Immunoreactive lactoferrin in resting, activated, and neoplastic lymphocytes

Leukemia Research Vo|. 14, No. 5, pp. 441-447, 1990. Printed in Great Britain.

0145-2126/90 $3.00+ .00 Pergamon Press plc

IMMUNOREACTIVE LACTOFERRIN IN RESTING, ACTIVATED, AND NEOPLASTIC LYMPHOCYTES THOMAS W. BUTLER,ill CARLO E. GROSSI,:~§ II¶ ANDREA CANESSA,:~§ [1¶ VITO PISTOIA:~§ II¶ and JAMES C. BARTONt II** Departments of tMedicine, SPathology and §Microbiology, IIComprehensive Cancer Center and ¶Cell Identification Laboratory, University of Alabama at Birmingham, Birmingham, Alabama 35294, U.S.A. and ** Veterans Administration Medical Center, Birmingham, Alabama 35233, U.S.A.

(Received 19 December 1989. Revision accepted 23 December 1989) Abstract--Lactoferrin (Lf) in lymphocytes was assessed with immunofluorescence/flow cytometric technique. Surface Lf was detected primarily among B-cell-enriched preparations. Tonsillar B-cells of different densities expressed surface Lf similarly. Very small percentages of CALLA + ALL, HCL, or EBV-transformed B-cells expressed surface Lf, whereas B-CLL lymphocytes had the highest percentages of surface Lf positivity. Few resting, cultured, or neoplastic T-lymphocytes expressed Lf. The pattern of immunofluorescence and analyses of surface and total cellular immunoreactive Lf indicated that Lf is associated primarily with the lymphocyte surface. The percentage and/or intensity of surface Lf-specific fluorescence were not significantly altered in B- or T-cells by incubation with physiologic concentrations of diferric Lf, and the percentages of Lf-positive cells detected in respective subjects remained stable over time. Surface Lf positivity was unrelated to the expression of other surface antigens (except those marking B- or T-cell lineage) or cell cycle. Expression and/or binding of Lf in B-lymphocytes may become increased during certain stages of cell maturation.

Key words: Lf, lymphocyte, cell surface, B-cell, T-cell, B-CLL.

In accordance with these observations, human Lf can be bound by mouse peritoneal lymphocytes and by the cells of some cultured human and murine lymphoid cell lines [10, 11]. However, the subsets of lymphocytes which bind Lf are not well defined, nor is the function of Lf in these cells understood. The purpose of the present work is to define better the subpopulations of human lymphocytes which express Lf using flow cytometric techniques.

INTRODUCTION LACTOFER~IN(Lf), a 78,000 mol. wt iron-binding glyeoprotein, is produced by neutrophilic granulocytes and glands of external secretion [1]. After its release from granulocytes, Lf is bound by putative specific receptors on the surface of neutrophils [2, 3] and monocytes [4--6]. Small quantities of~Lf have also been detected in human E - blood lymphocytes and non-adherent MNC using radioimmunoassay and immunofluorescence/flow cytometry technique [4, 7-9]. Further, studies with radiolabeled Lf [4] and Lf-specific immunofluorescence staining [9] suggest that Lf is localized primarily on lymphocyte surfaces.

M A T E R I A L S AND M E T H O D S

Subjects All specimens were obtained after approval of the study protocol by the Human Use Committee of the Institutional Review Board of the University of Alabama at Birmingham. Blood removed by antecubital venipuncture was collected in acid citrate dextrose (hydrous dextrose, 24.5 g/l, hydrous sodium citrate, 22.0 g/l, and anhydrous citric acid, 7.3 g/l, used at a 16% v/v concentration) from 12 normal adult volunteers, patients with chronic B-cell leukemias (26 CLL, 4 HCL), three patients with untreated CALLA + ALL, and patients with T-cell leukemias (two Stzary cell leukemias with T-helper immunophenotypes, four large granular lymphocyte expansions with T-suppressor immunophenotypes). Tonsil was obtained from two normal subjects undergoing routine tonsillectomy and adenoidectomy. Cells were also obtained from EBV-transformed B-cell cultures (n = 3) and from T-cell cultures (n = 8)

