Surface marker andother characteristics Gaucher's cells

J. clin. Path., 1977, 30, 981-988

Surface marker and other characteristics ofGaucher's cellsG. F. BURNS, J. C. CAWLEY, R. J. FLEMANS, K. E. HIGGY, C. P. WORMAN,C. R. BARKER, B. E. ROBERTS, AND F. G. J. HAYHOE

From the Department of Haematological Medicine, University of Cambridge and Department of Medicine,University ofLeeds, UK

SUMMARY A full surface marker study of the splenic storage cells in a case of Gaucher's diseaselargely substantiates the monocyte/histiocyte nature of Gaucher's cells. In addition, an apparentT-lymphocyte deficiency is demonstrated in the spleen and peripheral blood, and the possiblesignificance of this finding is discussed.

Gaucher's cells have been extensively investigatedmorphologically and biochemically, and it is nowthought that they represent histiocytic cells in whichlarge amounts of glucocerebroside have accumulatedbecause of a deficiency of ,B-glucosidase (Brady,1972). The histiocytic nature of Gaucher's cellshas been inferred from their phagocytic ability (Leeet al., 1967), high lysosomal enzyme content(Rozenszajn and Efrati, 1961), and distinctivelyreticuloendothelial distribution. However, this viewhas not been universally accepted (Roos et al., 1961;Fisher and Reidbord, 1962) and, in addition,immunological mechanisms have been invoked inthe pathogenesis of the disease (Pennelli et al., 1969).Immunological surface marker studies, which haveproved of great value in the characterisation of thepathological cells of a wide variety of other haemato-logical proliferations, should therefore contribute tofurther understanding of the nature of the pathog-nomonic cells of Gaucher's disease.No previous surface-marker study of Gaucher's

disease has been described, and in the present paperwe report such a study in a case of the non-neurono-pathic form of the disease.

Patient and methods

CASE HISTORYThis 8-year-old girl of Anglo-Saxon descent presen-ted with recurrent epistaxes. Examination revealedmassive splenomegaly and moderate hepatomegalybut no otherabnormality. Theserumacidphosphatasewas greatly increased at 17-2 KA units, and skeletal

Received for publication 7 March 1977

survey showed early, but definite, Ehrlenmeyer-flaskdeformity of the femora. There was a peripheralpancytopenia (Hb 10-7 g/dl, WBC 3'8 x 109/l,platelets 60 x 109/l) and differential white countshowed 44% neutrophils, 49% lymphocytes, and7% monocytes. A bone marrow aspirate was hyper-cellular and contained moderate numbers of typicalGaucher's cells; the cytology and cytochemistry ofthese are described in detail below. A diagnosis ofnon-neuropathic Gaucher's disease with hyper-splenism was made.The only other sib, a younger sister aged 54 years,

also has hepatosplenomegaly with peripheralpancytopenia and high serum acid phosphatase andalmost certainly has Gaucher's disease. However,since she is asymptomatic a bone marrow aspiratehas not been performed to confirm the diagnosis.There was no other family history of Gaucher'sdisease.

Because of the patient's thrombocytopenicbleeding attributable to hypersplenism, splenectomywas performed in December 1976 and this splenicmaterial formed the basis of our study. Postopera-tively, the patient has done well, and the pancyto-penia has disappeared with cessation of the thrombo-cytopenic epistaxes.

CYTOLOGY AND CYTOCHEMISTRYBone marrow and splenic impression smears wereexamined by a variety of routine methods includingpolarised and ultraviolet microscopy, Romanowsky,PAS, acid and alkaline phosphatase, iron, peroxidase,Sudan black, a-naphthyl butyrate esterase withsodium fluoride inhibition, and naphthol AS-Dchloroacetate esterase stains. Bone marrow andsplenic material, fixed and processed in the manner

981

on June 8, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.30.10.981 on 1 O

ctober 1977. Dow

nloaded from

http://jcp.bmj.com/

982

described below, were also examined by electronmicroscopy.

