The metabolismof vitamin B@in McA-RH7777 cells has been characterizedwith respect to pyridoxal 5'-phosphate (PlY) levels, and theactivities of pyridoxine (15N)kinase (EC 2.1.7.35) and pyridoxine 5'-phosphate (PNP) oxidase (EC 1.4.3.5). PLP levels (12.4 ±4.4 nvJmgprotein) wereat the lower endof the rangefound for Morris hepatomas,csrcinogen-inducedrathepatomas,andliverfromratsfeda PN-deflcientdiet. PN kinase activity was about one-third of that found in normal ratliver. PNP oxidase appeared to be absent in high-speed supernatants ofhomogenatesprepared from McA-RH7777 cells. The absence of PNPoxidase was supported by enzymatic and immunologicaldata. Thesefindingsresemblethose foundpreviouslyfor Morrishepatoma7777. Incontrast to rat liver, such preparations caused liffle or no release ofvolatile countsuponInCubatiOnwith either rH-C4'IPN or 13H-C4'JPNP.High-speed supernatants of homogenates prepared from both McARH7777 cells and Morris hepatoma 7777 were very much less capablethan similar preparations from rat liver in converting IG-3HIPN to PLPand pyridoxamine 5'-pbosphate. Despite the apparent absenceof PNPoxidase, intact confluentor log-phase McA-RH7777 cells were capableof convertingIG-3HIPNto PlY andpyridoxamine5'-phosphate.Thesefindings are discussed in terms of tumor nutrition and vitamin B@metabolism in a rat hepatomacell line.
INTRODUCTION
Vitamin B6 metabolism has been examined in a variety ofspecies and tissues (1). In mammals, the liver is the organ ofparticular interest owing to its central role in the conversion ofthe various B6 vitamer forms to coenzyinatically active PLP3and its export of PLP to extrahepatic tissues as a PLP-albumincomplex (2). Studies on vitamin B@uptake and metabolism inisolated rat hepatocytes have provided information on theseprocesses (3, 4). The uptake and metabolism of vitamin B6 bysubcellular organelles, i.e., rat liver mitochondria, also havebeen studied (5, 6).
There have been relatively few studies on vitamin B6 metabolism in mammalian cells capable of dividing in vitro. Eaglefirst established in 1956 that vitamin B6 was necessary for thesurvival and growth under tissue culture conditions of HeLahuman carcinoma cells and a mouse fibroblast cell line designated L (7). Subsequently, the uptake of PL and PLP (8) andthe enzymes involved in vitamin B@metabolism were studied incells grown as ascites tumors (9, 10). The effects of vitamin B6deprivation on the activities of PLP-dependent enzymes incultured human fibroblasts have been examined (i 1). In 1982,DiSorbo and Litwack reported that PN at millimolar concentrations was toxic to a rat hepatoma cell line grown in vitro
Received9/21/87; revised 11/20/87; accepted12/1/87.The costsof publicationof this articleweredefrayedin part by the payment
of pagecharges.This article must thereforebe herebymarkedadvertisementinaccordancewith 18U.S.C.Section1734solelyto indicatethis fact.
I Supported by United States Public Health Service Grant 35878 awarded by
the National CancerInstitute, Departmentof Health and Human Services,andUnited States Public Health Service Biomedical ResearchSupport Grant 5429-25/16 awardedbythe UniversityofVermont CollegeofMedicine. A preliminaryreportof this work waspresentedat theJune 1986meetingof the AmericanSociety of Biological Chemists, Washington, DC.
2 To whom requests for reprints should be addressed.
3 The abbreviations used are: PLP, pyridoxal 5'-phosphate; PL, pyridoxal;
(12). A cell line selected for its resistance to the toxic effects ofPN was examined with respect to its uptake of PN and its PLPcontent (13).
Previous reports from our laboratory have dealt with vitaminB6 metabolism in normal rat liver (14), regenerating rat liver(15), transplantable Morris rat hepatomas (16—18),and carcinogen-induced rat hepatomas (19). The present study is concerned with vitamin B@metabolism in cells grown in tissueculture, specifically, the McA-RH7777 rat hepatoma cell linederived from the corresponding transplantable Morris hepatoma (20).
MATERIALS AND METhODS
Materials. B6vitamers and PA werepurchasedfrom Sigma ChemicalCo. (St. Louis, MO). PLP concentrationsof stock solutions weredetermined spectrophotometrically after suitable dilution into 0.1 NKOH using an extinctioncoefficientof 6600 at 388 am (21). 1G-3H1PN hydrochloride(1.4 Ci/mmol), (G-3Hjtryptaminehydrochloride(3.5Ci/mmol), and @c:sliquid scintillationcounting fluid were from theAmersham Corp. (Arlington Heights, IL). N-(S'J-phosphoryridoxyl[G3Hjtryptamine was synthesized by the method of Langhani et a!. (22).[3H-C4'JPN and 13H-C4'JPNPwere from a previous study (14). L-(I‘4C@tyrosine(53 mCi/mmol) wasobtained from New England Nuclear(Boston,MA). Ion exchangeresins(AG I-X8, 200—400mesh,andAG50w-x8, 200-400 mesh) were purchasedfrom Bio-Rad (RockvilleCentre, NY). Tyrosine decarboxylaseapoenzyme used in the quantification of PLP was partially purified from dried Streptococcus faecaliscells (Sigma) by the procedure of Lumeng and Li (23). The sources ofantibodies, immunochemical reagents, and supplies used in an ELISAfor rat PM@oxidase (17) have been publishedpreviously(19) McARH7777 cells(ATCC No. CRL 1601)were obtained from the AmericanType Culture Collection (Rockville, MD). Components used in themedium for the growth of McA-RH7777 cells in tissue culture werepurchased from GIBCO Laboratories (Grand Island, NY). Falconplasticware used in tissue culture work was from Becton DickinsonLabware (Oxnard, CA). House-distilled water was further purified witha Milli-Q water purification system (Millipore Corp., Bedford, MA).All other chemicals and reagents used in these experiments were of thehighest quality commercial grade available.
