Erythrocyte Catabolism by Macrophages InVitro THE EFFECT OF HYDROCORTISONE ONERYTHROPHAGOCYTOSIS AND ON THEINDUCTION OF HEME OXYGENASE

Diethard Gemsa, … , H. Hugh Fudenberg, Rudi Schmid

J Clin Invest. 1973;52(4):812-822. https://doi.org/10.1172/JCI107245.

Phagocytosis of erythrocytes was studied in vitro in an incubation system consisting of ratperitoneal macrophages and antibody-coated 59Fe-labeled erythrocytes. The system wascharacterized in terms of the rate and magnitude of erythrophagocytosis, determined by theinteriorization of the 59Fe label. On incubation of 150 × 106 macrophages with 75 × 106

antibodycoated erythrocytes, erythrophagocytosis began within a few minutes and wasessentially completed after 2 h when 50% of the offered red cells had been ingested by themacrophages. Heme oxygenase (HO) activity, which is very low in native macrophages,increased 4- to 10- fold in response to the ingested erythrocytes; this enzyme stimulationoccurred with a delay of 3 h in relation to erythrophagocytosis. Actinomycin D or puromycinprevented the increase of HO activity without affecting erythrophagocytosis, which suggeststhat the enzyme stimulation was due to substrate-mediated enzyme induction.

Hydrocortisone (HC) (0.1 mg/ml medium) dissociated erythrophagocytosis from HOinduction, leaving the former unimpaired but completely suppressing the latter. Thesuppressive effect of HC on the enzyme induction was completely prevented by 5 mgglucose and 0.02 U insulin/ml of the medium. In macrophages engaged inerythrophagocytosis. HC also lowered glucose removal from the medium and reducedformation of 14CO2 from [1-14C]glucose.

These results suggest that induction of HO in macrophages by the hemoglobin of ingestederythrocytes requires intact […]

Research Article

Find the latest version:

http://jci.me/107245-pdf

Erythrocyte Catabolism by Macrophages In Vitro

THE EFFECTOF HYDROCORTISONEONERYTHROPHAGOCYTOSIS

ANDONTHE INDUCTION OF HEMEOXYGENASE

DIETHARDGEMSA,C. H. Woo, H. HUGHFUDENBERG,and RUDI SCHMm

From the Department of Medicine, University of California, San Francisco

Medical Center, San Francisco, California 94122

A B S T R A C T Phagocytosis of erythrocytes was studiedin vitro in an incubation system consisting of rat peri-toneal macrophages and antibody-coated 'Fe-labelederythrocytes. The system was characterized in terms ofthe rate and magnitude of erythrophagocytosis, deter-mined by the interiorization of the 'Fe label. On incu-bation of 150 X 106 macrophages with 75 X 10 antibody-coated erythrocytes, erythrophagocytosis began within afew minutes and was essentially completed after 2 hwhen 50% of the offered red cells had been ingested bythe macrophages. Hemeoxygenase (HO) activity, whichis very low in native macrophages, increased 4- to 10-fold in response to the ingested erythrocytes; this en-zyme stimulation occurred with a delay of 3 h in relationto erythrophagocytosis. Actinomycin D or puromycinprevented the increase of HOactivity without affectingerythrophagocytosis, which suggests that the enzymestimulation was due to substrate-mediated enzymeinduction.

Hydrocortisone (HC) (0.1 mg/ml medium) dissoci-ated erythrophagocytosis from HO induction, leavingthe former unimpaired but completely suppressing thelatter. The suppressive effect of HC on the enzyme in-duction was completely prevented by 5 mg glucose and0.02 U insulin/ml of the medium. In macrophages en-gaged in erythrophagocytosis, HC also lowered glucoseremoval from the medium and reduced formation of"4CO2 from [1-"C]glucose.

These results suggest that induction of HO in macro-

phages by the hemoglobin of ingested erythrocytes re-

quires intact transport or metabolism of glucose. Glucose

This work was presented in part at the Annual Meetingof the Association of American Physicians on 3 May 1972in Atlantic City, N. J.

Received for publication 6 October 1972 and in revisedform 22 November 1972.

utilization appears to be impaired by HC, but is re-stored by additional glucose and insulin. The findingssuggest that plasma steroid concentrations in the phar-macological range could reduce bilirubin formation inphagocytic cells in vivo without affecting the sequestra-tion and degradation of erythrocytes. This provides apossible explanation for the observation that in patientswith hepatogenous jaundice, steroids often lower theserum bilirubin level.

INTRODUCTIONCirculating erythrocytes which have reached the endof their physiological life-span are believed to be re-moved and subsequently degraded in mononuclear phago-cytes of the spleen, liver, bone marrow, and other tis-sues (1). Erythrocytes which have been damaged byphysical, chemical, or immunological means appear to behandled in a similar manner (2-4) although the relativeimportance of various tissue sites in the removal of thesedamaged cells may differ from that of physiologicallyaged erythrocytes (2, 5). Red cells extruded into extra-

vascular sites are phagocytized and degraded by mono-

nuclear macrophages which converge at these sites(6, 7). Details of the mechanisms by which these phago-cytic cells capture and sequester erythrocytes are un-

clear, and particularly the mechanism whereby mono-

nuclear phagocytes of the spleen, liver, and bone marrow

recognize senescent erythrocytes and remove them fromthe circulation remains obscure. Events involved in theintracellular degradation of ingested erythrocytes are

somewhat better understood. Preformed lysosomes havebeen shown to converge around, and eventually to fuse

with, the phagocytic vesicles containing the red cells(8, 9). During this digestive process, lysosomal hydro-lases increase (10), proteolytic products are released

812 The Journal of Clinical Investigation Volume 52 April 1973

into the surrounding extracellular space (11), and iron,presumably derived from the degraded hemoglobin, ac-cumulates within the phagocytes (12). In addition, hemeoxygenase (HO),' the enzyme system that converts hemeto bilirubin IXa (13, 14), is stimulated by hemoproteinsingested by the macrophages ( 15).

A more detailed characterization of the mechanismof the degradation of phagocytized erythrocytes and par-ticularly of the conversion of their hemoglobin-hemeto bilirubin required the development of a reproduciblein vitro system in which the individual steps of thedegradative process could be analyzed sequentially. Inthe present study, antibody-coated 'Fe-labeled rat eryth-rocytes were offered to a specified number of homolo-gous macrophages maintained in primary culture. Therate of erythrophagocytosis was measured by the inter-iorization of the 'Fe label and the rate of HO stimula-tion was determined by serial enzyme assay. By usinginhibitors of protein synthesis, steroids, or nutritionalmanipulations, it was possible to dissociate the twoevents. The findings obtained with these macrophagecultures permitted evaluation of the quantitative andtemporal relationship between the phagocytosis of eryth-rocytes and the stimulation of the principal enzymaticmechanism involved in the degradation of the ingestedhemoglobin-heme.

