cells and is formed through the decarboxylation of lysine by ornithine decarboxylase. The lack of measurable accumulation of intracellular cadaverine implies ...
THEJ O ~ AOFLBIOUXICAL CHEMISTRY
Vol. 269,No. 10,Issue of March 11, pp. 7412-7418, 1994 Printed in U.S.A.
0 1994 by The American Society for Biochemistry and Molecular Biology, Inc
Biosynthesis and Selective Export of 1,5-Diaminopentane (Cadaverine) in Mycoplasma-free Cultured Mammalian Cells* (Received for publication, July 7, 1993, and in revised form, November 22, 1993)
Leo Hawel 111, Raymond R. Tjandrawinata, Gary H. Fukumoto, and CraigV. ByusS From the Department of Biochemistv and the Division of Biomedical Sciences, Universityof California, Riverside, California 92521-0121
Macrophage-like RAW 264 and H36 hepatoma cells
The diamine cadaverine (1,Fi-diaminopentane) is formed by the decarboxylation of intracellular lysine by the enzyme lysine and an unidentified diamine into the culture medium. decarboxylase (EC4.1.1.18). This enzyme is found in prokaryUnlike putrescine, the unknown compound could be deotes sucha s bacteria (6)and mycoplasma (7). Mammalian cells tected only extracellularly. Analyses of dansylated poly- do not possess a unique lysine decarboxylase enzyme and, unamine standards and mass spectroscopy confirmed that der normal conditions, are not thought to synthesizecadaverthe unknown compound was cadaverine (1,6-diamino- ine. The presence of intracellular cadaverine in cultured mampentane). The cells were free of mycoplasma as evi- malian cells has been generally associated with the denced by a negative result using a probe specific for contamination by mycoplasma (7). Moreover, various DFMO1prokaryotic rRNA. After prophylactic treatments with resistant cultured mammalian cell lines in which ornithine two different mycoplasmacidal agents,the cells continued to export cadaverine. Attempts to “infect” a non- decarboxylase gene amplification led to markedly high consticadaverine-exportingcell line with culture medium and tutive levels of ornithine decarboxylase activity were shown to cell-free lysates proved unsuccessful, establishing that accumulate cadaverine (8-10). Intracellular cadaverine proaction of highly cadaverine was in fact a bona fideproduct of these mam- duction in thesecells could be attributed to the amplified ornithine decarboxylase activity on intracellular lymalian cells. sine (11,121. Kinetic studies performed with purified ornithine Cadaverine export by RAW 264 and H36 cellswas stimulated by lipopolysaccharide and insulin, respec- decarboxylase showed that the V,, of ornithine decarboxylase tively. However, administration of exogenous ornithine for lysine is about 4-fold lower than for ornithine, whereas the caused cadaverine export to decrease significantly with K , for lysine (9.2 mu) is 100 times greater than that of orniconcomitant increases in putrescine export. a-Difluoro- thine (0.09 mM) (11).However, “normal” mammalian cells and methylornithine, a selective inhibitor of ornithine de- tissues have not been reported to contain intracellular cadavcarboxylase, inhibited both cadaverine and putrescine erine even when ornithine decarboxylase is induced to high export. When cellswerelabeledwith[SH]lysine, the levels with hormones or growth factors. great majority of the radioactivity recovered in exThe export of polyaminesfrom inside the cells has been ported polyamines was foundin cadaverine. The cumu- shown to occur in a variety of cultured cells following selective lative data suggested that cadaverine formation may be pressure by certain drugs or growth limitation caused by concaused by the action of intracellular ornithine decar- fluence (5,10,13-15). We report herethat the macrophage-like boxylase upon lysine to produce cadaverine, which is RAW 264 and rat hepatoma H35 cells selectively exported relathen effluxed from the cell with a high degree of effi- tively large amounts of cadaverine in addition to putrescine ciency. into the culture medium. Unlike putrescine, which was also found inside thecells, cadaverine wasnot found intracellularly. With a variety of approaches, the cells were carefully demonThe polyamines (putrescine, spermidine, and spermine) are strated to be free of mycoplasma contamination. The studies highly charged molecules at physiological pH (1).The biosyn- show that cadaverine is a normal metabolite of mammalian following cells and is formed through the decarboxylation of lysine by thesis of the polyamines has been reported to increase stimulation by various growthfactors, mitogens,and hormones ornithine decarboxylase. The lackof measurable accumulation is emuxed through a n induction of the highly regulated enzyme ornithine of intracellular cadaverine implies that cadaverine decarboxylase (1, 2). Various studies using agentsthat inhibit more efficiently from mammalian cells than putrescine by an synthesis of polyamines as well as manipulationsof the genes undefined mechanism. grown under serum-freeconditions exported putrescine
encoding polyamine synthetic enzymes have beenemployed to reduce polyamine abundance in many different tissues (3-5). These studies haveunderscored the importanceof polyamines in the growth processes of mammalian cells.
