THE ANATOMICAL RECORD 221567-515 (1988) NeutrophiI Lactoferrin Content: Variation Among Mammals JAMES C. BARTON, RICHARD T. PARMLEY, THOMAS W. BUTLER, SUE WILLIAMSON. SANDRA MAcKENZIE, DAVID B. CHANDLER, WARREN BLACKBURN, AND LOUIS W. HECK, JR. Veterans Administration Medical Center (J.C.B., S. W ,D.B.C., WB., L W H.), and Department of Medicine, Comprehensive Cancer Center, and Multipurpose Arthritis Center, University ofAlabama at Birmingham, Birmingham, Alabama 35233 (J C.B., II WB., D.B.C., WB., L WH.);Department of Pediatrics, University of Texas Health Science Center at San Antonio, San Antonio, Texas 78229 (R.TP.);and Birmingham Zoo, Birmingham, Alabama 35223 (S.M.) ABSTRACT Lactoferrin (LfJ in blood andor marrow neutrophils was semiquantified using indirect immunofluorescence technique in nine mammalian species. Neutrophil iron-binding reactivity (NFeBR), which corresponds primarily to Lf, was also visualized and semiquantified using functional cytochemical (FeNTA-AF)technique at the light microscopic level in these nine and in a n additional fifteen mammalian species, and in selected species at the ultrastructural level. Neutrophil immunoreactive Lf was positively correlated with total cellular and granule content of NFeBR among these nine species, and with previously reported concentrations of neutrophil Lf quantified by radioimmunoassay. Relative levels of Lf in neutrophil extracts from rat, hamster, and human were confirmed using SDS-polyacrylamide gel electrophoresis and immunoblotting. Relatively high levels of immunoreactive neutrophil Lf andor NFeBR were observed in carnivores (ten species) and primates (six species). Among rodents (five species), the levels were variable, and the artiodactyls (four species) studied had low levels. These results demonstrate that neutrophi1 Lf levels vary widely among mammalian species. In addition, FeNTA-AF technique provides a rapid means of evaluating animals for relative quantities of neutrophil Lf. Lactoferrin (LO functions defined primarily in vitro include bacterial killing, feedback inhibition of granulopoiesis, and nonheme iron transport (Broxmeyer, 1984). Some functions of neutrophil Lf have also been inferred from observations in vivo, e.g., the increased frequency of bacterial infections among rare patients with congenital neutrophil Lf deficiency. However, because the neutrophils of these patients possess other anatomic, functional, and biochemical anomalies (Boxer et al., 1982), a n isolated neutrophil Lf deficiency state remains unknown among humans. Alternatively, animals with naturally low neutrophil Lf levels could be useful for studies of Lf function. The evaluation of prospective subjects by established means would require a n immunologic assay of neutrophil Lf, procedures for efficiently extracting neutrophil Lf, which may vary from species to species (Masson et al., 1969; Baggiolini et al., 1970; Kinkade et al., 19761, and, possibly, the production of species-specific, anti-Lf antibodies. By staining the blood cells of a variety of mammals with iron nitrilotriacetateacid ferrocyanide (FeNTA-AF) method and by using a cytochemical scoring technique, we visualized and semiquantified Lf as neutrophil iron-binding reactivity (NFeBR) (Parmley et al., 1982; Barton and Parmley, 1986; Barton et al., 1987). In mammals, NFeBR scores could be positively correlated with granule content of NFeBR at the ultrastructural level, with indirect im0 1988 ALAN R. LISS, INC. munofluorescence staining of neutrophil Lf, and with quantification of total neutrophil Lf by radioimmunoassay and immunoblotting in the respective species. The present methods demonstrate the interspecies variability in neutrophil Lf, and permit the rapid evaluation of animals for their relative quantities of neutrophil Lf. MATERIALS AND METHODS Blood and Marrow Collection Human blood and marrow specimens were acquired according to the Declaration of Helsinki and with the approval of the Human Use Committee of the University of Alabama a t Birmingham. Peripheral blood smears from humans were prepared at the time of finger puncture. Morphologically normal marrow was obtained from the posterior iliac crests of six patients undergoing marrow aspiration and biopsy for staging of malignancy; all other human and animal subjects were apparently healthy at the time of specimen collection. Principles of laboratory animal care as promulgated by the National Research Council were observed. Labora- Received August 19, 1987; accepted November 16, 1987. Address reprint requests to Dr. James C. Barton, Division of Hematology/Oncology, University of Alabama at Birmingham, University Station, Birmingham, AL 35294. 