ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 24, No. 6 Copyright © 1994, Institute for Clinical Science, Inc.

Flow Cytometric Staining of Pancreatic Endocrine Cells* LIBERTY WALKER, B.S. and CHARLES M. PETERSON, M.D. Sansurn Medical Research Foundation, Santa Barbara, CA 93105

ABSTRACT Few distinct surface markers have been identified for endocrine cells, but antibodies directed against hormonal epitopes are widely available. In this study, human fetal pancreatic (3 cells were identified on the flow cytometer by intracellular staining for insulin. A paraformaldehyde based permeabilization scheme was used which preserves surface staining and helps maintain cell integrity while allowing antibodies access to compart­ ments within the cell interior. Cells were first stained for surface antigens, fixed in paraformaldehyde, and then permeabilized in phosphate buffered saline containing 0.2 percent Tween. Intracellular staining was accom­ plished using a fluorescein conjugated anti-insulin monoclonal antibody. In this manner, major histocompatibility complex Class II antigen expres­ sion was documented on 3 cells exposed to interferon gamma. This study demonstrates the feasibility of using hormonal content to identify endo­ crine cells by flow cytometry while simultaneously staining for sur­ face antigens. Introduction Intracellular staining 1,2 permits the flow cytometric identification of cell types for which distinctive surface mark­ ers have not yet been identified. Endo­ crine cells are perfect candidates for such analysis because of their dense intracel­ lular hormonal content and because of available antibodies directed against hor­ monal epitopes .3 Surface antigen expres­ sion in pancreatic endocrine cells is of great interest to those searching for the autoimmune trigger in Type I diabetes .4

However, dual intracellular surface stain­ ing of endocrine cells has until now been limited to microscopic techniques where quantitative analysis is extremely labor intensive and less precise than flow cyto­ metric (FACS) analysis. Here simulta­ neous insulin and major histocompati­ bility com plex (MHC) stain ing is reported in cells isolated from human fetal pancreas using a modification of the procedure of Schmid et al.5 Materials and Methods Rea g en ts

and

A n t ib o d ie s

* Address reprint requests to: Charles M. Peter­ Phycoerythrin conjugated anti-human son, M.D., Sansum Medical Research Foundation, leukocyte antigen DR (HLA-DR) mono2219 Bath Street, Santa Barbara, CA 93105.

501

0091-7370/94/1100-0501 $01.20 © Institute for Clinical Science, Inc.

502

WALKER AND PETERSON

5 0

50

CM P a n c r e a s 3.3%

0 10

J

0

im r j-

10

“

10

10-

IFN Pancreas 1.9%

10*

10

“

iqä

10

10

-

Insulin RIA CM: 1.8 uU/cell IFN: 1.1 uU/cell

H

10

Insulin Insulin Figure 1. Insulin staining of human fetal pancreatic cells. Pancreas was digested and cultured 24 hours with 1000 U/ml IFN (IFN) or in media alone (CM). Precisely 0.5 x 106 cells were permeabilized and

stained with anti-insulin-FITC or analysed for insulin content by RIA. An IFN associated decline in the percent of insulin staining cells (CM/IFN = 1.7) can also be documented in the insulin per cell extracted (CM/IFN = 1.6).

clonal antibody phycoerythrin conju­ gated anti-HLA-DR monoclonal antibody (anti-HLA-DR-PE);* anti-insulin mono­ clonal antibody (mAb);t mouse IgG lt and IgG2§ isotype controls; polyclonal guinea pig 11or sheepH anti-insulin antibod­ ies; and a fluorescein isothiocynate (FITC) conjugation kit** were all purchased. The anti-insulin mAb was fluorescein conjugated according to the manufactur­ er’s instructions. The conjugation reac­ tion was performed at several fluorescein to protein (F/P) ratios ranging from 12.5 to 100 , resulting in antibody-fluorescein conjugates w ith a bound F/P ratio between 0.7 and 5.8. All conjugates were titered for optimal internal staining in the range of 0.25 to 4 mg per 5 x 105 cells. Precisely 0.5 |xg of the 1.3 F/P conjugate

