a time study of the incorporation of radiophosphorus...

A Time Study of the Incorporation of Radiophosphorus intothe Nucleic Acids and Other Compounds of a Trans

planted Mouse Mammary Carcinoma*

CYRUSP. BARNUM,ROBERTA. HUSEBY,!ANDHALVORVERMUND

(Departments of Physiological Chemistry, Physiology, and Radiology of the Medical School,Unirersity of Minnesota, Minneapolis, Minn.)

A previous study (2) with P32and with normalmouse liver tissue indicated the marked metabolicheterogeneity of the pentosenucleic acid (PNA)isolated from various cellular fractions. It washoped that by extending such a study to tumor tissue we could derive further information concerning possible metabolic interrelationships betweenthe PNA of various fractions and also betweenPNA and desoxypentosenucleic acid (DNA). Suchinformation should help us to better understandthe growth process.

MATERIALS AND METHODSSupply of tissue.â€”Sucha study required the periodic supply

of a uniform tumor tissue over a period of about 1 year. Forthis purpose a spontaneous mammary carcinoma that arosein an AZF! mouse was carried by successive transplantation inAZFi or AxZbFi mice. The transplantation was done by subcutaneous inoculation of a tumor cell suspension to the backsof the recipient animals. After several transplant generationsthe tumor appeared to have attained a uniform rate of growthwhich it maintained for 32 generations. The rate of growthwas such that tumors of 1-2 gm. could be removed 18-16 daysafter inoculation.

The radioactive tracer.â€”At this period, 2 weeks afterinoculation, groups of mice1 containing three or four animalsper group were given injections intraperitoneally of a solutioncontaining Pn as orthophosphate.1 The amount of Pa usedranged from 0.5 to 2 /iC/gm body weight.

Fractionation of tissue.â€”Attime intervals ranging from 15minutes to 16 hours after injection of P32the mice were sacrificed, and the tumors were homogenized and fractionated aspreviously described (1, 8, 14). The samples obtained were thenuclear fraction (V), the large granule or mitochondria!fraction (L),' the microsomes (M), the ultrasedimentable

* Aided by grants from the American Cancer Society andthe National Institutes of Health of the Public Health Service.

t Present address: Dept. of Surgery, University of ColoradoMedical School, Denver.

1The animals used in these studies were kindly supplied byDr. J. J. Bittner.

*PÂ° was supplied as HiPOÂ«on allocation from the OakRidge National Laboratory.

*It was recognized that saline is not ideal for isolation ofthe mitochondria! fraction, and in most experiments thisfraction was discarded. It was retained in a few experimentsfor comparative purposes only.

Received for publication July 13, 1953.

particles (I'), and the nonsedimentable supernatant fluid (S)

From each of these fractions the nucleic acids, phospholipids,and "phosphoprotems" were extracted and purified (2, 3, 8,

14). The purified organic compounds and the inorganic phosphorus from an aliquot of the original homogenate (11) werewet ashed and analyzed for P32and total P (2). The nuclearPNA obtained was that previously described (14) as fractionu, which accounts for about 80 per cent of the PNA of mousemammary tumor nuclei isolated by this citric acid procedure.The preparation of PNA from S has proved more difficultthan from the particulates, owing to the larger amounts ofprotein that fail to coagulate on the first extraction with hot10 per cent NaCl. Consequently, the alcohol precipitate fromthis first extract has been re-extracted with hot 10 per centNaCl and Â«precipitated with alcohol. At this point if the PNAis dissolved in water, much of it fails to precipitate withtrichloroacetic acid (TCA). Therefore, in about half of theexperiments, the second alcohol precipitate was washed withTCA directly and then hydrolyzed with TCA.

The DNA samples have shown an average N:P ratio of1.68 with a range of 1.60-1.67. The average N:P ratios forthe PNA from the cytoplasmic particulates have rangedfrom 1.75 to 1.79, that for the supernatant PNA 2.35, andthat for the nuclear PNA 4-5. Despite the high N:P ratios insome of these fractions, the pentose content based on theorcinol-HCl reaction (measuring purine-bound pentose only),expressed as the percentage of that calculated from the P content, has averaged from 49 to 51.5 per cent for the PNA fromthe cytoplasmic particulates, 52.5 per cent for the supernatant PNA, and 49.5 per cent for the nuclear PNA. It hasbeen concluded, therefore, that the bulk, if not all, of the Pin the PNA samples is nucleic acid P.

RESULTSIn Table 1 are presented the results of the chem

ical analyses on the whole homogenate of thistumor line and on the separated cellular fractions.It may be seen that the total DNA and the totalPNA are very similar in amount and that the bulkof the PNA is found associated with the particu-late fractions (primarily M and U ) derived fromthe cytoplasm. The value for S-PNA is probablysomewhat low because of the difficulties encountered in quantitatively precipitating this fractionin several experiments. Consequently, two experiments were carried out in which the cytoplasmicextract (after centrifuging the homogenate to remove nuclei) was spun directly at 100,000X gfor 1 hour, and all the cytoplasmic particulates

880

on June 20, 2018. © 1953 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

BARNUMet al.â€”Incorporation of P32 into Nucleic Acids 881

were sedimented together. The sediment and su-pernate were treated with cold 5 per cent TCAand then with hot TCA to remove the PNA. Onthe basis of orcinol-HCl tests the PNA P of thecombined participates averaged 493 Â¿ig/gmtissue,

