[CANCER RESEARCH 37, 1125-1132, April 1977) Production of Formaldehyde from N5-Methyltetrahydrofolate by Normal and Leukemic Leukocytes1

Janet Thorndike and William S. Beck

Department of Medicine, Harvard Medical School, and Hematology Research Laboratory, Massachusetts General Hospital, Boston, Massachusetts 02114

SUMMARY malian tissues (3, 14, 22, 25, 29, 46). It was first considered that the methyl group was transferred intact in analogy with Extracts of human normal and leukemic leukocytes con the methyl transfer known to take place from AdoMet to tam an enzyme that catalyzes a transfer of labeled methyl various amines (2). This early conclusion rested on the carbon from Ns@[i4C]methyltetnahydrofoIate to tryptamine. assumption that radioactive reaction products were N- Evidence is presented that this reaction is not attributable to methyl derivatives of the amine substrates (14, 22). How a methyltransferase but to the following reaction sequence: even, later studies identified the products as the come (a) an oxidation of N5-[―C]methyltetrahydrofolate to N5,N'°- sponding carbolines formed by spontaneous condensation [i4Clmethylenetetrahydrofolate that is catalyzed by N5,N10- of amines with HCHO (4, 17, 28, 29, 36, 42, 46). These data methylenetetrahydnofolate reductase (EC 1.1.1.68); (b) suggested that transfer of the methyl carbon from CH3- spontaneous release of [“C]fonmaldehydefrom N5,N10- H4folate to an aromatic amine occurs in 2 steps: an enzy [“C]methylenetetrahydrofolate; and (C) nonenzymatic con matic conversion pf CH3-H4folate to HCHO, followed by a densation of [‘4C]formaldehydewith tryptamine to form a nonenzymatic reaction of HCHO with amine. radioactive carboline derivative. The occurrence of this se It was then shown that, in assays for Ns,Nio@methyIene@ quence inleukocytesissuggestedby datathatshow that H4folate reductase (EC 1.1.1.68) by the method of Kutzbach the enzyme reaction is strongly stimulated by addition of and Stokstad (20), tissue extracts catalyze a conversion of flavin adenine dinucleotide and that the final product is CH3-H4folate to HCHO; i.e., CH3-H4folate is oxidized in the chromatognaphically identical to the adduct formed in the presence of an electron acceptor to N5,N'°-methylene reaction of [“C]fonmaldehydewith tryptamine. In the ab H4folate, which then dissociates to HCHO and H4folate (9, sence of tnyptamine, a product accumulates that can react 10, 20). This observation is of interest because transfer of with other HCHO acceptors, i.e., f3-phenylethylamine and the methyl carbon from CH3-H4folateto amines is stimulated dimedone; another reaction product is tetnahydnofolate. by electron acceptors such as FAD on menadione (15, 31, Production of formaldehyde is relatively more active in 44). normal lymphocytes than in normal granulocytes, but it is We have been prompted to study this reaction of CH3- even higher in lymphocytes of chronic lymphocytic leuke H4folate in human hematic tissues because of the particular mia. Activity in granulocytes from a subject with chronic importance of tissue folate metabolism and the controven myelocytic leukemia is also elevated but to a lessen extent sies that have surrounded the distinctive cobalamin-de than activity in lymphocytes of chronic lymphocytic leuke pendent reaction, in which the methyl group of CH3-H4folate mia. Formaldehyde production in leukocytes is only slightly is transferred to homocysteine (28, 39). The purposes of this stimulated by addition of various cobalamins, and activity is study are: (a) to determine whether a transfer of the methyl normal in leukocytes from a vitamin B12-deficient patient. carbon of CH3-H4folateto tryptamine occurs in human Ieu We conclude that the system is cobalamin independent. kocytes; (b) to characterize the nature of this reaction ; (c) to Thus, there exists an active pathway from N5-methyltetrahy compare levels of the reaction in leukocytes from normal drofolate to tetrahydrofolate other than the one catalyzed by individuals and subjects with leukemia and other disorders; cobalamin-dependent N5-methyltetnahydrofolate-homocys and (d) to determine to what extent a cobalamin-independ teinemethyltransfenase. ent conversion of CH3-H4folate to H4folate can occur in these cells. INTRODUCTION Transfer of the methyl carbon from CH3-H4folate2to ano MATERIALS AND METHODS matic amines has been demonstrated in a variety of mam Reagents. [‘4C]CH3-H4folate(54@Ci/@tmoIe),[2-'4C]folic I This work was supported by Research Grant CA-03728 from the National acid (58 @Ci/@moIe),and[‘4C]formaldehyde(2 @Ci/@moIe) Cancer Institute, NIH, and by the John Phyffe Richardson Fund. 2 The abbreviations used are: CH3-H4folate, N'.methyltetrahydrofolate; were obtained from Amersham/Searle Corp. , Arlington AdoMet, S-adenosylmethionine; N,N'0-methylene-H4folate, N5,N'0@methyl@ Heights, Ill. [i4C]CHHfolatewasstored undenN2at —20°as enetetrahydrofolate; H4folate, tetrahydrofolate; FAD, flavin adenine dinucleo a 4.5 mM solution in 1 M mercaptoethanol at pH 7. 3',5'- tide; N5-formyl-H4folate, N5-formyltetrahydrofolate; FMN, riboflavin 5'-phos phate; N, N'°-methenyl-H4folate, N5,N' °-methenyltetrahydrofolate ; CLL, [3HJfoIic acid (22 @Ci/@mole)waspurchased from Schwarz/ chronic lymphocytic leukemia; CML, chronic myelocytic leukemia. Ab Mann, Onangeburg, N. Y.; N5-formyl-H4folate, calcium Ieu breviatlons of cofactors are in accordance with standard usage in The covonin, was obtained from Lederle Laboratories, Pearl Journal of Biological Chemistry. Received August 2, 1976; accepted December 29, 1976. Riven, N. V.; tryptamine hydrochloride, N-methyltnyptamine,