PREPARATION OF CELL SUSPENSIONS

Splenic cellsA single cell suspension was obtained by forcingseveral pieces from different areas of freshly removedspleen through a 120 gauge stainless-steel mesh intoa Hepes-buffered Hanks balanced salt solution(HBSS). The cell suspension was kept on ice untilits return to the laboratory where the cells were

washed in HBSS, and mononuclear cells wereseparated by centrifugation over Ficoll-Hypaquegradients (Boyum, 1968). The cells were washed afurther three times in HBSS before being resus-pended in HBSS + 0-2% bovine serum albumin(BSA) to a final concentration of 2 x 106 cells/mi.For comparison, suspensions from a 'normal'

spleen (removed during gastrectomy for carcinomaof the stomach) and the spleen from a patient withhereditary spherocytosis (HS) were prepared in asimilar way.

Peripheral blood cellsThe patient's peripheral blood was examined ontwo occasions-initially, five days after splenectomyand, on a second occasion, 35 days post-splenectomy.On the latter occasion, the sister's blood was alsoexamined. In each case, the blood was collected intoheparin and the mononuclear cells were separated inthe manner described above.

In both the splenic and the peripheral blood cellsuspensions viability was assessed by the cells'ability to convert fluorescein diacetate to fluorescein(Celada and Rotman, 1967). Viability was greaterthan 95% in the peripheral blood, HS, and normalcell preparations. However, in the suspension fromthe Gaucher's spleen preparation the overall viabilitywas 80%, while the Gaucher's cells, which consti-tuted approximately 30% of this preparation, showedsome 50% viability, dead Gaucher's cells accountingfor the great majority of the non-viable cells.

ROSETTE TESTS

E-rosettesSheep erythrocyte rosette tests (E) were carried outusing aminoethylisothiouronium bromide-treatederythrocytes according to the method of Kaplanand Clark (1974).

Fc-rosettesThe method of Hallberg et al. (1973), using ox RBCssensitised with rabbit anti-ox RBC IgG antibody,was employed. Unsensitised washed ox RBCs were

G. F. Burns et al.

used as an Fc control.

EAIgM-rosettesOx RBCs coated with rabbit anti-ox RBC IgMantibody (EAIgM) were used by a method previouslydescribed by us (Barker et al., 1976; Burns et al.,1977). This test also served as a control for the C3rosettes.

C3-rosettes(a) C3bmo and C3dmol For C3bmo indicator cells,IgM-coated ox RBCs (0 5 ml) were pre-warmed to37°C and an equal volume of whole AKR mouseserum (1/10 dilution) was added. These were mixedfor 70 s at 37°C and 5 jil Antrypol(R) (50 mg/ml),which stabilises C3b against the action of KAF,was added and thoroughly mixed for 2 min (Lach-mann etal., 1973).As a source of C3dmo-coated cells, the reaction

was allowed to proceed at 37°C without addedAntrypol for 30 min during which time most C3bis converted to C3d by KAF.(b) C3bhu and C3dhu2 These two types of indicatorRBC were prepared in an identical way except thathuman serum previously exposed to zymosan (R3reagent) was used as the source of complement(McConnell and Hurd, 1976).

All the C3 rosette tests were performed by themethod of Ross et al. (1973).

Mouse (M-) rosettesThe method of Gupta et al. (1976), using fresh,washed, unsensitised CBA mouse erythrocytes wasemployed.

In all rosette tests, the indicator RBCs werewashed three times and resuspended to 1% inHBSS + 0-2% BSA before testing, and percentagesof rosetting cells were assessed using the fluoresceindiacetate method of Ramasamy (1974), scoring onlythe viable fluorescing cells under combined UV andphase contrast microscopy.

SURFACE IMMUNOGLOBULIN (SMIG)An indirect method was employed which usespolyvalent rabbit anti-human gammaglobulin raisedin the laboratory followed by fluoresceinated goatanti-rabbit IgG (Gibco Bio-Cult Ltd, Paisley,Scotland).

1C3bmo is the bound fragment of activated mousecomplement component C3, and C3dmo is the residualfragment of the inactivation of C3b by conglutinogenactivating factor (KAF).2C3bhu and C3dhu are the corresponding human com-plement components.


http://jcp.bmj.com

/J C


ctober 1977. Dow

nloaded from

http://jcp.bmj.com/

Surface marker and other characteristics of Gaucher's cells

,257k.4K*4.

1-,2!

Fig. 1 EM ofGaucher spleen ( x 8500). A matureplasma cell (P) is seen to be intimately associated with two typicalGaucher's cells (G).

983

t.

