[CANCER RESEARCH 49. 4859-4865. September I. 1989]
Biochemical Analysis of the Role of Transmethylation in the MethionineDependence of Tumor Cells1
Jean Gabriel Judder Martha Ellis, and Philip Frost1
Departments of Cell Biology fJ. G. J., M. E., P. F.] and Medicine ¡P.F.], The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030
ABSTRACT
The effects on methionine metabolism of the substitution of homocys-teine for methionine in vitro were investigated in normal and tumor celllines differing in their ability to utilize homocysteine for growth. Themajor finding of this study was that methionine-independent (Met-Indep)cell lines had much lower basal transmethylation rates than methionine-dependent (Met-Dep) cell lines. This was particularly evident in theparent SP1 cell line and its Met-Indep revenant, SP1-R. SP1-R compensated for a lack of methionine by reducing both its transmethylation andgrowth rates. An analysis of other potential differences in methioninemetabolism between Met-Dep and Met-Indep cell lines failed to demonstrate any consistent abnormalities in all but the absolutely Met-DepMDAY-D2 cell line. Thus, protein, 5-adenosylmethionine, and poly-amine synthesis were the same in Met-Dep and Met-Indep cell lines.
These results indicate that the major regulatory step in determiningthe Met-Dep phenotype is an inherent increase in the rate of transmethylation reactions. Cell lines with high basal transmethylation rates cannotcompensate for a relative deficiency of methionine and either ceasegrowing (MDAY-D2) or generate revertants (SP1-R) for which the basalrate of transmethylation is considerably reduced.
INTRODUCTION
Methionine auxotrophy (Met-Dep4) is defined as the inability
of a cell line to grow in medium devoid of methionine butcontaining its metabolic precursor homocysteine along withfolie acid and vitamin B,2 (Met~Hcy+ medium). The defect
appears to be exclusively associated with transformation(though most transformed cells are Met-Indep) since all normalcells tested thus far, including fibroblasts (1), liver and kidneycells (2, 3), and highly replicating mitogen-stimulated lymphocytes (4), can grow normally in Met~Hcy+ medium.
The biochemical basis of Met-Dep is only partially understood. In cultured mammalian cells methionine biosynthesis isaccomplished by remethylation of homocysteine by Met syn-thase (Fig. 1). It has been shown in some systems that Met-
Dep cells retain Met synthase activity at a level as high asnormal cells (Met-Indep) (5, 6) and that an increased level ofmethionine synthesis is not responsible for cells reverting fromMet-Dep to Met-Indep (7). However, Met-Dep SV40-trans-formed fibroblasts incubated in Met~Hcy+ medium labeled with['5S]homocysteine have reduced pools of free [<5S]methionine
compared to normal fibroblasts, even though similar amounts
Received 2/9/89; revised 5/25/89; accepted 6/8/89.The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported in part by Grants CA 39853 and CA 41525 of the
USPHS.2R. E. "Bob" Smith Fellow in the Department of Cell Biology*To whom requests for reprints should be addressed, at the Department of
Cell Biology (HMB 173), M. D. Anderson Cancer Center, 1515 HolcombeBoulevard, Houston. TX 77030.
4The abbreviations used are: Met-Dep, methionine-dependent: Met-Indep,methionine-independent; MerHcy* medium, medium containing homocysteinebut not methionine; SP1-R, SP1 revenant; Met*Hcy~ medium, medium contain
ing methionine but not homocysteine; AdoMet, 5-adenosylmethionine; Spm.spermine; Spd, spermidine; MTA, methylthioadenosine; AdoHcy, S-adenosyl-homocysteine; Met synthase, 5-methyltetrahydrofolate-homocysteine methyl-transferase; AdoMet synthase, methionine adenosyltransferase; HPLC, high-performance liquid chromatography.
of [15S]-methionine are incorporated into proteins (8). In addi
tion, (isotopie) AdoMet levels are decreased and AdoHcy levelsare increased in Met-Dep as opposed to Met-Indep cells (9).These studies indicate that defective methionine biosynthesis isnot the basis for Met-Dep in the cell lines examined and suggestthat the Met-Indep or Met-Dep phenotype may be determinedby the ability or inability of the cell to maintain a normalAdoMet/AdoHcy ratio in Met~Hcy* medium.
As shown in Fig. 1, the activity of AdoMet synthase, whichcatalyzes the transfer of the adenosyl moiety of ATP to thesulfur atom of methionine, and the transmethylation andpolyamine pathways are central to the regulation of cellularAdoMet levels. In fact, elevations of AdoMet-dependent transmethylation reactions (10) and polyamine synthesis (11) havebeen observed in tumor cell lines as have deficiencies in MTAphosphorylase (12), a key enzyme in the salvage of methioninethrough the polyamine pathway. Alterations in any of thesesteps in the methionine pathway could result in increasedAdoMet and methionine requirements in tumor cells.
To assess which step(s) are responsible for the acquisition ofthe Met-Dep phenotype by transformed cells, we undertook amore comprehensive analysis of methionine metabolism innormal and tumor cell lines the abilities of which to proliferatein Met Hey+ medium we recently characterized (4). Amongthese cell lines, a Met-Indep revenant isolated from a Met-Deptumor cell line provided us with a particularly good model foridentifying the biochemical basis for methionine auxotrophy.