APRIL 1977 1125

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. J. Thorndike and W. S. Beck f3-phenylethylamine, AdoMet, FAD, and folic acid were pro disrupted and centrifuged as described for blood leuko duced by Sigma Chemical Co., St. Louis, Mo. FMN, NADH, cytes. and NADPH were obtained from Calbiochem, La Jolla, - Assay of Formaldehyde ProductIon from CH3-H4folate. Calif.; NAD and NADP were from Boehringer/Mannheim, Cell extracts were incubated with [i4C]CH3@H4foIatein the Mannheim, West Germany; cyanocobalamin was from presence of a saturating concentration of tryptammne. Dun Schwarz/Mann. Silica gel and cellulose thin-layer Chroma ing this incubation, enzymatically produced HCHO was gram sheets were purchased from Eastman Kodak Co., converted to the corresponding carboline, which was ex Rochester, N. V. tracted in toluene:isoamyl alcohol (pH 10) (14). Standard Methylcobalamin and propylcobalamin were synthesized incubation mixtures contained (in a total volume of 0.25 ml): by reduction of cyaflocobalamin with sodium borohydride, [14C]CH3-H4folate,9 nmoles; tryptamine, 5 @moles;FAD,0.5 followed by alkylation with methyl iodide and propyl bro @mole;potassium phosphate (pH 6.5), 25 @moles;andcell mide, respectively (32). Dihydrofolate and H4folate were extract,0to0.5mg protein.Afterincubationfor30mm at37° prepared by reduction of folic acid with sodium hydrosulfite in the dark, 0.5 ml of 0.5 M sodium borate (pH 10) was (11) or sodium borohydride (40), respectively. N5,N10-meth added; tryptoline was extracted with 3.5 ml of tolu enyl-H4folate was prepared by boiling H4folate in 80% formic ene:isoamyl alcohol (97:3) (‘‘tolueneextract'‘and‘‘toluene acid:0.1 M mercaptoethanol for 5 mm (37) on by treat extractible―are used in subsequent discussion to refer to ment of N5-formyl-H4folate with HCI (33). N5,N'°-methenyl this extraction). After centnifugation for 5 mm at 400 x g, [3H]H4folate, prepared by reduction of [3H]folic acid with aliquots of the toluene extract were evaporated to dryness sodium borohydride (40) and boiling with formic acid (37), and taken up in 1 ml isoamyl alcohol. Radioactivity was was purified by chromatography on cellulose thin-layer assayed in Aquasol liquid scintillation fluid in a Beckman sheets in 1 N formic acid:0.2 M mercaptoethanol, followed LS-150 scintillation counter. Incubations lacking enzyme by elution with 0.1 M mercaptoethanol. Unlabeled N5,N'°- were used as controls. Enzyme activity was expressed as methenyl-H4folate (which gives a fluorescent spot at AF nmoles HCHO formed pen hr pen mg protein. 0.24) was used as reference standard. N5,N10-['4Cjmethyl Chromatographlc Characterization of Reaction Prod ene-H4folate was prepared from [14C]HCHO and H4folate ucts. Chromatognaphic methods differed depending on (18). CH3-[―C]H4folate was synthesized by reduction of whether substrates were [“C]CH3-H4folateon CH3-[2- [2-―C]folicacid with sodium borohydnide and methylation “C)H4folate.Whenthe substrate was [‘4C]CH3-H4folatein with unlabeled HCHO in the presence of borohydride standard incubation mixtures, aliquots of toluene extracts (8); samples were purified by chromatography on cellulose were applied to a silica gel thin-layer chromatography sheet thin-layer sheets in 0.1 M potassium phosphate:0.2 M men and developed in acetone:1 N ammonium hydroxide (10:3) captoethanol, eluted with 0.2 M mercaptoethanol, and (16). In a separate reference mixture, 0.05 ml of a 40 @M evaporated to a smaller volume under N2. [@C]CH3-H4folate solution of [14C]formaldehyde was mixed with 0.2 ml of 125 was used as reference standard. mM tryptamine and 0.5 ml of 0.5 M sodium borate (pH 10) PreparatIon of Lymphocytes and Granulocytes. Leuko and extracted with 3.5 ml of toluene:isoamyl alcohol (97:3). cytes were obtained from venous blood (10 to 60 ml) of Aliquots of this toluene extract were chromatographed on ostensibly normal subjects and patients with the following the same sheet. A reference sample of nonnadioactive N- various disorders: CLL, CML, unexplained monocytosis, methyltryptamine was also developed on the same sheet. and severe cobalamin deficiency (mean corpuscular vol Sheets were dried, examined under UV, cut into portions, ume, 117 cu @m;serumcobalamin, 20 pg/mI). Some of the and assayed for radioactivity in Aquasol liquid scintillation leukemic subjects were receiving prednisone and cyclo fluid. phosphamide (CLL) or busulfan (CML); none was receiving However, when the substrate was CH3-[2-'4C]H4folate,be folic acid. fore the addition of borate mixtures were boiled in 80% After sedimentation of erythrocytes for 1 hr at 37°follow formic acid:0.1 M mercaptoethanol for 5 mm. Aliquots were ing the addition of 0.1 volume of 6% dextran (41), platelet mixed with 24 nmoles of N5,N10-methenyl-[3H]H4folate, ap free leukocytes were isolated by centrifugation of the super plied to a cellulose thin-layer sheet, and chromatographed natant fraction at 140 x g for 10 mm. Lymphocytes and in 1 M formic acid:0.2 M mencaptoethanol. The sheet was granulocytes were purified from normal blood by fractiona dried, examined under UV, cut into portions, and eluted tion of defibrinated blood on a Ficoll-Hypaque gradient (1, with water. Eluates were centrifuged at 400 x g for 5 mm; 12). Residual erythrocytes were removed from leukocyte supernatant fractions were assayed for radioactivity with preparations by hypotonic hemolysis (7, 32). Cells were appropriate corrections for quenching of radioactivity by washed 3 times in 0.85% NaCI solution:10 mM phosphate water. The absorption spectra of eluates were determined (pH 7.4), sonically disrupted in 20 mM potassium phosphate with a Carey Model 15 recording spectrophotometen. (pH 7.0) (32), and centrifuged at 27,000 x g for 30 mm. The supennatant fraction was used in enzyme assays. Protein was assayed by the method of Lowry et al. (27) with bovine RESULTS serum albuminas a standard. Bone marrow obtained by aspiration was diluted in IdentIficatIon of Reaction Products. Extracts of various Ringer's solution containing 0.2% sodium lactate and hepa types of leukocytes were found to catalyze formation of a nfl (1 unit/mI). Cells were washed in 0.85% NaCI solution:10 radioactiveproductfrom[“C]CH3-H4folateandtryptamine, mM phosphate, pH 7.4; residual erythrocytes were removed according to the procedure originally proposed as an assay by hypotonic hemolysis(7). Remaining cells were sonically for aromatic alkylammne N-methyltransferase (14). Before