MATERIALS AND METHODS Chemicals and Supplies-Cell culture media were from Life Technologies, Inc. Dansyl chloride and polyamine standards were obtained from Fluka. All other chemicalsandsupplieswerepurchasedfrom either Sigma or Fisher. Cell Culture-The mouse monocyte-macrophage RAW 264 cell line was obtained from the American Type Culture Collection and certified as mycoplasma-free. The Reuber H35rat hepatoma andRAW 264 cells were grown to 25-35% confluence in Dulbecco’s modified Eagle’s me-
* This work was supported by National Institutes of Health Grant ES06128 and Department of Energy, Ofice of Energy Storage and costs of publication of Distribution, Contract DEAlAO-85CD76260. The this article were defrayed inpart by the payment of page charges. This in accordance article must thereforebe hereby marked “aduertisement” The abbreviations used in this manuscript are: DFMO, difluorowith 18 U.S.C. Section 1734 solelyto indicate this fact. To whom correspondence shouldbe addressed:Division of Biomedi- methylornithine; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl; cal Sciences, University of California, Riverside, CA 92521-0121. Tel.: DMEM, Dulbecco’s modified Eagle’s medium;LPS, lipopolysaccharide; HPLC, high pressure liquid chromatography. 909-787-4535; Fax: 909-787-5504.
7412
Mammalian Cells Synthesize and
Effectively Export
Cadaverine
7413
sham Corp.) were added to the plates. After 2 h in the radioactive dium (DMEM) (pH 7.4) plus 10% fetal bovine serum and 50 IU/ml penicillin, 50 pgml streptomycin, and 10 pg/ml gentamicin. Twelve media, the cells were washed thoroughly with phosphate-buffered saline before being harvested. Proteinfrom the [3Hlleucine-treated plates (RAW 264) or 24 h (H35)before starting an experiment, the media were replaced with fresh serum-free DMEM, allowing the cells to complete was precipitated, counted, and found to incorporate the labeled leucine for control cultures left in DMEM the cell cycle and become stationary in Go (at 5 5 4 5 % confluence). In at rates identical to those measured by culture this manner the RAW 264 cells have been shown to respond in a syn- (thus indicating that the cells were not artificially influenced in theRPMI with diminished lysine). The lysine- and ornithine-treated chronous manner to LPS inducing ornithine decarboxylase activity and plates were analyzed for polyamines following dansylation via HPLC, DNA synthesis (16). To begin an experiment, the media were again from the reverse phase column (0.75-ml fractions) were replaced with fresh serum-freeDMEM containing the compounds indi- and the eluants cated in the figures (LPS, insulin, ornithine, DFMO). Over the experi- collected and analyzed for radioactivity. Additionally, blank plates (no c3Hllyor mental time courseof 24 h, in the H35cell cultures, cell numbers (and cells) with culture media were also treated with [3Hlornithine protein) increased by 20% in the untreated and ornithine-treatedcul- sine and similarly analyzed to ensure that they did not contain any radioactive polyamines(from nonenzymatic decarboxylation reactions). tures and by 50%in cultures receiving insulin. All cells were grownat to free of polyamines and 37 “C, 5% C02, 100% humidity. Under these conditions, both cell lines All blank plates,upon analysis, were found be did not contain any radioactivity in the fractions where polyamines remained highly viable (>95% as judged by trypan blue exclusion) would be expected to be eluted fromthe HPLC. throughout a 36-h period. Use of Mass Spectroscopy forIdentification of the Unknown as The cells were subcultured no more than 10 times a t a dilution of 1:40 Cadaverine-Didansylputrescine and the unknown dansylated peak using 0.05% trypsin with 0.02% EDTA. Both cell lines were testedfor mycoplasma with a commercial kit capable of detecting as few as lo5 were isolated via reverse phase HPLC and concentrated to dryness by by the National Center colony-forming unitdml ofMycoplasma hominis (Gen Probe, San Diego) vacuum centrifugation. Analysis was performed For Toxicological Research (U.S. Department of Health and Human and were found to be mycoplasma-free. However, the stock cultures Services) Jefferson, AR. Fast atom bombardment-mass spectroscopy were further submitted to treatment with B-M Cycline (3 weeks), deion scanswereobtained from aFinnigan rivatives of pleuromutilin and tetracycline (Boehringer Mannheim) fol- spectraandproduct lowed by Mycoplasma Removal Agent (for 1week), a derivative of 4-oxo- MAITSQ-70 using an Ion Tech atom gun. Sample ionization was genquinoline-3-carboxylic acid (ICN Flow, Costa Mesa,CA) a s described by erated by xenon atoms accelerated t o %lo keV with thioglycerol as a ion spectra the manufacturers. After treatment with these agents, bothcell lines matrix. Fast atom bombardment full scans and product still tested negative for mycoplasma, were frozen in liquid nitrogen,were andtaken for all samples. served a s “stock” cultures for future experiments. RESULTS Determination of Polyamine Levels-To analyze intracellular polyamines, cells were rapidly( 4 s) flooded with a stream of ice-cold isoIdentification of Unknown Peak during Polyamine Analysis tonic phosphate-buffered saline and scraped freefrom the plate with a of Culture Media from Mouse Macrophage-like RAW 264 and rubber policeman. The pelleted cells were then treated with perchloric Rat Hepatoma H35 Cells-Intracellular polyamine analysis of acid, the polyamines derivatized with dansyl chloride, and submitted to H35 cells showed the presence of putrescine, spermidine, and analysis via reversed phase HPLC a s described in detail (17, 18). To determine exported