DNA Encoding Duffy 9pd Protein

1. Field of the Invention

The present invention relates to the gpD protein, which is the major subunit of the Duffy blood group antigen, and the use thereof in the detection and treatment of malaria.

2. Description of the Related Art

Malaria is the most prevalent infectious disease of mankind. Its widespread geographic distribution together with the severe pathologic consequences of the infection make malaria a major medical and financial burden for many of the developing nations.

There are several different kinds of malaria, one of which is caused by the parasite Plasmodium vivax, which attacks the red blood cells of susceptible individuals. A genetic trait of special interest with regard to P. vivax is the absence of antigens encoded by the blood group system called Duffy (F. B. Livingston, "The Duffy Blood Groups, Vivax Malaria and Malaria Sections in Human Populations: Review", Human Biol., 56, 413, (1984)). It has been shown that individuals whose red blood cells lack the product of the Duffy genes are not susceptible to the penetration of P. vivax owing to the fact that Duffy molecules serve as the receptor for the parasite. (L. H. Miller, H. J. Mason, D. F. Clyde and M. H. McGinnis, "The Resistance Factor to Plasmodium Vivax in Blacks, The Duffy Blood Group Genotype (a-b-)", N. Eng. J. Med., 295, 302, (1976)).

Malarial parasites are transmitted from host to host by blood sucking females of several species of the genus Anopheles. It is in the mosquito that the sexual phase of the life cycle of P. vivax takes place leading to the production of sporozoites. After their introduction into a "new" host, these sporozoites reside in the parenchymal cells of the liver and multiply asexually causing the eventual rupture of hepatic cells and the release of asexual forms (merozoites) into the blood stream. There the merozoites actively penetrate into red blood cells in a nearly synchronous fashion and because the rate of growth and cell division of P. vivax is essentially identical, the infected erythrocytes simultaneously reach the stage of parasite load at which they break. This produces the typical cycles of fever every 48 hours, hence the name of Tertian malaria.

P. vivax infection may persist without treatment for as long as five years. P. vivax parasitemias are relatively low-grade, primarily because the parasites favor the few young red blood cells or reticulocytes that exist in peripheral blood.

Immunity to P. vivax is commonly only partial in nature, which allows the occurrence of superinfections that evolve independently causing an overlap in the cycles of parasite release leading to the appearance of fever in shorter cycles. P. vivax exhibits considerable antigenic diversity and variation, as do other malarial Plasmodia (M. Hommel, Antigenic Variation in Malaria Parasites", Immunology Today, 6, 28, (1985)), although it has been recently shown that antigenic components of P. vivax sporozoites exist that are common to parasites from different isolates (F. Zavala, A. Masuda, P. M. Graves, V. Nussenzweig and R. Nussenzweig, "Ubiquity of the Repetitive Epitope of the CS Protein in Different Isolates of Human Malaria Parasites", J. Immunol., 135, 2790, (1985)).

In the context of the sources of antigenic differences between P. vivax isolates and their consequences with regard to vaccination, it is important that the merozoites of different strains of P. vivax share the same receptor for penetration into red blood cells, i.e., the Duffy molecule (Miller et al., N. Engl. J. Med., supra). In addition, regardless of its capacity to vary other antigenic molecules, the parasite recognition molecule, i.e., the molecule that binds to the Duffy molecule, must remain constant since it is the complementarity between it and the invariant receptor that allows the penetration of merozoites into erythrocytes and, thus, the continuity of the infection. Changes in the ligand specificity of this molecule would result in the loss of the parasite's capacity to infect, since P. vivax merozoites appear to be unable to utilize other human red blood cell receptors for their penetration in vivo, as shown by the resistance of Duffy negative erythrocytes.

The Duffy blood group system consists of two principal antigens Fy.sup.a and Fy.sup.b produced by Fy.sup.a and Fy.sup.b alleles. Antisera anti-Fy.sup.a and anti-Fy.sup.b defined four phenotypes, Fy(a+b-), Fy(a-b+), Fy(a+b+) and Fy(a-b-). W. L. Marsh, Crit. Rev. Clin. Lab. Sci., 5, 387 (1975). Neither antiserum agglutinates Duffy Fy(a-b-) cells, the predominate phenotype in blacks. Antisera defining the other phenotypes, Fy3, Fy4 and Fy5, are very rare. A murine monoclonal antibody, anti-Fy6, defined a new Duffy antigenic determinant present in all Duffy positive cells but absent in Fy(a-b-) cells. M. E. Nichols, P. Rubinstein, J. Barnwell, S. R. de Cordoba, and R. E. Rosenfield, J. Exp. Med., 166, 776 (1987). Blacks with Fy(a-b-) erythrocytes cannot be infected by P. vivax. These cells are also resistant to the in vitro invasion by P. knowlesi, a simian parasite that invades Fy(a+b-) and Fy(a-b+) human erythrocytes. L. H. Miller, S. J. Mason, J. A. Dvorak, M. H. McGinniss and K. I. Rothman, Science, 189, 561 (1985). Receptors for red cell invasion by these parasites, therefore, are related to the Duffy blood group system.

