BririshJournal oftlaematology, 1978.39, 509.

Lactoferrin Content of Peripheral Blood Cells ROBERT M. BENNETT A N D THERESA KOKOCINSKI

Department of Medicine, University of Oregon Health Sciences Center, Portland,,Orgeon, and Department ofMedicine, University of Chicago, Chicago, Illinois, U.S.A. (Received 22 September 1977; acceptedfor publication ibjanuary 1978) SUMMARY. The peripheral blood of 60 normal adults was separated into plasma, red cells, neutrophils, lymphocytes, monocytes and platelets. Lactoferrin concentrations were measured in the plasma and cell extracts and compared to those of lysozyme. The neutrophil lactoferrin content in males and post-menopausal females was found to be significantly higher than in pre-menopausal females. A small amount of lactoferrin was found in association with monocytes, but not with lymphocytes, erythrocytes and platelets. Neutrophil lysozyme concentrations did not exhibit any variation with sex and age; but the level in monocytes was higher than that in neutrophils. No correlation was observed between individual neutrophil lactoferrin values and the plasma level. Immunofluorescent studies showed neutrophils to have a lobulated pattern suggestive of nuclear staining. Monocytes did not show direct staining, but exhibited a peripheral pattern after prior exposure to lactoferrin-confirming the existence of a surface receptor. Gel chromatography indicated that neutrophil lactoferrin is in a polymerized or complexed form which elutes with the void volume on Sephadex G-zoo; serum lactoferrin consists of two forms, one of which also elutes with the void volume on Sephadex G-zoo. Recent studies have indicated that the iron binding protein, lactoferrin, is normally present in the plasma of healthy humans (Bennett & Mohla, 1976). Its major repository is in epithelial secretions (Masson & Heremans, 1966, 1968; Eddie-Quartey & Bennett, 1973; Bennett et al, 1973; Biserte et al, 1963; Bennett & Skosey, 19771, being found in greatest quantities in breast milk. Despite its widespread distribution a well-defined biological role for lactoferrin has not yet been elucidated. An important contribution to our knowledge of lactoferrin was the work of Masson et a1 (1969) which demonstrated its presence in neutrophils. Through the work of Baggiolini et a1 (1970) and Spitznagel et a1 (1974),lactoferrin has been shown to be associated with the secondary (specific) granules of neutrophils. The location of lactoferrin to specific granules has been supported by recent studies that have demonstrated its isochronous release with other lysosomal proteins during phagocytosis (Van Snick & Masson, 1974; Leffell & Spitznagel, 1975; Bennett & Skosey, 1977).At variance with these observations is the work of Green et a1 (1971)who demonstrated an association of lactoferrin with neutrophil nuclei, both by immunofluorescent and immunochemical methods. It is generally quoted, from the work Correspondence: Dr R. M. Bennett, Division of Immunology, Allergy and Rheumatology, University of Orgeon Health Sciences Center, Portland, Oregon 97201, U.S.A. 0007-1048/78/08000509$02.000 1 9 7 8 Blackwell Scientific Publications 509

510

Robert M . Bennett and Theresa Kokocinski

of Masson et a1 (1969), that of all the circulating blood cells, lactoferrin is only present in neutrophils, with a concentration ofabout 3 pg/Io6 cells. No studies have been reported on the distribution of neutrophil lactoferrin levels in a comprehensive population of healthy adults. Nor have the more sensitive techniques of radioimmunoassay been employed to explore the possibility that other circulating blood cells may contain small amounts of lactoferrin. The purpose of this study is to examine the evidence that lactoferrin is specific for neutrophils, determine its intracellular concentration in healthy adults with respect to age and sex and examine any possible correlation between its cellular level and its plasma concentration. Lysozyme, a well-studied protein of similar localization, has been studied simultaneously for comparison. Recent work has shown that the measurement of plasma lactoferrin may provide a useful indirect assessment of the total blood granulocyte pool (Hansen et al, 1975). Previous work on lysozyme (Osserman et al, 1974), which has a well-established association with circulating leucocytes (Hansen & Andersen, 1973; Briggs ef al, 1966; Hansen et al, 1972), has demonstrated the feasibility of using the data from protein turnover studies, cellular levels and plasma levels, in assessing neutrophil turnover (Hansen, I 974). The intriguing possibility that lactoferrin may be an even more specific marker for neutrophil turnover studies, demands a precise knowledge of its distribution in circulating blood cells and accurate data on its concentration in neutrophil leucocytes. METHODS