Abbreviations: Lf , lactoferrin; CALLA +, common acute lymphoblastic leukemia antigen-positive; ALL, acute lymphoblastic leukemia; HCL, hairy cell leukemia; EBV, Epstein-Barr virus; B-CLL, B-cell chronic lymphocytic leukemia; E-, sheep erythrocyte rosette-negative; MNC, mononuclear ceils; E +, sheep erythrocyte rosette-positive; FITC-GAM-IgG2b, fluoresceinated goat anti-mouse IgG2b; HBSS, Hank's balanced salt solution; CNPF, channel number of peak fluorescence; PHBR, peak height-tobase ratio; S.E.M., standard error of the mean. Correspondence to: Dr Thomas W. Butler, East Carolina University School of Medicine, Section of Hematology/Oncology, Brody Building 3E-106, Greenville, North Carolina 27858-4354, U.S.A. 441

442

T.W. BUTLERel aL

from phytohemagglutinin-stimulated T-cells maintained with interleukin-2.

Cell isolation and characterization Fresh tonsil in cold 0.9% NaC1 was trimmed to remove extraneous tissue, and the lymphoid tissue was finely minced with a stainless steel razor blade. The fragments were aspirated repeatedly through an 18-gauge needle to form a cell suspension. Blood and tonsil MNC were isolated using Histopaque (Sigma Chemical Co., St Louis, MO) or Percoll (Pharmacia Fine Chemicals, Piscataway, NJ) density gradient sedimentation as previously described [6, 9] and depleted of monocytes by plastic adherence. E ÷ and E- lymphocytes in MNC suspensions were isolated by rosetting E ÷ cells with aminoethylisothiouronium bromide hydrobromide-treated sheep erythrocytes at 4°C for 1 h, separating the rosettes using Histopaque density gradient sedimentation, and removing erythrocytes by hypotonic lysis. The results of preliminary experiments using cell isolates derived with and without hypotonic lysis of erythrocytes indicated that surface Lf expression by lymphoid cells is not significantly affected by this maneouver. Some E cells were further fractionated with Percoll gradients (045%, 45-50%, 50-55%, 55---60%, and greater than 60%) as previously described [12, 13]. Cell morphology was assessed using Wright's/Giemsa staining. Preparations from most patients with B- and T-cell neoplasms consisted of greater than 90% neoplastic cells. All cell preparations had greater than 90% viability by Trypan blue exclusion technique.

Lf purification Human colostral Lf obtained commercially (Sigma) was dissolved in HBSS, pH 7.2, and further purified by Sephadex G-150 (Pharmacia) column chromatography. Human neutrophil Lf from normal donors was also purified using an isolation procedure that included sequential sodium chloride extraction, heparin-Sepharose affinity chromatography, and AcA44 gel filtration chromatography [14]. Coomassie blue staining of the gel after sodium dodecyl sulfate-polyacrylamide gradient gel electrophoresis of the reduced purified protein demonstrated two polypeptides of Mr 80,000 and 78,000. Lf from both sources was dialyzed against 1 mM deferoxamine mesylate (Desferal®; Ciba Pharmaceutical Co., Div. of Ciba-Geigy Corp., Summit, NJ) or 1 mM citrate, pH 2.0, to remove any bound iron, and saturated with iron using a calculated quantity of iron as ferric ammonium citrate in phosphate-buffered saline, pH7.2, with added sodium bicarbonate. Preliminary experiments indicated that Lf from both sources yielded similar results, and therefore colostral Lf was used for the remaining experiments.