gave good peak separation (figures 1 and ) and minimal interaction with the dual stain phycoerythrin (PE) antibody. Sodium azide, bovine serum albumin (BSA), paraformaldehyde, Tween-20™ and Histopaque™ were all purchased.tt One percent paraformaldehyde was fil­ tered before use and prepared fresh daily by shaking in phosphate buffered saline (PBS) for 60 minutes at 56°C. Solid phase insulin radioimmunoassay (RIA) kitsti were purchased and used according to the manufacturer’s protocol. 2

Cell Culture

Human fetal pancreatic specimens, 16 to 22 weeks gestation, were supplied§§ and were cultured as described else­ where .6 Briefly, specimens were shipped overnight on ice in sterile Roswell Park Memorial Institute (RPMI) 1640 supple­ * Beckton Dickinson Immunocytometry, 2350 m ented w ith 100 U/ml p en icillin / Qume Dr., San Jose, CA 95131. t AMAC, Inc. 160 C Larrabee Rd., Westbrook, streptomycin, 2.5 |xg/ml Amphotericin B ME 04092. t Olympus Corp., 4 Nevada Dr., Lake Success, and 10 percent fetal calf serum (FCS). NY 11042. § Beckton Dickinson Immunocytometry, 2350 Qume Dr., San Jose, CA 95131. 1DAKO Co. 6392 Via Real Carpinteria, CA 93013. 11Binding Site Limited, 97 Vincent Dr., Edgbaston, Birmingham B15225Q, England. ** Boehringer Mannheim Co., 9115 Hague Rd, Indianapolis, IN 46250.

t t Sigma Chemical Co., PO Box 14508, St. Louis, MO 63178. t t Diagnostic Products Co., 5700 West 96th St, Los Angeles, CA 90045. §§ Advanced Bioscience Resources, 1516 Oak St #309 Alameda, CA 94501.

FLOW CYTOMETRIC STAINING OF PANCREATIC ENDOCRINE CELLS

503

FIGURE 2. HLA-DR or insulin staining in human fetal pancreatic cells cultured 48 hours in media alone (A) or 1000 U/ml IFN (B). Insulin staining in fresh PBL (C)._____ isotype control,..........prior to permeabilization after permeabilization.

The tissue was minced in sterile Hank’s balanced salt solution (HBSS) containing 1 percent bovine serum albumin (BSA) and digested in 1.5 mg/ml collagenase XI for ten minutes at 37°C with vigorous shaking. The resulting islet like cell clus­ ters (ILCC) were washed three times in HBSS and cultured 24 to 48 hours either in complete media (CM), that is RPMI supplemented with antibiotics and 10 percent human adult serum, or in CM containing 1000 U/ml recom binant human interferon-gamma (IFN). T r y p t ic D

ig e s t

After culture, islet-like cell clusters (ILCC) were dissociated as described earlier.6 Briefly, ILCC were digested in 0.05 percent trypsin for 2 minutes at 37°C and dispersed by repeated pipetting. The resulting cell suspension was washed three times in RPMI containing 10 per­ cent FCS and resuspended in pH 7.2 PBS containing 0.1 percent sodium azide and 2 percent FCS (PBS-FCS). P e r ip h

eral

Blo

od

L

ym phocytes

H e p a rin iz e d w h o le b lo o d was obtained by venipuncture from adult vol­ unteers following informed consent. Peripheral blood lymphocytes (PBL)

were isolated by Histopaque density cen­ trifugation according to the manufactur­ er’s instructions. STAINING