568 J.C. BARTON ET AL. tory animals were acquired from Southern Animal Farms, Prattville, AL (rats),Jackson Laboratories, Bar Harbor, ME (mice), Charles River Breeding Laboratories, Raleigh, NC, and Kingston, NY (guinea pigs and hamsters), and Myrtle's Rabbitry, Thompson Station, TN (rabbits). Blood was obtained from the retroorbital venous sinus from laboratory rodents under general anesthesia with a heparinized microhematocrit tube. Femoral marrow was collected from rats, rabbits, mice, and hamsters immediately after the animals had been sacrificed by cervical dislocation. In other animals, blood smears were prepared from fresh heparinized blood obtained by phlebotomy or heart puncture. Zoo animals were maintained under conditions specified by the Animal Welfare Act and the American Association of Zoological Parks and Aquariums. For these subjects, blood was obtained by venipuncture a t the time of routine health maintenance examinations; other specimens were acquired through veterinarians' clinical practices. Functional Cytochemical Technique For light microscopy, NFeBR was visualized in blood and marrow neutrophils as previously described in detail (Barton and Parmley, 1986). Briefly, thin smears prepared on standard glass slides were air-dried at room temperature overnight, fixed with freshly prepared buffered formol acetone (BFA) for 6 minutes, and rinsed in H2O. The smears were then incubated in freshly prepared 1%saponin (Fisher Scientific Co., Fairlawn, NJ) a t 50°C for 1 hour, and rinsed with 0.9% NaC1. The slides were incubated in freshly prepared 3 mM FeC13: 4 mM nitrilotriacetate (NTA) in 0.9% NaC1, pH 7.4 (FeNTA), at room temperature for 1 hour (Barton and Parmley, 1986). The slides were then rinsed in 0.9% NaC1, pH -5.5. In certain experiments, the slides were rinsed instead in 0.9% NaCl a t pH 4.0 or 8.0 after FeNTA incubation to determine whether there were cytochemically demonstrable pH-dependent differences in ironprotein binding affinity. The slides were then incubated in 1%acid ferrocyanide (AF) for 10 minutes, and rinsed in water (Barton and Parmley, 1986). The smears were counterstained in 1%Gurr's Nuclear Fast Red (Esbe Laboratory, Toronto, Canada) in 5% A12(S04)3.18 H20 for 5-10 minutes, rinsed in water, routinely dehydrated, cleared, and mounted. Normal and abnormal human control slides were included in each staining run. Blood smears similarly fixed with BFA were stained with AF technique alone for the assessment of native ferric iron in the cells, and to serve as a negative control for FeNTAAF staining. A blood smear from each subject was stained with WrightIGiemsa stain. One hundred consecutive mature neutrophils or heterophils (having two or more nuclear lobes) in the thin edge of each smear were rated 0 to 5 + on the basis on the distribution, color intensity, and granularity of their cytoplasmic stain deposits by light microscopy at 1,000 x magnification (Table 1, Fig. 1). A rating of 4 was defined as that appearance and quantity of FeBR visualized most commonly in the neutrophils of normal human subjects. The ratings were added to yield a NFeBR score (0-5001, which is almost linearly related to human neutrophil Lf content determined by immunoassay over a wide range of values (Barton et al., 1988). We also visualized NFeBR in ultrastructural specimens using a previously described method (Parmley et al., 1982). Briefly, venous blood samples were collected in heparinized tubes from six rats, four cows, four