THE DYNAMICS OF LEUKOPOIESIS AND LEUKOCYTOSIS, ASSTUDIED BY LEUKOPHERESIS AND ISOTOPIC

TECHNIQUES1

By CHARLES G. CRADDOCK, JR.,2 SEYMOUR PERRY, AND JOHN S. LAWRENCEWITH THE TECHNICAL ASSISTANCE OF MARY H. BAKER AND GLORIA PAUL

(From the Department of Medicine, School of Medicine, University of California, and theHematology Research Laboratory, Veterans Administration Center,

Los Angeles, Calif.)

(Submitted for publication July 25, 1955; accepted November 7, 1955)

Previous studies (1, 2) in normal and irradiateddogs employing the technique of leukopheresishave suggested some hitherto unknown relation-ships between the number of circulating leuko-cytes and the reserves of these cells elsewhere inthe body. By way of introduction and as back-ground for the data to be presented it is necessaryto review briefly and interpret some of these find-ings.The acute response to the production of severe

leukopenia by leukopheresis is qualitatively simi-lar from animal to animal and is divisible intothree phases. When leukopheresis is terminatedthere is a delay before the peripheral white bloodcell count begins to rise (phase I). Followingthis there occurs a linear rise in peripheral bloodcount over a six to eight-hour period (phase II).After a slower rise to a peak peripheral white cellcount, usually 150 to 200 per cent of baseline, thewhite cell count stabilizes at this new level forseveral hours (usually 24 to 36); this last periodis denoted as phase III. Phase I, or the delay pe-riod before the white blood cell count begins torise, is not observed unless the count is reduced to1,500 or 1,000 cells per mm3. Less marked de-grees of leukopenia, induced by slower rates ofclearance of white cells from the peripheral blood,may be maintained by continued leukopheresis forseveral hours but are not followed by the periodof sustained leukopenia (phase I). This is in spiteof the removal of many more cells than would beobtained during a short, rapid leukopheresis whichproduces marked leukopenia. On the other hand,phase I is observed after marked (1,500 per mm8or less) leukopenia no matter how long leuko-

1 This work was supported by grants from Parke-Davisand Co., the U. S. Public Health Service, and Gladys F.Bowyer Fund.

2 Markle Foundation Scholar in Medical Science.

pheresis is continued after leukopenia is induced.For these reasons phase I is thought to reflect amore rapid rate of dispersal of leukocytes to theperiphery which offsets the accelerated entry ofcells from the marrow into the circulation. Inother words, the evidence suggests that the changefrom a stationary leukopenia (phase I) to a rap-idly rising level (phase II) is due to a sudden al-teration in the rate of distribution of leukocytes toperipheral tissues. Once this "priming" has oc-curred in phase I, the blood white cell count be-gins to rise rapidly, reflecting the accelerated entryof leukocytes into the circulating blood.

Phase II represents a period of greatly acceler-ated entry of leukocytes into the peripheral blood,which continues until the new maximum isreached. Rapid leukopheresis with the removalof large numbers of cells may be carried out for aprolonged period without the production of leuko-penia if the procedure is instituted during phase II(see Figure 2).Phase III represents a period of entry of leuko-

cytes into the peripheral blood which has slowedto approximately that at time zero. Leukophere-sis instituted during phase III results in the samedegree of leukopenia produced by the same rateof leukopheresis as carried out initially. In mostanimals this type of response occurs repeatedlyuntil about the third or fourth day. By this timeacceleration in the growth of myeloid tissue hasoccurred so that the delay after cessation of leuko-pheresis is shorter and the leukocytosis developsmore rapidly. The maximal white cell count inthe peripheral blood after leukopheresis on thethird day varies from animal to animal; in most itis similar to the level after previous leukophere-ses, in some it is higher and in others it is lower.All animals show a slower, progressive rise to a

285

CHARLES G. CRADDOCK, JR., SEYMOUR PERRY, AND JOHN S. LAWRENCE

250

200uCD

150

00

50

3

2

36

TIME IN HOURSFIG. 1. SCHEMATIC REPRESENTATION OF THE TYPICAL RESPONSE OF THE

CIRCULATING LEUKOCYTE COUNT TO LEUKOPENIA INDUCED BY A SINGLELEUKOPHERESISRemoval of leukocytes terminated at beginning of Phase 1. See text for

explanation.