1126 CANCER RESEARCH VOL. 37

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. Formaldehyde Production by Leukocytes

attempting quantitative comparisons of various cell types, ing a plateau at 20 to 30 mM at which 75% of the [14C]HCHO we undertook to document the nature of the observed neac had reacted (Chart 1). For investigation of the acceptor tion with extracts of CLL cells as a convenient enzyme specificity of the reaction of [14C]HCHO with tryptamine, source. The following experiments were intended to show [14C]HCHO was incubated with other potential acceptors. that such extracts catalyze the conversion of [“C]CH3-With f3-phenylethylamine a toluene-extractible prod uct was H4folate to [‘4C]HCHO,whichthen condenses nonenzymati obtained , presumably methylene-f3-phenylethylamine (17), cally with tryptamine to form a toluene-extractible adduct, the formation of which was also influenced by acceptor Ei4Cltnyptolmne. concentration; however, the saturating concentration was It appeared that, if the reaction product were in fact above 60 mM at which solubility became limiting. A reaction [‘4C]HCHO,thenomission of tnyptammnefrom the incuba of [14C]HCHOwith H4folate was suggested by a decrease in tion mixture should lead to accumulation of a product that the radioactivity available to react with tryptamine. Fonma would react with subsequently added tryptamine or with the tion of this product, presumably N5,N10-methylene-H4folate, HCHO acceptor, dimedone (24). Results of such an expeni increased with the concentration of added H4folate; the ment, summarized in Table 1, show that a CLL cell extract concentration curve resembled that obtained when the ac did catalyze formation of a product that could react with cepton was tryptamine (Chart 1). tryptamine on dimedone. Dimedone was in fact a more For investigation of the fate of the folic acid moiety after efficient HCHO acceptor. release of the 1-carbon moiety, CLL cell extract was incu For determination of whether tryptoline formation is due bated in a mixture containingCH3-H4folatelabeledin the to spontaneous condensation of the enzymatic reaction ptenidine ring (CH3-[2-―C]H4folate)assubstrate. For protec product with tryptammne, we chnomatognaphed toluene ex tion of reduced folates from oxidation, incubations were tracts on silica gel thin-layer sheets to separate tryptoline conducted under N2. For determination of whether [“C] from N-methyltnyptamine, the product to be expected of a H4folate had been released from CH3-[―C]H4folate,the methyl transfer reaction (16). A radioactive peak was ob incubation mixture was boiled in 80% formic acid, which served at the same position (RF0.92) as that of the product converts H4folate to N5,N'°-methenyl-H4folate (37). When obtained when a mixture of [14CJHCHOand tryptammnewas chromatognaphed on cellulose thin-layer sheets in 1 N extracted with toluene in the presence of borate (pH 10) and formic acid:0.2 M mencaptoethanol, reference preparations similarly chromatographed . A simultaneously chromato of N5,N10-methenyl-H4folate, which were synthesized from graphed reference sample of N-methyltnyptamine had a mo formic acid and H4folate on from N5-fonmyl-H4folate, gave a bility of RF0.83. Thus, the radioactive product appeared to fluorescent spot at RF0.24, with the characteristic absonp result from a condensation of [14C]HCHO with tryptamine tion maximum at 350 mm (37) when eluted with 0.1 N HCI. rather than from a methyl transfer reaction. CH3-[―C]H4folateboiled in formic acid formed a radioactive The reaction was shown to be dependent on tryptamine; peak at RF0.78. Other folate derivatives treated in this way activity was decreased by more than 90% in the absence of did not migrate to RF 0.24. N5,N10-[―C]Methylene-H4folate tryptamine. Similarly, production of [14C]tryptolmne from gave a radioactive peak at RF0.87; folic acid and dihydnofol [14C]HCHO incubated with tnyptammne (in the absence of enzyme) was a function of tnyptamine concentration, reach

Table 1 100 Demonstrationoffrom(1°C]CH@-H4folate enzymaticproduction of (“C)formaldehyde extractResults by CLL cell @: 80 are expressedas radioactivity (cpm) extracted by tolu ene:isoamylstandardincubation alcohol after incubation of a CLL extract in a mixture containing [14C)CH3-H4folate but lacking trypt 60 amine.wasdetermined At the end of the incubation, accumulated [‘4C]HCHO (pH10), in 2 ways:(a) adding 0.5 ml of 0.5 Msodium borate oftoluene:isoamyl0.12ml of 125mMtryptamine(omitted in control), and4.5ml 40 alcohol (97:3); extracting; and assaying for radio @ activity0.4ml as described in “MaterialsandMethods―;or (b) adding 20 for5mmof0.015 Mdimedonein 1 Nsodiumacetate(pH4.5), heating assayingforat 95°,cooling, extracting with 4 ml of toluene, and radioactivity (20).Postincubation 0 0 20 40 60 80 additionsDuration CONCENTRATIONOFACCEPTOR(mM)

