Transport of Amino Acids by Confluent and Nonconfluent ST3 and Polyoma Virus-transformed ST3 Cells Growing on Glass Cover Slips’s

SUMMARY Transport of amino acids into 3T3 and Py3T3 (polyoma virus-transformed 3T3 cells) was measured at pre- and post-confluent stages of their growth on glass coverslips. The 3T3 line is from mouse embryo and selected because of its strong “contact inhibition”. These cells, respectively, are and are not sensitive to density-dependent inhibition of growth, a phenomenon characterized by inhibition of cell cells a medium a to either of growth or the between of

and are not sensitive to density-dependent inhibition of growth, a phenomenon characterized by inhibition of cell multiplication soon after cells reach confluency on a solid substratum.
The nonmetabolizable amino acids a-aminoisobutyric acid and cycloleucine (I-aminocyclopentane lcarboxylic acid) were accumulated about 30% less rapidly by confluent than by nonconfluent 3T3 cells. 3T3 cells grown in a medium containing a high concentration of serum failed to exhibit either density-dependent inhibition of growth or reduction in the transport of aminoisobutyric acid upon attainment of confluency.
The rates of accumulation of metabolizable amino acids such as glutamic acid, glutamine, and arginine were not lower in confluent than in nonconfluent 3T3 cells.
No differences between confluent and nonconfluent Py3T3 cells in rates of accumulation of either metabolizable or nonmetabolizable amino acids were observed. Py3T3 cells accumulated a-aminoisobutyric acid, cycloleucine, and glutamine about twice as rapidly as nonconfluent 3T3 cells, but glutamic acid and arginine were accumulated at the same rate by both 3T3 and Py3T3 cells.
The growth regulation apparently exemplified by densitydependent inhibition of growth in vitro seems not to be due to general alterations in membrane permeability and transport.
;Ut.crstions ill the: permeability of the plnsma mcmbrancs of animal cells may, it has been suggested (I, 2) ('rillurcs of cells rilriltipl~ing on :I solid substr:Itum such :I$ gl:~ss or ],l:istic seem to uffctr :L good system for testing this hypothrsis. 111 culturn of moat types of normal diploid cells and of some ot.her ccl1 lilies there is R m:wktttl reduction in the rntct of cell Illult.il)lic;ttiorl soon after the ccllr; brcomr com~Mt~l\-confluent. Growth tends to ?;t 01) wit.11 the formntion of :I confluwt nionol:qcr. .\ deprrssion iii thcl r:ltes of synthesis of l-)X;\, RS.1, :III~ l)rotciti :iccompallies the inhibition of growth @). The phcwomenon has been cnlled "contract inhibition of growth" or "densit?--dcl)rlldcnt illhihition of growth" The report. by Engle, I'iez, and Levy (5) of no regular differences iii lhe :ibilitiw of several lines of cells derived from normal and nlaligxtnt tissues to nccur~~ulate :rmino acids from n culture mcdiuni and the report by IIare (6) of no significant differences in the characterist-its of phenylalaninc transport into ~irns-t.ransforrncd hxmster cells and their precursors do not support the hypothesis.
In neither of these reports, however, is it clear at what stage iu the growth of the cells in culture the transport was rt.udied.
Presumably, the invcst,igators used cells in t.he logarithmic: phase of growth aud were primarily interested in intrinsic differences in amino acid transport by malignant or ~irlls-tranpforrncd cc4ls on t.he o11c hand and "normal" cells on the other.
'Ik~e studies have not excluded the interesting possibi1it.y that. changes in membrane permeability and transport may appear only upon contact or crowding of t,he cells.
3 '1'3 C&S Vol. 244,?co. 10 the transport of amino acids by cells of bot.h the density-inhibited and noninhibited types was mcasurcd at. stages in their growth both prior to and after cell couflrler~cy.  (7), was chosen for our st.udies because of its high sensitivity to density-dependent inhibition of growth. l'y3T3 cells (8), which do not exhibit density-dependent inhibition, were used for lmrposes of comparison with 31'3. Both cell lines were stated to be free of plcuropneumoniu-like organisms when obtained; they were tested by Dr. R. A. Tobey and found negative at the end of the set, of experiments.