polyamine levels, the culture media were ap-spermine incontrol cultures (Fig. 1). However, when the control culture medium samples were analyzed for extracellular polyplied to minicolumns containing 0.7 ml of Bio-Rex 70 in thesodium form (Bio-Rad) at room temperature asdescribed in detail (18). The column amines, a large, unidentified diamine peak was found, in adwas then washed with water followed by 0.145 M NH40H. The poly- dition to those of putrescine, spermidine, and Nl-acetylspermiamines were then eluted with 4.0 M NHIOH and evaporated to dryness dine. This unidentified peak (peak B, Fig. 1) appeared about 1 in a Speed-Vac (Savant). The dry pellets were resuspended and derivamin after thedldansylputrescine peak during C18 chromatogtized (as above) with dansyl chloride prior to analysis via reverse phase raphy on the HPLC. Prior to analysis, the media samples were HPLC. Polyamine levels were determined by comparison with appropriate processed using a cation exchange procedure, which recovers standards and monitored for recovery with an internal standard.Pro- only compounds with net charges of 2’ or greater (17). This tein values were determined by the method of Bradford (19). Polyamine procedure eliminates approximately 90% of the total dansyl concentrations were expressed in units ofnmoVmgof total cellular chloride-reacting materialpresentinculture medium (17). protein, which was directly proportional to cell number. We report the results in this fashion to analyze ratesof polyamine exporton a percell Comparison with known polyamine and acetylated polyamine of the unknown basis. This is important because the number of cells increased differ- standards revealed that the retention time entially, depending upon whether the cultures were treated with insu- compound corresponded to cadaverine only. When the media lin, ornithine, or LPS (see cell culture). Polyamine levels below 0.1 samples were spiked with increasing amounts of cadaverine nmollmg of total cellular protein were too low t o be quantitated accu- standard, the area underneath only the unknown peak became rately. The HPLC method(18),a modification of the original (17), was proportionally larger. capable of clearlyresolving the polyamines putrescine, cadaverine, Analysis of the Unknown Compound by Fast Atom Bombardspermidine,andspermine,as well as theacetylatedderivatives (N’-acetylputrescine,W-acetylspermidine,W-acetylspermidine, and ment-Mass Spectroscopy-Fast atom bombardment-mass spec(Nl-acetylspermine). troscopy was chosen as a method for confirming the identityof Determination of Intracellular AminoAcid Concentrations-Half of the unknown peak as cadaverine because of its ability to difthe cell lysate from intracellular polyamine analysis was neutralized ferentiate ion fragmentation processes. Full scan mass spectra with potassium bicarbonate and loaded a onto minicolumn containing1 ml of Bio-Rex 70 in thehydrogen form (Bio-Rad) as described previously showed that theunknown compound had equivalention forma(18).The column was then washed with waterfollowed by 0.15 M am- tion processes as did the didansylcadaverine standard: rnlz 569 monium hydroxide. The basic amino acids were eluted with0.5 M am- and 591 for the (M + H)’ and (M + Na)’ molecular ions, respecmonium hydroxide. The eluant was then lyophilized, resuspended in 0.1 tively. Product ion spectra of the unknown demonstrated ions M sodium bicarbonate, and derivatized with Fmoc C1 (N-(g-fluorenyl)a t mlz 84 and 101, both of which correspond to the cadaverine methoxycarbonyl chloride) prior to analysis by reverse phase HPLC and backbone. The putrescine backbone ions at mlz 70 were not quantitation againstk n o w n standards. As with the polyamines, amino detected in the unknown. The findings established that the acid values were expressed in unitsof nmoVmg of cellular protein. unknown diamine present in the culture media was cadaverRadiolabeling of Exported Polyamines-Reuber H35cellswere ine. grown in DMEM as described above before being stimulated withlo-’ M insulin. After 3 h, the media were removed and replaced with 10 ml Negative Evidence for Foreign Organisms in the Mammalian of insulin-containing RPMI 1640 specially formulated t o contain all Cell Cultures-It had been shown previously that cultured amino acids at the same concentrations found in DMEM except lysine, mammalian cells contained cadaverine only when they were which was adjusted 50 to p~ (versus 500 p~ in DMEM). This adjustment infected by mycoplasma (7). The presence of extracellular caof media lysine allowed us to maintain a higherspecific activity of the us to speculate daverine in theculture mediaprompted tritiated lysine and to employ a short labeling period. Seventy pCi of PHIlysine (83 CUmmol, Amersham Corp.), 40 pCi of [3H]ornithine (46.5 whether thecells were contaminated by mycoplasma. The RAW CUmmol, DuPont NEN), or 40 pCiof [3Hlleucine (150Ci/mmol, Amer- 264 and H35 cells did not show any hybridization of tritium-
7414
Mammalian Cells Synthesize and Effectively Export Cadaverine INTRACELLULAR
EXTRACELLULAR D
E
A
>-
L cn
FIG.1. Comparison of intracellular and extracellular polyamines via HPLC analysis. Reuber H35 cells were grown and stimulated with insulin as described under ‘Materials and Methods.” After 8 h, the cells and culture media werecollected and analyzed for polyamine content. Polyamines are indicated by the letters above the peaks. A, putrescine; B , cadaverine; C , internal standard (1,7-diaminoheptane);D , spermidine; E , spermine. The asterisks above the small peaks at 19 and 26 min are, respectively, trace amounts of Nl-acetylspermidine and Nl-acetylspermine.
