CA2199830C - Expression and export technology of proteins as immunofusins - Google Patents

Abstract

Disclosed are DNAs produced by recombinant techniques for inducing the expression and subsequent secretion of a target protein. The DNAs encode, in their 5' to 3' direction, a secretion cassette, including a signal sequence and an immunoglobulin Fc region, and a target protein. The DNAs can be transfected into a host cell for the expression, production and subsequent secretion of the target protein as a fusion protein. The secreted protein ca n be collected from the extracellular space, and further purified as desired. The secreted fusion protein additionally can be proteolytically cleaved to release the target protein from the secretion cassette.

Description

C.~ 02199830 1999-(0-l3 Ths invention relates to fusion protein expression syacems for use in mammalian cells that enhance the production of a given target protein. Mare opecifically, the invention relates to a secretion 10 cassette, eornpriaad of a mammalian signal peptide and a portion of mammalian inmunoglobulins. xhich, xhen used as the amino-terminal fusion partner to the target protein,,ganarally leads to high level expression and secretion of the lusion product. Such fusion protaina are useful, for example, Lor the production and extracellular 15 collection of target proteins without the need fo=' lysis of a host cell. The inventloa is perhaps moat uaetul for the expression of target proteins which are not normally secreted from a best cell, are secreted at low levels from a host cell, or are toxic ar otherwise deleterious to a hone cell.

Exprsssion systems employing gene fusion constructs have been used to enhance the produetien of proteins in bacteria. Employing a bacterial protein that is normally expressed at a very high level as the amino-teztainal fusion partner of a fusion protein helps co ensure 25 efficient transcription and translation of the message, and is some cases the secretion and solubilizacivn of the fusion protein tSmith and Johnson (19A8) Grease 69:311 Hopp et al. (19981 Biotechnology 6:1204; La Vallie et al. (1993) Biotechnology 11:187).
30 The major goal o! expression of zecomninant fusion proteins in mammalian cells has been to coaster novel properties to the hybrid molecules, e.g.. targeting at a cytokine or toxin inin vivo, Fc receptor binding, complement fixation, protein A binding, increasing the half lite, and crossing the blood-brain barrier. Examples of recombinant 35 fusion proteins produced in mammalian calls include cytokine C:A U21998:3U 1999-lU-13 immunoconjugate~ (Gillie~ et al.) (1992) Broc. Nat!. Aced.
Sci. USA 89:1428: Gillies et ai. (1993) Bioconjugate Chemistry 4:230), immunoadhesins (Capon et al. 1,1989) Nature 337:525), irnmunotoxins (Chaudhary et a1. (19890 Nature 339:394), and a nerve growth factor conjugate (Friden et al.
11993) Science 259:373). Proteins produced in mammalian cells often do not have the solubility and secretion problems encountered in bacterial expression. The use of gene fusion Constructs to enhance the production or secretion of a target protein in a mammalian system has not been explored fully.
It is the object of the invention to provide DNAs which facilitate the production and accretion of a target protein. In particular, objects of the invention are to provide novel DNAs which: facilitate IS efficient production and secretion of hard to express proteins, such as nuclear proteins, regulatory proteins and proteins waich otherwise may be toxic to a host cell, and can be adapted to any target polypeptide of interest which can be coded for and expressed in a boat organism; to provide DNA constructs for the rapid and efficient production and secretion of protaina in a variety of host cells; and to provide a method for the production, secretion and collection of genetically engineered proteins. including non-native, biosynthetic, or otherwise artificial proceina, such as proteins which have been created by rational design. Other objects o! the invention are to provide DNA
sequences which, rhea !used to a polynucleotide encoding a target protein. encode a fusion polypeptide which can be purified using common reagents and techniques, and to interpose a proceolytic cleavage site between the encoded secretion cassette and the encoded target protein such that the accretion cassette can be cleaved from the target protein and the target protein can ba purified independently. Still another object is to provide a procedure which is both efficient and inexpensive.
These and other objects of the invention will be apparent from the description, drawings, and claims that follow.

WO 96/085?0 PGTIilS93111720 ,g~mm~y Of The Invention The present invention features a DNA of general applicability for production and secretion of fusion proteins. The DNA comprises a S secretion cassette, as the amino-terminal fusion partner, and a target protein, and is termed herein an "immunofusin". The invention provides, in its various aspects, a recombinant DNA encoding the immunofusin, and methods of producing the encoded immunofusin protein.
The immunofusin is a DNA which comprises a polynucleotide encoding s IO secretion cassette, comprising in its S' to 3' d:i.rection a signal sequence and an immunoglobulin Fc region, and a polynucleotide encoding a target protein fused to the 3' end of the secretion cassette. A
secretion cassette of the invention, once constructed, can be fused to various target proteins. Additionally, one can optimize the sequences IS which regulate the expression of a secretion cassette, and hence the expression of the immunofusin. The resultant DNA can be expressed at high levels in a host cell, and the fusion protein is efficiently produced and secreted from the host cell. The secreted immunofusin can be collected from the culture media without the need for lysis of the 20 host cell, and can be assayed for activity or purified using common reagents as desired.
The portion of the DNA encoding the signal sequence preferably encodes a peptide segment which directs the secretion of the 25 immunofusin protein and is thereafter cleaved. As used in the specification and claims, "immunoglobulin Fc regi.on" means the carboxyl-terminal portion of an immunoglobulin heavy chain constant region. As is known, each immunoglobulin heavy chain constant region is comprised of four or five domains. T'he domains are named 30 sequentially as follows: CH1-hinge-CH2-CH3(-CH4), and the Fc region of each immunoglobulin subclass lacks at least the fHl domain. As is apparent from a review of the DNA sequences of th.e immunoglobulin . subclasses, the DNA sequences of the heavy chain domains have cross-hvmology among the immunoglobulin classes, e.g., the CH2 domain of IgG
35 is homologous to the CH2 domain of IgA and IgD, and to the CH3 domain of IgM and IgE. The portion of the DNA encoding the immunoglobulin Fc SUBSTITUTE SHEET RULE Zfi~

219~83~
WO 9b/08570 PGTIUS95/1I720 region preferably comprises at least a portion of a hinge domain, and a CH3 domain of Fcy or the homologous domains in any of IgA, IgD, IgE, or IgM. The immunoglobulin Fc region also preferably comprises at least a portion of the DNA encoding a hinge and a CH3 domain of Fcy or the homologous domains in any of IgA, IgD, IgE or FgM.
The currently preferred accretion cassette is a polynucleotide encoding, in its S' to 3' direction, the signal sequence of an immunoglobulin light chain gene and the Fcy1 region of the human immunoglobulin y1 gene. The Fcyi region of the immunoc~lobulin y1 gene includes at least a portion of the hinge domain and CH3 domain, or at least a portion of the hinge domain, CH2 domain and CH3 domain. The DNA encoding the secretion cassette can be in its genomic configuration or its cDNA configuration. However, the studies described below use secretion cassette in the genomic configuration. The use of human Fcyi as the Fc region sequence has several advantages. For example, if the fusion protein is to be used as a biopharmaceutical, the Fcyl domain may confer the effector function' activities to the fusion protein. The effector function activities include the biological activities such as complement fixation, antibody-directed cellular cytotoxicity, ability for placental transfer, and a longer serum half-life. The Fc domain also provides for detection by anti-Fc ELISA and purification through binding to ~~~hvlococcus sureus protein A ("Protein A"). In certain applications it may be desirable to delete specific effector functions from the Fc region, such as Fc receptor binding or complement fixation.
In another embodiment the Fc region can be a marine immunoglobulin gene. The use of marine Fc as the Fc region can have advantages. For example, if the fusion protein is to be used for the preparation of proteins in mice, then the marine Fc region will not elicit an immune response in the host animal. The Fc domain may confer the effector function activities to the fusion protein, and allow for detection of the fusion protein by anti-Fc ELISA and purification through binding to Protein A. In certain applications it may be desirable to delete specific effector functions from the Fc region.
SUBSTITUTE SHEET (RULE 26) '~1~98'3~