Table 2 are recorded the specific activities of theinorganic phosphate and of the various nucleicacid fractions. In seven experiments, performed atvarious time intervals, a large granule or "mito-chondrial" fraction was also isolated. In six of

TABLE1COMPOSITIONOFFRACTIONSDERIVEDFROMATRANSPLANTEDMOUSEMAMMARYCARCINOMA

FRACTION*

X (isolated)

N(correctedfor yield) ÃŽ

L

M

ACID-SOLUBLE615

(8)t[570-650]:581

(3)[560-620]INORGANIC326

(3)[310-340]301

(3)[290-310]DNA

PNA/igP/gm wet weight of tissue

834 (28) 783(28)[720-1,130]637

(33)[430-760]834[550-910]59(35)[33-90]7549(6)[35-75]253

(33)[170-385]193

(33)[130-265]91

(21)[54-120]661LlFID301

(2)[284-318]27

(29)[11-47]3597(7)[59-147]89

(33)[55-123]11

(33)[5-24]11

(32)[4-24]243PROTEINRESIDUE107

(3)[101-118]6(33)[3-14]84(7)6

(33)[2-14]3(33)[2-5]9(32)[5-19]30

PROTEINRESIDUE

nig N/gm13.9(5)

[12.5-15.5]3.1(5)

[2.7-3.3]4.7

1.3(5)[0.8-1.9]0.8(5)[0.5-1.0]0.75(5)[0.7-0.8]5.0(5)[4.0-5.9]12.6

S

2N.L.M,u, s* H = homogenate; N â€¢

supernatant fluid.t Values are means of the number of experiments shown in parentheses. Ranges are given in brackets below each mean.t Some nuclei are lost during purification. A correction factor is obtained from the ratio of DNA in the homogenate to that in the isolated

nuclei.

TABLE 2

â€¢nuclei; L â€”large granules or "mitochondria;" M = microsomes; U = ultrasedimentable particulates; S â€”

TIME AFTERP" INJEC

TION(HOURS)0.25

0.5

SPECIFICACTIVITY(BCC)* OFTHEPHOSPHORUSOFVARIOUSNUCLEICACIDFRACTIONS

INORGANICit7,

470 (5)Ã•[4,900-9,300]10,400(4)[8,600-11,600]9,640(4)[8,000-11,700]6,940(4)[6,100-8,000]4,920(4)[3,300-6,400]4,130(4)[3,200-4,700]2,670(4)[2,500-3,000]1,610(3)[1,530-1,700]DNAd5.5(5)[2.1-10.1]27.4(4)[22-36]84.5(4)[64-110]220

(4)[200-253]365

(4)[260-414]530

(4)[512-575]901

(4)[778-1,040]1,160(3)[1,100-1,230]N-PNAa131

(5)[65-216]633

(4)[540-760]1

,680(4)[1,470-2,170]2,690

(4)[2,370-3,370]2,840(4)[4,100-3,700]3,000(4)[2,480-3,380]2,360(4)[2,220-2,650]1

,640(S)[1,630-1,650]M-PNA3.3(5)[1.1-6.0]14.1(4)[11-18]69.6(3)[48-104]260

(4)[231-290]461

(4)[409-519]649

(4)[563-685]1,110(4)[1,020-1,210]1,520(3)[1,470-1,590]U-PNAa3.9(5)[1.0-8.8]15.2(4)[11-22]80.8(4)[49-117]242(4)[212-265]450

(4)[345-522]582

(4)[566-594]1,010(4)[853-1,110]1,360(3)[1,350-1,370]S-PNAt24.6(3)[16-38]62(3)[50-84]179

(4)[130-220]344

(4)[284-415]595

(4)[354-767]891

(4)[770-1,000]1,150(4)[974-1,440]1,210(3)[1,180-1,260]

1C

* See footnote 4.

t i, dt a, and 6 are symbols subsequently used for inorganic P, DNA P, nuclear PNA P, and supernatant PNA P, respectively, c is the symbol used forthe particulate PNA P, i.e., the P of M-PNA plus U-PNA.

ÃŽValues are means of the number of experiments shown in parentheses. Ranges are given in brackets below each mean.

while that of the supernate averaged 104. Thisgives a ratio of particulate to supernatant PNA inthe cytoplasm of 4.75, a figure that will be madeuse of later.

Specific activity* of nucleic acid fractions.â€”In4Throughout this paper specific activity is expressed as

the Biological Concentration Coefficient (BCC) as suggestedby Schulman and Falkenheim (12). It is defined as follows:

(counts/min/mM P) X 100BCC =counts/min administered/gm body weight

The relative specific activities reported in a previous paper(2) may be converted to BCC by multiplying them by 101.

these experiments the PNA isolated showed aspecific activity that was within the range observed for the PNA from the other particulatefractions of the cytoplasm. It will be noted fromTable 2 that the differences in the specific activities of the PNA isolated from the M and U particulate fractions of the cytoplasm are not such asto indicate that these nucleic acids are metaboli-cally distinct. This is particularly evident in thevalues recorded at the 15- and 30-minute time intervals when the differences might be expected tobe most marked. In this tumor tissue, therefore,



882 Cancer Research

it appears reasonable to visualize all the PNA ofthe cytoplasmic participates (P-PNA) as beingmetabolically homogeneous. In sharp contrast tosuch homogeneity in tumor are some recent unpublished results obtained with normal liver tissuein which the PNA isolated from the U fraction 15minutes after administration of P32had 4-5 timesthe specific activity of the M-PNA.