Cells. Plasma and peripheral blood cells were obtained from 60 healthy hospital personnel. In the case of females it was ascertained whether or not they were still menstruating. General. Acid washed, siliconed glassware or polypropylene was used in all procedures using cells. 40 ml of venous blood was collected into plastic centrifuge tubes containing 3 ml of 0.077 M EDTA. Separation ofplatelets. The anticoagulated blood was centrifuged at 50 g at 10°C for I 5 min. The upper two-thirds of the plasma (including platelets) was transferred to separate tubes and centrifuged at 400g at 4°C for 30 min. The resulting platelet pellet was resuspended in I ml of phosphate buffered saline (PBS). The platelet-free plasma was reconstituted with the blood cells from the first centrifugation step, prior to further purification. Separation of mononuclear cells. Mononuclear cells were separated from erythrocytes and granulocytes by Ficoll-Hypaque centrifugation (Summers et al, 1974). The plasma layer was removed and left at 37°C for further use (see below). The mononuclear cells were pipetted from the plasma-Ficoll-Hypaque interface. The Ficoll-Hypaque layer was discarded and the remaining erythrocytes and granulocytes were left for further separation. Separation of erythrocytes and granulocytes.* The combined erythrocyte-granulocyte suspension was added to 40 ml of a 30 g/l. solution of dextran 150(Sigma Chemicals) in a IOO ml stoppered cylinder. This was allowed to stand for 14 h a t 4”C, or until the erythrocytes had sedimented into a clearly defined layer. The granulocyte and erythrocyte layers were removed Granulocytes and neutrophils are used interchangeably-the exceed 6% in any of the donor samples.

eosinophil fraction of the granulocytes did not

Lactoferrin in Blood Cells

511

and further purified. The granulocyte containing layer was centrifuged at 400 g for 5 min at 4°C and washed twice with NaCl(9 g/l.). The pellet was resuspended in 2 ml of ice cold NaCl (9 g/l.) and 6 ml of ice-cold distilled water added and gently stirred for 30 s, isotonicity was restored by adding 2 ml of ice-cold NaCl(35 g/l.). Further centrifugation for 5 min at 4°C at 400 g, removed contaminating red cell debris, and two further washes with isotonic saline removed haemoglobin in solution. After the final wash the cell pellet was resuspended in I ml of PBS. The erythrocytes were similarly treated except for the hypotonic lysis step. Separation of lymphocytesfrorn monocytes. The mononuclear cell suspension was made up to 40 ml with isotonic saline and centrifuged for I 5 min at 4°C a t 400 g. The pellet was resuspended in 2.5 ml of the original plasma and incubated a t 37°C for 10min. This step restored to the monocytes the property of adhering to glass (Rabinowitz, 1964). The cells were further centrifuged for 5 min at room temperature at 400g and resuspended in 1.5 ml of plasma. A glass bead column was prepared from a modification of that described by Summers et al(1974). A glass column (zoo mm x 7.5 mm), acid washed and siliconized, was packed with 60-mesh glass beads previously soaked in concentrated nitric acid, thoroughly rinsed in distilled water and dried at 90°C. The bottom of the column contained about 10 mm of 3 mm diameter glass beads held in place by a fine nylon gauze. The top 10mm of the column was similarly packed with 3 mm glass beads. The suspension of mononuclear cells was applied to the column and allowed to run down to just past the top layer of 3 mm glass beads. The ‘primed’ column was then incubated a t 37°C for 30 min-to permit the monocytes to adhere to the glass beads. Thereafter 10 ml of plasma (previously kept at 37°C) was run into the column at the same rate as the effluent. This was followed by 20 ml of 40% plasma in osotonic saline at 37°C; the first I 5 ml of this effluent contained lymphocytes. The remaining effluent was discarded and the monocytes eluted with 35 ml of a solution of the following compositions: EDTA, 0.2 g, NaCl 8.0 g, KC1 0.2 g, Na2HP04,0.2 g and glucose 0.2 g, made up to I litre. Most of the monocytes eluted in the first 20 ml of effluent. Using centrifugation at 4008 ( 5 min at 4°C) the lymphocyte and monocyte fractions were washed twice in isotonic saline and each fraction was resuspended in I ml of PBS. Treatment of purijed cell preparations. Erythrocytes, granulocytes and lymphocytes were diluted IOO times and monocytes 10 times, prior to cell counting with a haemocytometer. A smear of each cell fraction was stained with Wright’s stain and the percentage of ‘contaminating’ cells was noted. A 50 pl suspension of each fraction, excepting the erythrocytes, was added to 0.5 ml of eosin (10 g/l.) in PBS and the viability of the cells measured in terms of eosin exclusion. From these results the total number of viable specific cells in each fraction was calculated, with adjustments being made for contaminating cells. After one further centrifugation the cell fractions were resuspended in I M NaCl and disrupted by sonication on ice (Sonifier Cell Disrupter W 1 8 5 D., Heat Systems Electronics Inc., Plainview, L.I., N.Y.).The duration of sonication was initially monitored by the completeness of cell disruption, as estimated by phase contrast microscopy. In five granulocyte suspensions the cell disruption was also done in 0.01 M citric acid and 0.15 M NaCl. In each instance sonication was compared to freezing and thawing six times. The disrupted cell extracts were centrifuged at Iooog at 4°C for 30 min and the supernatants assayed for lactoferrin and lysozyme-expressed as pg per 106 cells. It has been noted that more complete centrifugation