Flow cytometry methods Analysis of Lf. Lymphocyte surface Lf was assessed using indirect immunofluorescence/flow cytometric technique as described in detail elsewhere [9, 15]. MoAb5B2, a mouse monoclonal IgG2b antibody against mature human neutrophil (and milk) Lf [14] generously provided by Dr Louis W. Heck, and FITC-GAM-IgGEb second layer antibody (Southern Biotechnology Associates, Birmingham, AL) were used in 1:20 dilutions. In some experiments, lymphocytes were incubated in diferric Lf, 1 × 10-10 M in HBSS/5 mg/dl of CaCI2, for 1 h at 4°C prior to assessment of surface Lf to determine whether the percentage of Lfpositive cells and/or Lf-specific fluorescence intensity could

be increased. This Lf concentration simulates that in normal plasma and avoids non-specific Lf polymerization and cell binding which occurs at higher concentrations [4, 16, 17]. All specimens were analyzed using a FACScan (Becton Dickinson, Mountain View, CA). Cytometer gates were set to include lymphocytes as indicated by cell size and 90° incident light scatter; markers were set to include less than 5% false-positive cells using control preparations stained with FITC-GAM-IgG2b only. Manual evaluation of preparations yielded results similar to those obtained by FACScan analysis for both surface (data not shown) and total cell [9] Lf preparations, and revealed a surface pattern of specific immunofluorescence in both types of preparations. Cellular DNA analysis. DNA was evaluated using an RNAase/propidium iodide/flow cytometric method described in detail elsewhere [18]. Briefly, 2 x 106 lymphocytes in 0.2-0.5ml of phosphate-buffered saline/0.1% sodium azide/5% fetal calf serum (GIBCO, Grand Island, NY) were fixed by the dropwise addition of an equal volume of 100% ethanol with constant Vortex mixing, and the preparation was incubated on ice for 30 min. After centrifugation of the suspension at 1500 x g and removal of the supernatant, the cells were treated with 0.5 ml of RNAase (1 mg/ml; Sigma), resuspended using a Vortex mixer, and incubated for 20 min at 37°C. Propidium iodide (0.5 ml of a 40 ixl/ml solution) was added and the cells were incubated at 20° C for 10 minutes. The solution was then passed through a #41 mesh filter, centrifuged at 1500 x g for 1 minute, and one-half of the supernatant was removed and discarded. The resulting cell suspension was then analyzed with the FACScan.

Lymphocyte surface antigen phenotype analysis Surface antigen characterizations of lymphocytes were performed using routine diagnostic methods described in detail elsewhere [19]. Antibodies against the the following antigens were utilized: CD2 (Leu5), CD3 (Leu4), CD4 (Leu3), CD10 (CALLA), CD19 (Leul2), CD14 (My4), CD16 (Leull), CD20 (Leu 16), and HLA-DR (Becton Dickinson, Mountain View, CA; Coulter, Hialeah, FL).

Statistical considerations Data respresenting three or more determinations are expressed as the mean +-- 1 S.E.M. For light microscopic evaluations of cell preparation, 200 cells were analyzed. For FACScan determination, 5000-10,000 cells were analyzed from each sample and the results were corroborated with a direct inspection of the specimens by fluorescence microscopy. As defined in detail elsewhere [9], the difference (A) in channel number of peak fluorescence (CNPF) in .FACS-analyzed specimens (A = CNPFcomrol cells-CNPFexperimentalcells) was defined as a measure of the experimental cell fluorescence intensity in comparison with the CNPF of similar lymphoid cells prepared, stained with FITC-GAM-IgG2b only, and evaluated concurrently. Leftward and rightward shifts of the frequency distribution curves were defined to have negative and positive values of A, respectively. We also defined the ratio of the peak height of a frequency distribution curve to its base width (PHBR) as a measure of homogeneity of Lf content among the cells [9]. Higher values of PHBR signify greater homogeneity; lower values indicate greater heterogeneity. Analyses were performed using (a) the Student's t-test to evaluate the significances of differences between unpaired data; (b) Tukey's test to evaluate of all comparisons among

Lactoferrin and lymphocytes 40

443 300 E-

09 ._1 .J

18.2"/. p o s i t i v e 30

LU O I.LI

>

I-

20

ofl..

.5

10 J i,,,..

""

'"'"1

,

i,...

200

0 ALL

B-CLL NORMAL E"

HCL

CLL 41.1"/. p o s i t i v e

EBV

L | :

40"

,~" ! '

0") ._1 ,_1

~p,,,V'i'

30

LU 0 U.I I--

r.,,,~

20

! ~. '

. . . . ."'i .i ' "

', , ,1%,,

200 0 . 3 % positive

oQ.

