BritishJournal ofHaematology, 1976, 34, 489.
Folic Acid Metabolism in Normal, Folate Deficient and Alcoholic Man F. LANE,P. GOFF,R. MCGUFFIN, E. R. EICHNERAND R. S. HILLMAN Health Sciences Learning Resources Center, University of Washington, Seattle, U.S.A. (Received 23 January 1976;acceptedfor publication 5 April 1976)
SUMMARY.Folate metabolism was studied in normal, folate-deficient and alcoholic man by tracer measurements of plasma clearance, urinary excretion, tissue storage and release of folate using both [3H]pteroylglutamic acid (3H-PteGlu) and I4Cmethyl-H4PteGlu. Alcohol ingestion did not adversely affect tissue uptake of folates. Whether in normal or folate deficient subjects, the relative clearance rates of H-PteGlu and 4C-methyl-H4PteG1~were maintained in the face of alcohol ingestion and there was no evidence of increased urinary loss of intact vitamin or labelled breakdown products. As measured by the flushing technique, the rate of storage or tissue binding of 'H-PteGlu was not influenced by folate deficiency, folate store depletion or alcohol ingestion. However, alcohol may retard the release of methyl-H4PteGlu from tissue stores to plasma, A significantly greater recovery of ''C-methyl-H4PteGlu with flush was observed in those normal subjects who ingested alcohol for 6 d. A partial block in the rate of release of tissue folate stores would be a possible mechanism behind the rapid depression in serum methylH,PteGlu levels and early induction of megaloblastic erythropoiesis which has been observed following acute alcohol ingestion. Herbert (1962)first demonstrated induction of megaloblastic erythropoiesis in man by prolonged ingestion of a folate-deficient diet. Subsequent dietary studies have not only confirmed his work, but have also demonstrated the variability of body folate stores, the importance of stores and dietary intake in the development of folate deficiency in alcobolics, and a potential direct toxic effect of alcohol on erythropoiesis (Sullivan & Herbert, 1964; Eichner & Hillman, 1971;Eichner et al, 1971b;W u et al, 1975).Recently, alcohol has been shown to accelerate the appearance of folate deficiency in subjects with relatively normal folate stores by rapidly depressing the serum folate to a deficient level (Eichner & Hillman, 1973; Paine et al, 1973;McGu&n ct (11, 1975).The mechanism of this effect is as yet unknown, but could involve a specific alcohol-induced defect in folate metabolism or storage kinetics. Early studies of folate kinetics in man, first with large amounts of unlabelled pteroylglutamic acid (PteGlu) and later 'H-PteGlu and 14C-methyl-H4PteGlu, have demonstrated a rapid plasma clearance with uptake into an 'intracellular' compartment (Schweigert, 1948; Chanarin et a!, 1958,1963;Johns et al, 1961; Chanarin & Bannett, 1962;Herbert & Zalusky, 1962;Yoshino, 1968;Chanarin & Perry, 1968).The clearance of PteGlu from plasma is followed both by a rise in plasma methyl-H,PteGlu levels and the appearance of polyglutamate stores in organs such as the liver and kidneys. This sequence involves several biochemical Correspondence: Dr Robert S. Hillman, Health Sciences Learning Resources Center, University of Washington, Seattle 98195,U.S.A.
489
F. Lane et a1
490
conversions, including reduction and methylation of the PteGlu and subsequent polyglutamate formation. The avidity of tissue for folate is increased in folate-deficient subjects and patients with hyperthyroidism (Chanarin et al, 1958, 1963; Herbert & Zalusky, 1962; Lindcnbaum & Klipstein, 1964). However, in studies of a variety of other disorders, including vitamin-B, deficiency, leukaemia, aplastic anaemia, conditions with increased erythropoiesis and in liver disease, results have been variable (Chanarin et al, 1959, 1962; Girdwood & Delamore, 1961; Metz et al, 1961; Zail et al, 1963; Hogan et al, 1964). The effect of acute alcohol ingestion on folate kinetics has not been studied. For this reason, investigations were carried out on the role of disturbed tissue uptake, storage or release of folates as a cause of the sudden change in plasma folate levels with alcohol ingestion. Using 'H-PteGlu and I4Cmethyl-H,PteGlu, the plasma clearance, urinary excretion, tissue storage and intracellular metabolism were studied in normal, folate-deficient and alcoholic man. MATERIALS AND METHODS Forty-four volunteer subjects were studied in the Clinical Research Center at either Harborview Medical Center or the University Hospital, Seattle, Washington, observing the principles of the Helsinki Declaration for Human Research. On admission, a complete history was obtained from all subjects, and a physical and laboratory examination, including liver profile, was carried out. Routine haematological tests included a complete blood count, bone marrow aspirate for Wright's and Prussian blue stains, serum iron and total iron binding capacity (Bothwell & Mallett, 1955; Morgan & Carter, 1960), serum folate (Herbert, 1966) and vitamin-B, levels (Hillman ct al, 1969). No subject was admitted to the study until the evaluation confirmed normal liver function and general good health. Seven individuals demonstrated serum folate levels below 4 pg/l while two others showed mild megaloblastic erythropoiesis by marrow aspirate. Each had a history of poor diet and chronic alcohol ingestion. These subjects were studied immediately, prior to dietary correction of their folate deficiency. The remaining 37 subjects were judged to have normal folate balance on a normal dietary history, morning fasting L. cusei serum-folate values of 4-26 pg/l. on at least two occasions and normal haematology. Normal subjects were placed on either a regular or a low folate diet, containing no more than 5 &day L. casei-assayable folate (Eichner et ul, 1971a). The deficient diet was continued for 4-9 weeks until serum folate values were below 3 pgfl. Subsequently, several subjects from each category (normal diet and low folate diet) were also given I oz (28 ml) of pharmaceutical grade alcohol every 2 h to a total of eight doses per day for up to 6 d. Folate clearance from plasma was studied using both 3H-PteGlu, 30-50 Ci/mmole, and '4C-methyl-H4PteGlu, 50 Ci/molc (Amersham-Searle: Folic acid 3,5,9 [3H], potassium salt; ( */D.L.) 5-methyl ['4C]tetrahydr~folicacid, barium salt). The 'H-PteGlu was diluted in sterile water, divided into 50 pCi aliquots and frozen at - 70°C where it remained stable for up to 3 months. When chromatographed on DEAE (A-25) Sephadcx, 8 ~ 9 5 %of the tritium label was eluted in association with a single L. casci-active PteGlu peak, while the remaining activity was present on one or more contaminants, the most prominent being p-amino benzoylglutamate (Buehring ct al, 1973). At the time of each study, a 50 ,uCi aliquot was thawed, weighed, samples separated for standard count, repeat chromatography N
Folic Acid Metabolism and Alcoholism
49 1
and intravenous injection of 10-20 pCi (4 ng/kg 3H-PteGlu). All isotope studies were carried out at 09.00 hours with the subject fasting. Diluted in a sterile 0.1% ascorbate solution, l4C-methyl-H,PteGlu demonstrated more than 96% purity on DEAE chromatography and remained stable for more than 3 months at - 70°C. For each study, 10-20 pCi (4 pg/kg 4C-methyl-H4PteGlu) was injected intravenously while an aliquot was rechromatographed and counted as a standard. Following injection, serum L. casei values increased by 15-20 pg/l., effectively preventing a tracer measurement of plasma clearance. Venous samples were collected from the opposite arm at exactly 5 , 10,20, 30, 45 and 60 min, centrifuged and 0.5-1.0 ml serum suspended in 10 ml Aquasol (New England Nuclear Corporation) for counting. Since injected folate is not confined to the intravascular volume but immediately equilibrates with a volume equal to or greater than the extracellular fluid space (ECF; estimated as 16.6% of body weight), isotope clearance was calculated as per cent remaining in ECF with time. Plotted in this way, H-PteGlu clearance approaches an exponential function in normal subjects (Fig I). Following isotope injection, all urine was collected at intervals of 2 h or less, the volume measured, I ml pipetted into Aquasol for counting, and after addition of mercaptoethanol, a 30-1111 aliquot either frozen or immediately chromatographed to determine the character of excreted folates. All urine chromatography was carried out at 4°C in the dark using a 1.5 x 3 0 cm column, DEAE (A-25) Sephadex and a linear gradient of 0.1-1.0 M potassium phosphate buffer, pH 6.0 containing z m~ 2-mercaptoethanol. Effluent was collected in 5 ml fractions at a column flow rate of 40 ml/h. From each fraction, 1.0ml was then pipetted into Aquasol for counting. Labelled PteGlu, N-s methyl-H4PteGlu and paminobenzoylglutamate were employed as markers in selected chromatographs, with peak elutions occurring at tube numbers 130-140 with PteGlu, tube number 100-110 with niethyl-H,PteGlu and tube 50 with paminobenzoylglutamate. At varying times, ranging from o 11 to 5 d after isotope injection, an intravenous flushing dose of unlabelled PteGh was given as a single injection of 3oopg/kg followed by an infusion of 50 pg/kg/h for 6 h. In subjects receiving the I4C methyl-H,PteGlu, unlabelled methylH,PteGlu (Sigma Chemicals, Inc.) was given as freshly-prepared pulse doses of IOO pg/kg at 0, I, 2, 3 and 5 h, in order to avoid oxidation of the methylated folate. These rates of administration resulted in serum values in excess of 1000 ,ug/l. The ability of the flushing dose to clear unbound labelled folate was confirmed by initiating the flush immediately after an injection of3H-PteGlu. In three such studies, 85-95% ofthe injected isotope was recovered in the urinewithin 6 11. The efficiency ofthis clearancewas IO-I~% better than that previously reported by Johns at a1 (1961) at comparable levels of injected carrier. With the initiation of the flushing procedure, timed blood samples and all urines were collected for the next 24 h. Aliquots of the first two urine samples were immediately chroniatographed to determine the output of various labelled folates, whether intact PteGlu or methyl-H,PteGlu or breakdown products.
