A SPECIFIC INHIBITOR FORHUMANDESOXYRIBONUCLEASEANDAN INHIBITOR OF THE LUPUSERYTHEMATOSUS

CELL PHENOMENONFROMLEUCOCYTES1

By NATHANIEL B. KURNICK, LAWRENCEI. SCHWARTZ,SANFORDPARISER,'ANDSTANLEYL. LEE

(From the Department of Medicine, Tulane University School of Medicine, New Orleans,Louisiana; and the Hematology and Pathology Laboratories of the Mount Sinai

Hospital, New York, N. Y.)

(Submitted for publication September 29, 1952; accepted December 3, 1952)

In speculating upon the probable role of desoxy-ribonuclease (DNase) in the depolymerization ofdesoxyribonucleic acid (DNA), which character-izes (1) the "lupus erythematosus (L.E.) cell"(2), we have suggested (3), as have Klempererand his associates (4), that an increased serum en-zyme activity, potentiation of an intracellularDNase, or destruction of an intracellular inhibitorof DNase by a serum factor, might be responsible.In a previous communication (5) we reported evi-dence which excludes the serum DNase from re-sponsibility for the phenomenon. Studies on theinteraction of leucocytes and serum, in an ap-proach to the alternative possibilities, disclosed thepresence of a specific intracellular inhibitor of hu-man serum DNase and an inhibitor of the "L.E.cell" phenomenon (6, 7). In the present paper,we shall report upon some of the characteristics ofthese factors.

I. SERUMDNAsE INHIBITION

A. Preparation of leucocyte extractsLeucocytes (WBC) were prepared with the aid of

Bovine fraction I from heparinized blood as previouslydescribed (5) or by centrifugation at 0 to 2°C of fresh"ACD" bloods (120 ml. of 1.47 per cent dextrose, 1.32per cent sodium citrate, 0.48 per cent citric acid plus 480

ILThis work was aided by grants to Tulane UniversitySchool of Medicine (N. B. K.) from the National HeartInstitute, U.S.P.H.S. (No. H-714), the American HeartAssociation, American Cancer Society (recommended bythe Committee on Growth, National Research Council)and the Life Insurance Medical Research Fund, and tothe Mount Sinai Hospital from the Life Insurance Medi-cal Research Fund and A. A. List

2Anna Ruth Lowenberg Fellow in Hematology, MountSinai Hospital.

' Weare indebted to Dr. J. W. Davenport and Mr. JohnGavey of the Medical Department of the New OrleansRegional Blood Center, American Red Cross, for generoussupplies of human blood.

ml. whole blood) followed by collection of the WBClayer and repeated washing at 0 to 2°C with 0.14 MNaCIuntil a good separation of WBCfrom erythrocytes wasobtained. The latter technique was used with large vol-umes of human blood (2500-3000 ml.).

Extracts of WBCwere prepared by either of twomethods:

(a) Suspensions of 24 X 10, or more WBCper ml. in0.14 M NaCl or heparinized plasma were lysed by re-peated freezing in acetone-solid CO, mixtures followedby thawing by immersion in a 37°C water bath. Com-plete. disintegration of cells was accomplished by freezingand thawing four times. An aliquot was clarified by cen-trifugation at 2700 X G (measured at center of tube)for 30 minutes in a Sorvall angle centrifuge. The ma-terial was stored at - 20'C.

(b) Approximately 12 X 10' fresh WBCwere sus-pended in 30 ml. 0.14 MNaCl at 0 to 2°C and blended ina micro Waring Blendor equipped with an ice jacket (8),slightly modified by the addition of a rubber diaphragmseal for the ice jacket, to permit pouring from the Blendorwithout contamination of the material by the ice-salinemixture in the jacket, for 10 minutes (sufficient for com-plete fragmentation of the cells, as demonstrated bymicroscopic examination). The homogenate was cen-trifuged for 15 minutes at 430 X G. The supernatant,representing the bulk of the cytoplasmic material andsome soluble nuclear components, was labelled "cyto-plasmic fraction." This was separated by centrifugationfor 40 minutes at 2700 X G into a "cytoplasmic super-natant" and "cytoplasmic sediment." The sediment ob-tained from the whole homogenate at 430 X Gwas washedrepeatedly with saline in order to obtain a clean "chro-mosomal fraction" and fractionated into nucleohistone(DNH) and "residual chromosomes" after Mirsky andRis (9, 10). From DNH, purified desoxyribonucleic acid(DNA) was prepared according to Mirsky and Pollister(11). All fractions were resuspended in 0.14 MNaa toreconstitute, approximately, the original concentration(12 X 10' WBCin 30 ml.). All procedures were car-ried out at 0 to 2°C. Upon completion of the fractionation,sodium ethyl mercuri-thiosalicylate was added to a finalconcentration of 1: 100,000, and the material was storedat - 20'C. A pale pink color was imparted to the cyto-plasmic fraction and the supernatant obtained therefrom

193

N. B. KURNICK, L. I. SCHWARTZ,S. PARISER, AND S. L. LEE

by the hemoglobin of the RBCwhich remained entrappedin the WBClayers.