DOG 36-62 CONTINUOUS EXCHANGE -

DATE 2-23-55WT. 46 lbs.BLOOD VOL. 2.20 liters

500pc Pn GIVEN ONE HOUR BEFORE EXCHANGE

TOTAL CIRCULATING WBC (InitIal) 1.909 X 10

TOTAL WBC REMOVED 7.694 x le'%* TOTAL CIRCULATINGlINITIAL) VWCREMOED PER HOUR

0 0

//\ 6il

HOURS

DAY I 2 345676 I3

FIG. 2. INDUCTION OF LEUxOPENIA BY ACU IuKOPHEESIS FoLLowED BY PROLONGE, CONTINUED IUKO-PHEIESIS (Ovn 8 HouRs) INSTITUTED DURING TE PERIOD OF DEVLOPING LEUKoCYToSIS FOLLOWING THE INnIALDEPLETIONNote that leukopenia was not induced during this time despite removal of over 7 X IO' leukocytes. The data de-

rived from the specific activity of white cel DNA phosphorus (expressed as counts per minute per microgram ofDNA phosphorus) indicate that the removed cells were largely unlabeled.

286

wz-jw

4n

a.0

1--zhiU

aiA-

70.

41za

% 5

EU

4

.t 3a.

204

CO:

LEUKOPOIBSIS AS STUDIED BY LEUKOPHERESIS AND ISOTOPES

8Dom 3NI13SYS ;° 1N3OM3d

o

z

id t

I IA

4,

co

..

'~~~~~~~~~~~'

Sz

0 0e0)

UA C's

0

n e

0

0eC., o O

0 4)

U =

0-4)

, *

287

CHARLES G. CRA-DDOCK, JR., SEYMOUR PERRY, AND JOHN S. LAWRENCE

new maximum over the next two to three days(Figure 3). The slope of this delayed rise issimilar in normal animals regardless of the levelattained during phase II.The morphology of the peripheral blood leuko-

cytes and the marrow have been correlated withthese quantitative changes. There is no change in-dicative of increasing immaturity of the marrowor circulating leukocytes until the third or fourthday. Undoubtedly some stimulation to leuko-poiesis has occurred during this three-day period.

These data suggest that the initial acute re-sponse to leukopheresis involves the release ofpreformed leukocytes from areas of storage. Thisinitial response does not depend upon growthprocesses. Dogs exposed to LD-50 whole bodyx-ray behave similarly, though the leukocytosis isquantitatively less, until their marrow has be-come depleted of myeloid cells. Even after thedevelopment of radiation leukopenia, the response

i

R

0hia.

Hot=SDAY I DAY 2

to leukopheresis is qualitatively similar, as shownin Figure 5. A fundamental relationship seems toprevail between the leukocyte reserves and the de-gree of leukocytosis which develops after leuko-pheresis. The removal of leukocytes from theperipheral blood calls forth the release of cellsfrom the marrow until a new level has been at-tained. The height of the new level depends uponthe size of the reservoir of leukocytes. In the ir-radiated animal the level becomes progressivelylower as the marrow reserve falls. Nevertheless,the same pattern prevails and leukocytes are re-leased into the circulating blood in some sort ofproportion to the number of cells held in reserve.The reserves of preformed leukocytes in the

body assume fundamental importance as the sourceof supply of cells taking part in acute leukocytosisor pyogenic reactions of various sorts. It is,therefore, important to delineate the size and siteof such reserves of leukocytes. There are essen-

DAY 3 DAY4 6 8 H1 eS25 M

FIG. 4. RESPoNSE TO LEUROPHERESIS REPEATED AT 24-HOUR INTERVALS FOR THREE DAYS IN AN ANIMALEXPOSED TO LD-50 WHOLE BODY IRRADIATION ONE HOUR PRIOR TO THE INITIAL LEUKOPHERESIS

288

LEUKOPOIESIS AS STUDIED BY LEUKOPHERESIS AND ISOTOPES

DOG 35-71 IRRADIATED WITH 125 R

DATE 5-5-55 to 5-9-55

WT. 35 lbs.

BLOOD VOL. 1.6 liters

IS DAYS BEFORE EXCHANGE

TOTAL CIRCULATING WBC (initial) 0.33 x 10

TOTAL WBC REMOVED 0. 165 x 10

% of Totol Initial Circuloting WBC

Removed Per Hour

DAYS 3 5 7C9u a I 2 3 4 5 aa* aE 9X

FIG. 5. LEUKOPHERESIS OF AN ANIMAL RENDERED LEUKOPENIC BY 125r WHOLE BODY X-RAY Two WEEKSPREVIOUSLY

Note that the level of circulating white cells rises after each leukopheresis to a level exceeding baseline, althoughthe rise is much more delayed than in a normal animal.

tially three potential sites of such a "ready re-

serve": 1) Leukocytes sequestered in variouscompartments of the vascular bed, 2) leukocytesheld within the marrow organ, and 3) leukocyteswhich have migrated out of the vascular bed(peripheral tissues).