RESULTS Cleararrca of 3H-PteGlii
When administered in tracer amounts(4ng/kgL. casei-active folate) 3H-PteGluininiediately equilibrated with a volume somewhat larger than the extracellular fluid space and then was
492
F. Lane et af
Mites
FIG I. Tissue uptake of 3H-PteGlu was calculated as per cent original counts remaining in ECF with time from plasma count rates. Clearance approached an exponential function in normal subjects (o), Tt 5-8 niin. Folate-deficient subjects (0) demonstrated a more rapid clearance, Ti2-3 min.
cleared exponentially; Ti = 5-8 min in seven subjects maintained on a normal diet (Fig I). In two subjects with megaloblastic erythropoiesis and three individuals maintained on the low-folate diet until serum folate levels fell below 3 pg/l. clearance was more rapid, Ti being less than 2-3 niin (Figs I and 2 ) . At the same time, superimposed alcohol ingestion did
0
I
lormal
Normal
+
Alcohol
Low
folate
Low folote
+
Alcohol
FIG 2. Using the percentage of 3H-PteGlu remaining in ECF at min for comparison, normal individuals ingesting alcohol (0) failed to show a difference in 3H-PteGlu clearance from that seen in normal controls (0).Similarly, subjects from the folate-deficient diet group who also ingested alcohol (A) demonstrated the same rapid clearance of jH-PteGlu as appreciated in the two megaloblastic subjects (m) and two subjects on the folate defiaent diet (A).
Folic Acid Metabolism and Alcoholism
493
not appear to influence the clearance rates in either the folate-deficient or normal diet groups (Fig 2). The efficiency of tissue uptake following H-PteGlu injection was evaluated by monitoring urine for loss of the parent folate or breakdown products over a 32 h period. By 8 h, 85-goyo of the contaminant counts injected could be accounted for in the urine, while none of the original 3H-PteGlu was excreted. This pattern was sustained throughout the next 24 h ; less than 1% of the total counts injected appeared in the urine. Likewise, chromatograph studies of post-injection urines from folate-deficient and alcoholic subjects failed to reveal an abnormal loss of H-PteGh and the quantity of labelled p-aminobenzoyl-glutamate excreted did not exceed the original level of injected contaminant.
20
0
30 Minutes
6 0 2
4
Hwrs
6
I
2 k Y S
FIG 3. Following intravenous injection of ''C-methyl-H4PteGlu, clearance from plasma (0)was extremely rapid. At the same time, there was a linear excretion of 16-20% of the isotope during the first 24 h (o),primarily as a single early chromatograph peak. (See insert of example DEAE A-2s sephedex chromatograph of urine following injection. The position of intact, marker 14C-methyl-H4PteGlu is indicated by the dotted line.) This would suggest an oxidative destruction of the 14Cmethyl-H4PteGlu in either plasma or kidney, since great care was taken to maintain a reducing milieu in urine samples and carry out urine chromatographs immediately after void.
Clearance of C-methyl-H,PteGlu Following intravenous injection of 10-20pCi14C methyl-H,PteGlu, plasma clearance was extremely rapid (Fig 3). Less than 10% of the original counts remained in ECF by IS min, regardless of the folate status or alcohol intake of the subject. Moreover, accurate T, clearance and turnover rates could not be determined because of the amount of folate injected (4 &kg L. casei activity) and the resultant change in post-injection serum folate levels. However, studies of urinary excretion of 14C label demonstrates differences in methylH,PteGlu excretion in the first 24 h. As shown in Fig 3 , both normal and deficient subjects excreted 16-zo% of the total radioactivity injected during the first 24 h, primarily as a single early chromatograph peak. This loss was far in excess of the level of labelled contaminants present in the original injection and could only be explained as breakdown of methylH,PteGlu following injection. At the same time, the excretion of intact I4C-methylH,PteGlu was different for normal and folate-deficient subjects. While 0.8-2.9% of the
F. Lane et a1
494
I'
r 1:-
0
E C P b
2.0
0 0
b
8 1.5 -
0 Y)
0
-
E
1.0- 0
o
0.5-
0 I
I
Nwmol
Namol
-+
nicohoi
A
A
LOW folate
mate
LOW
t nicohol
FIG 4. Urinary loss of intact W-rneth~l-H~PteGlu in the first 24 h was significantly less in folate deficient and alcohol treated subjects. This was compatible with a more rapid uptake by tissue and greater efficiency of renal conservation in individuals showing low serum methyl-H4PteGlu levels, whether as a part of folate store depletion or alcohol depression.
original isotope dose was excreted as intact ''C-methyl-H4PteGlu in eight normals, less than 0.6% was excreted in subjects who were either folate-deficient secondary to store depletion or whose serum-folate level was acutely depressed by alcohol ingestion. After the first 24 h, urinary loss of I4C label fell to between 0.7 and 1.1% of counts remaining in vim per 24 h regardless of dietary group or alcohol intake (Fig 5 ) . In contrast to the first 24 h, there was no apparent difference in the excretion of intact '4C-methyl-H4PteGlu in normals, folate-deficient and alcohol-treated subjects. All subjects excreted negligible amounts of ''C-methyl-H4PteGlu, less than 0.2% of remaining counts per 24 h. Sirbsequent Metabolism nf3H-PteCltt The flushing technique of Johns et a1 (1961) was employed to study the behaviour of H-PteGlu and ''C-methyl-H4PteGlu following their clearance from the ECF. During the first 24 h after 3H-PteGlu injection, the infusion of large quantities of PteGlu resulted in an immediate rise in plasma counts and excretion of intact 3H-PteGlu monoglutamate in urine. Normal subjects who received a flushing dose together with their initial 3H-PteGlu injection excreted between 8 5 and 95% of the original 3H-PteGlu within 6 h, while only 70% of the injected dose was recovered in urine when the flush was delayed for 2 h. By 4 h, recovery decreased to 57% and by 24 h only 3 3% of the original H-PteGh was excreted (Fig 6 ) .This movement of H-PteGlu into a non-flushable, tissue-bound pool appeared to describe an exponential function with at least two components. Assuming an efficiency of 85% for the 6 h flushing technique, the initial slope demonstrated a T+clearance of 6 h, followed by a slower clearance with T+ c 23 h. Whether this reflects the presence of more than one tissue pool is not clear, since normal subjects were maintained on a regular diet and, therefore, experienced several new entries of dietary folate following the 3H-PteGlu injection. Folatedeficient subjects and both normal and deficient individuals ingesting alcohol were studied at I 5 h following €3-PteGlu injection. Their rate of clearance into non-flushable pools was neither faster nor slower when compared to the normal group.
Folic Acid Metabolism and Alcoholism
t
'"11
495
Normol
Normol
LOW
t
folole
Alcohol
LOW folote
+
Alcohol
FIG 5 . When excretion of I4C label was monitored over the next 4 d, less than 1% of the counts remaining in vivo were excreted per 24 h regardless of folate nutrition or alcohol ingestion. The major part of the activity excreted was identified by chromatography of one or more freshly voided urines to be breakdown activity (open bars). Less than 0.2% was excreted as intact 14C-methylH4PteGlu (hatched area).
0
8
16
24 Hours
FIG 6. Normal subjects (0)were flushed at intervals of 2,4,6,16and 24 h after 3H-PteGlu injection. Recoveries of intact "-PteGlu fell rapidly according to a complex function with at least two components (solid line, T+ = 6 h; broken line, T+ = 23 h). Studies of folate-deficient, megaloblastic subjects (m), folate-deficient-diet individuals (A) and deficient subjects and normals who ingested alcohol (A and 0 ) revealed urine recoveries which were identical to the normal control results. Thus, folate deficiency and alcohol ingestion did not appear to effect the rate of tissue binding of 3H-PteGlu.
Subsequerit Metabolism of'4C-methyl-H4PteGlti The recovery of ''C-methyl-H4PteGlu with flushing was studied on the fifth day following isotope administration. Similar to the experience with 3H-PteGlu, intravenous injection of pulse doses of IOO pg/kg methyl-H,PteGlu resulted in an immediate rise in plasma 14C counts and the excretion of intact '4C-methyl-H4PteGlu. While daily excretion of I4Crnethyl-H,Pte-Glu had not exceeded 0.2% of I4C counts remaining in uiuo per 24 h during the 3 d prior to flush, 20-27% of the remaining isotope was excreted in the first 6 h after flush in normal subjects (Fig 7). This labelled material was demonstrated by column chroniatography to be identical to the original 14C-methyl-H4PteGlu,better than 90% of the activity eluting in a single peak at tube numbers roo-110. Subjects on the low-folate diet with or
F. Lane et al
496
without alcohol supplementation demonstrated 19-22% excretion with flush. This was not significantly different than that seen in normals. In contrast, the four normal subjects who received supplements of alcohol excreted 32-3 8% of the remaining isotope as 14C-methylH,PteGlu. This represented a significant increase in recovery after flushing (P< 0.005) suggesting a decrease in ['4C]methyl-group utilization in the alcohol treated group.