TABLE3SPECIFICACTIVITY(BCC)*OFTHEPHOSPHORUSOF

LIPIDANDPROTEINRESIDUEFRACTIONSTIKE

AFTIBP"INJECTION(HOUBS)

0.25

0.5

LIPID

16

Hand U17 (9) t

[10-26]54(6)[42-36]150 (8)

[107-186]377 (8)[313-433]607 (8)[485-673]864 (7)[760-910]1,220(8)

[1,030-1,430]1,470(6)

[1,400-1,530]

99(3)[41-118]187 (3)

[83-302]284(3)

[190-430]486 (4)[458-525]771 (2)

[557-985]995 (4)[840-1,080]1,850(4)

[1,130-1,600]1,520(3)

[1,410-1,700]

PROTEINRESIDUE

M, U, and S

830 (11)[270-1,570]1,480(11)

[930-2,110]2,380(11)

[1,280-3,040]2,480(12)[1,690-3,360]2,400 (11)

[1,450-3,630]2,380(12)[1,140-2,970]1,990(11)

[1,400-2,670]1,700(8)

[1,510-1,950]* See footnote 4.t Values are means of number of observations shown in parentheses.

Ranges are shown in brackets below each mean.

It will be observed that the nuclear PNA showsby far the most rapid increase in specific activityof all the nucleic acid fractions isolated. This issimilar to previous observations for normal livertissue (2). Initially, the S-PNA has a specific activity much greater than that of the particulatePNA, but the latter rapidly approaches it.

Specific activity of phospholipid and "protein residue" fractions.â€”Table 3 shows the specific activities of the phospholipid P and the "proteinresidue" P. Since the phospholipid P of the M

and U fractions showed no significant differences,these are reported together. Similarly, the phospholipid P of the "mitochondrial" fraction ob

tained in seven experiments showed no differencesfrom that of the M and U fractions. On the otherhand, the S-phospholipid P shows a significantlyhigher specific activity at early time intervals, butthese values are approached by the particulatephospholipid values within 16 hours. The nuclearphospholipid has shown a specific activity of approximately 85 per cent of that of the cytoplasmicparticulates.

The "protein residue" P shows an extremely

rapid rise in specific activity, and no significantdifferences have been observed between any of the

cytoplasmic fractions studied. Again, however,the nuclear "protein residue" P has shown a spe

cific activity of approximately 85 per cent of thatof the other fractions.

Chart 1 illustrates the time course of the specificactivities of several of these fractions and demonstrates that by 16 hours they have very nearlyreached equilibrium.

DISCUSSIONThe specific activity of the "protein residue" P

shows an extremely rapid increase, comparable tothe few observations previously made on this fraction from liver (2). An inspection of Chart 1 indicates that the curve, though rising more rapidlythan that for the N-PNA, goes through its maximum at a value well below that for N-PNA. Thisseems to imply a heterogeneity of the P in thisfraction (10)6 with one part incorporating P32veryrapidly and a second part more slowly. Even theslower fraction, however, has come to equilibriumby 16 hours.

6 8 10Time in hours

14 16

CHART1.â€”Specificactivity-time curves for inorganic P,nuclear PNA (N-PNA), "protein residue," particulatephospholipid (P-PL), and particulate PNA (P-PNA).

BCC refers to the specific activity. See footnote 4.

The particulate phospholipid incorporates P32much less rapidly than the same fraction in liver(2). However, in tumor tissue the supernatantphospholipid shows a higher specific activity thanthe particulate phospholipid, whereas in liver nodifferences were observed.

Analysis of the specific activity-time curves fornucleic acid fractions.â€”Our particular interest hasbeen focused on the incorporation of P32into the

6Kennedy (10), in an oral report, presented evidence for theisolation, from a protein residue fraction, of serine and threo-nine phosphates with specific activities several times that forthe total P of this fraction.




various nucleic acids of the cells. One of our objectives was to find a tissue in which growth wassufficiently reproducible and constant over the experimental period that we could derive some usefulinformation by following the time-course of thespecific activity of the DNA and of the variousPNA fractions.

One important simplifying assumption which isbasic to all the calculations that follow was thateven this rapidly proliferating tumor tissue could,as a first approximation, be visualized as a tissuein a steady state. This assumption was made sinceit seemed probable, on preliminary inspection ofthe curves, that the rates at which most of the constituents were equilibrating with their precursorsgreatly overshadowed the rate of growth. In viewof the results reported in this paper the assumption of a steady state appears to have been a validapproximation.

Interrelationships between PNA fractions. â€”Onthe basis of previous observations (2) on mouseliver tissue a tentative hypothesis was suggestedwhich proposed that some unknown intermediate,x, which equilibrates very rapidly with inorganicP, i, is the immediate precursor of the phosphorusof nuclear PNA, a, and also of the cytoplasmicsupernatant PNA, 6, which in turn is the precursorof the particulate PNA, c, of the cytoplasm. Sucha scheme can be pictured as follows:

Â»Inorg.11

or I Â«11Â»,1. (1)b~Â±c

6

The justification for considering the particulatePNA as the end of the reaction chain â€”thus permitting the formulation of equation (4) below â€”isthat at early time intervals no P fraction that wehave isolated from this tissue has shown a lowerspecific activity than the P-PNA. Since the majorP fractions have been analyzed, it seems improbable that P-PNA could be the immediate precursor of any other appreciable P pool, which bya reversible reaction might affect its specific activity-time curve.