Cells. McA-RH7777 cells were grown in antibiotic-supplemented(alternatingpenicillin/streptomycinand gentamycin)Ham's F-i2 medium with glutamine,made 14 m@iin sodium bicarbonate,plus 10%fetalbovineserumina humidified95%air/5%CO2atmosphere(37'C);the doublingtimewas approximately24 h.
Cell Extracts. Extracts of McA-RH7777 cells for PLP analyses werepreparedas follows. Confluentcells were washedwith PBS, scrapedfrom the flasks, combined, and washed twice by suspension and centrifugation (10 miii, 1000 g) from PBS. The cell pellet obtained fromone T75 flask was suspendedin 1.0 ml of 80 m@isodium phosphatebuffer (pH 7.4). The suspension was homogenized with ice/watercooling(45 sec followedby 15sec)usinga Polytron PT-10 homogenizer(Brinkmann Instruments, Westbury, NY) at a setting of 10. Trichloroacetic acid (75% w/v) was added to the cell homogenate to a finalconcentration of7.5%. PLP was then quantified as describedelsewhere(14, 16, 18).
Extracts of McA-RH7777 cells used for assaysinvolving PNP oxidane were prepared as follows. Confluent or log-phase cells were scrapedfrom T75 flasks, collected,and washedwith PBS as describedabove.Cell pellets were suspended in 0.25 M sucrose and homogenized asdescribed above. The numbers ofceils per ml of0.25 M sucrose varied,depending on the particular experimental requirements. Homogenates
1080
. . . . 1
Vitamin B6 Metabohsm in McA-RH7777 Cells
Natalie T. Meisler and John W. Thanassi2
DepartmentofBiochemistry, University ofVermon: CollegeofMedicine, Burlington, Vermont05405
were centrifuged for 1 h at 100,000 g. Supernatants obtained in thisfashion were used in analyses for PNP oxidase activity, immunochemically reactive PNP oxidase protein, PN kinase activity, and PLPbinding proteins. PNP oxidase activity was determined by a sensitiveradiochemical method developed by Langham et a!. (22). The ELISAmethodforthe detectionofrat PNP oxidaseusingan antibodyfractionfrom immune rabbit serum has been described previously (17, 19). PNkinase activity was determined by a modification (14) of the procedureof Karawyaand Fonda(24). High-speedsupernatantsofextracts fromMcA-RH7777 cells also were used in experiments, described below, onthe conversion of [G-3HJPNto other B@vitamerforms. Protein wasdetermined by the method of Lowry et a!. (25) using bovine serumalbumin as the reference standard.
Metabolism [email protected] metabolism of[G-3HJPN was studiedusing high-speedsupernatants (100,000 x g-,1 h; 4C) of homogenatesfrom McA-RH7777 cells grown in culture, Moms hepatoma 7777grown in the hindlegs of Buffalo rats, liver of Morris 7777 hepatomabearing Buffalo rats, and homogenates of liver obtained from 75- to100-g male Sprague-Dawley Cr1:CD(SD)BD rats (Canadian BreedingLaboratories, St. Constant, Quebec,Canada). Morris 7777 hepatomabearing Buffalo rats were provided by Dr. Jen-Fu Chiu to whom theauthors are indebted. Methods for the preparation ofhepatoma extractshave been publishedpreviously(16). Rats were housed in the AnimalCare Facility of the University of Vermont College of Medicine wherethey were allowed free access to water and a nonpurified Purina Laboratory Chow 5001 (Ralston Purina Co., St. Louis, MO). Animals werekilled by a crushing blow to the skull. Livers were excised,minced, andhomogenized, as described above, in 9 volumes of 0.25 M sucrose.Incubation mixtures (2.0 ml) contained 0.125 Msucrose, 0.08 MMzincsulfate, 1 mM ATP, 3 @MFMN, and 3.33 @MIG-3HIPN (61.4 @sCi/zmol) in 80 mM potassium phosphate buffer (j,H 7.2; 37T). At thedesired times, 0.1 ml of 100% trichloroacetic acid (w/v) was added tostop the reactions. The acidified incubation mixtures were processedfor the identificationand quantificationof [G-3HJPN-derivedradioactive B@vitamers. These analyses were done, with minor modifications,by the open-column chromatography method describedby Lumeng andLi (3). Authentic B@vitamers and PA (1 mg of each) were used asmarkers in every experiment. They were added to the trichloroaceticacid-treated incubation mixtures. The elution ofthe various B6vitamersand PA were followed by measuring the absorptionat 290 nm ofacidified effluent fractions.
