) of pre-test urine to which tracer amounts of 3HPteGlu and 14C-CH4-3H-PteGlu had been added in vitro, is shown before and after 6 h incubation a t 36"C, and compared with chromatograms of the patient's plasma to which the same radiofolates had been added in uitro. (a) Plasma; (b) urine; (c) urine after 6 h incubation.
I50 100
50
Eluates
FIG4. The chromatograms (Sephadex DEAE-Aso) of plasma from patient L.L. taken 6 h after ingestion of 3H-PteGlu and l4C-CH3-H,PteGlu, is shown before and after 36 h dialysis in Visking casing. Protein elution (u.v. extinction at 280 nm) is indicated as a dotted curve. (a) Plasma; (b) plasma after dialysis.
422
F. P. Retiefet a1
an ct ,-macroglobulin and a 8-globulin, probably transferrin. Although Waxman & Schreibcr (1973) also favoured transfcrrin as a folate binder, this could not be confirmed by Jacob & Herbert (1974).On labclling the macromolecular folate binder in vivo wewere able to demonstrate only albumin-associated non-dialysablc (bound) radiofolate at 6 h (Fig 4). O u r previous inability to demonstrate radioactivity in the plasma binder 6 h after oral radio-PteGlu (Retief ct al, 1976a),is in accordance with the absence of 3 H (from 'H-PteGlu) from the nondialysable fraction in Fig 4, and the delayed appearance (at 24 11) o f 3 H in the HCCA resistant fraction in Table I. Radiofolate fractionated into four peaks in plasma (Fig I), and these fractions showed some rcseinblance to those described by Markkanen et al(197.2). Comparable radioactive peaks also appeared in urine. Protein-associated fractions in plasma may, however, represent physiological folate moieties with dissimilar elution propcrties rather than folate bound to different proteins. At present there is no conclusive evidence that the three folate carricrs described by Markkanen et al(1972)could not partly rcpresent protein association rather than protein binding. These workers did not attempt to separate bound from unbound plasma folatc prior to chromatography, although they did previously state that approximately 60% of plasma folate is not protein-bound (Markkanen & Peltola, 1971).The altcration of normal folate elution pattern described in various disease states (Markkanen et al, 1973) including pernicious anaemia (Markkanen ct a!, 1974), may then be due to abnormal folate metabolism and deranged physiological folates in plasma rather than a primary defect of binding proteins, unless it can be shown that alterations specifically occurred in the bound folate fraction. Waxinan (1975) has suggested in a review that the unsaturated serum folic acid binder (FABP) may have a functional role as withholder of folate from tissues, and relates this binder to intracellular and membrane binders, as well as to the well-known original milk binder (Ghitis, 1967). However, the amounts of FABP demonstrated in normal and even folatedeficient sera (Waxman & Schreiber, 1973, Retief et al, 1976a) are so small that the possibility of non-physiological in vitro artefacts must be seriously considered, especially in view of the fact that unphysiological folates are bound in preference to physiological folate (Waxman & Schreibcr, 1973; da Costa & Rothenberg, 1974). It is quite uncertain whether unsaturated and saturated binders are structurally or functionally related. Our findings suggest that plasma folate is attached to a binder during its passage through tissues (possibly the liver), but that this binder is virtually saturated. The binder may be of intracellular origin, but it is interesting to note that red cell biofolate (considered to be a valid indicator of folate stores) is extensively dialysable in Visking casing and thus not significantly attached to macromolecular binders (Retief& Huskisson, 1970). The minor urinary folate binder demonstrated in Table I1 could not be further characterized in this study. Although the present study was primarily designed to investigate in vivo plasma folate binding, interesting additional findings regarding folate metabolism may be commented on. Displacement of storage folate by absorbed folate (Whitehead & Cooper, 1967; Melikian et a!, 1971) was demonstrated by the initial divergence between total biofolate and radiofolate values (Table I). Plasma 3H-PteGlu (peak I) was totally converted to CH,-H,PteGh (peak 2) within 3 h after ingestion of 3H-PteGlu and then at 6 h 3 H appeared in the more proximal fraction, pcak 4 (Fig I). At 6 h peak 4 consisted predominantly of 3 H and at 24 h almost totally of 3H. This unexpected finding may perhaps be explained by postulating that the
