CA2020729A1 - Bone morphogenetic protein - Google Patents

Abstract

ABSTRACT

The purification and cloning of bone morphogenetic protein are disclosed, as well as production of BMP
and its analogs thereof by recombinant DNA
techniques. Pharmaceutical compositions comprising BMP and the use of such compositions are also disclosed.

Description

2 ~

BON~ MORP~OGENETIC PROTEIN

This inventi~n relates to a bone morphogenetic protein (BMP) whiGh initiates cartilage and bone growth. The full length coding sequence for BMP, a polypeptide, is provided~ The BMP is pr~vided by isolation fro~ bone sources and by synthesis using recombinant DNA techniques.

BACKGRC)UND OF THE INVENTIQN

It is known that demineralized bone ~atrix induces new bone formation when i~planted in the soft tissue by a process generally designated as matrix induced bone formation (see Urist, ~R., Scienc~, 150: 8~3-899 (1965)). There bave been numer~us efforts to extract and ident~fy the active ~aterial ~or materials) which induces this process, and it has been qenerally referred to in the literature as bone morphogenetic protein(s) (BMP), I~ is uncertain whether BMP ~s a single materi~l or a mixture o~
materials, and there does not appear to be agreemenS
among ~he investigators as to which ~ater~al, ~ any, is the bone ~orphog~netic protein. A~ discussed herein, the term BMP i~ used to describe the protein hav~ng th~ a~ino acid sequ-nce shown in FI~ 3 (bovine B~P) or FIG~ 5 (human BMP~, without the signal peptide.

; :

2~2~

The therapeutic use of BMP o~fers considerable advantages over use of traditional bone graft materials. ~hile not intended *o he limited by any theo~y, one hypothesis assumes that BMP initi~tes the differentiation o~ tissue cells into osteoblasts (cells that ~anu~acture bone). During a process that replica~es ~ormal hu~an ~etal develop~ent; ~MP-induced osteoblas~s ~orm cartilage which, over a period of several ~eeks yields ~olid bon~. Thus B~P
may be useful for replacing bone that has been destroyed by disease or accident, ~or use in treatment of scoliosis victims, for ~reatment o~ mal-or mis-formed bone, for use in healing of a fracture, dental reconstruction, hip replacement, bone remodeling, and control o~ osteoporosis.

It is thus ~n object of the present invention to produce a functional cartilage and bone growth ~actor or a component thereof, whic~ is a protein identified by its entire amino acid sequence, which ~as BMP
activity.

It is another object of the present invention to produce this biologically active BMP protein by recombinant DNA te~hnology.

2 ~ 2 ~

SUMMARY OF 5H~ VEN~ION

The present invention provides a class of mature native mammalian proteins texmed herein ~s ~bone m~rphogenetic protein~ or "~p-l, exemplified by tAe native hu~an and bovine BMP sequences described herein. Generally, ~his class of proteins induce bone growth in vivo or in Vi~rQ. The human and/or bovine BMP seguences are represen~ative of the class, and can be used to identify and isolate other mammalian BMP proteins, which will be at least partially or substantially homologous in nucleotide and amino acid sequences. Those o skill in the art will be able to readily identify other mammalian BMP's based on seguence homology to the human and bovine proteinC disclosed herein, as well as biolGgical acti~ity. It is rec~gnized that there may be allelic variations in BMP within a species, and such allelic variants are also within the scope of the class of proteins provided by the pre~ent invention.

The present invention further provides polypeptides which are analog~ of ~MP, such as BMP ~utei~s, fusion proteins comprising BMP or BMP domains, and BMP
fragments. Preferred analogs have BMP activity.
BMP mutein is a protein substantially ~omologous to a native BMP seguence (e~., a minimum of about ~5%, 85%, 90% or 95S homolcgous) wherein at least on~
amino acid is dif~erent. ~he ter~ fusion pr~tein includes a protein co~prising a complete BMP ~equence or a BMP domain, and a heterologou~ N- or C-termin~l sequence ~such as a ~ignal ~equence or ~e~uen~e wh~ch protects the protein ~rom degradation~. A ~P
fragment or doma~n $~ an amino acid &~guence o~
sufficient length ~ro~ a RMP protein ~uch t~at it is identifiable as having been derived ~rom such BMP

' 2 ~

protein. The origin o~ a particular peptide can be determined, for exa~ple by comparing it5 sequence to those in public databases.

The present invention provides in anot~er e~bodimen~
BMP having amino acid sequence shown in FIGS. 3 and 5 (which also show the 6ignal peptide~. The present invention alsQ provides methods oP preparing the BMP
by recombinant DNA techniques.

In yet another embodiment, the present invention provides the DNA sec~ence encoding ~MP or analogs thereof, which may be used to construct vectors ~or expression in host systems by racombinant DNA
techniques.

In still another e~bodiment, the present invention provides therapeutic compositions comprising BMP and, optionally, other osteoinductive associated actors such as matrix Gla protein (MGP~ and bone calcification factor (BCF) and ~ethods or forming cartilage and bone in vertebrates by introducing in vivo at the desired site an ef~ective bone initiatinq amount of BMP. The identity of MGP was first reported by Price, Urist ~nd Otawara in ~ioche~.
BioRhys. Res. CQmm. 117:765-771 (1983). ~he identity of BCF is disclosed ~n co~monly assigned copending Serial No~ 360,826, ~iled Jun¢ 2, 1989.

% ~ ~ ~ r~

BRIEF DESCRIPTION OF l~E DRAWI~G~;

FIG. 1 shows the amino acid seguences of eight peptides ro~ tryptic digestion of bovine ~MP:

FIG. 2 ~hows DNA and a~ino acid equences of two exon-containing reqions of the boYine B~P gene, ~IG. 3 shows the nucleotide se~uence o~ bovine BMP
cDNA and the deduced amino acid ~eguence of th~
precursor polypeptide;

FIGS. 4a through 4e show the ~equences of the exon~
containing regions of the human BMP gene.

FIG. 5 shows the nucleotide sequence of human BMP and the deduced a~ino acid sequence o~ the precursor polypeptide.

DETAILED DESCRIPTION OF THE INVENTION

The BMP according to the present invention ~ay be obtained, ~ree of other osteoind~cti~e associated factors, directly fro~ bone sources, by preparative peptide synthesis using chemical methods (such as the Merrifield synthesis method) or by recombinant DNA
technology.

As more particularly describad in Exa~pl~ MP ~ay be obtained from partially purifi~d hu~an, bov~n~, or other vertebrat~ bone extract~ by preparativ~ gel electrophoresis and electroelution of the protein.

BMP nucleic acid sequences may be obtained ~y recombinant DNA methods, such as by s¢reening reverse transcripts of ~RNA, or by 6creen~ng genomic 2~ ?~

libraries from any ~ell. The DNA laay also be obtained by syn~.hasizing the DNA using commonl2,~
available techniques and DNA synthesizing apparatus.
Synthe6is may be advantageou~; because unique restriction site~ ~ay be introduced at the Sime of preparing th~ DNA, thereby ~acilitating the use of the gene in YeCtOrs con~aining restriction sit s not otherwise present in the native source. Furthermore, any desired site modification in the D~A ~ay b~
10 introduced by sys~thesis, without the need to further modi~y the DNA by mutagerlesis~

In general, DNA encoding the BMP may be obtained from human, bovine or other sources by constructing a cDNA
library from mRNA isolated from vertebrate tissue;
and screening with labeled DNA prubes encoding portions of the human or boYine chains in order to detect clones in the cDNA library that contain homologous sequences; or by polym~rase chain reaction (PCR) ampli~ication o~ the cDNA (from mRNA~ and subcloning and screening with labeled DNA pro~es; and then analyzing the clones by restriction enzyme analysis and nucleic acid seguencing ~o as to identify ~ull length clones and, i~ ~ull-lenqth clones are not present in the library, recovering appropriate fragments rom ~he various clones and ligating them at restriction ~ites cvmmon to the clones to asse~ble a d one enc~ding ~ full-length molecule. Particularly preferred D~A probes ~re 8et forth in the accoMpanying examples. Any ~equences missing from the 5' end of the BMP cDNA ~ay be obtained by the 3' extension o~ ~h~ ~ynthetic oligonucleotides complementaxy to BNP sequences using ~RNA as a template. (so-called primer extens~on), or homolo~ous sequences may be ~upplied from known c~NAs derived fro~ human or bovine ~equences a~ ~hown in FIG. 3 or FIG. 5.

~2~2~

The practic~ of ~he present invention will employ, unless otherwis~ indica~ed, conventional molecular biology, microbiology, and recombinant DNA techniques within the skill of the art. Such techniques are explained fully in the li~erature. See ç.~., Maniatis, Fritsch & Sambrook, "Molecular Cloning: A
Laboratory Manual~ (1982); ~DNA Cloning: A Practical Approach," Volumes I and II (D.N. Glover ed. 1985~;
"Oligonucle~tide Synthesis" (~.J. Gait ed. 1984);
I'Nucleic Acid Hybridization~ (B.D. ~ames &
S.J. Higgins eds. 1985); ~Transcription And Translation" (~.D. Hames ~ S.J. Higgin~ eds. 1984):
"Animal Cell Culture" (R.I. Freshney ed. 1986);
"Immobilized Cells And Enzy~es" (IRL Press, 1986);
B. Perbal, "A Practical Guide To Molecular Cloning"
(1984).

In describing the present invention, the following terminology will be used in accordance with the defini~ions set out below.

The term "osteoinductiv~ associated factors" includes factors known in the art which are present in mammalian bone or other ~ammalian tissue and t~nd to co-purify with BMP or BMP activity. Such factors include, but are not limited to, protein~ which have been isolated ~rom bone having reported relat~ve molecular weights by migration on a SDS-PA~ of 34KD, 24KD, 18.5KD, 17.5KD, lS.5KD, 14XD taS cited in th~
U.S. Patent No. 4,7615471), and 6RD ~reported by Price, P.A., et ~1., ~rom ~rot. Natl ~zad. Sci~, 7~, pp. 1447-1451, 1376). All observed ~olecular weights are reported herein a~ relative molecular weight by migration on SDS-P~GE qel.

A "replicon" is any qenetic element ~.~.0 plas~id, chromosome, virus~ that functions as an autonomous 2,0~ 7~

unit of DNA replication ~ Y~Q; i~, capable of replication under its own control.

A "vector" is a replicon, such as plasmid, phage or co~idt to which another DNA ~eg~ent ~ay be attached so as ~o bring about the ~æplication of the attached segment.