It is well known that acute fluctuations in thenumber of circulating leukocytes occur followinghemodynamic alterations in such organs as thelung and spleen. Although the participation ofsuch factors in the response to leukopheresis can-

not be entirely eliminated, the uniformity and mag-

nitude of the response in anesthetized animalsmakes unlikely such events as the leukocytosis ofsplenic contraction as playing an important role.Epinephrine intravenously did not alter the re-

sponse (2), and the results in one splenectomizeddog (Figure 6) did not differ significantly fromnormal.The size of the marrow leukocyte pool has been

estimated to be as much as 100 times that of thecirculating blood (exclusive of marrow) (3, 4).

Certainly our earlier experiments in irradiateddogs (1, 2) suggest there are sufficient numbersin the marrow to replace those circulating many

times over.

Aside from the marrow the main source of leu-kocytes to be considered is the extramedullary, ex-

travascular mass of leukocytes. Osgood, for ex-

ample, (5) estimates that the numbers of leuko-cytes in such areas are so vast as to be many timesthe numbers in the marrow and blood. Whetherleukocytes which have migrated out of the vasculartree can reenter in sufficient numbers to affect theacute response to leukopheresis is unknown.However, this appears improbable from previouswork (2). A heavily irradiated dog, renderedleukopenic by repeated leukopheresis so that fixedagranulocytosis was present showed large col-lections of leukocytes in peripheral areas of infec-tion in the lung and elsewhere. These leukocytesdid not reenter the circulation during or afterleukopheresis.The present experiments were directed pri-

nw

hix 4b0.

_J 400

hoi

0°*. 350

Ji 300

C)I-4ahiJ 250

~2O

to0

I.-zhi 100

0.wIL

289

_ _ _ _ _ _ _ _ _

rs

CHARLES G. CRADDOCK, JR., SEYMOUR PERRY, AND JOHN S. LAWRENCE

marily at the question of the source of those leuko-cytes contributing to the leukocytosis followingleukopheresis. For this purpose the rate of en-trance of newly formed leukocytes from the mar-row has been studied, employing p32 incorporationinto leukocyte DNA. Previous works by Klineand Cliffton (6) and Ottesen (7) have indicatedthe feasibility of this method for determining themean age, life span and turnover rate of leuko-cytes. Both groups of investigators noted a delayperiod of four to six days before labeled myeloidleukocytes appeared in the circulation, attributedto the time required for cells capable of synthesiz-ing DNA at the time of maximum labeling to ma-ture to the point of being released. These workersutilized modifications of the Schmidt-Thannhauserprinciple of nucleic acid separation. Unfortu-

nately, we have been unable to obtain adequateyields of DNA phosphorus uncontaminated byphosphorus from other sources using this methodin the study of dog leukocytes. Similar difficul-ties have been encountered by others (8). Con-sequently, a different technique of DNA extrac-tion developed by Dr. Norman Simmons 8 hasbeen employed. The DNA obtained from leuko-cytes by this method is highly purified, as deter-mined by spectroscopic, ultracentrifugal, x-raydiffraction, and chromatographic techniques.The leukocyte response to leukopheresis in-

volves myeloid leukocytes almost exclusively (1,2). Lymphoid cells generally decrease so that thevast majority of leukocytes are of the neutrophilic

'Chief, Enzyme Chemistry Section, UCLA AtomicEnergy Project.

DOG 11-55 ('DATE 1-3-55WT. 32 lbs.

BLOOD VOL. I

TOTAL CIRCUL

TOTAL WBC F

% ofTvRemoved

SPLENECTOMI ZED)to 1-5-55

1.073 liters

ATING WBC(initial) 1.2 xtO'lREMOVED 3.374 x 10og

*otol Circulating (Initiol) WBCPer Hour

,\ X1 f

22 24 26 28 30 32 48 50

DAY

52 54 56 v

4 5678 910

FIG. 6. THE LEUKOCYTF RESPONSE OF A SPLENECTOmIZED DOG TO REPEATEDLEUKOPHERESIS

The pattern differs in no way from normal. Large numbers of nucleated red cellsappeared after the initial leukopheresis; the white cell determinations were correctedfor this.