In another embodiment the DNA sequence encodes a proteolytic cleavage site interposed between the secretion cassette and. the target protein. A cleavage site provides for the proteolytic cleavage of the encoded fusion protein thus separating the Fc domain from the target protein. As used herein, "proteolytic cleavage .site" is understood to mean the amino acid sequences which are cleaved by a proteolytic enzyme or other proteolytic cleavage agents. As will be described in more detail below, useful proteolytic cleavage sites include amino acids sequences which are recognized by prote~alytzc enaymes such as trypsin, plasmin or enterokinase K.
In a preferred embodiment the target protein sequence encodes prostate specific membrane antigen, PSMA. PSMA i:; a type II membrane protein, thus the extracellular domain, or soluble form of the, protein, is utilized as the target protein sequence. The encoded soluble form of PSMA can be a human sequence such as the sequence provided in Israeli et al. (1993) Cancer Res., 53:227-ff.
In another preferred embodiment the target protein sequence encodes the protein gp120. The envelope protein gp120 of human immunodeficiency virus is a glycoprotein which is expressed in infected cells as a polyprotein, gp160, and then cleaved by a cellular protease to gp120 and gp4l. The nucleotide sequence and amino acid sequence of gp120 is provided in Ratner et al., 1985. Nature, 313:277-ff.
In another aspect, the DNA sequence of the invention is integrated within a replicable expression vector. As used herein, "vector's is understood to mean any nucle.c acid comprising a nucleotide sequence of interest and competent to be incorporated into a host cell and to be recombined with and integrated into the host cell genome, or to replicate autonomously as an episome. Such vectors include linear nucleic acids, plasmids, phagemids, cosmids and the like. A preferred expression vector is pdC, in which the transcription of the immunofusin DNA is placed under the control of the enhancer and promoter of the human cytomegalovirus. The vector pdC was dex°ived from pdEMp, which is described in La et al. 1991, Biochim. Biophys. Acta 1088:712 (which SUBSTITLfTE SHEET (RULE 26) ~19~83~
WO 961Q8570 PCTlUS95111720 publication is incorporated herein by reference) as follows. The SalI-XhoI fragment containing the car3ginal enhancer and promoter sequence were replaced by the enhancer and promoter of ttue human cytomegalovirus by standard molecular biology techniques. The enhanc:er and promoter sequence of the human cytomegalovirus used was derived from nucleotides -601 to +7 of the sequence provided in Boshart et al., 1985, Cell 41:521, which is incorporated herein by reference. T'he vector also contains the mutant dihydrofolate reductase gene as a selection marker (Simonsen and Levinson (1983) Proc. Nat. Acad_ Sci. USA 80:2495, incorporated herein by reference).
An appropriate host cell can be transformed or transfected with the DNA sequence of the invention, and utilized for the expression and secretion of a target protein. Currently preferred host cells for use in the invention include immortal hybridoma cello, myeloma cells, 293 cells, Chinese hamster ovary cells, Hela cells. and COS cells. As used herein, "gene expression" or "expression of a target protein" is understood to refer.to the transcription of the DNA sequence, translation of the mRNA transcript, and secretion of the fusion protein product.
The method of the invention involves providing a I)NA sequence encoding an immunofusin, transfecting the DNA sequence: into a host cell by an available transfection or transformation technic,~ue, culturing the transfected host cell in a suitable media under conditions which promote the expression and secretion of the immunofusi.n, and collecting the fusion protein from the extracellular media, When desired, the target protein may be cleaved from the secretion cassette either before or after it is collected from the extracellular media.
Other advantages and features of the invention will be apparent from the description, drawings, and claims which follow.
SUBSTITUTE SHEET (RULE Z6) WO 96!08570 PGT/US95I1I7ZU
_ '7 ~r~,Q",~ Descryyotion of the Drawincr The Figures 1A-D are a schematic illustration of an immunofusin.
Figure 1A, "DNA," illustrates the DNA encoding an immunofusin protein.
Figure 1H, "Fused Protein 1," illustrates the immunofusin protein prior to cleavage of the signal sequence. Figure 1C, "Fused Protein 2,°
illustrates the immunofusin protein after cleavage of the signal sequence. Figure 1D, "Target Protein," illustrates the target protein portion of an immunofusin protein after cleavage of the immunofusin protein at the cleavage site which is interposed between the Fc region and the target protein.
SUBSTITUTE SHEET (RULE 2fi) WO 96/08570 ~ ~ PCTIUS95111720 _g_ The present invention is a DNA compris$ng a polynucleotide encoding, in the 5' to 3' direction ~e~signal sequence, an Fc region of .5 an immunoglohulin, and a target protein. This approach to the expression and subsequent secretion of a target protein is superior to the existing techniques because of the choice and the configuration of the secretion cassette which is placed at the 5' end of the fusion construct. Additionally, the regulatory sequences which direct the expression of the secretion cassette can be optimized, and the optimized secretion cassette can be paired with numerous target proteins, thus allowing for the efficient production of numerous fusion proteins.
The production of the immunofusin proteins is characterized as efficient and high level, because the target protein has been produced at the level of several micrograms/milliliter using the DNAs and methods according to the invention. Previously, workers in the art have rarely quantitated the expression levels of hard to express proteins due to the low levels of expression that are obtained in the known mammalian expression systems and the difficulties faced in quantitating proteins by techniques such as western blotting and RIA.
Prior to the teachings of this invention, expression of microgram per milliliter of hard to express proteins would often be attempted using bacterial expression systems.
This invention is based on the concept that the ease of production and collection of a target protein could be improved if the polypeptide of interest were linked to an immunoglobulin Fc domain and the fusion protein were expressed in a host cell, in particular a complementary host cell which naturally expresses the immunoglobulin, such that the fusion protein would be readily secreted from the host cell. In addition to promoting the secretion of the fusion protein from the host cell, the Fc region can further be exploited to aid in the purification 3$ of the fused polypeptide. The general approach of they invention involves the construction of recombinant DNA which encodes a fused SUBSTti'UTE SHEET (RULE 26~

21~98~3~J
. O 9Gf085i0 PGTIU593/11720 polypeptide, which upon expression, result: in expression of a secretion cassette linked to a target protein, ~..e., a protein of interest having potential ar demonstrable utility.
S The overall structure of the preferred DNA of the invention, the fusion protein it encodes, the form of the protein which is most often secreted and the target protein product after enzymatic cleavage are illustrated schematically in Figures ~A.-D, Reference characters in the DNA, Figure IA, are carried over into the protein, Figures 1B-D, as corresponding primed characters. The DNA which encodes the immunofusin is shown between the start and the stop markers on the illustrated DNA
sequence, Figure 1A. Upstream regulatory elements are shown at the 5 end of the DNA and are labeled "regulatory sequences". The DNA is composed of three distinct polynucleotides which are linked together.
In Figure 1A, 3~ of the regulatory sequences, which may be optimized for each secretion cassette, is a first DNA 8 which encodes a secretion cassette comprising two of the three polynucleotides: 1) a signal sequence 10, and 2) an immunoglobulin Fcy region 12. The immunoglobulin Fcy region is comprised of three subregions: 1) a hinge region 14, 2) a CH2 region 16, and a CH3 region 20. Attached to the 3' end of the DNA encoding the secretion cassette is the third polynucleotide, a DNA encoding the target protein 24. Optionally, DNA
encoding a proteolytic cleavage site 22 can be interposed between the DNA encoding the CH3 region of the immunoglobulin Fcy region and the ZS DNA encoding the target protein.
The encoded fused protein comprises the secretion cassette e' and the target protein 24', shown as Fused Protein 1 in Fig. 1B. Most often the signal peptide 10' will be enzymatically cleaved from the fusion protein by the host cell prior to the secretion of the immunofusin, and thus Fused Protein 2, shown in Figure 1C, shows the secreted fused protein which comprises the Fcy peptide 12' fused to the target polypeptide 24'. Both Fused Proteirx 1 and Fused Protein 2 show the optional interposition of a proteolytic cleavage site 22' between the CH3 domain 20' of the Fcy region 12' and the target protein 24'.
Cleavage of either Fused Protein with the appropriate proteolytic agent SUBSTITUTE SHEET (RULE 26) wo 96108s70 219 ~ 8 3 ~ PGTIIIS95/11720 at the cleavage site 22' results in the release of the target protein 24' from the Fc region 12', as shown in Figure 1D.
The processes for manipulating, amplif~~~,ng and recombining DNAS are generally well known in the art" and therefore are not described in detail herein. Methods of identifying and isolating genes encoding proteins of interest, or for constructing such genes, are well understood and developed. In general the methods involve selecting genetic material coding for amino acids which define the polypeptide of 11~ interest according to the genetic code.
Accordingly, the DNA construction principle disc7.osed herein can be exploited using known recombinant DNA techniques involving the use of various restriction enzymes which make sequence specific cuts in DNA to produce blunt ends or cohesive ends, DNA lipase techniques enabling enzymatic addition of sticky ends to blunt ended DNA, construction of synthetic DNAs by assembly of short oligonucleotides, cDNA synthesis techniques, polymerase chain reaction, and synthetic probes for isolating genes having a particular function, various promoter sequences and other regulatory DNA sequences used in achieving expression, and various types of host cells are also known and available. Conventional transfection techniques, and equally conventional techniques far cloning and subcloning DNA are useful in the practice of this invention and known to those skilled in the art.
2~ various types of vectors may be used such as plasmids and viruses including animal viruses. The vectors may exploit various marker genes which impart to a successfully transfected cell a detectable phenotypic property that can be used to identify which of a family of cells has successfully incorporated the recombinant DNA of the 'vector. Given the foregoing state of the genetic engineering art, skilled persons are enabled to practice the invention disclosed herein in view of this disclosure.
One method for obtaining the DNA encoding the various synthetic linkers disclosed herein is by assembly of synthetic oligonucleotides in a conventional, automated, po:lynucleotide synthesizer followed by SUBSTITUTE SHEET ~RUIE 2fi) c.v om9!~s:~o mq9-io-i3 ligation with a ligase. For example, the linkers can be synthesized as complementary DNA fragments using phosphoramidice chemistry.
The signal sequence of the invention is a polynucleotide which 5 encodes an amino acid sequence that initiates transport of a protein across the membrane of the endoplasmic reticulum. Signal sequences which will be useful in the invention include antibody light chain signal sequences, e.g., ancibady 1.18 tGillies et. al., 1989, Jour. of Immunol. Meth.. 125:191~102), antibody heavy chain signal sequences, 10 e.g., the MOPC141 antibody heavy chain signal sequence tSakano et al., 1980. Nature 286:5774), and any other signal sequences which are known in the art tsee for example, rlatsan, 1984, Nucleic Acids Research 12:5145). Signal sequences have been well characterized in the art and are known typically to contain 16 to 30 amino acid residues, and IS may contain greater oz fewer amino acid residues. A typical signal peptide consists of three regions: a basic N-terminal region, a central hydrophobic region, and a more polar C-terminal region. The central hydrophobic region contains ~ to 12 hydrophobic residues that anchor the siqrsal peptide across the membrane lipid bilayer during transport 20 0! the nascent polypeptide. Following iniciatian. the signal peptide is usually cleaved within the lumen of the endoplasmic reticulum by cellular enzymes known as signal peptidases, Patantial cleavage sites of the signal peptide generally follow the "t~3. -i) rule". Thus a typical signal peptide has small, neutral amino acid residues in 25 positions -1 and -3 and lacks proline residues in this region. The signal peptidase will cleave such a signal peptide between the -1 and ~1 amino acids. Thus, the portion of the DNA encoding the signal sequence may be cleaved from the amino-terminus of the imanrnofusin protein during secretion. This results in the secretion of a 30 immunofusin protein consisting of the Fc region and the target protein.
A detailed discussion of signal peptide sequences is provided by won Heijne (1986) Nucleic Acids Res., 14:4683. As would be apparent to one of skill in the art, the suitability of a particular signal sequence for use in the secretion cassette 35 may require same routine experimentation. Such ~19'~~~~(~

experimentation will include determining the ability of the signal sequence to direct the secretion of an immunafusin and also a determination of the optimal configuration, genomic or cDNA, of the sequence to be used in order to achieve efficientz secretion of S immunofusins. Additionally, one skilled in the art is capable of creating a synthetic signal peptide fol:l~owing the rules presented by von Heijne, referenced above, and testing for the eff-_icacy of such a synthetic signal sequence by routine experimentation. A signal sequence is also referred to as a "signal peptide", "leader sequence"
or "leader peptides" and each of these terms having meanings synonymous to signal sequence may be used herein.
The Fc region of an immunoglohulin is the amino acid sequence for the carboxyl-terminal portion of an immunoglobulin heavy chain constant region. The Fc regions are particularly important in determining the biological functions of the immunoglobulin and these biological functions are termed effector functions. As known, the heavy chains of the immunoglobulin subclasses comprise four or five domains: TgM and IgE have five heavy~chain domains, and IgA, IgD and IgG have four heavy chain domains. The Fc region of IgA, IgD and IgG is a dimer of the hinge-CH2-CH3 domains, and in IgM and IgE it is a dimer of the hinge-CH2-CH3-CH4 domains. Further the CH3 domain of IgM and IgE is structurally equivalent to the CH2 domain of IgG, and. the CH4 domain of IgM and IgE is~the homolog of the CH3 domain of IgG (aee, W.E.PauI, ed., 1993, Fundamental Immunology, Raven Press, New York, New York, which publication is incorporated herein by reference). Any of the known Fc regions would be useful as the Fc region of the secretion cassette. However, it is important that the binding sites for certain proteins be deleted from the Fe region during the construction of the secretion cassette. For example, since coexpression with r_he light chain is unnecessary, the binding site for the heavy chain binding protein, Bip (Hendershot et al. (1987) Immunol. Toc~av x:111-114), should be deleted from the CH2 domain of the Fc region of IgE, such that this site does not interfere with the efficient secretion of the immunofusin. Likewise, the cysteine residues present in the Fc regions which are responsible for binding to the light chain of the SUBSTITUTE SHEET (RULE 26~