A nDNA moleculen re~er~ to the polymeric form o~
deoxyribonucleotides (adeDine, ~uanine~ thymine, or cytosine~ in its either ~rgle stranded ~orm, or ~
double-stranded helix. Thi~ term refers only to the primary an~ secondary ~tructure o the molecule, and does not li~it it to any p~rticular tertiary for~s.
Thus, this term includes dcuble-stranded DNA found, inter alia, in linear D~A molecules (e.g., restriction fragments), vLruses, plasmids, and c~romosomes. In discus~ing the structure of particular double-stranded DNA ~olecules, sequences may be described herein according to th~ normal convention of giving only the seguence in the 5' to 3' direction along the no~ranscribed ~trand of DNA
(~.e., the strand havi~g a sequence homo~ogous to the mRNA).

A DNA "coding sequencea i~ a double-stranded DNA
sequence w~ich is transcribed and translated into a polypeptide in vivo when placed under the con~rol o~
appropriate regulatory ~eguences. The bound~ries o~
the coding seguence are determined by a start codon at the 5' ~a~ino~ terminus and a tran51~tion 6top codon at the 3' (carboxy) terminu~. A coding ~equence can $nclude, but iB not li~ited to, procaryotic seguences, c~NA ~rom eucaryotic ~RNA, genomic DNA sequences from eucaryoti~ (~
mammalian) DNA, and sven ~ynthetic DNA ~eguences. A
polyadenylation #~gnal a~d transcript~on termination 2 ~ 2 ~

sequence will usually be located 3' to ~he roding 6equence.

Transcripti~nal and translati~nal control ~equences are DNA regulatory sequences, such ~s promoter~, enhancers, polyadenylat~on ~ignals, ter~inators, and the like, that provide ~or the expression of a coding sequence in a host cell.

A "promoter seguence~ is a DNA regula~ory region capable of binding RNA polymerase in a cell and initiating transcription o~ a downstream (3' direction) coding sequence. For purposes of defining the present inventisn, the promoter sequence is bounded ~t its 3' terminus by the transcription initiation site and extend~ upstream (5' direction) to include the minimum number of bases or elements necessary to initiate transcription at levels detectable above background. Within the promoter seguence will be found a transcription initiation site (conveniently defined by ~apping w~th nuclease Sl), as well as protein binding domains (con~ensus sequences) responsible for the binding of RNA
polymerase. Eucaryotic promoters will often, but not always, contain "TATA" boxes and nCAT" boxes.
Procaryotic promoters contain Shine-Dalgarno seguences in addition to the -10 and -35 consensus sequences.

A coding sequence i~ "under the controln of transcriptional and translational control sequences in a cell when RNA polymeras~ transcribes the cod~ng sequence into mRNA, which i8 ~hen translated into the protein ~ncoded by the codiny ~equence.

A "signal sequence" can be included before the coding sequence. Th~s sequence encodes ~ ~ignal pept~de, N-~-10~
terminal to the polypeptide, that communicates to the host cell to direct t~e polypeptide to the cell surface or ~ecrete the polypeptide into the media, and this signal peptide i8 cl ipped off by the host cell before the protein leave~ ~he cell. Signal sequenc2s can be found associated with a variety of proteins native to prokaryotes and euXaryotes. FQr instance, alpha-factor, a native yea~t pro~einO is s~creted fro~ yeast, and its ~ignal sequence can be at~ached to heterologous proteins to be secreted into the media (See U.S. Patent 4,546,08~, ~P0 0 116 201, publication date 12 January 1983; U.S. Patent Appli~ation Ser. No. 522,909, ~iled 12 August 1983).
Further, the alpha factor and its analogs have been f~und to secrete heterologous proteins from a variety of yeast, such as Saccharomyces and Xluyveromyces, (EPO ~8312306.9 filed 23 December 1988; U.S. Patent Application Ser. No. 139,682, filed 30 December 1987, and EP0 Pub. No. 0 301 669, publication date 1 February 1989).

A cell has be~n "transformed" by exogenous or heterologous DNA when 6uch DNA has b~en introduced inside the cell. The trans~orming DN~ ~ay or may not be integrated (covalently linked~ into chromoso~al DNA making up the genome of the cell. In prokaryotes, yeast, and ma~malian cells for example, the transforming DNA may be maintained on an episomal element ~uch as a plasmid. ~ith respect to eukaryot~c cells, a stably transgormed cell is on~ ln which the transformin~ DNA has beco~e int¢grated ~nto a chromosoms so that it ~s inherited ~y daughter cells through chromosome replication. ~hi~ ætability is demonstrated by th~ ~bility o the eukarysti~ cell to establish cell lines or clones comprised o~ a population of daught~r cells containing the transforming DNA. A "clone~ i8 a population o~ cells ~2~

derived from a single cell or co~ on ancestor by ~itosi~. A ~cell line~ is a clone of a primary cell that is capable of ~table growth in Yi5~ ~or many gener~tions.

Two DNA sequences are ~substantially homologous" when at least about 85% ~preferably at least about 90~, and ~ost preferably at least about 95%) o~ ~he nucleotides ~atch over the defined lengt~ of the DNA
sequences. Sequences that are ~ubstantially homologous can be identified in a Southern hybridization experiment under, for example, stringent conditions as de~ined ~or that particular system. Defining appropriate ~ybridization conditions is within the s~ill of the art. See, e.g., Maniatis et al., supra; DNA Cloninq, Vol I &
II, supra; Nucl~ic Acid Hybridization, ~eE~-A "beterologous" region of the DNA construct is anidentifiable segment of DNA within a larger DNA
molecule that is not found in association with the larger molecule in nature. Thus, when the heterologous region encodes a ~am~alian gene, the gene will usually be flanked by DNA ~hat does not flank the ~ammalian genomic ~NA in the geno~e o~ the source organis~. Another exa~ple of a heterologous codin~ sequence is a construct where the coding sequence ltself is not ~ound ln nature ~ cDNA
w~ere the genomic coding sequ~nce conta~ns introns, or synthetic ~equences having codons di~ferent t~an ~he native gene). Allelic variations or naturally-occurring mutational events do not give ri~ toheterologous region o~ DNA as de~ined herein.

A composition comprising "A~ ~wher~ ~A~ is ~ ~ingle protein, DNA ~olecule, vector, etc.) ~5 ~ubstanti~lly free o~ "B" ~where "B" comprises one or more J ~d ~

conta~inating protein~, DNA ~olecules, vectors, etc.) when at least about 7~% by weight of the proteins, DNA, vectors (depending on ~he catego~y of ~pecies to which A and B belonq) in the composition is "A".
Preferably, ",~" comprises ~t least about 90% by weight of the A+B species in the compo~ition9 most preferably ~t least about 99~ by weiqht. It is also preferred that a composition, which is substantially free of contamination, con~ain only a single molecular weight spec~e~ having the activity or characteristic of the species of intere~t.

An "antibody~ is any immuno~lobulin, including antibodies and fragments ther~of, that binds a specific epitope. The term encompasses, inter alia, lS polyclonal, monoclonal, and chimeric antibodies.

For more about chimeric antibodies see U.S. Patents Nos., 4,816,397 and 4~816,567.

The intron-frae DNA provided by the present invention is novel, since it i5 believed ~hat the naturally~
occurrinq human a3~ ~ovine genes contain introns.
Hence, the term "intron-free" excludes the DNA
sequences which naturally occur in the chromosomes of human or bovine cells. The present invention also encompasses the intron-free cDNA sequences derivable from ~he DNA ~eguences disclosed herein.

As more particularly describ~d in th~ ~ollowing examples, ~uman and bovine cDNA libraries wer~
initially probed f~r seque~ces encod~ng ~P ~equences using la~eled oligodeoxyr~ucleotid~s whose seguences were based on a partial amino acld ~equence determined from analysis of puri~ied protein sa~ples derived from bone de~cri~ed herein. 80wever, it is realized that once being provided with intron-free .

2~2~

DNA encoding human and bovine BMP and their leader ~equences as de~cribed herein9 one sf ordinary skill in the art would recognize that oth~r precisely hybridizing pro~es may be prepared fro~ the described seguences in order to readily obtain ~he remainder of the desired human or bovine gen~.

Vectors are used to si~plify ~anipulation o~ the DNA
which encodes the BMP polypeptide, either for preparation of large guantities of DNA for further processin~ (clonin~ vectors) or for expre sion o~ the BMP polypeptid~ (expression vectors). Vectors comprise plasmids, viruses (including phage), and integratable DNA fragments, i.e" fxagments that ar integratable into the host genome by recombination.
Cloning vectors need not contain exprsssion control seguences. ~owev~r, control sequences in an expression vector include transcriptional and translational control sequences ~uch as ~
transcriptional promoter, an optional operator sequencP to control transcription, a seguence encodinq suitable ribosome binding sites (~or prokaryotic expression), and sequences which control termination o~ transcription and translation. The expression vector æhould pref~rably include a selection gene ~o ~acilitate the stable expression of BMP and/or to identify transforman~s. However, ~he selection gene for ~aintaining expression can be supplied by a separate Yector in cotran~or~ation systems using eukaryotic host cell~.

Suitable vect~rs generall~ will contain replicon (orig~ns of replication, for use in non-integrative vectors) and control ~equen~e~ which are d~rived ~ro~
species compatible with the ~ntend~d expression host.
By the term ~replicable" vector a~ used herein, it is intended to encompas~ vectors containing ~uch replicons as w011 as vectors which are replicated by integration into the ~o~t genome. Transformed host cells are cells which have been transfor~ed or transfected with vectors containing BMP encoding DNA.
5 The expressed BMP will be deposi~ed intracellularly or secreted into either the peripl~s~ic space or t~e culture ~upernatant, depending upon the host cell selected and the presence o~ suitable processing signals in the expressed peptide, e.g. homologous or heterologous signal sequences.

Suitable host cells are prokaryotes or eukaryotic cells. Pro~caryotes include Gram negatiYe or Gram positive organisms, for example ~. col$ or bacilli.
Eukaryotic c~lls ~nclude yeast or higher eukaryotic cells such as established cell lines of ~a~malian origin.

Expression vectors for host cells ordinarily include an origin of replication, a promoter located ups~ream from the BMP coding ~equence, together with a 20 ribosome binding ~ite, a polyadenylation site, and a transcriptional termination sequence. Those o~
ordinary skill will appreciate that certain o~ the~e sequences are no~ reguired for expression in certain hosts. An e~pression vector or use with microbes need only contain an origin of replicatio~ recognized by the host, a promoter which will function in tbe host and a selection gene.