A
Normal
Normal
-+
&coho1
:L1 Low folole
Low folote
+
Alcohol
FIG 7. With flush, normal subjects (e),folate-deficient-diet. subjects ( A ) and folate-deficient-plusalcohol individuals (A) demonstrated similar recoveries of ''C-methyl-H4PteGlu. However, normal subjects ingesting alcohol excreted 32-38% of the remaining isotope as intact l4C-methy1-H4PteGlu ( P < 0.00s).
DISCUSSION Normal man maintains limited tissue stores of folate. In the face of dietary deprivation or a defect in absorption, folate deficiency and megaloblastic erythropoiesis appear within a few weeks or months (Herbert, 1962;Eichner et al, 1971b;Eichner & Hillman, 1971).Recent reports have emphasized the importance of both store size and alcohol ingestion in deterrnining the rate of development of abnormal erythropoiesis. Patients with reduced folate stores because of an inadequate diet demonstrate a more rapid fall in serum folate values and an earlier onset of megaloblastic erythropoiesis compared to normal individuals (Eichner et al, 1971b). Alcohol ingestion dramatically depresses the serum-folate level in a matter of hours and, depending on the level of folate intake and store size, can induce megaloblastic erythropoiesis in less than 10 d (Sullivan & Herbert, 1964;Eichner & Hillman, 1971,1973;Paine et al, 1973).Though the exact mechanism of action is unknown, the ability of alcohol rapidly to depress circulating ievels of methyl-H4PteGlu suggests interference with normal mechanisms of tissue storage, release and transport to metabolically active cells (McGuffin et a!, 1975). To investigate this phenomenon, studies of plasma clearance, tissue uptake, storage and subsequent release were carried out using radio-labelled folic acid (PteGlu) and 5-methyltetrahydiofolate (methyl-H,PteGlu), in normal individuals, patients with folate-deficient niegaloblastic erythropoiesis, subjects maintained on a deficient diet for 3-6 weeks and in both normal and deficient subjects ingesting alcohol. Previous studies of tissue uptake and
Folic Acid Metabolism and Alcoholism
497
metabolism of PteGlu have used either large doses of unlabelled PteGlu or an 3H-PteGlu preparation of relatively low specific activity (Schweigert, 1948; Chanarin et al, 1958; Johns et al, 1961; Herbert & Zalusky, 1962). With ingestion or injection of sufficient unlabelled PteGh to change the L. casei serum folate level, plasma clearance is incomplete and unchanged PteGlu, formyl-H,PteGlu and methyl-H,PteGlu are excreted in urine (Baker et al, 1965). When 3H-PteGlu was injected at dosage levels of I, I S , 1 5 0 and 1430 pg/kg by Johns et a1 (1961),plasma clearance was rapid and complete at the I pg/kg level only. From this work it was concluded that plasma clearance involves a rapid equilibration with extracellular fluid and removal of folate into an intracellular tissue space. The present study used a dose of 4 ng/kg of high specific activity 'H-PteGlu. With this extremely small dose, normal subjects demonstrated immediate equilibration in a volume slightly greater than the estimated extracellular fluid space, followed by an exponential clearance into tissue, within T+ = 5-8 min (Fig I). This behaviour differed from that reported by Johns et al(1961).At their lowest dose, I pg/kg PteGlu, they found almost immediate equilibration with a space larger than total body water and a T, of less than I min. The present data was more typical of the ciearance curves observed by Johns et al(1961) with the 15 pg/kg dose and would be more compatible with transport by facilitated diffusion. The efficiency of tissue uptake was studied by monitoring urine for any loss of 3H-PteGlu. Only liniited amounts of labelled breakdown products, ['HIP-aniinobenzoyl glutamate and [3H]pterins appeared in urine in the 32 h following injection. Intact 3H-PteGlu was never detected. This was true for both normals and the folate-deficient alcohol group. It must be recognized that despite repeated chromatographic purification, up to I 5% of the radio-label in an H-PteGlu preparation may be on breakdown products, primarily p-aminobenzoyl glutamate (Buehring et al, 1973). Since p-aminobenzoyl glutamate does not gain entry into the intracellular space but is rapidly and quantitatively excreted in urine, the observed 3H-PteGlu clearance rate will be influenced by the clearance rate of this contaminant into urine (Johns et al, 1961). With limited amounts of labelled contaminant in the original injection this has only a minor effect on the clearance rate; the mean 3H-PteGlu T+clearance in our normal subjects changed by less than I min when corrected for clearance of 3H paminobenzoyl glutamate into urine. However, with a major breakdown of 3H-PteGlu, the effect could be significant. Previous clearance data should be interpreted in this light. The rate of uptake of the 3H-PteGlu into tissue was significantly influenced by the nutritional status of the individual. Both folate-deficient and megaloblastic subjects and subjects on folate-deficient diet demonstrated significant shortening of their clearance rates (Fig I). The T, clearance for the deficient subjects was in the range of 2-3 min, compared to 5-8 min for the normal group. It must be recognized, however, that neither this nor past studies have accurately measured the plasma concentration of PteGlu so as to permit the calculation of the actual PteGlu turnover (T,clearance relative to plasma concentration). Available bacteriological and isotope dilution assays of plasma folates lack the sensitivity required for detection and separation of methylated and nonmethylated folates at levels below I pg/l. Since a variation in plasma PteGlu content of as little as 50 ng/l. would be sufficient to explain the shortened T,, the apparent increased avidity of tissue for 3H-PteGlu may not represent an actual change in transport kinetics but only a slight reduction in a small, unmeasurable circulating PteGIu pool.
498
F. Lane et a1
In both the normal and folate-deficient subjects, the ingestion of alcohol failed to effect the
'H-PteGlu clearance (Fig 2 ) . Subjects from the normal group who received alcohol demon-
strated a normal clearance despite the fact that their serum methyl-H,PteGlu levels had been acutely depressed to values below 3 pg/l. Similarly, the more rapid clearance of the folatedeficient group was unchanged when these individuals received alcohol. Thus, a blocking effect of alcohol on tissue uptake of PteGlu could not be demonstrated. In addition, PteGlu clearance did not appear to be affected by the level of circulating methyl-H,PteGlu. This would tend to confirm the postulate of separate cellular transport mechanisms for PteGlu and methyl-H,PteGlu (Nahas et al, 1969; Das & Hoffbrand, 1970). Although a tracer measurement of rate of tissue uptake of ''C-methyl-H4PteGlu was not possible because of the amount injected, 4 pg/kg L. card activity, and the tendency toward early breakdown, it was apparent that the Clearance rate of the isotope from the extracellular fluid space exceeded that of PteGlu (Fig 3). At the same time, chromatography of urine collected in the first 24 h after injection revealed a significant difference in methyl-H,PteGlu excretion between normal subjects and subjects who were either folate-deficient or treated with alcohol (Fig 4). The efficiency of conservation of l4C-methyl-H4PteG1u by the kidney was greater in those individuals whose intial serum folates were below 3 pg/l. This suggests that the T+ clearance rate of methyl-H,PteGlu from plasma correlates with the size of the circulating methyl-H,PteGlu pool, and does so regardless of the reason for the reduction in serum folate. Tissue storage and subsequent metabolism of both PteGlu and methyl-H,PteGlu was studied by injecting large doses of unlabellcd folic acid to displace intact 'H-PteGlu and 4C-methyl-H4PteGlu from tissue. Sufficient PteGlu or methyl-H,PteGlu was infused to elevate the serum L. casei level to values exceeding 1.0mg/l. for a period of at least 6 h, a level which completely blocks tubular resorption and results in renal clearance of labelled folate according to the glomerular filtration rate (Johns et al, 1971; Goresky et al, 1963 ;Johns & Plenderleth, 1963). At the same time, clearance is not complete inasmuch as repeated flushes at 24 h intervals result in an additional excretion of small amounts of isotope. This implies either a variable degree of folate binding within cell cytoplasm and mitochrondria or a spectrum of binding affinities to different organs (Corrocher & Hoffbrand, 1972). In the face of the high level of unlabelled PteGlu used for flushing, it seems unlikely that it could be explained on the basis of plasma binding. However, in favour of the latter explanation,Johns' (1961) study of plasma binding and renal clearance demonstratesa fixed upper limit of clearance which is lower than the glomerular filtration rate, secondary to weak plasma protein binding of a constant 60-70% of circulating PteGlu regardless of plasma concentration. Despite the limitations of the flushing technique, it was possible to demonstrate a progressive decrease in the recoverability of 'H-PteGlu and 14C-methyl-H4PteGlu with time In the case of 'H-PteGIu, the fall in counts recovered in urine followed a complex curve with two or more components. Whether this finding implies more than one tissue PteGh pool with different clearance kinetics is unclear, Since normal subjects were permitted to return to their normal dietary regimen following the isotope injection, the recovery data must be interpreted in the light of an ininterrupted flow of dietary folate into tissue stores. An entry of even a small amount of unlabelled PteGlu into tissue could result in a progressive delay in 3H-PteGlu movement to a tissue-bound pool and a complex function such as that shown