Sodiumof Chart 1. Reaction of various acceptors with [‘4C)HCHO.Asample of incu-boratebation [‘4C]HCHO(20nmoles) was incubated with 25 @.cmoIespotassiumphosphate Sodium and trypt- Dime buffer (pH 6.5) in a volume of 0.25 ml at 37°.Varying amounts (1.2 to 19 (mm)done1 Incubation mixture borate amine @.cmoles)oftryptamine (•)or p-phenylethylamine ( 0) were added. After incubation for 30 mm, 0.5 ml of 0.5 N sodium borate (pH 10) and 3.5 ml of 5 Complete54,603Minus 654 21 826 toluene:isoamyl alcohol (97:3) were added. Aliquots of toluene extract were 25,440Differenceenzyme 42 3,416 then assayed for radioactivity. Ordinate, calculated percentage of total [“C) 29,16330 612 18,410 HCHO that had reacted with the amine. In 1 experiment 0.25 to 20 @@moIes of H4folate were added in place of an acceptor (tx). After incubation for 30 Complete85,563Minusenzyme 1,200 46,238 mm, 0.8 ml of 0.5 M borate (pH 10), 0.12 ml of 125 mM tryptamine, and 6.0 ml 28,110Difference 126 4,508 of toluene:isoamyl alcohol were added. Aliquots of toluene extract were then assayed for radioactivity; the percentage of [“C]HCHOreactedwas calcu 1,074 41,730 57,453 lated from decrease in toluene-extractible radioactivity.

APRIL 1977 1127

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. J. Thorndike and W. S. Beck ate gave UV-absorbing spots at the origin and AF 0.65, ityplussomedegradationtoafasten-movingcompound. respectively; untreated H4folate gave an UV-absonbing spot After correction for incomplete recovery of methenyl [email protected]. [3H]H4folate, it is possible to estimate that about 1 nmole of A CLL cell extract was incubated with CH3-[14C]H4folate H4folate had been produced from CH3-H4folate for 2.6 and FAD at 37°(Chart 2). After 1 hr, tryptamine was added, nmoles of [‘4C]HCHOproduced from [14C]CH3-H4folate.It mixtureswere boiledin formicacid,and aliquotswere was also observed that, after [3H]H4folatehad reacted with chromatographed . N5,N10-Methenyl-[3H]H4folate was added formic acid, only about one-half of the radioactivity was as an internal standard. The results showed that the bulk of recovered before chromatography. This connection brings the ‘4Cradioactivity formed a peak corresponding to un the observed molar ratio closer to 1:1. reacted CH3-[14C]H4folatewith a trace at RF0.24. Most of the Requirements and Kinetic Properties of Enzyme Reac 3H radioactivity (from the N5,N1°-methenyl-[3H]H4folate tion. The foregoing experiments showed that [‘4C]CH3- marker) formed a peak that coincided with the fluorescent H4folate is converted to [14C]HCHO and H4folate. Such a spot at [email protected]. After incubation with the cell extract, 14C reaction would require an oxidative step; therefore, we in radioactivity increased at RF0.24. Incubation mixtures that vestigated the effects of various electron acceptors and had not been boiled with formic acid gave no 14Cradioactiv other cofactons on the production of [14C]HCHO from ity at AF0.24. In all cases no radioactivity was found at the [14C]CH3-H4folate by leukocyte preparations. The results, origin. summarized in Table 2, show that added FAD increased These results show that the appearance of methenyl activity 6-fold. Added FMN, menadione, NADP, methylco [‘4C]H4folatewasdue to an enzyme-dependent formation of balamin, and propylcobalamin were stimulatory only to a H4folate from CH3-H4folate. The initially added ‘4Cradioac minor extent. Cyanobalammnhad no significant effect. In the tivity was recovered quantitatively (>95% of it as CH3- presence of methylcobalamin or menadione, the stimula [14C]H4folate). However, recovery of methenyl-[3H]H4folate tory effect of FAD was lessened considerably. NADPH and at AF0.24 was only partial and amounted to only 43% of the metabisulfite were inhibitory; NAD, NADH, AdoMet, anaeno amount added initially. This reflects total loss of madioactiv biosis, and addition of a boiled extract of mixed normal leukocytes were without significant effects. Dependence of the enzyme reaction on an electron ac cepton such as FAD suggests that in leukocytes [14C]HCHO 3H 800

Table 2 activityEnzymeEffect of added cofactors on enzyme 400 beendialyzedactivity was assayed in a CLL extract that had mMpotassiumfor 6 hr (with 2 changes)against 100 volumes of 20 percentageofphosphate,pH 7. Activity is expressedas a activity observed in a control incubation lacking added cofac tons. When a sequenceof figures is reported, each figure repre 0 sents relative activity at the concentration denoted by the corre spondingfigurewas0.43 in the left column. Actual activity in the control nmole HCHOper hr per mgprotein.Concentrations L activityAdditions Relative

L control)None, (mM) (% of 100Cyanocobalamincontrol 129Methylcobalamin 0.1, 0.4 111, 195Propylcobalamin 0.1, 0.4 221, @ 800 182NAD 0.1, 0.4 154, 122NADH 0.02 123NADP 0.2 132NADPH 0.02 10FAD 0.01,0.02,0.04 49,28, 200 718FMN 0.2, 1.0, 2.0 294, 628, 161AdoMet 0.2 124Potassium 0.05,0.2 119, 60EDTA metabisulfite 1.0 0 100Menadione 5.0 o_ 2 4 6 8 10 12 14 182Potassium 3.6 DISTANCE FROM OR/GIN , cm 57Plus metabisulfite 1.0 5.0MethylcobalaminEDTA Chart 2. Evidence that CLL extract catalyzes formation of H4folate from 239Plus 0.1 CH3H4folate.After incubation of CLL extract with CH3-[―CjH4folateandFA@ in the absence of tryptamine for 1 hr, 0.08 ml of 125 mu tryptamine was 2.0MenadioneFAD added. Addition of borate was omitted, and mixtures were boiled for 5 mm in 173PIusFAD 3.6 80% formic acid:0.1 M mercaptoethanol (37). Aliquots were mixed with 24 1.0Boiled nmoles of N,N10.methenyl-[3.HjH4folate and chromatographed on cellulose 97granulocytes°Anaerobiosis―extract of normal thin-layer sheets in 1 N formic acid:0.2 M mercaptoethanol. Sheets were examined under uv and eluted with water. Eluates were assayed for “C(0), 101 and 3H(•).A,complete incubation; B, position of fluorescent spot observed in complete and control incubations; C, control incubation in which enzyme a Boiled 27,000 x g supernatant fraction (0.3 mg). was added after incubation (immediately before boiling). b Incubated under N2.