For surface staining, 0.2 to 0.5 x 106 cells in 50 ml PBS-FCS were incubated for 30 minutes at 4°C with 0.25 mg of either anti-HLA-DR-PE antibody or IgG2-PE isotype control. Samples were gently washed first in PBS-FCS and then in PBS alone (400 X g for 3 minutes). Cells were then fixed for 60 minutes at room temperature in 500 ml PBS contain­ ing 2 percent paraformaldehyde. After a gentle wash (200 x g for 5 minutes) in PBS-FCS, cells were permeabilized for 15 minutes at 37°C in 0.5 ml PBS contain­ ing 0.2 percent Tween. After dilution to 1 ml with PBS and a gentle spin, the cells were incubated for 40 minutes at 4°C in a final volume of 70 ml with 0.5 mg of either anti-insulin-FITC or IgGl-FITC isotype control. Samples were gently washed once in PBS containing 0.1 per­ cent Tween and then in PBS-FCS. They were resuspended in a final volume of 300 ml and analyzed within 1 hour of staining. Aliquots of 0.5 x 106 cells reserved for insulin analysis w ere extracted overnight at 4°C in 100 ml acid ethanol and diluted 1:100, 1:200 and 1:

WALKER AND PETERSON

504

400 in PBS con tain in g 1 p ercen t color scheme where labelling of surface BSA prior to insulin radioimmunoanaly- antigens with mouse antibodies was per­ formed prior to permeabilization and sis (RIA). cytoplasmic staining. Fixations were carried out over a range F C of temperatures and concentrations to optimal conditions for maxi­ Ten thousand events per sample were ascertain mum intracellular and surface staining. A acquired and analyzed on a Becton Dick­ 0.25 percent paraformaldehyde inson FACScan flow cytometer using a yielded dim insulin staining whenfixation com­ 488 nm argon ion laser and Consort 30 pared to 2 percent fixation at an elevated software. The FITC emissions (500 to (37°C). The HLA-DR stain­ 560 nm) and PE emissions (543 to 627 temperature ing, on the other was reduced dra­ nm) were collected on a 4 decade log matically by 37°Chand, fixation when com­ scale. Forward and side scatter signals pared to 4°C (table I, experiment 1). An were collected on a linear scale. The for­ interm ediate fixation tem perature of ward scatter threshold was adjusted dur­ 22°C maintained insulin staining without ing acquisition to exclude debris and impairing HLA-DR staining (table I, scatter gates were employed during analy­ experiment 2). Finally, elevated Tween sis to select particular subsets of cells. concentrations up to 1 percent were not associated with enhanced insulin stain­ Results ing (data not shown). To confirm that two color events were Up to 6 percent of permeabilized fetal not clumps of cells or doublets, stained pancreatic cells stained positively for cells were filtered through 30 mm mesh insulin after 48 hours in culture. Intra­ prior to analysis. Seventy-five percent assay variation was less than 5 percent. of the doubly stained events remained Insulin staining was consistent with insu­ after filtration indicating minimal clump­ lin RIA. Both RIA and staining methods ing artifact. documented a relative decrease in insu­ The IFN induced substantial increases lin content following interferon treat­ in MHC Class II expression in the pan­ m ent (figure 1). In su lin stain ing creatic digests while simultaneously appeared in pancreatic cells only after decreasing insulin expression. The per­ permeabilization and did not appear in centage of DR bearing cells was signifi­ perm eabilized PBL (figure 2). One cantly increased after 48 hours exposure hundred percent permeabilization was to IFN when compared to media controls documented in all cell types by trypan (3 ± 0.9 percent vs 39 ± 9.0 percent, p < blue uptake. 0.007) while insulin containing cells Polyclonal insulin antisera suitable for were significantly decreased (5 ± 1.2 per­ histochemistry exhibited excessive non­ cent vs 2 ± 1.1 percent, p < 0.002). A specific binding in our FACS samples, representative experiment (n = 5) is even if cells were pre-incubated in non- shown in figure 3. immune serum. On the other hand, indi­ As seen in figure 2B, permeabilization rect insulin staining of permeabilized does not affect a sample’s HLA-DR pro­ pancreatic cells with mouse mAb gave file. Moreover, other cell types do not good resolution histograms (data not pick up insulin antibody non-specifically shown). However, such indirect cytoplas­ after permeabilization (figure 2C). Con­ mic staining was not suitable in a two sequently, serial anti-HLA-DR and anti­ low