METHODSPreparation of peritoneal nacrophages. In all experi-

ments, female Sprague-Dawley rats weighing 175-200 gwere used. The animals were fed standard laboratory chow.Peritoneal macrophages were obtained by stimulating theperitoneal cavity with a single intraperitoneal injection of20 ml of 1.2%o (wt/vol) sodium caseinate suspended insterile isotonic saline. 3 days later the rats were sacrificedand the peritoneal cavity was lavaged with cold isotonicsaline. All peritoneal cell preparations which appearedgrossly hemorrhagic or which by phase microscopy con-tained more than 15%o erythrocytes were discarded. Har-vested peritoneal cells from 10-15 rats were pooled andcentrifuged in a refrigerated centrifuge (Ivan Sorvall, Inc.,Norwalk, Conn.) for 5 min at 100 g. The pellet was sus-pended in 4 ml cold isotonic saline, contaminating erythro-cytes were lysed by brief addition of 12 ml cold distilledwater, and isotonicity was restored within 10 s with 4 mlof a 3.5%, saline. The cell preparation was again centrifugedand the supernate containing the dissolved hemoglobin wasdiscarded. This procedure did not affect the viability of themacrophages as determined by trypan blue exclusion. Thecells were then suspended in medium 199 (Grand IslandBiological Co., Grand Island, N. Y.), and total cell countswere performed in a Neubauer hemacytometer. An averageof 8 X 10' peritoneal cells were obtained from each rat.Differential counts were performed after staining withMay-Gruenwald-Giemsa dye.

The percentage of mononuclear macrophages ranged from85-90% of all peritoneal cells, the remainder being granulo-

'Abbreviations used i'i this paper: HO, heme oxygenase;HC, hydrocortisone.

cytes, lymphocytes, and occasional mast cells. Phagocyticcells were identified by incubation of representative cellsamples with suspensions of carbon particles measuring 450nm or more in diameter (Higgins India Ink, A. W. Faber-Castell Pencil Co., Inc., Brooklyn, N. Y.). By light micros-copy 90-95%o of the peritoneal cells contained carbon par-ticles; most of these cells were mononuclear macrophagesand a few granulocytes not exceeding 5% of the total cellpopulation.

Preparation of erythrocytes and dissolved hemoglobin.In adult Sprague-Dawley rats, 1.5 ml blood was removed bycardiac puncture and 0.12 mCi 59FeClh (specific activity, 48mCi/mg Fe) was administered intravenously through a tailvein. 3-20 days later, the rats were bled by cardiac punctureand the erythrocytes were washed three times in phosphate-buffered saline, pH 7.4. The 5'Fe-labeled erythrocytes werecounted in a Neubauer hemacytometer and their radio-activity was measured in a Nuclear-Chicago well-type scin-tillation counter (Nuclear-Chicago Corp., Des Plaines, Ill.).Dissolved 'Fe-labeled hemoglobin was prepared from theseerythrocytes by lysis in distilled water, followed by removalof stroma and membranes by centrifugation at 10,000 g.Preliminary experiments ascertained that under these con-ditions, as well as during incubation, hemoglobin did notcrystallize.

Antiserum to rat erythrocytes was prepared in rabbits asfollows. After siphoning off the buffy coat and washing theerythrocytes three times with phosphate-buffered saline,pH 7.4, 1 ml of packed rat erythrocytes was injected intra-peritoneally into six New Zealand rabbits at biweekly in-tervals. After 10 wk the rabbits were bled, the sera pooled,and decomplemented by heating for 1 h at 560C. The hemag-glutination titer of the antiserum was 1: 4096 as determinedby conventional methods (16). The 'Fe-labeled rat erythro-cytes were coated with the antibody by incubating 3 X 109cells with 0.12 ml of the antiserum for 30 min at roomtemperature. Preliminary experiments had indicated thatthis amount of antiserum was optimal for producing a maxi-mal degree of erythrophagocytosis without causing erythro-cyte agglutination. The antibody-coated erythrocytes werewashed twice with buffered saline.

Incubation of mnacrophages with erythrocytes or dissolvedhemoglobin. Unless stated otherwise, standard incubationsconsisted of 150 X 106 macrophages and 75 X 10 antibody-coated 'Fe-labeled erythrocytes or 10.5 mg dissolved '9Fe-labeled hemoglobin (which is equivalent to the hemoglobinin approximately 600 X 10' rat erythrocytes). The cells wereincubated in 50-ml Erlenmeyer flasks (Nalgene) containing15 ml medium 199 and 0.125 mg penicillin/ml. Incubationusually lasted 5 h at 37°C with the rotating water bath setat 70 cycles/min. The flasks were gassed with air containing5% C02 and covered with Parafilm. Under these conditions,the pH of the medium was maintained at 7.3-7.45 for theduration of the incubation which was terminated by im-mersing the flasks in ice water. Hemolysis during the 5 hof incubation never exceeded 2%o of the incubated erythro-cytes.

In some instances, glucose, fructose, or maltose (5 mg/ml, final concentration) was added to the incubation systemin addition to the medium 199 which contained 1 mg glu-cose/ml. In other instances, 0.02 U insulin (E. R. Squibb& Sons, Princeton, N. J.), 0.2 jug actinomycin D (SigmaChemical Co., St. Louis, Mo.), 2 ,ug puromycin (ICN Nu-tritional Biochemicals Div., Cleveland, Ohio), or 0.01-0.18mg hydrocortisone succinate (Sigma Chemical Co.) dis-solved in medium 199 was added per ml of the incubationmixture.