MATERIALS AND METHODS
Materials. Radioactive compounds were purchased from AmershamCorporation, Arlington, IL. [/nefA>7-'H]MTA was prepared from[methyl-'H]-S-adenosyl-L-methionine (83 Ci/mmol) by acid hydrolysis
and purified by HPLC (13, 14). The Dowex resins used for productseparation in the enzyme assays were obtained from Bio-Rad Laboratories (Richmond, CA). P81 ion exchange chromatography paper waspurchased from Whatman (Clinton, NJ) and 3-deazaadenosine wasfrom the Southern Research Institute (Birmingham, AL). Other biochemical reagents were obtained from Sigma Chemical Co. (St. Louis,MO). Media and materials for tissue culture were purchased from FlowLaboratories (Rockville, MD).
Cell Lines. The human HE lung embryo fibroblast cell line wasobtained from Flow Laboratories. The mouse SP1 mammary adeno-carcinoma (15), MDAY-D2 lymphoma (16), and human A375 melanoma (17) have been described elsewhere. SP1-R is a spontaneous Met-Indep revenant isolated from the SP1 cell line in this laboratory (4).For these studies, cells were grown in methionine-deficient RPMI 1640supplemented with penicillin (100 units/ml), streptomycin (0.1 mg/ml), 10% dialyzed fetal calf serum (shown to be free of methionine byhigh-voltage paper electrophoresis), 100 ^M folie acid, 1.5 ^M cyano-cobalamin, and 200 ft\i DL-methionine (Met*Hcy~ medium). The SP1-
R cell line was maintained in the same medium where methionine wasreplaced by 200 ¿IMOL-homocysteine thiolactone (Met~Hcy* medium).The cultures were kept in humidified 95% air-5% CO2 at 37°Cand
subcultured with 0.25% trypsin every 3 or 4 days. To study the effectsof homocysteine substitution for methionine, cells were plated inMet*Hcy~ medium and on the following day the medium was pouredoff and replaced by Met~Hcy* medium. The cells were then harvested
Fig. I. Methionine metabolism in mammalian cells. The numbers representthe enzymatic reactions and pathways which have been examined in this study:/, Met synthase; 2, AdoMet synthase; 3, AdoMet-dependent transmethylation
reaction: 4, AdoMet decarboxylase; 5, methylthioadenosine phosphorylase; 6,utilization of methionine for protein synthesis; 7. cystathionine synthase. PUT,putrescine; ORN. ornithine; MTR-l-P, 5-methylthioribose 1-phosphate: THF,
tetrahydrofolate.
Measurement of Intracellular AdoMet and AdoHcy. Cell cultureswere harvested by trypsinization and washed twice in ice-cold phosphate-buffered saline. The cell pellet was extracted on ice for 30 min in0.4M perchloric acid (2 x 107cells/ml)containing0.15% (w/v)Na2S2O5
and 0.05% (w/v) EDTA. After centrifugation, the acid supernatant wasfiltered through disposable 0.2-//m Spin-X filter units (Costar, Cambridge, MA) and stored at —¿�70°Cfor up to 1 month without appreciable
loss of activity. Perchloric acid extracts were analyzed by HPLC asdescribed previously (18) using a Beckman Ultrasphere ion-pair column(particle size, 5 urn). Samples were eluted at 1.5 ml/min using a lineargradient of 0.1 M NaH:PO4:acetonitrile (98:2, v/v) and 0.15 MNaH:PO4:acetonitrile (74:26, v/v) containing 8 x 10~' Moctanesulfonic
acid sodium salt (Kodak, Rochester, NY). AdoMet and AdoHcy werequantified by monitoring absorbance at 254 nm. External standardswere injected every ten samples and used for calibration.
Polyamine Analysis. Cells were trypsinized and washed twice withice-cold phosphate-buffered saline. The cell pellet was extracted on icefor l h with 200 n\ of 8% trichloroacetic acid for each 10" cells. The
precipitate was separated by centrifugation at 10,000 X g for 10 min,and aliquots of the supernatant were immediately used for derivatiza-
tion of polyamines. Fifty n\ of supernatant were mixed with 50 ^1 of 20¿IM1,6-hexanediamine in 0.1 N HC1 (internal standard), 200 ¿iI of asaturated Na:CO, solution, and 200 ^1 of 10 mg/ml dansyl chloride inacetone. After agitation for 15 s on a vortex machine the mixture washeated for 10 min at 70°C.The derivatized polyamines were then
adsorbed on Ci«Bond Elut columns (Analytichem International, Harbor City, CA), eluted with 0.5 ml absolute methanol, and stored at 4°C
for up to 2 weeks without appreciable loss of activity.The derivatized polyamines were assayed by HPLC as described
previously (19) using a Beckman Ultrasphere octadecylsilane reversephase column maintained at 50°Cand perfused at 2 ml/min with a
gradient of acetonitrile and 10 IHMNaH2PO4, pH 4.4. External standards of putrescine, spermidine, and sperm ine were derivatized and usedfor calibration.
Enzyme Assays. Cells were harvested by trypsinization and the pelletwas resuspended at 2 x IO7 cells/ml in 0.25 M sucrose. Sucrose-cellsuspensions could be stored at —¿�70°Cfor at least 3 weeks without loss
of enzyme activity. Cell extracts were prepared just before use by threerounds of freeze-thawing followed by centrifugation at 10,000 x g for15 min at 4°C.Determination of enzyme activities were done according
to previously published assays for 5-adenosylmethionine decarboxylase(20), MTA phosphorylase (21), Met synthase (22), AdoMet synthase(23), and cystathionine synthase (24). Enzyme assays were carried outunder conditions in which the activity was proportional to the amountof protein added and to the time of incubation. Results are expressedas nmol of product obtained in a 1-h incubation per mg of protein.