In experimentsdealing with the metabolismof [G-3HJPNby intactMcA-RH7777 cells, the growth medium contained 0.3 @M[G-3HJPN(247 @iCi/Mmolfor confluent cells; 491 pCi/@imol for log-phase growthcells). Confluent cells were incubated with the radioactive vitamer for2 h at 37C; cells in log-phase growth were incubated with the radioactive vitamer for 24 h. Cells were collected and washedessentially asdescribed above. Cell pellets obtained by centrifugation were homogenized with 7.5% trichloroacetic acid and processed for the analysis of(G-3HIPN-derived B6 vitamers by the open-column chromatographymethod. Controls comprised of radioactive vitamer in medium alonewerecarriedthroughthe same procedures.
Determination of Volatile Counts Released from 13H-C4'IPNP andrH-c4'IpN. Volatilecountsweredeterminedbyvacuumdistillationofincubation mixtures and collection ofthe distillate in a cold finger trap.Inexperimentswith[3H-C4']PNP,thesubstratewas3.33 @M(205 @Ci/@imol)in a reaction mixture (2.0 ml) containing liver or McA-RH7777
extracts, 3 zMFMN, 0.18 M Tris, and 80 mrvtpotassium phosphate(pH 7.2). The reactions were stopped by the addition of 0.1 ml 100%trichloroacetic acid. Denatured protein was removed by centrifugationand the volatile counts in the supernatant were determined. Boiledenzyme or no enzyme controls were run; there was no differencebetween them. In experiments with [3H-C4']PN, the substrate was 6.77
@M(87.3 zCi/@smol) in a reaction mixture (2.0 ml) containing liver or
McA-RH7777 extracts, 3 @iMFMN, 0.18 MTris, 80 MMzinc sulfate, 1mM AlP, and 80 mM potassium phosphate (pH 7.2).
RESULTS
PLP Content. The PLP content of confluent McA-RH7777cells was determined in this study to be 12.4 ±4.4 ng/mg
protein (n = 9). Using data obtained in previous studies (14,16, i8, 19), we calculate that the PLP content of Morrishepatoma 7777 grown in the hindlegs of Buffalo rats is in therange of 14 to 19 ng/mg protein. Similarly, the PLP content ofcarcinogen-induced rat hepatomas is in the range of 14 to 22ng/mg protein. These values compare with 37.0 ±2.5 and 19.8±2.7 ng PLP/mg protein for liver tissue obtained from ratsfed PN-sufficient and -deficient diets, respectively.
Activities of Enzymes Involved in the Metabolism of VitaminB6. The conversion of PN to coenzyinatically active PLP conventionally requires the sequential actions of PN kinase (EC2.1.7.35) and PNP oxidase (EC 1.4.3.5) (1). We have measuredthe activities of these enzymes in McA-RH7777 cells to obtaininformation on the capability of these cells to use these twoenzymes to form PLP from PN, and to compare the data withcorresponding data obtained previously for liver and liverderived tumors (14-19). Both PN kinase and PNP oxidaseactivities were determined by radiochemical methods.
In the case of PN kinase, the activity found in supernatantsof homogenates obtained from McA-RH7777 cells was 8. 1 ±1.1 nmol/h per mg protein (n = 7); this is approximately 36%of the activity found in supernatants of homogenates preparedfrom normal rat liver. In comparison, Morris hepatoma 7777grown in vivo has 28% of the PN kinase activity found innormal rat liver (i8, 19). Carcinogen-induced hepatomas andfetal rat liver also exhibit low PN kinase activities relative tonormal rat liver (18, 19).
We could find no PNP oxidase activity in extracts of eitherlog-phase or confluent McA-RH7777 cells using N-[5'J-phosphopyridoxyl[G-3Hjtryptamine as the substrate in a sensitiveradiochemical assay (22). Nor was there any immunologicallydetectable protein in extracts from either log-phase or confluentMcA-RH7777 cells that corresponded to rat liver PNP oxidase;the detection method involved a microtiter plate ELISA technique that used rabbit antiserum to rat liver PNP oxidase (Fig.1).
Release ofVolatile Counts from (3H-C4'JPN and f3H-C4')PNPby Extracts of McA-RH7777 Cells. Any conversion of[3H-C4'JPN or [3H-C4'JPNP to PLP necessarily must involve the releaseofvolatile counts, either as H202 or its decomposition product,H20 (26). Accordingly, release ofvolatile counts over and abovethat of the background provides a measure of the enzymatic
Protein (Micrograms)Fig. 1. Microtiter plate ELISA analysis for PNP oxidase in cytosolic extracts
of normal rat liver and log-phaseMcA-RH7777 cells.The controlwasrun witha normalrat liverextractand nonimmuneserum.Points,averagesof duplicatedeterminations.
oxidation of these substrates. In the case of (3H-C4'JPNP,volatile count release provides a measure of the activity of PNPoxidase (EC 1.4.3.5) and any other enzyme(s) capable of effecting the oxidation of the C4'-hydroxymethyl group of PNP tohigher oxidation states, i.e., carboxaldehyde and carboxyl. Theresults of these experiments are provided in Figs. 2 and 3.When 13H-C4'IPNP was used as the substrate and the enzymesource was an extract of normal rat liver, the release of volatilecounts was linear with respect to both time and protein concentration (Fig. 2). In contrast, extracts of McA-RH7777 cells hadlittle or no such activity. In the case of[3H-C4'IPN and normalrat liver extracts, there was also an increase in volatile countswith time (Fig. 3). Under the same conditions, there was onlya slight release of volatile counts when extracts of McARH7777 cells were used in the incubation mixture.