Erythrocyte Catabolism by Macrophages 813

2 30 f 04k

~~~~~~HEOXYGENASE30 -

10~~~~~~~~~~

0.212

10 O E

0 1 2 3 4 5 6 7

Hours of kIcubation

FIGURE 1 Erythrophagocytosis and stimulation of HO inmacrophages. In a standard incubation system containing150 X 106 macrophates and 75 X 100 antibody-coated erythro-cytes, erythrophagocytosis (closed circles) and HO activity(open squares) was determined at the indicated times afterthe start of the incubation.

Quantitation of endocytosis of erythrocytes or hemoglobin.At the end of the incubation period the macrophages, mostof which adhered to the bottom of the incubation flask,were carefully detached with a fine brush and centrifugedin the incubation medium at 3,000 g for 6 min. Erythrocytesthat had not been phagocytized were lysed by brief ex-posure to hypotonicity followed after 20-30 s by restorationof isotonicity with 3.5% saline. Macrophages which hadbeen incubated with dissolved hemoglobin instead of anti-body-coated erythrocytes were treated in the same manner.After centrifugation the macrophage pellet was suspendedin 2 ml cold 0.1 M K-phosphate buffer, pH 7.4, and the59Fe activity retained in the cells was determined in awell-type scintillation counter. All samples were countedat least 5 min which assured a standard error of the meanof less than 2%.

Assay of heme oxygenase (HO) activity. After comple-tion of the incubation the macrophages were disrupted bysonication for 10 s at 45 watts (Heat Systems-Ultrasonics,Inc., Plainview, N. Y.). The sonication disrupted morethan 98% of the cells as estimated microscopically in ahemacytometer. The homogenate was centrifuged at 20,000g for 10 min at 4°C and the supernate used as the enzymesource. Details of the assay for HO activity have been de-scribed elsewhere (13, 14). The standard assay was modifiedby reducing all constituents proportionally to a final volumeof 0.5 ml per cuvette, containing 0.7-1.3 mg protein/ml.

Analytical procedures. The conversion of [1-"C]glucoseto "CO2 by phagocytizing macrophages was determined asfollows. 2 gCi [1-"C]glucose (Tracerlab Div., LFE Elec-tronics, Boston, Mass., specific activity 3.0 mCi/nmol) wasadded to the standard incubations containing the macro-phages and antibody-coated erythrocytes. The flasks weregassed with air containing 5%, CO2 and closed with sleeve-type rubber stoppers equipped with a center well. At theend of the incubation, 0.4 ml 8 N H2SO4 was injected intothe medium through the rubber stopper and the center wellwas filled with 0.3 ml hyamine hydroxide. After furtherincubation for 45 min, the center well was removed andthe "CO2 trapped in the hyamine hydroxide was countedin a Packard liquid scintillation spectrometer (Packard In-strument Co., Inc., Downers Grove, Ill.) (17).

Glucose concentration in the incubation medium was de-termined by the glucose oxidase method (18). The proteinconcentration of the 20,000 g supernate of the homogenizedmacrophages was measured by the method of Lowry, Rose-brough, Farr, and Randall (19). Hemoglobin content oferythrocytes was estimated by the cyanmethemoglobinmethod (20) and the hemoglobin in the 20,000 g supernateby the method of Crosby and Furth (21).

Expression of results. In all instances, at least five in-dividual experiments were carried out under identical con-ditions with macrophage pools from 10 to 15 rats. Duplicateanalyses of individual experiments were possible only in afew instances because of the limited amount of macrophagesavailable, and because of the time required for the ana-lytical procedures, all of which were performed immediatelyafter completion of the incubation. The results are pre-sented either as the mean and range of all individual ex-periments or as individual data from a representative setof experiments.

Endocytosis of antibody-coated erythrocytes or dissolvedhemoglobin by the incubated macrophages was calculatedfrom their intracellular 6Fe activity and expressed as a per-centage of the total 'Fe activity of the erythrocytes orhemoglobin that had been offered to the macrophages. Thepercentage value obtained was converted to absolute num-bers of erythrocytes ingested by the macrophages. Glucoseoxidation in macrophages was measured by comparing theradioactivity of the "CO2 trapped in the hyamine hydroxidewith the total [1-"C]glucose added to the flask. The valuesobtained were converted to milligrams of glucose metabo-lized to C02.

RESULTS

Preliminary studies to define optimal conditions of in-cubation. When 150 X 106 peritoneal macrophages wereincubated with 75 X 10' antibody-coated erythrocytes,

sx

.!20u0

CL0

-EuLJ

E

0

cE

2

0

E

Erythrocytes Offered (x106)

FIGURE 2 Erythrophagocytosis (closed circles) and stimu-lation of HO (open squares) in 150 X 10' macrophages ex-posed to various amounts of antibody-coated erythrocytes.In this and all subsequent experiments, erythrophagocytosisand stimulation of HO was determined after 5 h of incu-bation.

814 D. Gemsa, C. H. Woo, H. H. Fudenberg, and R. Schmid

erythrophagocytosis began within the first few minutesof incubation and reached a maximum within the first2 h (Fig. 1). Thereafter, little erythrophagocytic ac-tivity was detectable despite the presence of additionalerythrocytes in the incubation medium. When a seconddose of 75 X 106 antibody-coated erythrocytes was addedto the incubation mixture after 2 h, erythrophagocytosisresumed to an extent comparable with the initial phago-cytic rate. On phase microscopy, phagocytized erythro-cytes appeared first as phase-dense rounded structures,which on further incuabtion gradually lost their density,until after 5-7 h only their remnants were detectable insmall intracellular vacuoles. HO activity in native mac-rophages or in macrophages incubated for 5 h withouterythrocytes ranged from 0.04 to 0.125 nmol bilirubinformed/10 mg protein per min. On incubation with 75 X106 erythrocytes, the enzyme activity increased steeplyafter an initial lag phase of approximately 3 h andreached a maximum after 5 h of incubation (Fig. 1).Maximal HO activity was regularly 4-10 times higherthan base activity. Since incubation beyond 5 h did notenhance HOactivity, 5 h of incubation was selected forall subsequent experiments.

When variable amounts of antibody-coated erythro-cytes were incubated with the standard number of 150 X106 macrophages, a relatively constant fraction of the redcells was phagocytized by the macrophages (Fig. 2).With erythrocyte doses ranging from 15 to 120 X 106cells, approximately 50% of the offered red cells wereingested, while with larger doses the percentage of in-teriorized cells dropped to 25% (Fig. 3). Phase micro-scopic examination disclosed that even with the highererythrocyte doses, only 45-55% of the incubated macro-phages were engaged in erythrophagocytosis, whereasvirtually all of the macrophages were found to interiorize

150_]-HEMEOXYGENE/Xd o~nd kmE

ERYTHROPHAGOCYTOSIS 0 E

2 ~~~~~~~~~~~~~~~~~~~~~~E

37 75 1030 0

Eryibocyes Offered (x106,

FIGURE 3 Stimulation of HO in macrophages offered largedoses of antibody-coated erythrocytes. Incubations wereperformed with (closed squares) or without (open squares)addition of glucose (5 mg/mI) and insulin (0.02 U/ml).Erythrophagocytosis (closed circles) was unaffected by thepresence of additional glucose and insulin.