1128 CANCERRESEARCHVOL. 37

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. Formaldehyde Production by Leukocytes

prod uction from [1‘C]CH3-H4folateinvolvesa reductase-cat alyzed oxidation of [14C]CH3-H4folate to N5,N10- [14C]methylene-H4folate. This spontaneous dissociation of N5,N'°-['4C]me@hyIene-H4folateto [‘4C]HCHOandH4folate presumably follows. The effect of FAD concentration was further studied in extracts from CLL lymphocytes (Chart 3A). Activity was strongly stimulated by FAD addition. The degree of stimula tion increased with the concentration of added cofactor and reached a plateau at concentrations of 0.8 to 2.0 mM. Simi lar results were obtained with extracts of other types of leukocytes. Enzyme activity in CLL cell extracts, assayed in the pres @ 0 0 ‘ ‘ 0 once of 1.5 mM FAD, was found to be influenced by the 5 6 7 8 9 0 20 40 60 00204060810 concentration of added CH3-H4folate with an activity maxi pH TIME (nun) PROTEIN (mgi Chart 4. Properties of HCHO-producing enzyme system in CLL extracts. mum at 30 to 40 j.@M(Chart 3B). By Lineweaven-Burk plot A, activity as a function of pH. Standard assay method was used except that (26), the Km for CH3-H4folate was 1 .1 x 10@ M. The rate of the pH was varied by replacement of phosphate buffer with sodium acetate tryptoline formation was a function of tryptamine concen below pH 6.5 and sodium glycinate above pH 8. B, time course of HCHO production from CH3-H4folate. CLL extract (0.5 mg) was incubated in stan tration with an activity maximum at 10 to 20 mM (Chart 3C). dard assay mixture ( 0) and in mixture lacking added FAD (•)for60 mm. C, In the absence of added tnyptamine, activity was less than effect of enzyme concentration (0 to 1.0 mg CLL extract protein) in standard 10% of that observed with 10 mr,i tryptamine. assay with ( 0) and without (•)addedFAD. A pH-activity curve (Chart 4A) revealed maximal activity at pH 6.5. HCHO production was also proportional to incuba centrifuged at 27,000 x g , 95% of the activity was recovered tion time between 0 and 60 mm (Chart 4B) and to the in the supernatant fraction. Activity was unaffected by ston amount of extract added to the assay mixture between 0.1 age at —50°forat least 30 days or by dialysis [against 20 mM and 0.5 mg protein (Chart 4C). When sonic extracts were potassium phosphate (pH 7)]. Similar conclusions were reached when CLL extracts were assayed in the absence of added FAD. When 20 m@if3-phenylethylammne was used as A HCHO acceptor in place of tryptammne, activity was consist ently lower, but kinetic properties were unaltered. 2 Comparison of Enzyme Activity in Leukocytes of Various Types. Table 3 summarizes a comparative study of enzyme

I activity (expressed as nmoles HCHO formed per hr pen mg protein) as assayed under standard conditions. Assays were also performed in which addition of tryptamine was do @ I I 1 I 0 0.2 0.6 1.0 20 ferred until after the end of the incubation and addition of PAL? (mM) borate. All activity was about 3 times higher when tnypt amine was added after the incubation rather than at the beginning. This appears to indicate that tryptamine exerts an inhibitory effect on HCHO production even though it 2@• serves as an HCHO acceptor. The enhancement caused by late tryptamine addition was similar for all cell types stud ied. The results in Table 3 show that mean activity in leuko cytes from CLL patients was considerably higher than that found in mixed leukocytes from normal subjects and was more than twice that found in purified normal lymphocytes. CHJ-FH@(,iM) Activity in granulocytes was relatively low. These differ ences are somewhat diminished if activity is expressed pen cell instead of per mg protein, because protein content is higher in granulocytes than in lymphocytes (5). Expeni ments in which extracts of normal lymphocytes and granu locytes were mixed together gave no evidence of the pres ence of enzyme activators or inhibitors. The results in Table 3 also show that activity in CML leukocytes was higher than that in isolated normal granulo 00 ‘ @0 ‘ 20 ‘ 30 cytes. However, extracts prepared from the bone marrow of TRYPTAM/NE(mM) a CML patient had lower activity. Activity in mononuclear Chart 3. Effect of concentration of various additions on HCHO production cells isolated on a Ficoll-Hypaque gradient from the blood by extracts from CLL lymphocytes. Standard enzyme assays were performed except that varying amounts of the following substances were added: A, of a patient with monocytosis (75%) approximated that FAD; B, CHrH4folate (CH3-FH4);C, tryptamine. found in normal lymphocytes. Activity in the mixed leuko

APRIL 1977 1129

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. J. Thorndike and W. S. Beck

Table3 variousleukocytesCellEnzymeactivity in extractsof activity (nmoles HCHO protein)Tryptamineformed/hr/mg

latebMixed type(s)DiagnosisSubjectsEnzyme early° Tryptamine ±0.31b 2.68 ±1.04 Mixed CML 2 0.71,1.12 1.81,2.86 Mixed Cobalamindeficiency 1 0.63 @ Lymphocytes Cobalamin deficiency 1 0.60 Lymphocytes Normal 4 0.96 ±0.34 3.08 ±1.09 Lymphocytes CLL 7 2.10 ±0.60 6.40 ±1.83 Granulocytes Normal 5 0.35 ±0.09 0.86 ±0.22 Mixed Monocytosis 1 1.10 Mononuclearcells Monocytosis 1 1.20 Mixed marrow cellsNormal CML5 10.80 0.15

a Tryptamine was either added at the beginning of the incubation (per the standard assay) or omitted during the incubation and added after addition of borate and toluene:isoamyl alcohol. b Mean ± S.D. given when more than 2 extracts were assayed.