512

Robert M . Bennett and Theresa Kokocinski

(128 ooo g for 18 h) to completely sediment organelles and membranes does not result in a significant change in the final lactoferrin concentration. The cell debris from sonication in 1.0 M NaCl was extracted a further three times in one cell preparation and the supernatant lactoferrin concentration measured after each extraction. In one instance a separate purification of neutrophils (the granulocyte fraction contained I O h of eosonophils) was made from IOO ml of freshly drawn venous blood and the extract from the disrupted cells was lyophilized; this will be referred to hereafter as neutrophil extract (NE). Plasma collection. Simultaneous samples of blood were anticoagulated with 3 mg EDTA per 5 ml of blood and the tubes kept on ice prior to separation of plasma. The tubes were centrifuged at 508 at 4°C for 20 min and the upper two-thirds of the plasma was removed for subsequent lactoferrin and lysozyme assays; this avoided contamination with any leucocytes at the plasma-red cell interface. T h e measurement of lactoferrin. Lactoferrin was isolated from human breast milk as previously described (Bennett & Mohla, 1976).Iron depleted lactoferrin (apolactoferrin) was prepared by dialysis of a 10g/l. lactoferrin solution against 20 vol of 0.1 M citric acid as described by Masson et a1 (1968). The purity of the lactoferrin and apolactoferrin was assessed by cationic polyacryamide gel disc electrophoresis using a 7.5% acrylamide gel with a running pH of 4.5 (Wrigley, I 962). A radioimmunoassay for the measurement of lactoferrin was employed using polypropylene counting vials (Beckman Biogamma Vials) coated with high affinity anti-lactoferrin antisera prepared in rabbits. The methodology of this assay has been described in detail elsewhere (Bennett & Mohla, 1976). Briefly, tracer labelled lz5I lactoferrin (about 3 0 pCi/pg) in 0.9 ml of barbital-BSA diluent was added to the antibody-coated vials along with 100 pl of appropriately diluted samples and lactoferrin standards. After overnight incubation at 3 7"C, the tubes were aspirated and washed four times with tap water prior to counting in a Biogamma Counter (Beckman Instruments, Fullerton, California). This is a competitive binding assay, the greater the amount of lactoferrin in the sample the less radioactivity is bound to the antibody-coated vials. T h e measurement of lysozyme. Lysozyme was measured by a modification of the turbidimetric assay of Shugar (1952). The lysozyme substrate was prepared from ultraviolet killed Micrococcus lysodeikticlrs (Difco Laboratories, Mich.). The substrate (80 mg) was homogenized in 25 ml ofo.1 M phosphate buffer pH 6.2 for 2 min and the volume then brought up to 300 ml. The suspension was kept at 37°C in a water bath. The appropriately diluted sample (0.5 ml) was added to 3 ml of substrate suspension and mixed by inversion. Readings of E,,, n m were taken at 37°C at intervals of I min for 6 min. The ratio of the 3 min and 6 min readings was compared to a similar ratio of standards prepared from egg white lysozyme (Difco Laboratories, Mich.). The choice of EDTA as an anticoagulant is critical, as heparin inhibits lysozyme activity (Kaiser, 1953) and also interferes with the lactoferrin immunoassay (Bennett & Mohla, 1976). Radioactive labelling of lactoferrin with l Z 5 Z and 5yFe. Iodination of lactoferrin with 1251 was accomplished by a modification of the chloramine-T method (Greenwood et al, 1963) using a 0.4 M borate buffer pH 8.0, as previously described (Bennett & Mohla, 1976). The incorporation of 59Feinto apolactoferrin employed the chelating effect of nitrilotriacetate (Bates et al, 1967). IOO p1 of I m M trisodium nitrilotriacetate in 0 . 2 M phosphate buffer pH 7.6 was added to