,.

ik 9,--

Nr


http://jcp.bmj.com

/J C


ctober 1977. Dow

nloaded from

http://jcp.bmj.com/

G. F. Burns et al.

4

Fig. 2 Fc-rosettes preparation ofsplenic Gaucher's cell suspension. The ability ofthe Gaucher cell toform Fc-rosettesis well shown at the level ofboth the light (inset) and electron microscope. In the electron micrograph ( x 12 000) theGaucher cell displays the imperfect fixation not infrequently observed in rosette preparations: this ispresumably attributableto handling beforefixation and was particularly marked in the case ofthe Gaucher cells.

The cytocentrifuge preparation (inset) shows the strong Fc-rosette typical ofthe intact Gaucher cells. The associatedplasmacytoid cell on the left is also clearly seen to beforming an Fc-rosette.

ELECTRON MICROSCOPY (EM)Bulk preparations of rosettes were formed byallowing 1 ml of the splenic cell suspension at 2 x106/ml to react with 1 ml indicator erythrocytes.After centrifugation, the pellets were fixed in 3%glutaraldehyde and processed for EM by routinemethods (Cawley and Hayhoe, 1973).

Results

CYTOLOGY AND CYTOCHEMISTRYBoth by light and electron microscopy, the largestorage cells contained the striated inclusions whichcharacterise Gaucher's disease (Figs 1 and 2). Thecytochemistry was also typical of Gaucher's diseasein showing strong PAS, acid phosphatase, andf3-naphthyl butyrate esterase activity (Fig. 3a),which was partially inhibited by sodium fluoride.

The cells showed weaker alkaline phosphatase, iron,and naphthol AS-D chloroacetate esterase reactivityand were peroxidase and Sudan black negative.Under polarised light the striatedinclusionsshowed

the form birefringence associated with some lipids(Bennett, 1950); in contrast, we have noted that theinclusions of sea-blue histiocytosis are not bire-fringent (R. J. Flemans; unpublished observation).Under UV illumination, the Gaucher's cells did notautofluoresce, again in contrast to both the storagecells of sea-blue histiocytosis and many bonemarrow reticuloendothelial cells in a variety of non-storage conditions which fluoresce a strong pink(R. J. Flemans; unpublished observations).

ROSETTE TESTSThe results of the various splenic-cell rosette testsare summarised in Table 1. It is seen that a high

984


http://jcp.bmj.com

/J C


ctober 1977. Dow

nloaded from

http://jcp.bmj.com/


Fig. 3 Gaucher's cells. (a) a-naphthyl butyrate esterase preparation. The strong positivity ofthe Gaucher's cell is wellshown. (b) E-rosette preparation; phase contrast, illumination. This positive rosette is typical in showing relativelyfewadherent sheep RBCs. (c) Fc-rosetteformation; phase contrast illumination. Thiis type ofstrong rosette with manyadherent cells is typical. (d) Mouse erythrocyte-rosette formation; phase contrast illumination. The moderate number ofadherent mouse RBCs in this example is typical. As is discussed in the text, this type ofrosette, although definitelypositive, tended to be less strong than the Fc rosette.

percentage of Gaucher's cells formed rosettes withunsensitised sheep (E) RBCs (Fig. 3b), and with oxindicator cells coated with either the Fc of IgG(Figs 2 and 3c) or with human C3 (both C3b andC3d). A high percentage of the Gaucher's cells alsoformed rosettes with unsensitised mouse (M)erythrocytes (Fig. 3d), while a much lower per-centage formed rosettes with ox indicator cellscoated with mouse C3d. No rosette formation wasseen with C3bmo-coated indicator cells. The Fccontrol (unsensitised ox RBCs) and the C3 control(EAIgM) were also completely negative.As is illustrated in Fig. 3 (b-d) the number of

adherent indicator cells varied in the differentrosette tests. Thus, in the Fc and C3hu (not illus-trated) rosettes, the large number of adherent RBCsprobably indicates high receptor density. In contrast,most Gaucher's cells which formed rosettes withsheep RBCs had only four to five adherent RBCs,although a number had more than 10 adherent cells.