An expression vector i6 constructed according to the present invention so that ~he ~MP coding ~equence is located in the vector wi~h ~he appropriate regulatory sequances, the positioning and orientation of the coding sequence with respect to the control sequences being such that the coding ~equence i8 tran~cribed and translated under the ~control" of the control 2~7~

sequences (l.ç., ~NA polymerase which binds to the DNA molecule at the control equences transcribes the coding seguence). The control sequences Day be liga~ed to the coding sequence prior to insertion into a ~ector, such as the cloning vectors described above. Alternati~ely, ~he coding sequence can be rloned directly into ~n expres~ion vector ~hlch already contains ~he control ~equences and an appropriate restriction site. For expression of BMP
in prokaryotes and yeast, the control sequences will necessarily be heterologous to the ~oding seguence.
If t~e host cell is a prokaryote, it is al50 necessary that the ~odin~ ~quence be Pree o~ introns (e.g., cDNA). If t~e selected host cell is a mammalian cell, the con~rol sequences can be heterologous or homologous to the BMP coding seguence, and the coding sequence can either be genomic DNA containing introns or cDNA. ~ither genomic or cDNA coding ~equences can be expressed in yeast.

Expression vectors must contain a promoter which is recognized by t~e host organism. Promoters commonly known and availabl~ which are used in recombinant DNA
construction include the ~-lactamase tpeni~illinase~
2s and lactose promoter systems, ~ tryptophan ~trp) promoter syste~ and the tac promoter. ~hile these are commonly used, other known ~icrobial promoters are suitabl~O

In addition to prokaryotes, eukaryotie cells such a yeast are transformed with BMP encoding vector~.
Saccharomyces çere~is~ae, or common baXer'~ yea~t, ~8 the most com~nly u~ed a~ong lower eukaryotic host microorganisms. ~owever, a number o~ other species ar~ commonly available ~d userul herein. Yea~t vectors generally will contain an origin of p~

replication from the 2 micron yeast plasmid or an autonomously replica~ing ~equence (ARS), a promoter, DNA encoding B~P, seguences for polyadenylation and transcription termination, and a selection gene.

- 5 Suitabl~ promoter fieque~es in yeast vectors include the promoters for the glycolytic enzymes 6uc~ ~6 enolase, 3-pho~phoglycerate k~nase, glyceraldehyde-3-phosphate dehydrogenase, hexo~inase, pyruvate decarboxylase, phosphoructokinase, glucose-6-phosphate isomerase, 3-phosp~oglycerate mutase, pyruvate kinase, triosephosphate isomeras~, phosphoglucose isomerase, and glucokinase.

Other yeast promoters, which have the additional advantage of transcription controlled by growth conditions ar the promoter regions ~or alcohol dehydrogenas~ ~ or 2, isocytochrome C, acid phosphatase, as w~ll as enzymes responsible fvr maltose and galactose utilization.

Higher eukaryotic cell cultures may be used, whether from vertebrate or invertebrate oells, inoluding insects, and the prooedures of propagation thereo~
are known. See, for example, Tissue Cultur~, Academic ~ress, Kruse and Patterson, editors (1973).

Suitable host cells for expressing BWP in higher eukaryotes include: monkey kidney CVI line transformed by SV40 (COS-7, A~CC ~L 16513; ~aby hamster kidney cells (BH~, ATCC ~RL 10): Ch~nes~
hamster ~vary-c~lls-DHFR ~describe~ by Urlaub and Chasin, PNAS (USA) 77: 4216 (1980)): ~ouse Sertoli cells (T~4, Mather, J.P., ~i~l. Repro~L 2~: 243-251 (1980)); monkey kidney cells (~VI A~ce cC~ 70) African green ~onkey kidney cells (VERO-76, ATCC C~L
1587): human cervical carcinoma cell~ ~HEL~, ATCC CCL

- 2~2~ 1~2~

2); canine kidney cells (MDCK, ATCC CCL 34): buffalo rat liver cells ~BRL 3At A~CC CRL 1442); hu~an lung c~lls ~W138, ATCC CCL 75); hum~ liY~r cells (Hep G2, ~B 8065); mou~e mammary tumor (MMT 06065~, ATCC CCL
51~; rat hepatoma cells (HTC, ~1~ 54, Bau~ann, ~0, ~t . Cell_Biol t_ 85: 1 8 (1980)) and ~RI c¢115 (Mather, J.P., e~ al., Annal~ ~.Y. Acad. S~i. 3~3:
44-68 (1982)). Co~monly used prom~ters are derived from polyoma, adenovirus 2, and simia~ virus 40 (SV40). It will be appreciated that when express~d in mammalian tissue, the recombinant BMP may have higher molecular weight due to glycosylation. It is therefore intended that partially or completely glycosylated forms o~ BMP having molecular weights greater than l9KD are within the scope o~ this invention as well ~s its unglycoeylated forms.

A number of procaryotic expression vectors ar~ known in the art. See, e.~., U~S. Patent Nos. 4,440,859:

4,436,815: 4,~31,740: 4,431,739; 4,428l941;
4~425~437: 4~418~149; 4~4~1~994; 4~366~246;
4,342,832; see also U.K. Pub. Nos. GB 2,121,054;
GB 2,008,123; GB 2,007,675; and European Pub.
No. 103,395. Preferred procaryotic expression sys~ems are in E. coli.

Other preferred expression vectors are those ~or use in eucaryo~ic systems. An exe~plary eucaryot~c expression system is that e~ploylng vac~inia virus~
which is well-known ln the art. See. ~.g., ~ac~et et ~ 1984) ~. V~rol. ~:857; ~DNA Clo~ing,~ Vol. II, pp. 191-211, ~upra; PCT Pub. No. W~ 86/07593. Yeast expression vectors are known in the art. See, U~. Patent Nos. ~,~46,235: 4,~43,S39~ 30,428;
31~ European Pub. No~O 103,409; 100,56~; 96~491.
Another preferred expre~s~on 8y5tem i~ ~ector pHSl, which transforms Chinese ha~ster ova~y cell~O See 2~7~

-1~
PCT Pub. No. Wo 87/02062. Mammalia~ tissue may be cotransformed ~ith DNA encoding a selectable marker such as dihydrofolate reductase (DHF~j or thymidine kinase and DN~ encoding ~P.

I~ wild type DHFR gene is employed, it i~ pre~erable ~o select a host cell which is deflcient in D~FR, thus per~itting the use o~ ~he ~HFR coding sequence as ~arker ~or successul trans~ection in hgt~ medium, which lacks hypoxanthine, glycine, and thymidine. An appropriate host cell in this case i the Chinese hamster ovary (CH0) cell line deficient in DHFR
activity, prepared and propagated ~s described by Urlaub and Chasin, 1980, Proc. ~at. Acad. Sci. (USA) 77: 4216. Expression vectors derived from baculovirus for use in insect cells are known and available in the art. See L~cklow and 5um~ers, Hiotechnolo~y, 6, p. 47-55.

Dapending on the expression syste~ and host selected, BMP is produced by growing host cells transformed by an exogenous or heterologous DN~ construct, such as an expression ~ector descrlbed above under conditions whereby the BMP protein is expressed. The ~NP is then isolated from the host cells and purifiedO If the expression system secretes BNP into growth media, the protein can be puri~ied directly from cell~ree media. If the BMP prot~in i~ not ~ecreted, it 1~
isolated from cell lysates. The selection of the appropr~ate growt~ conditions and recovery ~ethod~
arQ within the skill o~ ~he art.

The recombinantly made aMP i~ recovered from transformed cell~ in accordancz wi~h known procedures. Preferably, an expression ve~tor will ~e used which provides for &ecretion of BMP fro~ the transformed cells; thus the cells ~ay be ~eparat~d by 2 $ ~

centrifugation. The B~P typically is purified by general protein purificatio~ techniques, including, but not limited to, size exclusion, ion-exchange chromat~graphy, HPLC, and the like.

Once a ~oding sequence ~or BMP ha~ been pr~pared or isolated, it can be cloned into any suitable vector and thereby maintained in a composition o~ cell~
which ~s 6ubstantially ~ree o~ cell~ that do not contain a BMP codin~ seyuence ~e.~., free of other l~brary clones). Num~rou~ cloning vector~ are known to those of ~kill in the art. Exa~ples o recombinant DNA ~ectors for cloning and host cells which they can transorm include the various bacteriophage lam~da vector~ (~. coli), pBR322 (E. coli), pACYC1~7 (~. c~), pXT230 (gram-negativ~
bacteria), pGV1106 (gram-negative bacteria), pLRFR1 (gram-negative ~acteria), pME290 (non-E. coli gram-negativ~ bacteria), pHV14 (E. coli and Bacillus subtilis), pBD9 (Bacillus), pIJGl (Streptomyce~3, pUC6 (Streptomyc~s), actinophage, ~C31 (Streptomyces), YIp5 (Saccharo~yces), YCpl9 (SaccharomyceR), and bovine papillo~a virus (mammalian c~lls). ~ gDr31~y, DNA Cloning:
Vols. I & II, supra; T. Maniatis et a~., sup~;
B. Perbal, supr~.

Alternatively the BNP ~ay ~e ~ade by conventional peptide synthesis, for instance, by using th~
principles of the Merri~ield 6ynt~esis and using co~mercial automatic apparatuC de~igned to emplsy the ~ethods of the ~errifleld synthe~i~O PeptideY
prepar~d using convention~l peptide ~ynthesi~ ~ay be purified using conventional affin~ty chro~atography, qel filtrat~on and/or RP-HPLC.

2~2~ ~2~

Figure 3 ehows the nucleotide sequence o~ ~ovine B~P
cDNA and the deduc~d a~ino acid sequence or the precursor polypeptide. The putative signal peptidase cleavage site is ~oted ~). The cDNA sequ~nce was obtained from a 830 bp Bgl II insert of bovin~ BMP
cDNA clone #1~ which was isolated ~ro~ a cal~ liver cDNA library, as described belowO

Fiqure 5 shows the nucleotide s2quence of human BMP
cDNA and the d~duced a~ino ac~d seguence of the precursor polypeptide. The putative s~gnal peptidas~
cleavage site is noted (-). The cDNA sequence was obtained from a 600bp BamHI/HindIII inser~ of hu~an BMP cDNA clone A6, which was isolated ~ro~ a BMP/PCR
human kidney cDNA library, as described below.

It is further intended from the nucleotides seguences in Figures 3 and 5 that BMP analogs are within the scope of ~he present invention. Analogs, such as fragments, may be produced, for example, by pepsin digestion o~ BMP. Other analogs, such as muteins, can be produced by stan~ard site-directed mutagenesis of BMP codi~g sequences. Analogs exhibiting ~RMP
activity" ~ay be identified by the in vivo and/or in Yi~Q assays, preferably the i~ vi~ro cartilage inducing assay, Methods of ~n~y~olQ~y, 14~, pp 29~-312, (1987).