Z W I-
E
z C
W 0
Z W 0
tn W
CY
0
3
J
LL
A
~
1
6
1
I
I
I
;I
12 18 24 3 0 3 6
TIME I N MINUTES labeled single-stranded DNA probe specific for the rRNA of mycoplasma (see “Materials and Methods”). To minimize the possible presence of mycoplasma, the RAW cells were grown in three cycles (one cycle/week) of B-M Cycline and then in the Mycoplasma Removal Agent for 1week according to directions supplied by the manufacturers. Stock cultured cells treated in this manner stillexported cadaverine yet continued to show a negative result upon retesting for mycoplasma rRNA. To explore further thepossibility that theexported cadaverine originated from a contaminating microorganism and not from the H35or RAW 264 cells, attempts were made to“infect” a similar, but non-cadaverine-exportingcell line, thereby giving it theability to export cadaverine.To do this, theH35 cells were grown for three passages (i.e. 3 weeks) in the absence of any antibiotics to allow any foreign organisms that were present time to proliferate. The culture medium was then collected from the confluent H35 cultures and placed on growing cultures of RL-9 cells (another rat liver line that only minimally exported cadaverine). Additionally, to investigate whether a virus was responsible for the observed cadaverine, the H35 cells were lysed by sonication, and the resulting lysate was then filtered( t o remove cell debris) before being added to other presence of H35 cell medium RL-9 cultures. After 2 days in the or H35 cell-free lysate, all RL-9 cultures were stimulated with insulin for 8 h and the media analyzed for polyamines as described under “Materials andMethods.” The control cultures of RL-9 cells (Fig. 2 A ) exported slight amounts of putrescine andspermidine but only trace amounts of cadaverine after 8 h. The presence of the culture media or lysate from the H35 cells (Fig. 2, B and C) could not enhance the ability of the RL-9 cells to export cadaverine beyond what was observed in the control cultures. When the intracellular samples were analyzed, no cadaverine was detected in any of the RL-9 cultures (data not shown). The resultsof these experiments are additional evidence that the cadaverine exported by
5 C
I
I
I
I
I
6 12 18 24 30 3 TIME I N MINUTES
H35 and RAW 264 cultures wasnot produced by some contaminating agent, but instead by the RAW 264 and H35 cells directly. Radiolabeling of Cadaverine with PHILysine-To determine if lysine wasthe immediatemetabolic precursor to theexported cadaverine, a short term labeling experiment was conducted using r3H1lysine (see “Materials andMethods”). When [3Hlornithine was added to the H35 cells, the only radioactive polyamine detected was putrescine (data not shown). In cultures receiving 13H]lysine(Fig. 3), the only substantial peaks of radioactivity seen inpolyamines were those observed between 18 and 20 min after injection. These radioactive fractions were associated exclusively with the elution of cadaverine from the HPLC column. Furthermore, when the same sampleswere rerun and each peak was manually collected in its entirety, we found no significant radioactivityin anypolyamine peak except cadaverine (data notshown). These findings suggestedthat the precursor of the exported cadaverine was lysine. Time-dependent Accumulation of Extracellular Cadaverine following the Addition of LPS and Insulinin RAW 264 and H35 CeZls-Under transient serum-free conditions both the RAW 264 and H35cells exported cadaverineinto the culture medium constitutively (Fig. 4, A and B). Within a 24-h period, the amount of cadaverine emuxed by the RAW 264 cells increased from zero (fresh mediumcontained no cadaverine)to 2.5 nmoVmg of protein (Fig. 4A)with the most rapid rate of emux occurring during the initial8-h period. The control cultures of H35 hepatoma cells exported cadaverine constitutively to a maximum level of 1.7 nmoVmg of protein at 8 h and subsequently ceased to efflux the diamine. Upon the administration of LPS, a differentiation factor and activator of macrophages and the RAW 264 cells (16),the level of extracellular cadaverine accumulation increased by almost 7-fold to a value of 15.2 nmoVmg of protein in comparison to the controls (untreated cultures) at the 24-h point (Fig.4A).In thepresence of insulin,
Mammalian Cells Synthesize Effectively and Export Cadaverine A. CONTROL
7415
“/I
1
FIG.2. Attempts to induce cadaverine export in RL-9cells by infecting them with culture media or cell lysate from cadaverine-exporting Reuber H35 cells. HPLC tracings from: panel A, RL-9 culture medium (control); panel B, culture medium from RL-9 cells grown for 2 days in medium conditioned by H35 cells; panel C , culture medium
r
C
from RL-9 cells treated with the cell lysate from cadaverine-exportingH35 cells. Polyamines are indicated by the letters above the peaks. A, putrescine; B , cadaverine; C , internalstandard; D , spennidine. The asterisks above the small peaks at 19 and 26 min are, respectively, trace amounts of N’-acetylspennidine and N’-acetylspermine.