CA (121!79830 1')99-lil-13 - t3 , immunoglobulin should be deleted or substituted with another amino acid, such that these cysteine residues do not .interfere with the proper folding of the Fc region when it to produced na an immunotusin.
In the same manner, transmembrane domain sequences. such as chose present in Icfil, should be deleted such chat these sequences do not result in misdirecting the imrnunofuain to the membrane as a transmembrane protein.
Upon expression and production of the Fc region as a portion of the secretion cassette, it may retain some o! the biological properties, termed "effector lunctions~, which are native to the particular immunoglobulin class from which the Fc region is obtained. Useful efieccor functions include, for example. complement fixation. Fc receptor binding, binding to cell membranes, and placental transfer.
IS In some cases. it may be advantageous to modify or remove one or more of these effeccor functions, sucri as Fc receptor binding or complement fixation, using site directed mucagenesis or other well known molecular biology techniques. For example. ouncan et al. (Nature. 1988, 332:738) have mapped the amino acids responsible for the several of the immunoglobulin gamma etfector functions activities, see also. Ihincan et al., 1988, 332:563: Yasmeen et al.. Imn~unol., 1976, 116:518; Tao et al.. J. Immunol.. 1989. 113:2595. The amino acids or peptide segments responsible for rheas functions can be deleted thus removing chat portion of the Fc region, or substituted with sequences which would not confer the function using well known molecular biology techniques.
The currently preferred class o! imiausloglobulin from which the Fc region is derived is immunoglobulin gamma-1, because it has been well characterized and is efficiently secreted from most cell types. The F~_ 3Q region of the other subclasses of immunoglobulin gamma (gamma-2, gamma-3 and gamma-41 would function equally well in the secretion cassette.
The Fc region of immunoglobulin gamma-1 is preferably used in the secretion cassette includes at least part of the hinge region, CH2 region, and CH3 region. In addition, the Fc region of imeaunoglobulin ga~a-1 can be a CFi2-deleted-FC, which includes a part of a hinge WO 96108570 ~ 19 ~ ~ 3 0 PG"TIU595l11720 region and a CH3 region wherein the CH2 region has been deleted. A
CH2-deleted-FC has been described by Gillies et al., 1990, Hum.
Antibod. Hybridomas, 1:47, which publication is incorporated herein by reference, As is apparent from the above'discussion of Fc regions, the Fc regions from the other classes of immunoglobulins, IgA, IgD, IgE, and IgM, would also be useful as the Fc region of the secretion cassette.
Further, deletion constructs of these Fc regions, in which one or more of the constant domains are deleted would also be useful. One of ordinary skill in the art could prepare such deletion constructs using well known molecular biology techniques.
The identity of the target protein produced in accordance with the invention is essentially unlimited. Indeed, an important feature of the invention is that it provides a generalized DNA construct, and procedure which can be adapted to facilitate recombinant groduction of any desired target protein. For instance, the application of the invention to the expression of the regulatory proteins, such as transcription factors which are normally localized to the nucleus.
allows for the efficient secretion of such normally non-secreted proteins. In addition, regulatory proteins are in general difficult to express and the purification procedures are generally cumbersome (see, for example, Meisterernat et al. (1991? Cell 66:981). Therefore, it is 2~ esgecially desirable that such proteins be exported into the culture medium. Additionally, the invention can be used to enhance the production and secretion of proteins which are normally secreted at low levels. If a desired target protein includes sequences encoding a secretion signal or a transmembrane signal, these sequences can be 3~ removed from the target protein such that the secretion cassette directs the secretion of the fusion protein.
The optional proteolytic cleavage sits may be any amino acid sequence which is recognized by specific cleavage agents. The 35 specificity of cleavage agents is determined by the identity of the sequence of amino acids at or near the peptide band which is to be SUBSTITUTE SHEET (RULE 26) ~19~8.3~
v~ro 9s~o85~0 PCTlUS95lI17Zo -'d5..
hydrolyzed. A given cleavage agent may recognize the bond between two specific amino acids or may recognize a bond following one or a specific sequence of amino acids. ~'he specificity of many cleavage agents is known. Table 1 set forth below lists various known cleavage S agents and their primary (and in some cases secondary) sites of action.
TB~I

_C~,~:avaQe Agent Major its of ActionQther Sites of A~~t ion 1~Trypsin Arg, Lys ChymOtrypsin Trp, Phe, Tyr Leu, Met, His Elastase Neutral Aliphatic Residues Pepsin Phe, Leu, Trp Ala, Gly, Glu 15Papain Arg, Lys, Gly Wide specificity Subtilisin Aromatic and Various Aliphatic residues Thermolysin . Amino-linked bonds Ala, Phe of Aliphatic Residues 20S. aureus proteaseGlu Asp Endoproteinase Arg Arg C (Submaxillaris protease) Clostripain Arg 25Thrombin Arg Collagenase x-Gly-Pro X-Ala-Pro X-Gly-Thr Lysobacter Lys enzymogenes 30 (endoproteinase Lys-C) Mysobacter A1-1 Lys Protease Armillaria mellea Lys Flavobacterium Pro 35 meringosepticum Factor Xa lle-Glu-Gly-Arg SUBSTITUTE SHEET (RULE 26) WO 96/085?0 PCTIUS95I117Z0 CNBr Met BNPS-skatole Trp N-bromosuccinimide Tra O-iodosobenzoic Trp S acid Fi8 r jDMSO Trp NTCB Cys Sodium metal in liquid ammonia Pro 1U Fiydroxylamine Asn-Gly Dilute acid Asp-Pro Other cleavage agents are known. Those preferred for use in the invention are enzymes with a primary site of action which cleave at the 15 C-terminal aide of the cleavage site residue, The cleavage site in the fused protein generally can comprise any one or sequence of amino acids which can be cleaved by a cleavage agent specific for the site in an appropriate environment, Specificity of 20 cleavage can be increased, and likelihood of undesired cleavage within the target protein or elsewhere in the fused polypeptide can be decreased, by selecting the cleavage agent having a site of action which is absent from the target polypeptide. The fused polypeptide is preferably cleaved under conditions in which it has assumed its native 25 conformation. This has the effect of masking the presence of potential cleavage sites in the target polypeptide.
The invention is illustrated further by the following non-limiting examples.
Example 1. Conatri.~s~~on of a ,secretion Ca~ette The construction of an exemplary secretion cassette is described below. As would be appreciated by those of ordinary skill in the art, the signal sequence and the Fc region of an immunoglobulin could be other sequences than those described.
SUBSTITUTE SHEET (RULE 2fi~

The signal sequence of an immunoglobulin light chain of the 14.18 antibody was selected fox- use as the signal sequence of the secretion cassette. The sequence of the 14.36 antibody light chain is provided in Gillies et al., 1'389, Jour. Immunol. Meth., 125:191-202. The signal sequence was modified for ease of cloning as an XbaI-AflII fragment of the DNA. As wou"~d be apparent to those of skill in the art, the DNA
encoding a human signal sequence could also be used. Specifically, an XbaI site was introduced 5' of ~he translation. initiation codon and the consensus sequence far optimal ribosome binding (Kozak, 1984, Nature 308:241). An AflII site was introduced into the 3~ end of the signal sequence by mutagenizing the DNA coding for the penultimate amino acid residue of the signal peptide fram a serine to <~ leucine, thus the sequence ATC was mutagenized to TTA using side directed mutagenesis.
The Fc region of an immunoglobulin was selected to be the human Fcyl genomic DNA, including the genomic configuration of the hinge, CH2 and CH3 domains. The genomic sequence of 'seaman Fcyl is provided in Huck et al., (1986) Nucleic Acids Res. 14:17?9 and is incorporated herein by reference. As would be apparent to one of ordinary skill in the art, a CH2-deleted-Fc may also be used as the Fc region of the secretion cassette (see, Dillies et al., 1990, i~um. Antibod.
Hybridomas, 1:47), in which case the CH2 domain would be deleted from the Fc region using established molecular biology techniques during the construction of the secretion cassette. The genomic DNA of Fcyl was 2~ modified for ease of cloning as an AflII-XmaI fragment. The 5' end of the human Fc genomic DNA was mutagenized to an AflII site by performing a Polymerase Chain Reaction iPCR) using a 5~ sense primer with the following sequence (Sequence ID No. 1):

3() This primer introduced an AflII site (underlined) and a cysteine to serine mutation (TGT to TC'I, bold). The cysteine being mutated is the one that is normally involved in disulphide bonding with the light chain and thus does not affect the effector functions. of the Fc region.