Lactoferrin in Blood Cells

513

p~ of 59FeC13(o.j ml) containing 2 mCi of activity. 50 pl of this 59Fechelate was added to I mg of apolactoferrin dissolved in 20 p1 of 0.2 M phosphate buffer pH 7.6. The reaction mixture was incubated at 37°C for 3 h; unincorporated iron was separated by passage through a small column of Sephadex G-25 (Pharmacia Fine Chemicals, Piscataway, N.J. 08854) equilibrated with I mM tri-sodium nitrioloacetate in 0.2 M phosphate buffer pH 7.6. I mg of the neutrophil extract was similarly treated. Gel chromatography. Normal human serum, 1251 lactoferrin and 59Fe labelled neutrophil extract were chromatographed on a Sephadex G-200 column (200 x 2.5 cm) equilibrated with PBS and 3 . 0 ml fractions were collected. The percentage transmission at 280 nm was continuously recorded (Uvicord monitor, Pharmacia) during the serum elution. Aliquots of 0.5 ml from each fraction of the 1251 lactoferrin and of the 59Fe neutrophil extract were counted. I ml of each serum fraction was added to the antibody-coated radioimmunoassay tubes and incubated at 37°C for 24 h. The tubes were washed three times with NS and I ml of tracer labelled 1251 lactoferrin was added. After a further incubation overnight at 37°C the tubes were washed four times with tap water and counted. Aflnity chromatography. Purified antilactoferrin antibodies, for immunofluorescent studies, were prepared from cyanogen bromide activated sepharose 4B (CNBr-qB) coupled to lactoferrin, as described by Cuatrecasas et al (1968). Similarly high avidity anti-lactoferrin antiserum was coupled to CNBr-4B for isolation of lactoferrin containing fractions from the 59Feneutrophil extract. The latter was sequentially eluted with: (i) 0.5 M NaCl, pH 7.6; (ii) 0.2 M glycine+o.5 M NaCl, pH 2.8; (iii) 0.1 M acetic acid+o.5 M NaC1, pH 2 . 3 ; (iv) 3 . 5 M sodium thiocyanate, pH 6.8. Immunojuorescent studies. Specific antilactoferrin immunoglobulins prepared by affinity chromatography, were fluoresceinated to a fluorescein-protein ratio of 2.57 at E490/280 (Sternberger, 1974). Purified suspensions of neutrophils, lymphocytes, monocytes, erythrocytes and platelets were spread on microscope slides, dried and fixed in absolute ethanol for 5 min. The smear was then washed with PBS and stained with fluorescein-labelled antilactoferrin. Prior absorption of the fluoresceinated antisera with purified lactoferrin was used as a check on the specificity of the fluorescence. Further smears were made after the addition oflactoferrin ( I mg/ml) to the cell suspensions and incubation at room temperature for 30 min, as suggested by the work of Van Snick & Masson (1974). The smears were studied with an OA-10 Microstar American Optical microscope, using a high tension vapour lamp as a light source fitted with BG 12 and UGI filter. A dark field condenser was employed and photographs were taken using daylight high speed Ektachrome film, ASA 160. Statistical analyses. Results are expressed as the arithmetical mean & SE (standard error of the mean), unless otherwise indicated. Statistical differences were determined using Student’s t test (no significance, P> 0.051, comparison being of aggregate mean values. Linear regression lines were fitted to data points by the method of least squares. I