20
p@ 15
@0
x@10
Ez0.
0
Metabolism ofIG-3HIPN by Extracts of Rat Liver, Host Liverof Moms Hepatoma 7777-bearing Buffalo Rats, Morris Hepatoma 7777 Grown in the Hindlegs of Buffalo Rats, and McARH7777 Cells Grown in Tissue Culture. Fig. 4 provides theresults of experiments dealing with the metabolism of [G-3H]PN by extracts of rat liver as a function of time. At the lowerprotein concentration (2 mg/ml), one readily can infer precursor-product relationships from the data which indicate that thesequence of conversions is in the following order: PN —@PNP—4 PLP. The rate-limiting enzyme in this sequence must be
PNP oxidase. Were that not the case, one would anticipate thatPNP at all times would be either at very low concentrations ornot detectable. The experimental results at the higher proteinconcentration (9.3 mg/ml) reveal that PLP constitutes morethan 80% of the B6 vitamer forms present in the incubationmixture after 1 h. After 2 h, equilibrating transamination reactions convert PLP to PMP, the only other coenzymaticallyactive B6 vitamer form.
In contrast, incubation mixtures containing extracts (7.5 mgprotein/ml) of Morris hepatoma 7777 grown in the hindlegs ofBuffalo rats appear to be compromised in their ability to convertIG-3HIPN to coenzymatically active PLP and PMP (Fig. 5).For example, after a 1-h incubation, the labeled (PN +PNP):(PLP + PMP) ratio is approximately 2.9 for extracts ofMorris hepatoma 7777, compared to the corresponding ratios,using extracts of normal rat liver, of approximately 1.1 (2 mgprotein/ml) and 0.08 (9.3 mg protein/ml) (Fig. 4). Likewise,extracts of confluent McA-RH7777 cells poorly convert (G-3H]PN to PLP and PMP (Fig. 6). Thus, incubation mixtures
80@
@60
@: @jjJII H I H I H
60mm 120mm 180mmFig. 2. Releaseof volatile counts upon incubation of [3H-C4'JPNP (3.33 @sM,
205 zCi/@zmol)with extracts of normal rat liver (L) and McA-RH7777 rathepatomacells(H). Data areexpressedasDPM/ml of watertrappedandaretheaveragesof 3 or 4 determinations.Bars, differencesbetweenthe mean of thecompleteincubationmixturesand the controls(boiledextracts).P valuesfor thesignificance of the differences betweenthe mean for controls and liver were all<0.01 . P valuesfor the significanceof the differencesbetweenthe mean forcontrols and extracts from McA-RH7777 cells wereall <0.05 except for thepointat 120 mm (0.8 mg protein)whereP < 0.15.
Time (minutes)Fig. 4. Metabolism of IG-3HIPN (3.33 aiM, 61.4 @Ci/@&mol)to other B@
vitamers by extracts of normal rat liver. The different B@vitamers were separatedbyion-exchangechromatography(see“Materialsand Methods―).At thosetimesnot designatedby , the amountof proteinin the incubationmixtureswas2 mg/ml. , amountof proteinin the incubationmixtureswas9.3 mg/ml.
80
2
I— 40
>.@0
@20
I30
Time (minutes)Fig. 5. Metabolism of [G-3HJPN (3.33 MM, 61.4 MCi/,@mol)to other B
vitamers by extracts of Morris hepatoma 7777 grown in the hindlegs of Buffalorats. The amount of protein in the incubation mixtures was 7.5 mg/mI.
48
0
Fig. 3. Releaseof volatile counts upon incubation of 13H-C4'IPN (6.77 @M,87.3 ,@Ci/Mmol)with extracts of normal rat liver (L) and McA-RH7777 rathepatomacells(H). Data areexpressedasDPM/ml of watertrappedandare theaveragesof 3 determinations. Bars, differencesbetweenthe meanof the completeincubation mixtures and the controls (boiled extracts). All incubation mixturescontained0.6 mg protein/mI. P valuesfor the significanceof the differencesbetweenthemeanfor controlsandliverwereall <0.01.P valuesfor thesignificanceof the differences betweenthe mean for controls and extracts from McARH7777 cells were P< 0.3 at 60 mm and Pc 0.01 at 150 and 300 mm.
protein compared to 37.0 ±2.5 ng/mg protein for normalyoung adult rat liver. For isolated rat hepatocytes, the PLP
I pa content has been determined to be approximately 80 ng/mgprotein (27). The only mammalian cell lines other than McA
a@ RH7777studiedinthisregardareahumanfibroblastline(11)0 and two rat hepatoma cell lines selected for their resistance to
PL pyridoxine toxicity (1 3). In the former study, the PLP content
I p,p rangedfrom approximately27 ng/mg protein, for cellsgrownin PL-supplemented medium, to approximately 1—2ng/mg
0 i'n protein for cells that were passed 4 times in PL-depleted media.@ PPW In the latter study, PLP concentrations were reported to be
high, approximately 81 and 137 ng/mg protein for the two celllines studied. Thus, McA-RH7777 cells, with 12.4 ±4.4 ngPLP/mg protein, have very low levels of PLP, falling at thelower limits of PLP concentrations determined by us for Morrishepatomas grown in vivo, carcinogen-induced rat hepatomas,and liver from rats fed a vitamin B6-deficient diet (14, 16, 18,19).