450

400

350

co3.

z-JnU)

co

300

250

200

-

150

100

zLuJ

a.

o50

HOURS

290

LEUKOPOIESIS AS STUDIED BY LEUKOPHERESIS AND ISOTOPES

SOLVENT BOUNDARY

THYMINE

URACIL >

CYTOSINE

ADENINE

GUANINE E-

START -T-

r IG. /. rAPER lHROMATOGRAM :iHOWING rURITY(LEFT) OF DoG LEUKOcYTE DNA OBTAINED BY METHODDESCRIBED IN TEXT AS COMPARED TO VIRUS PREPARATIONUSING ScHMIDT-THANNHAUSER TECHNIQUE (RIGHT).

variety. No attempt has been made to obtainmore complete separation of lymphocytes andneutrophiles.

METHODS

Leukocyte suspensions were obtained froir dogs.Siliconed glassware was used throughout and thl entireprocedure was carried out at 40 C. In animals su. jectedto leukopheresis, sufficient numbers of leukocytes wereobtained so that separation by simple centrifugatioiu wasadequate. In the case of less rich sources, leukocyteswere separated from whole blood by means of dextran(Commercial Solvents, av. molecular wt. 120,000). Fif-teen ml. of 6 per cent dextran per 25 ml. blood was foundto give optimal separation. After sedimentation of thered cells the supernatant plasma was centrifuged at 800to 1,000 r.p.m. for 15 minutes in an International 2 cen-trifuge to pack the leukocytes and leave the platelets insuspension. After washing twice with saline the leuko-cytes were diluted to a specific volume, mixed thoroughlyand counted in quadruplicate. The theoretical amount ofDNA in the cell suspension could thus be calculated.

The leukocytes were then washed with 1 per cent citricacid until the supernate was clear and then homogenizedfor 5 minutes in 1 per cent citric acid. They were then

* washed twice, or until the supernatant fluid was clear,with 0.2 per cent citric acid and twice with cold saline.The separated nuclei were subjected to a procedureinvolving the dissociation of nucleoprotein with 30 percent sodium xyiene sulfonate at pH 7, deproteiniza-tion by dilution, filtration, and adjustment of pH to3.8 and precipitation of the DNA from neutral solu-tion with isopropyl alcohol. The details of this procedurewill be published subsequently by Dr. Simmons and willbe the subject of a separate report concerning leukocytenucleic acids from this laboratory.The yield of DNA was determined by the diphenyl-

amine reaction and by ultraviolet spectroscopy usingpure DNA supplied by Dr. Simmons as standard. Or-ganic and inorganic phosphorus determinations weremade on each specimen and in all instances the formerchecked closely with the yield of DNA. Inorganic phos-phorus amounted to less than 10 per cent and the countsof radioactivity were corrected for the small amountpresent. RNA contamination, as determined by theorcinol reaction, was less than 5 per cent. Quantitativechromatography 4 showed negligible RNA contamina-tion on selected samples. An example of a chromatogramis shown in Figure 7.The P' employed was obtained from Abbott Labora-

tories. P'2 counting was carried out in the RadioisotopeSection, Research Division, Sawtelle Veterans Admini-stration Hospital using a Tracerlab Superscaler and aTGC-2 Geiger-Muller Tube with a statistical reproduci-bility of + 1 to 2 per cent. All values have been correctedfor decay.

P3 was administered as disodium phosphate intra-venously in the amount of 500 /Ac. regardless of the sizeof the animal. (The dogs ranged in weight from 30 to 50

2* I x_ D06 3-6922 ,.w oDOG 14-64

v-1vDOG 2-63.2 A-- D0 14-61

6s.

14E

12.

10.

46

06

4

02

0912 34 56 7689 10TIME I N DAYS

FIG. 8. SPECIFic ACTIVTY OF DNA PHOSPHORUS INLEUKOCYTES SAMPLED AT INTERVALS AFTER ADMINISTRA-TION OF 500 MICROCURIES P' AT TIME 0 To FOUR NoRiALDOGSThese animals were not leukopheresed.4 Dr. Gian L. Turco, Research Fellow, California In-

stitute of Technology, graciously performed these studiesfor us.

291

L'i,r '7 A -rn- t--r ^.^ ^ 'C

I

iy.!-i.Agge

CHARLES G. CRADDOCK, JR., SEYMOUR PERRY, AND JOHN S. LAWRENCE

z

b 1001- 90.