Since the steady state is assumed in scheme 1,then Â»i(the rate at which P in compound x isincorporated into compound a) is equal to t>2(therate at which P returns from a to a;). Similarly,Â»3= Â«4and Â»5= Â»6.As previously described (2),one can define the time changes in the specific activities of compounds a, b, and c by means of thefollowing equations:

(2)

(3)

(4)

In these equations or, b, and C refer to the figPNA P/gm wet tissue to be found in nuclei, cytoplasmic supernate, and cytoplasmic particulates,respectively, while X, A, B, and C refer to thespecific activities of the compounds x, a, b, andc, respectively. The constants k, I, m, and n define that fraction of a given compound replacedeach hour. Thus, k is that fraction of the P presentin a which traverses the path from z to a eachhour.

Metabolic reactions involving DNA.â€”In attempting to analyze the data with respect to DNAwe have made the assumption that the incorporation of P32into DNA occurs only in cells that arepreparing to undergo division, i.e., that in most ofthe cells at any given time the DNA is metaboli-cally inert as far as its P is concerned. This assumption rests on the observation of Osgood et al.(11) that in three nondividing tissues (skeletalmuscle, cartilage, and brain cortex) of an adultpatient treated with P32 for 1 year the DNAshowed no incorporation of P32: on our observation (2) that the DNA of adult mouse liver cellsincorporates extremely little P32; and on the observation of Brues et al. (5) that when P32is incorporated into liver during regeneration it remains in the DNA for a long time, whereas itgradually leaves the other P compounds of theliver.

In a cell that is preparing to undergo divisionand is synthesizing an equivalent amount of newDNA, one could visualize several possibilities:(a) the pre-existing DNA of the mother cell mightremain metabolically inert, and therefore only halfof the DNA of the daughter cells would becomelabeled; (b) the pre-existing DNA might go completely into the pool of its precursors from which adouble amount of DNA, all labeled, wouldemerge; or (c) the pre-existing DNA might enterinto reversible reactions with its labeled environment at, or at about, the same time that an equivalent amount of new DNA was being synthesizedfrom the precursor pool. Regardless of which ofthese three possibilities is correct, the rate of incorporation of P32into DNA will be a function ofthe rate of synthesis of DNA. Furthermore, sincewe have found no significant differences in thecontent of DNA/gm of tissue between small andlarge tumors, one can also conclude that the incorporation of P32is a function of tumor growth.



884 Cancer Research

We may visualize a general reaction scheme asfollows :-Â»Inorg.P^z^DNA or -+i^Â±xlÂ±d (5)

I i

where Â»7is the rate at which P goes from a precursor, x, to DNA, d, and t>8is the rate of a possiblereverse reaction. This system is not in a steadystate, and we know that Â»7> t>8-If possibility (a)above is correct, then Â»8= 0; if possibility (6) iscorrect, then Â»8= c7/2; and if possibility (c) is correct, then 0 < va < t>7/2.

The quantity (Â»7â€”Â»8)/drepresents the fractional increase of DNA per unit time and will be referred to as the growth constant, g. If one assumesthat the number of cells in which DNA is beingsynthesized is proportional to the number of cellspresent (or to the total DNA present) at any giventime, and that this proportionality remains constant over the experimental period of 16 hours,6then one can write that d = dQe<"where d0 is theamount of DNA at time zero, d the amount ofDNA at any specified future time, g the growthconstant, and t the time in hours. Using this assumption of the constant percentage increase inDNA over the experimental period, one can derivethe following equation defining the time changein the specific activity of DNA :

(6)

and

Inor.

dt d

where

f-l-J~ d'

It is surprising at first to find the form of theequation identical to that for reversible, steady-state systems, but this is a direct result of the assumption that the rate at which DNA is formedin the tissue is proportional to the amount there.

Proposed interrelationships between DNA andthe various PNA fractions.â€”With equation (6) onecan readily demonstrate that neither inorganic Pnor any of the fractions of PNA P isolated can fulfill the role of x. Thus, as an initial assumption tobe tested, we have visualized the following:

*The validity of the assumption of a constant mitotic ratethroughout the 16-hour experimental period is strengthenedby two sets of observations. Several observers (4, 6) have notedthat for malignant tumors (epidermoid carcinoma and carcinoma of large intestine have been investigated) the mitoticrate remains constant and does not show the diurnal variationseen in normal tissues and in regenerating tissues. Unpublished observations in this laboratory indicate the lack of anysignificant difference in the rate of incorporation of P32 intothe DNA of a mouse mammary carcinoma when tested between7 and 9 A.M.as against 9-11 P.M.

DNAU

S-PNA?Â±P-PNA

or

(7)

In words, this assumption says that the phosphorus of the DNA, d, of the nuclear PNA, a, and ofthe cytoplasmic supernatant PNA, b, is all derivedfrom precursors that share a common metabolicbox. It is possible that the phosphorus precursorsof each of these nucleic acids are identical, thoughthis is not necessary and does not seem very probable. If the precursors of a, b, and d are so rapidlyequilibrated among themselves that their specificactivities are indistinguishable, then for our subsequent considerations x may be considered asingle substance.