O4r

a.i 0.3.E

E0.C.

E 0.2

.E

d 0.1

i-

-]14II

a54

* ACTINOMYCIN D

/, PUROMYCIN

1 2 3 4 5Addition of Actinomycin D or Puromycin after

Start of Incubation (hours)

FIGURE 4 Effect of inhibitors of protein synthesis on thestimulation of HO in macrophages. Actinomycin D, 0.2jug/ml (closed bars), or puromycin, 2 ,ug/ml (hatched bars),was added to the standard incubation system at various timeintervals after the start of the incubation. The open bars onthe left represent HO activity in native macrophages(basal) and in macrophages after erythrophagocytosis(stimulated) without added inhibitors.

carbon particles that had been added to the incubationmixture.

With erythrocyte doses ranging from 15 to 75 X 10fcells, stimulation of HO in the macrophages was strictlyproportional to the absolute number of red cells that hadbeen phagocytized (Fig. 2). When the number of offerederythrocytes exceeded 75 X 106 cells, no further stimula-tion of HOwas achieved. In fact, under standard con-ditions of incubation, larger doses of erythrocytes re-sulted in a progressive reduction of HOstimulation (Fig.3). This paradoxic relationship could be reversed almostcompletely by supplementing the incubation medium with5 mg glucose and 0.02 U insulin/ml (Fig. 3). Additionof glucose and/or insulin did not augment the enzymeresponse with the lower erythrocyte doses (Fig. 3).

Since 75 X 10' erythrocytes offered to 150 x 10' mac-rophages under standard conditions of incubation yieldedmaximal stimulation of HO, these conditions werechosen for all subsequent experiments.

When, instead of antibody-coated erythrocytes, macro-phages were incubated with hemoglobin dissolved in themedium, uptake was much lower than with intact cells,but increased linearly over the 5 h of incubation. With10.5 mg of hemoglobin in the incubation medium, whichis equivalent to the hemoglobin content of 600 X 106erythrocytes, uptake in 5 h was only 0.17 nmg or 1.6% ofthe offered dose. The hemoglobin interiorized underthis condition produced a three to fourfold stimulationof HO.

Effect of inhibition, of protein synthesis on stimulationof HO. Preliminary experiments indicated that 0.2 ,ug

Erythrocyte Catabolism by Macrophages 81.5

100 __ ERYTHROPHAGO-CYTOSIS

~60

40-

20 I

None 0.01 0.02 0.04 0.06 0.08 010 0.14 018

mg Hydrocortisone/ml medium

FIGURE 5 Effect of HC on erythrophagocytosis (openbars) and stimulation of HO (closed bars) in macrophages.Various concentrations of HC were added at the start ofthe incubation. The results are expressed as a percentageof the activity in macrophages incubated without HC.

actinomycin D or 2 isg puromycin/ml of the incubationmixture failed to exert a detectable effect on the rate orextent of erythrophagocytosis or on the viability of theincubated macrophages. Either compounds, when addedin the indicated concentration at the beginning of theincubation period completely suppressed the stimulationof HOby the phagocytized erythrocytes (Fig. 4). Whenactinomycin D was added to the medium at various timeintervals after the start of the incubation, its suppressiveeffect on the stimulation of HO was progressively re-duced, and by delaying the addition for 3 h or longer,the effect of the inhibitor on the enzyme stimulation wascompletely lost (Fig. 4). Puromycin was found to havea more protracted effect, in that addition of this com-pound 4 h after the beginning of the incubation still re-sulted in submaximal stimulation of HO in the macro-phages (Fig. 4).

Effect of hydrocortisone (HC) on erythrophagocyto-sis and stimulation of HO. Hydrocortisone (HG) in

100 _

80 _

6t 0 -

.t

;l 40 _

20_

Noa

p=UFTE OF

M

OXYGEfAI

).005 0.01 0.04 0.06 0.1 018

mg Hydrocortisone/ml medium

FIGURE 6 Effect of HG on pinocytosis of dissolved hemo-globin (open bars) and stimulation of HO (closed bars) inmacrophages. Results are expressed as in Fig. 5.

concentrations of up to 0.1 mg/ml in the medium pro-duced little interference with erythrophagocytosis in thestandard incubation system (Fig. 5). Larger concen-trations of the steroid, on the other hand, progressivelydepressed erythrophagocytosis, with 0.18 mg/ml leadingto almost 50% reduction in the phagocytosis of erythro-cytes. Comparable HC concentrations had been shownpreviously to impair phagocytosis of aggregated albuminparticles and bacteria (22). By contrast, when 10.5 mghemoglobin was dissolved in the medium, HC in theconcentrations tested did not affect the pinocytic ac-tivity (Fig. 6).

Stimulation of HOby the interiorized erythrocytes orhemoglobin was significantly suppressed by HC, evenwith steroid concentrations that did not interfere witherythrophagocytosis or with pinocytosis of dissolved he-moglobin (Figs. 5 and 6). With HCconcentrations rang-ing from 0.01 to 0.1 mg/ml, the suppressive effect of thesteroid on the HO stimulation was dose-related, withthe upper value resulting in complete abolition of theenzyme stimulation without significantly affecting eryth-rophagocytosis (Fig. 5). Because of this dissociation oferythrophagocytosis from HO stimulation by 0.1 mgHC/ml, this experimental condition was chosen for allsubsequent studies.

When HCwas added to the macrophages phagocytiz-ing erythrocytes at various time intervals after the startof the incubation (Fig. 7), the suppressive effect of thesteroid on the stimulation of HO was clearly time-re-lated in a manner similar to the effect of puromycin(Fig. 4). Complete suppression of enzyme stimulation

0.4

GE0

C.

0coE

-D

z

.I

0E

0.3-

02 _-

0.1 raA<to<

0

S4 1130 1 2 3 4 41/2 5

Addition of Hydrocortisone after

Start of Incubation (hours)

FIGURE 7 Effect of HC on stimulation of HO in macro-phages. HC (0.1 mg/ml) was added to the medium atvarious time intervals after the start of the incubation. Inthis and in Fig. 8, the open bars on the left represent HOactivity in incubated native macrophages (basal) and inmacrophages after erythrophagocytosis (stimulated) with-out the addition of HC.

816 D. Gemsa, C. H. Woo, H. H. Fudenberg, and R. Schmid

I

required the presence of the steroid during the 1st h of in-cubation (Fig. 7). Delayed addition of HC gave rise toprogressively less suppression, until after 4 h the steroidhad lost nearly all of its suppressive effect (Fig. 7).

Reversal of HC effect on stimulation of HO. Thesuppressive effect of HC on the stimulation of HO inmacrophages phagocytizing erythrocytes was partiallyreversed by supplementing the incubation medium withglucose, fructose (5 mg/ml), or insulin (0.02 U/ml)(Fig. 8). Addition of the maltose was without effect.The combination of 5 mg/ml glucose with 0.02 U/mlinsulin completely restored the stimulation of HO andthus effectively counteracted the suppressing effect of0.1 mg/ml HC. On the other hand, combination of in-sulin with fructose or maltose failed to yield an additiveeffect (Fig. 8).

The beneficial effect of glucose and insulin was evenobserved when their addition to the culture medium,containing HC from the start, was delayed for severalhours after the beginning of the incubation. Glucoseand insulin completely reversed the steroid-mediatedsuppression of HOstimulation when added 3 h after thestart of the incubation. With later addition, their re-versing effect was progressively reduced (Fig. 9).