RESULTS Purity of Cell Preparation The total number of cells in I ml PBS prior to sonication, expressed as a mean and standard deviation of all 60 specimens was: neutrophils, 93 x 106 (S, 31); lymphocytes, 52 x 1o6 (S, 19);

Robert M . Bennett and Theresa Kokocinski

514

monocytes, 1 . 4 x 106 (S, 0.8); erythrocytes, 1 2 0 0 ~106 (S, 113); platelets, 860 x 1 0 6 (S, 102). These figures have been corrected for contaminating and non-viable cells. The average purity of the lymphocytes and neutrophils was 98% with the major contaminants being monocytes and lymphocytes respectively. Monocyte preparations were less pure with an average lymphocyte contamination of I I %. Erythrocyte and platelet preparations had less than 0.I O h of contaminating cells. The viability of the cells, as assessed by eosin exclusion, was 99% for both neutrophils and lymphocytes, while monocyte preparations contained an average of 3 % of non-viable cells. Extraction of Lactoferr in fro m Neu trop hi Is Initial experiments were done to find out the best method of extraction as outlined in the Methods. These results are shown in Table I and indicate that sonication for 10s at 40 watts in I M NaCl yielded the best results. Repeated sonication and extraction of the cell pellets yielded an additional 2.03 pg of lactoferrin/Io6 cells, compared to 19.91 pg/Io6 cells after the first sonication. This is a 91% yield after the first extraction. TABLE I. Lactoferrin yield obtained with different methods of extraction Method of disruption (pgglio6 cells) Extractant

citric acid NaCl NaCl

Freeze

0.01 M

4.65

0 . 15 M

(4.6) 7.96 (36)

1.0M

1.01

(21)

Sonicate 12.74 ( 5 8 ) 6.31 (29) 19.91 (90)

Further extraction ofcell debris to ‘exhaustion’

ND ND 21.94

(100)

Effects of using different extractants and methods of cell disruption on the yield of lactoferrin from neutrophils-expressed as pg/ lo6 cells, parentheses enclose the percentage yield, based on a maximal neutrophi1 level of 21.94 pg/Io6 cells; the extra 2.03 pg resulted from multiple extraction of the cell debris.

Cell Lactoferrin Concentrations Lactoferrin was found predominately in neutrophils with a significant difference ( P < 0.002) between males and females over 50 years compared to pre-menopausal females; the mean concentrations in pg/Io6 Cells were: males, 17.8 (N=3 I , SE k0.96); pre-menopausal females, 10.0(N=19, SEk1.02); females over 50 years 20.8 (N=Io, SE+2.o5), see Fig I . All the females under 50 years of age (mean age 29) were premenopausal, two of the females over 50 years of age were still menstruating erratically, the lower two points in Fig I . A small amount of lactoferrin was found consistently in association with the monocyte preparations, 0.32 pg/Io6 cells (N= 19, SE+o.o03). There was no obvious difference in monocyte lactoferrin levels with age or sex, nor was there any discernible alteration with age in the neutrophil lactoferrin concentrations in men. 0ther cell preparations contained only minute amounts of lactoferrin, presumed due to contamination: lymphocytes, 0.0036 pg/Io6 cells; erythrocytes, 0.00087 pg/Io6 cells; platelets, 0.000027 pg/Io6 cells.

Lactoferrin in Blood Cells

30t :

0

30

0

0

2 J

t

0

0

25

W

,"20

0 \

5 15 W N

(0

0 v)

>

J

5

0 MALES FEMALES FEMALES M+F

Lactoferrin content of peripheral blood cells.

BririshJournal oftlaematology, 1978.39, 509. Lactoferrin Content of Peripheral Blood Cells ROBERT M. BENNETT A N D THERESA KOKOCINSKI Department of...
986KB Sizes 0 Downloads 0 Views