Both cell lines were cultured in J>ulhecco's modified Eagle medium (9), but 1vit.h a reduci.ion in the cowentration of NaIlCO~ to 2.5 g per liter so that a pH of 7.4 was obtained when the medium was equilibrated with an at.mos1ihere containing 5c70 CO*. In addition to penicillin and streptomycin at concentrations of 100 units and 100 pg per ml, reqiectively, the medium was suplilemented with 10% uninact.ivated calf serum obtained from Baltimore Biological Laboratory. Chinese hamster ovary cells (10) were obtained from Drs. L). F. Peterson and R. A. Tobey of the Los Alamos Scientific Laboratory.
Chinese hamster ovary cells were grown in Ham's F 10 medium (11) supplemented with penic*illin, streptomycin, and calf serum at the concentrations given above.
In an at.tempt to ~IISIIIT that throughout our studies the characteristics of the different cell lines did not change, cultures of cells were restarted from frozen stock at 4-to g-week intervals.
Unless otherwise specified, measurements of transport were made with cells which were growing on and attached to glass coverslips.
No. 2 glass coverslips, I5 mm in diameter, were cleaned essentially as described by ITnm and Puck (12) and were heat-sterilized.
Twenty-one coverslips were arranged in each of a number of IOO-mm plastic tissue cuhure dishes (Falcoii Plastics).
Cells were seeded as a suspension in growth medium at a density of 1 x IO5 to 5 X 10" cells per dish.
The cells were allowed to settle onto the covcwlips and after about. 15 min they adhered sufficiently so that moving the dish to an incubator did not cause the cells to wash about (which can result in uneven distribution of the cells). The cultures were incubated at 37" and the incubator was gassed with 5y0 (~02 in air. They were fed with fresh medium at 3&y intervals and on the day preceding their use in transport experiments.
Coverslips with attached cells were used 2 to 7 days after seeding.
Thr &ndard deviation of protein per coverslip or cells per coverslip ranged from about 10% of the mean at low densities of cells to less than 5yG of the mean with confluent cultures.
Cell numbers were e.stimated by removing the cells from coverslips with a small volume of trypsin solution (0.057; trypsin in calcium and magnesium-free PIW with 0..5 mbr EIYI'A (1.3)) and counting a portion of the result,ing cell suspension in a hemocytometer.
air-dried. Several cover~li1is from each dish were used for the cell counts and protein d(,terrnillatiolls.
Coversli1is from six to eight dishes seeded on the same day and at the same density of cells were usd for each series of transport espcrinwnta.
Transport was nwasured at several stages of growth prior to conflucncy and also after the cells had reached cwllflrlellcy.
The growth nirdium was removed by aspiration and the covcrsli1is were rinrrd twice by adding PBS (15) lo the dish.
1% with 0.17;. glurose was then added to the dish alid the coverslips with attached cells were incubated at. 37" for 1 to 1.5 hours to allow dq)letinn of intrarellular amino acids before trunqiort rxperimrnts were begun. The incubation medium for transport studies consisted of pII 7.4 PBS (15) with 0.1 yh glucose and a Y-labeled amino acid at a concentral ion generally of 2.0 rnM and spwific: activity 0.1 to 1 .O mCi per mmole.
lncubation was at 37 + 0.5" for periods usually up to 1 hour.
Coverslips were transfcrrcd to and from the incubation medium with forceps.
Following incubntioii in the isotol)e-corltainil~g IIICdium, a coverrlip was rinsed for a total of 10 sec. hy dipping it serially through three containers of PBS at 3i", drained by touching the edge of the coverslip to filter paper, and dropped into a wintillxtinn vial containing 0.5 ml of water and 15 ml of it was found that a satisfactory correction for the small amount of estrarellular isotope not removed by the rinsing of the coverslip c~oultl he made by using a 0.5~min incubation period as a control.