<70

i 604

50

40

, 30

4- 20

100'o 017

x-s Dog 3-690-0 Dog 14-64w-s Dog 12-63a-- Dog 14-61*-o Moen

PERCENT CHANGE in SPECIFIC ACTIVITY of

DNA PHOSPHORUS In 4 NORMAL DOGS

O 2 3 4 5 6 7 8 9 10 11 12

T ME IN DA YS

FIG. 9. SAME DATA SHOWN IN FIGURE 8 PLOTTED ASPER CENT OF THE MAXIMUM SPECIFIC ACTIVITY

lbs.) Those animals who were not subjected to leuko-pheresis were bled from the jugular vein under sterileconditions, into ACD-dextran solution. After separa-lion of the leukocytes, the red cells were readministeredat the time of the next bleeding. All animals withstoodthe procedure without showing ill effects. No hemato-logic depression was noticed. Although it is most prob-able that the amount of irradiation delivered by the PUhas some effect on the processes of cellular growth, theuse of the technique in acute studies of differences inentry rate, distribution, and source of leukocytes shouldbe valid.

RESULTS

Figures 8 and 9 show the results in four normaldogs whose leukocytes were sampled at intervalsafter administration of the isotope. These dogswere not subjected to leukopheresis. The curvesare not comparable from animal to animal in terms

LEUIIOPHERESIS

10

! 14

w 12

a'a

4

0-o DOG 35-52

¢XX DOG 60-60

_ MEAN, SP AC,DNA P4 NORMAL DOGS

2 3 4

TIME IN DAYS

FIG. 10. ACCELERATION IN THE RATE OF ENTRANCE OFLEUxOCYTES CONTAINING DNA LABELED WITH PI, 500i, PU ADMINISTERED AT TIME 0

Leukopheresis carried out in dog 35-52 and 6-60 atindicated points resulting in the typical normal leukocyteresponse.

of the amount of P32 incorporated into DNA sinceno attempt was made to administer p82 on a weightor blood volume basis. However, the form of thecurve for each animal is similar. The mean curve

(Figure 9) resembles those observed by others inhumans. The rise in specific activity is earlierthan in human experiments and the logarithmiccurve reaches its maximum on the fourth to fifthday, followed by a brief plateau. Sufficient obser-vations have not been made to describe fully thedescending portion of the curve and for the pur-poses of these experiments emphasis has beenplaced on the rising portion of the curve as show-ing the time of appearance of labeled cells in thecirculating blood. Calculations of mean cell age

DOG 5-66

100

)- 90-

70.0

hi50.

a

, 30.20

10.

n-s % Chaep Is SpeIdfle ActvlityDN4A Phosphorus previously

Loukophresed Do-%% Change Is Spoelifi Activity

4 Normal Dos

0 1 2 3 4 5 6 7 8 9 I

TIEE IN DAYS

FIG. 11. THE RATE OF APPEARANCE OF LEUKOCYTESWITH PU LABELED DNA INTO THE PERIPHERAL BLOOD INAN ANIMAL LEUKOPHERESED FOR THREE DAYS PRIOR TOADMINISTRATION OF 500 JAC. P' INTRAVENOUSLY AS COM-PARED TO THE CURVE FOR NORMAL DOGS NOT SUBJECTEDTO LEUKOPHERESIS

and survival time have not been attempted fromthese limited data, but they suggest a shorter lifespan for dog leukocytes than that reported forhuman white cells (6, 7).

If leukopheresis is carried out within two hoursafter administration of p32 the removed cells con-tain no labeled DNA. Furthermore the cells re-moved during leukopheresis 24 hours later are stillunlabeled, even if large numbers of cells are re-moved, as is shown in Figure 2. However, re-peated leukopheresis does lead to some accelera-tion in the time of appearance of significantly la-beled leukocytes, as is shown in Figure 10. Theseresults are compatible with a large marrow poolof mature leukocytes which precede the entry ofthose cells which incorporated p82 during the

292

LEUKOPOIESIS AS STUDIED BY LEUKOPHERESIS AND ISOTOPES

process of DNA synthesis. The same type of datais shown in more detail in Figure 2.

Figure 11 shows the curve of specific activityof DNA phosphorus of a normal dog previouslysubjected to leukopheresis for 3 days before giv-ing the isotope. The maximum of specific activityoccurred 24 hours earlier than that of normal dogsnot subjected to leukopheresis which suggests themore rapid release of leukocytes from the marrowat this time. This would be expected from thestimulation of leukopoiesis brought about by leu-kopheresis on three successive days.