=,

" ""'"'l

J

10

o~

t --,0

' "

'

~

NORMALE+ T-LEUKEMIA CULTUREDT

FIG. 1. Percent surface Lf-positive cells in B-cell-enriched (upper) and T-cell-enriched (lower) isolates.

0

i

10 °

means; and (c) the F test to evaluate variances among multiple classification data [20]. Some comparisons were also made using the correlation coefficient (r). For all tests, values of p less than 0.05 were defined as statistically significant.

i,

"

i

101

IlIHI

F

1FI,..

10 2

I

T II,.ll

10 3

10 4

FIG. 2. Representative flow cytometry histograms depicting surface Lf reactivity in normal blood E- and E + lymphocyte fractions, and in CLL blood lymphocytes. 30

RESULTS 09

Lf positivity in blood mononuclear cells When monocyte-depleted peripheral blood MNC preparations from 12 normal subjects were examined for surface-associated Lf, we observed 6.6 ± 1.1% Lf-positive cells. To determine whether these positive cells could be attributed to a distinct subset(s) of lymphocytes, we analyzed cell preparations enriched for B- and T-cells, respectively. To ascertain whether the cells capable of binding Lf could correspond to a maturation or activation stage(s) of B- and T-cells, we also examined E--enriched fractions of different density derived from normal tonsil, and a variety of lymphoproliferative disorders in which the neoplastic cells are presumed to simulate arrested maturation of their normal counterparts.

Lf positivity in B-cell-enriched preparations Residual monocytes not removed by plastic adherence were excluded from further analysis of E - cell

,_1 .-I

LU O U..I

20

>_ 1.09

o

o.

10

0-45

45-50

50-55

55-60

>60

FRACTION, % P E R C O L L

FIG. 3. Percent surface of Lf-positive E- cells in Percoll fractions from tonsil obtained from two normal subjects.

preparations by gating on small, non-granular cells. Surface Lf was detected in approximately 16% (range 15.1%-18.2%) of normal blood E - lymphocytes (Figs 1 and 2). E - tonsil lymphocytes from two subjects.were fractionated by density gradients to yield

444

T.W.

BUTLER et al.

five subpopulations, and there was moderate variation in percentage Lf positivity among the subpopulations (Fig. 3). Approximately one-third of lymphocytes from patients with B-CLL had detectable surface Lf, although this value varied widely (range 0.7%-88.6%). In contrast, few Lf-positive blood lymphocytes were found in specimens from patients with HCL or CALLA + ALL. Cultured Bcells (from 3 EBV-transformed lines) had very few surface Lf-positive cells (Figs 1 and 2). We also performed two-color analysis of B-CLL cells to determine a possible correlation between the expression of the myeloid marker My4 (found on a subset of BCLLs) [21] and Lf binding. The results of these studies showed a lack of significant correlation between the two markers. In B-cells, the expression of multiple diagnostic markers and surface immunoreactive Lf by B-CLL lymphocytes did not reveal statistically significant correlations (data not shown).

5 =3 0

:37.

I'-ej )

d

100 80 60 40

iii 0 t.tJ

i-.

20

,#-, •

O3

ota

,

°.°°'

0

I

0

20

40

60

8,.,

100

% POSITIVE CELLS WITHOUT Lf INCUBATION Z

---Q 4 0 z

///:/

n"

y:--:7+0.gx. Z

o o

20

'0"

10

Lf positivity in T-cell enriched preparations In contrast to the results obtained with B-cells, few T-cells had surface Lf positivity (Figs 1 and 2). Tcell preparations included normal blood E + cells, chronic T-cell leukemias, and eight cultured lines derived from normal T-cells.