This correction usually amounted to less than 57; of t.he total substrate awumulated in I hour. Isotopically labeled con~pountls \vere obtained from Calbiochcm or Kew England Nuclear and were used without. further purification.
:Zll other compounds wrrc of reagent grade and all solutions and media were prepared with glass-redistilled water.

RlSULTS
Shdnrdization oj Trawport J~eclsurellrentsTho comparison of transport rates by cells at differeiit densit.ics 011 glass coverslips rryuired a basis for qirrssing the measurements in terms related to the amount of ~11s on ewh coverslip. Fig. 1 shows that standardization on the basis of the number of cells is of little value.
XB the density of the cells inereascs, the average size of the cells as measured by the packed volume of 106 ~11~ decreases. Fig. 2 placed in a vial containing 0.5 ml of NCS solubilizer (Nuclear-Chicago).
Ten milliliters of t.oluene scintillation fluid were added to each vial, and 3H and 1Y: radioactivity were measured in a liquid scintillation spectrometer. Intracellular fluid space was calculated as the difference between apparent urea space and apparent mannitol space. The results of these experiments (Fig. 3) show a linear relationship between intracellular fluid space and protein per covemlip. Accordingly, we concluded that protein per coverslip ws a satisfactory basis on which to standardize transport measurements. As a check on these measurements of intracellular fluid space, the total volume of confluent 3T3 cells on a covcrslip was calculated from measurements of the diameters of 250 cells which had been removed from the glass and allowed to assume a spherical Cells were grown to different densities in plastic culture flasks of 75-cm* surface area, detached from the surface with trypsin solution, and suspended in growth medium.
An aliquot of the suspension was centrifuged at @Xl X g for 15 min in a siliconized centrifuge tube with a graduated capillary stem to determine packed cell volume Five cell counts for each suspension were averaged and used to calculate cells per cm2 of flask surface and the number of cells in the packed volume.
Each point represents data obtained with one to three flasks of cells.

Cells per coverslip x 10e4
FIG.
2. Relationship between the protein and the number of cells on 15-mm diamet.er glass coverslips: l , 3T3; 0, Py3T3. Cells were grown to different densities on coverslips arranged in culture dishes.
Protein and cell numbers were determined on separate coverslips KS described in the text.
Each poinl represents the average of several cell counts on two or three coverslips and protein on five coverslips.

forni.2
Intracellular fluid space as measured by the isotope distribution method was about. 75% of the total cell volume calculated from the measurements of cell diameters. Since this figure compares favorably with measurements of the water content of animal cells in general (17), the isotopic method which we used to determine intracellular fluid space seems to be satisfactory.
Transport by Cells in Suspension and by Cells At&&d to Glass--Chinese hamster ovary cells which grow eit.her in suspension or attached to a solid substratum were used to compare transport by suspended cells with respect t.o that of cells attached to glass coverslips.
Transport by cells at.tached to coverslips was measured as described under "Experimental Procedure." The Chinese hamster ovary cells used to study transport by suspended cells were grown in 200 ml of medium in a spinner flask (Bellco Glass, Inc.) rather than on a glass substratum.
The possibility of damaging the cells in removing them from a glass substratum was thus avoided.
Chinese hamster ovary cells in logarithmic growth in the spinner flask were collected from the growth medium by centrifugation for 4 min at 100 x g. They were washed three times with PBS cont,aining 0.1% glucose.