In Vivo Folate Binding 423 unidentified folate in peak 4 is gradually derived from CH,-H,PteGlu in peak 2, and that peak 2 incorporates folate from ingested PteGlu (,H) much more slowly than from readily available CH,-H,PteGh ('"C). Once formed peak 4 seems fairly stable. At 24 h radioactivity in this peak was comparable to that in peak 2 (Fig I). Another possible explanation is that PteGlu is less rapidly absorbed than CH3-H,PteGlu, and gradually metabolized in the in peak 4 may represent a nonphysiological metabolite af folic acid. intestine, so that the However, the presence of some I4C in peak 4 (in plasma and urine) would then have to indicate similar intestinal degradation of CH,-H,PteGlu. It is unlikely that peaks 3 and 4 represent in vitro folate change, due to oxidation in plasma or urine, as this would not explain in peak 4. PteGlu is a stable folate. the differential presence of In urine four peaks of radioactivity were evident: PteGlu (peak I), CH,-H,PteGlu (peak 2), a prominent fraction in which radiofolate made its first appearance at 3 0 min (peak 3), and a lesser fraction (peak 4) (Fig 2). A small peak 3 was also present in plasma (Fig I), but this was quite minor compared with the prominent urinary equivalent. McLean & Chanarin (1966) described three folate peaks in urine using DEAE-Sephadex-ASo after i.v. H-PteGlu, and suggested that this represented 10-forniyl-tetrahydrofolate (with p-amino-benzoyl fragments), CH,-H,PteGlu and PteGlu. The presence of lo-forniyl-folate and 5-10-methenyltetrahydrofolate in urine after oral PteGlu were first recognized by Silverman et al(1965) and Albreclit & Broquist (1956). The nature of peaks 3 and 4, demonstrated in the present study, is being further investigated, and may partly represent formyl- and methenyl-folates. Osborne-White & Smith (1973) showed 10-formyl-tetrahydrofolate to elute proximal to CH,-H,PteGlu on DEAE-A~s,in positions which could correspond to the peaks 4 (even 3) and 2 respectively. However, forniyl-folate is not a recognized constituent of plasma (Herbert et al, 1962), and the possibility exists that fractions 3 and 4 may even represent di- or triglutamates of CH,-H,PteGlu, although Osborne-White & Smith (1973) claimed that polyglutamates elute distal to monoglutamates in this system. One may also speculate that peak 3 (and possibly peak 4) are prominent in urine because they represent less essential physiological folates produced by the experimental conditions, and lost in urine because the kidney's folate receptors are primarily geared for retention of CH,-H,PteGlu (peak 2 ) . ACKNOWLEDGMENT
This study was supported by an ad hoc grant from the South African Medical Research Council. REFERENCES ALBRECHT, A.M. & BROQUIST, H.P. (1956) Evidence for occurrence of Io-formyletrahydrofolic acid in human urine. Proceedings of the Society for Experimental Biology and Medicine, 92, 158. CHANARIN, I. (1969) The Megaloblastic Anaemias, p 23 3 . Blackwell Scientific Publications, Oxford. DACOSTA, M. & ROTHENBERG, S.P. (1974)Appearance of a folate binder in leukocytes and serum of women who are pregnant or taking oral contraceptives. Journal of Laboratory and Clinical Medicine, 83, 207.
ELSBORG, L. (1972) Binding of folk acid to human plasma proteins. Acta Haematologica, 48, 207. GHITIS, J. (1967) The folate binding in milk. American Journal of Clinical Nutrition, 20, I. HERBERT, V. (1961) The assay and nature of folic acid activity in human serum. Journal of Clinical Investigation, 40. 81. HERBERT, V. (1966) Aseptic addition method for Lactobacillus casei assay of folate activity in human serum.Journal of Clinical Pathology, 19. 12.
