Isolation and Characterization of the Receptor on Human Neutrophils That Mediates Cellular Adherence*

The receptor on human neutrophils (polymorphonu- clear leukocytes or PMN) that mediates cellular adherence has been purified from the peripheral blood PMN obtained from an individual with chronic myelogenous leukemia (CML). This receptor consists of two noncovalently associated subunits, designated aM (Mac- la, CDllb) (Mr = 170,000) andB (Mac-18, CDwlS) (Mr = 100,000), respectively, which are identical on normal and CML PMN. The subunits were purified by mono- clonal antibody 60.1-Sepharose (anti-aM) affinity chromatography and separated in 5-nmol quantities by high pressure liquid chromatography on a TSK-4000 gel filtration column. Subunits were characterized by amino acid composition, NHz-terminal amino acid sequence, and carbohydrate content. The NHz-terminal sequence of the human PMN aM subunit contains regions of homology with the human platelet glycoprotein IIba. We conclude that nanomole amounts of indi- vidual (YM and subunits of the receptor on human PMN that mediates cellular adherence can be isolated and separated using CML PMN. To perform in the human neutrophil (polymorphonuclear

Two observations indicate that the aM . p complex plays a critical role in PMN physiology. First, monoclonal antibodies (mAb) directed against the subunits of this complex block PMN adherence-related activities (1)(2)(3)(4)(5)(6)(7)(8). Second, the PMN from children whose leukocytes are severely deficient in the member subunits of this glycoprotein family display profound defects in PMN adherence and adherence-related functions, including phagocytosis, in in uitro assays (3,9,10). Clinically, these children suffer from recurrent and severe bacterial infections that frequently culminate in death (3,9,10).
Two members of this leukocyte family, a L (LFA-1) and aM (Mac-1), have been isolated from the murine cell lines EL-4 and P388D1, respectively (11). Sequence analysis indicated that the terminal 19 residues of the two subunits contained 33% homology (11). Thus, the genes for these proteins might have risen by gene duplication from a primordial gene. The NH2-terminal amino acid sequence of the human platelet adherence glycoprotein IIba was demonstrated recently to contain regions of similarity to both murine a L and murine aM (12). Preliminary molecular characterization of these antigens, obtained by transfection of L-cells, suggested that the genes for aL and aM and platelet glycoprotein IIb/IIIa were located within a 20-kilobase segment of human DNA (13).
Purification of nanomole amounts of the PMN adherence glycoproteins aM and p for structural-functional relationships has proven difficult. In this report we describe the purification of nanomole amounts of both subunits from the peripheral blood PMN of an individual with chronic myelogenous leukemia (CML). Purification was performed with a one-step method using mAb 60.1-Sepharose (anti-aM) affinity chromatography. The individual subunits were separated with a TSK-4000 gel filtration column using high pressure liquid chromatography (HPLC). Sequence analysis indicated that the murine and human aM subunits are homologous and that the human a M subunit is related to human platelet glycoprotein IIba.
Gel Electrophoresis-Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed according to the method of Laemmli (14).
Cell Preparation-Neutrophils (PMN) were obtained from the peripheral blood of normal individuals and two individuals with chronic myelogenous leukemia (CML/chronic phase) (15) according to standard procedures (16). HL-60 promyelocytic leukemia cells (provided by Dr. Steven J. Collins, Fred Hutchinson Cancer Research Center, Seattle, WA) were grown in RPMI 1640 (Irving Scientific, Santa Ana, CA) supplemented with 10% heat-inactivated fetal bovine serum (Hyclone Laboratories, Logan, UT) (17). HL-60 cells were induced to differentiate by incubation for 3 days in the presence of 1.25% dimethyl sulfoxide (Me,SO; Sigma) (18).