Effect of Lf incubation, constancy of Lf expression, and relationship to total cell Lf To determine whether incubation with Lf could enhance the percent positivity and/or the intensity of fluorescence in lymphoid cells, and to determine whether cultured cells growing without Lf could bind Lf, cell isolates (n = 71) were incubated without and with exogenous diferric Lf. Values of percent Lfpositive cells were very similar (r = 0.54, p < 0.001; F test value of p > 0.40), and the parameters PHBR and ,4 (measures of homogeneity and fluorescence intensity, respectively) did not reveal significant differences either (Fig. 4; Table 1). Likewise, when subgroups (e.g. normal blood MNC, B-CLL, etc.) were analyzed in the same manner, the mean values of percent Lf-positive cells in isolates incubated without and with exogenous diferric Lf did not differ significantly (data not shown; Tukey's t-test value of p > 0.30 in all comparisons). Taken together, these results suggest that the percentages of cells capable of expressing surface Lf are not increased with Lf incubation. In addition, the quantity of Lf per positive cell is not significantly increased after incubation with physiologic concentrations of diferric Lf. To determine the constancy of surface Lf expression, Lf positivity in blood lymphocytes was quantified on two occasions at least 3 months apart in six subjects (one normal E - cells, two B-CLL, one B-cell lympho-

W

o

0

•

0

i

10

•

i

20

•

I

,

30

i

40

% POSITIVE CELLS, FIRST DETERMINATION

FIO. 4. Relationships of percent surface Lf-positive cells to incubation without or with exogenous 1 x 1 0 -10 M diferric Lf (upper) and of the percent surface Lf-positive cells on two separate occasions more than 3 months apart in different subjects (lower).

cytic lymphoma, one HCL, and one large granular lymphocyte expansions with T-suppressor immunophenotype). In four of the six subjects, percentage Lf positivity remained relatively unchanged, but in two patients (one B-CLL and one B-cell lymphocytic lymphoma), lower values of percentage Lf positivity were observed at a time when their neoplasms became clinically more aggressive (Fig. 4). In aggregate, the corresponding determinations were not significantly different (r = 0.98, p < 0.001; value ofp in F test > 0.30) (Fig. 4). To evaluate the relationship of cell surface Lf to total cellular Lf, we quantified both values using adherence-depleted MNC from four normal subjects, two patients with B-CLL, and two patients with large granular lymphocyte expansions with T-suppressor immunophenotypes using a previously described flow cytometric assessment for total cell Lf [9] and the present surface Lf technique. Similar values for percent Lf positivity in the same patients were obtained (14.5 _ 7.6 versus 13.6 - 2.0, respectively; r = 0.73, p < 0.05; F test value of p > 0.30).

Lactoferrin and lymphocytes

445

TABLE 1. SURFACE IMMUNO LF REACTIVITY WITHOUT AND WITH INCUBATION WITH EXOGENOUS

Positive cells, % (range) PHBR (range) A (range)

Lf*

Without Lf incubation

With Lf incubation

Correlation

Differencet

18.6 + 5.0(0-88.6) 2.0 - 0.2(0.1-7.3) 9.1 - 1.2(-1-33)

14.6 - 2.3(0-82.7) 2.2 --- 0.3(0.1-6.8) 9.3 --+ 1.2(-1-34)

r = 0.54, p < 0.001 r = 0.66, p < 0.001 r = 0.99, p < 0.001

NS NS NS

* These data were determined using 71 lymphoid cell isolates: 12 normal blood MNC, 11 normal tonsil fractions (10 E-, 1 E+), 25 B-CLL, 3 cultured EBV-transformed B-cell lines, 4 HCL, 2 CALLA + ALL, 6 chronic T-cell proliferations, and 8 T-cell cultures. Determinations were made with cells incubated in diferric human colostral Lf, 1 × 10-1° M in HBSS/5 mg/dl of CaC12, for 1 h at 4°C prior to assessment of surface Lf, or with aliquots of the same cell isolates similarly incubated with the omission of Lf. PHBR and A were determined using control cell preparations stained with FITC-GAM-IgGzb only. Details are presented in Materials and Methods. t NS = difference not significant at the 0.05 level of probability with use of F test.