I
0
6 12 18 24 30 36
I N M I N UTTIEMSE
T I M E I N M I N UTTIEMSE
I N MINUTES
HOURS AFTER MEDIA CHANGE
TIME I N
B
MINUTES
I
/c;
FIG.3. [sH]Lysine radiolabeling of cadaverine exported from H36 cells. Reuber H35 cells were grown and treatedas described under “Materials and Methods.” After 2 h of [3Hllysine labeling, cells and media were harvested and analyzed for polyamines via HPLC (solid line).Additionally, the HPLC eluant was collectedin 45-s fractions and analyzed for radioactivity (stippled bars).The figure is a composite of the culture medium HPLC tracing and the corresponding 3Helution profile.
a mitogenic stimulus for the H35 cells (20), the efflux of cadaverine was also stimulated above the constitutive level to a value of 5.5 nmoVmg of protein at 24 h (Fig. 4B). It should be noted LPS or insulin for 24 h, that even in the cultures treated with no detectable intracellular cadaverine was observed. Reduction in Cadaverine Exportfollowing Administration of Exogenous Ornithine in the RAW 264 and H35 Cells-It has been shown previously that ornithine plays an importantrole of polyamine metabolism and may be limiting in the regulation in cells and tissues in termsof saturating ornithine decarboxylase with substrate (20-23). In addition, recent studies reported that ornithine was able to trans-stimulate putrescine export in bacteria (23). We have observed recently that ornithine markedly stimulated putrescineexport in both the RAW 264 and H35 cells (data not shown). For these reasons, the effect of exogenous ornithine on cadaverine efflux was investigated. “he administration of ornithine (1.0 mM) to themedia of the
INSULIN UNTREATED INSULIN + ORNITHINE ORNITHINE ONLY
,a=8=6
0 6 12 24 18 HOURS AFTER MEDIA CHANGE
FIG.4. Export of cadaverine from RAW 264 (panel A ) and H35 (panel B ) cells following treatment with LPS, insulin, or 1 m~ ornithine. Reuber H35 and RAW 264 cells were grown as described under “Materials and Methods.” In panel A, RAW 264 cells werestimulated by LPS k ornithine; in panel E , H35cellsreceived insulin 2 ornithine. At the indicated times, cells and culture media were collected and analyzed by HPLC for polyamine content. 0, untreated; 0, LPS only; insulin only; 0 , ornithine only; A, ornithine + LPS; A,ornithine + insulin. The data are represented as the means of triplicate plates. Error bars are 2 S.E. In some cases, error bars are not shown because they are smaller than the symbols used.
.,
RAW 264 cells led to a marked decrease in theconstitutive rate of exported cadaverine (Fig.4A).After 4 h in cultures receiving exogenous ornithine alone, cadaverine efflux ceased. Ornithine
Mammalian Cells Synthesize and EffectivelyExport Cadaverine
7416 2.0
8 0 .
CADAVERINE PUTRESCINE
TABLEI Intracellular basic amino acid levels of RAW 264 and H35 cells in the presence and absence of ornithine Ornithine (0.1 m)was administered for a period of 4 h, and intracellular amino acid levels were determined as described under “Materials and Methods.” Data are represented as the mean f the S.E. of triplicate determinationsfrom three different dishes. ~
RAW 264 cells
+ L O m ornithine Control
Control
nmol Img protein
,
0.0’
P
I
I
I
-u I
0 0.1 5 1 10 EXOGENOUS ORNITHINE (rnM)
Arginine
L0
FIG.5. Cadaverine and putrescine exportfrom RAW 264 cells withvariousexogenousornithineconcentrations. Cells were grownand stimulated with LPS as describedunder“Materialsand Methods.” After 4 h in the presence of ornithine at the indicated concentrations, cells and media were collected and their polyamine contents determined.Cadaverine: 0,untreated; 0 , LPS. Putrescine: 0, untreated; LPS. The data are represented as means of triplicate plates. Error bars are 2 S.E. In some cases, error bars are not shown because they are smaller than the Symbols used.