Analysis of aM and j3 Subunits by Immunofluorescence and Immunoprecipitation-Indirect immunofluorescence was used to estimate the expression of the aM and subunits on the PMN from normal and CML individuals (6, 7) and on HL-60 human promyelocytic leukemia cells. Immunofluorescence of individual cells was quantitated using an Ortho 50H cytofluorograph interfaced to a model 2150 Computer (Ortho Diagnostic Instruments, Westwood, MA).
of normal PMN, CML PMN, and HL-60 cells. To create the cell The a M and j3 subunits were immunoprecipitated from cell lysates lysates, cells (5 x IO7) were solubilized by incubation at 4 'C for 30 min in a cell lysis buffer consisting of 0.5% Nonidet P-40, 0.15 M NaCl, 0.05 M Tris, 2 mM phenylmethylsulfonyl fluoride (Sigma), 50 mM iodoacetamide (Sigma), 1 p~ pepstatin (Sigma), and 1 mM diisopropyl fluorophosphate (Sigma) at pH 8.3. Nuclei were removed by centrifugation at 5000 X g at 4 "C for 30 min. The supernatant was further clarified by centrifugation at 100,000 X g at 4 "C for 60 min. The aM and j3 subunits were immunoprecipitated from the cell lysate by incubating the lysate with 0.1 ml of a 1:l liquidsolid solution of mAb 60.1-Sepharose at 4 "C for 60 min. The immunoprecipitates were collected by centrifugation and washed eight times with buffer containing 0.5% Nonidet P-40, 0.45 M NaC1, 0.05 M Tris, 0.1% SDS, at pH 8.3. Prior to gel electrophoresis samples were reduced with 5% (v/v) 2-mercaptoethanol at 100 "C for 5 min. Proteins were visualized by Coomassie Blue staining of the gels.
Purification of the aM and j3 Subunits-For purification of quantities of aM and j3 greater than 0.1 mg, the starting material consisted of 10-50 g of PMN obtained by leukapheresis of individuals with CML and leukocytosis (white blood cell count > 50,000/mm3). The supernatant (200-800 ml) from the final centrifugation at 100,000 X g was applied directly to a 2-ml mAb 60.1-Sepharose column at a flow rate of 15 ml/h. Following washing with cell lysis buffer at pH 8.0, the column was washed with 0.010 M Tris, 0.1% deoxycholate, 2 mM phenylmethylsulfonyl fluoride. The column was eluted with 0.050 M triethylamine, 0.1% deoxycholate, 2 mM phenylmethylsulfonyl fluoride, pH 11.5, at a flow rate of 15 ml/h. Preliminary studies, using a gradient of increasing pH, had demonstrated that the majority of aM/j3 eluted from mAb 60.1-Sepharose at pH 11.5 (19). Fractions of 1.8 ml were collected into 0.2 ml of 1 M Tris, pH 6.8. The protein content of fractions was estimated by measuring absorbance at 280 nm (Arn) and fractions with an Am 5 0.1 were pooled and dialyzed overnight at 4 'C in 0.001 M Tris, 0.01% deoxycholate, pH 8.2. Following dialysis, the eluate was lyophilized.
The aM and j3 subunits were separated using HPLC. The dialyzed and lyophilized eluate from the mAb 60.1-Sepharose column was dissolved in 6 M guanidine containing 0.050 M dithiothreitol at 37 "C for 4 h and alkylated by incubation with iodoacetic acid for 25 min. The reduced and alkylated proteins were dialyzed, lyophilized, resuspended in 2% SDS to achieve a final concentration of 1 mg of protein/ ml of 2% SDS, and injected onto a TSK-4000 preparative (2.1 X 60cm) gel filtration column at a flow rate of 2 ml/min. Eluted proteins were detected by absorbance at 280 nm. Fractions encompassing the peaks were pooled, lyophilized, and analyzed for purity using SDS-PAGE followed by Coomassie Blue and silver staining. Fractions of aM and j3 collected from the initial HPLC were subjected to a repeat HPLC separation procedure.