In addition to the low PLP content of McA-RH7777 cells,there is a much lower level of activity of the two enzymesresponsible for the conventional conversion of PN to PLP, i.e.,PN kinase (EC 2.1.7.35) and PNP oxidase (EC 1.4.3.5). Bothof these enzyme activities were determined by sensitive andspecific radiochemical methods. In the case of PN kinase, wehave determined that the activity ofthis enzyme in supernatantsofhomogenates obtained from McA-RH7777 cells is only aboutone-third that measured in normal rat liver. More striking isthe apparent absence of PNP oxidase activity which can beattributed to the absence or near absence of the enzyme asdetermined by an immunochemical ELISA detection methodwhich uses antibodies to authentic rat liver PNP oxidase (Fig.1). Similar observations have been made for Morris hepatoma7777 and carcinogen-induced rat hepatomas (17, 19).
In related experiments, we measured the appearance of volatile counts in incubation mixtures containing high-speed supernatants of homogenates obtained from McA-RH7777 cellsor normal rat liver, using as substrates PNP or PN specificallylabeled with tritium at C4' (Figs. 2 and 3). If the conventionalbiosynthetic pathway of PLP from PN, i.e., PN to PNP toPLP, were operative, oxidation of a C4'-hydroxymethyl to a -carboxaldehyde group would necessarily lead to the appearanceof volatile counts in the medium. This is because PNP oxidaseis a flavin-dependent enzyme which produces volatile hydrogenperoxide, a species which would be expected to undergo decomposition in the incubation mixture to oxygen and anothervolatile product, water. The results provided in Fig. 2 clearlyindicate that normal liver extracts, upon incubation with 13H-C4'IPNP, cause the release of volatile counts in a time- andprotein-dependent fashion. In contrast, under otherwise identical conditions, high-speed supernatants of homogenates ohtamed from McA-RH7777 cells cause the release offew volatilecounts. While such counts are for the most part statisticallysignificant relative to the controls, there is no apparent dependency on time or protein concentration for their release. When[3H-C4'JPN is used as the substrate and high-speed supernatants of homogenates of McA-RH7777 cell as the source ofenzyme, there appears to be a time dependency associated withthe small release of volatile counts but it is difficult to assessthe meaningfulness of these data (Fig. 3).
Extracts of normal rat liver efficiently converted IG-3HIPNto PLP via PNP; PLP was then equilibrated with PMP bytransamination reactions (Fig. 4). At longer incubation timesand higher protein concentrations, PLP and PMP constitutedthe majority (approximately 90%) of the six B6 vitamer forms.
1083
p1-p
I
‘s)E0
>
0I'
p,lp
@25@_________ ________@@ @E,i.I@
VITAMIN B6 METABOLISM IN McA-RH7777 CELLS
80
60
40
20
060 120
Time (minutes)
Fig. 6. Metabolism of IG-3HIPN (3.33 MM, 61.4 MCi/@mol) to other B6vitamers by extracts of confluent McA-RH7777 cells. The amount of protein inthe incubationmixtureswas 1.9 mg/mI.
so
CONFLUENT CELLS LOG PHASE CELLS
Fig. 7. Metabolism of (G-'HIPN to other B vitamers by intact confluentMcA-RH7777 cells ((G-3HJPN, 0.3 aiM, 247 @iCi/@@mol;2-h incubation) and byintactMcA-RH7777cellsin log-phasegrowth(IG-'HJPN,0.3MM,491 @iCi/@mol;24-h incubation).
containing extracts of McA-RH7777 cells (I .9 mg protein/ml)show a (PN 4 PNP):(PLP + PMP) ratio of greater than 10after a 1-h incubation with radioactive substrate, approximately9 times the ratio found for normal rat liver extracts under thesame conditions.
Metabolism of tG-3HIPN by Intact McA-RH7777 Cells. Theability of intact McA-RH7777 cells to convert IG-3HJPN to theother B6 vitamer forms was investigated. The results of thesestudies are provided in Fig. 7. It is evident that, despite theapparent absence of PNP oxidase and the demonstrated relatively poor conversion of[G-3HJPN to PLP by extracts of McARH7777 cells, intact cells clearly were capable of producingsignificant amounts of PLP from PN. Thus, after a 2-h incubation ofconfluent McA-RH7777 cells with IG-3HIPN at 37°C,PLP constituted almost one-half of the B6 vitamer forms. After24 h ofgrowth in medium containing radioactive PN, log-phaseMcA-RH7777 cells had more than 85% of their IG-3HIPNderived B6 vitamers in the forms of PLP and PMP.
DISCUSSION
To the best ofour knowledge, these studies represent the firstdetailed report on vitamin B6 metabolism in mammalian cellsgrown in tissue culture. The PLP concentration in McARH7777 cells was determined by us to be 12.4 ±4.4 ng/mg
These results are in general accord with previous studies on themetabolism ofradioactive pyridoxine in mice (28, 29), rats (30—33), perfused rat liver (34), and isolated rat hepatocytes (35; seealso Ref. 36).