Leukopheresis was carried out in three normalanimals at a time when those leukocytes whoseDNA was maximally labeled with p32 were stillin the marrow but shortly before they would benormally released in significant numbers. Thesedogs were leukopheresed on the third day afterbeing given 500 jc. P32 intravenously. At thispoint those cells coming from the marrow shouldhave a high specific activity, while cells whichwere in peripheral areas of storage should con-

tain negligible amounts of P32 labeled DNA. Be-cause of this difference in degree of labeling ofmarrow and peripheral leukocyte pools it shouldbe possible to determine the source of leukocytescontributing to the response to leukopheresis.

Figures 12 and 13 show the results in two ofthese animals. The third showed an identical re-sponse to Figure 13. All dogs showed a relativelylow specific activity of leukocyte DNA P in thosecells removed during the first leukopheresis.There was a marked rise in specific activity of cir-culating leukocyte DNA P during the next fewhours when the leukocyte count rose after cessa-tion of leukopheresis. It is clear, therefore, thata high percentage of the cells contributing to phaseII came from the labeled marrow leukocyte pool.Both dogs showed a fall in specific activity of leu-kocyte DNA P from this peak the next day. Dog20-72 (Figure 12) showed very little change inspecific activity incident to the second or thirdleukopheresis. However, Dog 36-70 (Figure 13),after the second leukopheresis, showed a repetition

DOG 20- 72 GIVEN SOO51c pFr THREE DAYS BEFORE EXCHANGE

DATE 5-16-55 to 5 -18 -55 TOTAL CIRCU

WT. 34 lbs. TOTAL WBC

40 4001 BLOOD VOL. 2.167 liters

35

30

s25<L2-9

, is

< a0 ?

EWBC

JLATING W8C (Initial) Z8l7 x 10

REMOVED 2.450 x 1O0

f Total Initial CirclatingRemowd Per Hour

DAY0 1 2 3 4 5 aS XaE " a a to

FIG. 12. THE RESPONSE OF A NORMAL ANIMAL TO REPEATED LEUKOPHERESIS INSTITUTED THREE DAYS AFTERADMINISTRATION OF 500 /Ac. P33 INTRAVENOUSLY

Note the marked rise in specific activity of DNA P in those cells contributing to the leukocytosis after the firstleukopheresis.

293

II:91

CHARLES G. CRADDOCK, JR., SEYMOUR PERRY, AND JOHN S. LAWRENCE

DOG 36-70 4-18 - 55 BL.VOL 2 2 LSOOuc PE given Total Ciro. WBC (initial) 1.8 X l0o* % Ciro. Cells

9 at 0 time Total WUC Removed 3 06 x id v,w R

8 400

7 350Sz s

6 i300 S >

IE5 uj 250 >

DAYS 2

FIG. 13. THE RESPONSE OF A NORMAL ANIMAL TO

REPEATED LEUKOPHERESIS INSTITUTED THREE DAYS

AFTER ADMINISTRATION OF5050c. PNote the rise in specific activity of DNA P during

the period of developing leukocytosis after the first two

leukophereses.

of the rise in specific activity observed after the

first leukopheresis. Whether or not most of the

maximally labeled cells appear after the first leu-

kopheresis or later, as in the case of this dog prob-

ably depends upon the timing of the procedure.

Both animals exhibited a fall in specific activity

of leukocyte DNA P during leukopheresis. This

suggests that leukocytes which are older than the

average circulating cell population contribute to

the mass of cells removed during leukopheresis.

The marked rise in specific activity of DNA P

after leukopheresis on one occasion in Dog 20-72

and on two occasions in Dog 36-70 indicate that

cells were contributed from the highly labeled

marrow pooi in much higher percentage than from

the relatively unlabeled peripheral tissues.

DISCUSSION

These data indicate the feasibility of experi-

ments on leukocytes employing a method of DNA

extraction which we believe to be superior to the

Schmidt-Thannhauser principle used by most

other workers.

Preliminary results indicate a life span for dog

leukocytes considerably shorter than those re-

ported for humans.

The changes in the concentration in the periph-

eral blood of white cells having DNA labeled with

p32 support the concept, presented earlier in this

paper, that the bulk of cells contributing to the

leukocytosis after leukopheresis are released from

the marrow and do not come from extravascular

and extramedullary areas. Therefore, the re-

sponse to leukopheresis is believed to representan accurate measure of the capacity of the marrowto supply the peripheral blood with leukocytes.These findings may be correlated with previousdata in irradiated animals (2) to indicate that themarrow represents the main, if not the only, re-serve supply of leukocytes available for releaseinto the circulation during leukocytosis of the typeobserved here.