The amino acid compositions of the peptides were determined on acid hydrolysates of the samples using an updated single column Beckman 120C automatic amino acid analyzer (20). NH2-terminal amino acid sequence of the aM and j 3 subunits was determined by automated Edman degradation with the use of an Applied Biosystems (Foster City, CA) model 470A gas-phase sequenator (20).
Neutral and amino sugars were determined by gas chromatography as described (21). Analyses were performed with a Varian 3700 gas chromatograph with an OV 101 WCOT column and flame ionization detector using the on-column technique of Grob and Grob (22).

Expression of PMN Adherence Complex on Human PMN, CML PMN, and HL-60 Promyelocytic Leukemia Cells-In
initial studies we analyzed the expression of the a M and p subunits on the peripheral blood PMN from normal individuals, the PMN from two individuals with CML, and myeloid cells of the HL-60 human promyelocytic leukemia cell line.
Subunit expression was quantitated using subunit-specific mAb and indirect immunofluorescence followed by flow cytometry. The mean fluorescence of each subunit was nearly identical on normal human PMN and CML PMN ( Table I). The a M subunit was not expressed on undifferentiated HL-60 cells, but it was expressed following differentiation with Me2S0 (23). However, the surface density of a M on Me2SOtreated HL-60 cells never reached the levels present on normal or CML PMN (Table I). These experiments, which were conducted to determine a source of each subunit for largescale purification, indicated that CML PMN were a potential source of both a M and p.  Purification and Separation of the aM and j3 Subunits-mAb 60.1-Sepharose (anti-aM) affinity chromatography was utilized to purify both subunits from a cell lysate created from 25 g of CML PMN ("Materials and Methods"). Localization of the aM and j3 subunits to discrete fractions of the eluate is shown (Fig. 1). Proteins eluted from the mAb 60.1-Sepharose column were injected onto a preparative TSK-4000 HPLC gel filtration column and five discrete peaks were recorded by measurement of Apno (Fig. 2). The aM and j3 subunits eluted in peaks designated A and B, respectively (Figs. 2 and 3). Coomassie Blue stainingof an overloaded gel (Fig. 3, overload) as well as silver staining of an SDS-PAGE gel (data not shown) both demonstrated that a M (170 kDa) contained approximately 5% of j3 (120 kDa), and vice versa. The fractions containing the aM and j3 subunits in the initial HPLC gel filtration were each subjected to repeat HPLC gel filtration to further purify each subunit (Fig. 2, panels A and B ) . The protein peaks collected from the second HPLC run were used to determine amino acid composition, NH2-terminal amino acid sequence, and carbohydrate content.
Amino acid compositions were determined for each subunit. The number of residues for a M was estimated from the M , of the aM glycoprotein (170,000) and the aM protein (130,000) (24). Similarly, the number of residues for j3 was estimated from the M , of the @ glycoprotein (100,000) and the j3 protein (72,000) (24). Using this data the number of residues for a M was 1169 and the number of residues for B was 639. The absorption coefficient E was estimated from the absorbance of purified protein in solution and the amino acid content; for the aM subunit E = 5.49 and for j3 E = 8.70 (25).
Results of the amino acid composition of crM and ( 3 indicated that the proteins were similar to each other (Table 11). The remarkable difference lies in the increased cysteine content in j3 compared to aM. This high cysteine content of @ is consistent with the observation that the higher M , of this protein on SDS-PAGE after reduction is due to the presence of intra-chain disulfide bonds.