The following amino acids may be indicated by the following 3- or 1-letter codes elsewhere in the specification:

Using an anti-Fy6 monoclonal antibody, there has now been developed a procedure for purification of Duffy antigens in human red cells. Duffy antigens appear to be multimeric red cell membrane proteins composed of different subunits. A glycoprotein, named gpD, of a molecular mass of 35-45 kDa, is the major subunit of the protein complex, and has the antigenic determinants defined by anti-Fy.sup.a, anti-Fy.sup.b, and anti-Fy6 antibodies. The characterization, at the molecular level, of this novel protein is crucial in finding its function on the red cell membrane, in understanding the parasite-erythrocyte recognition process, and in resolving the molecular mechanism of parasite invasion. The present invention relates to the isolation, sequence analysis and tissue expression of a mRNA encoding the gpD protein.

gpD protein has significant sequence homology with human and rabbit interleukin-8 receptors and, therefore, gpD protein most probably is a new class of chemoattractant cytokines receptor. Also, gpD protein cDNA has a quasi-total homology with a human hippocampus cDNA clone HHCMF86 and, therefore, it is highly probable that gpD protein or a homologous protein is present as a neuropeptide receptor in brain. gpD is present in all red cell progenitors and the possibility exists that it may function as a receptor for cell proliferation and/or differentiation. gpD protein cDNA identifies in human kidney a mRNA of the same size as the bone marrow. Since the kidney is not and has no potential to become an erythropoietic organ, it is possible that this putative chemoattractant receptor has essential renal functions.

gpD protein has therapeutic value in the prevention of malaria and in the regulation of essential erythrocyte, neural and renal functions and can be combined with physiologically acceptable diluents to yield a therapeutic agent suitable for these purposes.

Peptides corresponding to a portion of gpD protein that contains the receptor also have been synthesized. Such peptides have therapeutic usefulness identical to that of gpD protein itself and, as is the case with gpD protein, the synthetic peptides can be combined with physiologically acceptable diluents to yield a vaccine against malaria or a therapeutic agent useful to regulate essential erythrocyte, neural and renal functions.

gpD protein and synthetic peptides corresponding to a portion of gpD protein also have utility in the production of therapeutics, e.g., antibodies, complementary peptides, and drugs modeled on the tertiary structure of the gpD protein or synthetic peptides, which are also of therapeutic value in the treatment of malaria and in the regulation of essential erythrocyte, neural and renal functions.

The invention will now be described in detail with reference to the drawings, wherein:

FIG. 1a is the schematic representation and partial restriction map of the two longest gpD protein cDNA clones. FIG. 1b is the schematic of the nucleotide and amino acid sequences of the combined Fyb71-81 cDNA clones encoding gpD protein. FIG. 1b corresponds to SEQ ID NO: 1.

FIG. 2a is a hydropathy plot of the gpD protein sequence. FIG. 2b is a proposed model for the membrane orientation of the gpD protein.

FIG. 3 is Northern (FIG. 3a) and Southern blots (FIG. 3b) probed with either the Fyb71 or Fyb81 insert.

FIG. 4 is a Northern blot analysis of poly(A).sup.+ RNA obtained from human tissues probed with the insert of Fyb81 clone.

FIG. 5 is a chart showing the DNA sequence homology between gpD protein and hippocampus cDNA clone HHCMF86. The nucleic acid sequence encoding HHCMF86 is set forth by SEQ ID NO: 15. The nucleic acid sequence encoding GpD Protein is set forth by SEQ ID NO: 1.

Four cDNA clones that encode gpD protein, the major subunit of the Duffy blood group antigenic system, have been isolated. From these four cDNA clones, the nucleotide sequence of a structural gene encoding gpD protein has been determined. The sequence is shown in FIG. 1 and is designated SEQ ID No: 1. Due to the degeneracy of the genetic code, there may exist other natural DNA sequences that encode the identical gpD protein. The present invention, therefore, extends to such other natural DNA sequences as well as to synthetic DNA sequences having the same DNA sequence as shown in SEQ ID No: 1 and such other natural DNA sequences. The synthesis of DNA can be by any conventional means.

FIG. 1 also shows the amino acid sequence of gpD protein. As was the case with respect to the DNA sequence of the structural gene, the present invention extends to gpD isolated from natural sources or to gpD prepared by chemical synthesis. The chemical synthesis can be by any conventional means.

In FIG. 1, amino acid residues are numbered on the left; nucleotide positions are numbered on the right. The positions of peptides that match predicted amino acids, are shown by solid single lines. The two potential carbohydrate binding sites to asparagine residues, are shown by up arrows. The third glycosylation site, the asparagine at position 37, is unlikely to occur since it is followed by aspartic acid. See R. D. Marshall, Annu. Rev. Biochem., 41, 673 (1972). Double underline at the 5' end is the sequence used to primer extend the 5' end, and at the 3' end is the consensus poly(A) addition sequence.

gpD protein is a highly hydrophobic intramembrane glycoprotein with nine putative transmembrane .alpha.-helices. The cognate gene is present in Duffy positive and negative individuals, but the bone marrow of Duffy negative individuals does not synthesize gpD specific mRNA. In adult kidney, spleen and fetal liver, the mRNA has the same size as gpD mRNA; however, in brain, the mRNA is much larger. The clones that have been characterized will provide the elements to investigate: (i) the structural components of gpD genes, (ii) the biosynthesis and expression of gpD protein in human bone marrow and other tissues, (iii) the structure-function of this novel red cell membrane protein that might exist in other cell types and may function as a chemokine receptor, and (iv) the role of gpD protein as the receptor for P. vivax merozoite invasion.