~n example of a BMP analog are the yeast cleavageproducts produced in vivo fro~ the full-length ~ovine or human BMP expressio~ product. The processing of tha bovine expression product results in two roug~ly 16KD polypeptides, described ~n Example 9 below, ~nd one or both of ~hese analogs ~re referred to her~in as the "16KD yeast cleavage analog.~

2~3~7~

--~ 1 As mentioned a~ove, a DNA seguence encoding BMP can be prepared ~ynthetically rather than cloned. The DNA sequence can be designed with the appropriate codons for t~e BMP amino acid sequence. In general, one will sel~ct preferred codons for ~he intended host i~ the sequence will be used Por expres~ion.
The co~plete ~egusnce is assembled from overlapping oligonucleotides prepared by standard methods and asse~bled into a co~plete coding sequence. fi~, ~.g., Edge (1981) Nature ~ 756: Na~bair, et a~.
~1984) Science ~ 299; Jay e~ ~1. (19843 io~
Çh~m. ~ 6311.

Synthetic DNA sequences allow co~venient construction o~ genes which will express BMP analog~ or "muteins".
~lternatively, DNA ~ncoding muteins can be ~ade by site-directed mutagenesis of native ~MP genes or cD~As, and muteins can be made directly using conventional polypeptide synthesis.

Site-directed mutagenesis is conducted using a primer synthetic oligonucle~tide comple~entary to a sinqle stranded phage DNA to be mutagenized exc~pt ~or limited ~ismatching, representing the desired mutation. Brie~ly, the syntheti~ oli~onucleotide is used as a primer to direct synthesis ~ a strand complementary to the phage, and the resulting double-stranded DNA ~s transformed into a phage-supporting host bac~eriu~. Cultures o~ th~ transformed bacteria are plated ln top agar, permitting plaque formation from single cells which harbor the phag~.

Theoretically, 50~ of the n~w plague~ ~ill contain the phage having, ~8 a 6ingle strand, the ~utat~
form; 50% will have the original sequence. The resulting plaques are hy~ridized with kinased synthetic primer at a te~perature which p~r~

2 ~

hybridization of an exact ~atch, but at which th~
mismatches with the original strand are ~uficien~ to prevent hybridization. Plaques which hybridize with th~ probe are ~hen picked, cultured, and the DNA
recovered.

A general method for ~ite-specific incorporation o~
unnatural a~ino acids lnto prot~ins is described in Christopher J. Noren, Spencer ~. Anthony-Cahill, Michael C.. Griffith, Peter G. Schultz, (April 1989), Science, Vol 244, pp 182-1~8. This ~ethod may be used to create analogs with unnatural a~ino acids.

Preparations of BMP and its analogs ~ay be assayed vivo according tQ the method described by Urist, e~
al., Methods in Enzymo~o~y (D. Barnes and D. A.
Sirbaska, eds.), vol. 146, pp. 294 312, Acade~ic Press, N.Y. (1987), and 1~ Q by the method o~
Sato and Urist, Clin. Orthop., 183:180-187 (1984) as modified by Kawamura and Uri~t, ~ev. ~iol., 130.435-442 (1988), all of which are incorporated by reference herein.

Substantially pure ~MP, hlqher molecular glycosyla~ed for~s thereof, or active ~ragments t~ereof, or the nontoxic salts thereof, combined with a pharmaceutically acceptable carrier to form a phar~aceut~cal composit~on, ~ay be ad~ini6tared to mammals, ~ncluding humans, either intravenously, subcutaneously, percutaneously, intramuscularly or orally.

Such proteins are often administered in th~ ~or~ o~
pharmaceutically acceptable nontox~c salt~, ~uoh ~s acid addition salts vr ~etal complexes, e.g., with zinc, iron or the like (which are considered as salts for purposes of ~his application). Illustrative o~

~23- 2 such ac~d addition ~lts ar~ hydrochl~r~d~, ~ydrobromid~, ~ulphat~, phosp~at~, ~aleat~, acetate, citr~t~, benzoat~, ~uccln~t~, ~alat~, ~scorbat~, tartrat~ and th~ llk~. If ~he act~ve lngredient 18 to be ~dministered ln tabl~t for~, ~he tablet may contaln a b~nd~r, ~uch a~ tragacanth, co~n starch or gel~tin; ~ di~intQgratlnq agQn~, ~uch ~s alg1nic acid; and ~ lubricant, ~uch ~5 ~agnesiu~ ~tearat~.
I~ ad~inistration ;n l~qu~d for~ ~ de3~r~d, 6weeten~ng and/or flavoring ~ay b~ use~, and lntravenous ad~ni~tratlon ln icotonlc ~alln~, phosphate buf~er solut~ons ~r the llXe may b8 affect~d.

Pharmaceutical compositions will usually contain an 15 e~fective amount o BM~ in con~unction witb ~
convention~l, phar~saceutic~lly accept~bl~ carri~rO
The dosage will vary dependinq upon the sp~c~ ~s:
purpose for wh~ch the pr~tein ~ being a~in~tered, and dosage levels ln t~e range of about 001 sJg to 20 about 100 ~ gra~D,s per ~g. of b/~dy ~Qre~gl~t may bg used .
Thc present invantion further ~ontampl~tes the combination of BCF, part~cularly hu~an or bo~in~3, with B~P ~8 a therapeut~c ~:o~po8~Lt~on or ~n~tiat~ng l:one as fo~mation both in vitro and ln YiVo.
~Dplants of rsco~>irant B~P, alon~ or wl~sn ~lxed witl~
one or ~ore osteolnductive assoelated ~actors ~ay b~
used to init~ate cartilage and ~ne growth. Tha lmplants can be time-release compositlon 30 encapsulated"ror lnstancQ~ ln liposome~ or o~er e~me-release lDembranes~ na'cural or synthetlc:, whlch are absorbabl2 by the host su~ect. Ta~ y b~
e~t~er human, bovine or any o~er D~a~allan ~orla, o~
~oixtures thereo~, To initia1;e bon~ growth, ~ y 35 be ~nixed wit~ any comblnatlon O~r one or a~or~ otb~r protein~, partlcularly, wit~ on~ or mor~ oth~r prote~ns derived :~ro~ bon~. Such mixtur~s ~o~y initia~e cartilzlg~ formation~ follow~d by bon~ I
growth. Implant~ o~ aay ~e u~e~ to indu~

2 ~
.

cartilage and bone growth in the quadriceps compartment.

The purification protocols, de~cribed in detail below, allow for the fir~ ti~e ~he puri~icatlon o~
native B~P in ~uf~icient quantity and a~ a high enough purity to perfflit accura~e amino acid seguencing. The amino acid seguences derived from the purified ~MP ~llow for the design of probes to aid in the isolation of nativ~ ~MP nucleic acid sequence, or the design of ~ynthet~ nuclei~ acid sequences sncoding the amino acid ~equence of BMP, as well as allowing both diagnost~c and therapeuti~
antibodies to BMP and its analogs to be produced for the first time. BMP DNA genes or fragments may also be utilized in a diagnostic test ~or identifying subjects having defective BMP-genes.

Specific anti-sera or monoclonal antibodies (described below) can be ~ade to a synthetic BMP
peptide having the sequence or fragments of the sequence of amino aci~ residues, such as those shown in Figures 3 or 5. Examples are th~ tryptic fragments shown in FIG. 1, and antibodies thereto can be used to immunoprecipitate ~ny BMP present in selected tissue, cell extract, or body fluid.
2S Purified BMP from this sourc¢ can t~en be seguenced and used as a basis for designing speci~ic probes as dsscribed above. Antibodie5 to other regions that diverge ~rom known aMP ca~ also be u~ed. Also u~eful as antigens are purified natiYe or recombinant ~P.

Natlv~, recombinant or ~ynthetlc ~MP peptides (Pull length or subunits) can be used to produce bo~h pclyclonal and monoclonal antibod~es. I~ polyclonal antibodies are desired9 puri~ied BMP peptide is used to im~un~ze a selected ~ammal (~.~., ~ous~, rabbit, (~ f . 2q~ v~

goa~, horse, etc.) and serum from the immunized animal lat~r collected and treated according to known procedures. Compositions contain~ng polycl~nal antibodies to a variety of antigen ~n addition to BMP can be made substantially ree of antibod~es which ~re not anti-BMP by i~munoaffinity chromatograp~y.

Monoclonal anti-BMP antibodies can also be readily produced by one skilled in the art ~rou the disclosure herein. The general methcdology ~or making monoclonal antibodies by hybridomas is well known. I~ortal, antibody-producing cell lines can also be created by techniques other than fusion, such as direct transformation of B lymphocytes with oncogenic DNA, or transfection with Epstein-Barr virus. See, e.~., M. Schreier e al., "Hybrido~a Techniques" (1980); Hammerling et ~ onoclona~
Antibodies And T-cell Hybridomas" (1981); Rennett e al., "Monoclonal Antibodies6' (1980); se~ also U.S.
Pat~nt Nos. 4,341,761, 4,399,121: 4,427,783;
4,444,887; 4,451,570; 4,~66,917; 4,472~500;
4,491,632: 4,493,890.

Panels of monoclonal antibodies produced against 3MP
peptides can be screened ~or various propertie~:
~.~., isotyp~, epitope, affinity, etc. 0~ particular interest are monoclonal antibvdies that neutralize ~he activity of BMP, Such monoclonals can be readily identi~ied in ~MP activity assays. ~igh a~finity antibodies are also use~ul ~n ~mmunoa~~nity purif~cation o native or reco~bi~nt ~P.

Antibodies to B~P (both polyclonal and ~on~clonal) forms described herein may be ~ed to inhibit or to reverse various clinical indications of bone disease such as osteoporosls, osteoarthritls, etc. One 2 .~ r~

therapeutic approach would be to treat ~he patient with an effective dos~ of anti-BMP ~ntibodies thr~ugh a c~nventional intravenous route. ~P antagonists or agonists, such as ~NP ~uteins, could also be used in place of antibodies. These anti-BMP compositions may also be useful in dec~ecting or inhi~it~nq various forms of tu~ors, since some tu~ors are ~now~ to be induced by growth ~actor~.