B. Inhibition of serum DNase

Each of the WBCextracts prepared by thefreezing-thawing method was tested for DNaseactivity by the methyl green method (3) as modi-fied for serum (12). In no case was any DNaseactivity detected under these conditions of ionicstrength and pH. Tests for DNase activities underother conditions (13-17) were not made, sincethese would not interfere with our experiments.Tests for DNase inhibition against human serumwere performed by substituting the fraction to betested for part or all of the 3 ml. saline used in di-luting the substrate (12). Simultaneous observa-tions on the inhibition of the "L.E. cell" phe-nomenon by each of the fractions are reported be-

E.520 . B

.480 a\''c,440 \-

.400 -\

.360 *\ \'

.320 \\ \

.280 \ >>.E

.240 A

.200

o 30 60 90 120 150 180 210 840MINUITES

FIG. 1. HUMANSERUMDNASEINHIBMON BY HUMANLEucocvT (WBC) EXTRACTS

Each point represents optical density of 3 ml. sub-strate-inhibitor-serum mixture added to 0.5 ml. 0.33 Msodium citrate after incubation for intervals indicatedat 37°C. Each 19 ml. of the incubated mixture contains1 ml. serum and 1 ml. of the following saline WBCextracts: A. None (1 ml. 0.14 M saline), B. Supernatantfrom 6 X 10' frozen and thawed cells, C. "Cytoplasmicsupernatant," D. Unwashed "cytoplasmic sediment," E.Chromosome fraction.

low. The observations on the two types of inhibi-tion parallel each other closely. Rarely, prepara-tions which had previously been found to containinhibitor, failed to inhibit the "L.E. cell" phe-nomenon and only weakly inhibited human serumDNase. As will be seen, these failures were prob-ably due to proteolysis (autolysis), whereas the dis-crepancies result from the greater sensitivity of theDNase system.

Inhibition of human serum DNase was observedwith the frozen and thawed extracts, both beforeand after centrifugation, with the total "cyto-plasmic fraction," and the "cytoplasmic super-natant" prepared from blended WBC. Neitherthe "cytoplasmic sediment" nor the "chromosomalfraction" (nor fractions prepared therefrom)demonstrated inhibition (Fig. 1).

The relationship between the amount of in-hibitor present and the degree of inhibition is ap-parently not linear (Fig. 2), and resembles thecurve reported by Zamenhof and Chargaff ( 13) fortheir yeast DNase inhibitor.

C. Physico-chemical properties of inhibitorProlonged dialysis against frequent changes of

0.14 M saline at 0°C did not alter the inhibitoryactivity of extract fractions. No loss of inhibitionresulted from heating at 56°C for 2 hours (bothunbuffered and buffered at pH 7.5 with tri-methylol-amino-methane buffer [18]). Heatingto 100°C in a boiling water bath for 30 minutescompletely destroyed all inhibitory effect. Theaddition of citric acid to bring the final concentra-tion to 1 per cent, followed by dialysis againstsaline at 0°C (to remove citrate), resulted in lossof inhibitory activity by "cytoplasmic supernatant."This was probably a pH effect, since treatment ofinhibitors prepared by freezing and thawing cellsinto saline with 0.1 N NaOHor 0.06 N HCl alsodestroyed their inhibitory effect. However,0.00025 N HCl (as used in the pepsin digestiondescribed below) did not affect inhibition.

Since the citrate ion is a potent inhibitor ofDNase, it seemed possible that traces of this ionfrom the ACDsolution might have been responsi-ble for the inhibition of serum DNase by the "cyto-plasmic supernatant." The following observations,previously noted, militate against this possibility:

(a) Potent inhibitors were prepared fromheparinized blood, (b) 100°C destroyed the in-

194

SPECIFIC INHIBITOR OF DESOXYRIBONUCLEASEAND "L. E. CELL"

0coE 60

Z 400-

20

o

a e4xioiwec/mI,fro3 swthaud tnp(s-avs serunrO 'is n 0 1I serulnro t N) Ul} of " serurnTIA -go X io*wc/ml. fro3Wnr thcAed in saLuits sVSmwe !