Protein concentration was determined by the method of Lowry et al.(25) using lysozyme as standard.
Measurement of Transmethylation Rates. The rate of AdoMet-dependent transmethylation reactions was assayed by measuring the time-dependent accumulation of AdoHcy in cells treated with periodale-oxidized 3-deazaadenosine, a potent inhibitor of AdoHcy hydrolase(26). Periodate-oxidized 3-deazaadenosine was prepared just before useby the reaction of 3-deazaadenosine with sodium periodate in 0.1 Macetate, pH 4, as described previously (10). Cells in exponential growthwere incubated with 10 ¿IMinhibitor, a concentration that producescomplete inhibition of AdoHcy hydrolase in cultured cells (10), andharvested every 30 min for HPLC analysis of cellular AdoHcy content.
RESULTS
Cell Growth Studies. As shown in Fig. 2, the five cell linesunder study differed markedly in their ability to proliferate inMet~Hcy+ medium. The HE lung fibroblasts showed no differ
ence in growth rate when homocysteine replaced methionine inthe medium. However, they had a doubling time of 130 hcompared to 13.8 h for the SP1 mammary adenocarcinoma.The Met-Indep A375 human melanoma cell line had a slightlyreduced growth rate during the first 3 days in Met~Hcy+ me
dium and then resumed growth at a normal rate. In contrast,the absolutely Met-Dep MDAY-D2 lymphoma and SP1 adenocarcinoma did not grow in Met~Hcy*. However, although
we are unable to obtain Met-Indep revenants from MDAY-D2despite repeated attempts at selection, SP1 Met-Indep rever-tants (SP1-R) were obtained that could proliferate in MerHcy*medium (SP1-R cell line). Although SP1-R cells could growpermanently in Met~Hcy+ medium, their doubling time increased from 14.5 h in Met+Hcy~ medium to 32 h in Met~Hcy+
medium. Because of the significant degree of cell death whenMDAY-D2 and SP1 (to a lesser extent) were placed inMet~Hcy+ medium for 48 h, we limited our analysis to the first2 days of in vitro culture in Met~Hcy+ medium.
Transmethylation Rates. AdoMet-dependent methyl transferreactions consume about 70% of AdoMet in growing cells (27,28). In addition, the overall rate of AdoMet-dependent trans-
A375 MDAY-D2
012301234
DaysFig. 2. Growth of the cell lines in Met*Hcy" (•)and Met'Hcy* (O) media.
METHIONINE DEPENDENCE AND TRANSMETHYLATION IN NEOPLASIA
methylation reactions is increased in tumor cell lines comparedto normal fibroblasts (10), as is the methylating capacity ofDNA (29) and tRNA (30) methyltransferases. In order to testthe hypothesis that Met-Dep could be related to an increase inAdoMet requirements through transmethylation reactions, weused the rate of AdoHcy accumulation in cells treated with thepotent AdoHcy hydrolase inhibitor periodate-oxidized 3-dea-zaadenosine as a measurement of AdoMet-dependent transmethylation rates (10). As shown in Fig. 3, the Met-Dep SP1and MDAY-D2 cells accumulated AdoHcy approximately 5-fold faster than the Met-Indep HE and A375 cells. Interestingly,the SP1-R Met-Indep revenant had a low transmethylation ratecompared with the parent SP1 cell line.
Protein Synthesis. In preliminary experiments we had foundthat after a 4-h incubation of cells with [<5S]methionine, 90%
of the radioactivity was incorporated into proteins. We therefore could use the relatively safer radiolabeled ['HJmethionine
as a measure of methionine consumption in protein synthesis.Fig. 4A demonstrates that the incorporation of ['Hjmethionine
into acid-insoluble material occurred at similar rates for all thecell lines tested. We next examined the effect of the substitutionof methionine by homocysteine on the overall rate of proteinsynthesis as measured by the incorporation of ['Hjleucine intoacid-insoluble material from cells incubated in Met~Hcy+ me
dium. As shown in Fig. 4Ä,protein synthesis remained unaffected over a 24-h period for the Met-Indep HE and A375 celllines. While ['Hjleucine incorporation in the Met-Dep SP1 cell
line decreased 50% at 4 h, it returned to normal at 24 h. Incontrast, MDAY-D2 showed a continuous decrease in proteinsynthesis to 20% of background by 24 h correlating with celldeath.
Effect of Met~Hcy+ Medium on AdoMet, AdoHcy, and
Polyamine Pools. Previous studies using radioisotopic methodsindicated that after 24 h incubation in Met~Hcy* medium,
AdoMet levels decreased and AdoHcy levels increased in Met-Dep tumors compared with Met-Indep tumors (9). We thusused a HPLC method to compare the AdoMet and AdoHcypools of cell lines grown in Met'Hcy* medium. The results are
shown in Fig. 5. In the normal Met-Indep HE fibroblast cell
30 60Minutes
Fig. 3. Rates of transmethylation. HE (•).A375 (O), SP1 (A), and MDAY-D2 (•)cells were placed in Met*Hcy~ medium and SP-1 R cells (A) in Met'Hcy*
medium. At zero time, periodate-oxidized 3-deazaadenosine was added at aconcentration of 10 JIM and the intracellular AdoHcy pools were measured byHPLC at each time point. Values are the average of triplicate determinationswith less than 10% SD.