P. Rubinstein et al., U.S. Pat. No. 5,101,017, have disclosed a monoclonal antibody specific for gpD protein (hereinafter "the Rubinstein antibody"). The Rubinstein antibody blocks the penetration of the P. vivax malaria parasite into human red blood cells by virtue of effective blocking of the target molecule of the P. vivax malaria parasite. It is likely that the Rubinstein antibody has a combining site that has the same stereochemistry as the ligand site on P. vivax, and elicits anti-idiotypic antibodies that react with the parasite. As a result of these properties, the Rubinstein antibody is useful, for example, for the immunodiagnosis of P. vivax infection, to induce anti-idiotypic responses, as alluded to above, that protect against these parasites, and directly in vivo to block the red cell receptors for the parasites. Details of these and other uses of the Rubinstein et al. antibody are disclosed in the patent, the entire contents of which are hereby incorporated by reference.

The Rubinstein antibody was prepared essentially by the method of G. Kohler and C. Milstein (Nature, 256, 495 (1975)). Mice were immunized with human red cells of the Fy(a+b+) type, the spleen removed and hybridized with a suspension of mouse myeloma cell line P3/NSO-Ag4-1 (NS-O) (obtained from the ATCC (American Type Culture Collection, Rockville, Md.)), and the hybridomas tested for the secretion of antibodies that bound to human red cells. One well was found to contain antibodies that bound to erythrocytes of the Fy(a+b-) and Fy(a-b+) types, but not to erythrocytes of the Fy(a-b-) type. The cellular contents of that well were recovered, subjected to dilution and cloned. One of the clones was found to secrete the Rubinstein antibody.

As noted above, the Rubinstein antibody is specific for the gpD protein, now cloned. Accordingly, the inventive gpD protein is also useful to prepare monoclonal antibodies having the same specificity as the Rubinstein antibody. The procedure for preparing such antibodies is essentially the same procedure that Rubinstein et al. employed except that instead of immunizing mice with human red cells, immunization is with gpD protein.

Additionally, the N-terminal (exocellular) region of the gpD protein has been identified as being involved in the interaction of the malaria parasite with the red cell. Work is currently ongoing with synthetic peptides to identify the precise amino acid residues involved in this interaction. The following peptides have been found to bind the Rubinstein antibody in an ELISA assay:

(1) MASSGYVLQAELSPSTENSSQLDFEDVWNSSYGVNDSFPDGDYDAN LEAAAPCHSCNLLDDSALPF, which has been designated SEQ ID No: 8;

(2) MASSGYVLQAELSPSTENSSQLDFEDVWNSSYGVNDSFPDGDYD, which has been designated SEQ ID No: 9; and

(3) AELSPSTENSSQLDFEDVWNSSYGVNDSFPDGDYD, which has been designated SEQ ID No: 10.

The following peptides do not bind to the Rubinstein antibody in an ELISA assay:

(4) DFEDVWNSSYGVNDSFPDGDYD, which has been designated SEQ ID No: 11;

(5) ANLEAAAPCHSCNLLDDSALPF, which has been designated SEQ ID No: 12; and

(6) AELSPSTENSSQL, which has been designated SEQ ID No: 13.

The fact that peptide (3) binds the Rubinstein antibody whereas peptides (4) and (6) do not suggests that the junction between the C-terminal of peptide (6) and the N-terminal of peptide (4) is important for binding. The amino acid sequence AELSPSTENSSQLDFEDVWNSS, which has been designated SEQ ID No: 14, likely contains the epitope for the Rubinstein antibody. Therefore, the invention also extends to peptides comprising the amino acid sequence depicted in SEQ ID No: 14.

The peptides designated SEQ ID No: 8, SEQ ID No: 9, SEQ ID No: 10, and SEQ ID No: 14, as well as peptides containing SEQ ID No: 14, bind the parasite in vivo and, therefore, can be used as the immunogen in a vaccine against malaria. The present invention is, consequently, also directed to such vaccines as well as to a method of protecting a warm-blooded animal, especially, a human, against infection by P. vivax by administering to such animal an effective amount of said immunogen. Since synthetic peptides are typically poor immunogens, it may be advantageous to conjugate the peptides to a carrier, e.g., a protein carrier such as tetanus toxoid, keyhole lympet hemocyanin (KLH), etc., or a lipid carrier such as those taught by R. Neurath et al. in U.S. Pat. Nos. 4,847,080 and 5,204,096, the disclosures of which are hereby incorporated by reference, or by T. Hoppe in U.S. Pat. No. 5,019,383, the disclosure of which is also hereby incorporated by reference.

The inventive peptides can be formed from natural sources, i.e., by proteolytic cleavage of gpD protein, or through chemical synthesis. Details on chemical synthesis of peptides and linkage of peptides to carriers can be found in U.S. Pat. Nos. 4,847,080 and 5,204,096.

In the inventive vaccines, the inventive peptides will usually be present along with a physiologically acceptable diluent (medium), e.g., phosphate buffered saline containing an adjuvant. Generally speaking, the amount of the peptide in the physiologically acceptable diluent will be between approximately 1 .mu.g and 1 mg per dose.

A suitable vaccine can also be formulated using gpD protein in similar amounts in similar diluents, although for most applications the use of synthetic peptides will be more practical.

In either case, the inventive vaccine can be administered by subcutaneous, intravenous, intradermal or intramuscular injection. While the preferred route would depend upon the particular vaccine, intramuscular injection will be generally suitable. Frequency of administration will vary depending on the vaccine.