Evaluation of proposed mechanism.â€”The firststep in testing the above hypothesis involvesanalysis of the postulated chain of reactions fromx to b to c as shown in equation (7). Calculationson the cytoplasmic constituents have been carriedthrough the 8-hour interval. From Chart 2 it maybe seen that it was possible to draw a smoothcurve through, or very close to, all the experimental points for C, i.e., the points for the specificactivity of the particulate PNA of the cytoplasm.In the case of the supernatant PNA, the experimental points for specific activity, B, were lessconsistent, and a smooth curve, B\, was drawnamong the points as a first approximation. At thisstage the areas under curves BI and C (i.e., ft and7, respectively) were evaluated and introducedinto the integrated form of equation (4) :

C=Â»(0i-7). (8)

By plotting C against (ft â€”7), it was possible todraw the best straight line through the points.The slope of this line would represent the mostprobable value of TO,which was evaluated as 1.18.Using this value for m, along with the fixed valuesfor C and 7, equation (8) can be used to re-evaluatej3, this time calling it /3j.Finally, the values for ÃŸtat the various time intervals may be numericallydifferentiated, and a smooth curve drawn throughthese differentials to give B$ as shown in Chart 2.It may be seen, then, that a curve Bz, varyingonly a little from the curve B\ first drawn amongthe experimental points, will fit exactly into therole of precursor for the particulate PNA.



BARNUM et al.â€”Incorporation of JP32into Nucleic Acids 885

The next step involves writing equations (6), If we now plot B/A against (o â€”ÃŸ)/Aand A/D(2), and (3) in their integrated forms.

A = *(Ã-a)

B = l(i-ÃŸ) -n(ÃŸ-y).

(9)

(10)

(U)

In these equations a, ÃŸ,S, and Â£represent the integrals of A, B, D, and X, i.e., they represent theareas under the respective specific activity-timecurves up to any given point. Since (ÃŸâ€”7) =C/m, as seen from equation (8), and since n/m =

1200

1000

800

600JaD400

200

0 345

Time in hours

CHART 2.â€”Specific activity-time curves for the super

natant PNA (B) and the participate PNA (C) of the cytoplasm.

BI was first smoothed curve drawn among the B pointswhile Bj defines precisely the time course of a fraction thatcould be the immediate precursor of the particulate PNA.

BCC refers to the specific activity. See footnote 4.

c/b = 4.757 as seen from equations (3) and (4),equation (11) may be written as

B=l(e-ÃŸ) -4.75C. (Ila)

Finally, equation Ila may be rearranged to read5 + 4.75C=A = /({-0). (116)

The hypothesis to be tested assumes that Â£isidentical in equations (9), (10), and (lib). Thus,from equations (10) and (116) we have

S = a + k-1A=ÃŸ + l-1ÃŸ. (12)

Dividing through by A and rearranging we get

!Â±=^â€”/-i =-*-'. (12a)

Similarly, from equations (9) and (10) we get

D D(13)

* 4.75 is the ratio of the particulate PNA in the cytoplasmto the supernatant PNA. See first paragraph under "Results."

against (a â€”S)/D, as shown in Charts 3 and 4, wecan evaluate l~l, k~l and/"1 from the proper slopes

2.4

2.0

B.A

1.6

1.2

0.8

0.4

a-ÃŸ \ B.=__LA JL ' A A

JL- 0.38

A- 1.5

VCHART8.â€”Graphicalsolution of equation (12a).Errors of B//l_calculated as V(Z(x â€”x)*]/(N - 1) where

i is the ratio of B/A in a given experiment, z the mean B/Aratio at a given time point, and N the number of observationsat a given time point (N = S or 4).

Errors of (o â€”ÃŸ)/Aare crudely estimated on the basisof the assumption that a value of A higher than the valueat that point on the smoothed curve is associated with thevalue of a proportionately higher than the o derived fromthe smoothed curve. Similar reasoning applies to B and ÃŸ.This allows the calculation of o and ÃŸfor each observed valueof A and B, and therefore gives three or four "observations"of the ratio (a â€”ÃŸ)/Afor each time point and allows theestimation of a standard deviation as above.

20

16

10

0 4 8 12 A 16 20 24 28

CHART 4.â€”Graphical solution of equation (13). SeeChart 3 for method of estimating errors.

and intercepts, provided the points can reasonablybe considered to fall on straight lines. Since Ar1must be the same for both plots, then the selectionof best values for all three constants is influenced



886 Cancer Research

by all the experimental data, thus minimizingerrors inherent in the data for any single timecurve. The values for I, k, and /, as evaluated bythis graphical method, are 0.38, 1.5, and 0.05,respectively. Using these constants we can calculate Â£Dfrom equation (9), Â£Afrom equation (10),and Â£â€žfrom equation (Ila) and compare them asis done in Table 4. In this calculation the actual

TABLE 4CALCULATEDAREASUNDERSPECIFICACTIVITY-TIME

CURVESOFPOSTULATEDNUCLEICACIDPRECURSORS

TIMEAfTER P"

INJECTION(BOOHS)0.250.512:i4816I.*1105511,7204,5707,77011,50021,90035,400&1025191,8004,7707,76010,90021,20036,400f.1123641,5104,4308,17011,70021,800XÃŒ7201,8502,7003,2203,4403,2502,0901,640

*Â£Â»,Ã•A,and {â€¢calculated from equations 9, 10, and 11Â«respectively.Values of the constants, obtained from the graphical solution, were 1.5, 0.05,and O.S8 for i,/, and I, respectively. Values of A, B, C, and D used were thoseexperimentally observed, but values of the areas o, f>,and S were obtainedfrom the smoothed time curves of A. B, and D.

t X is derived by numerical differentiation of the time curve of the average Â£values.

experimental values for A, B, C, and D at thevarious time points were used, but it was necessary to use the areas o, ÃŸ,and 5 derived from thesmoothed curves. Inspection of Table 4 shows thatthe Â£values calculated from the three reactionpathways arc in good agreement, indicating theessentially identical course of the specific activity-time curves for the precursors of DNA, nuclearPNA, and cytoplasmic supernatant PNA. Also,shown in Table 4, are values of X, the specific activity of these precursors at the various time intervals. These values of X are obtained by graphical differentiation of a smooth curve drawnthrough the average Â£values. The relationship ofX to several other fractions is shown graphicallyin Chart 5.