0 Q15 0.25 0.5 IVOLUMEINHIBITOR IN ML.(Finalvl,9ml.)

FIG. 2. DEGREEOF INHIBITION OF SERUMDNASEVS. VOLUME(CONCENTRATION)OF INHIBITOR (19 ML. FINAL VOLUMEOF SUBSTRATE-INHIBITOR-SERUM MIXTURECONTAINING 1 ML. SERUM)

hibition, (c) 1 per cent citric acid, NaOH, andHCI destroyed inhibition, and (d) dialysis did notreduce inhibition. This possibility was furtherexcluded by a direct test:

To 20 ml. of a "whole chromosome" suspension, and ofa "cytoplasmic sediment" suspension, 5 ml. of "ACD"solution were added. After an hour at room temperature,the material was dialyzed against 0.14 M NaCl at 0-2Cfor 24 hours, with frequent changes of the bath. No in-hibition of human serum DNase was detected with eithermaterial before or after the treatment with "ACD."

D. Effects of enzymes"Cytoplasmic supernatant" inhibitors were in-

cubated for 2 hours at 37°C with an equal volumeof solutions containing 3 mg. per ml. of the follow-ing enzymes 4 in 0.01 MpH 7.5 trimethylol-amino-methane buffer (18): ribonuclease, trypsin, chymo-trypsin, papain, and yeast protease. Ribonucleasewas also used at 56°C for 2 hours. Pepsin wassimilarly used, but dissolved in water (final pH4.7) or in 0.0005 N HCG. Two ml. of each mix-ture was tested for inhibitory effect against humanserum DNase (final volume 19 ml. as in [12]).Controls consisted of the inhibitor incubated withthe solvent (i.e., without enzyme) and of the en-zyme incubated with saline (i.e., without inhibi-tor). The inhibitor controls showed unimpairedinhibitory action and the enzyme controls had noeffect on serum DNase activity except in the case

4 Crystalline enzymes obtained from General Biochemi-cals, Inc., Chagrin Falls, Ohio.

of yeast protease, which, alone, revealed DNaseactivity, and produced an additive DNase activitywith sera. All of the proteolytic enzymes com-pletely destroyed the inhibition of serum DNase,while ribonuclease was without effect upon theinhibitor.

E. Permeability of cell membrane to inhibitorSince WBCwere found to contain an inhibitor

for human serum DNase, it appeared possiblethat the very low serum DNase values commonlyencountered in normal sera (12) might resultfrom exposure of the serum to the WBCof theblood before separation, particularly if separationwas not carried out promptly. Freshly clottedblood was allowed to stand at room temperatureand at 37°C. Samples of sera were removed atintervals from 15 minutes to 6 hours after col-lection. Similar experiments were performedwith heparinized blood. No change in DNase ac-tivity was observed over this period. After 24hours at 37°C, there was moderate reduction ofDNase activity, but this was no greater than wasnoted when separated serum was allowed to standfor the same interval. The fact that no significantamount of inhibitor "leaked"' into autologous se-rum or plasma was also supported by the observa-tion that in no case was inactive serum found toinhibit the DNase of an active serum. However,0.14 Msaline, allowed to stand in contact with iso-lated WBCat 0 to 2°C for 24 to 48 hours, ex-

195

N. B. KURNICK, L. I. SCHWARTZ, S. PARISER, AND S. L. LEE

TABLE I

Inhibition of srum desoxyribonucease by varsous ceU types

Approx. No. cells X 100 per ml.No. of ceLlsper ml. of Lympho- Granulo- Mono- Blast Inhibition.

Diagnosis saline cytes cytes cytes cels per cent

1) Lymph. Leukemia 20 X 10' 14 6 0 0 322) Normal 35 X 10 1S 18 2 0 393) Monocytosis 170 X 106 17 68 8S 0 414) Normal 174 X 10 75 84 1S 0 665) Ac. Myel. Leukemia 240 X 10" 5 35 0 200 566) Chronic Myel. Leukemia 300 X 10 2 290 0 8 36

tracted a significant amount of inhibitor fromthem, as demonstrated by inhibition of serumDNase by the supernatant. Furthermore, 60 x10' WBCwashed four times with 0.14 M NaCl,then frozen and thawed into 1 ml. of the test se-rum, produced only slight inhibition of DNase,while unwashed WBC, similarly treated, producedcomplete inhibition.

-420

0 50 60 9 12D 15080 21 240 0 30 60 9O120 150 80aIO 240MINUTES MAINJUTES

FIG. 3. INHImON OF VAnous DESOXYRIONUcASESBY HUMANWBC"CYToPLAsMIc SUPENATANT"

Solid line represents activity of substrate-serum alone;dotted line represents activity of substrate-serum mix-ture plus 1 ml. "cytoplasmic supernatant" (final volume19 ml.). 1, 2. Dog serum, 3, 4. Rabbit serum, 5, 6. Guineapig serum, 7, 8. Cat serum, 9, 10. Rat serum, 11, 12. Ham-ster serum, 13,14. Human serum, 15,16. Bovine pancreaticDNase.