Removal of glucose from the incubation medium. Inorder to estimate the amount of glucose metabolized bythe macrophages in 5 h of incubation, glucose removalfrom the incubation medium was measured under thefollowing conditions of incubation: macrophages alone;macrophages with the standard amount of antibody-coated erythrocytes with or without additional glucoseand insulin, or HC, or both. As shown in Fig. 10, glu-cose removal was not appreciably enhanced by erythro-phagocytosis, but was increased approximately one-thirdby addition of glucose and insulin. HC in a concentra-

0.4

E

a 0.3

.C

a

a 02.01-

_

2F

I00r-

80

60

40

20 Io '' 3 A 4/2 5

Addition of Glucose and InsulinAfter Start of Incubation (hours)

FIGURE 9 Reversal by glucose and insulin of the inhibitoryeffect of HC on the stimulation of HO in macrophages.Standard incubations contained 0.1 mg/ml HC. Glucose (5mg/ml) and insulin (0.02 U/ml) were added at varioustime intervals after the start of the incubation. Results areexpressed as percent reversal of the inhibition produced byHC in the absence of added glucose and insulin.

tion of 0.1 mg/ml greatly reduced glucose removal butthis steroid-mediated repression was partly reversed byaddition of glucose and insulin (Fig. 10).

A similar pattern was obtained by analysis of "4CO2formed from [1-14C]glucose added to the incubation mix-

16

14

E 12

° 100

E0 8

2

4

2

!_

I-

IOAdditions Glucose

Insulin

0 Glucose Insulin Frucose Mohose Glucose Fruciose Mobose0 0 0 0 0 4- Insulin-- p-

FIGURE 8 Reversal by hexoses with or without insulin ofthe inhibitory effect of HC on the stimulation of HO inmacrophages. Standard incubations contained 0.1 mg/mlHC and either glucose, fructose or maltose (5 mg/ml) withor without insulin (0.02 U/ml).

Hydrocortisone Hydrocortisone

0 Glucose0 Insulin

FIGURE 10 Removal of glucose from the incubation me-dium by macrophages during 5 h of incubation. The stan-dard incubation system (15 ml) containing 15 mg of glucosewas supplemented with 75 mg glucose, 0.3 U insulin, or1.5 mg HC in various combinations. The open bars on theleft represent glucose removal in the standard incubationsystem without additions. Each bar represents the averagevalue of two separate incubations, the brackets indicatingthe range of values.

Erythrocyte Catabolism by Macrophages 817

a

>1ac

mh

.a

<0do

OA

(N0u

v0SI

0v

0E1

031-

021

0.1

<b-

vmllllJ 0 Hydrocortisone Hydrocortisone

Additions Glucose 0 Glucose

Insulin 0 Insulin

FIGURE 11 Conversion of [1-14C]glucose to CO2 in macro-phages during 3 h of incubation. The standard incubationsystem (15 ml) containing 15 mg of glucose was supple-mented with 75 mg glucose, 0.3 U insulin, or 1.5 mg HCin various combinations. The open bars on the left representglucose conversion in the standard incubation system with-out additions. Each bar represents the average value ofthree separate incubations, the brackets indicating the rangeof values.

ture (Fig. 11). During the first 3 h of incubation,macrophages phagocytizing erythrocytes in the presenceof HC produced substantially less '4CO2 than when thesteroid was absent. Glucose and insulin added to the in-cubation reversed this steroid effect. After 5 h of incuba-tion, the results obtained were essentially similar, exceptthat glucose and insulin were moderately less effectivein reversing the suppressive effect of HC.

DISCUSSION

There is extensive morphologic (23, 24) and experi-mental (10, 25) evidence that macrophages and othermononuclear phagocytes possess the capacity to engulfintact or fragmented erythrocytes and to digest them.Erythrophagocytosis by these cells appears to be an im-portant, and perhaps the primary mechanism by whichphysiologically aged or damaged red blood cells areremoved from the circulation (1, 26-28). This processcan be reproduced in vitro, in that in a variety of cul-ture systems isolated macrophages may be observed in-gesting and degrading offered erythrocytes that previ-ously had been injured by chemical or physical means(10, 29). Antibody-coated red cells are also ingestedand degraded at 370C, presumably as the result of bind-ing of the Fc fragment of the IgG to the monocyte-macrophage receptor for IgG (30). In the present study,this technique has been refined and controlled in sucha way that a standard number of peritoneal macrophagesin primary culture reproducibly phagocytized a predic-

table number of antibody-coated autologous erythrocytesin a specified time. Several observations made with thisprimary culture system are worthy of comment.

Although under standard conditions, the macrophageswere incubated with erythrocytes for a total period of 5h, it was apparent that the rate of erythrophagocytosiswas most rapid during the 1st h, progressively slowedduring the 2nd h, and approached zero after the 3rd hof incubation (Fig. 1). This decline in the rate of eryth-rophagocytosis occurred despite the availability of ampleunagglutinated erythrocytes in the culture medium. It isunlikely that this phenomenon reflected potential "satu-ration" of the macrophages with red blood cells, as an in-crease in the number of offered erythrocytes resulted inan absolute and proportional rise of the number of cellsthat were ingested by the macrophages (Fig. 2). Fur-thermore, the cultured macrophages appeared to remainfully viable throughout the entire period of incubation,as shown by their continued adherence to the surface ofthe incubation flask, their active glucose assimilation andC02 production, their trypan blue exclusion, and theirability to respond to stimulation with formation of newenzyme protein. Although the exact cause of this time-dependent progressive reduction in erythrophagocytosisremains undetermined, possible explanations includequantitative differences in the degree of antibody coatingof individual erythrocytes, which may be related to cellage (31, 32), or progressive elution of antibody fromerythrocytes during the course of the incubation. De-pletion of factors essential for phagocytosis in the incu-bation medium or in the macrophages seems unlikely, asaddition of a second fresh dose of antibody-coatederythrocytes to the macrophage culture 2 h after the startof the incubation reactivated erythrophagocytosis.

It also was noted with this culture system that only45-55% of the incubated macrophages actually partici-pated in the ingestion of erythrocytes, whereas carbonparticles were interiorized by virtually all macrophages.Even when the number of offered erythrocytes greatlyexceeded that of macrophages, microscopic inspectionrevealed a relatively constant proportion of macro-phages that were devoid of ingested erythrocytes, whereasothers contained several red cells. Otherwise, no morpho-logic differences could be detected between macrophagesthat had, or had not, phagocytized erythrocytes. Whereasno plausible explanation can be offered for this functionaldifference among the incubated macrophages, the find-ings are reminiscent of similar observations with thesinusoidal cell population of rat liver (33). These phago-cytes also appear to be heterogeneous in composition inthat they include a subpopulation of cells recognizableby histochemical and enzymatic methods, which is in-volved specifically in erythrophagocytosis. It is possiblethat macrophages harvested after chemical irritation of