To define these two proteins more precisely, sequence analysis was performed. Comparison of 13 NH2-terminal amino acid residues of the human a M subunit with the murine aM (Mac-l) (11) indicated only two substitutions (Table 111). In the human sequence Asn replaced His at position 7 and Ala Approximately 1000 pg of reduced, alkylated, "C-labeled protein from the mAb-Sepharose column was dissolved in 2% SDS and injected onto a TSK-4000 preparative (2.1 X 60-cm) gel filtration column a t a flow rate of 2.0 ml/min under conditions described in the text. Eluted proteins were detected by absorbance a t 280 nm. Fractions encompassing the peaks were pooled and lyophilized. The two tracings at the right were obtained when the peaks labeled A and B from the first HPLC run were subjected to repeat chromatography using the same scales for elution volumes and Am. replaced Pro at position 8, although some ambiguity may exist in assigning Pro at position 8 of the murine protein (11). In addition, human PMN aM contained approximately 40% sequence homology with another hematopoietic cell adherence glycoprotein, that of platelet glycoprotein IIba (12). No amino acid sequence could be obtained from 2 nmol of PMN p subunit, suggesting the presence of a blocked NH2 terminus.

FRACTIONS OVERLOAD
The carbohydrate composition for a M and j3, respectively, were: fucose 1.5 and 1.2, mannose 16.5 and 3.1, galactose 8.4 and 2.6, and N-acetylglucosamine, 7.1 and 13.6, expressed as nanomole of carbohydrate/nanomole of protein. Sialic acid residues were present in both subunits but were not quantitated.

DISCUSSION
We purified and separated the aM and /3 subunits of the receptor on human PMN that mediates cellular adherence using a cell lysate created from the peripheral blood PMN of an individual with CML. The initial isolation utilized mAb 60.1-Sepharose (anti-aM) affinity chromatography. Separation of the subunits required HPLC using a TSK-4000 gel filtration column. The subunits were characterized by amino acid composition, amino acid sequence, and carbohydrate content. Initial studies were designed to identify a cell line or cell population capable of serving as a source of antigen for largescale purification. PMN from individuals with CML/chronic phase (15) were demonstrated to express the aM and /3 subunits in amounts identical to those present on normal PMN. Our experiments using immunoprecipitations from PMN obtained from normal individuals and those with CML/ chronic phase demonstrated that the aM and / 3 subunits were identical as assessed by one-dimensional SDS-PAGE. Individuals with CML/chronic phase provided an excellent source of starting material for antigen preparation since: 1) they !iating Cellular Adherence 5579 maintain peripheral white blood cell counts 5-20 times greater than normal individuals; 2) their cells are primarily mature PMN; and 3) they are generally in otherwise good health, thus tolerating leukapheresis for removal of white blood cells (15). mAb 60.1-Sepharose affinity chromatography provided a one-step purification of aM and / 3 using a cell lysate created from 25 g of PMN. Although milligram quantities of protein could be eluted from the mAb 60.1-Sepharose column, the individual aM and /3 subunits could not be separated using conventional gel filtration. The reduced and alkylated aM and /3 subunits were resolved using a TSK-4000 HPLC gel filtration column. The aM and /3 subunit each contained less than 5% contamination with the other subunit after gel filtration. By subjecting the fractions of aM and /3 collected from the initial HPLC to a second HPLC separation, a further purification was achieved. Proteins isolated from the second HPLC filtration were used to analyze amino acid compositions, NH2-terminal amino acid sequence, and carbohydrate content.
Sequencing of the NH, terminus of the aM subunit provided insight into the relationship of this subunit to other adherence glycoproteins. The aM subunit contained only two substitutions (at residues 7 and 8) when compared to the murine aM (Mac-1) (11). Both substitutions were relatively conserved, single base pair changes. Furthermore, nearly 40% sequence homology was demonstrated between human PMN aM and human platelet IIba. This relationship between platelet IIba and murine aM (Mac-1) was previously described by Charo et al. (12).
The carbohydrate compositions do not account for the molecular weight difference between the glycosylated and nonglycosylated subunits described for the murine proteins (24). This result may arise due to somewhat low recovery of carbohydrate from subnanomolar amounts of glycoprotein, or from the spuriously slow migration of glycoprotein subunits during SDS-PAGE.
The development of molecular probes for these proteins will be required for more extensive comparisons of the adherence glycoproteins of both mouse and man. In addition, knowledge of the molecular structure of these antigens will facilitate understanding of their structure-function relationships.