ytom etry

FLOW CYTOMETRIC STAINING OF PANCREATIC ENDOCRINE CELLS

505

TABLE I Optimizing Fixative Temperature and Concentration

Fixation Temperature Paraformaldehyde Concentration

4° C 2%

0.25%

22° C 2%

37° C 2%

37° C 0.25%

EXPERIMENT 1 % Insulin MF Insulin % HLA-DR % MF DR

3.2 436 14.2 392

4.6 348 20.4 142

% Insulin M F Insulin % H LA-DR MF DR

ND ND 5.4 393

ND ND ND ND

ND

ND ND ND

6.2 1207 5.9 509

4.5 448 8.0 757

EXPERIMENT II 4.8 397 5.4 397

5.0 538 3.9 269

ND ND ND ND

Human fetal pancreatic cells were cultured 24 hours in 1,000 U/ml of human interferon gamma. Cells were surface stained with anti H L A -D R -P E (% DR), fixed, permeabilized, and stained with anti-insulin-fluorescein isothiocyanate (% insulin). Mean fluorescence (MF) is the arithmetic mean of the peak fluorescence intensity for a given stain. ND is not done.

insulin staining by the method described now be identified based on hormonal allows accurate visualization of Class II content and simultaneously stained for bearing (3 cells (figure 3). surface markers as well. Many methods exist to permeabilize cells for cytoplasmic or nuclear flow cyto­ Discussion metric staining. Membrane disruption can be accom plished w ith d e te r­ Insulin, an intracellular antigen, has gents,7,8,9,23 with hypotonic solutions,34 been used to identify P cells by flow sonication,35 or even electroporation cytometry in a heterogeneous pancreatic which is sufficiently gentle to allow the cell digest. Moreover, MHC Class II recovery of viable cells after staining.36 expression has been documented in (3 Fixation techniques using alcohols1,13 cells exposed to IFN. Cytoplasmic mark­ have largely been supplanted by meth­ ers to date have been used in flow cytome­ ods using paraformaldehyde in combina­ try to identify cellular subsets,7’8’9,10 tion with detergent.3,9,15,22,30,32,33,37 stages of cellular differentiation,11,12,13 Paraformaldehyde tends to stabilize and gene products14,15,is,17,18,19,20,21,22,23 cells in the presence of permeabilizing or metabolic changes24,25,26,27,28 present agents by cross linking amino groups on during cellular activation. Moreover, target proteins.30,37 The loss of cellular DNA staining or cell cycle analysis has contents is minimized during subsequent often been performed simultaneously washings32 without impeding antibody with cytoplasmic staining to further char­ access to the cytoplasm or decreasing acterize cellular activation in vivo29 or in immunoreactivity of antigenic epitopes.5 vitro.1'5'30’31'32’33 With the method pre­ H ow ever, stain ed cells fixed in sented here, uncloned target cells in paraformaldehyde are no longer viable, a endocrine autoimmune disorders can distinct disadvantage when compared to

WALKER AND PETERSON

506

PiQ 1-1 S3

iau

10*

IQ

10-

1GT

IQ

IQ

1QJ

iQ

In s u lin

F igure 3. Simultaneous HLA-DR and insulin staining in human fetal pancreatic cells cultured 48 hours in 1000 U/ml IFN (B) or in media alone (A).