The cells were suspended at a concentration of 4 x lo6 cells per ml in incubation medium identical with that used with cells on coverslips and the suspension was shaken gently during incubation at 37". At intervals 0.5-ml aliquots of the suspension were removed and immediately chilled on ice. The cells were maintained at 2-4" during three washes wit.h PBS to remove extracellular isotope. The washed cell pellet was extracted with 90% ethanol and an aliquot of the extract was assayed for radioactivity. Fig. 4   AIU and cycloleucine are apparently nof. metabolized by animal cells (18,19). To determine whether the results obtained with these amino acids extended to rnctubolizable amino acids, the transport of a number of naturally occurring L-amino acids carrying different charges was studied. Some of the results of these st.udies arc giwn in Figs. 8 and 9. The rates of accumulntion of glutamic acid (Fig. 8, upper curve), argininc (Fig. 8, lower curve), and glutamine ( Fig. 9,  glut.amine was accumulated mow rapidly by l'y3T3 cells than on glass coverslips: 0, 1.1 X 10' cells (nonwnfuent) and 10 rg of by 3T3 cells (Fig. 9), glulamic acid and argininc were accumuprotein per coverslip; A, 3.7 X 10' cells (nonconfuent) and 27 pg latcd at the same rate by both ccl1 lines. The significance of the broken horizontal line is dcsrribcd in the cumulated AIB both prior to and after confluency (Fig. 10). In medium containing 50yG serum the cells grew well beyond confluency and were still multiplying at the time that they were used in the transport experiments. These cells, in contrast to cells grown in 10% serum, did not exhibit a reduction in AIU transport upon attainment of confluency. Reproduc-ibility of Results-For unknown reasons, the rates of amino acid accumulation by both 3T3 and Py31'3 cells varied considerably in different series of experiments. This variation can be seen by comparing in Figs. 5 and 10 the rates and levels of AIB accumulation by nonconfluent 3T3 cells. However, within a series of experiments with 3T3 and Py31'3 cells seeded on the same day and cultured in the same batch of gron-th medium, the results depicted in Figs. 5 to 10 were rrproducible with respect to the presence or absence of differences bet.ween confluent and nonconfluent cells. IXfferenccs between 3T3 and Pg3T3 cells were less reproducible.
For example, in one of four series of expcrirnents, nonconfluent 3T3 cells accumulated .\I13 nenrly as rapidly as Py3T3 cells.
In each of three series of experiments glutamic acid and arginine were accumulated as rapidly by nonconfluent 3T3 cells as by Pg3T3 cells.
The present invest igat.ions were initiated to determine whether alterations in the permeability of the cell plasma mcmhraiie could have a significant role in the rcgulat.ion of cell growth and mult,iplication.
If, as has been widely assumed, density-rlcpendent inhibition of growth in cell cultures is an exluwsion of some basic growth regulatory phenomenon, and if changes iii mcmbrsne permeability and t.ransport. were important in growth regulation, one might expect to find significant differences in membrane permeability between cells which do and thwe which do not eshibit density-dependent inhibition. These differences might be inherent to the different cell lines or they might appear only after cell contact or crowding occurs.
The latter condit.ions are particularly significant because t,hcy most resemble the cells' state in viuo.
The results of our studies seem not to support. the hypothesis that general alterations in membrane permeability and transport are related to gr0wt.h regulation.
With nonconfluent cells we found no consistent differences in t.he abilities of the densityinhibimble 3T3 line and the noninhibitable Py3T3 line to tmnsliort a variety of amino acids.
Accumulation of AIB, cycloleutine, and glutamine was usually about twice as rapid in Py3T3 cells as in nonconfluent 3'1'3 cells, but this difference was not found for glutmic acid and arginine. These results are in agreement with previous studies (5, 6).
Although it. is perhaps significant that only with the two nonmctabolizable amino acids tested was there a reduction in transport when 3T3 cells reached confluency, it is not clear how the presence or absencr of reduced transport activity in confluent 3T3 cells might relate t.o the ability of the ~11~ to metabolize the substrate.
Whether the reductions in transport activity for AIR and cycloleucine upon attainmrnt of confluency with 3T3 cells are of sufficient magnitude to be of significance in growth regulation is unknown.
The presence of such changes suggests, however, the interesting possibility that there may be growth regulatory mech-anisms which depend on specific changes of permeability t.o possible growth regulatory substances. The basis for the rcduct.ions in AIL< and cycloleucine transport upon att.ainment of confluency by 3'1'3 cells is unknown. These reductions seem not to be a consequence of cell crowding per se. Crowding of 3T3 ~11~ upon attainment of conflucncy did not reduce the tranq~ort of glutamine, glutnmic acid, and argininc ( Figs. 8 and 9), nor did it reduce AIU transport by 3T3 cells grown to beyond confluency in medium containing 50~~ swum (Fig. 10). I\Ieasurements of the kinetic constants, K, and I',,,, under conditions approsimnting initial rates of transport of AlJ{ and cycloleucine into the cells did not reveal any major changes in these constants upon attainment of confluency.