The data obtained from similar studies using high-speedsupernatants of homogenates of Morris hepatoma 7777 andconfluent McA-RH7777 cells are plotted in Figs. 5 and 6,respectively. It is readily apparent that extracts of both Morrishepatoma 7777 grown in vivo and the derivative McA-RH7777cell line grown in vitro differed markedly from liver extractswith respect to their metabolism of IG-3HIPN. In the case ofMorris hepatoma 7777, PNP constituted more than 80% of thesix B6 vitamer forms after 15 mm and declined only to about70% after 60 min even when a relatively high protein concentration of 7.5 mg/mI was used in the incubation mixture; PLPand PMP levels rose only very slowly during the 60-mm incubation (Fig. 5). In the case ofextracts from McA-RH7777 cells(1.9 mg protein/mI), PN and PNP constituted 84% of the B6vitamer forms after a 1-h incubation, with PLP and PMPaccounting for a total of only 8% (Fig. 6). Under similarconditions, incubation mixtures containing liver extracts (2 mgprotein/mi) had 52% ofthe B6 vitamers in the forms of PN andPNP and 45% in the forms of PLP and PMP (Fig. 4). Thus,extracts of both Morris hepatoma and McA-RH7777 cells areable to convert [G-3HJPN to radioactive PLP and PMP in smallbut measurable amounts by mechanisms that appear to differsignificantly from those that operate in the parent tissue fromwhich they were derived.
The metabolism of [G-3HJPN by both confluent and logphase McA-RH7777 cells was also studied. The results of theseexperiments are provided in Fig. 7 which reveals that intactMcA-RH7777 cells indeed were able to effect the conversionof PN to PLP and PMP. PLP accounted for almost one-halfof the intracellular B6 vitamer pool in confluent cells incubatedfor 2 h with radioactive PN. In the case of log-phase cellsincubated for 24 h with radioactive PN, PLP and PMP comprised 87% of the radioactive intracellular B6 vitamer pool, adistribution resembling that found in normal rat liver.
If EC 1.4.3.5 PNP oxidase is missing, it follows that theremust be an alternate pathway for PLP synthesis from PN inintact McA-RH7777 cells since both intact confluent and logphase McA-RH7777 cells are capable ofconverting PN to PLP(Fig. 7). It is likely that selection pressures attendant to thegrowth of these cells in culture necessitate the phenotypicproperty of independent production of the coenzymaticallyactive form of vitamin B6 from its coenzymatically inactiveprecursors.
All of the studies to date involving vitamin B@metabolism inMorris hepatomas (16—18)and carcinogen-induced rat hepatomas (19) indicate that rat hepatomas have very low PLPlevels, low levels of PN kinase, and very low to absent levels ofPNP oxidase. The present studies on vitamin B6 metabolism inMcA-RH7777 cells grown in tissue culture are in accord withthese findings but indicate as well that intact rat hepatoma cellsgrowing in tissue culture can make PLP from PN. Therefore,hepatomas arising in vivo may, in principle, obtain their PLPeither by acquisition, as demonstrated for at least one subcellular organelle, i.e., mitochondria (5, 6), or by limited synthesiswhich is quite sufficient for its needs, as demonstrated in thepresent study for McA-RH7777 cells.4
4 Growth ofrat hepatoma-derived HTC cells in suspension culture has enabled
usrecentlyto obtainrelativelylargeamountsof cells.We haveshown,usingthesensitiveradiochemicalPNP oxidaseassayof Langhamet aL (22) with highconcentrationsof proteinin theassaymixture,that thereisa lowbutmeasurable
That cells growing in culture may require only very lowintracellular levels of PLP is demonstrated dramatically by theexperiments of Lipson et a!. (1 1) who showed that reduction ofthe intracellular PLP content of cultured human fibroblasts toapproximately 1 ng/mg protein (3% of control levels) resultedin no evident changes in the morphology or growth rate ofthese cells even after 20 days (four passages) in culture. Conservation of the residual PLP no doubt involves its stabilizationas PLP-protein complexes (2).
Given these observations, it seems unlikely that a generalphysiological deprivation of vitamin B@or administration ofagents which lead to a generalized vitamin B@deficiency will beeffective as nutrition-based interventions in the treatment ofcancer. Such possibilities were explored originally almost 25 yrago and the potential difficulties were appreciated at that time(37, 38). Also conceptually unattractive are approaches thatinvolve long-term oral administration of B@vitamers in veryhigh doses in the hope that there will be selective toxicity ontumors in the host organism. In the first place, high concentrations (mM) ofPN are required to affect cell growth in vitro (12).It is now clearly established that high dietary intake of PN(>500 mg/day; approximately 2 mmol as PN hydrochloride) ishazardous, particularly affecting sensory nerves (for a review,see Ref. 39). Secondly, all B@vitamers are pyridine derivatives.Pyridine itself has been recognized by toxicologists for decadesas an industrial hazard that can cause central nervous systemdepression (40); its maximal acceptable concentration is 5 partsper million (41). Thirdly, owing to their free carboxaldehydegroups, PL and PLP, in particular, have special nondiscriminating chemical reactivity toward lysine side chains in proteins.This property indeed has led to the use of PLP as a proteinlabeling reagent and reporter group (42). Finally, it has beenshown that PN-resistant cell lines will grow under the selectivepressure of high concentrations of PN (13), an adaptation thatis unlikely to be available to normal cells. On the other hand,targeted approaches aimed at specific inhibition of PLP-dependent enzymes that are Critical to cell viability or cell division,such as ornithine decarboxylase, have considerable attraction(43, 44). In addition, vitamin B6 status may exert effects ontumor growth in an indirect fashion, such as by enhancementofimmune system function (45). Finally, there is the possibilitythat there may be specific instances in which there reason tobelieve that vitamin B6 status may exert effects on carcinogenesis and tumor growth (46).