This does not necessarily mean that leukocytosiscan occur only by means of accelerated release ofcells from the marrow. There is ample evidenceindicating the importance of redistribution of leu-kocytes within the vascular compartment secondaryto various hemodynamic alterations, causing fluc-tuations in peripheral blood white cell concentra-tion (e.g., exercise, epinephrine). Other types ofleukocytosis have not been studied by techniquessimilar to those employed here. It may be that theleukocytosis subsequent to leukopheresis is uniquein its mechanisms. Nevertheless, it has revealedan interesting and intricate relationship betweenthe utilization rate of peripheral blood leukocytesand the release of cells from the marrow. Thisconcept is presented in diagrammilatic form inFigure 14.The rate of development and degree of leuko-

cytosis after leukopheresis depend upon the nu-merical size of the leukocyte reserve. The evi-dence presented supports the belief that this re-serve is primarily within the marrow. Normallythe marrow reserve is replaced at such a rate thatrepeated or continuous leukopheresis as presentlyapplied fails to greatly affect its size. However,damage by irradiation which reduces the replace-ment of the reserve by growth depresses the re-sponse to repeated leukopheresis. The level at-tained in the blood after leukopheresis becomesprogressively less as the size of marrow reservebecomes reduced.

This relationship between the blood level of leu-kocytes attained after leukopheresis and the sizeof the marrow leukocyte reserve suggests that theadequacy of the marrow reserve and replacementby growth may be indirectly assessed by thismeans. Whether other types of stimulation to leu-kocytosis such as inflammation or infection are as-sociated with a similar relationship between bloodand marrow leukocytes remains to be investigated.The data suggest that the release of cells from themarrow is geared to the utilization rate of leuko-

294

LEUKOPOIESIS AS STUDIED BY LEUKOPHERESIS AND ISOTOPES

PERIPHERAL

A BLOO D

[ANauS// brm 1~01 TISSUES

100%

3.

RELATION OF MARROW AND PERIPHERAL BLOOD LEUKOCYTE LEVELS

IN LEUKOPHERESIS

NORMAL RESTING STATE. FLOW AT As. FLOW AT B

. IMMEDIATELY AFTER LEUKOPHERESIS

3. PHASE I PRIMING OF PERIPHERAL BLOOD AND TISSUES

4. PHASE x PERIPHERAL LEUKOCYTOSIS A> B

6. PHASE x STABILIZATION OF BLOOD WBC AT NEW LEVEL. A .

SLEUKOCYTOSIS DUE TO MARROW MYELOID HYPERPLASIA

_,A * g

FIG. 14. DIAGRAM

cytes in the periphery. Such utilization may in-volve actual migration of cells from the blood intothe tissues or destruction of leukocytes in the cir-culation by agglutination, lysis, etc. The leuko-cyte count at a given moment does not revealwhether the rate of utilization is high, low, or nor-

mal, and hence cannot yield information as to therate of entry of leukocytes from the marrow.

However, a persistently elevated neutrophile countmust reflect an adequate marrow reserve of leuko-cytes. Conversely, a low neutrophile count, de-spite obvious stimulation to neutrophilic leuko-cytosis, must mean a depleted marrow reserve,

as in the leukopenia of overwhelming pyogenic in-fection. Presumably in this latter situation thecontinued demand for leukocytes in the peripheryoutstrips the ability of the growth processes toreplace those cells released from the marrow res-

ervoir at a rate sufficient to maintain the volumeof the marrow reserve. This phenomenon couldresult from a great demand for leukocytes im-posed on a marrow whose reserves were alreadydepleted by previous disease even if the actualrate of cell synthesis were continuing at a normal

or accelerated rate. A small percentage of mar-

row myeloid cells are myeloblasts, and since theserepresent the ultimate source of mature leukocytes,depletion of the marrow of its contents of maturecells requires time for replacement. Assumingthat the level of peripheral leukocytes in the.

presence of a stimulus to leukocytosis dependsupon the size of the marrow reserve, as suggestedby our experiments, the blood leukocyte level willremain low despite intense leukopoiesis if the re-

serve has been greatly reduced.Looked upon in this way, the large size of the

marrow reserve of mature neutrophilic leukocytesin proportion to those circulating becomes an in-tegral factor in leukocyte physiology. The pres-

ence of this reserve, described so clearly by Kin-dred and Yoffey (3, 4), is necessary to preventthe occurrence of marrow depletion by the sud-den imposition of a greatly increased rate ofperipheral utilization of leukocytes. The forma-tion of new leukocytes is sufficient to balance thedemand for leukocytes under resting conditions.However, one can visualize a situation wheregrowth would be inadequate to sustain the need

MARROW

2.