methods which use autofluorescence,38 References passive dye uptake,39 or surface markers Jacobberger WJ. Cell cycle expression of to identify intact cells.40 Zinc specific 1. nuclear proteins. In: Yen A, ed. Flow Cytome­ dyes allow live P cell labelling in hetero­ try: Advanced Research and Clinical Applica­ tions, vol. 1. Boca Raton: CRC Press, 1989:305geneous cell preparations,41 however, in 26. cell types for which monoclonal antibod­ 2. Turner JR, Tartakoff AM, Berger M. A flow ies are still the most effective means of cytometric method for the quantitative analysis of intracellular and surface membrane antigens. intracellular staining, some permeabiliIn: Tartakoff AM, ed. Methods in Cell Biology zation is required. The method presented 32, Part B: Vesicular Transport. New York: Aca­ here is considerably faster than methods demic Press, 1989:351—63. used in the past to stain endocrine cells 3. Hatfield JM, Hymer WC. Flow cytometric immunofluorescence of rat anterior pituitary for flow cytometry.3 cells. Cytometry 1985;6:137-42. In our hands, surface staining followed 4. Castano L, Eisenbarth GE. Type I diabetes: a chronic autoimmune disease of human, mouse by perm eabilization and then direct and rat. Ann Rev Immunol 1990;8:647-79. intracellular staining gave the lowest 5. Schmid I, Uittenbogaart C, Giorgi J. A gentle level of non-specific binding and double fixation and permeabilization method for com­ bined cell surface and intracellular staining staining artifact. Surface and intracellular with improved precision in DNA quantifica­ staining are both maximized in this rapid tion. Cytometry 1991;12:279-85. and technically simple method to which 6 . Ruhland B, Walker L, Wollitzer A, Peterson C. Ethanol influences Class I and Class II MHC other cell types should also be amenable. Acknowledgments This work was supported in part by AMAC Cor­ poration, Westbrook, ME and by the Diabetes Research and Education Foundation.

antigen expression on human fetal islet like cell clusters. Alcohol Clin Exp Res 1991;15:745-7. 7. Anderson P, Nagler-Anderson C, O’Brien C, Levine H, Watkins S, Slayter HS, Blue ML, Schlossmann SF. A monoclonal antibody reac­ tive with a 15 kDa cytoplasmic granule associ­ ated protein defines a subpopulation of CD 8 + T lymphocytes. J Immunol 1990;144:574-82.

FLOW CYTOMETRIC STAINING OF PANCREATIC ENDOCRINE CELLS 8.

9. 10.

11. 12. 13. 14.

15. 16.

17.

18.

19.

Bardales RH, Al-Katib AM, Carrato A, Koziner B. D etection of intracytoplasm ic immuno­ globulin by flow cytometry in (3 cell malignan­ cies. J Histochem Cytochem 1989;37:83-9. Becker RS, Weaver CW, Knight KL. Cytoplas­ mic immunofluorescence and light scatter anal­ ysis of lamina propria plasma cells by flow cytometry. J Immunol Methods 1986;89:159— 64. Slaper-Cortenbach IC, Admiral LG, Kerr JM, van Leeuwen EF, von dem Borne AE, Tetteroo PA. Flow-cytometric detection of terminal deoxynucleotidyl transferase and other intra­ cellular antigens in combination with mem­ brane antigens in acute lymphatic leukemias. Blood 1988;72:1639-44. Kastan MB, Schlaffer E, Russo JE, Calvin DM, Civin Cl, Hilton J. Direct demonstration of ele­ vated aldehyde dehydrogenase in human hematopoietic progenitor cells. Blood 1990;75: 1947-50. Leung MF, Sokoloski JA, Sartorelli AC. Changes in microtubules, microtubule-associated pro­ teins, and intermediate filaments during the differentiation of HL-60 leukemia cells. Cancer Research 1992;52:949-54. Levitt D, King M. Methanol fixation permits flow cytometric analysis of immunofluorescent stained intracellular antigens. J Imm unol Methods 1987;96:233-7. Caruso A, Terlenghi L, Scalzini A, Verardi R, Foresti I, Pollara P, Bonfanti C, Ravizzola G, Manca N, Turano A. Evaluation of the expres­ sion of IFN-gamma in lymphocytes using a monoclonal antibody and flow cytometry. J Immunol Methods 1988;113:37-43. Chikanza IC, Corrigal V, Kingsley G, Panayi GS. Enumeration of interleukin-1 alpha and beta producing cells by flow cytometry. J Immunol Methods 1992;154:173-8. De Caestecker MP, Telfer BA, Hutchinson IV, Ballardie FW. The detection of intracytoplas­ mic interleukin -1 alpha, interleukin -1 beta and tumour necrosis factor alpha expression in human monocytes using two colour immuno­ fluorescence flow cytometry. J Immunol Meth­ ods 1992;154:11-20. Gazitt Y, He YJ, Erdos GW, Chang L, Ashktorab H, Cohen RJ. Development of a two color immunofluorescence stain and immunolocalization method for N-myc and c-myc oncopro­ teins with a newly generated mouse IgM anti N-myc antibody. J Immunol Methods 1992;48: 159-69. Hayden GE, Walker KZ, Miller JF, Wither­ spoon JS, Raison RL. Simultaneous cytometric analysis for the expression of cytoplasmic and surface antigens in activated T cells. Cytometry 1988;9:44-51. Himmelfarb J, Hakim RM, Holbrook DG, Leeber DA, Ault KA. Detection of granulocyte reactive oxygen species formation in whole bloodusing flow cytometry. Cytometry 1992; 13:83-9.