2199~3fl~
WO 96108570 PC1'IUS95I11720 The deletion of this cysteine may serve to enhance the production of the Fcyl region as the efficient productio:~ of this modified Fcyl region will not require the coe.xpression ofi fihe immunoglobulin light chain. This cysteine was also removed such that it does not interfere with the proper folding of the Fcy1 region or the fused target protein.
The 3' end of the Fc~l1 genomic DNA encodes for two XmaI restriction sites. They axe located at a0 and 280 by upstream of the translation step codon in the CH3 domain. The distal Xmal site was destroyed by introducing a silent mutation, using site directed mutagenesis, (TCC to TCA, where the CC were the first two bases of the Xmal site) so that the XmaT site 10 by upstream of the stop codon became unique.
The XbaI-AflII restriction fragment encoding the light chain signal peptide was then ligated to the AflII-Xmal restriction fragment I~ encoding the Fc region. The resultant XbaI-XmaT restriction fragment therefore encodes the secretion cassette, and the gene encoding the target protein of interest can be ligated to the 3' end of the secretion cassette via the Xmal site.
In general, the DNA encoding the target protein can be ligated to the unique Xmal site through the use of a linker-adaptor, such a linker-adaptor may also include restriction endonuclease sites ir_~
addition to an Xmal site. The use of a linker-adaptor has the additional feature in that it can encode a proteolytic cleavage site for subsequent use in cleaving the target protein from the secretion cassette after production and secretion of the fusion protein. :or examgle, the linker-adaptor can encode a lysine residue at the junction of the fusion protein, which provides the option of cleaving the target protein from the Fc domain by proteolytic enzymes such as trypsin or plasmin. Similarly, the linker adaptor can include a DNA encoding the cleavage site of enterokinase K tAsp-Asp-Asp-Asp~Lys) in order to provide for the specific cleavage of the secreted fusion protein by enterakinase K.
Example 2. ~vnst tin o~ an Immy~,~fusin SUBSTITUTE 5HE~T(RULE26) 21~~8~0.
w0 96f08570 PfT/IJS95lI17ZU
_ Ig _ 'The construction of an exemplary immunofusin, including a secretion cassette and a target protein is described below. As would be apparent to those of ordinary skill in the art, other target proteins can be fused to a secretion cassette using the same or other molecular cloning techniques.
The target protein for Che exemplary immunofusin was chosen to be CDZ6, which is a type II membrane protein having its active site within the carboxyl-terminal region of the protein which. is the extracellular domain. During the construction of a CD26 immunofusin, the cytoplasmic and transmembrane domains of CD26 were deleted so that they would not interfere with the secretion of the immunofusin by the secretion cassette. The 5' end of the cDNA encoding the extracellular domain was modified for ease of cloning to include a Xma2 site, which was introduced via a linker-adaptor. The 3' end of the CD26 cDNA was also modified for ease of cloning to include a Xhoi site, which could be introduced downstream of the translation stop codon either by PCR or by linker-adaptor ligation.
Various linker-adaptors can be used depending upon the desire for introduction of a proteolytic cleavage site between the DNA encoding for the Fc region and the CD26 cDNA. For example, one linker-adaptor which can be used for CD26 is:
5' CCG GGT (AAA) GGC ACA GAT GAT GCT ACA G
. 3' CA (TTT) TTG TGT CTA CTA CGA TGT C
as provided in Sequence ID Nos. 2 and 3. The first three codons in the top strand encode the last three amino acid residues of the CH3 domain, and starting with the codon GGC is the gene sequence of the extracellular domain of CD26. This linker-adaptor had the cohesive end of an XmaI site at its 5' end and the blunt end o:E a PvuI2 site at its 3' end, the blunt ended PvuII site being a convenient site for reconstruction with the rest of the CD26 cDNA. The lysine codon (AAA, in parenthesis) in the linker-adaptor is but one of many optional amino acid sequences which are useful to provide for a proteolytic cleavage site by cleavage agents. For example, this lysine residue can be cleaved by enzymes such as trypsin or plasmin.
SUBSTITUTE SHEET (RULE 26) ('A (1~t99S30 1999-10-l3 .zo-Alternatively, for rnora specific proteolytfc cleavage by eneerokinase K, the gene sequence encoding the enterokinasa K cleavaqe site can be introduced via the following linker-adapcor:
S 5 ' Cw GGT TCA GGG GAT G7~C G71T C71C QJLZ J1 3 ' CA AGT CCC CTR CTQ C?J1 CTO CTJ1 TTC 07l as provided in Seq. ID Nos. 4 and 5. The nucleatidas in bold encode the amino acid residues IAsp)4-Lys, which is the recognicivn site of antarokinaae K. The linker-edaptor ends with a HindIII site, to which the CD~6 gene or ocher target protein gene sequences can be joined.
Example 3. Ho~dt CellZ and Tran~~~ezion The preferred host cell lines Include the mouse myeloma for hybridoma) NS/0 and Sp2/0 Agl4 cells. The myeloma cells were transfected by protoplast fusion and selected in D~slbecco's modified Eagle's medium (Gibcol containing IOt fatal bovine serum and 100 cu~1 methotrexate. as described by Gillies ec al.. 196 9 8ioTechnology, 7:799, which publication is incorporated herein by reference.
20 Transfectaats secreting the fmmunofuains were identified by anti-Fc ELISA. as described by Gillies et a1. t19A9) J. ia~ol. Methods 125:191. The highest producers were adapted to media containing 1 uM MTX and subcloned by limiting dilutions. For the production of immunofusins, the cells were grown in Hybridoma Serum-Free Media (HSFM~ Gibco) c~ontaaning 1~ fetal.
bovine serum and 1 uM MTK.
The ocher preferred recipient cell line is the human kidney Z93 cells, which is useful for both transient and stable expression. Other cells, such as the Beta and the Chinese hamster ovary (CHC7> cells, also worked in our system. The preferred method of translection for these adherent cells is by coprecipication of plasmid DNA with calcium phosphate, and other methods include lipofection and electroporation.
For a description of these machods and other useful transfection 35 methods see. Sambrook et al. (19691 l~,o"1!~cular ~onina--A Laboratory ~ysy, Cold Spring Harbor, NY. incorporated herein by reference.

~:v ozr9~783(1 1919-r0-r3 -21 ~
Example 4. characterization and P~riticacion at .I~~i~S~t~ili 5 For routine characterization by gel electrophoresis, immunofusins in the conditioned media were first captured on Protein A Sepharose*
tRepligen, Cambridge, MA! and then eluted by boiling in protein sample buffer with or without 2~mercaptoethanol. After electrophoresis on an SDS-gal, the protnin bands were visualized by Coomassie staining. For 10 example, the IL2 immuriofusin, see example 5, gave a band having the molecular weight of ~5 kD under reducing conditions and a band having the molecular weight of 90 kD under non-reducing conditions, showing that the ZL2 lmmunofusin was produced as a dimer, presumably through disulphide banding in the hinge domain of the Fc region.

For purification. the call culture media was collected and then the immunofusins were bound on Protein A Sapharose. The iasnunofusins were subsequently eluted from the Protein A in a sodfum citrate buffer tl0e mM, pH 4). The elusce was then immediately neutralized with 0.1 volume 20 of 1 M Tris-hydrochloride, pH 8. In the case of CL126 immunofusin, it was shorn that such an elucian procedure resulted i.n greater than sOt recovery of the CD26 immunofusin with no lnss of enzyme activity Example 5. ~p,ef IL2 T~
ZS
The cDNA of mature III protein was modified for ease of cloning tc have a 5' XmaI restriction endorzuclease sites and a 3' XhoI restriction endonuclease site using well known molecular techniques, such as chase which were as described in example 2. The sequence of the mature IL2 30 cDNA is provided in Taniguchi et al., 1983, Nature, 302:305 and is incorporated herein by reference. The cDEtA of the mature IL2 protein was constructed using recombinant techniques as a synthetic gene in order to optimize colon usage and to introduce desirable restriction endonuclease cleavage sites. The synthetic gene was created using 35 conventional DNA manipulation techniques. Once the synthetic IL2 cDNA
was constructed, the 5' XmaI site of the xL2 eDNA was ligaced to the 3' *Traaemark 21~a~~a WO 96/08570 PCT/iJS95I11720 XmaI site of the secretion cassette, described in Example 1. The IL2 immunofusin, was then cloned into the expression vector pdC. The IL2 immunofusin expression vector was transfected into NS/0 and Sp2/0 as host cells by protoplast fusion, as is described by Gillies et al., S 1989, Biotechnology, 7:799.
Two to three weeks after transfection, MTX-resistant NS/0 and Sp2/0 clones appeared. The initial clones were screened by anti-Fc ELISA.
The IL2 immunofusin protein was collected from the media. An appropriate assay for the biological activity cf IL2 was the standard T-cell proliferation assays according to dillies et al. (Proc. Natl. .
Aced. Sci. (1992) 89:1428), which is incorporated herein by reference.
The spent culture of the best clone contained about 100 N.g/ml of IL2 immunofusin. The host cell clones which efficiently produced and secreted the IL2 immunofusin protein were subcloned in media containing 100 nM MTX, and the best subclone ,produced about 200 ~.g/ml of protein in spent culture. When MTX was left out of the media in the subcloning, the best subclone thus isolated produced about 180 wg/ml in spent culture. Thus, the construction ef an IL2 immunofusin 2l? unexpectedly provided for the production of IL2 at a level which is about 80 times that which can be achieved by the expression of IL2 alone using the pdElHp vector (unpublished data), and many times o~ that bf the IL2 that was expressed in mammalian cells (COnradt et al., 1989, J, Biol. Chem.; 264:17368) and in yeast (Ernst et al., 1989, Biotechnology, 7:716). As mentioned in example 9, Ih-2 immunofusin was produced as a homo-dimer of molecular weight of 90 kD, presumably through disulphide bonding in the hinge domain of the 45 kD monomers.
Example 6 . ~_x~ess~ on of ~~,~F ~ mmunofus ~ n The construction of CD26 as an immunofusin was undertaken to demonstrate that the invention is applicable to the expression of membrane anchored proteins such as type zI membrane proteins. A type IT membrane protein displays the carboxyl-terminal domain on the 3S extracellular surface, and most often includes its active region within this carboxyl-terminal domain. The joining of a fusion polypeptide to SUBSTITUTE SHEET (RULE 26) W O 96!08570 ~ 19 ~ 8 a 1 ~ pGT/US95ll I 720 the carboxyl-terminal region o~ such a protein may interfere with the proper folding of the active site, and thus reduce or prevent the production of active protein.
CD26 is a type II membrane protein comprising 766 amino acid residues. The biological function of CD26 is as a T cell activation antigen and the putative coreceptor for entry of HIV in CD4+ cells (Callebaut et al. (1993) Science 262:2045). The CD26 protein is anchored to the lipid bilayer of the plasma membrane through a hydrophobic domain between residues 7 and 2~3 at t-he N-terminus. Amino acids 1 to 6 form a short cytoplasrnic tail. The rest of the protein, between residues 29 and 766, is extracellular and includes several potential N-glycosylation sites and the active site of the enzyme (Tanaka et al. (1992) J. Immunoi. 149:481). The 728 carboxyl-terminal residues in CD26 protrude from the membrane surface and the C-terminus is free. A soluble CD26 expressed as an immunoadhesin, will have a conformation different from that of the native CI726, because the carboxyl-terminus in an immunoadhesin CD26 protein is not free but connected to antibody sequence. On the other hand, if we engineer an immunofusin in which the antibody sequence is am~:no-terminal to the target protein, such as CD26, the native conformation of CD26 will be preserved, I.e. the C-terminus is free, and the antibody sequence, herein an Fc region, takes the place of the membi:ane to which CD26 is normally anchored. The enzymatic and biological activities of such a soluble CD26 immunofusin will not be compromised.. In addition, CD26 is a protease and its expression may be deleterious to the host cell.
Thus by efficiently exporting the CD26 protease outside of the host cell in the form of an immunofusin, a higher level of expression can be achieved.
A 2.3 kb cDNA fragment encoding the extracellular domain of CD26 was used to Construct the CD26 immunofusin expression vector. The DNA
sequence of CD26 is provided in Tanaka et al.., 1992, J. Immunol., 149:481 and is incorporated herein by reference. CD26 was fused 3' of the secretion cassette as described above in example 2, and then the secretion cassette and CD26 target protein were cloned into the SUBSTITUTE SHEET (RULE 26~