0.30 f 0.03
Ornithine 0.062 0.00
0.46 2 0.03
Lysine
0.36 f 0.02
0.30 2 0.04
0
+ L O m ornithine
2.0 2 0.5 0.25
UNTREATED LPS OFMO LPS t DFMD
1’
~
nmol lmg protein
0.31 2 0.02
.,
also inhibited completely the stimulationof cadaverine export brought about by LPS. In the H35 hepatoma cells, ornithine caused a marked reduction in cadaverine exportwithin 4 hand then inhibited efflux completely (Fig. 4B). Additionally, the administration of ornithinefurther prevented theinsulinstimulated increase in cadaverine export seenin theH35 hepatoma cells and reduced exportto levels observed with ornithine alone. To determine the degree of inhibition of cadaverine export brought about by ornithine, a 2-log range of exogenous ornithine (from 0 to 10 mM) was administeredto the RAW 264 cells with or without LPS. In the absence or presence of LPS, the level of cadaverine efflwed dropped 5-7-fold in a dose-dependent manner as themedia ornithine increased from 0 to 10 mM (Fig. 5). In this regard, 1mM exogenous ornithine wassufficient to inhibit cadaverine export maximally. In the same culture medium, putrescine exportwas stimulated by exogenous ornithine in a similar dose-response manner in the presence or absence of LPS (Fig. 5 ) . Comparisons of the Levels of Basic Amino Acids in RAW 264 and H35 Cells-The intracellular concentrations of ornithine and lysine weremeasured in culturesincubated with andwithof out exogenous ornithine to determine the relative saturation ornithine decarboxylase with these 2 amino acids inside thecell (Table I). In theRAW 264 and H35 cells, the levels of intracellular lysine were 5- and 10-fold higher, respectively, than the level of ornithine. Thepresence of exogenous ornithine (1.0 n m ) in theRAW 264 and H35 cells increased intracellular ornithine levels 5-8-fold within a 4-h period but did not change the amount of either lysine or arginine. Changes in theLevels of Cadaverine Exported in Response to a-DFMO in the RAW 264 and H35 Cells-The effect of a-DFMO, a highly selective enzyme-activated irreversible inhibitor of ornithine decarboxylase, on extracellular cadaverine accumulation was investigated. In the presence of DFMO the constitutive export of cadaverine was reduced to a value 2.5fold lower than the control cultures (i.e. about 0.7 nmol/mg of protein) within the24-h period in both RAW 264 and H35 cells (Fig. 6, A and E?). Furthermore, DFMO dramatically inhibited the stimulation of cadaverine export brought about by LPS in the RAW 264 cells and by insulin in the H35 cells (Fig. 6, A and B ) . Under the same conditions, DFMO markedly reduced the accumulation of intracellular and exported putrescine in both cell lines (data notshown). Given the high degree of selectivity of DFMO for ornithine decarboxylase under in vivo conditions
~~
H35 cells
1.9 f 0.3
* 0.05
1.3 ? 0.2
2.4 2 0.5
2.2 f 0.2
P
P
d
024 618 12 HOURS AFTER MEDIA CHANGE
I
I
A INSULIN 9 - 0 UNTREATED A INSULIN + OFMO 0 OFMO
6-
HOURS AFTER MEDIA CHANGE
FIG.6. Export of cadaverine from RAW 264 (panel A) and H3:5 (panelB ) cells stimulated with LPS or insulin 0.1 m~ a-DFM(.1. a W 264 and ReuberH35 cells were grown andstimulated with LPS or insulin in the presence or absence of 0.1 m a-DFMO. Atthe indicated times, cells and culture media were harvested and analyzed for polyamines using HPLC. 0, untreated; 0, DFMO only; 0, LPS; LPS + DFMO; A, insulin; A, insulin + DFMO. The data are represented as means of triplicate plates. Error bars are 2 S.E. In some cases, error bars are not shown because they are smaller than the symbols used.
.,
and the inhibitory effect of exogenous ornithine, the datasuggest thatcadaverine biosynthesis by these mammalian cells is caused by the enzymatic decarboxylation of lysine by normal or altered forms of ornithine decarboxylase inside the cells (see “Discussion”). Comparisons of the Levels of Exported Cadaverine and Putrescine in RAW 264 and Reuber H35 Cells-Table I1 summarizes the levels of exported putrescine incomparison to cadaverine in the presence of either LPS, insulin, ornithine, or the combinations indicated within the 24-h period of culture. Cultured RAW 264 and H35 cells grown under serum-free conditions selectively exported putrescine along with cadaverine. (The RAW 264 cells also exported small amountsof N1-acetylspermidine.) The constitutive levels of exportedputrescine were higher in the RAW 264 cells than that in theH35 cells, whereas thecontrol levels of cadaverine exported by both cells were about the same.At 15.2 nmoVmg of protein, LPS-stimulated RAW 264 cells exported cadaverine to the highest level
Mammalian Cells Synthesize and
Effectively Export
Cadaverine
7417
TABLEI1 Comparison in the levels of exported putrescine and Cadaverine in the culture media of RAW 264 and H35 cells LPS was used to stimulate RAW 264 cell, and insulin was used tostimulate H35 cells. All treatments were administeredfor a period of 24 h. Data are represented as the mean of triplicate determinations from three different dishes * the S.E. RAW 264 cells Putrescine
Cadaverine
H35 cells N'-Acetylspermidine
nmol I mg protein
Controls LPS or insulin Ornithine LPS + ornithine or insulin
+ ornithine
10.8 f 0.1 17.2 0.6 17.6 0.2 20.6 2.8
* *
2.5 2 0.2 15.2 2 .09 0.80 2 0.04 2.2 2 0.1
Putrescine
Cadaverine
nmol lmg protein
0.65 2 0.01 1.56 2 0.05 0.55 2 0.04 2.23 2 0.06
2.0 2 0.2 7.2 2 0.1 6.0 2 0.5 30.0 2 1.2
1.6 2 0.3 5.5 2 0.2 0.25 2 0.04 0.50 2 0.05