The determination of he appropriate treatment regimen ~i-e-, dosage, frequency of administration, systemic vs. local, etc.) is within the skill of the art. For administration, the antibodies will be formulated in a unit dosage injectable form (solution, suspension, emulsion, etc.) in association with a pharmaceutically acceptable parenteral vehicle. Such vehicle are usually non~oxic and nontherapeutic. Examples of such vehicles are wat~r, saline, Ringer's solution, dextrose solution, and Hank's solution. Nonaqueous vehicles such ~s Pixed oils and ethyl oleate ~ay also be used. A preerred vehicle is 5% (w/w) human albumin in saline. The vehicle may contain minor amounts of additivesO ~u~h as substances t~at ~nhance isotonicity and che~ical stability, e.~., bu~fers and preservatives. The antibody is typically ~or~ulated in such ~ehicles at concentrations of about 1 ~g/ml to 10 ~g/ml.

Anti-BMP antibodies will also be useful ~n diagnostic applications. The presen~ invention conte~plates a method, part~cularly a diagnostic ~ethod, ~n which 6ample fro~ a human ~or o~her ma~al) 1~ provided, and the amount of BNP ~ quantitativ~ly ~easured in an assay. For example, ~mploy~ng anti-BMP ~ntibodies in a quantitative im~unoassay could be used to detect genetic deficienoy in BMP. Antibody specif~c for BMP
could be formulated into any conventional i~munoassay f i ~2~ ~?ri?~

for~at,o ~., Ao~o~eneous ox heterogeneous, radioim~unoassay or ELISA. The variou format~ are well Xnown to those skilled in the art. See, e.~., "Immunoassay~ A PractiGal Guide" (D.W. Chan and M.T.
P~rl6tein eds. 1987) the disclosure of which is incorporated her~in by referenceO

In general, production of r~combinant BMP can provide compositions of that polypeptide 6ub~tantially ~ree of contaminating proteins. The ability to obtain high levels o~ purity is a r~sult o~ recombinant expression syste~s which can produce BMP ~n substantial quantities YiS-a-ViS in ViYo sources.
Thus, by applying conventional techniques to recombinant cul~ures, BMP compositions can be produced that are substantially more pure than the compositions available from bone sources.

Purified BMP will be particularly us0ful as a tool in the dçsign and screening of cartilage or bone growth inhibitors. First, ~illigram a~ounts of the material are obtainable according to the present invention.
Milligram amounts are capable of crystallization to permit thrae dimensional ~tudies using X-ray diffraction and computer analysis. This may permit deduction concerning the ~hape of the ~olecule, thus defininq proper shapes ~or substances usable as inhibitors of the activity nor~ally exhibited by BMP.
Generally, antagonists have bee~ pep~ides whose interactions wi~h an a polypeptide, tbe activity o~
which is inhibited, are stabilized by modification of 30 the ~Iresidues~ participat$ng ;n the peptide bond ~o as to enhanc~ the ability of the pept~d~ to interact specif~cally with the ~nzym~, receptor, or co-~actor, ~;uch as osteoinductive associated actors in the case of BMP. ~hus the peptide bond jo;ns specifically 35 chosen carboxyl i c acid~ and amines (not necessarily 2 ~ ?~ ~

amino acids). These peptides ar~ conigured in a three dimensional array ~o as to comple~ent the contours of the intended target. A similar lock and key spatial arrange~ent may result ~ro~ molecules designed complementary to the surface eontours of the BMP o~ the lnvention. I~ ~s understood that ~surface~ includes convolu~lons which ~ay face inward, and specifically includes the active ~ite.
Furthermore, "coMplemen~ary" ls understood to ~ean that, in addition to spatial conformations which ~ f itn ~ interactions bet~en the protein and the mol~cule which ~atches its surface contours are attractive and positive. These interactions may be hydrogen bonding, ionic, or hydrophobic affinity.

Accordingly, the invention contemplates peptide antagonists and agonists (2-15 amino acids) to BMP
which are characterized by three di~ensivnal contours complementary to the three dimensional contours on the surface of recombinant ~M~. By peptide in this context i5 meant that the antagonist or agonist contains carboxylic aci~ amide bonds. The carboxylic acid and amine particip~nts need hOt be ~-amino acids.

Second, even wi hout th~ assistance of a three dimensional structure d~termination, purified BMP of the invention is useful as ~ reagent in screening BMP
inhibitors ~ E~ as ~n ~d h~ approach to evaluat~on. Impure BMP preparation~ currently available yield con~us~ng data due to i~pur~tie~.
For example, contaminants which turn out to be themselves inhibitor~, ~ctivators, or ~ubstrat~s ~or BMP may interfere with ~he evaluation. Thus, substantial improvement..in current screening technigues gor BMP agonists and an~agonists would ~e 2 ~

effected by the availability of the puri~ied BMP
protein~

It will b~ understood that this description and disclosure ~f the invention i~ intended to cover all S chang~s and mvdifications of the invention which are within the spirit and ~cope og ~he invention. It is within the knowledge og the art to insert, delete or substitute amino acids within the amino acid sequence of a BMP without ~ubstantially affecting the calcification and bone growth inducing activity of the ~olecule. Thus, the invention includes such deletions, additions or subst~tution~. Further~ore, it is recognized that one skilled in the art could recombinan ly produce such modified proteins.

The following examples are provided by way of illustration but are not intended o limit the invention in any way.

~ XAMPI~E.. l purification of BMP Fro~_Bone The BMP proteins o~ interest, partially purified from hu~an (17KD) and bovine (l9~D) sources as described by Ur$st, et ~1., Proc. Nat.~cad. ~ 371-375 ~1984), were ~urther puri~ied to hoDogeneity by preparat~ve gel electrophoresi~ and electroelutio~
IM.W. Hunkapiller, E. Lujan, F. Ostrander and L.~.
Hood, Method~ i~ Enzymoloqy, ~1: 227-236 ~1983)).
This purification showed that the init~al part$ally pur~fied samples contained, in addition to the l9~D
BMP, other mammalian proteins at 34RD, 22XD, 14XD and 6KD. After precipitation with acetone (~
Ronigsberg ~nd L. Hender~on, ~ethods ~n ~nzymQl~gy, ~1: 254-259 (1983~) and quantitation by amino acid analysi~ (B.A. Bidlingmeyer, S.A~ ~ohen and T.L.

~ 2~2~P~

30~
TarYin, ~Q~rnal~of Ch~o~ato~rap~y, ~3Ç: 93-104 (1984)), the materi~l was reduced under denaturing conditions wi~h 2~mercaptoethanol and cysteine residues were derivatized with ~-~inyl-pyridine (~.
Friedman, L.G. Rrull and J~Fo Cavins, ~çYrn~l_~f Biol wical rhemis~y, ~45: 386~-3871 (1970))0 A~ter exhaustive dialysis to remove ~he denaturant, protein recovery was assessed by a repetition of amino acid analysis. The proteins were di~ested with TPCX-trypsin in the presence of 2~ urea ~o generateunblocked peptide fragments suitable ~or seguence analysis (G. Allen, Se~uencinq o~ Proteins ~n~
Peptide~s., pages 51-62 (1981~, Else~ier/North ~clland Publishing Company, A~sterda~, Holland). Products of the digestion were resolved by reverse-phase high performance liquid chromatography using gradients of acetonitrile or acetonitrile/isopropanol in aqueous trifluoroacetic a~d (J.E. Shively, Methods of protein Microcharacterization, pages 41-87 (1986), ~u~ana Pre~s, Clifton, New Jersey). Peptide fractions were sub;ected to automated Ed~an degradation using an Applied Biosystems 470A protein seguencer (M.W. Hunkapiller, R.M. ~ewick, W.J. Dreyer and L.E. Hood, Methods in Enzymoloqy, ~1: 399-413 (1983)). The phenylthiohydantoin a~ino acid derivatives were identi~ied by chromatography on an Applied Biosystems 120~ PTH analyzer (M.W.
Hunkapiller, pplied Biosystem~ er Bulletin Number 14 (1985), Applied Biosystems, Fost~r C~ty, California).

E~
~NA Isolation nnd~ Ll~oe~

Cell cultu~. 293, ~n embryonic h~an kidney cell line obtained ~ro~ the ~merican Type ~ulture Collection (ATCC No. CR~ 1S73) w~s cultured in 2 0 2 ~ I r I ~

Dulbecco's modified Ea~le's medium containing 109~
fetal c:alf serum, 100 U/~nl penicillin and 100 f,g/ml streptomycin a~ 37 C in 5~ G02 0 RNA isolatio~. RNA was isolated from calf liver (obtained from JR Scientific, ~oodland, CA~ by the guanidin~um thiocyanate/CsCl m~th~d (Maniatis, T.
Fritsch, E.F. and Sa~brook, J~ (1982) ~olecular Cloninq: A ~aboratory Manu~l ~Cold Spring Harbor Lab., Cold Spring Harbor, NY) and Free~an, G.J., Clayberger, C., DeRruyff, ~., Rosenblum, D.S. and Cantor, H. (1983) PXOC. Na~ cad. Sci~: USA 80:
4094-4098). Poly(A)~ ~NA was p~ri~ied by a single fractionation over oligo(dT)-cellulose (Maniati-, T.
Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloninq: a Laboratory ~anual (Cold Spring Harbor Lab., Cold Spring Harbor, NY)).

Oliaonucleotide synthesis. Oligonucleotide adapters, primers and probes were synthesized by the phosphoramidite method with an Applied Biosystems tFoster City, CA) model 380A sy~thesizer, purified by polyacrylamide ge1 el~ctrophor~sis and desalted on a Waters SEP-PAX (C18) cartridge.
(a) Adap~er~. ~ 14 mer oligonucleotide (5' CCTGTAGATCTCCG 3') and a 18-mer oliqonucleotide ~5' AATTCGGAGATCTAC`AGG 3') were synthesized and used as the Eco~I adapter5 for the cDNA l~brary constructed ~n lambda ZAPII. The 14-mer was phosphorylated (Maniatis, T. Fritsch, E.F. and sambroQk9 J. (1982) Molecular Cloning: ~ ~aborato~ Manual (Cold Spring 30 Harbor Lab., Cold Spring Harbor, NY)) and subsequently heated to 95''C for 15 ~in. to inactivate polynucleotide kinase, prior to annealing with the 18-mer. These asymmetric:ally pho~phoryla1:ed adapter~
also contain ar~ internal BglII restriction enzyme 35 site.

(b) PCR primer~i. Two 47-mer oligonucleotides were synthesized for PCR ampli~ication of the complete c:oding region of ~uma~l ~MP c~NA rom kidney cell line 293 ~JA and subsequent cloning into lambda ZAPII. The 5' ense primer ~5' TCCGGACTCGAGGAATTC AAACAA~CATGA m CCAGAATGGAG~AG 3'~
and 3' antisens~ prl~er ~5' GACTCGAGGAATTCGTCGAC~CCTCAAGGCCGTTACTCAAA GTCAGT 3') are based on the sequenc~ o~ the human gene exons 1 and 7, respectively. Both primers have 5' extensions (adapters) that contain several restriction enzyme sites.