rabbits, four mice, and four humans. The samples were centrifuged at 1,500 x g for 3 minutes, and the buffy coat was removed with a pipette. The cells were resuspended in 3% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.35, for 1 hour. After rinsing in 0.1 M cacodylate7% sucrose buffer, the cells were resuspended for 1hour in 1% saponin at 50-55°C. The cells were then rinsed twice in cacodylate-sucrose buffer and once in 0.9% NaC1, exposed to FeNTA solution for 1hour at room temperature, rinsed three times in cacodylate-sucrose buffer, and incubated for 30 minutes in 1% AF. After three additional rinses in cacodylate-sucrose buffer, the cells were postfixed in 1% 0 ~ 0 4routinely , dehydrated, and embedded in S p u r low viscosity medium. Thin sections (5070 nm thick) were examined without counterstains with a Zeiss 109 electron microscope a t a n accelerating voltage of 50 kV. TABLE 1. Scoring criteria for NFeBR' Individual cell rating 0 Cytoplasmic pattern - Granularity - Color - Granules - Relative FeBR, quantityicell2 1+ 2+ 3+ Diffuse + 2 25% focal Weak Weak Weak + strong Pale grayish Pale grayish Pale grayish blue blue + medium blue blue None Rare Rare-few Focal (any) Undetectable Severely deficient Diffuse Moderately deficient Low normal 4+ Diffuse Strong Medium blue 5+ Diffuse + 2 25% focal Strong + very strong Medium blue + blue-black Few-moderate Moderate, many overlying nucleus Normal High normal, increased '100 consecutive mature neutrophils (possessing 2 or more nuclear lobes) on the thin edge of a peripheral blood smear are scored after normal and abnormal control slides stained simultaneously are reviewed. Total score is the sum of 100 individual cell ratings (range of scores 0-500); see also Figure 1. 'In comparison with normal human peripheral blood neutrophils. 569 NEUTROPHIL LACTOFERRIN IN MAMMALS 0 1+ 2+ 3+ 4+ 5+ Fig. 1. Cartoons of mature blood neutrophils depicting the relationships of the intensity and distribution of cytoplasmic FeNTA-AF stain deposits to individual cell rating; see also Table 1. Indirect lmmunofluorescence Technique fluoresceinated second antibody layer only. After a final Fresh heparinized blood or suspended marrow was washing in PBS and mounting in Fluoromount-G centrifuged at 1,500 x g for 15 minutes, and b a y coat (Southern Biotechnology Associates, Inc.), the preparaa Leitz orthoplan/*rthomat cells were collected. After the cells were washed either tions were in cold PBS, p~ 7.4, or in Hank’s balanced salt solution, microscope. Immunofluorescence reactivity was rated pH 7.2 cytocentrifuged Smears were prepared on glass according to these criteria: 0 = no Staining; k = faint 5 minutes(cytospin 2; &andon staining in occasional cells; 1 + = faint; 2 + = weak; slides gt 200 wm Southern Instruments. Inc.. Sewicklev. PA). The smears 3+ = moderate; and 4 + = ”, were air-dried for 24 hours,’fixed in fresh cold BFA for 1 Purification of Human Neutrophil Lactoferrin minute, and rinsed in water. The slides were then incuHuman neutrophil Lf from normal donors was purified bated in 1% saponin at 50°C for 1 hour, and rinsed using a n isolation procedure that included sequential thoroughly with 0.9% NaCl and then with PBS to reduce nonspecific, but not specific, Lf staining. The cells were sodium chloride extraction, heparin-Sepharose affinity then overlayered with 10 pl of polyspecific rabbit anti- ‘chromatography, and AcA44 gel filtration chromatograhuman milk Lf (lot #011063; Calbiochem-Behring Corp., ,phy (Heck, L.W., unpublished results). Coomassie blue La Jolla, CA) diluted 1:lO in 2% fetal calf serum/l% :staining of the gel after sodium dodecyl sulfate-polysodium azide, for 30 minutes at 4°C. After washing, the acrylamide gradient gel electrophoresis of the reduced cells were overlayered with fluorescein-labeled goat an- purified protein demonstrated two polypeptides of Mr tirabbit IgG (lot #J5A026; Southern Biotechnology As- 80,000 and 78,000. sociates, Inc., Birmingham, AL), 1:lO dilution, for 30 Production of Antibodies to Human Neutrophil Lactoferrin minutes a t 4°C. Similar