~la~~s~L~
WO 96/085?0 PGTlUS95/11720 expression vector pdC using the XbaI restriction endonuclease site 5' of the light chain signal sequence and the Xhoi restriction endonuclease site 3~ of the CD26 protein r3s described in example 2 above. The resultant CD26 immunofusin expression vector was S transfected into a host cell as described in Pxample 3 above. MTX-resistant clones from transfected NS/0 and Sp2/0 cells were screened by anti-Fc ELISA and DPPIV activity assay. CD26 is also known as DPPIV, which is an exopeptidase that cleaves after amino-terminal X-P (X can.
be any amino acid residue, and P is praline). IOPPIV enzyme activity of the CD26 immunofusin was assayed according to Tanaka et al., Proc.
Natl. Acad. Sci., 1993, 90:4586, incorporated herein by reference, using glycylproline p-nitroanilide tosylate (Gly-Pro-pNA) as a eubst,xate. The best NS/0 clone produced about 3.5 ~g/ml of CD26 immunofusin. The DPPIV moiety of the protein product was determined to IS be fully active, having KM and kcat values similar to those of the native CD26. Furthermore the enzymatic activity of CD26 immunofusin was inhibited by known peptide inhibitors in a dose-dependent manner.
The peptide inhibitoz~s tested included the tripeptides IPI and VPL and APL, each of which inhibited the CD26 enzyme activity greater than 30%
at 0.15 mM, greater than 70% at 1 mM and greater than 90% at 4 mM. AS
a control known non-inhibitor peptides were also tested for their effect upon CD26 enzyme activity and the known non-inhibitors, GGG and GPHyP (wherein HyP is hydroxproline), were found to have no effect on the CD26 activity when incubated with the CD26 immunofusin at concentrations ranging between 0.01 mM and 11 mM.
Example 7. ~g~re~~ioy of T~: immunofusin The invention was also applied to the expression of regulatory proteins which are normally localized to the nucleus, Because regulatory proteins are in general difficult to express and purify, it is especially desirable to devise a method by which such proteins can be efficiently secreted from a host cell. Immunofusin constructs of Tat and Rev (described in example 8), which are two proteins encoded by the human immunodeficiency virus (Hiv) that regulate expression of viral proteins in the cell nucleus, were made in order to determine the SUBSTITUTE SHEET (RULE 26) ~m~s3o PG'TlUS95I11720 "~'O 96!08570 efficiency with which these protez.ns can be expressed and collected.
We obtained high level expression and secretion of the Tat and Rev immunofusins, and readily purified the imrtmnofusins in a single step.
A 260 base-pair cDNA fragment encoding "pat was cloned into the XmaI
and XhoI sites of the pdC expression ve:a or by modification of the 5 and 3' ends of the Tat protein using recombinant DNA techniques as described above. The sequence of the cDNA encoding the Tat protein is provided in Ratner et al., 1985, Nature, 313:27, and is incorporated herein by reference. Specifically, the sequence at the 5' end was modified to, Seq. ID No. 6. $'~~'G GGT CGC ATG GACr . . . . , where the underlined sequence is the Xmal site and the ATG in bold is the translation start colon of the Tat gene. At the 3' end, an Xhol site was introduced immediately downstream of the translation stop codon by standard PCR techniques. The Tat immunofusin expression vector was then transfected into a host cell, as described above, and the host cells were analyzed for production of Tat immunofusin protein. High level expression was obtained in transiently transfected 293 cells and stably transfected NS/o cells. Stable NS/0 clones produced about 3 pg/ml of a 48 kD protein, analyzed on a SDS-gel under reducing conditions. This protein was confirmed to be Tat immunofusin by an anti-Tat antibody tCat. #'1001, American SioTechnologies, Cambridge, MA) .
The Tat immunofusin was shown to be active by the following transient expression experiment in 293 cells, the results of which are presented below in Table 2. The expression vector for Tat immunofusin was cotransfected with a separate vector containing LTR-TAR-Kappa, where LTR-TAR is the~long terminal repeat DNA sequence of HIV that is transactivated by the Tat protein, and Kappa is tlae gene sequence encoding the Kappa light chain of immunoglobulin. To measure expression levels of Fc-Tat (Tat immunofusin) and Kappa, the supernatants were assayed by anti-Fc and anti-Kappa ELISA respectively.
In Table 2, pdC-Fc-Tat represents the pdC expression vector for Tat immunofusin; LTR-TAR-Kappa represents the expression vector for Kappa light chain, in which the LTR-TAR regulatory region can be transactivated by Tat: and pCEP-Tat is an expression vector for Tat, SUBSTITUTE SHEET (RULE 26) WO 96/0857Q PCTlUS95/11720 whose transcription is under the control of the human cytomegalovirus enhancer and promoter. pCEF-Tat was used.as a positive control to monitor the transactivation of the L'~RrTAi~:-Kappa by Tat protein. As a negative control LTR-TAR-Kappa ufas transfected alone to demonstrate that it is not transactivated in the absence of Tat protein or Tat immunofusin. As shown in Table 2, high level expression of the Tat immunofusin was observed in transfection 1, high level expression of both Tat and Kappa light chain were observed in the cotransfection experiment, transfection 2. Transactivation of Kappa by Tat was seen in the positive control, transfection 3, as expected. Little or no expression of Kappa was seen in the negative control, transfection 4, also as expected. Therefore, the Kappa light chain is expressed only through transactivation of the LTR-TAR region by a functional Tat protein, and the Tat immunofusin provides a functional Tat protein which is readily secreted from the host cell. This result also demonstrates that the secretion cassette is able to direct the secretion of a protein which is normally transported to the nucleus of the host cell.
- Table 2 EI3',,~ ~ n /q ml; 1 DI3A ~ ed ~ n transfect~,on fc Kant~a 1.) pdC-Fc-Tat >300U 0 2.) pdC-Fc-Tat, LTR-TAR-Kappa 16UU 160 3.) PCEP-Tat, LTR-TAR-Kappa , 0 277 4,) LTR-TAR-Kappa U 3 Example 8. F~ress~on o ~gv Imm~,~yOf,,~sin A 350 base-pair cDNA fragment encoding Rev was modified to include a 5' Xmal site and a 3' XhoI site and then ligated 3' of the described secretion cassette in the pdC expression vector. The sequence of the cDNA encoding the Rev protein is provided in Rather et al., 1985, Nature, 313:277, and is incorporated herein by reference. , Specifically, the 5' end of the cDNA was modified to $~ fCG GGT CGC ATt3 GCA . . .. . (Seq. ID No. 7), where the underlined sequence is~the Xmal .
site and the ATG in bold is the translation start codon of the Rev SUBSTITUTE SHEET (RULE 2fi) fib 96!08570 ~ PCTlUS95/11720 gene. At the 3' end, an XhoI site was introduced immediately downstream of the translation stop codon by standard PCR techniques.
High level expression was obtained in transiently transfected 293 cells and stably transfected ATS/0 cells. Stable NS/0 clones produced about 3 wg/106 cells/day of the Rev immunofusin, which has a molecular weight of about 50 kD when analyzed on a SDS-gel under reducing conditions.
Example 9. ~i e-speci 3,5 $~-oreolyt?,~ Cleavage of Ir~muinaf.~sin An exemplary cleavage of an immunofusin is described below, as would be apparent to one of ordinary skill in the art, each of the above described immunofusins could be cleaved from their respective secretion cassettes using the same method or an analogous method.
A CD26 immunofusin having a lysine residue ("Fc(Lys)-CD26 immunofusin"), introduced by linker adaptor during construction of the immunofusin between the Fc region and the CD26 target protein sequence was cleaved using txypsin. To cleave the Fc(Lys)-CD26 immunofusin, the immunofusin was bound on Protein A Sepharose and cleaved at the desired lysine position by ti-ypsin to release CD26 as follows: Fc(Lys)-CD26 immunofusin bound on Protein A Sepharose was incubated with a 1~
trypsin solution at 37°C for 2 hr, Trypsin inhibitor (Sigma) was then added to stop any further digestion. The supernatant was then removed and analyzed on an SDS-gel under reducing conditions. After Coomassie staining, a band having a molecular weight of 110 kD, which corresponds to the size of CD26 without the secretion cassette, was obtained. As a control, CD26 immunofusin, without the lysine residue at the junction of the fusion between the Fc domain and the CD26 target protein ("Fc-CD26 immunofusin"), was bound on Protein A Sepharose and similarly treated. The CD26 was found to not be released from the secretion cassette of the Fc-CD26 immunofusin, as was expected, and this also confirmed the specific cleavage of the immunofusin at the amino acid lysine which was inserted between the CH3 domain of the Fc region and the target CD26 protein. As a further control, a.n identical aliquot of Fc-CD26 immunofusin which was bound to Protein A Sepharose was boiled SUBSTITUTE SHEET (RULE 26) r.A ozi99ssa t9v9-~o-m in the protein san~le butter and SDS-gel analysis of the supernatant showed a 140 kD band corresponding to the full length CD26 immunotusin protein monomer.