A X ) . These observations support the conclusion that the exported cadaverine was not produced by a contaminating organism, but instead by the cultured mammaliancells. The metabolic precursor of the effluxed cadaverine was established when r3H]lysine (but not [3H]ornithine) selectively labeled the cadaverine exported by the H35 cells (Fig. 3). The specific activity of this exported cadaverine was determinedto be 8,180 cpdnmol, i.e. lower than the specific activity of the extracellular lysine pool. However, the lower specific activity was to be expected considering that the radiolabeling experiments were performed for a short timeperiod and did not allow for equilibration between the intracellular and extracellular lysine pools. Even though the exported cadaverine appeared to be decarboxylated lysine, it has been well established that a lysinespecific lysine decarboxylase does not exist inmammalian cells (11). However, ornithine decarboxylase has been shown to deDISCUSSION carboxylate lysine in vitro (11, 12) but at a slower rate than is The presence of an unidentified polyamine in media from seen with ornithine as a substrate (V,,, of 4.6 versus 16.9 cultured mammalian cells was a consistent observation in our pmovmgh (11)). Additionally, the K,,, of ornithine decarboxylinvestigations of polyamine export. The cumulative data sup- ase for lysine is about 100 times greater than theK , for orniporting the identification of the unknown peak as cadaverine thine (9,200 versus 90 PM (11)). Given that the ability of the RAW 264 and H35 cells to produce cadaverine was severely include the following. (i) The pre-HPLCclean-upprocedure eliminated greater than 90% of the total dansyl-reacting ma- diminished when DFMO (a specific and selective inhibitor of terial present in the culture media (18). The unknown was ornithine decarboxylase) was administered (Fig. 6, A and B ) , bound tightly by this cation exchangeresin underconditions in we believe it is likely that the exported cadaverine was prowhich only molecules with at least a 2' charge remainedbound. duced by ornithine decarboxylase acting upon lysine. (ii) During the HPLC analysis procedure the unknown comSuch a conclusion is supportedby the dose-dependent inhibipound coeluted with a didansylcadaverine standard that was tory effect of ornithine upon cadaverine efflux (Figs. 4 and 5). capable of clearly separating all known polyamines and their Since (i) the intracellular ornithine concentrations for the RAW acetylated derivatives. (iii) Mass spectroscopic analysis of the cells and theH35 cells were low relative to the K,,, of ornithine unknown (purified from a biological sample) unambiguously decarboxylase for ornithine (measured in Table I and calcuidentified it as didansylcadaverine. As a positive control, mass lated to be 5 and 21 PM, respectively); and (ii) intracellular spectroscopy of the putrescine peak (purified from the same lysine was 10-fold higher than ornithine(Table I), some decarbiological sample) correctly identified it as didansylputrescine boxylation of lysine by ornithine decarboxylase might be ex(data not shown). (iv) Finally, the unknown was specifically pected to occur based upon kinetic grounds alone. As intracelradiolabeled by incubation of the culturedcells with [3H]lysine lular ornithine concentrations increased with the but not by [3Hlornithine (Fig. 3). administration of exogenous ornithine, this aminoacid would Otherresearchershave previously noted the presence of effectively compete with lysine as a substrate for ornithine trace levels of cadaverine in mammalian tissues that have un-decarboxylase. Extending thisline of reasoning, we have made dergone ornithine decarboxylase gene amplification (8,9,24, a careful kinetic analysis of the relative velocities for putres25). However, the detection of cadaverine in normal cultured cine and cadaverine biosynthesis by ornithine decarboxylase in mammalian cells has historically prompted the suspicion of the RAW and H35cells. In theabsence of exogenous ornithine, mycoplasma contamination (7).Even though the H35 and RAW 15-50-fold more cadaverine relativeto putrescinewas observed 264 cultures were found to be negative for prokaryotic ribo- to be synthesized in these cells than was calculated to be prosomal RNA in our hands, it was possible that this test was duced using the measured substrate concentrations and pubincapable of detecting the presence of all strains ofmycolished kinetic constants for ornithine decarboxylase (11, 26). plasma. However, following treatment with two effective my- Although there are many assumptions that limitconclusion the coplasmacidal agents, both cell lines still actively exported ca- to be drawn from such analyses, furtherinvestigations into the daverine. Additionally, cells infected with mycoplasma mechanism of cadaverinebiosynthesis inmammalian cells generally have significant levels of intracellular cadaverine (73, seems warranted. which was not the case in cultures of the RAW and H35 cells. A Regardless of the mechanism by which cadaverine was synnon-cadaverine-exportingcell line could not be infected and thesized, it was detectable at only trace levels within the cell. induced to exportcadaverine by the application of media or cell The fact that cadaverine was found exclusively in the extracellysate from H35 cells that actively exported cadaverine (Fig.2, lular medium indicated that in the H35 and RAW 264 cells
seen inboth cells. The additionof exogenous ornithine reduced the amount of cadaverine exported while simultaneously increasing the amount of putrescine effluxed from both cell lines. The greatest increase in putrescineexport (as compared with control values), occurred in the H35 cells when bothinsulin and exogenous ornithine were present in culture media. Ornithine was significantly more effective in stimulating putrescine export in the H35 cells than in the RAW cells. In this regard, ornithine appeared tobe less a limiting factor in the export of putrescine from the RAW 264 cells than from the H35 cells even though the measured basallevel of ornithine in theRAW cells was lower (Table I). It should be further noted that the ability of cadaverine in of insulin andLPS to stimulate the appearance the culture medium may be caused by increases in ornithine decarboxylase activities seen in these cells and not by any direct increase in a specific export process.