424
F. P. Retiefet a1
HERBERT, V., LARRABEE, A.R. & BUCHANAN, J.M. (1962) Studies on the identification of a folate compound of human serum. journal oj’ Clinical Investigation, 41, 1134. HINES, J.D., KAMEN, B. & CASTON, D. (1973) Abnormal folate binding protein(s) in azotemic patients. (Abstract). Blood, 42, 997. JACOB,E. 81 HERBERT, V. (1974) Evidence against transferrin as a binder of either vitamin Blz or folic acid. Blood, 43, 767. JOHNS, D.G.,SPERTI, S. & BURGEN, A.S.V. (1961) The metabolism of tritiated folic acid in man. Jortrnal of Clinical Investigation, 40, 1684. MCLEAN, A. & CHANARIN, I. (1966) Urinary excretion of 5-methyl-tetrahydrofolate in man. Blood, 27, 386. MARKKANEN, T. (1968) Pteroylglutamic acid (PGA) activity of serum in gel filtration. Life Sciences, 7. 11, 887. MARKKANEN, T. & PELTOLA, 0.(1971) Carrier proteins of folic acid activity in human serum. Acta ffaematologica, 45, 106. MARKKANEN, T., VIRTANEN, S., HIMANEN, P.& PAJULA, R.-L. (1972) Transferrin, the third carrier protein of folic acid activity in human serum. Acta HaematolOgb.3, 48, 213. MARKANNEN, T., PAJULA,R.-L., HIMANEN, P. & VIRTANEN, S. (1973) Serum folk and activity (L. casei) in Sephadex gel chromatography. Journal qf Clinical Pathology, 26, 486. MARKKANEN, T., HIMANEN, P. & PAJULA, R.-L. (1974) Binding of folic acid to serum proteins. 111. The effect of pernicious anaemia. Acta Haematologica, 51, 193. MELIKIAN, V., PATON, A., LEEMING, R.J. & PORTMANGRAHAM, H. (1971) Site of reduction and methylation of folk acid in man. Lancet, ii, 955. METZ,J. & HERBERT, V. (1967) Folate binders in milk and human serum; their use in coated charcoal assay of folic acid; their possible physiologic role. (Abstract). Journal of Clinical Investigation, 46, 1 ~ 9 6 .
OSBORNE-WHITE, W.S. & SMITH,R.M. (1973) Identification and measurement of the folates in sheep liver. Biochemical Journal 136. 265. RETIBF,F.P. & HUSKISSON, Y.J. (1969) Serum and urinary folate in liver disease. British Medical Journal, ii, 150. RETIEF, F.P. & HUSKISSON, Y.J. (1970) Folate binders in body fluids.Journal ofClinica1 Pathology, 23, 703. RETEIEF,F.P., HEYNS,A. DU P., OOSTHUIZEN, M. & BADENHORST, C.J. (1975) Plasma folate binding. (Abstract). South African Medical Journal, 49, 1060. A. DU P., OOSTHUIZEN, M., VAN RETIEF, F.P., HEYNS, REENEN, O.R. & BADENHORST, C.J. (1976a) In vitro binding of folates by body fluids. British Journal of’ Haematology, 32, I I 3. RETIBF, F.P., HEWS,A. DU P., OOSTHUIZEN, M. & VAN REENEN, O.R. (1976b) Aspects of radiofolate absorption, metabolism and plasma binding. South African Medical Journal, 50, 212. SILVERMAN, M., EBAUGH, F.G., JR, & GARDINER, R.C. (1956) The nature of labile citrovorum factor in human urine. Journal of Biological Chemistry, 223, 259. WAXMAN, S. (1975) Folate binding proteins. British Journal ofHaematology, a9, 23. WAXMAN, S. & SCHREIBER, C. (1972) Further studies of serum folk acid binding protein (FABP). (Abstract). American Journal of Clinical Nutrition, 25, 450. WAXMAN, S. & SCHREIBER, C. (1973) Characteristics of folk acid-binding protein in folate-deficient serum. Blood, 42, 291. WEIR,D.G., BROWN, J.P., FREEDMAN, D.S. & SCOTT, J.M. (1973) The absorption of the diastereoisomers of 5-methyltetrahydropteroylglutamate in man: a carrier-mediated process. Clinical Science and Molecular Medicine, 45, 625. WHITEHEAD, V.M. & COOPER, B.A. (1967) Absorption of unaltered folic acid from the gastro-intestinal tract in man. British Journal offfaematology, 13, 679.