procedures were employed for immunostaining using MoAb-5B2, a mouse monoclonal Murine monoclonal antibodies to neutrophil Lf were IgG2B antihuman neutrophil Lf (used as clarified as- produced in the Hybridoma Core Facility a t the Univercites), and fluorescein-labeled goat antimouse IgGZB (lot sity of Alabama at Birmingham. AIJ mice were immu#F5X085; Southern Biotechnology Associates, Inc.). nized with a total of 600 pg of the purified human Control slides were prepared using either a first layer neutrophil Lf according to the protocol of Lieberman and antibody of isotype unrelated to the second layer, or the colleagues (1972). After the fourth injection of antigen 570 J.C. BARTON ET AL. Fig. 2. FeNTA-AF (left) and indirect immunofluorescence (right) (A); cytoplasmic immunoreactive Lf staining was also weak (B). Pig staining of Lf in neutrophils, arranged in increasing order of cyto- blood (C,D); hamster marrow (E,F); dog blood (G,H); and human blood bar = 10 pm. plasmic staining intensity. In the cow, the nuclei stained more in- (I,.J). ~1,400; tensely with nuclear fast red than did the cytoplasm with FeNTA-AF in saline, the popliteal, inguinal, subaxillary, and brachial lymph node lymphocytes were fused with the nonsecretory myeloma Pcx63-Ag 8.653 (Kearney and Lawton, 1975). Hybridomas secreting antibodies with specificity for human neutrophil Lf were selected using an ELISA assay (Engvall and Pearlmann, 1972). Eight hybridoma positive wells containing antibodies against human neutrophil Lf were identified and cloned in tis- sue culture. Subsequently, three stable hybridoma [ Z IgM and 1 IgG (y 2b-(5B2)]were injected into Pristaneprimed CAF mice. Ascites fluid was collected and the antibodies purified by sequential DE-52 cellulose ionexchange chromatography followed by Sephadex G150 gel filtration chromatography. The specificity of these antibodies has been verified by solid-phase binding assays and immunoblotting using iodinated antibodies Fig 3. FeNTA-AF staining of neutrophils at the ultrastructural level. A. Rat. B. Cow. C. Mouse. D. Human. N = nucleus. ~18,000; bar = 1 Fm. (Dunn et al., 19851, and immunofluorescence. A poly- trophoretic transfer of proteins to nitrocellulose and inclonal monospecific goat antibody to human milk lacto- cubation of nitrocellulose with iodinated antibodies were ferrin was provided by Dr. Jiri Mestecky. performed as previously described (Dunn et al., 1985). lmmunoblotting Procedure RESULTS Samples of rat and hamster bone marrow, human Functional Cytochemical Technique peripheral blood neutrophils, and purified human neuNeutrophils or heterophils in blood films contained trophil Lf were prepared to minimize epitope denatura- granular cytoplasmic FeNTA-AF stain deposits that vartion (Daniel et al., 1983) and separated using SDS- ied widely in relative amounts among species (Table 2, polyacrylamide gradient gel electrophoresis. The elec- Fig. 2). Within the same species or individual, the distri- 572 J.C. BARTON ET AL. TABLE 2. Interspecies comparison of neutrophil lactoferrin determined by functional cytochemical and immunologic methods NFeBR score' (n) Rat, albino Wistar Cow, Holstein Rabbit, New Zealand white Pig, Sinelair miniature Hamster, golden Syrian Mouse, C57B1/6J Dog, mixed breed Cat, domestic shorthair Human 40 160 169 176 264 320 371 385 403 k 0 (3)4,5 + 27 (3) k 14 (414 f 18(3) 54 (7)4 + 0 (3I4z5 + 44(5) 9(3) k 3 (5114 + + Granule NFeBR' 1 2 3 4 5 Immunostaining Immunoassay ~ g / l O 7cells3 k 1+ 2+ 1+ 2-3 3+ 4+ 4+ 4+ 34,35.8 + 20-40 77,100,178 'Values are expressed as the mean 5 standard deviation of the mean. 'Ratings of FeNTA-AF stain deposits in neutrophil cytoplasmic granules observed at the ultrastructural level in order of increasing intensity. 3These data represent values obtained with different sources of neutrophils (peripheral blood, marrow, peritoneal exudate), neutrophii isolation procedures, and immunologic assays. Each assay was performed with species-specific antibody. Rabbit heterophil Lf was quantified by van Snick et al. (1974) and Boxer (1985), respectively; mouse neutrophil Lf by Segars and Kinkade (1977); and human neutrophil Lf by de Vet and ten Hoopen (1978), Oseas et al. (19811, and Bennett and Kokocinski (1978), respectively. The variation among values for human seems largely attributable to methodological differences. Assay results in other species are unknown to the authors. 