It is known, for example, from J. H. Adams et al., Proc. Natl. Acad. Sci. USA, 89, 7085 (1992), that the binding regions identified in all Plasmodium malaria proteins are homologous. Therefore, gpD protein and the inventive peptides bind to and interfere with the erythrocyte binding of all Plasmodia. This means that the inventive vaccine will be generally useful against all types of malaria and the invention, thus, extends to a method of protecting a warm-blooded animal against malaria due to any Plasmodia.

gpD protein shows significant homology to interleukin-8 (IL-8) receptors on erythrocytes. This is consistent with a recent report suggesting that the Duffy blood group antigen and the erythrocyte chemokine receptor are the same protein. R. Horuk et al., "A Receptor for the Malarial Parasite Plasmodium vivax: The Erythrocyte Chemokine Receptor", Science, 261, 1182 (1993). The erythrocyte receptor apparently differs from the IL-8 receptors, IL-8RA and IL-8RB on neutrophils. The erythrocyte receptor binds a family of chemotactic and proinflammatory soluble peptides, including IL-8, melanoma growth stimulatory activity (MGSA), monocyte chemotactic protein 1 (MCP-1), and regulated on activation, normal T expressed and secreted (RANTES) protein. Administration of gpD protein (or the inventive synthetic peptides) interferes with the normal binding of these proteins to the erythrocyte receptor and, consequently, is useful to regulate the physiological effects of the secretion of these proteins. For example, it has been postulated that the erythrocyte receptor acts as a scavenger for certain inflammatory mediators, including IL-8. Administration of gpD protein (or the inventive synthetic peptides), therefore, would be expected to enhance scavenging of IL-8, thereby, lessening any IL-8 induced inflammation. For this purpose, the inventive vaccine, as described above, is suitable as a therapeutic agent.

gpD protein also shows significant homology to a human hippocampus cDNA clone HHCMF86 and, therefore, it is highly probable that gpD protein or a homologous protein is present as a neuropeptide receptor in brain. gpD protein cDNA identifies in human kidney with a mRNA of the same size as the bone marrow and, therefore, gpD protein or a homologous protein plays some role in the regulation of renal function. The inventive therapeutic agent will, accordingly, also find use in the regulation of these neural and renal functions.

Proteins that are complementary to gpD protein or the inventive synthetic peptides, e.g., antibodies specific to gpD, will block the natural receptor(s) and, consequently, will also have the therapeutic utilities outlined above. In the preparation of such complementary proteins, the use of gpD protein or the inventive synthetic peptides are of obvious value.

The invention will be further described with reference to the following non-limiting examples.

Red cells (Fy(a-b+)) were washed three times in cold phosphate-buffered saline (PBS) (pH 7.4), resuspended in the same solution, and mixed continuously overnight at 4.degree. C. with the Rubinstein antibody at a concentration of 10 .mu.g/ml of packed red cells. (This concentration determined with radioiodinated antibody, exceeds the concentration required to saturate Duffy antigen sites.) Unbound antibody was removed by washing the red cells with cold PBS. Red cell ghosts were prepared by hypotonic lysis with 20 volumes of cold 5 mM sodium phosphate buffer (pH 7.4) containing 1 mM phenylmethylsulfonyl fluoride and 100 kallikrein-inactivating units/ml Trasylol.TM. (aprotinin). Then the ghosts were washed exhaustively until they were light pink in color. Ghosts were centrifuged for 30 min at 43,000 x g; supernatant was decanted, and the pellet was made to 50 mM Hepes-NaOH, pH 8.0, 1 mM phenylmethylsulfonyl fluoride, 100 kallikrein-inactivating units/ml Trasylol, and frozen at -20.degree. C.

Frozen ghosts were later thawed and centrifuged for 30 min at 43,000 x g. The pellet was resuspended in 50 mM Hepes-NaOH, pH 8.0, 1 mM phenylmethylsulfonyl fluoride, 100 kallikrein-inactivating units/ml Trasylol.TM. to three times the initial volume of packed red cells. Triton X-100.TM. (peroxide free) detergent was added to a final concentration of 1%, and the solution was mixed gently for 1 h at room temperature. Shells were removed by centrifugation for 30 min at 43,000 x g. The supernatant was concentrated 4-fold in an Amicon concentrator using a PM Y10 filter (Amicon Corp.) under nitrogen pressure.

A 0.1 volume of PBS solution, 10 times the normal concentration, was added to the detergent extract. The detergent extract was then incubated with Sepharose 4B.TM. beads coupled to anti-mouse IgG for 1 h at room temperature. The ratio of beads to detergent extract was 1:100 (v/v). The anti-mouse IgG-Sepharose beads were removed by centrifugation, and washed in a solution containing PBS and 0.5% Triton X-100 at a 1:20 (v/v) ratio of beads to washing solution. The washings were done at room temperature and repeated three times. Elution was done by incubating the beads in a solution containing 62.5 mM Tris-HCl (pH 6.8), 0.5% sodium dodecyl sulfate (SDS) at a 1:2 (v/v) ratio beads to eluant. The incubation was at 65.degree. C. for 10 min and repeated three times. The eluted material was concentrated in an Amicon concentrator with PM Y10.TM. filter (Amicon Corp.) under nitrogen pressure.

Polyacrylamide gel electrophoresis (PAGE) in the presence of 0.1% SDS was performed according to U. K. Laemmli, Nature, 227, 680 (1970), with the following modifications: the acrylamide concentration was 10%, polymerization was done overnight to destroy oxidizing reagents, and 0.1 mM thioglycolate was added in the upper chamber. To the concentrated solution of affinity purified material, the following chemicals were added: urea to 4M, SDS to 2%, and .beta.-mercapoethanol to 5%. After electrophoresis, the gels were fixed for 30 min in 10% isoamyl alcohol and 5% acetic acid and stained with 0.002% Coomassie Blue R-250 until marker protein bands were seen. Regions that corresponded between the 36-46 kDa region and above the 96 kDa region were excised, destained with several changes of 5% acetic acid, and washed with distilled water. Gel pieces were stored at -20.degree. C. or used immediately.