Two additional sets of PCR pri~ers/adapters were synthesized for ampli~ying bovine and human ~MP cDNA
clones tencoding only the mature protein) and subsequent cloning into pAB125, a plasmid containing the yeast ~-factor leader seguence. Th~ 5' sense primer (42-mer) for the human BMP cDMA (5' ACTACTGGT
CTAGATAAAAGATTCCCAGTGTACGACTACGAT 3'3 and the 5' sense primer (41-~er) Por the bovine B~P cDNA (5' GGAACCCTCTCTAGATAAAAGATTCCCGGTGTATGACTATG 3') contain an XbaI ~ite to facil~tate ligation o the PCR
amplified cDNA to ~e n-~actor leader sequence. TAe 3' antisense primer (37-:ner) for the human BMP cDNA
25 (5' TCTTACCqGTCGACTATTACTCAA~GTCAGTATTTAT 3') and the 3' antisense primer (41-mer) for the bovine B~P cDNA
( 5 ' AATG&CCG~CGAClæI U~CTC~AAGCCAGGGTTTACTCIY~G 3 ' ) contain a Sal~ sit~ ~irectly after the stop eodon.
(c) Probes. Oligonucleotide probes, ~a~ed on 30 the tryptic peptide sequence~ of bov~ne :~P, w~re synthesized ~or 2;cr~enirlg the bovine g~nomis:: library (Probes A-D) and are ~;hown in Figure 1. The actual probes are complementary to the sequences shown. The sequences of ~our additional BMP tryptic peptides 35 were determined (E-H) and are also shown in F~g. 1 With some codons, two nuc:leotides w~re ~ncluded at ~ ~ 2 ~

degenerate p~sitions to increase tha probability of a correct guess. At serine residues, two separate oligonucleotides for each probe ~diferins only at the serine codon) had to be 6ynthesized due to the unlque nature o~ the serine codon degeneracy.

Two partially overlapping 5~-~er oligonucleotides, based on the bovine BMP exon 3 6equ~nce, wer~
~ynthesized ~or ~creening the cal~ liver cDNA library tProbe I) and are shown in Fig. 2a (underlined3.
Four oliqonucleotide probes were syn~hesized to identify exons 1, 5, 6, and 7 of the ~uman B~P gene (see Table 1~. Probe BMP103 ~5' CCAGTGGT~CATTCCAAGGA
CAAATATCACCAACATCTTCATCGCCATCTTCTCC~T 3') is a 57-mer complementary to bovine BMP exon 1 coding seguences.
Probe BMP104 (5' TCACTCAAAGCCAGGG m ACTCTTGCTCTGGGCC
ACGGGTTCGAGTACCTTCTATT 3 t ~ is also a 57-m~r complementary to bovin~ BMP exon 7 coding sequences.
BMP 116 (5' GAGACCAAATAGATAATTG m CTCCATT~ATGAGATCC
3'~ is a 39-~er complementa~y to human BMP ~xon 5 coding sequences. BMP 117 (5' CAAGTGACCGATCATAAAA
TTGTTCACTTATGGACTCGTCTGARATGAGA 3'~ is a 51-mer complementary to human EMP exon 6 coding sequences.
A 2~-mer oligonucleotide (5' TAGTCCCTCTGGAAGGCACATGTAGC 3') co~pl~en~a~y to human BMP exon 3 coding sequences (Probe J) was synthesized for screening the human BNP/PCR cDNA library.

EXAMPLE
~o~struction and Sc~Feenin~ o~ cDNA Librar~

~onstru~tion o th2 c~A lib~3~iÇ~. (a) Cal~ liver cDNA library. First ~trand cDNA wa~ synthesized ~ro~
calf liver poly(A)~ RNA u~ing conditions ~imil~r to Okayama and Berg tO~ayama~ H. and Berg, P. ~Qlec._and Cell Biol. 3:4094-4098 (1983)]. About 5 ~g oP
poly(A)~ RNA in 20 ~1 5 mM Tri~-hydrochlorid~ (pH

2 0 2, ~ 7 ,~ ~

-3~
7.5) was heat~d to 65-C Por 3 ~in., then quick cooled on wet ice and immediately adju~ted (at room temperature) to contain 50 mM Tris-hydrochloride (pH
8.3 at ~2~c), 8 m~ MgC12, 30 m~ RCl, 10 m~
dithiothreitol, 2 mM each of dA~P, dGTP, dTTP and ~-32~dCTP(-300cp~fpmol), 60 ~ RNasin, and 2.5 ~g of oligo (dT)12-18 (t~al volu~e 4~ ~1). Th~ reaction was initiated by the addition o~ 50-60 U of cloned moloney ~urine leuXemia virus reverse tran~criptase and continued ~or 60 ~in. a~ 4~-C. The seeond cDNA
strand was synthesized by the m~thod of Gubler and Hoffman tGubler, U. and B~f~r,~B.J. Gene ~5:263-269 tl983)] as modified by Axuf~o and Seed [Aru~fo, A.
and Seed, B. Proc. Natl. Acad. SCi.: US~ 74:8573 8577 (1987)]. The ds cDNA was then ligated to asymmetrically (hemi) phosphorylated EcoRI adapters tsee oligonucleotide ~ynthesis) as described by Aruffo and Seed, supr~, phosphoryla~ed with T~
polynucleotide kinase ~aniatis, et ~1., supra3, adjusted to 0.5 ~ NaCl~ 25 mM EDT~ and heated at 75-C
for 15 ~in. to inactivate ~he pDlynucleotide kinase.
The ds cDNA was separated from ~ligated adapt~rs by chromatography on Biogel A-15 3~ and recovered by ethanol precipitation. DNA wa~ ligated to EcoRI-cut lambda ZAPII (Stratage~e) with ~4 DNA ligase ~New England Biolabs) as descr~bed by supplier, ~ut included 15% polyethylene glycol ~PEG) 8000 ~Sig~a), a modifica~ion described by ~heif~er and Zimmer~an tPheiffer, B.H. and Zi~mer~an, S.B. ~ucl. Acids. ~
11:7853-787~ (1983)]. The liq~ted DN~ was recovered by centrifugation (12,000 xg), ~ashed w~th c~loroform, dried, resuspend~d ~ 4 pl wat~r and incubated with an in ~ Q packaging extr~ct (Stratagene) according to ~upplier. Reco~binant phage was propagated in E. Ç~l~ XLI-Blue ~Stratagene).

2 ~ 2 ~ ~ h ?~

(b~ ~MP primed P~R amplified ~BMP/PCR~ human kidney cDNA library. The PCR xeactions were perfor~ed as des~ribed ~y ~he suppliers of the PC~
Xit (PerkiR/El~er/Cetu~). Two 6ynthetic 47-mer oli~onucleotide pri~ers whose equences wer~ derived ~ro~ exon 1 (~ense pri~er) and 7 ~antis~nse pri~er) of ~he hu~an R~P gene and contained restrict~on ~ite adapt~rs ~ui~able for cloning wer~ used ~t ~ final concen~ration of 1 ~ eachO The PCR pri~er~ ~lanX
the complete cod~ng region of hBMP ~RNA. The template cDNA ~as synthesized ~ro~ 2.5 ~g of embryonic human kidney cell line 293 poly(A~ RN~.
T~e cond~tions of cDNA synthesi~ were identical to the above tPart A) except that the reaction volume was 20 ~1. T~e cDNA was fractionated on Biogel A-15m, also as above, recovered by ethanol precipitation and resuspended in lG0 ~1 of sterile water. 1-10 ~1 of cDNA template were used for each PCR reaction. 40 cyclec o~ PCR were performed in a Perkin/El~er/Cetus DNA the~mal ~ycler. Each cycl~
consist~d of a g4'C, 1 min. denaturation step, a 55-C, 2 ~in. annealing step and a 72-C, 3 min.
extension step. The extension step in cycle 40 was 15 ~in. instead of 3 min. Samples were extracted once wit~ phenol/ chloroform/IAA (1:1:0.04) once with chlorofor~/IAA ~24:1), recovered by ethanol precipitation, di~ested with EcoR~ and fractionated by electrophore~is on a 7~ acryla~ide, lx TB~ gel.
DNA migrating b~tween 400-800 b.p. was exci~ed fro~
the gel, purified by passage over an Elutip-d CO1UmnD
ligated to Eco-RI cut la~bda ZAPII, pacXaged and propagated as above (Part A).

Screening o~ the_l~brar~L~. (a) Bovine genomic library. Approximately 106 recombin~nt phage from a bovine geno~c libra~y ~Clontech) w~re plated (20,000 phage/137 ~m d~a plate) in ~. ç~l~ LE3S2, and grown -3fi-for 5-6 hour~ ~t 37^C. The phage were transferred onto nitrocellulose filters (Millipore, HATF13~) processed according to ~enton and Davis ~8) and screened with probe A ~Fi~. 1). Th~ probes were end-labeled with T4 polynucleot~de klnase and (s32_p~ ATPl) to a specific activity of 1~2 x 108 cmp/~g. The ~ilters were prehybridized ~or 1~2 h at 37-C ln 5x SSC (lx SSC ~ 0.15 N sodium chloride/0.015 ~ ~d~um citrate, pH 7~, 5x Denhardt~ solution (lx ~enhardt's solution - 0.02% polyYinylpyrrol~done/0.02%
Ficoll/0.02S ~ovine serum albu~in~, 10~ dextran sulfate, 5~ mM eodiu~ pho~phate pH 6.8, 1 ~M sodium pyrophosphate, 0.1~ NaDodS04 and 50 ~g/~l denatured . salmon sperm DNA. Labeled probe was added to a concentration of 106 cp~/ml and hybridization was continued overnight at 37~C witb ~entle shaking. The filters were washed in 2xSSC, O.lS NaDodSO~ at 55-C, and exposed to Kodak XAR-2 fil~ with a DuPont Lightning Plus intensifying screen overnight at -80-C. After develop~ent, the proke was remoYed from the filters by washing in O.lxSSC, 0.1% NaDodSO~ at 65-C. One set of filters was then hybridized with prob~ B ~Fig. 1) and washed and exposed a~ above.
Area~ of plaques giving signal6 with probes A and B
were picked, replated, transferred onto nitrocellulose in quadruplicate, ampli~ied according to Woo (21) and screened with probes A D (Fig. 1) (one pro~e per filter). Filters were washed and exposed to ~ilm as above. A plaque giving signals with three (A-C) of the four probes wa~ puri~ied ~y ~n additional round o~ plating and screening.
(b) Calf liver cDNA l~brary. Approximately 192,000 recombinant phage were plated ~16,000 phage/137 mm dia plate) ~n ~0 ÇQli XLl-~lu~, processed as above and ~creened with ~ i probe I
(Fig. 2a). The probe was labeled with DN~ polymerase I (Klenow fragment~ and (~32-P)-d CTP (9) to ~