REFERENCES
1. Snell,E. E., andHaskell,B. E. The metabolismofvitamin B. In: M. Florkinand E. H. Stotz (eds.),ComprehensiveBiOchemistry,Vol. 21, pp. 47-71.Amsterdam: Elsevier, 1971.
2. Lumeng, L L, and Li, T-K. Mammalian Vitmein B metabolism: regulatoryrole of protein binding and the hydrolysis of pyridoxal 5'-phosphate instorageand transport in: G. P. TryflateS, (ed.), Vitamin B@Metabolism andRole in Growth, pp. 27-51, Westport, CT: Food and Nutrition Press, 1980.
3. Lumeng, L L, and Li, T-IL Plasmacontent of B@vitamers and its relationship to hepatic vitamin B@metabolism. J. Clin. Invest, 66: 688-695, 1980.
4. Kozik, A., and McCormick, D. B. Mechanism of pyTidOXineuptake byisolated rat liver cells. Arch. Biochem. Biophys., 229: 187—193,1984.
5. LU1,A., Lumeng, L L,and Li,T-K. Metabolismofvitamin B@inratlivermitochondria. J. BiOI.Chem., 256: 6041—6046,1981.
6. Lw, A., Lumeng, L L, and Li, T-K. Transport ofpyridoxine and pyridoxal5'-pbosphate in isolated rat liver mitochondria. J. Biol. Chem., 2S7: 14903-14906,1982.
7. Eagle, H. Relative growth-promoting activity in tissue culture of co-factors
amountof activityin high-speedsupernatantsof HTC cellextracts.In addition,there isasimilarlysmall but measurableactivity in cell pellets. When supernatantsandpelletsarecombined,the activityappearsto be morethan additive.Onepossibility is that a microsomal cytochrome P450 system may be at least in partresponsiblefor thisapparentPNPoxidaseactivity.
and the parentvitamins.Proc.Soc.Exp. BioL Med., 91: 358-361, 1956.8. Pal, P. R., and Christensen,H. N. Uptake of pyridoxal and pyridoxal
phosphate by Ehrlich ascites tumor cells. J. Biol. Chem., 236: 894—897,1961.
9. Ito, IC, Nakahara, I., and Sakamoto,Y. Studieson vitamin B@metabolismof cancercellsand tumor-bearingrat liver. I. Enzyme activitiesrelated tovitamin B6metabolism. Gann, 55: 373—378,1964.
10. Ito, K@,Nakahara, I., and Sakamoto,Y. Studieson vitamin B@metabolismof cancercellsand tumor-bearingrat liver. II. Uptake of pyridoxinederivatives by tumor cells and the liver oftumor-bearing rats. Gann, 55: 378—385,1964.
I 1. Lipson,M. H., Kraus,J. P., Solomon,L R., and Rosenberg,L E. Depletionofcultured human fibroblasts of pyridoxal 5'-phosphate: effect on activitiesof aspartateaminotransferase,alanineaminotransferase,and cystathioninefl-synthase.Arch. Biochem. Biophys., 204: 486-493, 1980.
12. DiSorbo, D. M., and Litwack, G. Vitamin B@kills hepatoma cells in culture.Nutr. Cancer, 3: 216—222, 1982.
13. DiSOrbO,D. M., Paavola,L G., and Litwack, G. Pyridoxineresistancein arat hepatoma cell line. Cancer Res., 42: 2362—2370,1982.
14. Meisler, N. T., andThanassi,J. W. PyridOxinekinase,pyridoxinephosphatephosphatase,and pyridoxine phosphate oxidase activities in control and B@deficientrat liverandbrain.J. Nutr., 110 1965—1975,1980.
15. MeiSIer, N. T., and Thanassi, J. W. Vitamin B@metabolism and its relationto ornithinedecarboxylaseactivityin regeneratingrat liver.J. Nutr., 112:314—323,1982.
16. Thanassi,2. W., Nutter, L M., Meisler, N. T., Corners,P., and Chiu, J-F.Vitamin B6metabolismin Morris hepatomas.J. Biol. Chem.,256:3370—3375, 1981.
17. Nutter, L M., Meisler, N. T., and Thanassi,J. W. Absenceof pyridoxine(pyridoxamine) 5'-phosphate oxidase in Morris hepatoma 7777. Biochemistry, 22: 1599—1604,1983.
18. Meisler, N. T., Nutter, L M., and Thanassi, J. W. Vitamin B@metabolismin liver and liver-derived tumors. Cancer Rca., 42: 3538—3543,1982.
19. Kittler,J. M., Meisler, N. T., and Thanassi,J. W. Azodyc-induced alterationsin vitamin B. metabolismand in pyridoxal5'-phosphatebindingproteinsinrat liver. J. Nutr., 116: 588—598,1983.
20. White, M. F., and Christensen,H. N. CatiOniCacid transportinto culturedanimal cells. II. Transport system barely detectable in ordinary hepatocytes,but active in hepatoma cell lines. J. BiOL Cbem., 257: 4450-4457, 1982.