The mouse rosettes were of intermediate appearance,and this may indicate intermediate receptor density.It should be stressed that only viable cells were scoredin all these rosette tests.The percentage total spleen cells forming E-

rosettes in the Gaucher's preparation was remark-ably low: indeed, none of the many small lympho-cytes present was seen to form E-rosettes. This wasin marked contrast to the findings in the two otherspleens examined (Table 1).The results of our studies in the T- and B-cell

content of peripheral blood are summarised inTable 2. In both the patient and her sister, thepercentage of peripheral blood T-cells appearedmarkedly reduced, whether estimated by E-rosetteformation or by an esterase method recentlydescribed in this laboratory (Higgy et al., 1977).

In both the whole spleen tissue and the spleniccell suspension, the Gaucher's cells were frequentlyfound to be closely associated with one or more

985


http://jcp.bmj.com

/J C


ctober 1977. Dow

nloaded from

http://jcp.bmj.com/

G. F. Burns et al.

Table I Rosetting properties of splenic cell suspensions

Rosette test % Rosette formation by viable splenicleucocytes

Gaucher's disease 'Normal' HS

Total Gaucher'scells cells

E 5 54 51 45Fc 71 59 34 35EAlgM 0 0 5 4

C3bmo 0 0 ND NDC3moC

C3dmo 32 8 ND IC3bhu 68 60 ND ND

C3huC3dhu 58 34 ND ND

M 7 53 2 ND

ND = not done

Table 2 Peripheral blood rosetting data

Patient % Rosetteformation Esterase stainingby mononuclear cells

E Fc T-cell B-cell

Propositus 5 dayspost-splenectomy 24 44 26 71

Propositus 35 dayspost-splenectomy 40 58 45 34

Sister with Gaucher'sdisease 53 45 54 35

Normal 7-year-old girl 71 24 ND ND

ND = not done

plasmacytoid cells (Figs 1 and 2 inset). Theseassociated plasmacytoid cells formed strong Fc(Fig. 2 inset), and mouse, but no C3 or E rosettes.

SURFACE IMMUNOGLOBULIN

Whereas some small lymphocytes in the splenic cellpreparation were positive for Smig, both theGaucher's cells and the associated plasmacytoidcells were uniformly negative for SmIg.

Discussion

The morphology of Gaucher's cells is so distortedby the accumulated glucocerebroside that firmconclusions about the basic nature of the cell are notpossible at the level of either the light or electronmicroscopes. In addition, unlike typical marrow

histiocytes, the Gaucher's cells do not autofluoresceunder UV light. However, our cytochemical studies,like those of others (Rozenszajn and Efrati, 1961),are consistent with the assumed monocytic origin ofthese cells; in this regard, the a-naphthyl butyrateesterase positivity, inhibited by sodium fluoride, isparticularly suggestive of a monocytic origin.Our demonstration of strong Fc-rosette formation

by Gaucher's cells, in the absence of Smlg, is also

in keeping with a monocyte/histiocyte nature. Thepresence of a high density of receptor for human C3(both C3b and C3d) is again in accord with amonocytic/histiocytic origin of the Gaucher cell(Ross and Polley, 1975). The failure to form C3bmorosettes, and only a low percentage of rosetteformation with C3dmo, is difficult to interpret butmay reflect either a lower density of C3 products onthose indicator cells or the ability of the Gaucher'scells to distinguish between human and mouse C3.The ability of Gaucher's cells to form rosettes withsheep erythrocytes is also difficult to interpret.Normally, E-rosettes formed by T-lymphocytes ofless than 10 ,t diameter consist of a tight morulamade up of several layers of indicator cells. Incontrast, the much larger Gaucher's cells (> 40 ILdiameter) usually displayed only relatively fewadherent cells sparsely distributed over the cellsurface. Although it is possible that Gaucher's cellspossess a receptor similar to that of T-lymphocytes,in view of the other characteristics of the cells, it isdifficult to imagine that the formation of rosetteswith sheep RBCs is a true expression of a T-lympho-cyte phenotype. Histiocytes are known to be capableof recognising and removing effete erythrocytes, andthe rosettes seen with sheep RBCs may be related tothis process. Some support is lent to this inter-pretation by our observation of weak E-res-tteformation by the storage cells of a single case ofsea-blue histiocytosis (Burns, unpublished obser-vations). Because the mouse rosette test is generallyconsidered to identify specifically a subpopulationof B-lymphocytes (Forbes and Zalewski, 1976) andthe leukaemic cells in CLL (Stathopoulos andElliott, 1974), the reactivity of the Gaucher's cellmay also be attributed to the adherence of effeteRBCs. However, in our hands, some normal mono-cytes and neutrophils form M-rosettes, and thus theability of Gaucher's cells to form these rosettes maybe fully compatible with a monocyte-histiocytelineage.