While these calculations indicate that the experimental data are compatible with the reactionscheme outlined in equation (7), they do not ruleout the possibility that the data might be equallycompatible with some other scheme. A currentlywidely accepted alternative scheme has been presented by Jeener and Szafarz (9), and they havemarshalled the arguments supporting it. In thisscheme it is proposed that nuclear PNA is a precursor of the nonsedimentable PNA of the cytoplasm. If the values of B + 4.75(7 are plottedagainst (Â£â€”ÃŸ),it may be seen from equation

(Ilo) that a straight line of slope I should result.

Such a plot, using the average Â£values derivedabove, gives an excellent straight line passingthrough the origin. Regardless of what the precursor of this cytoplasmic PNA (B) might be, ifthe area under its curve were substituted for Â£inequation (lie) and a similar plot made, a straightline through the origin should result. However,when the values for a, the area under the nuclearPNA curve, were substituted for Â£,the resultingplot described a curve for which the points werevery far from falling on a straight line. Thus, atleast in this tissue, the nuclear PNA isolated cannot be considered the precursor of the supernatantPNA of the cytoplasm.

Comparison of f with the growth rate of the tumor.â€”Of all the constants that emerge from these

calculations, / has special interest because it isrelated to the growth constant, g, which can beapproximated by independent means. Unfortunately, the tumor transplant line used in thesestudies was accidentally lost before several contemplated experiments could be undertaken. However, three independent evaluations of increase intumor volume by caliper measurements had previously been done on this tumor for other experi-

6 8 10Time in hours

12 14 16

CHABT 5.â€”Specific activity-time curves for inorganicP (7), nuclear PNA (A), DNA (D), and a postulated precursor(X).

BCC refers to specific activity. See footnote 4.

ments. Two of these evaluations involved twentyand thirteen tumors, respectively, that had beentransplanted to the thigh regions (14), while thethird one involved 24 tumors transplanted to theback. In these evaluations the tumor volume wascrudely estimated as the volume of a sphere whosediameter was the average of the two external dimensions of the tumor minus double the thicknessof the skin. In the first experiment, the crudegrowth rate per hour was 2.0 per cent between the




12th and 17th days, in the second experiment 2.0per cent between the 13th and 18th days, and inthe last experiment 2.6 per cent between the 9thand 12th days.

A growth rate of 5 per cent per hour, which wasthe calculated value of /, implies that the tumorshould triple its size each 24 hours, and this tumorwas clearly not growing that fast. From the crudeestimation of its actual growth rate, g, it would appear that this might be one-half of /. Referringback to equation (6) it would appear clear thenthat Â»g/dor h does have a finite value and that thebest approximation currently possible is that it isequal to g. Thus, it is tentatively concluded that,in a cell preparing to synthesize new DNA, thepre-existing DNA does enter into metabolic reactions with its environment and that this preexisting DNA goes completely (so far as its P isconcerned) into the metabolic pool of its precursor.

This same conclusion has been recently reachedby Stevens, Daoust, and Leblond (13) fromstudies on rat liver and intestine in which theycompared the "rate" of incorporation of P32into

DNA to evaluations of the rate of new cell formation in these tissues. They calculate the rate ofincorporation of P32 on the assumption that theprecursor of DNA has a specific activity-timecourse which can be represented by that for theacid-soluble P of their tissues. Their values onthe rate of incorporation are expressed as DNAspecific activity at the time of sacrifice as a percentage of the "average specific activity" of the

acid-soluble P from the time of injection to thatof sacrifice. They do not explain how they calculate the average specific activity of the acid-soluble P, but from their data one can calculate8 thatthe value they use for the average activity from0 to 3 hours is the observed value at 3 hours andthat the value they use for the average from 0 to 6hours is the mean of the observed values at 3 and6 hours. From our previous study (2) and fromunpublished data pertaining to acid-soluble P obtained in the same study, we can calculate thatfor mice injected intraperitoneally with P32 theobserved specific activity of liver acid-soluble P at3 hours is 73 per cent of the average activity overthe 3 hours. We assume that a somewhat comparable relation exists for the liver acid-soluble Pin their rats which were injected subcutaneously.Furthermore, their calculation of the 6-hour average tacitly assumes that the average between 3and 6 hours is equal to the value observed at 6

*A personal communication from Dr. Leblond confirmsthat the calculation described here is equivalent to the oneused by them.

hours, rather than to some value intermediate between the 3- and 6-hour values as would be necessary on this descending portion of their curve. Wehave recalculated their data, assuming that the3-hour value for acid-soluble P should be dividedby 0.73 to give the average over the first 3 hours,and using, as a first approximation, the mean ofany two adjacent values as the average activitythroughout that time interval. On this basis wefind that their liver DNA at 24 hours has a specificactivity that is 1.11 per cent that of the averagefor acid-soluble P instead of 1.36 per cent as theyreport. This makes what they call the "rate of formation of DNA" considerably less than twice the

rate of new cell formation which they give as0.71 per cent per day. On the other hand, the observations reported in this paper indicate that thepostulated precursor of DNA in mammary tumortissue has a specific activity-time course well below that for inorganic P (and also below that fortotal acid-soluble P) throughout the first 8 or 12hours. This implies that if Stevens et al. had beenable to calculate the average specific activity ofthe actual precursor they would have obtained avalue for DNA formation greater than 1.11 percent per day and probably in the range of doublethe rate they report for new cell formationâ€”asthey claim their study demonstrates. It would appear that they reached their conclusion because ofa canceling of errors.