F. Inhibitor preparations from different cell typesThe presence of inhibitor in different types of

cells was studied by freezing and thawing in0.14 M saline, WBCisolated from a normal per-son, a patient with post-splenectomy monocytosis(40,000 WBCper cmm. capillary blood, 50%smonocytes), a patient with chronic myelogenousleukemia, one with lymphosarcoma and lymphaticleukemia (707% lymphocytes), and one with acutemyelogenous leukemia. The effect of 0.5 ml. ofthe supernatant from each was tested against 1 ml.of a single serum sample (Table I).

The data suggest that lymphocytes may be richerin inhibitor than granulocytes (cf. 1 and 6) andthat monocytes may be poor in inhibitor (cf. 3 and4). The relationship found between the inhibitionby 35 x 10' and by 174 x 10' normal WBCap-pears consistent with the data in Fig. 2. Imma-ture leucocytes appear to be richer in inhibitorthan mature WBC(cf. 5 and 6). Caution isnecessary in interpreting these observations, how-ever, since completeness of inhibitor extractionmay vary among the cell types.

G. Species specificity of inhibitorNine different inhibitors prepared from human

WBCwere tested against the sera of dog (diluted3: 4 with saline), cat (diluted 3:100), rabbit(diluted 1: 2), hamster (undiluted), guinea pig(diluted 3:4), rat (undiluted), and mouse (un-diluted). Dilutions were selected so that 1 ml.of the diluted serum approximated the DNase ac-tivity of active human serum (about 140 x 10-'units per ml. serum [see (3) for the definition ofDNase unit]). Bovine pancreatic DNase (19)'was similarly tested in a final concentration of0.005 Ig. per ml. The results of a typical experi-

5 Worthington Laboratories, Freehold, N. J.

196

SPECIFIC INHIBITOR OF DESOXYRIBONUCLEASEAND "L. E. CELL

TABLE IS

Enhanced susceptibility of washed eucocytes to "L.E. ceW' transformaion

Unwashed WBC Washed (4 X) WBCL.E. serum

dilution "L.E. "L.E.(initial) "Globs" cel" Rosettes Total "Globs" cels" Rosettes Total

Undil. 2 2 1 5 2 3 3 81:2 0 1 1 2 2 2 3 71:4 0 0 0 0 2 2 2 61:8 0 0 0 0 1 0 1 21:16 0 0 0 0 0 0 0 0

ment are shown in Fig. 3. As indicated by thecurves, only the human and rabbit DNase werecompletely inhibited. There was moderate inhibi-tion of guinea pig and hamster serum DNase,questionable inhibition of canine DNase, and noinhibition of rat, cat, and mouse serum and bovinepancreatic DNase.

II. "L.E. CELL" PHENOMENONINHIBITOR

A. Quantitation of "L.E. cell" phenomenonSera with "L.E. phenomenon activity" (L.E. serum)

were the same as those reported previously (5). Leuco-cyte (WBC) "buttons" of 10-15 X 10. WBCwere pre-pared as in that paper (5). These WBC"buttons" wereused as substrate for the L.E. phenomenon without fur-ther treatment or after washing four times by resuspensionin about 7 ml. isotonic saline and re-centrifugation. The"buttons" were suspended in 0.5 ml. saline (or inhibitorto be tested), 0.5 ml. of L.E. serum (diluted with normalserum) was added, and the mixture incubated at 37°C for1 hour. The WBCwere then sedimented by centrifuga-tion for 5 minutes at low speed, smeared upon glass slides,fixed in absolute methyl alcohol for 5 minutes, andstained in dilute Giemsa solution for 30 minutes. Thestained smears of a given experiment (with appropriatecontrols) were examined by a single observer who (toinsure objectivity) did not know the labelling code.Each of the three morphological manifestations of the

LE. phenomenon (a. classical "L.E. cell" with singlelarge inclusion, or "droplet cell" with several small in-clusions, b. rosettes, c. "globs" [5]) was graded from 0to 4 on the basis of frequency of occurrence.

B. Increased susceptibility of washed WBCto"L.E. cell" transformation

It was regularly found that washed WBCweremore susceptible to "L.E. cell" transformation thanwere unwashed WBC, as measured both by theintensity of the "L.E. cell" phenomenon at anygiven dilution of L.E. serum and by the titer towhich the L.E. serum would tolerate dilution be-fore loss of demonstrable L.E. activity (Table II).To exclude the possibilities (a) that the removalof the native plasma or the heparin and fibrinogenincluded therein and (b) that the traumatizationof the WBCby repeated centrifugation had causedtheir greater susceptibility to the "L.E. cell" trans-formation, two controls were added. In one, thewashed WBCwere resuspended in their originalplasma, re-centrifuged, and tested. No diminu-tion in the enhancement of the "L.E. cell" phe-nomenon produced by washing was observed.In the second control, unwashed WBCwere re-suspended in the original plasma and re-centrifuged