In view of the immunological abnormalities thatmay be associated with Gaucher's disease (Pratt etal., 1968), the severe deficiency of splenic T-cellsdemonstrated by the strikingly low percentage ofE-rosette formation by the numerous lymphocytespresent in the patient's spleen is of interest. Thisdeficiency is clearly not attributable to samplingerror since our splenic suspension was prepared fromseveral pieces of tissue from widely different areasof the spleen. This unexpected finding prompted usto examine the T-cell content of the patient'speripheral blood. Interestingly, both the patient andher sister displayed a definite reduction in thepercentage of circulating T-cells. These resultscannot be simply attributed to a defect in our

986


http://jcp.bmj.com

/J C


ctober 1977. Dow

nloaded from

http://jcp.bmj.com/


E-rosette technique because it yielded an entirelynormal result in a control child and because similarresults were obtained by a completely differentmethod employing esterase cytochemistry (Higgyet al., 1977). Furthermore, the detectionofT-lympho-penia in the sister argues against the finding in thispatient being simply a response to the operationand/or its attendant anaesthetic. The possibility thatthe T-lymphopenia represents a manifestation ofhypersplenism or an unrelated chance associationcannot, however, be excluded.The association between plasma cells and both

normal reticuloendothelial cells and Gaucher's cellsis well recognised, and the latter observation isconfirmed in the present study. However, since theassociated cells closely resembled mature plasmacells morphologically, and lacked SmIg, the presenceof an Fc receptor on their surface is an unusualfinding since plasma cells usually lack the Fcreceptor (McConnell, 1975). We are unable to offerany definitive explanation for this finding. However,it has been suggested that plasma cells activelysecreting immunoglobulin lack the Fc receptorwhile inactive plasma cells possess this marker(Ramasamy et al., 1974); the Gaucher-associatedplasma cells may therefore be inactive cells.To conclude, the present marker study largely

substantiates the assumed relationship betweenGaucher's cells and cells of the monocyte/histiocyteseries. However, some possible differences betweenGaucher's and monocytoid cells were observed, buttheir significance awaits further investigations ofboth more cases of Gaucher's disease and thesurface marker properties of the reticuloendothelialcells of normal spleen, for which no data arecurrently available. Moreover, the demonstrationof an apparent T-cell deficiency in the presentpatient's spleen and peripheral blood suggests thatexamination of T-cell function in further cases ofGaucher's disease should be undertaken.

It is a pleasure to acknowledge the help and coopera-tion of Professor R. W. Smithells and Mr J. M.Beck, of Leeds General Infirmary, and Dr R. J. M.Bell, of Scunthorpe General Hospital, all of whomassisted us in studying this family under their care.We are grateful to Mr G. Binns for technicalassistance.GFB and KEH gratefully acknowledge support

from the Medical Research Council (UK) and thegovernment of the United Arab Republic of Egyptrespectively.

References

Barker, C. R., Bums, G. F., Cawley, J. C., and Hayhoe,

F. G. J. (1976). IgM receptors on the surface of hairycells of leukaemic reticuloendotheliosis (Letter).Lancet, 1, 1303.

Bennett, H. S. (1950). The microscopical investigation ofbiological materials with polarized light. In McClung'sHandbook of Microscopical Technique, 3rd edition,edited by R. McClung Jones, pp. 591-677. Hoeber,New York.

Boyum, A. (1968). A one-stage procedure for isolationof granulocytes and lymphocytes from human blood.Scandinavian Journal of Clinical and Laboratory Inves-tigation, 21, Supplement 97, 51-76.

Brady, R. 0. (1972). Biochemical and metabolic basis offamilial sphingolipidoses. Seminars in Haematology,9,273-284.

Burns, G. F., Cawley, J. C, Barker, C. R., Goldstone,A. H., and Hayhoe, F. G. J. (1977). New evidence re-lating to the nature and origin of the hairy cell ofleukaemic reticuloendotheliosis. British Journal ofHaematology, 36, 71-84.