507

20. Kataoka K, Naomoto Y, Kojima K, Horiki S, Hizuta A, Tanaka N, Yahagi N, Miyazaki Y, Okumura K, Orita K. Flow cytometric analysis on perforin induction in peripheral blood m ononuclear cells w ith interleu kin -2 or OK-432. J Immunotherapy 1992;11:249-56. 21. Lacal P, Pulido R, Sanchez-Madrid F, Mollinedo F. Intracellular location of T200 and Mol glycoproteins in human neutrophils. J Biol Chem 1988;263:9946-51. 22. Lundhal J, Hallden G, Hed J, Johansson SG. A flow cytometric method to measure the stimu­ lated mobilization and the intracellular pool of the adhesion promoting glucoprotein MAC-1. APMIS 1991;99:139-46. 23. Sumner H, Abraham D, Bori-Gharios G, PlaterZyberk C, Olsen I. Simultaneous measurement of cell surface and intracellular antigens by multiple flow cytometry. J Immunol Methods 1991;136:259-67. 24. Glab N, Petit PX, Slonimski PP. Mitochondrial dysfunction in yeast expressing the cytoplasmic male sterility T-urfl3 gene from maize: analysis at the population and individual cell level. Molecular Gen Genet 1993;236:299-308. 25. Hassall DG. Three probe flow cytometry of a human foam-cell forming macrophage. Cytome­ try 1992;13:381-8. 26. Hedley D, Chow S. Flow cytometric measure­ ment of lipid peroxidation in vital cells using parinaric acid. Cytometry 1992;13:686-92. 27. Janeczek AH, Van Alten PJ, Reyes HM, Walter RJ. Modulation of the cytoskeleton and intra­ cellular calcium in leukocytes exhibiting a can­ cer-associated chemotaxis defect. J Leuk Biol 1993;54:351-9. 28. Liddle RA, Misukonis MA, Pacy L, Balber AE. Cholecystokinin cells purified by fluorescenceactivated cell sorting respond to monitor pep­ tide with an increase in intracellular calcium. Proc Nat Acad Sci 1992;89:5147-51. 29. Chan CS, Wormsley SB, Peter JB, Schechter GP. Dual parameter analysis of myeloma cells by flow cytometry. DNA content of cells con­ taining monotypic cytoplasmic immunoglobu­ lin. Am J Clin Pathol 1989;91:12-7. 30. Dent GA, Leglise MC, Pryzwansky KB, Ross DW. Simultaneous paired analysis by flow cytometry of surface markers, cytoplasmic anti­ gens, or oncogene expression with DNA con­ tent. Cytometry 1989;10:192-8. 31. Gazitt Y, Erdos GW. Fluctuations and ultrastructural localization of oncoproteins and cell cycle regulatory proteins during growth and apoptosis of synchronized AGF cells. Cancer Res 1994;54:950-6. 32. Mann GJ, Dyne M, Musgrave EA. Immunofluo­ rescence quantification of ribonucleotide reductase M l subunit and correlation with DNA content by flow cytometry. Cytometry 1987;8:509-17. 33. Zocchi MR, Marelli F, Poggi A. Simultaneous cytofluorometric analysis for the expression of

508

34. 35. 36. 37.