cytes of a severely cobalamin-deficient patient was not no HCHO production by leukocyte extracts. The extent of inhi tably different from that of a normal patient. bition by this compound may depend on the method of No activity was found in plasma from hepaninized normal enzyme preparation, as has been shown in liven extracts blood. ABC disrupted by sonic oscillation showed some (20). The noteworthy enhancement in activity by late addi activity (10 pmoles/min/ml of ABC) only after dialysis. This tion of tnyptammneresembles the observations of Stebbins et value is insignificantly low when related to cellular protein. al.(42)inplatelets.Wecannotaccountforthe inhibitory effect of tryptamine during the incubation. The results furnish another example of an enzyme that is DISCUSSION relatively more active in lymphocytes than in gnanulocytes and that is higher in CLL lymphocytes than in normal lym The resultsareof interestfromseveralpointsofview. phocytes. Indeed, the comparative activity pattern closely First, they demonstrate that leukocytes contain an enzyme resembles that described for CH3-H4folate-homocysteine that causes the conversion of CH3-H4folate to HCHO and methyltnansferase (32). Having earlier concluded that activ H4folate. Activity levels in leukocytes from normal blood are ity levels of the latter enzyme can be correlated with the similar to values observed in some mammalian tissues, e.g., proliferative potential of a particular cell line (32, 39), we are brain (44), kidney (23), and platelets (42), but are lower than led to a similar conclusion for the HCHO-producing enzyme those in liver (20, 21) and various cultured mammalian cells system in leukocytes. It is of interest to compare the activity (35, 43). The data show that in leukocytes, as in other levels of the 2 enzymes in various leukocyte types. When tissues, transfer of methyl carbon from CH3-H4folateto var activities of HCHO-pnoducing enzyme, i.e., N5,N'°-methyl ious amines is not due to a methyltransferase reaction. It ene-H4folate neductase, and CH3-H4folate-homocysteine appears rather to be due to an enzymatic oxidation of CH3- methyltransferase (32) are both expressed in terms of H4folate to N5,N'°-methylene-H4folate that is catalyzed by nmoles of CH3-H4folateutilized per hr pen mg, respective N5,N'°-methylene-H4folatereductase.The immediate oxida levels are: in CLL lymphocytes, 2.10 and 2.15; in normal tion product dissociates to H4folate and HCHO, although lymphocytes 0.96 and 0.95; in normal granulocytes, 0.35 the equilibrium constant is not in favor of dissociation in the and 0.15. It may be inferred that the enzymatic machinery of absence of a HCHO acceptor (19). Liberated HCHO then folate metabolism is elevated in toto in CLL lymphocytes reacts nonenzymatically with an acceptor such as trypt and other cells with high proliferative potential. amine to form an adduct. We have shown in leukocytes that Finally, the results are of interest in the context of a long free HCHO arises from CH3-H4folate in a reaction sequence standing controversy oven the validity of the so-called that is stimulated by FAD and, hence, must include an “methylfolatetrap―theory of vitamin B12action in animal oxidative step. Other electron acceptors, i.e., FMN, menadi cells (13, 30); i.e., cobalamin participates in the biosyn one, and NADP, are less stimulatony. It is unclear why stimu thesis of DNA by promotion of activity of the CH3-H4folate Iation by FAD is mitigated by the simultaneous addition of homocysteine methyltransferase reaction. This in turn pen menadione. The reducing cofactor, NADPH, is inhibitory. mits conversion of CH3-H4folateto H4folate, which can serve Presumably, such inhibition is due to the ability of NADPH as a direct precursor of N5,N'°-methylene-H4folate, a coen to force the equilibrium of the N5,N'°-methylene-H4folate zyme of thymidylate synthetase. In this view methylcobala reductasereactioninthe reductivedirection(21,31).In mm mediates the “escape―ofCH3-H4folatefrom the “trap― contrast, NADP is slightly stimulatory. Because enzyme ac in which it is said to be sequestered. An essential prenequi tivity is not depressed by NADH, leukocyte enzyme presum site to this theory has been the assumption that no cobala ably has a higher affinity for NADPH, as is the case with pig mm-independent avenue exists in man for the conversion of liver N5,N'°-methylene-H4folate neductase (21). CH3-H4folate to H4folate. Although other evidence has An inhibitory effect of AdoMet on reductase activity has placed the methylfolate trap theory in question (6), we con been reported in liver (20, 31), but AdoMet had no effect on sider that the present data, obtained in hematic tissues,

1130 CANCERRESEARCHVOL. 37

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. Formaldehyde Production by Leukocytes