80248 24
Hours
Fig. 4. (A) Rate of methionine incorporation into proteins. At zero time cellswere placed in Met*Hcy~ medium containing 25 >JMi.-['H|methionine (I /iCi/ml). The radioactivity in trichloroacetic acid-insoluble material was determinedat each time from triplicate cultures of HE (•).A375 (O), SP1 (A), and MDAY-D2 (A) cells. (B) Effect of methionine substitution by Hey on protein synthesis.At zero time Met*Hcy~ medium was replaced with MerHcy* medium and thecultures were pulsed at each time point with 25 /JM['Hjleucine (1 i<Ci/ml) for 30min. Each point represents the radioactivity of trichloroacetic acid-insolublematerial in HE (•).A375 (O), SP1 (A), and MDAY-D2 (A) cells.
Fig. 5. Effect of methionine substitution by homocysteine on the intracellularpools of AdoMet (•)and AdoHcy (O). At zero time Met^Hcy medium wasreplaced by Met'Hcy* medium and the cellular pools of AdoMet and AdoHcy
were determined by HPLC. Each point is the average of triplicate determinationswith less than 10% SD.
line, AdoMet levels remained essentially constant for the first8 h in MerHcy* medium and then increased about 4-fold by24h. In the Med-Indep A375 and Met-Dep SP1 cell lines,AdoMet levels declined during the first 4-8 h and then increased about 2.5-fold by 24 h. In contrast, the AdoMet poolin the Met-Dep MDAY-D2 cell line was almost completelydepleted by 8 h (20 pmol/10" cells compared to 6400 pmol/106cells in Met+Hcy~ medium) and remained extremely low at 24h ( 100 pmol/10" cells).
For all cell lines the AdoHcy pool was increased 2-4-foldduring the first 2-4 h in Met~Hcy* medium. This increase may
have resulted from a partial reversal of the hydrolysis ofAdoHcy in the presence of excess homocysteine (31 ). After 24h, AdoHcy levels were either normal in HE and MDAY-D2cells or increased by about 3-fold in A375 and SP1 cells.
The variations in the AdoMet/AdoHcy ratios followed asimilar pattern in HE, A375, and SP1 cells with a transient butsignificant drop during the first 2-4 h followed by a return tonormal by 24 h. In MDAY-D2 cells, the AdoMet/AdoHcy ratiodecreased dramatically from 119 in Met*Hcy~ medium to 1.0after 8 h and 1.8 after 24 h in MerHcy* medium. This patternin MDAY-D2 resulted entirely from the depletion of theAdoMet pool. The Met-Indep revenant SP1-R had AdoMetand AdoHcy levels in Met~Hcy+ medium similar to those of
the parent SP1 cell line (data not shown).In addition to its role as a methyl donor in transmethylation
METHIONINE DEPENDENCE AND TRANSMETHYLAT1ON IN NEOPLASIA
reactions, AdoMet provides the aminopropyl moiety necessaryfor the synthesis of the polyamines Spd and spermine. It istherefore possible that a polyamine sink could deplete AdoMetin Met-Dep cells. Alternatively, low AdoMet levels could resultin polyamine depletion and inhibition of cell growth. To assessthese possibilities we examined the levels of cellular polyaminesin cells grown in Met'Hcy* medium and determined their
relationship to AdoMet levels. As shown in Fig. 6, all tumorcell lines had similar pools of Spd and Spm in Met+Hcy~medium (between 6 and 7 mmol/106 cells), while the HE
fibroblast cell line had a slightly lower polyamine content (4.3mmol/106 cells). However, the ratio of Spd to Spm markedly
differed between the cell lines. HE and A375 had a Spd/Spmratio of about 1, compared to 3 for MDAY-D2 and SP1. TheSpd/Spm ratio correlated with the growth rate: cells having afast growth rate had a high Spd/Spm ratio. This higher Spd/Spm ratio in fast growing cells has been described previouslyin a comparative study of 3T3 and SV40-transformed 3T3 cells(32).
The fluctuations of the polyamine pools in response to me-thionine substitution by homocysteine followed a similar pattern for all cell lines. There was a rapid but transient decreasein both Spd and Spm between 2 and 4 h, one of which wasmarginal in HE cells and most pronounced in MDAY-D2 cells(59% of the Spd plus Spm content in Met+Hcy~ medium). The
polyamine content then returned to normal or supranormallevels (60% increase in SP1 cells) by 4 to 8 h and remained atthese levels during the 48-h observation period. In Met Hey*medium, the SP1-R revertant had 6.4 nmol Spd/106 cells and1.6 nmol Spm/106 cells compared to 7.0 nmol Spd/106 and 1.5
A375
4 B 24 24 4848 048Hours
Fig. 6. Effect of methionine substitution by homocysteine on the intracellularpools of spermidine (•).spermine (O). and spermidine plus spermine (A). At zerotime Met*Hcy- medium was replaced by Met Hey* medium and cellular polyam
ine pools were determined by HPLC. Each point is the average of triplicatedeterminations with less than 10% SD.
nmol Spm/106 for the parent SP1 cell line. Thus, the early
decrease in the polyamine content corresponded to a decreasein AdoMet pools during the first hours in Met~Hcy+ medium
and probably resulted from a lack of AdoMet substrate forAdoMet decarboxylase.