5 The resulca from the gal electrophoresis experiment were confirmed by DPPIV activity assays o! the Cryptic digests. Quantitative recovery of the DPPIV enzymatic activity was obtained in the aupernatanc when the FctLys)-CD26 immunotusin bound to Protein A Sepharose was treated with trypsin. In the parallel experiment with Fc-CDZ6 immunoEusin, 10 there was no DPPIV activfty in the supernatant, because the t'D26 protein was not released from the Protein A Sepharose.
Example 10. ~,~a.~-!pin o~ 4~F_,~2 imrsn nnfLin 15 OSF-2 is a 80-kD secretory protein that is involved in the ossification process. The sequence the DNA encodinq OSF-2 is provided in Takeshita et al., 1993. Biochem. J. 294:271.
The cDNA encoding the OSF-2 protein with its signal peptide was cloned into the expression vector pdC. NSlO cells were 20 used for stable tzansfection and 293 cells were used for transient expression; but in neither case was the OSF-2 protein detected.
The OSP~2 eDNA was then adapted to be expressed as an ie~unotusin.
At the 3' and. the Xbal size ac tho translation atop codon was 25 converted to an Xhoi sits by linker ligation. At the 5' end the Following linker-adaptor was used:
' ~,~T AAA ,7NIC JI,iI,T G? TJ1T Gilt 71~
3' ~A TTT n'G TG G'E71 11,T11 CTG Z'fC TJ14 as provided in Seq. ID Nos. 8 and 9. The nucleotides in bold encode 30 the N-terminus at the mature OSF-2 protein, ending with BglII cohesive ends. These BglII cohesive ends were ligared to the BglII-Xhoi fragment o! the OSF-2 cDNA. The Xrnal cohesive ends ac the 5~ end of the linker~adaptor (underlined! were ligated co the unique XmaI site in the immunotusin expression vector.

cry o~t9~sao t~!~~-tn-i:~
~29~
High level expression was obtained in transiently cransfeccad 293 cells and stabiy cransfaeted NS/0 cells. Stable NS/0 clones produced about 5 to 7 ycg/ml of a 110 kD protein, when analyzed on a SDS-gel under reducing conditions. This protein was eontirmed to be the OSF-2 5 immunofuain by Western blotting with an anti-OSF-Z antibody.
It was also found that the expression of OSF~2 as an immunotuain in a mammalian system was superior to the expression of OSF-2 in the thioredoxin gene fusion expreseian syatam 1n E. coli tLaVailie et al., 10 1993, Biotechnology, 11:18''). The thioredoxin gene fusion system was designed to circumvent the formation of inclusion bodies because fusion to thioredoxln increases tine salability of many heterologous proteins produced in the E. coli cytoplasm. To cast chin system for the expression o! OSF-2, the cDNA encoding the mature OSF-2 was inserted 15 into the SmaI site of the pTrxFus vector*4lnvitrog~en. San Diego, C~li.
thus creating a thioredoxin OSF-2 fusion protein. The supplier's protocol for the expression of the fusion proteins was followed. The thioredoxin oSF-2 fusion protein was expressed, and, as a control, the thioredoxin protein was expressed alone without a fusion partner. The 20 results showed that although thioredoxin alone could be produced as a soluble protein at a high level, the thloradoxin OSF-Z fusion protein was present only in the insoluble fracclon. Therefore, in addition to the lack of post-translational modification in bacterial expression, a relatively cortrplex mammalian protein such as OSF-Z was not synthesized 25 as a soluble protein when fused to thioredoxin.
Fsxampie 11. ~ e~s o~o,~' ,~I~~ inm~no~, ~~~sin ~iIG-H3, a gene product which is induced by transforming 30 growth factor-p, is a 68-kD secretory protein that shares sequence homology with OS!~-2. The sequence of cDNA encoding ~iIG-H3 is provided in Skonier et al. (1992) DNA and Cell Biology, 11:511. The cDNA encoding the native pIG-H3 was 35 cloned into the expression vector pdC; but attempts to obtain stable transfectants producing pIG-H3 were unsuccessful.

C~<\ 02194830 1499- 10- 1 3 - :30 -The pIG-H3 cDNA was then adapted to be expressed as an immunofusin.
At the 3' end, the Hsml site downstream of the translation stop codon was converted co an Xhol site by linker ligation. At the 5' end, the following linker-adaptor was used:
5 5 ' CCG GGT AAA c3CC CTO tSGC C
3 ' CA TTT COC (711C
(Seq ID. Nos. 10 and 11). The nucleotides in bold encode the N-texzninus of the mature SIG-H3 protein. The linker-adaptor had XmaI
cohesive ends for ligating to the expression vector as described in the 10 above examples, sad Apai cohesive ends for ligating to the ApaI site at the 5' end of the cDNA sequence encoding the mature SIG-H3.
High level expression was obtained in transiently transfected 293 cells and stably transfected NS/0 cells. Stable NS/0 clones produced 15 about 3.5 1g/106 cells/day of a 100 kD protein when analysed on a SDS-gel under reducing conditions. This protein was confirmed to be the SIG-H3 imntunofusin by Western blotting with anti-SIG-H3 antibody.
Example 12. ~mreerien ~ the ael'h~orm of IcE

The high affinity IgE receptor alpha subunit (IgE-R), the DNA sequence of which can be found in Kochan et al. (1988) Nucleic Acids Res. 16: 3584 was constructed as an immunofusin 25 as follows: An XmaI site was introduced to the 5' end of the cDNA encoding the mature IgE-R so that the sequence at the junction of the fusion was C CCG GGT GTC CCT CAG --- (Seq. ID
No. 12), where the XmaI site is underlined and the three codons in bold are the first three amino acid residues of the 30 mature IgE-R. At the 3' end of the IgE-R, the cDNA encoding the transmembrane domain and the rest of the C-terminus was deleted and a translation stop codon was placed after the last codon of the extrace:llular domain. The sequence of the IgE-R
immunofusin at the 3' end was thus TAC TGG CTA TAA CTC GAG

(Seq. ID No. 13), where the three codons in bold were the last three amino acid residues of the extracel:Lular domain of the IgE-R, and they were followed by a stop codon and an XhoI site (underlined).

WO 96108570 ~ ~ PC3'lLTS951117Z0 The pdC expression vector containing the IgE-R immunofusin was transfected into 293 cells and NS/0 cells. High levels of expression (3 to 5 Hg/ml) of the IgE-R immunofusin were detected in the cell culture media by anti(Fc) EZ,ISA, SDS-gel analysis under reducing conditions showed a band of the expected size of 70 kD. The partially purified protein (on Protein A Sepharose) was shown to bind IgE in an IgE-R/IgE ELISA.
Example 13. ~~ressiom of Fc:rl Fcy1 was expressed by itself without a C-terminal target protein.
This was achieved by ligating the following linker (having XmaI and Xho2 cohesive ends) IS 5' CCG GGT AAA TAG C
3' CA TTT ATC GAG CT
(Seq. ID Nos. 14 and 15), to the XmaI and Xhol sites of the pdC to reconstruct the coding region of Fe. High levels of expression was detected by anti(fc) ELISA in the cell culture media of the transiently transfected 293 cells (5 to 7 N.g/ml) and stably transfected NS/0 clones (5 to l0 ~.g/ml). SDS-gel analysis under reducing conditions showed an Fc band of the expected size of 31 kD.
Example 14 . ~y~ressiq~ of PSMA,_ ~pmunofusin PSMA, prostate specific membrane antigen, i.s a type II membrane protein having a molecular weight of greater than 100 kD. PSMA is an integral membrane protein, and as such i.t is an attractive target for imaging and immunoconjugate delivery. To facilitate the expression of sign ficant quantities of PSMA, we subcloned the extracellular domain of PSMA (the soluble form) and expressed~this domain of PSMA as an immunofusin. A portion of the extracellular domain of PSMA, which is a soluble form of PSMA, can be produced as an immunofusin.
The eDNA encoding the full length PSMA was cloned from a human prostate carcinoma cell line LNCaP [Israeli et al. (1993) Cancer Res., SUBSTITUTE SHEET {RULE 2fi) CA 0219'1830 1999-10-!3 53:22?. The portion of the PSMA cDNA correapondinq to the extracellular domain was adapted to be expressed as an immunofusin by Polymerase Chain Reaction using the following primers:
5 N-terminal: 5' ,a8~ JW1 TCC TCC J~1T G71a1 t~C
C ~ t erminal : 5 ° ~ TT71 GGC TJ1C tTC JILT C'~U1 71G
(Seq. ID Nos. lfi and 17f. The two primers provide the HindIII and the ~ChoI sites (underlined) for cloning intn the immunotusin expression vector. In the N-terminal primer, the HindIII site is followed by the l0 coding sequence of the extracellular domain v! PSMA (in bold!
immediately niter the transmembrane region. In the C-terminal primer, the XhoI site is followed by the anticodon o! the STOP codon and cha C-tertninal coding sequence of PSMA (in bold). Ths amine acid sequence of the extracellular domain of PSMA is shown in Stq. ID No. 18.

High level expression was obtained in scably tzanstected Z90 and Sp2/0 cells. The PSMA immunotusin secreted inca the cell culture media was purified by Protein A Sepharose, Treatment of the immunotusin with the protease plasmin quantitatively converted the 130-kD Fc-PSMA into 30 two products: the i00-kD PSMA extraceliular domain and the 31-kD Fc.
The Fc was then removed from the solution oy adsorption onto Protein A
Sepharose. The soluble PSMA was purified and used 'to immunize mice.
It is expected that an antibody specific only to PSMA should facilitate diagnosis and therapy of prostate cancer.