7418
Mammalian Cells Synthesize and Effectively Export Cadaverine
cadaverine, formed intracellularly, was either exported with a high degreeof efficiency or rapidly convertedmetabolically into some other product. Cadaverine hasbeen shown to be capable of replacing putrescine asa precursor in polyamine biosynthesis (27, 28). The lack of a measurable steady-state level of cadaverine is a n interesting observation, considering that putrescine wasalso emuxed from the cells yet achieves significant intracellular levels. The mechanism by which the RAW 264 and H35 cells efflux cadaverine and putrescine is unknown. A nonsaturable process (diffusion?) has been implicated in theefflux of putrescine from cultured Neurospora cells (29,30). It isalso conceivable that theexport of these diamines involves an integral plasma membrane-associated protein(s).In this regard, we are attempting to characterize more fully these export processes using isolated inside-out membrane vesicles. Acknowledgments-We gratefully acknowledge the efforts and expertise of Drs. M. Paul Chiarelli and Jackson0. Lay at theFood and Drug Administration, National Center forToxicological Research, Jefferson, AR. Their analysis of the mass spectroscopy spectra and subsequent interpretation greatly aided the identification of the unknown compound as cadaverine. REFERENCES 1. Tabor, C. W., and Tabor, H. (1976)Annu. Rev. Biochem. 45,285-306
2. Pegg, A. E., and McCann, P.P. (1988) IS1 Atlas of Science: Biochemistry, pp. 11-18, Institute for Scientific Information, Philadelphia, Pa 3. Weeks, C. E., Hermann, A. L., Nelson,F. R., and Slaga,T.J. (1982)Proc. Natl. Acad. Sci. U. S. A. 79, 6028-6032
4. Takigawa, M., Verma, A. K., Simsiman, R. C., and Boutwell, R.K. (1983) Cancer Res. 43,3723-3738 5. Pegg, A. E. (1988) Cancer Res. 48, 759-774 6. Fothergill, J. C., and Guest, J. R. (1977) J . Gen. Microbiol. 99, 139-155 7. Alhonen-Hongisto, L., Veijalainen,P., Ek-Kommonen, C., and Janne,J. (1982) Biochem. J. 202,267-270 8. Holtta, E., and Pohjanpelto, P. (1983)Biochem. J. 210,945-948 9. Pohjanpelto, P., Holtta, E., Janne, 0.A., Knuutila, S., and Alitalo, K. (1985)J. Siol. Chem. 260,85324537 10. Pegg, A. E., Wechter, R., Pakala, R., and Bergeron, R.J. (1989)J. Biol. Chem. 264, 11744-11749 11. Pegg, A. E., and McGill, S. (1979) Biochim. Biophys. Acta 568,416427 12. Persson, L. (1977)Acta Physiol. Scand. 100,424429 13. Pegg, A. E., Pakala, R., and Bergeron, R. J. (1990) Biochem. J. 267,331-338 14. Porter, C. W., Ganis, B., Libby, P. R., and Bergeron,R.J . (1991)Cancer Res. S1, 3715-3720 15. Wallace, H. M., and Keir, H. M. (1986)FEBS Lett. 194,60-63 16. Taffet, S. M., and Haddox, M.K. (1985) J. Cell. Physiol. 122, 215-220 17. Kabra, P. M., Lee, H. K., Lubich, W. P., and Marton,L. J. (1986)J . Chromatogr 380,19-32 18. Gilbert, R. S., Gonzalez, G. G., Hawel, L., and Byus, C. V. (1991)Anal. Biochem. 199,86-92 19. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 20. Wu, V.S., and Byus, C. V. (1984) Biochim. Biophys. Acta 804,89-99 21. Byus, C. V., and Wu, V. S. (1991)J . Cell. Physiol. 149, 9-17 22. Gonzalez, G. G., and Byus, C. V. (1991) Cancer Res. 51,2932-2939 23. Kashiwagi, K., Miyamoto, S., Suzuki, F., Kobayashi, H., and Igarashi, K. (1992) Proc. Natl. Acad. Sci. U. S. A. 89,45294533 24. Pegg, A. E., Secrist, J.A,, and Madhubala,R. (1988) Cancer Res. 48,2678-2682 25. Bachrach, U.,and Shtorch, A. (1985) Cancer Res. 45,2159-2164 26. Segel, I. H. (1975) Enzyme Kinetics: Behavior and Analysisof Rapid Equilibrium and Steady-state Systems,John Wiley & Sons, New York 27. Alhonen-Hongisto, L., Seppanen, P., Holtta, E., and Janne,J. (1982)Biochem. Biophys. Res. Commun. 106,291-297 28. Albonen-Hongisto, L., Hirvonen, A., Sinenirta, R., and Janne, J. (1987) Biochem. J . 247,651-655 29. Davis, R. H., and Ristow, J . L. (1989)Arch. Biochem. Biophys. 271,315322 30. Davis, R. H., Morris, D. R., and Coffino, P. (1992)Microbiol. Reu. 56,280-290