TABLE III

Inhibon of "L.E. ceU" phenomenon by kucocyte homogenat*

Control (normal plasma) Inhibited (WBChomogenate)"L.E. serum

dilution "L.E. "L.E.(initial) "Globs" cells" Rosetts Total "Globs" cedls" Rosettes Total

Undil. 2 3 3 8 2 1 1 41:2 1 2 3 6 1 1 1 31:4 1 2 1 4 1 1 1 31:8 1 1 1 3 0 0 0 01:16 1 1 0 2 0 0 0 01:32 1 0 0 1 0 0 0 01:64 0 0 0 0 0 0 0 0

Washed WBC(15 X 106) plus 0.5 ml. normal plasma or WBChomogenate plus 0.5 ml. L.E. serum (diluted withnormal serum).

** WBChomogenate consists of 40 X 10 WBCper ml. frozen and thawed in normal plasma.

197

N. B. -KURNICK, L. I. SCHWARTZ,S. PARISER, AND S. L. LEE

'and resuspended four times. Such unwashed, butrepeatedly resuspended and re-centrifuged WBCrevealed no increased susceptibility to "L.E. cell"transformation. Controls in which normal sera,with and without DNase activity, were substitutedfor active L.E. sera revealed no "L.E. cell" induc-tion in either washed or unwashed WBC.

C. "L.E. cell" phenomenon inhibitionThe probability that the increased susceptibility

of washed WBCto "L.E. cell" transformation wasdue to the partial removal from the WBCof in-hibitor was confirmed by replacing the inhibitor.

8

7

6

4

3

\.---\ ---GUI; ----.

\I

'N~

v-.ashed oaLl tLreanwashQd ceLl tiirev,ashodce11 tf'nhibibtThLan,shed call +*inhibitortitim

1:2 1:4 1:8 1:16 1:3. 164 1:128 1:256 1:512F1NALvi tLUTxC*Jor LE SERCUM

FIG. 4. EFFECT OF SUPERNATANTFROM24 x 10' WBCnL. FROZEN AND THAWEDIN NORMALSERUMON

INTENSITY OF "L.E. CELL" PHENOMENONAT SEVERALDILUTIONS OF L.E. SERUM(IN NORMALSERUM)

The washed WBCwere resuspended in a WBChomogenate or a WBCfraction (see Part I, A fordescription of fractions). The whole homogenates,and each of the fractions which contained DNaseinhibitor ("cytoplasmic fraction," "cytoplasmicsupernatant"), but not those without DNase in-hibitor activity (chromosomal fraction, cytoplas-mic sediment, DNH, DNA, histone, "residualchromosomes"), reduced the intensity of the "L.E.cell" phenomenon in washed WBC(Table III).Slight inhibition of "L.E. cell" transformation ofunwashed WBCwas also demonstrated (Fig. 4).

In order to test the relationship between inhi-bition of the "L.E. cell" phenomenon and the con-

- a contro[- - inhibitor 1:8

6 ' V4

undlkuLed

-3~ ~ ~ ~~N

156 l12 124 IAIBLE. SEWI DIXUlmON (Fial)

FIG. 5. EFcEr OF CONCENTRATIONOF INHIBrIOR ONTHE INTENSITY OF THE "L.E. CELL" PHENOMENONPRO-DUCEDIN WASHEDWBCBY SEVERAL DILUTIONS OF L.E.SERUM

centration of the inhibitor, an experiment was per-formed in which the inhibitor, a homogenate of24 x 106 WBC per ml. plasma (frozen andthawed), was serially diluted with normal plasma.The control contained normal plasma in place ofthe inhibitor. The data are presented in Fig. 5,and the percentage inhibition in Fig. 6. The simi-larity between Fig. 6 and Fig. 2 is apparent. Itshould be borne in mind that the percentage inhibi-tions shown in Fig. 6 represent the ratios of thetotal grading units of the inhibited and uninhibitedexperiments. This is only a very rough estimateof the degree of inhibition, therefore. Since the

l00O

80

o 60

1 40

20

o

a

0

* 1:4 LE: serrn1:8 " of

A 1:16 " .,D 1: X .

0 .1 .2 .3 .4 .5\ULUME ilRlUBITOP. (in frbl VoL 1.5 ml.)

FIG. 6. DEGRmOF INHIBMON OF "L.E. CELI' PHE-NOMENONVS. VOLUME(CONCENTRATION) OF INHIBITOR(FROM DATA IN FIG. 5)

Z

0

ID

-i

1-

198

I

SPECIFIC INHIBITOR OF DESOXYRIBONUCLEASEAND "L. E. CELL 19

final volume is approximately 1.5 ml. in the "L.E.cell" experiments, whereas it is 19 ml. in theDNase experiments, the greater sensitivity of theDNase system to the inhibitor is obvious.