4Similar staining intensity was observed in mature blood and marrow neutrophils in these species. 5Similar values were obtained for other strains of these respective species, the hooded Lister rat and the WCBGF1 J + / +mouse. TABLE 3. Neutrophil iron-binding reactivity (NFeBR) scores in other mammals Order, species NFeBR Score (n)' Marsupiala Didelphys virginiana (opossum) Carnivora Eira barbara (tayra) Lutra canadensis (river otter) Ursus americanus (American black bear) Paradoxurus hermaphroditus (palm civet) Suricata suricatta (meerkat) Panthera tigris altaica (Siberian tiger) Panthera onca (jaguar) Felis concolor (cougar) Artiodactyla Capra hircus (domestic pygmy goat) Odocoileus uirginianus (white-tailed deer) Perissodactyla Equus caballus (Tennessee walking horse) Edentata Dasypus nouemcinctus (armadillo) Rodentia Cauia porcellus (Guinea pig, Charles River strain I1 tricolor) Dasyprocta cristata (golden-rumped agouti) Primates Cercopithecus patas (Patas monkey) Macaca mulatta (macaque) Papio sphinx (mandrill baboon) Papio leucacephaeus (drill baboon) Pongo pygmaeus (orangutan) 284 277 k 6 (3) 316 284 348 420 392 364 400,398 72 105,160 321 k 16(3) 236 112 110 396 385,394 378,425 398 & 18(5) 387 k 18(3) 'Values for three or more subjects are expressed as the mean f 1standard deviation of the mean. bution of FeBR among neutrophils was very homogeneous. Stain deposits observed at the ultrastructural level occurred in increasing order of intensity in the cytoplasmic granules of rat, cow, rabbit, mouse, and human neutrophils (Table 2, Fig. 3). This order corresponded to values of NFeBR scores in the respective species (Table 2). The positively stained granules were generally of an intermediate size range and varied among species, with the smallest population of positively stained granules observed in cows and the largest in rabbits. Values of NFeBR scores in the mouse, guinea pig, and human were also positively related to previously published values of neutrophil Lf for these respective animals (Table 2). Staining with AF alone revealed no positivity in neutrophils (Koszewski et al., 1967; Barton and Parmley, 1986). Values of NFeBR scores in 20 additional 573 NEUTROPHIL LACTOFERRIN IN MAMMALS cytoplasmic FeNTA-AF positivity, similar to that observed with AF technique alone, was observed in some species in some mid- and late erythroblasts, and is consistent with the presence of ferritin in these cells (Parmley et al., 1982; Barton and Parmley, 1986). In light and electron microscopic specimens, eosinophils, basophils, and lymphocytes had no staining; nuclear staining was not observed. The severely FeBR-deficient neutrophils present in some species were readily distinguished from eosinophils and basophils at the light microscopic level by the nuclear configuration and distinctive cytoplasmic granules clearly visualized by the counterstain in the latter cells. Blood leukocyte counts (both differential and estimated absolute counts) and the morphologic and tinctorial features of leukocytes, erythrocytes, and platelets observed on Wright/Giemsa-stainedblood films were similar to those previously described for the same or closely related species (Andrew, 1965; Schalm et al., 1975; Archer et al., 1977; Smith, 1983). Indirect lmmunofluorescence Technique Fig. 4. Autoradiography of detergent-solubilized hamster and human neutrophils and purified human neutrophil Lf using iodinated polyclonal anti-human neutrophil Lf (A) and MoAb-5B2 (B) run under nonreducing conditions. The protein samples loaded on this gel are: lane 1, extract of 7.5 x lo5 hamster neutrophils; lane 2, extract of 7.5 x lo5 human neutrophils; lane 3, 10 pg of purified human neutrophil Lf. After electrophoresis, the proteins were transferred to