295

CHARLES G. CRADDOCK, JR., SEYMOUR PERRY, AND JOHN S. LAWRENCE

for myeloid leukocytes if a sudden demand, manytimes greater than the resting requirement, wereimposed before marrow myeloid hyperplasia hadtaken place. The marrow reserve affords a hugebuffer between supply by growth, which is fixedwithin certain limits, and demand, which may varywidely. The rate and degree of replenishment ofthe peripheral blood with leukocytes under con-ditions of increased need is initially geared to themarrow reserve and not to the rate of growth ofmyeloid tissue. Thus, heavily irradiated animalsrespond to leukopheresis with a rapid and markedleukocytosis, until the marrow reserves become de-pleted. Even after marked depletion of the mar-row and peripheral leukopenia, the stimulus as-sociated with leukopheresis leads to release ofcells from the marrow, but the rise peripherallyis very slow and the maximum is very low. Themechanisms involved in the accelerated release ofcells from the marrow reserve induced by leuko-pheresis remain to be investigated. Whether theprimary stimulus is a change in the level or con-centration of leukocytes in the peripheral bloodor the actual increased removal of cells is not en-tirely clear. However, the results in irradiateddogs (see Figure 5, for example) suggest thatthe removal of cells represents a stimulus over andabove that of leukopenia itself. Thus, in this ani-mal who was leukopenic at the time leukopheresiswas instituted (2,070 per mm.3), the marrow re-sponded by releasing cells more rapidly until anew maximum was reached. The mode of trans-mission of the stimulus to the marrow resultingin accelerated release of leukocytes from the re-serve is presumed to be humoral. However, at-tempts to detect such a humoral agent have beenfruitless thus far.

These experiments have provided some back-ground for further study of leukocyte physiology.It is hoped that the information relating leukopoie-sis, marrow reserve and peripheral blood leuko-cytes will aid those interested in derangements inleukopoiesis and leukemia.

SUMMARY

1. The concept is presented that the majority ofleukocytes which contribute to an acute leukocyto-sis are released from areas of storage. Increasedproduction of leukocytes by acceleration of leuko-poiesis occurs only after considerable delay, which

in the case of the dog subjected to repeated leuko-pheresis is about three to four days.

2. Evidence is presented that the chief source ofmature leukocytes taking part in an acute leuko-cytosis, such as that observed following leuko-pheresis, is the marrow reserve. This evidencewas obtained by taking advantage of the delaybefore leukocytes whose DNA becomes labeledwith P32 during growth move from the marrowinto the peripheral blood.

3. Preliminary results of the rate of DNA syn-thesis in dog neutrophile leukocytes are presented,employing a new method of desoxyribose nucleicacid extraction. These results suggest a shorterlife span for this cell in the dog than that reportedby others for man.

4. The data are briefly discussed and a con-cept is formulated which clarifies the relationshipof the marrow reserve of leukocytes to those invascular and extravascular areas.

ACKNOWLEDGMENTS

Sincere appreciation is expressed for the unlimitedhelp given by Mr. Max Field and Miss Panchita Thomasof the Radioisotope Unit, Research Division, SawtelleVeterans Administration Hospital under the directionof Dr. Franz Bauer. We are also indebted to Dr. Nor-man Simmons for providing us with his method of DNAextraction and to Dr. Gian D. Turco for his chromato-graphic analyses.

REFERENCES

1. Lawrence, J. S., Physiology and functions of thewhite blood cells. The Minot Lecture. J.A.M.A.,1955, 157, 1212.

2. Craddock, C. G., Jr., Adams, W. S., Perry, S., Skoog,W. A., and Lawrence, J. S., Studies of leukopoiesis.The technique of leukopheresis and the responseof myeloid tissue in normal and irradiated dogs.J. Lab. & Clin. Med., 1955, 45, 881.

3. Kindred, J. E., A quantitative study of the hemopoieticorgans of young adult albino rats. Am. J. Anat.,1942, 71, 207.

4. Yoffey, J. M., The quantitative study of the leukocytes.Ann. New York Acad. Sc., 1955, 59, 928.

5. Osgood, E. E., Number and distribution of humanhemic cells. Blood, 1954, 9, 1141.

6. Kline, D. L., and Cliffton, E. E., Life span of leuco-cytes in man. J. Applied Physiol., 1952, 5, 79.

7. Ottesen, J., On the age of human white cells in periph-eral blood. Acta physiol. Scandinav., 1954, 32, 75.

8. Drasher, M. L., A criticism of the indiscriminate useof the Schmidt-Thannhauser method for the frac-tionation of nucleic acids in biological material.Science, 1953, 118, 181.

296