D. Physico-chemical properties of "L.E. cell"phenomenon inhibitor

Like the DNase inhibitor, the "L.E. cell" phe-nomenon inhibitor was found to be unaffected bya temperature of 56°C for 30 minutes, but wasdestroyed at 100°C for 15 minutes. Similarly,dialysis did not impair the inhibition of the "L.E.cell" phenomenon. After repeated use over a pe-riod of weeks, which required bringing the materialto room temperature for short intervals and re-freezing, inhibition of both the serum DNase andof the "L.E. cell" phenomenon became progres-sively less effective. This was presumed to be

TABLE IV

Effect of anions on "L.E. ceU" phenomenon

Result*Concen-

Salt tration Undil.* 1:4 1:16 1:64

Sodium Chloride 0.14M 10 6 3 1Sodium Arsenate 0.2 M 10 4 0 0Potassium Oxalate 0.2 M 6 1 0 0Sodium Citrate 0.2 M 5 2 1 0

* Total grading score. Maximum is 12 (4 each for "L.E.cells," rosettes, "globs")

** Dilutions indicated are those of the L.E. serum(diluted with normal serum)

due to proteolysis, since, under the conditions de-scribed in Part I, D, the "L.E. cell" inhibitor wasalso destroyed by proteolytic enzymes.

E. Effect of magnesium ion binding anions on"L.E. cell" phenomenon

Since intracellular DNase is thought to playa part in the production of the "L.E. cell" phe-nomenon, anionic DNase inhibitors were tested.These ions are thought to act by binding Mg++,which is required for optimal DNase activity. TheWBCbuttons were suspended in 0.5 ml. of theelectrolyte solution, followed by 0.5 ml. L.E. se-rum. Thus, the final concentration of the salt wasapproximately one-third that indicated in TableIV, which presents the observed results. Definiteinhibition by citrate, oxalate, and arsenate, in orderof decreasing effect was apparent.

DISCUSSION

The experimental data demonstrate the presencein human leucocytes of an inhibitor for humanserum DNase, and an inhibitor of the "L.E. cell"phenomenon.

These inhibitors do not permeate a cellophanemembrane, are destroyed by heating to 100°C butare stable at 56°C, are stable at - 20°C for sev-eral months and at 0-2oC for at least several days.They are destroyed by a number of proteolytic en-zymes, but not nucleolytic enzymes. Extremesof pH also destroy the DNase inhibitor. Effortsat purification and further physico-chemical char-acterization of the factor or factors are in progress.

The DNase inhibitor reveals species specificityto some extent (only in the rabbit is the same de-gree of inhibition obtained as in the human). Itmay be suggested that the specificity is due to therequirement for a hypothetical serum co-factor.However, inhibitors prepared by freezing andthawing WBCin their own plasma, which wouldcontain the hypothetical co-factor, reveal the samespecificity as do saline extracts. Thus, the speci-ficity is probably a function of differences in DNasefrom various species. Immunologic differences inDNases have already been demonstrated by Mc-Carty (20).

That the inhibitors are present in the leucocytesis apparent from the method of preparation. Theonly other elements regularly present are throm-bocytes. However, potent inhibitors were obtainedfrom the WBCof a patient with severe thrombo-cytopenia. The inhibitors are found in the super-natant upon rupturing the cells in saline. Al-though this fraction is ordinarily regarded as cyto-plasmic, it undoubtedly includes some proteins ofnuclear origin. The cytological localization of theinhibitors is, therefore, unsettled.

Desoxyribonuclease inhibition by cell extractshas also been reported by others. The inhibitorof yeast DNase which Zamenhof and Chargaff(13), obtained from yeast, resembles the inhibitorin WBCdescribed in this paper in its species speci-ficity (for yeast), in its instantaneous effect uponthe enzyme activity, and in its protein nature, butis less stable (destroyed at 56°C in 5 minutes).Dabrowska, Cooper and Laskowski (21) andCooper, Trautmann and Laskowski (22) pre-pared an inhibitor of bovine DNase from several

199

N. B. KURNICK,- L. I. SCHWARTZ, S. PARISER, AND S. L. LEE

avian and mammalian organs. These inhibitorswere also proteins and were very unstable. Thisinstability interferes with the separation of excessDNase activity from the inhibitor content of cellextracts since both DNase activity and inhibitorare destroyed at 56°C. Thus, these inhibitorsdiffer from ours both in lack of species specificityand in stability.

Henstell and Freedman (23, 24) have recentlyreported on the preparation of an inhibitor of bo-vine DNase from leucocytes by the dilution ofwhole blood 1: 100 or 1: 1000 with water. Theirinhibitor appears to be very similar to that of Las-kowski's group in its non-specificity and instability.It differs from that reported by us, because it ispotent in very great dilution and is less, ratherthan more, abundant in immature cells. It is ap-parent that we failed to observe the inhibitor re-ported by these authors because of its instability,since several hours were required for the prepara-tion of the inhibitor in leucocytes, and it was nottested on the day of preparation.