'
Folic Acid Metabolism and Alcoholism
499
in Fig 6. Folate-deficient subjects were studied at 15 h following injections of 3H-PteGlu. Urine recoveries of intact H-PteGlu were not significantly different from that seen in normal subjects. This suggests that folate deficiency and depletion of stores do not increase the rate of utilization or conversion of 3H-PteGlu to a tissue-bound storage form. In addition, alcohol did not affect the rate of conversion regardless of the state of folate nutrition. In contrast to the study of Cherrick et a1 (1965), in which patients with alcoholic cirrhosis demonstrated a much greater release of 3H-PteGlu into plasma with flush, acute alcohol ingestion in the present study did not result in a greater recovery. Moreover, the report of Brown et a1 (1973) that alcohol interferes with liver polyglutamate formation could not be confirmed. If the rate of tissue binding is determined by polyglutamate formation, then it would have been predicted that the recovery of 3H-PteGlu with flush would be increased in alcoholic patients. This was not the case in the present study. Studies of 14C-methyl-H4PteGlu recovery in urine after flush showed a difference in alcohol-treated subjects. As shown in Fig 7, normal, folate-deficient subjects and folatedeficient individuals treated with alcohol showed similar recoveries of ''C-methyl-H4PteGlu with flush. At the same time, there was a significant increase in the amount of isotope recovered in the four normal subjects who received alcohol supplements. This suggests a blocking effect of alcohol on the mobilization and transport of methyl-H,PteGlu from tissue stores to plasma in the acutely alcoholic individual. Obviously, this would be an attractive way to explain the sudden fall in serum folate levels which is observed with the acute ingestion of alcohol. However, it would have to be a relatively weak effect inasmuch as the folatedeficient individuals who received alcohol did not show a similar increase in recoverable C-methyl-H,PteGlu. Further studies will be necessary to confirm this possibility ACKNOWLEDGMENTS
This work was supported by U.S. Public Health Service Grants No. AM-13732, AA-00196 and RR-133, and by Alcoholism and Drug Abuse Institute (Parent Account), State of Washington No. 63-0660. REFERENCES BAKER,H., FRANK,O., FEINGOLD, S., ZIFFER,H., GELLENE, R.A., LEEVY, C.M. & SOBOTKA, H. (1965) The fate of orally and parenterally administered folates. AmericanJournul of Clinical Nutrition, 17,88. BOTHWELL, T.H. & MALLBTT, B. (1955) The determination of iron in plasma or serum. Biochemical Jofrrnal, 59, 599. BROWN, J.P., DAVIDSON, G.E., SCOTT,J.M. & WEIR, D.G. (1973) Effect of diphenylhydantoin and ethanol feeding on the synthesis of rat liver folates from exogenous pteroylglutainate [3H]. Biochemical Pharmacology, 22, 3287. BUEHRING, K.U., TAMURA, T. & STOKSTAD, E.L.R. (1973) Folate coenzymes of Lactobacillus casei and Streptococcus faecalis. Journal of Biological Chemistry, 249, 1081. CHANARIN, I. & BENNETT, M. (1962) The disposal of small doses of intravenously injected folic acid. BritishJournal of Haernatology, 8, 28.
I. & PERRY,J. (1968) Metabolism of 5CHANARIN, methyltetrahydrofolate in pernicious anaemia. British Journal of Haernatology, 14,297. CHANARIN, I., BELCHER, E.H. & BERRY, V. (1963) The utilization of tritium-labelled folic acid in megaloblastic anaemia. BritishJournal ofHaematology, 9,456. CHANARIN, I., MOLLIN, D.L. & ANDERSON, B.B. (1958) The clearance from the plasma of folic acid injected intravenously in normal subjects with megaloblastic anaemia. British Journal of Haernatology, 4, 435.
CHANARIN, I., DACIE,J.V. & MOLLIN,D.L. (1959) Folic-acid deficiency in haemolytic anaemia. British Journal of Haematology, 5, 245. CHANARIN, I., BENNETT,M.C. & BERRY,V. (1962) Urinary excretion of histidine derivatives in megaloblastic anaemia and other conditions and a comparison with the folic acid clearance test. Journal of Clinical Pathology, 15, 269.
500
F. Lane et a1
CHERRICK, G.R., BAKER, H., FRANK, 0.& LEEW,C.M. (1965) Observations on hepatic avidity for folate in Laennec's cirrhosis. Journal of Laboratory and Clinical Medicine, 66, 446. CORROCHER, R. & HOFFBRAND, A.V. (1972) Subcellular localization of folate and effect of methotrexate on the incorporation of radioactive folic acid into guinea-pig liver folate. Clinical Science, 43, 81 5. DAS, K.C. & HOFFBRAND, A.V. (1970) Studies of folate uptake by phytohaemagglutinin-stimulated lymphocytes. BritishJournal ofHaematology, 19. 203. EICHNER, E.R. & HILLMAN, R.S. (1971) The evolution of anemia in alcoholic patients. American Journal of Medicine, 50, 218. EICHNER,E.R. & HILLMAN,R.S. (1973) Effect of alcohol on serum folate level. Journal of Clinical Investigation, 52, 584. EICHNER, E.R., BUERGEL, N. & HILLMAN, R.S.(1g71a) Experience with an appetizing, high protein, low folate diet in man. American Jownal of Clinical Nutrition, 24. 1337. EICHNER, E.R., PIERCE,I. & HILLMAN, R.S. (1g71b) Folate balance in dietary-induced megaloblastic anemia. New EnglandJournal of Medicine, 284, 93 3. GIRDWOOD, R.H. & DELAMORE, I.W. (1961) Observations on tests of folic acid absorption and clearance. Scottish Medical Journal, 6 , 4. GORESKY, C.A., WATANABE, H. &JOHNS,D.G. (1963) The renal excretion of folk acid. Journal of Clinical Investigation, 42, 1841. HERBERT,V. (1962) Experimental nutritional folate deficiency in man. Transactions of the Association of American Physicians, 75, 307. HERBERT, V. (1966) Aseptic addition method for Lactobacillrrs casei assay of folate activity in human serum. Jotrmal of Clinical Pathology, 19, 12. HERBERT, V. & ZALUSKY, R. (1962) Interrelations of vitamin BIZ and folic acid metabolism: folic acid clearance studies.Journal of Clinical Investigation, 41, 1263.
HILLMAN,R.S., OAKES,M. & FINHOLT.C. (1969) Hemoglobin-coated charcoal radioassay for serum vitamin Biz. A simple modification to improve intrinsic factor reliability. Blood, 34, 385. HOGAN, J.A., MANIATIS, A. & MALONEY, W.C. (1964) The serum Lactobacillus casei folate clearance test in various hematologic disorders. Blood, 24, I 87. JOHNS,D.G. & PLENDERLETH, I.H. (1963) Folk aciddisplacement in man. Biochemical Pharmacology, 12, 1071.
JOHNS,D.G., SPERTI,S. & BURGEN, A.S.V. (1961)The
metabolism of tritiated folic acid in man. Joirrnal of Clinical Investigation, 40, 1684. LINDENBAUM, J. & KLIPSTBIN, F.A. (1964) Folic acid clearances and basal serum folate levels in patients with thyroid disease. Journal of Clinical Pathology, 17, 666.
MCGUFFIN, R., GOFF, P. & HILLMAN,R.S.(1975) The effect of diet and alcohol on the development of folate deficiency in the rat. BritishJournal ofHaematology, 31,185. METZ,J., STEVENS, K., KRAVITZ,S. & BRANDT,V. (1961) The plasma clearance of injected doses of folic acid as an index of folic acid deficiency. Journal of Clinical Pathology, 14, 622. MORGAN, E.H. & CARTER, G. (1960) Plasma iron and iron-binding capacity levels in health and disease: with an improved method for estimation of plasma iron concentration and total iron-binding capacity. Australasian Annals qf Medicine, 9, zog. NAHAS,A., NIXON, P.F. & BBRTINO, J.R. (1969) Transport of 5-methyltetrahydrofolate by LIZIO mouse leukemia cells. (Abstract). Federation Proceedings, 28, 389. PAINE,C.J., EICHNER,E.R. & DICKSON,V. (1973) Concordance of radioassay and microbiological assay in the study of the ethanol-induced fall in serum folate level. American Journal of the Medical Sciences, 266, 135. SCHWEIGERT, B.S. (1948) Folic acid metabolism studies. 111. Intravenous administration of pteroylglutamic acid and pteroyltriglutamic acid. Journal of laboratory and Clinical Medicine, 33. 1271. SULLIVAN, L.W. & HERBERT, V. (1964) Suppression of hematopoiesis by ethanol. Journal of Clinical Investigation, 43, 2048. Wu, A., CHANARIN,I., SLAVIN.G. & LEVX, A.J. (197s) Folate deficiency in the alcoholic-its relationship to clinical and haematological abnormalities, liver disease and folate stores. British Journal of Haematolo$)', 2% 469. YOSHINO,T. (1968) The clinical and experimental studies on the metabolism of folic acid using tritiated folic acid. 111. Plasma clearance in man and organ distribution in rat following intravenous administration of tritiated folk acid. Journal of Vitaminology, 14, 49. ZAIL,S., METZ,J. & BRANDT, V. (1963) The plasma clearance of injected doses of folic acid at varying rates of erythropoiesis. South African Jorrrnal of Medical Science, 28, 121.