2 ~ ~ ~J r~

~37;
specific activity o 2 x 109 cpm/~g. The ~ilters were scr~ened as above (a) but with the ~ollowing changes. ~1) The hybridization olution contained 40% formamide and ~2) th~ Pilters were washed in 2xSSC, 0.1% NaDodS04 at 65C7 (c) Hu~an gen~ic library. ~pproximat~ly 106 recombinant phage ~ro~ ~ human y~nomic library (Stratagene) were plated (50,000 phage/plate3 in ~.
oli LE392, processed and screened with the ~Qbp BglII insert of bB~P cDNA#l (F~g. 3). The probe was labeled (9) and the filters were hybridized as above ~ee calf liver cDNA library). The filters were washed in 2xSSC, 0.1% NaDodS0~ at 60-C. Rositive plaques were purified by replating and rescreening.
(d) (BNP/PCR) human kidney cDNA library.
Approximately 1000 reco~binant pha~e were plated (500 mm dia plate) in ~. Goli. XLl B1Ue~ processed, hybridized with probe J (Fig. 4~, underlined) and washed as in (a). Positive plasyes were purified by replating and rescreening.

~ ~ P~E 4 Plasmid and ~haqe ~NA---x-s-QLl~ion~
Subclonina! Sequençin~ and analysi~

Plasmid DNA was isolated by the alkaline lysis method ~Maniatis, e~ al., supr~) and la~bda DN~ was isolated by a phage miniprep procedure described by Jones and Rawls, tJones, X.W. and Rawls, 3.~ ~ene~ Q 733-742 (1988)~.

Bluescript S~(-3 pla~mids containi~g BMP cDN~ ~re relea~ed fro~ la~bda ZAP by the ~13 rescue/exci~ion protocol described by the ~uppli~r (Stratagene). B~P
gene fragments were released fro~ the EMB~3 lambd~
vector by SalI or other appropriate restriction enzyme digestions (~ee Fig. 2 and ~able 1). ~MP cDNA

2 ~ 2 ~ ~ ~

inserts were excised from the Bl~escript SK~-~ vector by a BglII (bBMP cD~A) or a Bam~I/H~ndIII (hB~P cDNA~
digestion. The DNA frag~en~s were purified by polyacrylamid~ or agarose gel electrophoresls (Maniatis, t ~~ Y~) and pa~sage over an Elutip-d column (Schleicher and Schuell) and were then subcloned into pUC 19 or ~13 ~equencing vector~
tYamisch-Perron, C" Vieira, J. and ~essing, J~, Gene ~3:103-119 (1985)]. DNA seguenc~ng was performed by t~e dideoxy chain termination method (5angerd F., Nicklen, S. and Coulson, A.R.~ P~oc! Natl. ~cad. Sc~.
USA 74:5463 67 ~1977)~ using ~13 pri~ers as well as specific internal prlmers. Ambiguous regions were resol~ed using 7-deaza-2-deoxyguanidine-triphosphate ~Barr, P.J., Thayer, R~M., L2ybourn, P., Najarian, R.C., Seela, F., and Tolan, D., ~iotechniques 4:428-32 (1986)] and sequenase (U.S.
Biochemicals).

Northern blot analys~. Poly(A)+ RNA was fractionated on a 1.4% agarose gel in the presence o~
formaldehyde (Lehrach, H., Diamond, D., Wo~ney, J.~.
and Boedt~er, H., Bi~che~is~ry 16:4743 51 tl977~] and directly transferred to nitrocellulose according to Thomas (Tho~as, P., Proc. ~atl. ~cad. Sci. ~S~
77:5~01-5 (1980)].

Filters were hybridized with prob~ I or bBMP cDNA ~1 and washed as previously described above in 6creening of the librarles, section ~.

Southern blot analysi6. Genomic blots$ 10 pg o~
human, bov~ne and mouse genomic DNA (Clontech) wa~
digested with EcoRI, ~ractionated on ~ 0.7% agarose gel and transferre~ to nitrocellulose (Maniatis, al., supra). ~ybridizaklon and washing were identical to those described under ~Northern blot ~2~2~

~39 ; analysi~.~ Clone and PC~ blot8: DNA from genomic clones, cDN~ clones or P~R reactions wer~ digest~d with various restriction enz~mes, ~ractionated o~ 1%
agarose gels and transferred to nitrocellulose (Maniatis, ~ nl-, ~Y~) E~
Expression o~ ~ovln~ and Human BMP in Yeas~

T~e PCR pri~ers described ~n detail above were used to generate DNA seguenc~s wit~ Xba-l and Sall restricti~ sites for direct in-frame clon~ng into pAB125, a vector containing the o-fackor leader ~described in EP0 P~b. No. 0 116 201, published 22 August 1984) sequence fused to the ADH2/GAPDH
promoter ~described in a oop~ndinq U.S. Patent Application Ser. No. 760,197, filed 29 July 1985).
The resulting promoter/leader/gene cassettes were excised with BamHl and Sal-l and cloned into the yeast expression vector pBS24.1. The plas~id pBS24.1 is a derivative a pBS24 which is described in copending ~.S. Patent AppliG~tion SerO No. 138,894, filed 24 December 1~87. The BamHI-SalI vector fragment of pBS24 was ligated to h 65 base pair human basic FGF BamHI-SalI ~ragment to create pBS2~
This 65 base pair fragment will b~ replaced by the BamHI-SalI promoter/leader/B~P cassettes which are ligated in~o pBS24.1 The resulting plas~id~
pBS24b~MP and pBS24h~ were u~ed to transforD~ yeas~c strain ABllO w~th genotype 2~ata, ura3-5~ u2-04 c~r both leu2-3 and leu2-1~2, pep4-3, ~i~4-580, c~r-, 3 o under leuoine celection media, ~uch as th~ ~;ynthetio complete ~nedia without leucine in ~thods ~n Yeas~
Genet~cs, Cold Spring Harbor La~orator$e~, 1986, ~.
Sherman, G.R. Fink and J.B. ~Iicks. For inducti~n of expression, cell~ were grown Por 48h the uracil 35 deficient ~edia below:

20 g Casamino Acids 5 g A~moni~L~ Sulfate ~ g Potassium Pho~phate 0. 5~ ~Sagnesiu~D Sulfat~
O. lg Sodium Chl~ride 0. lg Calcium Chloride 0. 04~D1 Trace Element~ M~xture 41 1 2% S~dium PSolybdate 0.015g Vitamin ~ix 0 ~ 03 S~ Pan~ot~ena~e 0. 03 g ~nositol 70 ml 50~ Glucose q. ~. ~o a ~OOml an~ pH to ~ O
ll.S to ~000 ~1 15 Trace Elements Plixture 5 g B~ric Acid 0. 4g Cupri~ Sulfate g Pc~tassium Iodide 2 g Ferric-Chloride 20 4 g Manganese Sulate 2 g Sodiu~ Molybdate 4 g Z inc Sul fate q. s. to 1OOOID1 and Rterili2e by filtratioal.

Vitamin Mix 3 g Myo-Inositol 3 g ThiaD~ine 3 g Pyridoxine 3 g Calcium Paratotherlate 0. 2g Blotin 2 g p-A~inobenzoic: Acid 2 g R~bo~ rin 0. 2g ~olic acid 3 g ~iacill , ~
2 ~ 2 ~ ~ 2 r9 Yeast cells were re~oved rom the culture media by cen~rifugation and the proteins ~n the supern~tants analyzed after precipitation with 10~ trichloroacetic acid/deoxycholate ~0.4 mg/~l). ~he pellets were wa~hed with a~etone and loaded onto 15~ SDS-polyacrylamide gel6 t~gether ~i~h appropriate control~. Proteins were vi~ualized by Coomass~e ~lue staininy. A samples of S. cerev~sl~e strain ABllO
containing pBS24.lbRMP or pBS24.lhBMP were deposited w$t~ the ATC~ on 1 June 1989 under Accession ~os.
~0949 and 20950.

Recombinant Bovine BMP Purificatio~

Medium was removed f rom yeast cells by centrifugation, and concentrated approximately ten fold. The pH ~as adjusted to 4.5 and the concentrate diluted to a conductivity below 5 ~S/cm. This was applied to Fast Flow S ion-exchange re~in (Pharmacia) pre-equilibrated with 50 m~ sodium acetate, 1 m~
EDTA, 1 mM PNSF lpH 4.5~. The column was washed with one volume of the above buffer and eluted using a 0-1 M sodiu~ chloride gradient in the above bufer. The l9XD and 16XD proteins eluted at a conductivity of 5-25 mS/c~, as confirmed by SDS-P~GE. The fractions containing predominantly the l9~D prote~n ~ere pooled, adjusted to pH ~.5 and 4.5 M wi~h respect to urea, concentrated, and applied to an S-100 ~izing column in 4M urea, 100 mM Tri~ l, 1 ~M EDTA, 1 mM
PMSF (p~ 7.5). Fractions containing l9RD ~nd 16RD
proteins were identi~ied by SDS-P~GE, pooled separately, concentrat~d and dialyzed against ~ater~
prior to lyophilization.

, . , ~

2~2~f~

~2~
ASSA~ EQR ~M~-~5r~ y Samples of puri~ied recombinant ~P were added to CM~-1066 (GIBCO) culture mediu~D oontaining ~etal rat 5 midbelly triceps brachii DUSC~ ragments.
Connective tissue outgrowt~s were cultured on a substratu~ o~ BMP-free matrix ~ro~ the diaphyses o~
long bones o~ adult male Sprague-Dawley rats. This assay is detalled more ~ully by Kawa~ura and Urist~
10 Developmental Biology, ~Q, 435-4~.2 (1982). The inductive activity is measured by ~3~]thymidine incorporation into DNA, [35S~ulfate incorporation into glycosaminoglycans and conflrmation o~
chondrogenesis (cartilage formation) by histolo~y.
Recombinant BMP was tested at concentrations of 200 ng/ml to 5 g/ml in the 2 ml culture syste~. The same level o~ po~itive results w~s observed with recombinant ~MP as was previously noted with native 8~P at equivalent concentrations. Higher [3H]thymidine and [35S~sulfate incorporation and new cartilage and cbondrocytes were observed in the recombinant BMP induced cultures. No cartilage or chondrocytes were seen in control cultures without recombinant B~P~

EXAMP~E 8 Isolation and Anal~sis o~ ~he BMP ~ ~ngL~n~_c~
I. The Bovine BMP qene.
Probes ~ and ~ (Fig. 1) were ueed to isolate 1 strongly and 7 weakly hybridizing clones fro~ 106 recombinants of a bovine genom~c library. ~he one strongly hybridi~ing clone (29-bg-3a was ~ubjected to a second round of creening with probe~ A~D SFig. 1).
Three of the probes (A-C) hybridlzed to 29 ~g-3 Southern blot analysis o~ puri2ied 29-bg-3 DNA
localized probes A and B to a l.9kb EcoRI ~ragment e.