Enzymatic Studies. The activities of Met synthase, AdoMetsynthase, and MTA phosphorylase in extracts from cells incubated for 24 h in either Met+Hcy~ or MerHcy* medium are
shown in Table 1. All cell lines except MDAY-D2 had similarMet synthase activities in Met+Hcy~ medium and showed asignificant but modest increase in activity in Met~Hcy+ medium. The SP1-R revertant grown in Met~Hcy+ medium had
an only slightly higher Met synthase activity than the parentSP1 cell line. Thus, the ability of HE, A375, and SP1-R cellsto grow in Met^Hcy* medium did not correlate with an increase
in methionine biosynthesis when compared with the Met-DepSP1 cells. However, MDAY-D2 cells had a very low basal Metsynthase activity in Met*Hcy~ medium that further decreasedin MerHcy* medium. These differences in methionine biosyn
thesis could not be explained by differences in cystathioninesynthase activity from the same cell extracts (Table 1).
The variations in AdoMet synthase activity for each cell linewere consistent with the changes observed in the AdoMet pools.There was a 3-, 7-, and 6.6-fold increase in AdoMet synthaseactivity in HE, A375, and SP1 cells, respectively, after 24 h inMet~Hcy+ medium. The SP1-R revertant also had elevatedAdoMet synthase activity in Met~Hcy+ medium, although it
was only 60% of the activity in the parent SP1 cell line in thesame medium. In Met+Hcy~ medium, SP1-R showed a decrease
in AdoMet synthase activity to basal levels. In MDAY-D2,AdoMet synthase activity was depressed after 24 h in Met~Hcy+
medium, which accounted for the inability of this cell line tomaintain its AdoMet pool in this medium.
The utilization of AdoMet for polyamine synthesis yields thethioether nucleoside MTA, which is metabolized by MTAphosphorylase to adenine and 5-methylthioribose 1-phosphate(33). The 5-methylthioribose 1-phosphate is then reconvertedto methionine (34). Since MTA phosphorylase deficiency hasbeen reported in some tumors (12) and would represent a lossof methionine by an interruption of the methionine salvagepathway, we also measured this enzyme activity. Methioninesubstitution by homocysteine had no significant effect on MTAphosphorylase activity and all cell lines had sufficient activityto degrade the MTA produced during polyamine synthesis.
Table Effect of methionine substitution by homocysteine on specific enzymeactivities of the methionine pathway0
METHIONINE DEPENDENCE AND TRANSMETHYLATION IN NEOPLASIA
In response to methionine substitution by homocysteine, theactivity of AdoMet decarboxylase increased severalfold in allcell lines, reaching a peak at 8 h for HE, A375, and MDAY-D2 and at 24 h for SP1 (Fig. 7). AdoMet decarboxylase activitythen returned to normal at 48 h except in SP1 where it remainedelevated. SP1-R had a basal AdoMet decarboxylase activitysimilar to the parent SP1 cell line grown in Met+Hcy~ medium.It thus appears that in Met~Hcy+ medium, a rapid and signifi
cant increase in AdoMet decarboxylase activity allows cells tomaintain their polyamine pools in the face of decreasingAdoMet levels.
Cystathionine synthase levels were also measured in Met-Dep and Met-Indep cells. No significant differences were foundbetween the two phenotypes (Table 1).
DISCUSSION
We have presented evidence to indicate that the critical factordetermining methionine dependence in tumor cells related totheir inherent rates of transmethylation. In cells with hightransmethylation rates, failure to increase AdoMet synthaseactivity in Met~Hcy+ medium resulted in a rapid depletion of
the AdoMet pool and cell death (MDAY-D2). In cells able tomaintain AdoMet levels by increasing AdoMet synthesis, thebalance between methionine synthesis from its precursor homocysteine and the methionine requirements for AdoMet-de-pendent transmethylation reactions ultimately determined thegrowth rate of tumor cells in Met~Hcy+ medium.
Our ability to define the biochemical parameters of methionine dependency is due to the identification of specific cell lineswith unique Met-Dep phenotypes (4). Thus, we were able tocompare normal fibroblasts and Met-Indep tumor cells to absolutely Met-Dep cells that do not survive in Met~Hcy+ medium
(MDAY-D2) and to cells that although initially Met-Dep areable to generate Met-Indep revenants (SP1-SP1-R).
The metabolism of methionine by normal HE fibroblasts isessentially the same whether the cells are grown in Met+Hcy~or Met~Hcy+ media. This is likely because these cells grow
slowly and have low transmethylation rates. HE fibroblasts do,however, undergo several metabolic changes during their first4 h of growth in Met~Hcy+ medium. These changes are mani
fested as a decrease in AdoMet and polyamine levels and anincrease in AdoHcy levels that may result from a partial reversalof the AdoHcy hydrolase reaction towards AdoHcy synthesis,because of an increase in cellular homocysteine levels (31). Thecells compensate for these changes with time by a 3-fold increase in AdoMet synthase activity that allows them to grownormally in Met~Hcy+ medium.
Fig. 7. Effect of methionine substitution by homocysteine on the activity ofAdoMet decarboxylase. At zero time Met*Hcy~ medium was replaced byMet~Hcy* medium and cell cultures were harvested at different times for determination of specific enzyme activities in HE (•),A375 (O), SP1 (A), and MDAY-D2 (•)cell lines.