WALKER AND PETERSON

cytoplasmic antigens and DNA content in CD3 human thymocytes. Cytometry 1990;11:883-7. Tate EH, Wilder ME, Cram LS, Wharton W. A method for staining 3T3 cell nuclei with propidium iodide in hypotonic solution. Cytome­ try 1983;4:211-5. Fechleimer M, Denny C, Murphy RF, Taylor DL. Measurement of cytoplasmic pH in Dictyostelium discoideum by using a new method for introducing macromolecules into living cells. Eur J Cell Biol 1986;40:242-7. Berglund DL, Starkey JR. Isolation of viable tumor cells following introduction of labelled antibody to an intracellular oncogene product using electroporation. J Immunol Methods 1989;125:79-87. Hallden G, Andersson U, Hed J, Johansson SG. A new membrane permeabilization method for the detection of intracellular antigens by flow

38. 39. 40.

41.

cytometry. J Immunol M ethods 1989; 124: 103-9. Pipeleers DG.: Islet Cell Purification. In: Lamer J, ed. Methods in Diabetes Research, New York: John Wiley and Sons, 1984:185-211. Weston SA, Parish CR. New fluorescent dyes for lymphocytic migration studies. Analysis by flow cytometry and fluorescence microscopy. J Immunol Methods 1990;133:87-97. De Krijger RR, Aanstoot HJ, Kranenburg G, Verkerk A, Jongkind JF, van Strik R, Lafferty KJ, Bruining GJ. Enrichment of beta cells from the human fetal pancreas by fluorescence acti­ vated cell sorting with a new monoclonal anti­ body. Diabetologia 1992;35:436-43. Jiao L, Gray DWR, Gohde W, Flynn GJ, Morris PJ. In vitro staining of islets of Langerhans for fluorescence activated cell sorting. Transplan­ tation 1991;52:450-2.

STATEMENT OF OWNERSHIP, MANAGEMENT, AND CIRCULATION (Act of October 23,1962, Section 4369, Title 39, United States Code) Date of Filing - September 26,1994 Tide of Publication - ANNALS OF CLINICAL AND LABORATORY SCIENCE Frequency of Issue - Bimonthly Location of Known Office of Publication - 301 South Eighth Street, Philadelphia, PA 19106-4014 Location of the Headquarters or General Business Offices of the Publisher - Same as above Publisher - Institute for Clinical Science, Inc. Editor - F. William Sundcrman, M.D., PhD. Managing Editor - Same as above Owner - Institute for Clinical Science, Inc. Known Bondholders, Mortgages, and Other Security Holders Owing or Holding 1 Percent or More of Total Amount of Bonds, Mortgages or Other Securities - None Average No. Copies Actual No. Copies Each Issue During of Single Issue Published Preceding 12Months NeaiesttoFilingDate A* Total No. Copies Printer (Net Press Run) 979 975 B. Paid Circulation 1. Sales Through Dealers and Carriers, Street Vendors and Counter Sales None None 2. Mail Subscriptions 874 868 C. Total Paid Circulation 874 868 D. Free Distribution by Mail, Carrier or Other Means, Samples, Complimentary, and Other Free Copies None None E. Total Distribution (Sum of C and D) 874 868 F. Copies not Distributed 1. Office Use, Left-over, Unaccounted Spoiled After Printing 105 107 2. Returns from News Agents None None G. Total (Sum ofE and F - should equal net press run shown in A) 979 975 I certify that the statements made by me are correct and complete - F. WILLIAM SUNDERMAN, MD., PhJ).