818 D. Gemsa, C. H. Woo, H. H. Fudenberg, and R. Schmid

11;

the peritoneal cavity are similarly heterogeneous and thatonly a fraction of these cells is equipped for phagocyto-sis of antibody-coated erythrocytes. It also should benoted that since in the present experiments, macrophageswere not preincubated with serum-containing medium(10, 25), the cultures may have contained cells of dif-ferent maturity.

The precise mechanisms involved in the disruptionand subsequent degradation of the ingested erythrocytesare incompletely understood. Morphological and bio-chemical (10) evidence suggests that the interiorized redcells are attacked by lysosomal enzymes which are re-leased into the phagocytic vesicles after they have fusedwith the lysosomes (34). This process is presumed to beanalogous to the sequential degradation of phagocytizedbacteria (35). The identification of a variety of proteo-lytic enzymes in lysosomes facilitated a more detailed in-vestigation of this degradative mechanism. Axline andCohn (10) recently reported that peritoneal macrophagesincubated for 24 h after ingestion of formaldehyde-treatederythrocytes exhibited a net increase in lysosomal en-zyme activity. This increase in proteolytic enzymes waspresumed to reflect enzyme induction mediated by themembrane proteins and the globin moiety of the phagocy-tized erythrocytes. The fate of the heme moiety of theingested erythrocyte-hemoglobin was not determined, butprevious observations suggested that it is converted tobile pigment (36). The mechanism of this conversionwas obscure until the recent demonstration by Pimstone,Tenhunen, Seitz, Marver, and Schmid (15), that macro-phages possess HO, which is the enzyme system catalyz-ing the conversion of ferroprotoporphyrin (heme) tobilirubin IXa (13, 14). Although HO activity is verylow in native peritoneal macrophages, presumably be-cause they normally lack contact with heme compounds,a striking stimulation of enzyme activity was observedafter macrophages had been exposed to methemalbuminor hemoglobin in the peritoneal cavity (15).

In the present study, the low HO activity of nativeperitoneal macrophages or of macrophages incubatedwithout red cells has been confirmed. In contrast, whenthe macrophages were offered antibody-coated erythro-cytes in the culture, their HOactivity regularly rose inproportion to the absolute number of red cells ingested,at least with the lower red cell doses (Fig. 2). There wasa time lag of about 3 h between erythrophagocytosis andstimulation of HO (Fig. 1), which is strikingly similarto the time interval that is required in vivo for theconversion of the hemoglobin of sequestered erythro-cytes to bilirubin in intact rats (36). The stimulationof HO in the cultured macrophages was paralleled bymorphologic evidence that after 4-5 h of incubation, theingested erythrocytes became visibly pale and had ob-viously undergone disintegration. It is noteworthy that

the response of microsomal HO to the ingestion oferythrocytes was much more rapid and more pronouncedthan that reported for lysosomal enzymes (10). Althoughthis may reflect the different intracellular localization ofthese enzyme systems, it is more likely that the retardedand attenuated stimulation of the lysosomal enzymes wasdue to the pretreatment of the ingested erythrocytes withformaldehyde, which may have reduced the digestibilityof the red cells or their components. Indeed, when macro-phages were incubated with formaldehyde-fixed erythro-cytes, there was a marked delay in the stimulation ofHO. Moreover, erythrophagocytosis occurred at a lowerrate, but over a more extended period of time, and a

larger percentage of the macrophages appeared to be en-

gaged in erythrophagocytosis, as compared with incu-bations containing antibody-coated red cells.2

In addition to intact antibody-coated erythrocytes,hemoglobin dissolved in the culture medium also stimu-lated HOactivity in the cultured macrophages (Fig. 6).However, in order to achieve comparable enzyme stimu-lation, far higher concentrations of dissolved hemo-globin were required than the hemoglobin equivalent tothe incubated intact erythrocytes. This difference prob-ably reflects the relatively low pinocytotic uptake ofdissolved hemoglobin by the macrophages (11) whichappears to be proportional to the extracellular concen-tration of hemoglobin. The different handling of dis-solved hemoglobin and of intact erythrocytes by themacrophages is further emphasized by the observationthat even the highest concentrations of HC added to theculture medium failed to depress pinocytosis of henmo-globin (Fig. 6), whereas with these high steroid con-centrations phagocytosis of intact erythrocytes was sig-nificantly reduced (Fig. 5). The preference of macro-phages for the ingestion of intact red cells, as contrastedto dissolved hemoglobin, is analogous to recent observa-tions in the liver where erythrocytes are phagocytizedexclusively by a subpopulation of sinusoidal cells, whereasdissolved hemoglobin is taken up mainly by hepaticparenchymal cells (37).

The stimulation of HO appeared to depend on newprotein synthesis, as actinomycin D and puromycinstrikingly inhibited the increase in enzyme activity with-out reducing the rate or magnitude of erythrophagocyto-sis (Fig. 4). These findings and previous observationsof a similar nature in rat kidney (38) strongly suggestthat HO is induced by its substrate, heme. Moreover,when these two inhibitors of protein synthesis vereadded to the macrophage cultures at various time inter-vals after the beginning of the incubation, a distinctdifference in the effect patterns was discernable (Fig.

2 Gemsa, D., H. H. Fudenberg, and R. Schmid. Unpub-lished observations.

Erythrocyte Catabolism by Macrophages 819

4). The suppressive effect of actinomycin D on enzymestimulation was more limited in time as compared withthe more protracted action of puromycin, consistent withthe concept that stimulation of HO by the ingestederythrocyte-hemoglobitn requires formation of additionalmessenger RNA.

Induction of HO in macrophages after erythrophago-cytosis also appeared to be critically dependent on anadequate supply of exogenous glucose. This became evi-dent when large doses of antibody-coated erythrocyteswere offered to the standard number of 150 X 106 macro-phages. Whereas with relatively small red cell doses inthe range of 15-75 X 106 cells, enzyme induction wasproportional to the absolute number of ingested erythro-cytes (Fig. 2), with larger doses no further rise in HOactivity was observed despite the increase in the numberof phagocytized cells (Fig. 3). As seen in Fig. 2, withthe standard incubation medium containing 1 mg glu-cose/ml, the response of HO to erythrocyte doses ex-ceeding 75 X 106 cells was suboptimal. This apparentparadox was explained when it was found that additionof glucose and insulin to the incubation mixture re-paired this defective enzyme response (Fig. 3). Sup-plementation of the incubation medium with either glu-cose or insulin alone restored the enzyme induction onlypartially, which may suggest that under the "stress" oflarge loads of ingested erythrocytes, insulin-dependentglucose metabolism may become essential for optimalmacrophage function.