Similarly, Met-Indep tumor cells such as A375 had low basalrates of transmethylation (and therefore low AdoMet requirements) and compensated for the substitution of homocysteinefor methionine by increasing methionine and AdoMet synthaseactivities. These cells were thus able to survive and grow at anormal rate in Met~Hcy+ medium. However, the AdoMet pools
were initially much more affected in A375 than in the normalHE fibroblast cell line, and the compensatory increase inAdoMet synthase was also more significant, suggesting theAdoMet requirements in A375 are higher. Although both celllines have low transmethylation rates, the polyamine levels,AdoMet decarboxylase activity, and rate of methionine incorporation into proteins were all slightly higher in A375 cells.This is probably due to the higher growth rate of A375 andmay account for the more pronounced effect of Met Hey*
medium on AdoMet synthesis in A375 cells.In contrast, MDAY-D2 tumor cells are absolutely Met-Dep
and manifest profound changes in many parameters of themethionine pathway when placed in Met'Hcy* medium.
MDAY-D2 cells have 5-fold higher transmethylation rates thanHE or A375 cells. Thus, the AdoMet requirement in MDAY-D2 cells is increased and the growth of such cells in MerHcy*
medium would require dramatic compensatory biochemicalchanges so as to enable them to maintain their high transmethylation rate. In fact, MDAY-D2 cells grown in MerHcy+
medium are unable to increase either methionine or AdoMetsynthase activities resulting in a near total depletion of cellularAdoMet. The AdoMet/AdoHcy ratio in MDAY-D2 cellsdropped from 119 in methionine-containing medium to 1.0after 8 h in Met~Hcy+ medium, presumably resulting in the
inhibition of essential transmethylation reactions (35). This ledto suppression of protein synthesis and cell death within 24 h.Our data do not provide an explanation for the inability ofthese cells to increase either Met synthase or AdoMet synthaseactivities in Met~Hcy+ medium. However, the fact that the
AdoMet decarboxylase activity of MDAY-D2 cells inMet"Hcy+ medium increases in order to maintain polyamine
levels and that MTA phosphorylase also increases indicatesthat there is no general inhibition of enzyme activity secondaryto the observed decrease in protein synthesis.
Of particular interest are the SP1 cells which, when observedinitially, were thought to be absolutely Met-Dep. These cellsalso have high transmethylation rates. However, SP1 cellsplaced in Met~Hcy+ medium become cytostatic, but after ap
proximately 1 week they begin dividing and generate a Met-Indep revertan! cell line, SP1-R (4). The frequency of reversionin SP1 cells can be as high as 10%. Although SP1 cells stoppedgrowing in Met~Hcy+ medium, biochemical analyses showed
that following an initial decrease, a considerable increase inAdoMet levels occurred after 24 h in this medium. This changewas due to a 6-7-fold increase in AdoMet synthase activity anda significant but less dramatic increase in Met synthase activity.Thus, although SP1 cells cannot initially divide in Met~Hcy+
medium, in contrast to MDAY-D2 cells they can survive via acompensatory increase in AdoMet synthase activity allowingthem to maintain a high AdoMet/AdoHcy ratio.
Although our data do not identify the actual mechanismaffecting cell division in SP1, several lines of evidence stronglysuggest that it is related to the high rate of transmethylation ofthese cells, (a) The major biochemical difference between theSP1-R and parent SP1 cells is a marked decrease in the transmethylation rate in SP1-R that results in a decreased requirement for AdoMet. This seems to be the major compensatorymechanism allowing SP1-R cells to regain the ability to divide
METHIONINE DEPENDENCE AND TRANSMETHYLATION IN NEOPLASIA
in Met Hcy+ medium, albeit at a slower rate, (b) SP1 and A375
cells differ primarily in their rates of transmethylation, againsuggesting that a lower AdoMet demand allows cells to compensate more readily for a relative methionine deficiency. It ispossible that since SP1 cells must compensate for their hightransmethylation rate by a corresponding increase in AdoMetsynthesis, a relative state of ATP deficiency could occur, resulting in cytostasis. Such a mechanism would be consistent withthe metabolic studies of Hardwick et al. (36) on methionine-induced toxicity in guinea pigs, which attributed methioninetoxicity to hepatic ATP insufficiency resulting from increasedAdoMet synthesis (36).
All the cell lines we studied maintained their polyaminecontent despite extensive AdoMet depletion (MDAY-D2 cells)by increasing AdoMet decarboxylase activity. This observationconfirms that polyamine synthesis is tightly regulated andagrees with studies using AdoMet synthase inhibitors, whichindicated that depletion of AdoMet pools suppressed cellgrowth by inhibiting transmethylation reactions without affecting polyamine levels (37). In addition, besides having similarpools of Spd and Spm, none of the cell lines studied wasdeficient in MTA phosphorylase activity, making the existenceof a polyamine sink unlikely. Therefore, the polyamine pathwaydoes not appear to contribute significantly to the Met-Depphenotype in the cell lines examined.