Cawley, J. C. and Hayhoe, F. G. J. (1973). UltrastructureofHaemic Cells. W. B. Saunders, London.

Celada, F. and Rotman, B. (1967). A fluorochromatictest for immunocytotoxicity against tumor cells andleucocytes in agarose plates. Proceedings ofthe NationalAcademy of Sciences, 57, 630-636.

Fisher, E. R. and Reidbord, H. (1962). Gaucher's disease:pathogenetic considerations based on electron micro-scopic and histochemical observations. AmericanJournalofPathology, 41, 679-691.

Forbes, I. J. and Zalewski, P. D. (1976). A subpopulationof human B lymphocytes that rosette with mouseerythrocytes. Clinical and Experimental Immunology,26,99-107.

Gupta, S., Good, R. A., and Siegal, F. P. (1976). Rosette-formation with mouse erythrocytes. II. A marker forhuman B and non-T lymphocytes. Clinical and Experi-mental Immunology, 25, 319-327.

Hallberg, T., Gurner, B. W., and Coombs, R. R. A.(1973). Opsonic adherence of sensitized ox red cells tohuman lymphocytes as measured by rosette formation.International Archives ofAllergy, 44, 500-513.

Higgy, K. E., Burns, G. F., and Hayhoe, F. G. J. (1977).Discrimination of B, T and null lymphocytes byesterase cytochemistry. Scandinavian Journal ofHaematology, 18, 437-448.

Kaplan, M. E. and Clark, C. (1974). An improvedrosetting assay for detection of human T lymphocytes.Journal ofImmunological Methods, 5, 131-135.

Lachmann, P. J., Hobart, M. H., and Aston, W. P.(1973). Complement technology. In Handbook ofExperimental Immunology, 2nd edition, edited byD. M. Weir. Blackwell, Oxford.

Lee, R. E., Balcerzak, S. P., and Westerman, M. P.(1967). Gaucher's disease. A morphologic study andmeasurements of iron metabolism. American Journal ofMedicine, 42, 891-898.

McConnell, I. (1975). T and B lymphocytes. In TheImmune System, edited by M. J. Hobart and I. Mc-Connell, p. 98. Blackwell, Oxford.

McConnell, I. and Hurd, C. M. (1976). Lymphocytereceptors. It. Receptors for rabbit IgM on human T

987


http://jcp.bmj.com

/J C


ctober 1977. Dow

nloaded from

http://jcp.bmj.com/

G. F. Burns et al.

lymphocytes. Immunology, 30, 835-839.Pennelli, N., Scaravilli, F., and Zacchello, F. (1969).Th- morphogenesis of Gaucher cells investigated byelectron microscopy. Blood, 34, 331-347.

Pratt, P. W., Estren, S., and Kochwa, S. (1968). Immuno-globulin abnormalities in Gaucher's disease. Reportof 16 cases. Blood, 31, 633-640.

Ramasamy, R. (1974). A fluorescent stain for viablerosette-forming cells. JoiurnalofImmunological Methods,5, 305-306.

Ramasamy, R., Munro, A., and Milstein, C. (1974).Possible role for the Fc receptor on B lymphocytes.Nature, 249, 573-574.

Roos, B., Cottier, H., and Rossi, E. (1961). Electron-optical observations in storage diseases (Abstract).Experientia, 17, 430.

Ross, G. D., Rabellino, E. M., Polley, M. J., and Grey,H. M. (1973). Combined studies of complement receptorand surface immunoglobulin-bearing cells and sheeperythrocyte rosette-forming cells in normal andleukemic human lymphocytes. Journal of ClinicalInvestigation, 52, 377-385.

Ross, G. D. and Polley, M. J. (1975). Specificity ofhuman lymphocyte complement receptors. Journal ofExperimental Medicine, 141, 1163-1180.

Rozenszajn, L. and Efrati, P. (1961). Cytochemical andphase-contrast observations on Gaucher cells. ActaHaematologica, 25, 43-48.

Stathopoulos, G. and Elliott, E. V. (1974). Formation ofmouse or sheep red-blood-cell rosettes by lymphocytesfrom normal and leukaemic individuals. Lancet, 1,600-601.

988


http://jcp.bmj.com

/J C


ctober 1977. Dow

nloaded from

http://jcp.bmj.com/

Documents

Surface marker andother characteristics Gaucher's cells