establish that H4folate can anise freely from CH:tH4folate 14. Hsu, L. L., and Mandell, A. J. Multiple N-Methyltransferases for Aromatic Alkylamines in Brain. Life Sci., 13: 847-858, 1973. without the intervention of a cobalamin-dependent methyl 15. Hsu, L. L., and Mandell, A. J. Stimulation of Brain Aromatic Alkylamine transferase. Similar observations have been reported in N-Methyltransferase Activity by FAD and Methylcobalamin. Life Sci., 14: platelets (42). Further, we find that HCHO production by 877-885, 1974. 16. Hsu, L. L., and Mandell, A. J. Enzymatic Formation of Tetrahydro-@3- leukocytes is close to normal in a vitamin B12-deficient pa Carboline from Tryptamine and 5-Methyltetrahydrofolic Acid in Rat Brain tient and that addition of various cobalamins in vitro has Fractions: Regional and Subcellular Distribution. J. Neurochem. , 24: little influence on the reaction. 631-636, 1975. 17. Hsu, L. L., and Mandell, A. J. Methylene-fJ-Phenylethylamine Formation These results are also of interest in light of recent evi from 5-Methyltetrahydrofolic Acid and f3-Phenylethylamine. Life Sci., 17: dence that shows diverse effects of vitamin B,2 deficiency 387-396,1975. 18. Huennekens, F. M. Preparation and Properties of “ActiveFormalde on the activities of other folate-linked enzymes. It has been hyde―and“ActiveFormate.―Methods Enzymol., 6: 806-811 , 1963. found, for example, that conversion in vivo of [‘4C]formate 19. KalIen, R. G. Tetrahydrofolic Acid and Formaldehyde. Methods En to [‘4C]senineisdepressed in vitamin B,2-deficient leuko zymol.,18B: 705-716, 1971. 20. Kutzbach, C., and Stokstad, E. L. R. Feedback Inhibition of Methylene cytes (34) and that thymidylate synthetase activity rises in Tetrahydrofolate Reductase in Rat Liver by S-Adenosylmethionine. extracts of vitamin B12-deficient bone marrow (38). Presum Biochim. Biophys. Acta, 139: 217-220, 1967. ably, these effects are related in some way to repression on 21. Kutzbach, C., and Stokstad, E. L. R. Mammalian Methylenetetrahydrofo late Reductase. Partial Purification, Properties, and Inhibition by 5- denepression of various enzymes in the pathway of DNA Adenosylmethionine. Biochim. Biophys. Acta, 250: 459-477, 1971. synthesis, as an indirect result of vitamin B,2 deficiency. 22. Laduron, P. N-Methylation of Dopamine to Epinine in Brain Tissue Using N-Methyltetrahydrofolic Acid as the Methyl Donor. Nature New Biol., The physiological role of N5,N10-methylene-H4folate me 238: 212-213, 1972. ductase activity, as it expresses itself when assayed with 23. Laduron, P. M., Verwimp, M. F., Janssen, P. F. M., and Gommeren, W. CH3-H4folateand tryptamine, was originally believed related R. Tissue Fractionation in Rat Brain, Kidney and Liver. I. Intracellular Localization of a 5-Methyltetrahydrofolic Acid Requiring Enzyme. Bio to the metabolism of aromatic amines, particularly in brain chimie, 57: 253-260, 1975. tissue (3, 14, 22). However, little is known of the natural 24. Leysen, J. , and Laduron, P. Characterization of an Enzyme Yielding occurrence or function of tryptolines on other HCHO-ano Formaldehyde from 5-Methyltetrahydrofolic Acid. Federation European Biochem. Soc. Letters, 47: 299-303, 1974. matic amine condensation products. The nonspecificity of 25. Lin, R. L., and Narasimhachari, N. Evidence for the Absence of Amine-N- HCHO binding and the wide variety of possible HCHO ac methylation and 0-methylation in Indolethylamines with Methyltetrahy drofolic Acid-Dependent N-Methyltransferase. Res. Commun. Chem. Pa ceptons make it difficult to ascertain the actual in vivo role of thol. Pharmacol., 8: 535-542, 1974. the HCHO-pnoducing reaction in leukocytes. We are in 26. Lineweaver, H.. and Burk, D. The Determination of Enzyme Dissociation tnigued by evidence suggesting that enzymatically pro Constants. J. Am. Chem. Soc., 56: 658-666, 1934. 27. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. Protein duced HCHO can react with H4folate itself. The extent to Measurements with the Folin Phenol Reagent. J. Biol. Chem. , 193: 265- which such a train of events may participate in folate metab 275,1951. olism in vivo remains to be explored. 28. Mandel, L. R., Rosegay, A., Walker, R. W., VandenHeuvel, W. J. A., and Rokach, J. 5-Methyltetrahydrofolic Acid as a Mediator in the Formation of Pyridoindoles. Science, 186: 741-743, 1974. 29. MelIer, E., Rosengarten, H., Friedhoff, A. J. , Stebbins, R. D., and Silber, REFERENCES R. 5-Methyltetrahydrofolic Acid Is Not a Methyl Donor for Biogenic Amines: Enzymatic Formation of Formaldehyde. Science, 187: 171-173, 1. Aisenberg, A. C., and Bloch, K. J. Immunoglobulins on the Surface of 1975. Neoplastic Lymphocytes. New EngI. J. Med., 287: 272-276, 1972. 30. Noronha, J. M., and Silverman, M. On Folic Acid, Vitamin B,2, Methio 2. Axelrod, J. The Enzymatic N-Methylation of Serotonin and Other nine and Formiminoglutamic Acid Metabolism. In: H. C. Heinrich (ed), Amines. J. Pharmacol. Exptl. Therap., 138: 28-33, 1962. Vitamin B,2 und Intrinsic Factor, 2d EuropSisches Symposion, pp. 728— 3. Banerjee, S. P., and Snyder, S. H. Methyltetrahydrofolic Acid Mediates 736.Stuttgart:FerdinandEnkeVerlag,1972. N- and 0-Methylation of Biogenic Amines. Science, 182: 74-75, 1973. 31. Pearson, A. G. M., and Turner, A. J. Folate-dependent One-Carbon 4. Barchas, J. D., Elliott, G. R., DoAmaral, J., Erdelyi, E., O'Connor, S., Transfer to Biofenic Amines Mediated by Methylenetetrahydrofolate Re Bowden, M., Brodie, H., Berger, P. A., Renson, J. , Wyatt, R. J. Trypto ductase. Nature, 258: 173-174, 1975. lines. Formation from Tryptamines and 5-MTHF by Human Platelets. 32. Peytremann, R., Thorndike, J., and Beck, W. S. Studies on N5-Methyl Arch. Gen. Psychiat., 31: 863-867, 1974. tetrahydrofolate-Homocysteine Methyltransferase in Normal and Leu 5. Beck, W. S. The Control of Leukocyte Glycolysis. J. Biol. Chem., 232: kemic Leukocytes. J. Clin. Invest., 56: 1293-1301, 1975. 251-270,1958. 33. Rabinovitz, J. C. Preparation and Properties of 5,10-Methenyltetrahydro 6. Beck, W. S. Metabolic Features of Cobalamin Deficiency in Man. In: B. folic Acid and lO-Formyltetrahydrofolic Acid. Methods Enzymol., 6: 814- M. Babior (ed), Cobalamin—Biochemistry and Pathophysiology, pp. 815,1963. 403-450. New York: John Wiley & Sons, Inc., 1975. 34. Robertson, J. S., Hsia, V. E., and Scully, K. J. Defective Leukocyte 7. Bertino, J. R., Silber, R., Freeman, M., Alenty, A. , Albrecht, M., Gabrio, Metabolism in Human Cobalamin Deficiency: Impaired Propionate Oxi B. W. , and Huennekens, F. M. Studies on Normal and Leukemic Leuko dation and Serine Biosynthesis Reversible by Cyanocobalamin Therapy. cytes. IV. Tetrahydrofolate-dependent Enzyme Systems and Dihydrofolic J. Lab. Clin. Med., 87: 89-97, 1976. Reductase. J. Clin. Invest., 42: 1899-1907, 1963. 35. Rosenblatt, D. S., and Erbe, R. W. Reciprocal Changes in the Levels of 8. Blair, J. A. , and Saunders, K. J. A Convenient Method for the Preparation Functionally Related Folate Enzymes during the Culture Cycle in Human of dl-5-Methyltetrahydrofolic Acid (dl-5-methyl-5,6,7,8-tetrahydropteroyl Fibroblasts. Biochem. Biophys. Res. Commun., 54: 1627-1633, 1973. L-monoglutamic Acid). Anal. Biochem., 34: 376-381 , 1970. 36. Rosengarten, H., Meller, E., and Friedhoff, A. J. Synthesis of Tetrahydro 9. Donaldson, K. 0., and Keresztesy, J. C. Naturally Occurring Forms of /3-Carbolines from Indoleamines via Enzymatic Formation of Formalde Folic Acid. I. “PrefolicA―:Preparation of Concentrate and Enzymatic hyde from 5-Methyltetrahydrofolic Acid. Biochem. Pharmacol., 24: 1759- Conversion to Citrovorum Factor. J. Biol. Chem., 234: 3235-3240, 1959. 1762, 1975. 10. Donaldson, K. 0., and Keresztesy, J. C. Further Evidence on the Nature 37. Rosenthal, S., Smith, L. C., and Buchanan, J. M. Enzymatic Synthesis of of Prefolic A. Biochem. Biophys. Res. Commun., 5: 289-292, 1961. the Methyl Group of Methionine. lx. Transmethylation from S-Adenosyl 11. Futterman, S. Preparation and Properties of Dihydrofolic Acid. Methods methionine and 5-Methyltetrahydrofolate to 2-Mercaptoethanol and Ho Enzymol., 6: 801-802, 1963. mocysteine. J. Biol. Chem., 240: 836-843, 1965. 12. Handin, R. I., Piessens, W. F., and Moloney, W. C. Stimulation of 38. Sakamoto, S., Niina, M., and Takaku, F. Thymidylate Synthetase Activity Nonimmunized Lymphocytes by Platelet-Antibody Complexes in Idi in Bone Marrow Cells in Pernicious Anemia. Blood, 46: 699-704, 1975. opathic Thrombocytopenic Purpura. New EngI. J. Med., 289: 714-718, 39. Sauer, H., Wilms, K., and Jaenicke, L. Die Aktivitát der Methionin 1973. Synthetase (5-Methyl-5,6,7,7-tetrahydrofolsâure:Homocysteine Methyl 13. Herbert, V., and Zalusky, R. Interrelations of Vitamin B,2and Folic Acid transferase) als Proliferationsparameter in Wachsenden Zellen. Acta Metabolism: Folic Acid Clearance Studies. J. Clin. Invest., 41: 1263- Haematol., 49: 200-210, 1973. 1276,1962. 40. Scnimgeour, K. G., and Vitols, K. S. The Reduction of Folate by Borohy