A question raised by these studies is why high rates oftransmethylation deplete the AdoMet pool in Met Hcy+ me
dium, since the methionine pathway uses homocysteine obtained from AdoHcy hydrolysis for recycling into methioninevia Met synthase. One possibility is that the cystathioninesynthase reaction, by constituting the only outlet of homocysteine from the cycle, could shunt homocysteine toward cysteineproduction (38). However, we found no increase in the activityof this enzyme in Met-Dep cells. This suggests that the Met
synthase reaction is unable to salvage excess homocysteine incells with high transmethylation rates grown in MerHcy+
medium. In support of this view are studies in patients withhomocystinuria who lack cystathionine synthase and in whomhomocysteine methylases are insufficient to metabolize homocysteine derived from transmethylation reactions (39). Furthermore, rats fed choline-deficient diets show a decrease in hepaticAdoMet levels and an increase in AdoHcy levels, indicatingthat the conservation of methionine by the methyltetrahydro-folate-homocysteine methyltransferase reaction alone, may not
be possible (40).Our finding that the critical event in Met-Dep is a basal
increase in transmethylation rates confirms previous studies (inunrelated cell lines) suggesting a possible link between alteredtransmethylation in tumor cells and the Met-Dep phenotype(10, 41). Some cells with high basal transmethylation rates cancompensate for the substitution of homocysteine for methionineby increasing AdoMet synthesis and ultimately generating re-vertants; others cannot. Cells with low transmethylation ratescompensate even more readily for the substitution of homocysteine for methionine since their demands for AdoMet areconsiderably lower and they need not generate revertants tosurvive. We would emphasize that we have found very littleevidence for absolute Met-Dep among tumor cells (4), and thatmost Met-Dep tumor cell lines are able to generate Met-Indeprevertants (7). Thus, tumors like MDAY-D2 that do not com
pensate for high transmethylation rates are probably very rare.Although rare, in human cancer this phenotype could representa target for therapy. Whether specific types or subtypes of
tumors with a high frequency of Met-Dep exist remains to bedetermined.
REFERENCES
1. Halpern, B. C, Clark, B. R., Hardy, D. N., Halpern, R. M., and Smith, R.A. The effect of replacement of methionine by homocystine on survival ofmalignant and normal adult mammalian cells in culture. Proc. Nati. Acad.Sci. USA, 71:1133-1136, 1974.
2. Wilson, M. J., and Poirier, L. A. An increased requirement for methionineby transformed rat liver epithelial cells in vitro. Exp. Cell Res., ///: 397-400, 1978.
3. Jacobson, S. J., North, J. A., Rao, N. A., and Mangum, J. H. 5-Methyl-tetrahydrofolate: synthesis and utilization in normal and SV40-transformedBHK-21 cells. Biochem. Biophys. Res. Commun., 76:46-53, 1977.
4. Judde, J. G., and Frost, P. Patterns of methionine auxtrophy in normal andneoplastic cells: the methionine independence of lymphocyte mitogenesisand low frequency of the methionine-dependent phenotype in human tumors.Cancer Res., 48: 6775-6779, 1988.
5. Hoffman, R. M., and Erbe, R. W. High in vivo rates of methionine biosynthesis in transformed human and malignant rat cells auxotrophic for methionine. Proc. Nati. Acad. Sci. USA, 73: 1523-1527, 1976.
6. Tisdale, M. J. Methionine metabolism in Walker carcinoma in vitro. Eur. J.Cancer, 16: 407-414, 1980.
7. Hoffman, R. M., Jacobsen, S. J., and Erbe, R. W. Reversion to methionineindependence by malignant rat and SV40-transformed human fibroblasts.Biochem. Biophys. Res. Commun., 82: 228-234, 1978.
8. Stern, P. H., Mecham, J. O., Wallace, D. C., and Hoffman, R. M. Reducedfree-methionine in methionine-dependent SV40-transformed human fibroblasts synthesizing apparently normal amounts of methionine. J. Cell. Phys-iol., //7.-9-14, 1983.
9. Stern, P. H., Wallace, C. D., and Hoffman, R. M. Altered methioninemetabolism occurs in all members of a set of diverse human tumor cell lines.J. Cell. Physiol., 7/7:9-14, 1983.
10. Stern, P. H., and Hoffman, R. M. Elevated overall rates of transmethylationin cell lines from diverse human tumors. In Vitro (Rockville), 20: 663-670,1984.
11. Pegg, A. E. Polyamine metabolism and its importance in neoplastic growthand as a target for chemotherapy. Cancer Res., 48: 759-774, 1988.
12. Kamatani, N., Nelson-Rees, W. A., and Carson, D. A. Selective killing ofhuman malignant cells deficient in MTA phosphorylase, a purine metabolicenzyme. Proc. Nati. Acad. Sci. USA, 78: 1219-1223, 1981.
13. Parks, L. W., and Schlenk, F. Formation of a-amino->-butyrolactone from5-adenosylmethionine. Arch. Biochem. Biophys., 75: 291-292, 1958.
14. Kamatani, N., Willis, E. H., and Carson, D. A. Sequential metabolism of 5'-isobutylthioadenosine by methylthioadenosine phosphorylase and purine-nucleoside phosphorylase in viable human cells. Biochem. Biophys. Res.Commun., 104: 1335-1342, 1982.
15. Frost, P., Kerbel, R. S., Hunt, B., Man, S., and Pathak, S. Selection ofmetastatic variants with identifiable karyotypic changes from a nonmetastaticmurine tumor after treatment with 2'-deoxy-5-azacytidine or hydroxyurea:
implications for the mechanisms of tumor progression. Cancer Res., 47:2690-2695, 1987.
16. Frost, P., Kerbel, R. S., and Tartamella-Biondo, R. Generation of highlymetastatic tumors in DBA/2 mice. Invasion Metastasis, /: 22-33, 1981.
17. Giard, D. J., Aaronson, S. A., and Todaro, G. J. In vitro cultivation of humantumors: establishment of cell lines derived from a series of solid tumors. J.Nati. Cancer Inst., 51: 1417-1423, 1978.
18. Wagner, J., Danzin, C., Huot-Olivier, S., Claverie, N., and Palfreyman, M.G. High-performance liquid Chromatographie analysis of S-adenosylme-thionine and its metabolites in rat tissues: interrelationship with changes inbiogenic catechol levels following treatment with L-DOPA. J. Chromatogr.,290:247-262, 1984.