Further support for the essential role of glucosemetabolism in the substrate-mediated induction of HOin macrophages was obtained in experiments with HC.Munck and Young (39, 40) presented convincing evi-dence that HC markedly interferes with glucose trans-port and utilization in thymus cell cultures. In culturedlymphocytes the steroid reportedly depresses proteinsynthesis (41, 42), but this may be only a secondary ef-fect resulting from impaired glucose metabolism. Be-cause of these observations, HC in various doses was

added to standard incubations of 150 X 106 macrophagesand 75 X 106 antibody-coated erythrocytes. In the con-

centration range from 0.01 to 0.1 mg HC/ml medium,the steroid produced a progressive, dose-related repres-sion of HOinduction (Fig. 5). With a HC concentra-tion of 0.1 mg/ml, enzyme induction was completely sup-pressed while erythrophagocytosis was not significantlyimpaired. When addition of the steroid was delayed af-ter the start of the incubation (Fig. 7), the response pat-tern of HOinduction was similar to that observed withpuromycin, which may suggest that HC ultimately af-fects enzyme protein synthesis at the translational level.

The suppressive effect of HC on HO induction wasreadily reversible by supplementing the incubation me-

dium with additional glucose and insulin (Fig. 8). Thesuboptimal response of the enzyme activity to either glu-cose or insulin alone suggests a synergistic action of thetwo. This is reminiscent of the additive effect of glucoseand insulin on macrophages which had been offered verylarge doses of erythrocytes without the addition ofsteroid (Fig. 3). The importance of glucose for HOinduction was further emphasized when addition of glu-cose and insulin to the HC-containing cultures wasdelayed for several hours. The finding that glucose andinsulin added 3 h after the start of the incubation com-pletely reversed the inhibitory effect of HC (Fig. 9) isconsistent with the concept that the exogenous glucose isrequired for the synthesis of new enzyme protein. More-over, when glucose was replaced by the monosaccharidefructose, whose metabolism is insulin-independent (43),the induction of HOwas only partially repaired regard-less of the presence or absence of insulin (Fig. 8). Thedisaccharide maltose, although it may be metabolized inmacrophages (44), was without beneficial effect. Mal-tose combined with insulin yielded a degree of enzymestimulation comparable with that of insulin alone. Fi-nally, HC reduced glucose removal from the medium bythe phagocytizing macrophages (Fig. 10) and loweredproduction of C02 from glucose via the hexose mono-phosphate shunt (Fig. 11); both effects were partiallyor completely reversed by insulin and additional glucose.

These findings clearly indicate that in macrophagesthe induction of HO by the hemoglobin of ingestederythrocytes is dependent on adequate glucose transportor metabolism, and that HC interferes with this process.This may be analogous to the reported observation thatin thymus cell cultures, the steroid impairs the forma-tion of glucose-6-phosphate (39). The information pres-ently available is insufficient to determine whether thissteroid effect consists of partial inhibition of glucosetransport into the cell or reflects interference with earlysteps of intracellular glucose metabolism. Likewise, it isunknown how this impairment of carbohydrate metabo-lism expresses itself in reduced enzyme synthesis. It isnoteworthy, however, that under these selected experi-mental conditions, complete dissociation of erythrophago-cytosis from HO induction was achieved with HC con-centrations of the order of magnitude of those obtainedin the plasma under pharmacological conditions. More-over, it was evident on microscopic inspection that withthese steroid concentrations, not only was erythrophago-cytosis unaffected but, despite the lack of a HO response,digestion of ingested erythrocytes appeared to proceedwithout noticeable impairment. This apparent disparitypermits, three alternative interpretations, namely (a)that HOactivity is not critical for the conversion of theingested hemoglobin-heme to bilirubin IXa; (b) that

820 D. Gemsa, C. H. Woo, H. H. Fudenberg, and R. Schmid

the heme of the degraded hemoglobin simply accumulatesin the macrophages without being catabolized; or (c)that in the absence of increased HOactivity, hemoglobin-heme is degraded by alternate catabolic pathways lead-ing to metabolites that differ from bilirubin. The firstinterpretation appears unlikely, as the HOsystem is theonly mechanism known to produce exclusively bili-rubin IXa, which is the bile pigment isomer formed invivo (14). Moreover, in a variety of tissues, HO ac-tivity was shown to be stimulated in response to an in-creased supply of heme, derived from endogenoussources or administered parenterally (45). The secondpossibility, while perhaps unlikely on theoretical grounds,cannot be excluded with the presently available experi-mental means. The third explanation postulating opera-tion of alternate pathways for heme degradation ap-pears the most plausible, since the existence of suchdegradative pathways which do not lead to formation ofbilirubin has already been demonstrated (46, 47). Al-though the sequence of the chemical reactions and thenature of the intermediates involved are as yet unknown,it has previously been reported that on incubation ofheme with homogenates of stimulated macrophages,heme disappearance substantially exceeds the rate ofbilirubin formation (15).

If this last interpretation should prove to be correct, itmay provide a possible explanation for the beneficial ef-fect of steroid treatment in patients with jaundice due toparenchymal or obstructive liver disease. In thesepatients, steroids tend to produce a well documented(48, 49), but unexplained (50), reduction in the serumbilirubin level. If the present findings are applicableto erythrophagocytic cells in general, it is possible thatsteroids may reduce the amount of bilirubin formed fromsequestered erythrocytes without impairing the physio-logical removal mechanism of senescent red cells fromthe circulation. In order to prove this concept, it would benecessary to characterize the alternate pathways of hemedegradation and to demonstrate experimentally that theybecome functional under steroid therapy. Furthermore,since the present findings suggest that glucose metabo-lism becomes insulin-dependent in macrophages de-grading a large load of erythrocytes, it would be of in-terest to determine whether insulin deficiency states orhypercortisonism nmay be associated with defective mac-rolhage function.

ACKNOWLEDGMENTSThe authors are grateful to Miss Lydia Hammaker forvaluable advice and to Doctors D. Stites and D. Hanes forreviewing the manuscript.

This work was supported by Research Grants AM-11275and HD-05894, Training Grants AM-05598 and HL-05677from the National Institutes of Health, and by the WalterC. Pew Fund for Gastrointestinal Research.

REFERENCES

1. Harris, J. WV., and R. WV. Kellermeyer. 1970. The RedCell. Harvard University Press, Cambridge. Revisededition. 524.

2. Jandl, J. H., A. R. Jones, and W. B. Castle. 1957. Thedestruction of red cells by antibodies in man. I. Ob-servations on the sequestration and lysis of red cellsaltered by immune mechanisms. J. Clin. Invest. 36: 1428.