HOWever, since our method of measuring these constants was not sufficiently senGtive to wtablish t.he significance of small differcnccs, WC cannot esclude the possibility that, a small change in K,, I',,,,, or both underlies the reduction in AIR or cycloleucine transport which occurs when 3T3 cells reach confluency. The action of higher than normal concentrations of serum in the gro1vt.h medium in causing 3'1'3 cells to grow t.o beyond confluency and thus not to exhibit. dellsitJ--deI)elldent inhibition of gro1vt.h has been interpreted in different ways. Ilolley and Kicrnan (21) csplain the appearance of density-dependent inhibition in cultures of 3T3 cells as an artifact due to exhaustion of esscntial growth factors prrsent in serum.
Eagle and Levine (22) suggest,, however, that supprwsion or reversal of dcnsity-dependent inhibition by high concent.ration of strum may merely indicate a flexibility in the mechanism of the inhibition. In our experiments in which transport of 111) into confluent and non conflurnt 3T3 ~11s grown in medium containing 10yG or 5Oc/o serum was mewurcd (Fig. lo), cells grown at t.he higher concentration of serum not only grew to beyond confluency, but also failed to cshibit a reduction in MB transport when they became confluent.
These cells, in contrast to cells grown to confluency in lOy0 serum, were still mult.iplying rapidly at t.hc time that they were taken for USC in the transport experiments. It seems thcreforc that the reduction of AUS transport in densit.yinhibited cultures of 31'3 cells occurs nhen growth has stopped. If other changes in memhmne lwmoability and transport arc found to accompany density-dependent inhibition, it will be important.
to establish whether such changes are the cause or the effect of inhibition of growth. It should be noted that. transport was not measured in the prcwnce of serum because competit.ion between serum component.s and the substrate might complicate the results. iV:hat effects noultl rwult from internct~ions between serum and mcmbrane was not. determined.
Cells attached t.o the surface of a solid substratum presumuhl~ present a considerably larger area of plasma membrane for solute transport than do cells which have assumed a spherical shape because they have been grown in suspension or removed from a surface.
This difference may account for the rate of accumulation of Ail< by Chinese hamster ovary cells on glass coverslips being about 2.5 times greater t.htm the rate of AIB accumulation by suspensions of thtse cells (Fig. 4).
The use of roverslips with attached cells as a unit of material for studying tranqort or metabolism (23, 24) by cultured cells has several advantages over techniques which use either cells in suspension or cells attached to culture containers. In most cases in which a reaction such as uphill t.ransport is being followed by guest on March 23, 2020 http://www.jbc.org/ Downloaded from Iesuc of May 26, lS(iY nith isotopically labeled solute?;, as fC!v as IO4 cells on a roverslip provide sufficient material. This is of practical siguificance for it permits a considerable reduction in the time und espe~~se devoted to growing suflicient. amounts of cells for studies in ivhich numerous samples arc required. The coverslip technique also makes the separation of t.he cells from the incubation medium and their bvashing easier and more rapid than is the cast lvith other techniques. Furthermore, lvhen radioisotopically labeled substrates are used to follo\v transport or metabolism, :I cover-sliJ) \vith attached celLs can be placed directly into a vial for measurement of radioactivity by liquid scintillation counting. The small amount of cell material added to the vial produces no detectable reduction in t,he efficiency of counting of 14C or tritium. The coverslip technique therefore eliminates the extraction procedures usually necessary lvhen relat.ively large amounts of cells are used in measuring trnnq)ort or metabolism. It also mininlizes damage to the cells during manipulations.
The coverslip technique should find lvide utility in metabolic stud& nith cultured cells.
,IcknowZedgment-We thank Miss Ethcria Robinson for skillful and conscientious technical assistance.