19. Kabra, P. M., Lee, H. K., Lubich, W. P., and Marton, L. J. Solid-phaseextraction and determination of dansyl derivatives of unconjugated andacetylated polyamines by reverse phase liquid chromatography: improvedseparation systems of polyamines in cerebrospinal fluid, urine and tissue. J.Chromatogr., 380: 19-32, 1986.
20. Pegg, A. E. The role of polyamine depletion and accumulation of decarbox-ylated 5-adenosylmethionine in the inhibition of growth of SV-3T3 cellswith a-difluoromethylornithine. Biochem. J., 224: 29-38, 1984.
21. Pegg, A. E., Borchardt, R. T., and Coward, J. K. Effects of inhibitors ofspermidine and spermine synthesis on polyamine concentrations and growthof transformed mouse fibroblasts. Biochem. J., 194: 79-89, 1981.
22. Kamely, D.. Littlefield, J. W., and Erbe, R. W. Regulation of 5-methyl-tetrahydrofolate:homocysteine methyltransferase activity by methionine, vitamin B,2, and folate in cultured baby hamster kidney cells. Proc. Nati. Acad.Sci. USA, 70: 2585-2589, 1973.
23. Okada, G., Teraoka, H., and Tsukada, K. Multiple species of mammalian S-adenosylmethionine synthetase. Partial purification and characterization.Biochemistry, 20: 924-940, 1981.
24. Mudd, S. H., Finkelstein, J. D., Irreverre, F., and Laster, L. Transsulfurationin mammals. Microassays and tissue distribution of three enzymes of thepathway. J. Biol. Chem., 240: 4382-4392, 1965.
25. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. Protein
27. German, C. D., Bloch, C. A., and Kredich, N. M. Measurements of S-adenosylmethionine and L-homocysteine metabolism in cultured humanlymphoid cells. J. Biol. Chem., 25«:10997-11003, 1983.
28. lisaza, T., and Carson, D. A. Differential regulation of polyamine synthesisand transmethylation reactions in methylthioadenosine phosphorylase deficient mammalian cells. Biochim. Biophys. Acta, 844: 280-287, 1985.
29. Kautiainen, T. L., and Jones, P. A. DNA methyltransferase levels in tumor-igenic and nontumorigenic cells in culture. J. Biol. Chem., 261: 1594-1598,
1986.30. Brunke, K. J., Strickler, G., and Leboy, P. S. Elevated methylation capacity
of selected transfer RNA methyltransferases from 9,10-dimethyl-l,2-benzan-thracene-induced rat mammary tumors. Cancer Res., 40: 417-423, 1980.
31. Ueland, P. M. Pharmacological and biochemical aspects of 5-adenosylhom-ocysteine and S-adenosylhomocysteine hydrolase. J. Biol. Chem., 258:10997-11003, 1983.
32. Bethell, D. R., Hibasami, H., and Pegg, A. E. Regulation of polyaminecontent in cultured fibroblasts. Am. J. Physiol., 243: C262-C269, 1982.
33. Pegg, A. E., and Williams-Ashman, H. G. On the role of 5-adenosyl-L-
methionine in the biosynthesis of spermidine by rat prostate. J. Biol. Chem.,244: 682-693, 1968.
34. Backlund. P. S.. Jr., Chang, C. P., and Smith, R. A. Identification of 2-keto-4-methylthiobutyrate as an intermediate compound in methionine synthesisfrom 5'-methylthioadenosine. J. Biol. Chem., 257:4196-4202. 1982.
35. Hoffman, D. R., Marion. D. W., Cornatzer, W. E., and Duerre, J. A. 5-Adenosylmethionine and S-adenosylhomocysteine metabolism in isolated ratliver. J. Biol. Chem., 255: 10822-10827, 1980.
36. Hardwick, D. F., Applegarth, D. A.. Cockcroft. D. M., Ross, P. M., andCalder, R. J. Pathogenesis of methionine-induced toxicity. Metabolism (Baltimore), / 9: 381-391.
37. Porter, C. W., Sufrin, J. R., and Keith, D. D. Growth inhibition by methionine analog inhibitors of 5-adenosylmethionine biosynthesis in the absenceof polyamine depletion. Biochem. Biophys. Res. Commun.. 722: 350-357,1984.
38. Finkelstein, J. D., Martin, J. J., and Harris, B. J. Methionine metabolism inmammals: the methionine-sparing effect of cystine. J. Biol. Chem., 263:11750-11754, 1988.
39. Mudd, S. H., and Levy, H. L. In: }. B. Stanbury, J. B. Wyngaarden, D. S.Frederickson, J. L. Goldstein, and M. S. Brown (eds.). The Metabolic Basisof Inherited Disease. Ed. 5. pp. 552-559. New York: McGraw-Hill.
40. Finkelstein. J. D., Martin, J. J., Harris, B. J., and Kyle, W. E. Regulation ofbetaine content of rat liver. Arch. Biochem. Biophys.. 218: 169-173, 1982.
41. Hoffman. R. M. Altered methionine metabolism and transmethylation incancer. Anticancer Res., 5: 1-30, 1985.
1989;49:4859-4865. Cancer Res Jean Gabriel Judde, Martha Ellis and Philip Frost Methionine Dependence of Tumor CellsBiochemical Analysis of the Role of Transmethylation in the