A similar recalculation of their data on intestinal DNA leads to a value for DNA at 3 hours of10.75 per cent that of the recalculated average activity of acid-soluble P. This may be expressedas 86 per cent per day, whereas they report 117.6per cent. Again, one is justified in assuming thatthe actual immediate precursor of the P in intestinal DNA has an average specific activity overthe first 3 hours that is substantially below theaverage for acid-soluble P and would thus lead toa calculation of DNA formed per day substantiallygreater than 86 per cent. In mouse mammary tumor the average specific activity of the postulatedDNA precursor during the first 3 hours is only 50per cent that of the acid-soluble P during the sametime. If this same correction factor should applyto intestine, it would imply a daily rate of DNAformation of 170 per cent, thus making it muchmore than twice their calculated rate of new cellsformed per day of 52.7 per cent. Whether this discrepancyâ€”(a) results from the possibility that thecorrection factor to be applied to intestine shouldbe much less than would apply in mammary carcinoma, (6) indicates need for re-evaluation of the52.7 per cent per day of cell increase which wasbased on the piling up of cells in mitosis over a



888 Cancer Research

4-hour period after the administration of colchi-cine, or (c) implies that the thesis advanced by usas well as by Stevens, Daoust, and Leblond is invalidâ€”must await the accumulation of additionalexperimental evidence.

This conclusion, that the pre-existing DNA of acell that is preparing to divide does incorporateP32,is also consistent with the suggestion made byv. Euler and Hevesy (7) and with the data presented in a previous paper by the present authors(14). In our studies on mouse mammary carcinoma it was observed that a dose of x-radiation, sufficient to cause complete cessation of tumor growthand complete inhibition of new cells enteringmitosis, inhibited the incorporation of P32into the

TABLE 5

CONSTANTSFORREACTIONSSHOWNINEQUATIONS(7), (8), (9), (10), AND(n)

Mg NAP/gmwettissue0=75b

=104c=495d=8S4ReactioninvolvedXâ€”

Â»a3^>bb-*cc->6x-^ddâ€”>xFractionalturnover

oÃproduct/hrÂ¿

=1.5i=0.38m=1.18n=5.6/=0.05Rate

ofreaction(Wt

P/gm/hr)Fi=113t>,=

40us=580t>6=580Â»7=

42r8=21

DNA to only 50 per cent of that seen in the non-irradiated control tissue. It was proposed that x-radiation blocked the net synthesis of DNA butdid not affect the "turnover" of the pre-exist

ing DNA.Significance of constants and implications con

cerning steady state.â€”The values for the variousconstants together with the amount of P in eachof the nucleic acid fractions, as taken from Table1, enable us to calculate the rate at which P ispassing via the various pathways in equation (7).These rates are recorded in Table 5.

As may be seen in Table 5, the amount of nucleic acid P, in this case presumably intact PNA,that is going back and forth each hour between anonsedimentable form in the supernate (6) andthe particulate lipoprotein-nucleoprotein form (c)is 75 per cent of all the PNA P in the cell. It showsalso that the fractional turnover between S-PNAand P-PNA is nearly as great as that between precursor and N-PNA, and should make one cautious about concluding that a certain fraction ismetabolically less active than some other merelybecause its specific activity is much lower. Table 5also appears to justify the original approximationthat this system could be treated as a steady state,since even the slowest reaction (x ;=Â±6) has a turnover of 38 per cent/hour where the net increase

due to growth is of the order of 2-2.6 per cent/hour.

SUMMARY AND CONCLUSIONSMice bearing a transplanted mammary carci

noma have been sacrificed at time intervals varying from 15 minutes to 16 hours after the intraperi-toneal injection of P32. The tumors have beenfractionated by differential centrifugation to yieldnuclei (N), microsomes (M), an ultrasedimentableparticulate fraction (U), and a supernatant fluid(S). From each of these fractions the nucleicacids, phospholipids, and "protein residues" have

been isolated and the specific activities of the P ofthese constituents measured along with the specific activity of the inorganic P derived from thewhole homogenate.

Next to inorganic P, of those fractions studiedthe "protein residue" P showed the most rapid

initial rise in specific activity. However, the shapeof its specific activity-time curve leads us to conclude that it is a heterogeneous fraction with respect to its P, part of which equilibrates veryrapidly with inorganic P and the rest more slowly.Nevertheless, it has all equilibrated by 16 hours.No significant differences were observed in thespecific activities of this "protein residue" P de

rived from any of the cytoplasmic fractions, butthe "protein residue" P from the nuclei consistent

ly showed a slightly lower specific activity.The phospholipid P from the cytoplasmic par

ticulate fractions was indistinguishable on thebasis of specific activity. However, the supernatant phospholipid P showed a significantly higherspecific activity at early time intervals but itstime curve was approached by the curve of theparticulate phospholipid P as both of them approached equilibrium with the inorganic P at 16hours. Again, the nuclear phospholipid P consistently showed a slightly lower specific activitythan that from the other fractions.