Folic Acid Metabolism in Normal, Folate Deficient and Alcoholic Man F. LANE,P. GOFF,R. MCGUFFIN, E. R. EICHNERAND R. S. HILLMAN Health Sciences Learning Resources Center, University of Washington, Seattle, U.S.A. (Received 23 January 1976;acceptedfor publication 5 April 1976)
SUMMARY.Folate metabolism was studied in normal, folate-deficient and alcoholic man by tracer measurements of plasma clearance, urinary excretion, tissue storage and release of folate using both [3H]pteroylglutamic acid (3H-PteGlu) and I4Cmethyl-H4PteGlu. Alcohol ingestion did not adversely affect tissue uptake of folates. Whether in normal or folate deficient subjects, the relative clearance rates of H-PteGlu and 4C-methyl-H4PteG1~were maintained in the face of alcohol ingestion and there was no evidence of increased urinary loss of intact vitamin or labelled breakdown products. As measured by the flushing technique, the rate of storage or tissue binding of 'H-PteGlu was not influenced by folate deficiency, folate store depletion or alcohol ingestion. However, alcohol may retard the release of methyl-H4PteGlu from tissue stores to plasma, A significantly greater recovery of ''C-methyl-H4PteGlu with flush was observed in those normal subjects who ingested alcohol for 6 d. A partial block in the rate of release of tissue folate stores would be a possible mechanism behind the rapid depression in serum methylH,PteGlu levels and early induction of megaloblastic erythropoiesis which has been observed following acute alcohol ingestion. Herbert (1962)first demonstrated induction of megaloblastic erythropoiesis in man by prolonged ingestion of a folate-deficient diet. Subsequent dietary studies have not only confirmed his work, but have also demonstrated the variability of body folate stores, the importance of stores and dietary intake in the development of folate deficiency in alcobolics, and a potential direct toxic effect of alcohol on erythropoiesis (Sullivan & Herbert, 1964; Eichner & Hillman, 1971;Eichner et al, 1971b;W u et al, 1975).Recently, alcohol has been shown to accelerate the appearance of folate deficiency in subjects with relatively normal folate stores by rapidly depressing the serum folate to a deficient level (Eichner & Hillman, 1973; Paine et al, 1973;McGu&n ct (11, 1975).The mechanism of this effect is as yet unknown, but could involve a specific alcohol-induced defect in folate metabolism or storage kinetics. Early studies of folate kinetics in man, first with large amounts of unlabelled pteroylglutamic acid (PteGlu) and later 'H-PteGlu and 14C-methyl-H4PteGlu, have demonstrated a rapid plasma clearance with uptake into an 'intracellular' compartment (Schweigert, 1948; Chanarin et a!, 1958,1963;Johns et al, 1961; Chanarin & Bannett, 1962;Herbert & Zalusky, 1962;Yoshino, 1968;Chanarin & Perry, 1968).The clearance of PteGlu from plasma is followed both by a rise in plasma methyl-H,PteGlu levels and the appearance of polyglutamate stores in organs such as the liver and kidneys. This sequence involves several biochemical Correspondence: Dr Robert S. Hillman, Health Sciences Learning Resources Center, University of Washington, Seattle 98195,U.S.A.
489
F. Lane et a1
490
conversions, including reduction and methylation of the PteGlu and subsequent polyglutamate formation. The avidity of tissue for folate is increased in folate-deficient subjects and patients with hyperthyroidism (Chanarin et al, 1958, 1963; Herbert & Zalusky, 1962; Lindcnbaum & Klipstein, 1964). However, in studies of a variety of other disorders, including vitamin-B, deficiency, leukaemia, aplastic anaemia, conditions with increased erythropoiesis and in liver disease, results have been variable (Chanarin et al, 1959, 1962; Girdwood & Delamore, 1961; Metz et al, 1961; Zail et al, 1963; Hogan et al, 1964). The effect of acute alcohol ingestion on folate kinetics has not been studied. For this reason, investigations were carried out on the role of disturbed tissue uptake, storage or release of folates as a cause of the sudden change in plasma folate levels with alcohol ingestion. Using 'H-PteGlu and I4Cmethyl-H,PteGlu, the plasma clearance, urinary excretion, tissue storage and intracellular metabolism were studied in normal, folate-deficient and alcoholic man. MATERIALS AND METHODS Forty-four volunteer subjects were studied in the Clinical Research Center at either Harborview Medical Center or the University Hospital, Seattle, Washington, observing the principles of the Helsinki Declaration for Human Research. On admission, a complete history was obtained from all subjects, and a physical and laboratory examination, including liver profile, was carried out. Routine haematological tests included a complete blood count, bone marrow aspirate for Wright's and Prussian blue stains, serum iron and total iron binding capacity (Bothwell & Mallett, 1955; Morgan & Carter, 1960), serum folate (Herbert, 1966) and vitamin-B, levels (Hillman ct al, 1969). No subject was admitted to the study until the evaluation confirmed normal liver function and general good health. Seven individuals demonstrated serum folate levels below 4 pg/l while two others showed mild megaloblastic erythropoiesis by marrow aspirate. Each had a history of poor diet and chronic alcohol ingestion. These subjects were studied immediately, prior to dietary correction of their folate deficiency. The remaining 37 subjects were judged to have normal folate balance on a normal dietary history, morning fasting L. cusei serum-folate values of 4-26 pg/l. on at least two occasions and normal haematology. Normal subjects were placed on either a regular or a low folate diet, containing no more than 5 &day L. casei-assayable folate (Eichner et ul, 1971a). The deficient diet was continued for 4-9 weeks until serum folate values were below 3 pgfl. Subsequently, several subjects from each category (normal diet and low folate diet) were also given I oz (28 ml) of pharmaceutical grade alcohol every 2 h to a total of eight doses per day for up to 6 d. Folate clearance from plasma was studied using both 3H-PteGlu, 30-50 Ci/mmole, and '4C-methyl-H4PteGlu, 50 Ci/molc (Amersham-Searle: Folic acid 3,5,9 [3H], potassium salt; ( */D.L.) 5-methyl ['4C]tetrahydr~folicacid, barium salt). The 'H-PteGlu was diluted in sterile water, divided into 50 pCi aliquots and frozen at - 70°C where it remained stable for up to 3 months. When chromatographed on DEAE (A-25) Sephadcx, 8 ~ 9 5 %of the tritium label was eluted in association with a single L. casci-active PteGlu peak, while the remaining activity was present on one or more contaminants, the most prominent being p-amino benzoylglutamate (Buehring ct al, 1973). At the time of each study, a 50 ,uCi aliquot was thawed, weighed, samples separated for standard count, repeat chromatography N
Folic Acid Metabolism and Alcoholism
49 1
and intravenous injection of 10-20 pCi (4 ng/kg 3H-PteGlu). All isotope studies were carried out at 09.00 hours with the subject fasting. Diluted in a sterile 0.1% ascorbate solution, l4C-methyl-H,PteGlu demonstrated more than 96% purity on DEAE chromatography and remained stable for more than 3 months at - 70°C. For each study, 10-20 pCi (4 pg/kg 4C-methyl-H4PteGlu) was injected intravenously while an aliquot was rechromatographed and counted as a standard. Following injection, serum L. casei values increased by 15-20 pg/l., effectively preventing a tracer measurement of plasma clearance. Venous samples were collected from the opposite arm at exactly 5 , 10,20, 30, 45 and 60 min, centrifuged and 0.5-1.0 ml serum suspended in 10 ml Aquasol (New England Nuclear Corporation) for counting. Since injected folate is not confined to the intravascular volume but immediately equilibrates with a volume equal to or greater than the extracellular fluid space (ECF; estimated as 16.6% of body weight), isotope clearance was calculated as per cent remaining in ECF with time. Plotted in this way, H-PteGlu clearance approaches an exponential function in normal subjects (Fig I). Following isotope injection, all urine was collected at intervals of 2 h or less, the volume measured, I ml pipetted into Aquasol for counting, and after addition of mercaptoethanol, a 30-1111 aliquot either frozen or immediately chromatographed to determine the character of excreted folates. All urine chromatography was carried out at 4°C in the dark using a 1.5 x 3 0 cm column, DEAE (A-25) Sephadex and a linear gradient of 0.1-1.0 M potassium phosphate buffer, pH 6.0 containing z m~ 2-mercaptoethanol. Effluent was collected in 5 ml fractions at a column flow rate of 40 ml/h. From each fraction, 1.0ml was then pipetted into Aquasol for counting. Labelled PteGlu, N-s methyl-H4PteGlu and paminobenzoylglutamate were employed as markers in selected chromatographs, with peak elutions occurring at tube numbers 130-140 with PteGlu, tube number 100-110 with niethyl-H,PteGlu and tube 50 with paminobenzoylglutamate. At varying times, ranging from o 11 to 5 d after isotope injection, an intravenous flushing dose of unlabelled PteGh was given as a single injection of 3oopg/kg followed by an infusion of 50 pg/kg/h for 6 h. In subjects receiving the I4C methyl-H,PteGlu, unlabelled methylH,PteGlu (Sigma Chemicals, Inc.) was given as freshly-prepared pulse doses of IOO pg/kg at 0, I, 2, 3 and 5 h, in order to avoid oxidation of the methylated folate. These rates of administration resulted in serum values in excess of 1000 ,ug/l. The ability of the flushing dose to clear unbound labelled folate was confirmed by initiating the flush immediately after an injection of3H-PteGlu. In three such studies, 85-95% ofthe injected isotope was recovered in the urinewithin 6 11. The efficiency ofthis clearancewas IO-I~% better than that previously reported by Johns at a1 (1961) at comparable levels of injected carrier. With the initiation of the flushing procedure, timed blood samples and all urines were collected for the next 24 h. Aliquots of the first two urine samples were immediately chroniatographed to determine the output of various labelled folates, whether intact PteGlu or methyl-H,PteGlu or breakdown products.