Example 15. p,~,=~a~ urine Pc The Fc region of marina 7~Za was prepared for expression as an imenunotusin. Since the marine Pc region will not be immunogenic to 30 mice, such an inenunofusin containing the marine Fc followed by, for example, a human protein fusion partner can be used to immunize mice directly without prior cleavage to get rid of the Fc, The marine Fc was cloned into our immunofusin expression vector as described below, and was expressed at a high level under our expression conditions.

e~~ uzrgvxao i9y9-~o-The murlne Fc Y2a donsain, preceded by the signal peptide described above, was cloned into an expression vector, pdC, and was expressed without fusion to a target protein. Marine Fc yZa cDNA lSikorav et al.. 1980, Nucleic Acids Res., 8:313-3155, which publication is 5 incorporated herein by raterence> was adapted fo:~ cloning into the expression vector by Polymerase Chain Reaction using the following primers:
N-terminal: 5' Q,~C G7la CCC ~ GGG CCC aG
C-terminal: 5' ~C ?Cx TTt J1CC CGia ALT? CCC
10 (Seq. ID Nos. 19 and 20). The N-terminal primer contains an AtlII site (underlined) for ligating to the AtlII site at the 3~ end of the signal peptide. described above. The sequence following the Aflil site tin bold) encodes the amino acid residues in the hinge region a! marine yza gene. The C-terminal primer contains an XhoI sites for cloning into the 15 expression vector. :ollowed by the anticodons of the translation STdP
codon and tha carboxyl and of marine rZa iin bold).
High level expression at the marine Fcr2a region was demonstrated in Z93 cells by SDS gel analysis followed by i~estern blotting with an 20 anti-marine IgG antibody.
Example 16. $xnre~_~jr~n .tL~..Qp~120 The envelope protein gp120 of human immlnodaficieney virus (Hiv) is 25 a glycoprotnin having a molecular weight of lZOkD, and is expressed on tha surface of HIV particles and HIV infected cells. The protein 9p120 is originally expreuad in infected calls as a polyprotein, gp160.
which is then cleaved by a cellular protein to gpl~fl and gp~l. qp120 was prepared as an imemznotusin and determined that the gp120 3Q immunotusin was axpressed at a very high level. Any desired portion of gp120 may also be prepared as immunofusin. The Fc moiety of the gpi20 immunofusin could be cleaved off and gp120 was purifiad.
The complete nucleotide sequence of HIV has been published 35 in Ratner et al. (1985) Nature, 313:277. To prepare the gp120 immunofusin, a 2199F~D

translation STOP codon followed by an Xhol restriction site was introduced to the gp120-gp41 aunction after ammo acid Arg-518 of gp160 using standard molecular biology techniques, e.g., palymerase chain reaction. The existing Ndel restriction site present at nucleotide 5979, which is within the amino terminal portion of gp120, was converted to a HindTII restriction site through linker-adaptor ligation to generate and in-frame fusion. The resultant HindIII-XhoI fragment (1.36 kilobase pairs? encoding gp120 was then cloned into the immunofusin expression vector, pdC, as described above.
The gp120 immunofusin expression vector was expressed in stably transfected 293 cells according to the methods described above, and high Level expression of the gp120 immunofusin was obt=ained. The gp120 immunofusin was functionally active, as determined by binding to CD4 in an ETaISA. The gp120 immunofusin was also determined to be quantitatively cleaved by enterokinase to release gp120 and the Fc region.
0~,~er Emb~,iiments 2C) The invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are therefore to be considered in all respects as illustrative and non-restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing , description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.
SUBSTITUTE SHEET (RULE Z6) wo 9sro~s7u ~ 19 ~ ~ 3 0 p~.~gg,~117Z0 - JS
SEQUENCE LTSTING
( 1 ) GENERAL. INFORMATION

(I) APPLICANT:

(A) NAME: FUJI TMMUNOPHARMACEUTICALS
COR:P.

(H) STREET: 12S HARTWELL AVENUE

(C) CITY: LEXINGTON

(D) STATE: MA

( E ) COUNTRY : T.JSA

(F) POSTAL CODE: 02173 (G) TELEPHONE: (617) 861-5300 (H) TELEFAX: (617) 861-5301 1$ (I) TELEX:

(ii) TITLE OF TNVENTION: EXPRESSION AND EXPORT TECHNOLOGY OF

(iii) NUMBER OF SEQUENCES: 20 (iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: PATENT ADMINISTRATOR, TESTA, HURWITZ &

THIBEAULT

2$ (B) STREET: 125 HIGH STREET

(C) CITY: BOSTON

(D) STATE: MA

(E) COUNTRY: USA

(F) ZIP: 02110 (v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk (B) COMPUTER: IBM PC Compatible (C) OPERATING SYSTEM: PC-DOSiMS-DOS

3$ (D) SOFTWARE: Patentln Release #1.0, Version #1.25 (vi) CURRENT APPLICATION DATA;

(A) APPLICATION NUMBER: US

(B) FILING DATE:

(C) CLASSIFICATION:

(viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: PITCHER. EDMUND R.

(B) REGISTRATION NUMBER; 27,829 4$ (C) REFERENCE/DOCKET NUMBER: FIP-001 (ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: 617-248-7000 (B) TELEFAX: 617-248-7100 $0 (2) INFORMATION
FOR
SEQ
ID
NO:
l:

(I) SEQUENCE CHARACTERISTICS:

55 (A) LENGTH: 41 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear 60 (ii) MOLECULE TYPE: cDNA

6$
(xi) SEQUENCE DESCRIPTION: SEQ ID NO:1:
SUBSTITUTE SHEET (RULE Zfi) ~~9~~3~

GAGAATTCTT AAGCGAGCCC AAATCTTCTG ACAAAACTCA C

(2) INFORMATION FOR SEQ TD N0:2:
{i) SEQUENCE CHARACTERISTICS:
(Ay LENGTH: 28 base pairs (B) TYPE: nucleic acid tC) STRANDEDNESS: single 1U (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
IS
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:2:
CCGGGTAAAG GCACAGATGA TGCTACAG

(2y INFORMATION FOR SEQ ID NO:3:
(i) SEQUENCE CHARACTERISTICS:
(A1 LENGTH: 24 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:3:
CTGTAGCATC ATCTGTGTTT TTAC

t2) INFORMATION FOR SEQ ID N0;4:
(i) SEQUENCE CHARACTERISTICS:
(Ay LENGTH: 28 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
SO (xi) SEQUENCE DESCRIPTION: SEQ ID N0:4:
CCGGGTTCAG GGGATGACGA TGACGATA

SS (2) INFORMATION FOR SEQ ID N0:5:
( i ) SEQUENCE CHAR~~CTERISTICS :
(A) LENGTH: 28 base pairs (B) TYPE: nucleic acid 60 tC) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: CDNA
SU6STiTUTE SHEET tRULE 26) wo 96ross7o 219 9 ~ 3 4 PCTlUS95II~72Q
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:5:
AGCTTATCGT CATCGTCATC CCCTGAAC
2$
(2) INFORMATION FOR SEQ ID NO:6:
(i) SEQUENCE CHARACTERISTICS;
(A) LENGTH: 16 base pairs (B) TYPE: nucleic ;acid (C) STR.ANDEDNESS: single (D) TOPOLOGY: linear 1S (ii! MOLECULE TYPE: cDNA
2O (xi) SEQUENCE DESCRIPTION: SEQ ID N0:6:
CCCGGGTCGC ATGGAG

zS (2) INFORMATION FOR SEQ ID N0:7:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 16 base pairs (B) TYPE: nucleic acid (C! STRANDEDNESS: single 30 (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0;7:
CCCGGGTCGC ATGGCA

(2) INFORMATION FOR SEQ ID NO: B:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 26 base pairs (B! TYPE: nucleic acid (C! STR.ANDEDNESS: single (D).TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:8:
SS CCGGGTAAAA ACAATCATTA TGACAA

(2) INFORMATION FOR SEQ ID NO:9:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 26 base pairs (B) TYPE: nucleic acid iC) STRANDEDNESS: single (D) TOPOLOGY: linear SUBSTITUTE SHEET (RULE 26) (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION; SEQ TD N0:9;
GATCTTGTCA TAATGATTGT TTTTAC

)0 (2) INFORMATION FOR SEQ ID N0:~.0:
(I) SEQUENCE CHARACTERTSTICS;
(A) LENGTH: 19 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE; cDNA
(xi7 SEQUENCE DESCRIPTION: SEQ ID NO:10:
CCGGGTAAAG CCCTGGGCC
2.5 1 s (2) INFORMATION FOR SEQ ID N0:11:
3Q (I) SEQUENCE CHARACTERTSTICS:
(A) LENGTH: i1 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear 35 (ii) MOLECULE TYPE: cDNA
40 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:11:
CAGGGCTTTA C

45 (2) INFORMATION FOR SEQ ID N0:12:
(I) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 16 base pairs (B) TYPE: nucleic acid 50 (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:12;
CCCGGGTGTC CCTCAG

(2) INFORMATION FOR SEQ ID N0:13:
(I) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 18 base pairs (8) TYPE: nucleic acid SUBSTITUTE SHEET (RULE 2fi) VJO 96!08570 3 ~ PGT/IJS95JI17Z0 (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID NO:13:
TACTGGCTAT AACTCGAG

(2) INFORMATION FOR SEQ ID N0:14:
IS (i) SEQUENCE CHARACTERISTICS:
(A) LENGTH. 13 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:14:
CCGGGTAAAT AGC

(2} INFORMATION FOR SEQ ID N0:15:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 13 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: 5EQ ID NO:15:
TCGAGCTATT TAC
i3 (2) INFORMATION FOR SEQ ID N0:16:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 23 base gairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear S$ (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:16:
AAGCTTAAAT CCTCCAATGA AGC

(2) INFORMATION FOR 5EQ ID N0:17:
SUBSTITUTE 5H~ET {RULE 26) WO 961085'70 PGTIUS95l117Z0 _4p., (i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 26 laase pairs (B) TYPE: nucleic acid (C) STRANDEDNES5: single (D) TOPOLOGY: linear Iii) MOLECULE TYPE: cDNA
1 t) (xi) SEQUENCE DESCRIPTION: SEQ TD N0:17 CTCGAGTTAG GCTACTTCAC TCAAAG

IS
(2), INFORMATION FOR SEQ ID N0:18:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 707 amino acids 20 t8) TYPE: amino acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear 25 (ii) MOLECULE 'TYPE: protein (ix) FEATURE:

(A)NAME/ItEY: Protein (B)LOCATION: 1..70?