Lack of correlation of Lf expression with immunophenotype and cell cycle In lymphocytes from subjects with B-cell hematopoietic malignancies, there was no significant relationship of the percentage of Lf-positive cells with the presence or absence of surface immunoglobulin, immunoglobulin isotype, or light chain type. There was no significant correlation of Lf positivity with the expression of any surface marker antigen, except specific B- or T-cell lineage markers. Similarly, lymphocytes in T-cell preparations expressed a variety of T-cell antigens, but there was no surface antigen where its expression was significantly correlated with Lf positivity. We determined the cell cycle phase in nine subjects (six B-CLL, one HCL, and two T-cell leukemias). Most of the cells (87%98%) were in G0-G1 and the percentages of cells expressing surface Lf ranged from 0.1% to 35.5%. DISCUSSION Among peripheral blood non-adherent mononuclear cells, expression of surface Lf occurs primarily among B-lymphocytes, as demonstrated by the present results and our previous quantification of total cellular immunoreactive Lf [9]. The percentages of positive cells and/or intensity of surface Lf-specific fluorescence on these cells is not significantly altered by incubation with exogenous diferric Lf. A corresponding result was obtained by other investigators in blood mononuclear cells from small numbers of normal subjects and patients with C L L using radioiodinated Lf [22]. We also observed that the percentage of Lf-positive MNC remains stable over time in vivo. Although Lf positivity appears to be found mostly on the surfaces of B-cells, Lf positivity is not related to the presence of specific cell surface antigens (other than those markings B- or T-cell lineage) or cell cycle. The degree of cell activation is probably not related to the cell's ability to bind Lf, because tonsillar B-cells

of different densities express surface Lf similarly. However, the results of our analyses of B-cell malignancies comprising cells representing different maturational stages [23], suggest that the capacity to express Lf can vary according to maturation. Early B-cell progenitors (i.e. C A L L A + A L L cells) and cells at pre-plasmocytic stages of activation (i.e. H C L cells) do not express surface Lf whereas immature Blymphocytes (i.e. B-CLL cells) display the highest percentages of surface Lf-positive cells. These data are also supported by our observations of EBVtransformed B-cell lines (representing activated Bcells) in which surface Lf is virtually undetectable. The lack of expression of surface Lf in all of the cells within each B-CLL clone could represent intraclonal heterogeneity in B-CLL which has been demonstrated for other phenotypic features, e.g. expression of lysosomal enzymes or the ability to synthesize, secrete, or degrade immunoglobulin [24, 25]. Thus B-CLL cells with surface Lf could represent the least mature elements within the clone. In sharp contrast to the results obtained with B-cells, minimal quantities of Lf were detected in association with resting, cultured, or neoplastic T-lymphocytes, consistent with the findings of previous studies using radioimmunoassay techniques [4, 7]. Other investigators have reported apparent Lf binding by cells of some T-cell lines, although disparate results were obtained with rosette and immunofluorescence assay techniques, and some of the Lf binding could have been non-specific, related to the high concentrations of Lf in which the cells were incubated [11]. Functions attributed to Lf include regulation of granulopoiesis [26], bactericidal activity [27], and iron transport [28, 29]. The functional relationship(s) of Lf to lymphocytes is unclear. Diferric but not apoLf is capable of inhibiting the mixed lymphocyte reaction [30], but Lf does not donate iron to proliferating lymphocytes [31, 32], does not affect nonspecific cytotoxic functions mediated by lymphocytes [33], and does not influence colony-stimulating

446

T.W. BUTLERet al.

activity production by human lymphocytes in vitro [34, 35]. Although the physiologic or pathologic significance of surface-bound Lf on lymphoid cells cannot be readily explained with information presently available, our study shows that Lf expression is related to lymphoid lineage (i.e. B-cells) and to stage of maturation (i.e. immature B-cells), but not to cell proliferation, activation, a n d / o r cell cycle. As for other hematopoietic cell lineages such as granulocytes, the ability to bind Lf could play a role for the final stages of B-cell maturation. Acknowledgements--This work was supported by Veterans Administration Medical Research Funds, National Institutes of Health Hematology Training Program Grant 5-T32-AM07488 (NIADDKD), and Grants AM 03555, AI 18745, and CA 13148. Dr Butler is the recipient of a Frommeyer Fellowship. The authors appreciate the technical assistance of Ms Gail Phillips, Ms LiUian Blutcher, Ms Margaret Blackmon, Ms Peggy Powell, and Ms Donna Crabb, and the manuscript preparation and proofreading skills of Ms Mary Patton and Ms Helen Cooner. Dr Butler's current address is East Carolina University School of Medicine, Section of Hematology/Oncology, Brody Building 3E-106, Greenville, North Carolina 27858-4354.