RESULTS Cleararrca of 3H-PteGlii
When administered in tracer amounts(4ng/kgL. casei-active folate) 3H-PteGluininiediately equilibrated with a volume somewhat larger than the extracellular fluid space and then was
492
F. Lane et af
Mites
FIG I. Tissue uptake of 3H-PteGlu was calculated as per cent original counts remaining in ECF with time from plasma count rates. Clearance approached an exponential function in normal subjects (o), Tt 5-8 niin. Folate-deficient subjects (0) demonstrated a more rapid clearance, Ti2-3 min.
cleared exponentially; Ti = 5-8 min in seven subjects maintained on a normal diet (Fig I). In two subjects with megaloblastic erythropoiesis and three individuals maintained on the low-folate diet until serum folate levels fell below 3 pg/l. clearance was more rapid, Ti being less than 2-3 niin (Figs I and 2 ) . At the same time, superimposed alcohol ingestion did
0
I
lormal
Normal
+
Alcohol
Low
folate
Low folote
+
Alcohol
FIG 2. Using the percentage of 3H-PteGlu remaining in ECF at min for comparison, normal individuals ingesting alcohol (0) failed to show a difference in 3H-PteGlu clearance from that seen in normal controls (0).Similarly, subjects from the folate-deficient diet group who also ingested alcohol (A) demonstrated the same rapid clearance of jH-PteGlu as appreciated in the two megaloblastic subjects (m) and two subjects on the folate defiaent diet (A).
Folic Acid Metabolism and Alcoholism
493
not appear to influence the clearance rates in either the folate-deficient or normal diet groups (Fig 2). The efficiency of tissue uptake following H-PteGlu injection was evaluated by monitoring urine for loss of the parent folate or breakdown products over a 32 h period. By 8 h, 85-goyo of the contaminant counts injected could be accounted for in the urine, while none of the original 3H-PteGlu was excreted. This pattern was sustained throughout the next 24 h ; less than 1% of the total counts injected appeared in the urine. Likewise, chromatograph studies of post-injection urines from folate-deficient and alcoholic subjects failed to reveal an abnormal loss of H-PteGh and the quantity of labelled p-aminobenzoyl-glutamate excreted did not exceed the original level of injected contaminant.
20
0
30 Minutes
6 0 2
4
Hwrs
6
I
2 k Y S
FIG 3. Following intravenous injection of ''C-methyl-H4PteGlu, clearance from plasma (0)was extremely rapid. At the same time, there was a linear excretion of 16-20% of the isotope during the first 24 h (o),primarily as a single early chromatograph peak. (See insert of example DEAE A-2s sephedex chromatograph of urine following injection. The position of intact, marker 14C-methyl-H4PteGlu is indicated by the dotted line.) This would suggest an oxidative destruction of the 14Cmethyl-H4PteGlu in either plasma or kidney, since great care was taken to maintain a reducing milieu in urine samples and carry out urine chromatographs immediately after void.
Clearance of C-methyl-H,PteGlu Following intravenous injection of 10-20pCi14C methyl-H,PteGlu, plasma clearance was extremely rapid (Fig 3). Less than 10% of the original counts remained in ECF by IS min, regardless of the folate status or alcohol intake of the subject. Moreover, accurate T, clearance and turnover rates could not be determined because of the amount of folate injected (4 &kg L. casei activity) and the resultant change in post-injection serum folate levels. However, studies of urinary excretion of 14C label demonstrates differences in methylH,PteGlu excretion in the first 24 h. As shown in Fig 3 , both normal and deficient subjects excreted 16-zo% of the total radioactivity injected during the first 24 h, primarily as a single early chromatograph peak. This loss was far in excess of the level of labelled contaminants present in the original injection and could only be explained as breakdown of methylH,PteGlu following injection. At the same time, the excretion of intact I4C-methylH,PteGlu was different for normal and folate-deficient subjects. While 0.8-2.9% of the
F. Lane et a1
494
I'
r 1:-
0
E C P b
2.0
0 0
b
8 1.5 -
0 Y)
0
-
E
1.0- 0
o
0.5-
0 I
I
Nwmol
Namol
-+
nicohoi
A
A
LOW folate
mate
LOW
t nicohol
FIG 4. Urinary loss of intact W-rneth~l-H~PteGlu in the first 24 h was significantly less in folate deficient and alcohol treated subjects. This was compatible with a more rapid uptake by tissue and greater efficiency of renal conservation in individuals showing low serum methyl-H4PteGlu levels, whether as a part of folate store depletion or alcohol depression.
original isotope dose was excreted as intact ''C-methyl-H4PteGlu in eight normals, less than 0.6% was excreted in subjects who were either folate-deficient secondary to store depletion or whose serum-folate level was acutely depressed by alcohol ingestion. After the first 24 h, urinary loss of I4C label fell to between 0.7 and 1.1% of counts remaining in vim per 24 h regardless of dietary group or alcohol intake (Fig 5 ) . In contrast to the first 24 h, there was no apparent difference in the excretion of intact '4C-methyl-H4PteGlu in normals, folate-deficient and alcohol-treated subjects. All subjects excreted negligible amounts of ''C-methyl-H4PteGlu, less than 0.2% of remaining counts per 24 h. Sirbsequent Metabolism nf3H-PteCltt The flushing technique of Johns et a1 (1961) was employed to study the behaviour of H-PteGlu and ''C-methyl-H4PteGlu following their clearance from the ECF. During the first 24 h after 3H-PteGlu injection, the infusion of large quantities of PteGlu resulted in an immediate rise in plasma counts and excretion of intact 3H-PteGlu monoglutamate in urine. Normal subjects who received a flushing dose together with their initial 3H-PteGlu injection excreted between 8 5 and 95% of the original 3H-PteGlu within 6 h, while only 70% of the injected dose was recovered in urine when the flush was delayed for 2 h. By 4 h, recovery decreased to 57% and by 24 h only 3 3% of the original H-PteGh was excreted (Fig 6 ) .This movement of H-PteGlu into a non-flushable, tissue-bound pool appeared to describe an exponential function with at least two components. Assuming an efficiency of 85% for the 6 h flushing technique, the initial slope demonstrated a T+clearance of 6 h, followed by a slower clearance with T+ c 23 h. Whether this reflects the presence of more than one tissue pool is not clear, since normal subjects were maintained on a regular diet and, therefore, experienced several new entries of dietary folate following the 3H-PteGlu injection. Folatedeficient subjects and both normal and deficient individuals ingesting alcohol were studied at I 5 h following €3-PteGlu injection. Their rate of clearance into non-flushable pools was neither faster nor slower when compared to the normal group.
Folic Acid Metabolism and Alcoholism
t
'"11
495
Normol
Normol
LOW
t
folole
Alcohol
LOW folote
+
Alcohol
FIG 5 . When excretion of I4C label was monitored over the next 4 d, less than 1% of the counts remaining in vivo were excreted per 24 h regardless of folate nutrition or alcohol ingestion. The major part of the activity excreted was identified by chromatography of one or more freshly voided urines to be breakdown activity (open bars). Less than 0.2% was excreted as intact 14C-methylH4PteGlu (hatched area).
0
8
16
24 Hours
FIG 6. Normal subjects (0)were flushed at intervals of 2,4,6,16and 24 h after 3H-PteGlu injection. Recoveries of intact "-PteGlu fell rapidly according to a complex function with at least two components (solid line, T+ = 6 h; broken line, T+ = 23 h). Studies of folate-deficient, megaloblastic subjects (m), folate-deficient-diet individuals (A) and deficient subjects and normals who ingested alcohol (A and 0 ) revealed urine recoveries which were identical to the normal control results. Thus, folate deficiency and alcohol ingestion did not appear to effect the rate of tissue binding of 3H-PteGlu.
Subsequerit Metabolism of'4C-methyl-H4PteGlti The recovery of ''C-methyl-H4PteGlu with flushing was studied on the fifth day following isotope administration. Similar to the experience with 3H-PteGlu, intravenous injection of pulse doses of IOO pg/kg methyl-H,PteGlu resulted in an immediate rise in plasma 14C counts and the excretion of intact '4C-methyl-H4PteGlu. While daily excretion of I4Crnethyl-H,Pte-Glu had not exceeded 0.2% of I4C counts remaining in uiuo per 24 h during the 3 d prior to flush, 20-27% of the remaining isotope was excreted in the first 6 h after flush in normal subjects (Fig 7). This labelled material was demonstrated by column chroniatography to be identical to the original 14C-methyl-H4PteGlu,better than 90% of the activity eluting in a single peak at tube numbers roo-110. Subjects on the low-folate diet with or
F. Lane et al
496
without alcohol supplementation demonstrated 19-22% excretion with flush. This was not significantly different than that seen in normals. In contrast, the four normal subjects who received supplements of alcohol excreted 32-3 8% of the remaining isotope as 14C-methylH,PteGlu. This represented a significant increase in recovery after flushing (P< 0.005) suggesting a decrease in ['4C]methyl-group utilization in the alcohol treated group.