~~30 and probe C ~o ~ 1.2kb Eco~I-SalI fragment. Bo~h fra~ments were equenced and shown to contain regions, in the top reading frame, encoding their respective tryptic peptides (Fig. 2; boxed amino S acids). Interestingly, tryptic peptide D was also present but was spl~t between two exons, thus explaining the lack of hybridization with probe D.
Intron-Exon boundaries ccnform to ~he G~-AG rule and are denoted by vert~cal llne~. Addit~onal bov~ne BMP
exons wer~ not ~solated. Attempt were directed at isolating a bovine ~MP cDNA clone using as a probe exon sequences deriv~d ~rom the bovine ~P gene.

II. The Bovine B~P cD~A
Northern blot analysis o~ poly(A)+ RNA from se~eral bovine tissues using probe I revealed calf liver as a good source of bovine BMP mRNA tdata not ~ho~n).
Probe I was then used to isolate putative bovine BMP
cDNA clones ~rom a calf liver cDNA library. Two clones (#l and ~7) were sequenced, and shown to contain identical overlappin~ seguences as well as the expected encoded tryptic peptides (A-D) ~Fig. 3).
The additicnal BMP tryptic peptides (E-~, Fig. 1) were also found encoded in the bovine ~P cDNA.

III. The Human B~P ~ene Northern blot analysis of poly(A)+ RNA fro~ sev~ral human tissues and cell lines, including l~ver, ~idney and osteosarcoma, (using bovine B~P cD~A tl as a probe) failed to detect human BMP. However, a Southern blot, performed under identical hybridization and wash condit~ons detected hu~an W
gene fragmente (data not ~hown~.

Based on the Northern ~nd Southern re.sult~, th~
following ~trategy was adopted to (1~ Glone and ~equence the human BNP gene, then (2) identify a 2~$~

tissue source of hu~an BMP mRNA by PCR amplificat~on (using human BMP based primer$: BMP/PCR cDNA) and S~uthern blot a~aly~is of the products 3) clo~e and isolate the PCR generated h~an B~P cDNA.

Bovine ~MP cDNA #1 was used as a probe ~o isolate 7 s~rongly and 5 wea~ly hybridizing clones from 106 recomb~nant~ of ~ human gen~ic llbrary. Southern blot analysis of purified DNA fro~ 11 of the 12 ~lones identi~ied two hybridizing HindIII fragments (1.7 kb and 2.0 Xb) common to ~hree clones ~G4t 5 and 9. The 1.7 kb and 2.Q kb HindIII fragments fro~ HG9 were sequenced and shown to ha~ a 63.4% and 62.2%
amino acid h~mology to the ~ovine BMP exons shown on Figur~ 2A and ~B respectiv~ly. Table 1 summarizes these results as we~l as the So~thern blot results for the remaining human BMP exons. Th~ seguence of the exon-containing region oP each subclone is shown in Figure 4a-c. The 5' end of exon 1 (Cap site) and the 3' end of exon 7 (poly(A)+ addition site) are unXnown. The intron/exon boun~aries follow ~he GT-AG
rule ~nd are denoted by vertic31 lines. The DNA was translated in all three reading frame~. The ~iddle and top frames contain the correct amino acids for exons 1-6 and exon 7, respecti~ely (box~d~.

2S IY. The Human BMP cDNA
Xidney tcell line, 293) wa~ identif~ed as a ~ource o~
hBMP mRNA by PCR ampli~icat~on ~nd Southern blot analy~is. 103 recombinant~ o~ ~ cDNA library ~ade from the hBNP/PCR cDNA, wer~ screened wit~ probe ~
~human exon 3 probe) and 12 putatlve hBM~ cDNA clones were isolated. Agaro e gel electrophore~ ollowing a BamHI/HindIII digestion o~ t~e clones ~howed ~hat each clone contained either ~ 700bp or a 600bp cDNA
~nsert. DNA ~equencing revealed that the 700bp species was the expected full length human BmP cDNA

2~2~2~

(Fig. 5~ while the 600bp ~pecie~ was a ~runca~ed human BMP cDNA, missing exon 2 (not ~hown~. This ~horter cDNA species most likely represents an aberrantly ~pliced, non-functional mRNA ~inc~ a translat~onal frameshift occurs at the newly ormed exon l/exon 3 ~unction re~ulting in a prema~ure termination codon.

~ P~ 9 Analysis o ~uman and ~ov~e ~MPs ~x~ressed in Yeast hB~P is expressed in yeast as a mixture of thr~e approximately 17XD proteins, whereas bBMP is expressed as a ~ixture of 3 species with ~olecular weights of l9KD and a 16KD doublet. The ~ize heterogeneity is most probably a result of processing of each BMP by yea~t ~ncoded enzyme(s) during secretion. Amino ter~inal amino acid sequencing of the bBMP ~ixture gave a single amino terminus (Ph~)~
indicating that the proce~sing occurs in t~e carboxyl regi~n o~ the protein, yielding tbe 16RD yeast cleavage analogs. Several paired basic amino acid residues ~n thi~ region are likely candidates as proteolysis ~ites during ~ecretion. Thi~
observation, together with the fact that each ~MP
cDNA encodes a putative protein product of ~olecular weight greater than 20kD, also underscores th~
possibility tha~ BMP, 1~ YiYQ~ i8 the product o proteolytic processing fro~ a larger precursor.

~ .' 2~2~9 ~6--A Summary of the Southern Blot Results Identifying Human BMP Exon-Containing DNA Pro~ Genomic Clones.

Restri~tion_Fr~e~tl Geno~ic s ~Qa Sl~e (~ ~n~Y~ ç~Qn~2 ~E~k~
1 0.6 Pst I Hg 43 ~MP 103 2 0.6 P~t I ~G 43 hBMP cDNA
3 1.7 Rind III HG 9 bBMP cDNA
4 2.0 ~l~d III HG 9 bBMP cDNA
0.4 Bgl II ~G 5 BMP 116 6 3.0 Bgl II HG 53 B~ 117 7 1.1 ~gl II HG 5 ~P 104 1 Exon-containing DNA restriction fragment.
2 Geno~ic clone fr~m which restriction fragment was derived.
3 Exon 1 and 2 DNA's were obtained fro~ a ~ kb Sal I fragmen~ o~ BG 4 ~ubcloned into pUC 19, Both exons ~ere found on the ~a~e Pst I fragment, but identi~ied with different probes. ~xon 6 was obtained from a 15Xb Sal I insert o~ HG5 subcloned into pUCl9.
4 Pxobe used to identi~y exon-containin~ DNA
restriction fragment. Probes are d~scribed in ~Oligonucleotide Synthesis~ section o ~ater~als and Methods. ~ybridization and washi~g conditions are described un~er Exampl~ 3 ~n the ~Screening of the libraries ~c) hu~an geno~ic l~brary~ section o~ the examples.

Claims (26)

1. A composition comprising a polypeptide selected from the group consisting of mammalian bone morphogenetic protein (BMP) and analogs thereof substantially free of other osteoinductive associated factors.

2. A composition according to Claim 1, wherein said mammalian bone morphogenetic protein is human BMP.

3. A composition according to Claim 1, wherein said mammalian bone morphogenetic protein is bovine BMP.

4. A composition according to Claim 1 wherein said polypeptide comprises an amino acid sequence aa1-aa182 as shown in FIG. 5 or a fragment thereof,

5. A composition according to Claim 1 wherein said polypeptide comprises an amino acid sequence aa1-aa180 as shown in FIG. 3 or a fragment thereof.

6. A composition according to Claim 5 wherein said polypeptide comprises of a 16KD yeast cleavage analog.

7. An intron-free DNA sequence or its complement encoding a polypeptide comprising an amino acid sequence of mammalian BMP and analogs thereof.

8. A DNA according to claim 7, wherein said DNA
encodes human BMP.

9. A DNA according to claim 7, wherein said DNA
encodes bovine BMP.

10. A DNA sequence according to claim 7, wherein said polypeptide comprises the mammalian BMP signal sequence.

11. A DNA sequence according to claim 7 wherein said polypeptide does not comprise the mammalian BMP
signal sequence.

12. A DNA sequence according to claim 7 wherein said polypeptide further comprises an N-terminal yeast alpha-factor signal sequence that provides for secretion in a yeast host.

13. A replicon comprising a DNA sequence according to claim 8.

14. A vector comprising a DNA sequence according to claim 8.

15. A cell comprising a replicon according to claim 13.

16. A cell comprising a vector according to claim 14.

17. A method of producing recombinant mammalian BMP
or an analog thereof comprising:
(a) providing a population of cells comprising a heterologous DNA sequence, wherein said DNA
sequence comprises (i) transcriptional and translational control sequences functional in said cells, and (ii) a coding sequence under the control of said transcriptional and translational sequences that encodes a polypeptide comprising mammalian BMP
and analogs thereof; and (b) growing said population of cells under conditions whereby said polypeptide is expressed.

18. The method of claim 17 wherein said polypeptide is human BMP.

19. The method of claim 17, wherein said polypeptide is bovine BMP.

20. The method of claim 17, wherein said population of cells are microorganisms.

21. The method of claim 17, wherein said population of cells are yeast.

22. The method of claim 17, wherein said population of cell are E. coli.

23. The method of claim 18, wherein said polypeptide further comprises sequence encodes a signal sequence at the N-terminus of said polypeptide, and said polypeptide is secreted from the said cells.

24. The method of claim 23 wherein said cells are yeast cell and said signal sequence is a yeast alpha-factor signal sequence.

25. A method for inducing bone growth comprising the step of contacting bone tissue with an effective amount of a composition of mammalian BMP or an active analog thereof.

26. A monoclonal antibody to a mammalian bone morphogenetic protein.