Gel pieces, cut into 4.times.4 mm cubes were delivered into the elution chamber of an Elutrap.TM. apparatus (Schleicher and Schuell) and eluted overnight in 50 mM ammonium bicarbonate, 0.1% SDS solution at 100 volts (constant). Fresh 50 mM ammonium bicarbonate, 0.1% SDS solution was added, and electroelution was continued for an additional 6-8 h. Eluted material was concentrated by Centricon.TM. microconcentrator (Amicon Corp.).

The purified protein was alkylated and cleaved with cyanogen bromide (CNBr) as follows: The purified protein was precipitated with cold acetone in the presence of 1 mM HCl at -20.degree. C. for 2 hours. Precipitate was washed with 100% cold acetone, evaporated to dryness at room temperature and was dissolved in 0.1M Tris-HCl, pH 8.0 plus 0.5% SDS. Solid DTT was added to the solution to make final DTT concentration of 10 mg/ml and the solution was reduced for 2 hours at 85.degree. C. One tenth volume of 2.68M Iodoacetic acetamide was added to the solution and the tube was flushed with nitrogen, incubated for 30 minutes at room temperature in the dark. After incubation, solid DTT was added to 10 mg/ml and dialyzed overnight against 0.1M Tris-HCl (pH 8.0) plus 0.5% SDS. Protein was precipitated with acetone as above and air dried. Precipitate was dissolved in 96 .mu.l of 70% formic acid and 4 .mu.l of 1M CNBr solution in 70% formic acid and incubated at room temperature for 48 hours in the dark. Acid was evaporated to dryness and the pellet was washed with water and evaporated to dryness for several times. The digested protein was subjected to either high performance liquid chromatography (HPLC) or polyacrylamide gel fractionation.

Pe 1 peptide was obtained by sequencing the non-fractionated CNBr digest using the O-phthalaldehyde (OPA) blocking reagent (see, A. W. Brauer et al., Anal. Biochem., 137, 134 (1983)). Pe 5 peptide was the partial sequence of the only fragment (.about.4 kDa) that separated very well from the CNBr digest run of the three layer SDS-PAGE system (see, H. Shagger et al., Anal. Biochem., 168, 368 (1987)). After the run, the peptide fragment was electroblotted onto ProBlott.TM. (Applied Biosystems) and sequenced (see, N. LeGendre et al., A Practical Guide to Protein and Peptide Purification for Microsequencing, P. T. Matsudaira, ed., (Academic Press, New York), 49-69 (1989)). Another aliquot was digested with pepsin (50/1 ratio) at 37.degree. C. overnight and the fragments were separated by reverse-phase HPLC using Vydac C-18 column. Pe 2, Pe 3, Pe 4 and Pe 6 peptides, which eluted as single peaks from a reverse-phase HPLC column and were the few pepsin peptides yielded by reverse-phase HPLC, were sequenced. Applied Biosystems Protein/Peptide Sequencer.TM., Model 470 or 477, was used according to the manufacturer's recommendations. Pepsin digestions at 100/1 ratio and 4.degree. C. for 30 or 60 min, did not generate larger peptides.

Pe 5 was the most promising for generating a probe for the selection of gpD protein clones. Pe 2, Pe 3, Pe 4 and Pe 6 peptides were too short for PCR amplification, while the Pe 1 peptide was larger but it had three ill-defined residues.

The nucleotide sequence of the primers (23 mer each) was deduced from the N-terminal and C-terminal amino acid sequences of Pe 5 (see FIG. 1b). Since the Pe 5 peptide was produced by CNBr cleavage, a methionine was included at the N terminus to increase the length of the peptide to 24 residues. Bases were chosen according to the codon preference described by R. Lathe in J. Mol. Biol., 183, 111 (1985), the contents of which are incorporated herein by reference; and deoxyinosine (I) was incorporated at the position where degeneracy exceeded more than three fold except towards the 3' end.

Two generated primers for amino acids at the N-terminal (primer A) and amino acids at the C-terminal (primer B) were chemically synthesized and used to amplify the coding sequence of Pe 5 peptide from pooled human bone marrow mRNA of Fy(a-b+) individuals. Primer A (sense) was specific for residues 245 to 252 (see FIG. 1b) and consisted of 12 fold degeneracy 5'-ATGAAXATYXTITGGGCITGGTT (where I=deoxyinosine; X=C or T; and Y=C, T or A). Primer B (antisense) was specific for residues 261-268 (see FIG. 1b) and consisted of 32 fold degeneracy 5'-ACIAGMAAMTCIAGICCIAMNAC (where I=deoxyinosine; M=A or G; and N=G, A, T or C).

First strand cDNA was synthesized from Fy(a-b+) phenotype mRNA using the preamplification kit from BRL (Bethesda, Md.) and oligo dT as primer. For enzymatic amplification, cDNA, Primer A, Primer B and Taq polymerase (Stratagene) were incubated in a Perkin-Elmer thermal DNA cycler. The amplification product of expected size (72 bp) was subcloned in pBluescript-SK vector (Stratagene). The deduced amino acid sequence of the insert, matched the sequence of Pe 5 peptide (see FIG. 1b). From the sequence of peptide Pe 5 (see FIG. 1b) starting with the amino acid residues W F I and ending with W P H, the oligonucleotide TGGTTTATTTTCTGGTGGCCTCAT was chemically synthesized, .sup.32 P labeled at the 5' end with T4 polynucleotide kinase (New England Biolabs), and used as a probe to screen a human bone marrow cDNA library (see below). The 24 mer oligonucleotide probe having codon usage for amino acids 251 to 258, successfully identified true gpD protein cDNA clones.