nitrocellulose paper: lanes 1-3, (A), were incubated with iodinated goat antibody to human neutrophil Lf; lanes 1-3 (B) were incubated with iodinated MoAb-5B2. species are displayed in Table 3. Of all animals studied, carnivores (ten species) and primates (six species) had relatively high cytochemical scores, whereas the artiodactyls (four species) studied had relatively low scores. The neutrophil FeNTA-AF reactivity among rodent species (five species) was highly variable. No definite relationship could be found between the relative quantities of neutrophil Lf estimated by functional cytochemical or indirect immunofluorescence staining and the absolute numbers of granulocytes normally present in the blood of the respective species (Andrew, 1965; Schalm et al., 1975; Archer et al., 1977; Smith, 1983). There are also significant interspecies variations in the contents a n d or activities of neutrophil enzymes (Jain, 1967, 1968; Padgett and Hirsch, 1967; Prieur et al., 1974; Rausch and Moore, 1975). However, the relative amounts of NFeBR and immunoreactive Lf observed in the present study were found to have no obvious relationship to previously reported concentrations of these other substances. Occasional monocytes in some species had weak diffuse FeNTA-AF staining. This was particularly prominent in the human, the Patas monkey, and the mandrill, suggesting the presence of surface andor cytoplasmic Lf in these cells (Bennett and Kokocinski, 1978; Barton and Parmley, 1986). One or two fine cytoplasmic granules similar to those observed with AF technique alone were observed in occasional monocytes in many species, suggesting the presence of cytoplasmic ferritin. Focal Similar results were obtained with polyclonal and monoclonal antibodies. Great variability in immunoreactive neutrophil Lf among species was demonstrable by immunostaining, but there was a homogeneous distribution of staining among neutrophils within the same species andor subject. Fluorescence intensity was positively related to values of NFeBR scores, to the rating of neutrophil granule FeNTA-AF stain deposits assessed at the ultrastructural level, and to concentrations of neutrophil Lf quantified by radioimmunoassay in the respective species studied (Table 2, Fig. 3). lmmunoblotting Technique Extracts of hamster neutrophils showed moderately decreased immunoreactive Lf (Mr 76,000) in comparison with human neutrophil extracts (Mr 78,000 and 80,000) using polyclonal and monoclonal immunoblotting analysis (Fig. 4). Electrophoresis of rat neutrophil extracts had a protein band corresponding to Lf (Mr 76,0001, which required prolonged autoradiography for visualization; because of its faintness, this result is not displayed. Differences among the estimated molecular weights of these Lfs are probably attributable to differences in glycosylation of the polypeptide chains. DISCUSSION The results of this study demonstrate that substantial differences in neutrophil Lf content exist among mammals. Neutrophil immunoreactive Lf and NFeBR semiquantified in the present study were positively related for each species so studied. Both of these parameters were positively related to concentrations of neutrophil Lf previously determined by immunoassay in rabbits, mice, and humans, respectively (van Snick et al., 1974; Segars and Kinkade, 1977; Bennett and Kokocinski, 1978; de Vet and ten Hoopen, 1978; Oseas et al., 1981; Boxer, pers. comm., 1985). Interspecies variation in values of NFeBR scores, particularly low values, could be due to qualitative differences in Lf. Such alterations could include the presence of a single iron-binding site on the Lf molecule, by analogy to certain transferrins (Martin et al., 19841, or other changes in Lf or in its associated granules that simultaneously diminish chem- 5 74 J.C. BARTON ET AL. ical stability, immunoreactivity, and/or iron binding of Lf. However, all known Lfs are relatively thermostable proteins with similar values of molecular weight and charge that have interspecies immunologic cross-reactivity, possess a 2 : l molar ratio of iron:protein