A
Normal
Normal
-+
&coho1
:L1 Low folole
Low folote
+
Alcohol
FIG 7. With flush, normal subjects (e),folate-deficient-diet. subjects ( A ) and folate-deficient-plusalcohol individuals (A) demonstrated similar recoveries of ''C-methyl-H4PteGlu. However, normal subjects ingesting alcohol excreted 32-38% of the remaining isotope as intact l4C-methy1-H4PteGlu ( P < 0.00s).
DISCUSSION Normal man maintains limited tissue stores of folate. In the face of dietary deprivation or a defect in absorption, folate deficiency and megaloblastic erythropoiesis appear within a few weeks or months (Herbert, 1962;Eichner et al, 1971b;Eichner & Hillman, 1971).Recent reports have emphasized the importance of both store size and alcohol ingestion in deterrnining the rate of development of abnormal erythropoiesis. Patients with reduced folate stores because of an inadequate diet demonstrate a more rapid fall in serum folate values and an earlier onset of megaloblastic erythropoiesis compared to normal individuals (Eichner et al, 1971b). Alcohol ingestion dramatically depresses the serum-folate level in a matter of hours and, depending on the level of folate intake and store size, can induce megaloblastic erythropoiesis in less than 10 d (Sullivan & Herbert, 1964;Eichner & Hillman, 1971,1973;Paine et al, 1973).Though the exact mechanism of action is unknown, the ability of alcohol rapidly to depress circulating ievels of methyl-H4PteGlu suggests interference with normal mechanisms of tissue storage, release and transport to metabolically active cells (McGuffin et a!, 1975). To investigate this phenomenon, studies of plasma clearance, tissue uptake, storage and subsequent release were carried out using radio-labelled folic acid (PteGlu) and 5-methyltetrahydiofolate (methyl-H,PteGlu), in normal individuals, patients with folate-deficient niegaloblastic erythropoiesis, subjects maintained on a deficient diet for 3-6 weeks and in both normal and deficient subjects ingesting alcohol. Previous studies of tissue uptake and
Folic Acid Metabolism and Alcoholism
497
metabolism of PteGlu have used either large doses of unlabelled PteGlu or an 3H-PteGlu preparation of relatively low specific activity (Schweigert, 1948; Chanarin et al, 1958; Johns et al, 1961; Herbert & Zalusky, 1962). With ingestion or injection of sufficient unlabelled PteGh to change the L. casei serum folate level, plasma clearance is incomplete and unchanged PteGlu, formyl-H,PteGlu and methyl-H,PteGlu are excreted in urine (Baker et al, 1965). When 3H-PteGlu was injected at dosage levels of I, I S , 1 5 0 and 1430 pg/kg by Johns et a1 (1961),plasma clearance was rapid and complete at the I pg/kg level only. From this work it was concluded that plasma clearance involves a rapid equilibration with extracellular fluid and removal of folate into an intracellular tissue space. The present study used a dose of 4 ng/kg of high specific activity 'H-PteGlu. With this extremely small dose, normal subjects demonstrated immediate equilibration in a volume slightly greater than the estimated extracellular fluid space, followed by an exponential clearance into tissue, within T+ = 5-8 min (Fig I). This behaviour differed from that reported by Johns et al(1961).At their lowest dose, I pg/kg PteGlu, they found almost immediate equilibration with a space larger than total body water and a T, of less than I min. The present data was more typical of the ciearance curves observed by Johns et al(1961) with the 15 pg/kg dose and would be more compatible with transport by facilitated diffusion. The efficiency of tissue uptake was studied by monitoring urine for any loss of 3H-PteGlu. Only liniited amounts of labelled breakdown products, ['HIP-aniinobenzoyl glutamate and [3H]pterins appeared in urine in the 32 h following injection. Intact 3H-PteGlu was never detected. This was true for both normals and the folate-deficient alcohol group. It must be recognized that despite repeated chromatographic purification, up to I 5% of the radio-label in an H-PteGlu preparation may be on breakdown products, primarily p-aminobenzoyl glutamate (Buehring et al, 1973). Since p-aminobenzoyl glutamate does not gain entry into the intracellular space but is rapidly and quantitatively excreted in urine, the observed 3H-PteGlu clearance rate will be influenced by the clearance rate of this contaminant into urine (Johns et al, 1961). With limited amounts of labelled contaminant in the original injection this has only a minor effect on the clearance rate; the mean 3H-PteGlu T+clearance in our normal subjects changed by less than I min when corrected for clearance of 3H paminobenzoyl glutamate into urine. However, with a major breakdown of 3H-PteGlu, the effect could be significant. Previous clearance data should be interpreted in this light. The rate of uptake of the 3H-PteGlu into tissue was significantly influenced by the nutritional status of the individual. Both folate-deficient and megaloblastic subjects and subjects on folate-deficient diet demonstrated significant shortening of their clearance rates (Fig I). The T, clearance for the deficient subjects was in the range of 2-3 min, compared to 5-8 min for the normal group. It must be recognized, however, that neither this nor past studies have accurately measured the plasma concentration of PteGlu so as to permit the calculation of the actual PteGlu turnover (T,clearance relative to plasma concentration). Available bacteriological and isotope dilution assays of plasma folates lack the sensitivity required for detection and separation of methylated and nonmethylated folates at levels below I pg/l. Since a variation in plasma PteGlu content of as little as 50 ng/l. would be sufficient to explain the shortened T,, the apparent increased avidity of tissue for 3H-PteGlu may not represent an actual change in transport kinetics but only a slight reduction in a small, unmeasurable circulating PteGIu pool.
498
F. Lane et a1
In both the normal and folate-deficient subjects, the ingestion of alcohol failed to effect the
'H-PteGlu clearance (Fig 2 ) . Subjects from the normal group who received alcohol demon-
strated a normal clearance despite the fact that their serum methyl-H,PteGlu levels had been acutely depressed to values below 3 pg/l. Similarly, the more rapid clearance of the folatedeficient group was unchanged when these individuals received alcohol. Thus, a blocking effect of alcohol on tissue uptake of PteGlu could not be demonstrated. In addition, PteGlu clearance did not appear to be affected by the level of circulating methyl-H,PteGlu. This would tend to confirm the postulate of separate cellular transport mechanisms for PteGlu and methyl-H,PteGlu (Nahas et al, 1969; Das & Hoffbrand, 1970). Although a tracer measurement of rate of tissue uptake of ''C-methyl-H4PteGlu was not possible because of the amount injected, 4 pg/kg L. card activity, and the tendency toward early breakdown, it was apparent that the Clearance rate of the isotope from the extracellular fluid space exceeded that of PteGlu (Fig 3). At the same time, chromatography of urine collected in the first 24 h after injection revealed a significant difference in methyl-H,PteGlu excretion between normal subjects and subjects who were either folate-deficient or treated with alcohol (Fig 4). The efficiency of conservation of l4C-methyl-H4PteG1u by the kidney was greater in those individuals whose intial serum folates were below 3 pg/l. This suggests that the T+ clearance rate of methyl-H,PteGlu from plasma correlates with the size of the circulating methyl-H,PteGlu pool, and does so regardless of the reason for the reduction in serum folate. Tissue storage and subsequent metabolism of both PteGlu and methyl-H,PteGlu was studied by injecting large doses of unlabellcd folic acid to displace intact 'H-PteGlu and 4C-methyl-H4PteGlu from tissue. Sufficient PteGlu or methyl-H,PteGlu was infused to elevate the serum L. casei level to values exceeding 1.0mg/l. for a period of at least 6 h, a level which completely blocks tubular resorption and results in renal clearance of labelled folate according to the glomerular filtration rate (Johns et al, 1971; Goresky et al, 1963 ;Johns & Plenderleth, 1963). At the same time, clearance is not complete inasmuch as repeated flushes at 24 h intervals result in an additional excretion of small amounts of isotope. This implies either a variable degree of folate binding within cell cytoplasm and mitochrondria or a spectrum of binding affinities to different organs (Corrocher & Hoffbrand, 1972). In the face of the high level of unlabelled PteGlu used for flushing, it seems unlikely that it could be explained on the basis of plasma binding. However, in favour of the latter explanation,Johns' (1961) study of plasma binding and renal clearance demonstratesa fixed upper limit of clearance which is lower than the glomerular filtration rate, secondary to weak plasma protein binding of a constant 60-70% of circulating PteGlu regardless of plasma concentration. Despite the limitations of the flushing technique, it was possible to demonstrate a progressive decrease in the recoverability of 'H-PteGlu and 14C-methyl-H4PteGlu with time In the case of 'H-PteGIu, the fall in counts recovered in urine followed a complex curve with two or more components. Whether this finding implies more than one tissue PteGh pool with different clearance kinetics is unclear, Since normal subjects were permitted to return to their normal dietary regimen following the isotope injection, the recovery data must be interpreted in the light of an ininterrupted flow of dietary folate into tissue stores. An entry of even a small amount of unlabelled PteGlu into tissue could result in a progressive delay in 3H-PteGlu movement to a tissue-bound pool and a complex function such as that shown