3. Jandl, J. H. 1967. The spleen and reticuloendothelialsystem. In Pathologic Physiology: Mechanisms of Dis-ease. WV. A. Sodeman and W. A. Sodeman, Jr., editors.W. B. Saunders Co., Philadelphia. 4th edition. 897.

4. Keene, WV. R., and J. H. Jandl. 1965. Studies of thereticuloendothelial mass and sequestering function of ratbone marrow. Blood. 26: 157.

5. Wagner, H. N., Jr., M. A. Razzak, R. A. Gaertner,WV. P. Caine, Jr., and 0. T. Feagin. 1962. Removal oferythrocytes from the circulation. Arch. Intern. Med.110: 90.

6. Virchow, R. 1847. Die pathologischen Pigmente. Arch.Pathol. Anat. Physiol. Klin. Med. (Virchous). 1: 379.

7. Langhans, T. 1869. Beobachtungen fiber Resorption derExtravasate und Pigmentbildung in denselben. Arch.Pathol. Anat. Phvsiol. Kinm. M1ed. (Virchows). 49: 66.

8. Essner, E. 1960. An electron microscopic study of eryth-rophagocytosis. J. Biophys. Biochein. Cytol. 7: 329.

9. North, R. J. 1966. The localization by electron micros-copy of acid phosphatase activity in guinea pig macro-phages. J. Ultrastrzict. Res. 16: 96.

10. Axline, S. G., and Z. A. Cohn. 1970. In vitro inductionof lysosomal enzymes by phagocytosis. J. Exp. Med.131: 1239.

11. Ehrenreich, B. A., and Z. A. Cohn. 1968. Fate ofhemoglobin pinocytosed by macrophages in vitro. J.Cell Biol. 38: 244.

12. Petrik, P., and A. J. Collet. 1971. Electron microscopicstudy of the in vivo erythrophagocytosis by alveolarmacrophages of the cat. II. Late period: digestion. Z.Zellforsch. Mikrosk. Anat. 116: 477.

13. Tenhunen, R., H. S. Marver, and R. Schmid. 1968. Theenzymatic conversion of heme to bilirubin by micro-somal heme oxygenase. Proc. Natl. Acad. Sci. U. S. A.61: 748.

14. Tenhunen, R., H. S. Marver, and R. Schmid. 1969.Microsomal heme oxygenase. Characterization of the en-zyme. J. Biol. Chem. 244: 6388.

15. Pimstone, N. R., R. Tenhunen, P. T. Seitz, H. S.Marver, and R. Schmid. 1971. The enzymatic degrada-tion of hemoglobin to bile pigments by macrophages.J. Exp. Med. 133: 1264.

16. Mollison, P. L. 1967. Blood Transfusion in ClinicalMedicine. F. A. Davis Co., Philadelphia. 4th edition. 398.

17. Baehner, R. L., N. Gilman, and M. L. Karnovsky. 1970.Respiration and glucose oxidation in human and guineapig leukocytes: comparative studies. J. Clin. Invcst. 49:692.

18. Raabo, E., and T. C. Terkildsen. 1960. On the enzy-matic determination of blood glucose. Scand. J. Clin.Lab. Invest. 12: 402.

19. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193: 265.

20. Crosby, W. H., and D. N. Houchin. 1957. Preparingstandard solutions of cyanmethemoglohin. Blood. 12%1132.

Erythrocyte Catabolism by Macrophages 821

21. Crosby, W. H., and F. W. Furth. 1956. A modificationof the benzidine method for measurement of hemoglobinin plasma and urine. Blood. 11: 380.

22. Wiener, E., Y. Marmary, and Z. Curelaru. 1972. Thein vitro effect of hydrocortisone on the uptake andintracellular digestion of particulate matter by macro-phages in culture. Lab. Invest. 26: 220.

23. Rifkind, R. A. 1966. Destruction of injured red cell invivo. Am. J. Med. 41: 711.

24. Weiss, L., and M. Tavassoli. 1971. Anatomical hazardsto the passage of erythrocytes through the spleen.Semin. Hematol. 7: 372.

25. Rabinovitch, M. 1967. The dissociation of the attach-ment and ingestion phases of phagocytosis by macro-phages. Exp. Cell. Res. 46: 19.

26. Weed, R. I., and C. F. Reed. 1966. Membrane altera-tions leading to red cell destruction. Am. J. Med. 41:681.

27. Bunn, H. F. 1972. Erythrocyte destruction and hemo-globin catabolism. Semin. Hematol. 9: 3.

28. Jandl, J. H., N. M. Files, S. B. Barnett, and R. A.MacDonald. 1965. Proliferative response of the spleenand liver to hemolysis. J. Exp. Med. 122: 299.

29. Rabinovitch, M. 1969. Phagocytosis of modified eryth-rocytes by macrophages and L2 cells. Exp. Cell Res.56: 326.

30. Huber, H., and H. H. Fudenberg. 1968. Receptor sitesof human monocytes for IgG. Int. Arch. Allergy. Appl.Immunol. 34: 18.

31. Marikovsky, Y., and D. Danon. 1971. Agglutination ofyoung and old human red cells by blood group anti-bodies. Vox Sang. 20: 174.

32. Greenwalt, T. J., L. L. Flory, and E. A. Steane. 1970.Quantitative haemagglutination. III. Studies of sepa-rated populations of human red blood cells of differentdensities. Br. J. Haematol. 1-9: 701.

33. Bissell, D. M., L. Hammaker, and R. Schmid. 1972.Liver sinusoidal cells. Identification of a subpopulationfor erythrocyte catabolism. J. Cell Biol. 54: 107.

34. Cohn, Z. A., and E. Wiener. -1963. The particulatehydrolases of macrophages. I. Comparative enzymology,isolation, and properties. J. Exp. Med. 118: 991.

35. Cline, M. J. 1970. Leukocyte function in inflammation:the ingestion, killing, and digestion of microorganisms.Ser. Haematol. 3 (No. 2): 3.

36. Ostrow, J. D., J. H. Jandl, and R. Schmid. 1962. Theformation of bilirubin from hemoglobin in vivo. J. Clin.Invest. 41: 1628.

37. Bissell, D. M., L. Hammaker, and R. Schmid. 192.Hemoglobin and erythrocyte catabolism in rat liver: theseparate roles of parenchymal and sinusoidal cells.Blood. 40: 812.

38. Pimstone, N. R., P. Engel, R. Tenhunen, P. T. Seitz,H. S. Marver, and R. Schmid. 1971. Inducible hemeoxygenase in the kidney: a model for the homeostaticcontrol of hemoglobin catabolism. J. Clin. Invest. 50:2042.

39. Munck, A. 1968. Metabolic site and time course of cor-tisol action on glucose uptake, lactic acid output, andglucose 6-phosphate levels of rat thymus cells in vitro.J. Biol. Chem. 243: 1039.

40. Young, D. A. 1969. Glucocorticoid action on rat thymuscells. Interrelationships between carbohydrate, protein,and adenine nucleotide metabolism and cortisol effectson these functions in vitro. J. Biol. Chem. 244: 2210.

41. Mayman, M. H., B. Dvorkin, and A. White. 1968. In-fluence of cortisol on the utilization, of precursors ofnucleic acids and protein by lymphoid cells in vitro.J. Biol. Chem. 243: -1485.

42. Werthamer, S., C. Hicks, and L. Amaral. 1969. Proteinsynthesis in human leukocytes and lymphocytes. I.Effect of steroids and sterols. Blood. 34: 348.

43. Nocenti, M. R. 1968. Endocrine Functions of the Pan-creas. In Medical Physiology. V. B. Mountcastle, editor.The C. V. Mosby Co., St. Louis. 12th edition. 1029.

44. Cohn, Z. A., and B. A. Ehrenreich. 1969. The uptake,storage, and intracellular hydrolysis of carbohydratesby macrophages. J. Exp. M1ed. 129: 201.

45. Tenhunen, R., H. S. Marver, and R. Schmid. 1970. Theenzymatic catabolism of hemoglobin: stimulation ofmicrosomal heme oxygenase by hemin. J. Lab. Clin. Med.75: 410.

46. Goldstein, G. W., L. Hammaker, and R. Schmnid.1968. The catabolism of Heinz bodies: an experimentalmodel demonstrating conversion to non-bilirubin catabo-lites. Blood. 31: 388.

47. Landaw, S. A., E. W. Callahan, Jr., and R. Schmid.1970. Catabolism of heme in vivo: comparison of thesimultaneous production of bilirubin and carbon mon-oxide. J. Clin. Invest. 49: 914.

48. Katz, R., H. Ducci, and H. Alessandri. 1957. Influenceof cortisone on hyperbilirubinemia. J. Clin. Invest. 36:1370.

49. Schiff, L. 1966. The use of steroids in liver disease.Medicine (Baltimore). 45: 565.

50. Williams, R., and B. H. Billing. 1961. Action of steroidtherapy in jaundice. Lancet. II: 392.

D. Gemsa, C. H. Woo, H. H. Fudenberg, and R. Schmid822