APRIL 1977 1131

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. J. Thorndike and W. S. Beck

dride. Biochemistry, 5: 1438-1443, 1966. Arch. Biochem. Biophys., 172: 507-520, 1975. 41. Skoog, W. A., and Beck, W. S. Studies on the Fibrinogen, Dextran and 44. Taylor, R. T., and Hanna, M. L. 5-Methyltetrahydrofolate Aromatic Alkyl Phytohemagglutinin Methods of Isolating Leukocytes. Blood, 11: 436- amine N-Methyltransferase: An Artefact of 5,10-Methylenetetrahydrofol 454, 1956. ate Reductase Activity. Life Sd., 17: 111-120, 1975. 42. Stebbins, R. D., Meller, E., Rosengarten, H., Friedhoff, A., and Silber, A. 45. Taylor, R. T., Hanna, M. L., and Hutton, J. J. 5-Methyltetrahydrofolate Identification of N5,N'0-Methylene Tetrahydrofolate Reductase as the Homocysteine Cobalamin Methyltransferase in Human Bone Marrow and Enzyme Involved in the 5-Methyl Tetrahydrofolate-Dependent Formation Its Relationship to Pernicious Anemia. Arch. Biochem. Biophys. , 165: of a 13-CarbolineDerivation of 5-Hydroxytryptamine in Human Platelets. 787-795, 1974. Arch. Biochem. Biophys., 173: 673-679, 1976. 46. Wyatt, R. J., Erdelyi, E., DoAmaral, J. R., Elliott, G. R., Renson, J., and 43. Taylor, R. T., and Hanna, M. L. Folate-Dependent Enzymes in Cultured Barchas, J. D. Tryptoline Formation by a Preparation from Brain with 5- Chinese Hamster Ovary Cells: Induction of 5-Methyltetrahydrofolate Ho- Methyltetrahydrofolic Acid and Tryptamine. Science, 187: 853-855, mocysteine Cobalamin Methyltransferase by Folate and Methionine. 1975.

1132 CANCERRESEARCHVOL. 37

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research. Production of Formaldehyde from N5-Methyltetrahydrofolate by Normal and Leukemic Leukocytes

Janet Thorndike and William S. Beck

Cancer Res 1977;37:1125-1132.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/37/4/1125

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/37/4/1125. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1977 American Association for Cancer Research.