(D)OTHER jnote~ EXTRACELLULAFt INFORMATION: " DOMAIN
OF

PSMA"

(Xi) SEQC3ENCE PTION: 18:
DESCRI SEQ
ID
N0:

Lys SerSer GluAla ThrAsnIle ThrProLysHis AsnMetLys Asn Ala PheLeu GluLeu LysAlaGlu AsnIleLysLys PheLeuTyr Asp . Asn PheThr IlePro HisLeuAla GlyThrGluGln AsnPheGln Gln 35' 40 45 45 Leu AlaLys IleGln SerG1nTrp LysGluPheGly LeuAspSer Gln Val GluLeu HisTyr AspValLeu LeuSerTyrPro AsnLysThr Ala His ProAsn IleSer IleIleAsn GluAspG1yAsn GluIlePhe Tyr Asn ThrSer PheGlu ProProPro ProGlyTyrGlu AsnValSer Leu 55 100 1os 110 Asp IleVal ProPhe SerAlaPhe SerProGlnGly MetProGlu Pro 60 Gly AspLeu TyrVal AsnTyrAla ArgThrGluAsp PhePheLys Val Leu GluArg MetLys IleAsnCys SerGlyLysIle ValIleAla , Asp SUBSTITUTE SHEET (RULE 26) WO 96!08570 ~ PCTlIIS9a!11720 -Arg Tyr GlyLysVal PheArgG1~AsnLys ValLysAsnAla GlnLeu Ib5 170 175 A1a Gly AlaLysGly Va1TleLeuTyrSer AspProAlaAsp TyrPhe _ 180 185 190 Ala Pro GlyValLys 5erTyrProAspGly TrpAsnLeuPro GlyGly Gly Val GlnArgGly AsnIleLeuAsnLeu AsnGlyAlaGly AspPro Leu Thr ProGly'lyrProA1aAsnGluTyr AlaTyrArgArg GlyIle Ala Glu AlaValGly LeuProSerIlePra ValHisProIle GlyTyr Tyr Asp AlaGlnLys LeuLeuGluLysMet GlyGlySerAla ProPro Asp Ser SerTrpArg GlySerLeuLysval ProTyrAsnVal GlyPro Gly Phe ThrGlyAsn Phe5erThrGlnLys ValLysMetHis IleHis Ser Thr AsnGluVal ThrArgIleTyrAsn ValIleGlyThr LeuArg Gly Ala ValGluPro AspArgTyrValIle LeuGlyGlyHis ArgAsp Ser Trp ValPheGly GlyIleAspProGln 5erGlyAlaAla ValVal His Glu IleValArg SerPheGlyThrLeu LysLysGluGly TrpArg Pro Arg ArgThrIle LeuPheAlaSerTrp AspAlaGluGlu PheGly ' 370 375 380, Leu Leu~GlySerThr GluTrpAlaGluGlu Asn5erArgLeu LeuGln Glu Arg GlyValAla TyrIleAsnAlaAsp SerSerIleGlu GlyAsn Tyr Thr LeuArgVal AspCysThrProLeu MetTyrSerLeu ValHis Asn Leu ThrLysGlu LeuLysSerProAsp GluGlyPheGlu GlyLys Ser Leu TyrGluSer TrpThrLysLys~Ser ProSerProGlu PheSer Gly Met ProArgile SerLysLeuG1ySer GlyAsnAspPhe GluVal Phe Phe GlnArgLeu GlyIleAlaSerG1y ArgAlaArgTyr ThrLys Asn Trp GluThrAsn LysPheSerGlyTyr ProLeuTyrHis SerVal SUBSTITUTE SHEET (RULE 26) WO 96/08570 PCTlUS95/I1710 Tyr Glu Thr Tyr Glu Leu Val Glu Lys Phe Tyr Asp Pro Me~ Phe Lys S Tyr His Leu Thr Val GlnValArgGly GlyMet.ValPhe G1uLeu Ala Ala Asn Ser Ile. Val ProPheAspC'ysAr~r.~gpTyrAla ValVal Leu ~

~

Leu Arg Lys Tyr Ala LysIleTyrSer BileSerMetLys HisPro Asp 565 X70 5?5 Gln Glu Met Lys Thr SerVa1SerPhe AspSerLeuPhe SerAla Tyr 1~ 580 585 590 Val Lys Asn Phe Thr IleAlaSerLys PheSerGluArg LeuGln Glu 20Asp Phe Asp Lys Ser ProIleValLeu ArgMetMetAsn AspGln Asn Leu Met Phe Leu Glu AlaPheIleAsp ProLeuGlyLeu ProAsp Arg Arg Pro Phe Tyr Arg ValIleTyrAla.ProSexSerHis AsnLys His Tyr Ala Gly Glu Ser ProGlyIleTyr AspAlaLeuPhe AspIle Phe Glu Ser Lys Val Asp SerLysAlaTrp GlyGlutialLys ArgGln Pro 35Ile Tyr Va1 Ala Ala ThrValGlnAla AlaAlaGluThr LeuSer Phe Glu Val Ala (2) INFORMATION
FOR
SEQ
ID N0:19:

(i) SEQUENCE
CHARACTERISTICS:

(A) LENGTH: 25 base pairs 4$ (B) TYPE: nucleic acid (C) STRANDEDNESS:
single (D) TOPOLOGY:
linear (ii) MOLECULE
TYPE:
CDNA

$0 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:19:
S$ CTTAAGCGAG CCCAGAGGGC CCACA
(2) INFORMATION FOR 5EQ ID N0:20:
60 (i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 25 base pairs (B) TYPfi: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear SUBSTITUTE SHEET (RULE 2f) 43 _ (ii) MOLECULE TYPE: cDNA
(xi) SEQUENCE DESCRIPTION: SEQ ID NO::?C~;
CTCGAGCTCA TTTACCCGGA GTCCG

SUBSTITUTE SHEET (RULE 26)

Claims (30)

The embodiments of the invention in which exclusive property or privilege is claimed are defined as follows:

1. A DNA encoding a secreted fusion protein, said DNA comprising a polynucleotide encoding, from its 5' to 3' direction:

(a) a signal sequence;

(b) an immunoglobulin Fc region which lacks the CH1 domain; and (c) a target protein sequence.

2. The DNA of claim 1 wherein the signal sequence encodes a signal peptide which directs secretion of the fusion protein and is then removed by enzymatic cleavage.

3. The DNA of claim 1 wherein the Fc region is altered to delete at least one effector function activity.

4. The DNA of claim 1 wherein the Fc region comprises a hinge, a CH2 domain and a CH3 domain of immunoglobulin gamma.

5. The DNA of claim 1 wherein the Fc region comprises a hinge region and a CH3 domain of immunoglobulin gamma.

6. The DNA of claim 1 further comprising a polynucleotide encoding a proteolytic cleavage site interposed 3' of said polynucleotide encoding said immunoglobulin Fc region and 5' of said polynucleotide encoding said target protein.

7. A replicable expression vector for transfecting a mammalian cell, said vector comprising the DNA of claim 1.

8. A host cell which expresses the DNA of claim 1.

9. A method of producing a fusion protein comprising the steps of:

1) culturing the host cell of claim 8 in a medium under conditions to promote expression and secretion of said fusion protein; and 2) collecting the fusion protein from the medium.

10. The method of claim 9 wherein the fusion protein has bioactivity of the target protein.

11. The method of claim 9 further comprising the additional steps, after step 2, of:

3) cleaving the Fc region from the target protein; and 4) collecting the target protein.

12. A method of producing a fusion protein comprising the steps of:

1) culturing the host cell of claim 8 in a medium under conditions to promote expression and secretion of said fusion protein; and 2) cleaving the Fc region from the fusion protein;
and 3) collecting the target protein.

13. A DNA produced by recombinant DNA techniques for inducing expression and subsequent secretion of a target protein, said sequence being free of immunoglobulin CH1 and comprising a polynucleotide encoding, from its 5' to 3' direction:

(a) a secretion cassette which comprises a signal sequence;
an immunoglobulin Fc region; and (b) a target protein sequence encoding a portion of gp120 protein.

14. The DNA of claim 13 wherein the signal sequence encodes a signal peptide which directs secretion of the target protein and is then removed by enzymatic cleavage.

15. The DNA of claim 13 wherein the Fc region is altered to delete at least one effector function activity.

16. The DNA of claim 13 wherein the Fc region comprises a hinge, a CH2 domain and a CH3 domain of immunoglobulin gamma.

17. The DNA of claim 13 wherein the Fc region comprises a hinge region and a CH3 domain of immunoglobulin gamma.

18. The DNA of claim 13 further comprising a proteolytic cleavage site interposed 3' of a portion of said polynucleotide encoding said immunoglobulin Fc region and 5' of a portion of said polynucleotide encoding said entire target protein.

19. A replicable expression vector for transfecting a mammalian cell, said vector comprising the DNA of claim 13.

20. A host cell transformed with the DNA of claim 13.

21. The DNA of any one of claims 1 to 6, wherein said signal sequence is selected from the group consisting of an antibody light chain signal sequence and an antibody heavy chain signal sequence.

22. The DNA of any one of claims 1 to 6, wherein said signal sequence encodes a signal peptide containing 16 to 30 amino acid residues.

23. The DNA of any one of claims 13 to 18, wherein said signal sequence is selected from the group consisting of an antibody light chain signal sequence and an antibody heavy chain signal sequence.

24. The DNA of any one of claims 13 to 18, wherein said signal sequence encodes a signal peptide containing 16 to 30 amino acid residues.

25. The DNA of any one of claims 1 to 6, wherein said Fc region encodes a carboxy-terminal portion of an immunoglobulin heavy chain constant region.

26. The DNA of any one of claims 13 to 18, wherein said Fc region encodes a carboxy-terminal portion of an immunoglobulin heavy chain constant region.

27. The DNA of claim 3, wherein said effector function activity is selected from the group consisting of Fc receptor binding and complement fixation.

28. The DNA of claim 15, wherein said effector function activity is selected from the group consisting of Fc receptor binding and complement fixation.

29. The DNA of any one of claims 1 to 6, wherein said target protein sequence encodes a regulatory protein.

30. The DNA of claim 29, wherein said regulatory protein is a transcription factor.