The pentosenucleic acids could be placed inthree metabolically distinct fractionsâ€”the nuclear(N-PNA), the cytoplasmic supernatant (S-PNA),and the cytoplasmic particulate (P-PNA). The specific activity-time curves for these three fractionsand for the DNA were consistent with the assumption that the P of DNA, N-PNA, and S-PNA camefrom precursors whose specific activities had essentially identical time courses. This is interpreted tomean that the immediate precursors (perhaps amixture of desoxypentose- and pentose-mononu-cleotides) were all in rapid equilibrium with somecommon P donor (such as ATP). The data werealso consistent with the assumption that S-PNAwas the immediate precursor of P-PNA. On the




other hand, the data were not consistent with theprevalent assumption that N-PNA is the precursorof either fraction of cytoplasmic PNA.

Turnover constants, expressed in terms of thatpercentage of the P in a product which passes fromprecursor to product each hour, were calculated as150 per cent for N-PNA, 38 per cent for S-PNA,and 118 per cent for P-PNA. In the case of DNA,the amount of P incorporated per hour was calculated as 5 per cent of that present, whereas thenet increase of DNA was estimated to be 2-2.6per cent/hour. This led to the tentative conclusion that, in a cell preparing to synthesize newDNA, the pre-existing DNA enters into metabolicreactions with its environment and that this preformed DNA goes completely (so far as its P isconcerned) into the metabolic pool of its precursor.

ACKNOWLEDGMENTSIt is a pleasure to acknowledge our indebtedness to Mrs-

Florence Carney and Mrs. Ollie Olsen for the many chemicalanalyses reported and also our particular indebtedness to Dr.Joseph W. Weinberg for his extensive and invaluable helpwith the mathematical analyses.

REFERENCES1. BARNUM,C. P., and HUSEBY,R. A. Some Quantitative

Analyses of the Particulate Fractions from Mouse LiverCell Cytoplasm. Arch. Biochem., 19:17-23, 1948.

2. â€” .^The Intracellular Heterogeneity of Pen toseNucleic Acid as Evidenced by the Incorporation of Radio-phosphorus. Ibid., 29:7-26, 1950.

S. BARNUM,C. P.; NASH, C. W.; JENNINGS,E.; NTOAAKD,O.; and VERMUND,H. The Separation of Pentose and

Desoxypentose Nucleic Acids from Isolated Mouse LiverCell Nuclei. Arch. Biochem., 26:376-83, 1950.

4. BLUMENFELD,C. M. Studies of Normal and AbnormalMitotic Activity. Arch. Path., 36:667-73, 1943.

5. BRUES,A. M.; TRACY,M. M.; and COHN,W. E. NucleicAcids of Rat Liver and Hepatoma: Their Metabolic Turnover in Relation to Growth. J. Biol. Chem., 166:619-33,1944.

6. DUBLIN, W. B.; GREGG,R. O.; and BRODERS,A. C.Mitosis in Specimens Removed during Day and Nightfrom Carcinoma of Large Intestine. Arch. Path., 30:893-95, 1940.

7. EULEH, H. v., and HEVESY,G. Wirkung der RÃ¶ntgenstrahlen auf den Umsatz der NukleinsÃ¤ure im Jensen-Sarkom. II. Arkiv Kemi, Mineral., Geol., Vol. 17A, No.30, 1944.

8. HUSEBY,R. A., and BAHNUM,C. P. Investigation of thePhosphorus-Containing Constituents of CentrifugallyPrepared Fractions from Mouse Liver Cell Cytoplasm.Arch. Biochem., 26:187-98, 1950.

9. JEENER,R., and SZAFARZ,D. Relations between the Rateof Renewal and the Intracellular Localization of Ribo-nucleic Acid. Arch. Biochem., 26:54-67, 1950.

10. KENNEDY,E. P. Phosphoprotein of Ehrlich Ascites Tumor.Proc. Am. Assoc. Cancer Research, 1:29, 1953.

11. OSGOOD,E. E.; Li, J. G.; TIVEY,H.; DUERST,M. L.; andSEAMAN,A. J. Growth of Human Leukemic Leucocytesin Vitro and in Vino as Measured by Uptake of P32 inDesoxyribose Nucleic Acid. Science, 114:95-98, 1951.

12. SCHULMAN,J., and FALKENHEIM,M. Review of Conventions in Radiotracer Studies. Nucleonics, 3:13-23, 1948.

13. STEVENS,C. E.; DAOUST,R.; and LEBLOND,C. P. Rateof Synthesis of Desoxyribonucleic Acid and Mitotic Ratein Liver and Intestine. J. Biol. Chem., 202:177-86, 1953.

14. VEHMUND,H.; BARNUM,C. P.; HUSEBY, R. A.; andSTENSTROM,K. W. The Effect of Roentgen Radiation onthe Incorporation of Radiophosphorus into NucleicAcids and Other Constituents of Mouse Mammary Carcinoma. Cancer Research, 13:633-38, 1953.



1953;13:880-889. Cancer Res Cyrus P. Barnum, Robert A. Huseby and Halvor Vermund Mammary CarcinomaNucleic Acids and Other Compounds of a Transplanted Mouse A Time Study of the Incorporation of Radiophosphorus into the

Updated version

http://cancerres.aacrjournals.org/content/13/12/880

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/13/12/880To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



a time study of the incorporation of radiophosphorus...

Documents