Poly(A).sup.+ RNA was isolated as follows: Human bone marrow aspirates were washed and cells were lysed in a solution of 5% beta-mercapoethanol plus 6M guanidine thiocyanate, 25 mM sodium citrate (pH 7.0), 50 mM ethylenediaminetetraacetic acid (EDTA) so that the final guanidine concentration became 5M. The solution was passed through a 25G hypoderminc needle to shear DNA and Sarkosyl was added to 2%. The solution was spun on a 5.7M CsCl, 50 mM EDTA (pH 7.0) cushion in a SW 41 rotor at 32K rpm at 20.degree. C. for 18 hours. The pellet was washed with 6M guanidine hydrochloride and finally with absolute ethanol cooled in dry ice. The pellet was resuspended in diethylpyrocarbonate treated water, adjusted to 0.3M sodium acetate (ph 5.2) and ethanol precipitated. The pellet was resuspended in proteinase K digestion buffer and digested for 2 hours at 37.degree. C., phenol-chloroform extracted and ethanol precipitated. The pellet was then dissolved in water, adjusted to 1.times.DNase digestion buffer and treated with RNase free DNase (BRL). PolyA+ RNA was isolated with mRNA isolation kit FAST TRACK.TM. for Invitrogen according to the manufacturer's protocol. mRNA from white adult liver, spleen, kidney, brain and fetal liver, as well as erythroleukemia cells K562 were obtained from CLONTECH LABORATORIES. DNA was obtained from peripheral blood white cells of the four Duffy phenotypes by lysing the red blood cells in a whole blood unit with 0.83% NH.sub.4 Cl pH 7.4 followed by the standard DNA extraction procedure described by T. Maniatis, E. F. Fritsch, and J. Sambrook, Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Lab., Cold Spring Harbor, N.Y.) (1982), the entire contents of which are hereby incorporated by reference.

A non-amplified human bone marrow cDNA library constructed from pooled mRNA of Fy(a-b+) individuals, was screened with the 24 mer probe. Of 1.9.times.10.sup.6 recombinant .lambda.ZAP II.TM. phage, four positive clones were selected and sequenced. All clones had overlapping sequences but did not fully extend gpD cDNA. Fyb81 of 1085 bp, which was the only clone that included the ultimate 5' end, and Fyb71 of 1083 bp that extended from nucleotide position 185 to the poly(A).sup.+ tail, were the longest. Fyb31 of 989 bp and Fyb82 of 726 bp extended from nucleotide positions 275 and 527 respectively to the poly(A) tail. Combination of Fyb81 with any other clone, generated the full-length cDNA of gpD protein. FIG. 1a shows the overlapping and the combination of the two longest clones.

The joined Fyb71-81 clone predicted an open reading frame that started at position 176 and stopped at position 1192 encoding a polypeptide of 339 amino acid residues (FIG. 1b). A GenBank.TM. sequence search (release 77) at the NCBI using the BLAST.TM. network service, yielded a significant protein sequence homology to human and rabbit interleukin-8 receptors and quasi-total nucleotide sequence homology with a human hippocampus cDNA clone HHCMF86 (see below). The extended product of an antisense primer (from position 57 to 80, FIG. 1b), yielded a sequence of 80 nucleotides which matched exactly with the predicted size at the 5' end of the Fyb81 clone (not shown). At positions 176-178, the initiation codon is not embedded within a sequence context most frequently associated with mammalian translation initiation. See M. Kozak, Nucleic Acids Res., 87, 8125 (1987). It was assumed, however, that it is the true initiation codon for the following reasons: (i) it is the only ATG codon at the 5' end; and (ii) from the first methionine residue, the polypeptide encoded by the combined clones, has the same molecular mass as that of de-glycosylated gpD protein. See, A. Chaudhuri and A. O. Pogo (in press) in Blood Cell Biochemistry, eds. J. P. Cartron and P. Rouger. (Plenum Press, New York) Vol. 6. At the 3' end, clone Fyb71-81 included the consensus poly(A) addition signal AATTAAA (FIG. 1b).

Both clones had a perfect nucleotide sequence match except at the 5' end where several base substitutions yielded six different amino acid predictions. These discrepancies were not a sequencing error, since both DNA strands were sequenced several times. They were a consequence of protein heterogeneity, since the cDNA library was constructed from the mRNA of several Fy(a-b+) individuals.

To establish that Fyb71-81 had a coding sequence specific to gpD protein, the translated sequence was compared with the partial amino acid sequence data obtained from the six peptides Pe 1-Pe 6 (see FIG. 1b). Portions of the predicted amino acid sequence matched with Pe 1 peptide sequenced by the OPA reagent, with four peptides isolated by reverse-phase HPLC (Pe 2, Pe 3, Pe 4 and Pe 6 peptides), and with the Pe 5 peptide isolated by SDS-PAGE. However, two of a total of 62 residues did not match. Thus, the residues at positions 92 and 327 were tryptophan by codon sequence analysis, but they were isoleucine and arginine, respectively, by amino sequence determination. Since tryptophan is a very unstable residue, the discrepancies may be a technical problem in amine acid sequence analysis. On the other hand, they may be due to the heterogeneity of the gpD protein.