21. Peterson,E. A., andSober,H. A. Preparationofcrystallinephosphorylatedderivatives ofvitamin B. J. Am. Chem. Soc., 76: 169—175,1954.
22. Langham, L, Garber, B. M., Roe, D. A., and Kazarinoff, M. N. A radiometricassayof pyridoxamine (pyridoxine) 5'-phosphate oxidase. Anal. Biochem.,125: 329—334,1982.
24. Karawya, E., and Fonda, M. Purification, assayand kinetic propertiesofsheepliver pyridoxal kinase. Anal. Biochem., 90: 525-533, 1978.
25. Lowry, 0. H., Rosebrough,N. J., Farr, A. L, and Randall, R. J. Proteinmeasurementwith the Folin phenol reagent. J. Biol. Chem., 193: 265—275,1951.
26. Kazarinoff, M. N., and McCormick, D. B. Rabbit liver pyridoxamine(pyridoxine)5'-phosphateoxidase.Purificationandproperties.J. BiOLChem.,250: 3436—3442,1975.
27. Lumeng, L The role of acetaldehydein mediating the deleterious effect ofethanol on pyridoxal 5'-phosphate metabolism. 3. Chin. Invest., 62: 286—293, 1978.
28. Johansson, S., Lindstedt, S., and Tiselius, H-G. Metabolism of (3H@Jpyridoxine in mice. Biochemistry, 7: 2327—2332,1968.
29. Columbini, C. E., and McCoy, E. E. The utilization of (‘4C)pyridoxinebythe normal mouse.Biochemistry, 9: 533—538,1970.
30. Shane, B. Vitamin B@metabolism and turnover in the ethanol-fed rat. J.Nutr., 112:610—618,1982.
31. Segalman,T-Y. K., and Brown, R. R. The metabolismof 3H-pyridoxineinrat liver and brain.Am. J. Clin. Nutr., 34: 1321—1327,1981.
32. Tryflates, G. P., and Saus, F. L. Metabolism of pyridoxine in the liver ofvitamin B@deficient rats. Biochim. Biophys. Acts, 451: 333—341,1976.
33. Contractor, S. F., and Sham, B. F. Metabolism of [‘4CJpyridoxolin thepregnant rat. Biochim. Biophys. Acts, 230: 127—136, 1971.
34. Mehansho, H., Buss,D. B., Hamm, M. W., and Henderson, L M. Transportandmetabolismofpyridoxine in rat liver. Biochim.Biophys.Acts,631: 112—123, 1980.
35. Lumeng, L, Liii, A., and Li, T-K. Plasmacontentof B@vitamersand itsrelationship to hepatic vitamin B@metabolism. J. Clin. Invest., 66: 688—695,1980.
36. Li, T-K., and Lumeng, L Metabolism of vitamin B@and pyridoxal 5'-phosphate. In: P. Christen and D. E. Metzler (ads.), Transaminases, pp.563-586. New York: Wiley and Sons, Inc., 1985.
37. Rosen,F., Mihich, E., and Nichol, C. A. Selectivemetabolic and chemotherapeuticeffectsof vitamin 8.6 antimetabolites.Vit. Horm., 22: 609—641,1964.
38. Thanassi,J. W., Meisler, N. M., and Kittler, J. M. Vitamin B-6 metabolismand cancer.in: R. D. Reynoldsand J. E. Leklem (eds.), Vitamin 5-6: ItsRole in Health and Disease.,pp. 319—336.New York: Alan R. Liss, Inc.,1985.
39. Cohen, M., and Bendich,A. Safetyof pyridoxine—areviewof human andanimal studies.Tox. Lett., 34: 129—139,1986.
40. The Merck Index, Tenth Edition, pp. 1149—1150. Rahway, NJ: Merck andCo., Inc., 1983.
41. L. J. CasarettandJ. Doull (eds.).Toxicology,the BasicScienceof Poisons,p. 92. New York: Macmillan, 1975.
42. lUttler, J. M., Meisler, N. T., Viceps-Madore, D., Cidlowski, J. A., andThanassi, J. W. A general immunochemical method for detecting proteinsonblots.Anal.Biochem.,137:210—216,1984.
43. Metcalf, B. W., Bey, P., Danzin, C., Jung,M. J., Casara,P., and Vevert, J.P. Catalytic irreversibleinhibition of mammalian ornithine decarboxylase(EC 4.1.1.17) by substrateand productanalogs.J. Am. Chem. Soc., 100:2551—2553,1978.
44. Fozard, J. R., and Koch-Weser, J. Pharmacological consequencesof inhibition of polyamine biosynthesiswith DL-alpha-difluoromethyl ornithine.TrendsPharmacoLSd.,3: 107—110,1982.
45. Gridley, D. S., Stickney,D. B..,Nutter, R. L., Slater, J. M., and Shultz, T.D. Suppressionof tumor growth and enhancementof immune statuswithhigh levelsof dietaryvitaminB@in BALB/c mice.J. Natl. CancerInst., 78:951—959,1987.
46. Birt, D. F., Julius, A. D., Hasegawa, R., St. John, M., and Cohen, S. M.Effect of L-tryptophanand vitamin B deficiencyon rat urinary bladdercancer promotion. Cancer Res.,47: 1244—1250,1987.