REFERENCES 1. Masson P. L., Heremans J. F. & Schonne E. (1969) Lactoferrin, an iron-binding protein in neutrophilic leukocytes. J. exp. Med. 130, 643. 2. Boxer L. A., Haak R. A., Yang H. H., Wolach J. B., Whitcomb J. A., Butterick C. J. & Baehner R. L. (1982) Membrane-bound lactoferrin alters the surface properties of polymorphonuclear leukocytes. J. clin. Invest. 70, 1049. 3. Maneva A. I., Sirakov L. M. & Manev V. V. (1983) Lactoferrin binding to neutrophilic polymorphonuclear leucocytes. Int. J. Biochem. 15, 981. 4. Bennett R. M. & Davis J. (1981) Lactoferrin binding to human peripheral blood cells: an interaction with a B-enriched population of lymphocytes and a subpopulation of adherent mononuclear cells. J. Immun. 127, 1211. 5. Birgens H. S., Hansen N. E., Karle H. & Kristensen L. O. (1983) Receptor binding of lactoferrin by human monocytes. Br. J. Haemat. 54, 383. 6. Butler T. W., Heck L. W., Berkow R. L., Prasthofer E., Grossi C. E. & Barton J. C. (1986) Quantification of monocyte-associated lactoferrin. Clin. Res. 34, 185A (abstract). 7. Bennett R. M. & Kokocinski T. (1978) Lactoferrin content of peripheral blood cells. Br. J. Haemat. 39, 509. 8. Bennett R. M., Davis J., Campbell S. & Portnoff S. (1983) Lactoferrin binds to cell membrane DNA. Association of surface DNA with and enriched population of B cells and monocytes. J. clin. Invest. 71,611. 9. Butler T. W., Heck L. W., Huster W. J., Grossi C. E. & Barton J. C. (1988) Assessment of total immunoreactive lactoferrin in hematopoietic cells using flow cytometry. J. Immunol. Methods 108, 159.