Additional evidence that Fyb71-81 encoded gpD protein, was provided by Northern blot and ELISA analysis. Fyb81 did not detect any mRNA in Duffy negative individuals, but it detected a .about.1.27 kilobase (kb) transcript representing the full-length of gpD mRNA in Duffy positive individuals (FIG. 3a). Anti-Fy6 antibody reacted with a 35 mer synthetic peptide (residues 9 to 44, see FIG. 1b), predicted by the Fyb71-81 clone (not shown). The absence of gpD protein specific mRNA in Fy(a-b-) phenotypes (see below) and the reaction of anti-Fy6 with a peptide derived from a gpD cDNA, are strong indications that the clones that were isolated are true Duffy clones.

The predicted translation product of the Fyb71-81 clone, is an acidic protein of isoelectric point 5.65 and molecular mass M.sub.r 35,733. The protein carries at the amine terminal only two potential canonical sequences for N-glycosylation to asparagine residues. See R. D. Marshall, Annu. Rev. Biochem., 41, 673 (1972). This agrees with previous investigations that N-glycosidase F digestion increases gpD mobility on SDS-PAGE and with the chemical detection of N-acetylglucosamine. See A. Chaudhuri and A. O. Pogo, supra; M. J. A. Tanner, D. J. Anstee, G. Mallison, K. Ridgwell, P. G. Martin, N. D. Aventi, and S. F. Parsons, Carbohydr. Res., 178, 203 (1988); and K. Wasniowaska, P. Eichenberger, F. Kugele, and T. J. Hadley, Biochem. Biophy. Res. Commun., 192, 366 (1993).

Predictions of transmembrane helices locations from sequence data using the hydropathy map of Engelman et al. (Ann. Rev. Biophys. Chem., 15, 321 (1986)) and a scanning window of 20 residues show that the bulk of the protein is embedded in the membrane (FIG. 2). Nine transmembrane .alpha.-helices, a hydrophilic domain of 66 residues at the N-terminus, a hydrophilic domain of 25 residues at the C terminus, and short protruding hydrophilic connecting segments, were predicted. The pair of helices, D and E, is so closely spaced that they may be arranged as coupled anti-parallel helices. A schematic illustration of gpD protein topology, is shown in FIG. 2.

The charge-difference rule proposed by Hartmann et al. (Proc. Natl. Acad. Sci. USA, 86, 5786 (1989)), predicts that the N terminus is on the exocellular side and the C terminus of the protein is on the cytoplasmic side of the membrane. The N-terminal prediction is validated by the finding of the two potential N-glycosylation sites on the N-terminus. Moreover, the reaction of anti-Fy6 with a synthetic peptide deduced from this domain, establishes its exocellular location experimentally since the antibody binds to erythrocytes. The signal-anchor sequence for membrane insertion probably lies in the first transmembrane .alpha.-helix that follows the N-terminal domain. See H. P. Wessels and M. Spies, Cell, 55, 61 (1988); and G. Blobel, Proc. Natl. Acad. Sci. USA, 77, 1496 (1980). From there on, the protein is deeply buried in the membrane and exists at residue 314 on the cytoplasmic side of the membrane (FIG. 2). The topological predictions of helices, hydrophilic connecting segments and the location of the C-terminal fragment, should be substantiated by direct biochemical and immunochemical analysis.

Duffy gpD protein is deeply buried in the membrane like the membrane-associated fragment of Band 3 (see D. Jay, Annu. Rev. Biochem., 55, 511 (1986)), the human blood group Rh polypeptide (see, B. Ch erif-Zahar et al., Proc. Natl. Acad. Sci. USA, 87, 6243 (1990); and N. D. Avent et al., Biochem, J., 271, 821 (1990)), bacteriorhodopsin (see P. Carlton et al., EMBO J., 4, 1593 (1985)), and lipophilin (see W. Stoffel et al., Hoppe-Seyter' Z. Physiol. Chem., 364, 1455 (1983)). The significant homology of gpD protein with interleukin-8 receptors is very intriguing. See W. E. Holmes et al., Science, 253, 1278 (1991); and P. M. Murphy et al., Science, 253, 1280 (1991). If dpD protein bind chemokines and has the ability to activate a signal transduction cascade, this gives rise to gpD protein as a new class of proinflammatory mediators. Thus, gpD protein is not present in white blood cells, since a rabbit polyclonal antibody (anti-gpD) against purified and denatured gpD protein that reacts with erythrocytes and their precursors does not react with any white blood cells (unpublished results).

Poly(A).sup.+ RNAs were run on formaldehyde/agarose gel and transferred onto Hybond.TM. N+ nylon membranes (Amersham Corp.). They were hybridized in QuickHyb.TM. (Stratagene) and washed according to the manufacturer's instructions.

On Northern blot analysis Fyb71 or Fyb81 clone detected a .about.1.27 kb mRNA species in the bone marrow of the three Duffy positive phenotypes but not in individuals of Fy(a-b-) phenotype (FIG. 3a). The absence of gpD mRNA was consistent with the absence of gpD protein in Duffy negative individuals. Anti-gpD antibody did not react with any red cell membrane protein of Fy(a-b-) erythrocytes (not shown). Duffy negative individuals did not express gpD protein, because they do not synthesize Duffy specific mRNA.

In FIG. 3a, lane 1 contained 10 big of Fy(a-b-) mRNA, lanes 2 and 3 contained 5 .mu.g of Fy(a+b-) mRNA and Ry(a-b+) mRNA respectively, and lane 4 contained 2 .mu.g of Fy(a+b+) mRNA. They were resolved on a 2% denaturing agarose gel, blotted, hybridized, and autoradiographed for 72 h at -80.degree. C. RNA size markers shown: human 28S (5.1 kb) and 18S (2.0 kb) rRNA, and the 1.35 kb GIBCOBRL marker (LIFE TECHNOLOGIES), were used to calculate the size of gpD mRNA. The actin probe at the bottom was used as control of sample loading. RNA integrity was indicated by the presence of the two rRNA in the poly(A)+ fraction and the actin probe.