binding, and maintain iron binding a t acid values of pH (Masson et al., 1969; Baggiolini et al., 1970;Kinkade et al., 1976). Our present immunoblotting results using polyclonal and monoclonal antilactoferrins demonstrate substantial differences among species in quantities of neutrophil Lf. I n ongoing studies of human neutrophil Lf in a variety of physiologic and pathologic states, we have also found excellent correlations between NFeBR, immunoreactive Lf, and quantities of Lf determined by radioimmunoassay (Barton et al., 1988). Therefore, the corresponding intensities of NFeBR and immunoreactive Lf staining that varied among species are best explained by the occurrence of quantitative interspecies differences in neutrophil Lf. Marked interspecies differences have been observed for a variety of neutrophil granule components, e.g., myeloperoxidase is markedly reduced in cattle (Gennaro et al., 1983) and goats (Rausch and Moore, 1975); lysozyme activity is practically undetectable in Rhesus monkeys, cats, cows, goats, sheep, and hamsters (Rausch and Moore, 1975); and alkaline phosphatase activity is very low in Rhesus monkeys, cats, and mice (Rausch and Moore, 1975). Not surprisingly, the interspecies differences of these enzymes cannot be correlated with the interspecies differences of Lf observed in the present study, suggesting different regulatory mechanisms for synthesis of granule proteins. Carnivores and primates appeared to have greater NFeBR and neutrophil Lf content than herbivores, but a similar correlation with phylogeny has not been observed for neutrophil myeloperoxidase, lysozyme, or alkaline phosphatase contentslactivities. Although the mechanisms that regulate Lf production are poorly understood, the control of Lf synthesis in mammals may similarly affect the neutrophil and the breast glandular cell, because milk Lf concentrations determined for the dog, goat, cow, mare, rat, mouse, guinea pig, rabbit, and human (Masson and Heremans, 1971) generally correspond to the relative amounts of neutrophil Lf in these respective animals observed in the present study. ACKNOWLEDGMENTS The authors recognize the following individuals for their cooperation and efforts in acquisition of many of the blood specimens studied for this project: Drs. Charles R. Becker, Simon Gelman, Larry Boots, Larry Britt, Kenneth Zuckerman, Richard T. Gualtieri, and Ken Boschert, and Mr. Edward Dillard. Mr. Frank Denys assisted in making our photographic results. Ms. Diane Cerna in the Cell Identification Laboratory of the University of Alabama at Birmingham (under the direction of Dr. C.E. Grossi) performed the immunofluorescence staining. Dr. Grossi and Dr. Edgar F. Prasthofer provided advice concerning the immunofluorescence technique. Dr. Mary Ann Accivitti is director of the Hybridoma Core Facility where the monoclonal antibodies were produced. This work was supported by Veterans Administration Medical Research Funds and National Institutes of Health Hematology Training Program Grant 5T32-AMO7488(NIADDKD). T.W.B. is the reciuient of a Frommeyer Fellowship. I LITERATURE CITED Andrew, W. (1965) Mammalia. In: Comparative Hematology. Grune and Stratton, New York, pp. 136-160. Archer, R.K., L.B. Jeffcott, and H. Lehmann (1977) In: Comparative Clinical Haematology. Blackwell Scientific Publications, Oxford. Baggiolini, M., C. de Duve, P.L. Masson, and J.F. Heremans (1970) Association of lactoferrin with specific granules in rabbit heterophil leukocytes. J. Exp. Med., 131.559-570. Barton, J.C., and R.T. Parmley (1986) Light microscopic, non-immunologic demonstration of iron-binding proteins in hematopoietic cells. J. Histochem. Cytochem., 34:299-305. Barton, J.C., W.J. Huster, and R.T. Parmley (1988) Iron-binding reactivity in mature neutrophils: Relative cell content quantification by cytochemical scoring. J. Histochem. Cytochem. (in press). Bennett, R.M., and T. Kokocinski (1978) Lactoferrin content of peripheral blood cells. Br. J. Haematol., 39.509-521. Boxer, L.A., T.D. Coates, R.A. Haak, J.B. Wolach, S. Hoffstein, and R.L. Baehner (1982) Lactoferrin deficiency associated with altered granulocyte function. N. Engl. J. Med., 307:404410. Broxmeyer, H.E. (1984)Negative regulators of hematopoiesis. In: LongTerm Bone Marrow Culture. D.H. Wright and J.S. Greenberger, eds. Alan R. Liss, Inc., New York, pp. 363-397. Daniel. T.O.. W.J. Schneider. J.L. Goldstein. and M.S. Brown (1983) Visualization of lipoprotein receptors b y ligand blotting. J. Biol. Chem., 258:4606-4611. de Vet, B.J.C.M., and C.H. ten Hoopen (1978) Lactoferrin in human neutrophilic polymorphonuclear ieukocytes in relation to iron metabolism. Acta Medica Scand., 203:197-203. Dunn, T.L., W.D. Blackburn, W.J. Koopman, and L.W. Heck (1985) Solid-phase radioimmunoassay for human neutrophil elastase: A sensitive method for determining secreted and cell-associated enzyme. Anal. Biochem., 150:18-25. Engvall, E., and P. Pearlmann (1972) Enzyme-linked immunosorbent antibodies by enzyme-labeled anti-immunoglobulin in antigencoated tubes. J. Immunol., 109t129-135. Gennaro, R., B. Dewald, U. Horisberger, H.U. Gabler, and M. Baggioline (1983) A novel type of cytoplasmic granule in bovine neutrophils. J. Cell Biol., 96:1651-1661. Jain, N.C. (1967) Peroxidase activity in leukocytes of some animal species. Folia Haematol. (Leipzig), 88:297-304. Jain, N.C. (1968)Alkaline phosphatase activity in leukocytes in some animal species. Acta Haematol. (Basel), 39.51-59. Kearney, J.F., and A.R. Lawton (1975) B lymphocyte differentiation induced by lipopolysaccharide. I. Generation of cells synthesizing four major immunoglobulin classes. J. Immunol., 115(3):671-676. Kinkade, J.M. Jr., W.W.K. Miller 111, and F.M. Segars (1976) Isolation and characterization of murine lactoferrin. Biochim. Biophys. Acta, 446:407-418. Koszewski, B.J., H. Vahabzadeh, and S.Willrodt (1967) Hemosiderin content of leukocytes in animals and man and its significance in the physiology of granulocytes. Am. J. Clin. Pathol., 48t474-483. Lieberman, R., W.E. Paul, W. Humphrey, and J.N. Stimpfling (1972) H-2-Linked immune response (Ir)genes. Independent loci for Ir-IgG and Ir-IgA genes. J. Exp. Med., 136t1231-1240. Martin, A.W., E. Huebers, H. Huebers, J. Webb, and C.A. Finch (1984) A mono-sited transferrin from a rearesentative deuterostome: The ascidian Pyura stolonifera (subphylum Urochordata). Blood, 64.1047-1052 Masson, P.L. and J.F. Heremans (1971)Lactoferrin in milk from different species. Comp. Biochem. Physiol., 39B:119-129. Masson, P.L., J.F. Heremans, and E. Schonne (1969) Lactoferrin, an iron-binding protein in neutrophilic leukocytes. J. Exp. Med., 130:643-658. Oseas, R., H.-H. Yang, R.L. Baehner, and L.A. Boxer (1981) Lactoferrin: A promoter of polymorphonuclear leukocyte adhesiveness. Blood, 57:939-945. Padgett, G.A., and J.G. Hirsch (1967) Lysozyme: Its absence in tears and leukocytes of cattle. Aust. J. Exp. Biol. Med. Sci., 45.569-570. Parmley, R.T., M. Takagi, J.C. Barton, L.A. Boxer, and R.L. Austin (1982) Ultrastructural localization of lactoferrin and iron-binding protein in human neutrophils and rabbit heterophils. Am. J . Pathol., 109:343-358. Prieur, D.J., H.M. Olson, and D.M. Young (1974)Lysozyme deficiencya n inherited disorder of rabbits. Am. J. Pathol., 77t283-298. Rausch, P.G., and T.G. Moore (1975) Granule enzymes of polymorphonuclear neutrophils: A phylogenetic comparison. Blood, 46:913919. Schalm, O.W., N.C. Jain, and E.J. Carroll (1975) Normal values in blood of laboratory, fur-bearing, and miscellaneous zoo and wild animals. In: Veterinary Hematology, -_ 3rd ed. Lea and Febiaer. - Philadelphia, pp. 219-283.Segars, F.M., and J.M. Kinkade, Jr. (1977) Radioimmunoassay for NEUTROPHIL LACTOFERRIN IN MAMMALS murine lactoferrin, a protein marker of myeloid and mammary epithelial secretory cell differentiation. J. Immunol. Methods, 14:l- li. Smith, J.E. (1983)Comparative Hematology. In: Hematology, 3rd ed. R.P. McGraw and M. Lerner, eds. McGraw Hill, New York, pp. 117-128. 575 van Snick, J.L., P.L. Masson, and J.F. Heremans (1974) The involvement of lactoferrin in the hyposiderernia of acute inflammation. J. Exp. Med., 140r1068-1084.