10. van Snick J. L. & Masson P. L. (1976) The binding of human lactoferrin to mouse peritoneal cells. J. exp. Med. 144, 1568. 11. ¥amada Y., Amagasaki T., Jacobsen D. W. & Green R. (1987) Lactoferrin binding by leukemia cell lines. Blood 70, 264. 12. Dagg M. K. & Levitt D. (1981) Human B-lymphocyte subpopulations. I. Differentiation of density-separated B lymphocytes. Clin. Immun. Immunopath. 21, 39. 13. Gregory C. D., Tursz T., Edwards C. F., Tetaud C., Talbot M., Caillou B., Rickinson A. B. & Lipinski M. (1987) Identification of a subset of normal B cells with a Burkitt's lymphoma (BL)-like phenotype. J. Immun. 139, 313. 14. Barton J. C., Parmley R. T., Butler T. W., Williamson S. E., MacKenzie S., Chandler D. B., Blackburn W. & Heck L. W. Jr. (1988) Neutrophil lactoferrin content: variation among mammals. Anat. Rec. 221, 567. 15. Barton J. C., Parmley R. T., Butler T. W., Williamson S. E., Lilly M. B., Gualtieri R. J. & Heck L. W. (1988) Differential staining of neutrophils and monocytes: surface and cytoplasmic iron-binding proteins. Histochem. J. 20, 147. 16. Bennett R. M. & Mohla C. (1976) A solid-phase radioimmunoassay for the measurement of lactoferrin in human plasma: variations with age, sex, and disease. J. Lab. clin. Med. 88, 156. 17. Bennet R. M., Bagby G. C. & Davis J. (1981) Calciumdependent polymerization of lactoferrin. Biochem. biophys. Res. Commun. 101, 88. 18. Dean P. N., Gray J. W. & Dolbeare F. A. (1982) The analysis and interpretation of DNA distributions measured by flow cytometry. Cytometry 3, 188. 19. Barton J. C., Prasthofer E. F., Egan M. L., Heck L. W., Koopman W. J. & Grossi C. E. (1986) Rheumatoid arthritis associated with expanded populations of large granular lymphocytes. Ann. intern. Med. 104, 314. 20. Snedecor G. W. & Cochran W. G. (1956) Statistical Methods (5th edn), pp. 45--46; 244-250. The Iowa State University Press, Ames. 21. Morabito F., Prasthofer E. F., Dunlap N. E., Grossi C. E. & Tilden A. B. (1987) Expression of myelomoncytic antigens on chronic lymphocytic leukemia B cells correlates with their ability to produce interleukin 1. Blood 70, 1750. 22. Birgens H. S., Karle H., Hansen N. E. & Kristensen L. O. (1984) Lactoferrin receptors in normal and leukaemic human blood cells. Scand. J. Haemat. 33, 275. 23. Greaves M. F. (1986) Differentiation-linked leukemogenesis in lymphocytes. Science 234, 697. 24. Grossi C. E., Zicca A., Leprini A., Cadoni A., Pistoia V. & Ferrarini M. (1982) Acid hydrolases as markers of maturation in B-cell chronic lymphocytic leukemia. Blood 60, 220. 25. Rubartelli A., Sitia R., Zicca A., Grossi C. E. & Ferrarini M. (1983) Differentiation of chronic lymphocytic leukemia cells: correlation between the synthesis and secretion of immunoglobulins and the ultrastructure of the malignant cells. Blood 62, 495. 26. Broxmeyer H. E., Smithyman A., Eger R. R., Meyers P. A. & de Sousa M. (1978) Identification of lactoferrin as the granulocyte-derived inhibitor of colony-stimulating activity production. J. exp. Med. 148, 1052. 27. Arnold R. R., Cole M. F. & McGhee J. R. (1977) A bactericidal effect for human lactoferrin. Science 197, 263.

Lactoferrin and lymphocytes 28. van Snick J. L., Masson P. L. & Heremans J. F. (1974) The involvement of lactoferrin in the hyposideremia of acute inflammation. J. exp. Med. 140, 1068. 29. Moguilevsky N., Masson P.-L. & Courtoy P.-J. (1987) Lactoferrin uptake and iron processing into macrophages: a study in familial haemochromatosis. Br. J. Haemat. 66, 129. 30. Slater K. & Fletcher J. (1987) Lactoferrin derived from neutrophils inhibits the mixed lymphocyte reaction. Blood 69, 1328. 31. Brock J. H., Mainou-Fowler T. & Webster L. M. (1986) Evidence that transferrin may function exclusively as an iron donor in promoting lymphocyte proliferation. Immunology 57, 105. 32. Good M. F., Powell L. W. & Halliday J. W. (1987) The effect of non-transferrin-bound iron on murine T

447

lymphocyte subsets: analysis by clonal techniques. Clin. exp. Immun. 70, 164. 33. Nishiya K. & Horwitz D. A. (1982) Contrasting effects of lactoferrin on human lymphocyte and monocyte natural killer activity and antibody-dependent cellmediated cytotoxicity. J. Immun. 129, 2519. 34. Bagby G. C. Jr., Rigas V. D., Bennett R. M., Vandenbark A. A. & Garewal H. S. (1981) Interaction of lactoferrin, monocytes, and T lymphocyte subsets in the regulation of steady-state granulopoiesis in vitro. J. clin. Invest. 68, 56. 35. Strykmans P., Delforge A., Amson R. B., Prieels J. P., Telerman A., Bieva C., Deschuyteneer M. & RongeCollard E. (1984) Lactoferrin: no evidence for its role in regulation of CSA production by human lymphocytes and monocytes. Blood Cells 10, 369.