All restriction enzyme digestions were performed according to the conditions suggested by the supplier (New England Biolabs). Digested DNA was size-fractionated on 0.8% agarose gel and blotted as described for Northern analysis. Hybridization in QuickHyb.TM. solution was carried out at 68.degree. C. for 1 hour according to the manufacturer's instructions.

On Southern blot analysis Fyb71 or Fyb81 probe hybridized with DNA of Duffy positive and negative individuals (FIG. 3b). They identified a single band of 6.5 kb in Bam HI, two bands of 12 kb and 2 kb in EcoRI, and two bands of 3.5 kb and 1.4 kb in Pst I digested DNA. These findings agree with the restriction map of the Fyb71 and Fyb 81 clones and show a single copy gene. Determination of the structural differences among the genes of Duffy positive and negative individuals, should clarify the mechanism of gpD gene repression in negative individuals. A functional silencer element described in other systems may selectively repress transcription of gpD gene in the erythrocytes of Fy(a-b-) individuals. See L. Li, T. Suzuki, N. Mori, and P. Greengard, Proc. Natl. Acad. Sci. USA, 90, 1460 (1993). The Duffy system is different from the ABO (F. Yamamoto et al., Nature, 345, 229 (1990)) and Kell systems where mRNA has been found in individuals who do not express the blood group determinants.

In FIG. 3b, each lane contained 10 .mu.g of digested DNA; lanes 1-4 contained Fy(a-b-), lanes 5-8 contained Fy (a+b-), and lanes 9 to 12 contained Fy(a-b+) DNA. The enzyme digestions were as follows: lanes 1, 5 and 9 Bam H1; lanes 2, 6 and 10 Eco R1; lanes 3, 7 and 11 Hinf1; and lanes 4, 8 and 12 Pst 1. They were resolved on 0.8% agarose gel, blotted, hybridized, and autoradiographed for seven days at -80.degree. C. Sizes were calculated from the positions of the GIBCOBRL DNA markers.

As indicated in FIG. 4, a 1.27 kb mRNA species was found in adult spleen and kidney, fetal liver but not in adult liver and K562 erythroleukemia cells. Hybridization with the .beta.-globin probe showed a strong signal in bone marrow and fetal liver; it showed a weak signal in adult spleen but no signal in adult liver, brain and kidney (not shown). The presence of gpD mRNA in fetal liver was expected since the fetal liver is an erythropoietic organ. In human brain, a strong band of 8.5 kb and a faint band of 2.2 kb were detected. This bears some interesting possibilities. It indicates that in brain, there is a Duffy related protein. Also, this idea is supported by the quasi-total homology between Fyb71-81 clone and a human hippocampus cDNA clone HHCMF86, which was recently identified (see M. D. Adams et al., Nature, 355, 632 (1992)) and is designated SEQ ID No: 15. However, it is unlikely that the 8.5 kb brain mRNA codes for a Duffy protein with long 5' and 3' untranslated sequences. It is possible that the brain mRNA codes for a larger protein which has extensive homology with gpD protein. The homologies of these mRNA species with gpD specific mRNA, remain to be demonstrated by sequence analysis; however, the findings strongly indicate that gpD protein or a similar protein is produced in kidney, non-hemopoietic spleen cells, and probably in brain.

A chart showing the DNA sequence homology between gpD protein and human hippocampus cDNA clone HHCMF86 is presented as FIG. 5. The HHCMF86 cDNA clone was obtained from a 2 year old Caucasian female (Adams et al., Nature, 355, 632 (1992)). There are several undefined bases in HHCMF86 and the two clones have the same ORF up to position 623 in gpD protein (290 in HHCMF86). HHCMF86 has an extra adenine after this position producing a frame shift in the ORF. It is likely that the extra adenine may be a sequence error in HHCMF86 cDNA.

In FIG. 4, lanes 1, 3, 5 and 7 contained 2 .mu.g of Fy(a-b+) bone marrow, fetal liver, adult spleen and erythroleukemia (K562) mRNAs respectively. Lanes 2, 4 and 6 contained 7 .mu.g of total brain, adult liver, and adult kidney mRNA, respectively. They were resolved on a 1.5% denaturing agarose gel and autoradiographed for five days at -80.degree. C.

A mixture of mRNA of several Fy(a-b+) individuals, the BRL Superscript Choice.TM. System and oligo dT as a primer, were used to prepare cDNA. The cDNA was ligated into .lambda.ZAP II.TM. vector and packaged with Gigapack Gold.TM. (Stratagene) extract. About 1.9.times.10.sup.6 unamplified cDNA clones were screened with the .sup.32 P-labeled probe described above. cDNA inserts into pBluescript were isolated by the plasmid rescue method according to manufacturer's protocol. Both DNA strands were sequenced using vector primers; and primers were designed from the sequenced regions of the transcript.

A .sup.32 P-labeled 24-mer antisense primer from nucleotides 57 to 80 of the coding strand (FIG. 1b) was extended on Fy(a-b+) mRNA using the Preamplification Kit (BRL) and the products were separated on a 6% sequencing gel. The M13 sequence ladder was used to determine sizes of products.

It will be appreciated that the instant specification and claims are set forth by way of illustration and not limitation, and that various modifications and changes may be made without departing from the spirit and scope of the present invention.