'
Folic Acid Metabolism and Alcoholism
499
in Fig 6. Folate-deficient subjects were studied at 15 h following injections of 3H-PteGlu. Urine recoveries of intact H-PteGlu were not significantly different from that seen in normal subjects. This suggests that folate deficiency and depletion of stores do not increase the rate of utilization or conversion of 3H-PteGlu to a tissue-bound storage form. In addition, alcohol did not affect the rate of conversion regardless of the state of folate nutrition. In contrast to the study of Cherrick et a1 (1965), in which patients with alcoholic cirrhosis demonstrated a much greater release of 3H-PteGlu into plasma with flush, acute alcohol ingestion in the present study did not result in a greater recovery. Moreover, the report of Brown et a1 (1973) that alcohol interferes with liver polyglutamate formation could not be confirmed. If the rate of tissue binding is determined by polyglutamate formation, then it would have been predicted that the recovery of 3H-PteGlu with flush would be increased in alcoholic patients. This was not the case in the present study. Studies of 14C-methyl-H4PteGlu recovery in urine after flush showed a difference in alcohol-treated subjects. As shown in Fig 7, normal, folate-deficient subjects and folatedeficient individuals treated with alcohol showed similar recoveries of ''C-methyl-H4PteGlu with flush. At the same time, there was a significant increase in the amount of isotope recovered in the four normal subjects who received alcohol supplements. This suggests a blocking effect of alcohol on the mobilization and transport of methyl-H,PteGlu from tissue stores to plasma in the acutely alcoholic individual. Obviously, this would be an attractive way to explain the sudden fall in serum folate levels which is observed with the acute ingestion of alcohol. However, it would have to be a relatively weak effect inasmuch as the folatedeficient individuals who received alcohol did not show a similar increase in recoverable C-methyl-H,PteGlu. Further studies will be necessary to confirm this possibility ACKNOWLEDGMENTS
This work was supported by U.S. Public Health Service Grants No. AM-13732, AA-00196 and RR-133, and by Alcoholism and Drug Abuse Institute (Parent Account), State of Washington No. 63-0660. REFERENCES BAKER,H., FRANK,O., FEINGOLD, S., ZIFFER,H., GELLENE, R.A., LEEVY, C.M. & SOBOTKA, H. (1965) The fate of orally and parenterally administered folates. AmericanJournul of Clinical Nutrition, 17,88. BOTHWELL, T.H. & MALLBTT, B. (1955) The determination of iron in plasma or serum. Biochemical Jofrrnal, 59, 599. BROWN, J.P., DAVIDSON, G.E., SCOTT,J.M. & WEIR, D.G. (1973) Effect of diphenylhydantoin and ethanol feeding on the synthesis of rat liver folates from exogenous pteroylglutainate [3H]. Biochemical Pharmacology, 22, 3287. BUEHRING, K.U., TAMURA, T. & STOKSTAD, E.L.R. (1973) Folate coenzymes of Lactobacillus casei and Streptococcus faecalis. Journal of Biological Chemistry, 249, 1081. CHANARIN, I. & BENNETT, M. (1962) The disposal of small doses of intravenously injected folic acid. BritishJournal of Haernatology, 8, 28.
I. & PERRY,J. (1968) Metabolism of 5CHANARIN, methyltetrahydrofolate in pernicious anaemia. British Journal of Haernatology, 14,297. CHANARIN, I., BELCHER, E.H. & BERRY, V. (1963) The utilization of tritium-labelled folic acid in megaloblastic anaemia. BritishJournal ofHaematology, 9,456. CHANARIN, I., MOLLIN, D.L. & ANDERSON, B.B. (1958) The clearance from the plasma of folic acid injected intravenously in normal subjects with megaloblastic anaemia. British Journal of Haernatology, 4, 435.
CHANARIN, I., DACIE,J.V. & MOLLIN,D.L. (1959) Folic-acid deficiency in haemolytic anaemia. British Journal of Haematology, 5, 245. CHANARIN, I., BENNETT,M.C. & BERRY,V. (1962) Urinary excretion of histidine derivatives in megaloblastic anaemia and other conditions and a comparison with the folic acid clearance test. Journal of Clinical Pathology, 15, 269.
500
F. Lane et a1
CHERRICK, G.R., BAKER, H., FRANK, 0.& LEEW,C.M. (1965) Observations on hepatic avidity for folate in Laennec's cirrhosis. Journal of Laboratory and Clinical Medicine, 66, 446. CORROCHER, R. & HOFFBRAND, A.V. (1972) Subcellular localization of folate and effect of methotrexate on the incorporation of radioactive folic acid into guinea-pig liver folate. Clinical Science, 43, 81 5. DAS, K.C. & HOFFBRAND, A.V. (1970) Studies of folate uptake by phytohaemagglutinin-stimulated lymphocytes. BritishJournal ofHaematology, 19. 203. EICHNER, E.R. & HILLMAN, R.S. (1971) The evolution of anemia in alcoholic patients. American Journal of Medicine, 50, 218. EICHNER,E.R. & HILLMAN,R.S. (1973) Effect of alcohol on serum folate level. Journal of Clinical Investigation, 52, 584. EICHNER, E.R., BUERGEL, N. & HILLMAN, R.S.(1g71a) Experience with an appetizing, high protein, low folate diet in man. American Jownal of Clinical Nutrition, 24. 1337. EICHNER, E.R., PIERCE,I. & HILLMAN, R.S. (1g71b) Folate balance in dietary-induced megaloblastic anemia. New EnglandJournal of Medicine, 284, 93 3. GIRDWOOD, R.H. & DELAMORE, I.W. (1961) Observations on tests of folic acid absorption and clearance. Scottish Medical Journal, 6 , 4. GORESKY, C.A., WATANABE, H. &JOHNS,D.G. (1963) The renal excretion of folk acid. Journal of Clinical Investigation, 42, 1841. HERBERT,V. (1962) Experimental nutritional folate deficiency in man. Transactions of the Association of American Physicians, 75, 307. HERBERT, V. (1966) Aseptic addition method for Lactobacillrrs casei assay of folate activity in human serum. Jotrmal of Clinical Pathology, 19, 12. HERBERT, V. & ZALUSKY, R. (1962) Interrelations of vitamin BIZ and folic acid metabolism: folic acid clearance studies.Journal of Clinical Investigation, 41, 1263.
HILLMAN,R.S., OAKES,M. & FINHOLT.C. (1969) Hemoglobin-coated charcoal radioassay for serum vitamin Biz. A simple modification to improve intrinsic factor reliability. Blood, 34, 385. HOGAN, J.A., MANIATIS, A. & MALONEY, W.C. (1964) The serum Lactobacillus casei folate clearance test in various hematologic disorders. Blood, 24, I 87. JOHNS,D.G. & PLENDERLETH, I.H. (1963) Folk aciddisplacement in man. Biochemical Pharmacology, 12, 1071.
JOHNS,D.G., SPERTI,S. & BURGEN, A.S.V. (1961)The
metabolism of tritiated folic acid in man. Joirrnal of Clinical Investigation, 40, 1684. LINDENBAUM, J. & KLIPSTBIN, F.A. (1964) Folic acid clearances and basal serum folate levels in patients with thyroid disease. Journal of Clinical Pathology, 17, 666.
MCGUFFIN, R., GOFF, P. & HILLMAN,R.S.(1975) The effect of diet and alcohol on the development of folate deficiency in the rat. BritishJournal ofHaematology, 31,185. METZ,J., STEVENS, K., KRAVITZ,S. & BRANDT,V. (1961) The plasma clearance of injected doses of folic acid as an index of folic acid deficiency. Journal of Clinical Pathology, 14, 622. MORGAN, E.H. & CARTER, G. (1960) Plasma iron and iron-binding capacity levels in health and disease: with an improved method for estimation of plasma iron concentration and total iron-binding capacity. Australasian Annals qf Medicine, 9, zog. NAHAS,A., NIXON, P.F. & BBRTINO, J.R. (1969) Transport of 5-methyltetrahydrofolate by LIZIO mouse leukemia cells. (Abstract). Federation Proceedings, 28, 389. PAINE,C.J., EICHNER,E.R. & DICKSON,V. (1973) Concordance of radioassay and microbiological assay in the study of the ethanol-induced fall in serum folate level. American Journal of the Medical Sciences, 266, 135. SCHWEIGERT, B.S. (1948) Folic acid metabolism studies. 111. Intravenous administration of pteroylglutamic acid and pteroyltriglutamic acid. Journal of laboratory and Clinical Medicine, 33. 1271. SULLIVAN, L.W. & HERBERT, V. (1964) Suppression of hematopoiesis by ethanol. Journal of Clinical Investigation, 43, 2048. Wu, A., CHANARIN,I., SLAVIN.G. & LEVX, A.J. (197s) Folate deficiency in the alcoholic-its relationship to clinical and haematological abnormalities, liver disease and folate stores. British Journal of Haematolo$)', 2% 469. YOSHINO,T. (1968) The clinical and experimental studies on the metabolism of folic acid using tritiated folic acid. 111. Plasma clearance in man and organ distribution in rat following intravenous administration of tritiated folk acid. Journal of Vitaminology, 14, 49. ZAIL,S., METZ,J. & BRANDT, V. (1963) The plasma clearance of injected doses of folic acid at varying rates of erythropoiesis. South African Jorrrnal of Medical Science, 28, 121.