Bernheimer and Ruffier (25) isolated a specificDNase inhibitor from streptococci. This inhibitor,however, differs from that described here in be-ing destroyed by ribonuclease but not by proteoly-tic enzymes, thus indicating its non-protein na-ture. It appears, then, that we have observed anew inhibitor from mammalian tissues, most closelyresembling that described by Zamenhof and Char-gaff (13) in yeast.

Close parallelism between the inhibitor of serumDNase described by us and the inhibitor of the"L.E. cell" phenomenon suggest their identity.The two activities are located in the same WBCfraction. They show the same heat stability andare non-dialyzable. Complete correspondence isnot observed in the intensity of the effect, in thatmore nearly complete inhibition of serum DNasethan of the "L.E. cell" phenomenon is usually ob-tained. This, however, may not indicate lack ofidentity of the inhibitors, since the intracellularDNase is probably less accessible to inhibitoradded! in vitro, and the intra- and extracellularDNases may not be identical in susceptibility tothe inhibitor. Proof of the identity of the inhibi-tors must await their further purification.

Since the evidence suggests the identity of theintracellular DNase inhibitor and the inhibitor ofthe "L.E. cell" phenomenon, it is probable that

the depolymerization of DNAwhich characterizesthe "L.E. cell" is secondary to a derangement ofthe intracellular DNase-DNase inhibitor system.This abnormality is not primary in the cell, sinceit can be induced in leucocytes from a normal in-dividual (26) by serum from a patient with sys-temic lupus erythematosus. Wesuggest that theL.E. plasma factor (27), which itself has noDNase activity (5), causes the release of the in-tracellular DNase from inhibition. The inhibitionof the "L.E. cell" phenomenon by anionic DNaseinhibitors, as well as by an extract which containsa naturally occurring DNase inhibitor, supportsthis hypothesis.

This hypothesis led to the development of a"sensitized" "L.E. cell" test, using WBC, whichhad been washed to remove inhibitor. False posi-tive tests are not produced, perhaps because un-broken WBCcontain some inhibitor even afterrepeated washing. This was demonstrated for theDNase inhibitor. Thus, the L.E. serum factor isstill required to elicit the phenomenon.

The fact that mature granulocytes show greatersusceptibility to the "L.E. cell" transformationthan lymphocytes (28) may be due to the apparentgreater inhibitor content of the latter noted in ourstudies (Table I). Likewise, the lesser suscepti-bility of immature granulocytes may be attributedto their greater DNase inhibitor content as com-pared with mature leucocytes.

Observations on the DNase inhibitor content ofvarious tissues, normal and pathological, and atvarious stages of development are in progress.Some observations on lymphocytes, blasts, andgranulocytes have already been referred to above.The possible role of a specific inhibitor of DNasein cellular division, growth, senescence, and thedevelopment of tumors, which suggests itself, isunder further investigation.

SUMMARY

1. An inhibitor for human serum DNase hasbeen observed in human leucocytes. It is a pro-tein, soluble in saline, stable at 560C. Testedagainst several species, it is equally effective onlyagainst rabbit serum DNase and has little or noinfluence on some others. In its stability andspecificity it differs from other DNase inhibitorspreviously reported.

200

SPECIFIC INHIBITOR OF DESOXYRIBONUCLEASEAND "L. E. CELL 201

2. An inhibitor for the "L.E. cell" phenomenonhas been demonstrated in leucocytes. Evidencewhich suggests its identity with the DNase in-hibitor is presented.

3. A sensitive "L.E. cell" test using washedWBCis described.

4. A mechanism for the "L.E. cell" phenomenonis discussed.

ACKNOWLEDGMENTS

The technical assistance of Mrs. Mary Eason is grate-fully acknowledged.

The encouragement of Dr. Paul Klemperer throughoutthese studies is deeply appreciated.

REFERENCES

1. Lee, S. L., Michael, S. R., and Vural, I. L., The L.E.(Lupus erythematosus) cell. Clinical and chemi-cal studies. Am. J. Med., 195.1, 10, 446.

2. Hargraves, M. M., Richmond, H., and Morton, R.,Presentation of two bone marrow elements: the"tart" cell and the "L.E." cell. Proc. Staff Meet.,Mayo Clin., 1948, 23, 25.

3. Kurnick, N. B., The determination of desoxyribonu-clease activity by methyl green; Application to se-rum. Arch. Biochem., 1950, 29, 41.

4. Klemperer, P., Gueft, B., Lee, S. L., Leuchtenberger,C., and Pollister, A. W., Cytochemical changes ofacute lupus erythematosus. Arch. Path., 1950, 49,503.

5. Kurnick, N. B., Pariser, S., Schwartz, L. J., Lee, S. L.,and Irvine, W., Studies on desoxyribonuclease insystemic lupus erythematosus: Non-participationof serum desoxyribonuclease in the L.E. phe-nomenon. J. Clin. Invest., 1952, 31, 1036.

6. Kurnick, N. B., discussion of paper by Sparrow, A. H.,Moses, M. J., and Dubow, R. J., Relationships be-tween ionizing radiation, chromosome breakageand certain other nuclear disturbances. Exptl. CellRes., 1952, 2, 266.

7. Kurnick, N. B., Schwartz, L. J., Pariser, S., Lee, S.,and Irvine, W., The role of desoxyribonuclease anda nuclease inhibitor from leucocytes in the lupuserythematosus cell phenomenon. J. Clin. Invest.,1952, 31, 645.

8. Sorof, S., and Cohen, P. P., Modified semi-micro-and micro-Waring blendors for low temperatureuse. Exptl. Cell Res., 1951, 2, 299.

9. Mirsky, A. E., and Ris, H., Isolated chromosomes.J. Gen. Physiol., 1947, 31, 1.

10. Mirsky, A. E., and Ris, H., The chemical compositionof isolated chromosomes. J. Gen. Physiol., 1947,31, 7.

11. Mirsky, A. E., and Pollister, A. W., Chromosin, adesoxyribose nucleoprotein complex of the cell nu-cleus. J. Gen. Physiol., 1946, 30, 117.

12. Kurnick, N. B., Desoxyribonuclease activity of seraof man and some other species. Arch. Biochem.and Biophys., in press.

13. Zamenhof, S., and Chargaff, E., Studies on the desoxy-pentose-nuclease of yeast and its specific cellularregulation. J. Biol. Chem., 1949, 180, 727.

14. Catcheside, D. G., and Holmes, B., The action of en-zymes on chromosomes. Symp. Soc. Exptl. Biol.,Cambridge, 1947, 1, Nucleic Acid, 225.

15. Maver, M. E., and Greco, A. E., The nuclease activi-ties of cathepsin preparations from calf spleen andthymus. J. Biol. Chem., 1949, 181, 861.

16. Brown, K. D., Jacobs, G., and Laskowski, M., Thedistribution of nucleodepolymerases in calf thy-mus fractions. J. Biol. Chem., 1952, 194, 445.

17. Webb, M., Use of desoxyribonuclease inhibitors in theisolation of desoxyribonucleic acids. Nature, 1952,169, 417.

18. Gomori, G., Buffers in the range of pH 6.5 to 9.6.Proc. Soc. Exper. Biol. & Med., 1946, 62, 33.

19. Kunitz, M., Isolation of crystalline desoxyribonu-clease from beef pancreas. Science, 1948, 108, 19.

20. McCarty, M., Purification and properties of desoxyri-bonuclease isolated from beef pancreas. J. Gen.Physiol., 1946, 29, 123.

21. Dabrowska, W., Cooper, E. J., and Laskowski, M.,A specific inhibitor for desoxyribonuclease. J. Biol.Chem., 1949, 177, 991.

22. Cooper, E. J., Trautmann, M. L., and Laskowski, M.,Occurrence and distribution of an inhibitor for de-soxyribonuclease in animal tissues. Proc. Soc.Exper. Biol. & Med., 1950, 73, 219.

23. Henstell, H. H., and Freedman, R. I., The viscosi-metric determination of desoxyribonuclease inhibi-tion. Cancer Res., 1952, 12, 341.

24. Henstell, H. H., Freedman, R. I., and Ginsburg, B.,An inhibitor of desoxyribonuclease in human whiteblood cells and bone marrow cells and its relation-ship to cellular maturity. Cancer Res., 1952, 12,346.

25. Bernheimer, A. W., and Ruffier, N. K., Elaborationof desoxyribonuclease by streptococci in the rest-ing state and inhibition of the enzyme by a sub-stance extractable from the cocci. J. Exp. Med.,1951, 93, 399.

26. Berman, L., Axelrod, A. R., Goodman, H. L., andMcClaughry, R. I., So-called "lupus erythematosusinclusion phenomenon" of bone marrow and blood.Morphologic and serologic studies. Am. J. Chin.Path., 1950, 20, 403.

27. Haserick, J. R., Lewis, L. A., and Bortz, D. W.,Blood factor in acute disseminated lupus erythemato-sus; I. Determination of gammaglobulin as specificplasma fraction. Am. J. M. Sc., 1950, 219, 660.

28. Rohn, R. J., and Bond, W. H., Some supravital ob-servations on the "L.E." phenomenon. Am. J.Med., 1952, 12, 422.

201