Note: Descriptions are shown in the official language in which they were submitted.
2~7~5~
BONE CALCIFICATION FArTOR
This invention relates to a cla~s of mature native
mammalian proteins which initiates calcification and
which is named herein as bone calc~fication factor
(BCF). Representative of this class are human and
bovine BCF, for which the full lenqth coding
seguences are provided herein. The BCF is provided
by isolation from bone sources and by synthesis using
recombinant DNA technigues.
BACKGROUND OF THE INVENTION
It is known that demi~eralized bone matrix induces
new bone formation when implanted in the soft ti6sue
by a process generally designated as matrix induced
bone formation ~see Urist, M.R., Science, 150: 893-
899 (1965)). There have been nu~erous efforts to
extract and identify the active material (or
materials) which induces this procQss, and it has
been generally referred to in the l$terature ag bone
morphogenetic protein~s) (BMP). It $~ uncertain
whether BMP iB a single material or a mixture o~
materials, and there does not ~ppear to be agreement
among the investigators a8 to which ~ateri~l, iP any,
is the bone morphogenetic protein.
The therapeutic use of BMP offers considerable
advantages over use of traditional ~one graft
'~17~g
materials. While not intended to be limited by any
theory, one hypothesis assumes that BMP transforms
tissue cells into o~teoblasts (cells that manufacture
bone). During a process that replicates normal human
fetal development, ~MP-induced osteoblasts form
cartilage which, over a period of several months,
evolve into solid ~one. Ihus ~MP may be useful for
replacing bone that has ~een destroyed ~y disease or
accident, for use in treatment of scoliosis victims,
for treatment of mal- or mis-formed bone, for use in
healing of a fracture, etc.
It is thus an abject of the present invention to
produce a functional bone calcification factor or a
component thereof, which i8 a 22 XD protein
identified by its entire amino acid sequence, which
initiates calcification.
It is another object of the present invention to
produce this biologically active 22 XD protein by
recombinant DNA technology.
It is yet another object of the present invention to
construct nucleic ac~d screening probes for isolation
of the gene comprising the 22 XD BCF.
~t is yet another object of the present invention to
provide an amino acid seguence of mature 22 KD BCF
which can be thus prepared by direct biochemical
synthesis or from constituent amino acids by peptide
synthesis, for example as by the Merrifield method,
and particularly ~q ~e of automated peptide
synthesis technDlogy.
These and other ob~ects of the invention will be
apparent from the following description of the
2017~
preferred embodiments and from practice of the
invention.
~UMMARY OF THE INVE~TION
The present invention provides a class of mature
native mammalian proteins (the class is termed herein
as BCF), represented by native human and bovine BCF
described herein, which initiate calcification ~n
~ivo which is important for formation of bone. The
human and/or bovine BCF can be used to identify and
isolate other mammalian BCF proteins which may or may
not be homologous (in their nucleotide and amino acid
sequences) to human or bovine BCF and which exhibit
the BCF biological activity. It is recognized that
there may be al.elic variations in BCF within a
species, and such allelic variants are also within
the scope of the class of proteins provided by the
present invention.
The present invention further provides polypeptides
which are analogs of BCF, such as BCF muteins, fusion
proteins, comprising BCF or BCF domains, and BCF
fragments. The term fusion protein includes a
protein comprising ~ complete 8CF seguence or a BCF
domain, and a heterologous N- or C-terminal seguence
(such as a signal sequence or seguence which protects
the protein from degradation). A BCF mutein is a
protein substantially homologous to a nat$ve BCF
6eguence (e.~., a minimum of about 75%, 85%, 90% or
95% homologous) wherein at least on~ amino acid i~
different. A BCF fragment or domain i8 an amino acid
seguence of sufficient length from a BCF protein such
that it is identif~able a8 having been derived from
such BCF protein. The origin of a particular peptide
can be determined, for example, by comparing its
sequence to those in public databases.
'
-4-
The present invention further provides a 22 KD bone
calcification factor having the human and bovine
amino acid sequences shown in FIG. 1. The present
invention also provides methods of preparing the 22
XD bone calcification factor (BCF) by recombinant DNA
technigues.
~he present invention pr~vides ~he DNA 6equence
encoding BCF, which may be used to construct vectors
for expression in host systems hy recombinant DNA
techniques.
The present invention also prov$des therapeutic
compositions comprising BCF and matrix Gla protein
(MGP) for initiating calcification and ~ethods for
inducing calcification in vertebrates by introducing
in vivo at the desired ~ite an effective
calcifica~ion initiating amount of BCF and MGP. The
identity of MGP was first reported by Price, Urist
and Otawara in ~iochem. Biophys. Res. Comm. 117:765-
~71 ~1983).
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 depicts the DNA seguences and encoded amino
acid sequences of human and bovine BCF including
their ~ignal sequences;
FIG. lA depicts the ~mino acid sequence of human BCF
(hBCF) without its signal peptide,
FIG. lB depicts the amino acid sequence of bovine
BCF (bBCF) without its 6igDal peptide:
FIG. lC depicts the DNA sequence encod$ng hBCF
without its signal sequence;
FIG. lD depicts the DNA seguence encoding bBCF
without its signal ~eguence;
FIG. 2 illustrates t~e seguence of human BCF tryptic
fragment no. 41, a t~o-fold degenerate 45-mer
oligonucleotide probe (probe A), and a ~econd probe
B, designed therefrom, consisting of 64 18-mers which
are complementary to all possible codons shown.
FIG. 3 illustrates probe C (derived from clone 0st 3-
7) and four BCF cDNA clones isolated from a bovine
cDNA library (bbl.1-7) and two human osteosarcoma
cDNA libraries (Ost 1-7, 0st 3-7 and 0st 3-17). The
length, coding region (boxed) and partial restriction
map of the clones is included. The NcoI and SpeI
~ites (~) are only present in the human BCF
sequences.
FIG. 4 illustrates oligonucleotide adapters (boxed)
used to prepare BCF expression vectors which are
secreted from yeast using the alpha-factor signal
peptide.
FIG. 4A illustrates the junction between the ~CF-
encoding DNA and promoter in an expression vector
used to express unsecreted BCF in yeast.
FIGS. 5, 6 ~nd 7 are photomicrographs of the
quadriceps pouches of ~ice 21 days after implantation
of a composite of recombinant hBCF and MGP, showing
initiation of calcification.
DETAILED DESCRIPTI~N OF THE INVE~IIQN
The BCF according tv the present ~nvention may be
obtained, free of other osteoinductive associated
factors, directly ~rom bone sources, by preparative
201r~L6~
peptide synthesis using chemical methods (such as the
Merrifield synthesis method) or by recombinant DNA
technology.
As more particularly described in Example 1, BCF may
be obtained ~y purification from human, bovine, or
other vertebrate bone from partially purified
extracts ~.g., U.S. Patent 4,795,804 and references
cited therein) by preparative gel electrophoresi6 and
electroelution of the 22 X protein.
BCF may also be obtained by recombinant DNA methods,
such as by screening reverse transcripts of mRNA, or
by screening genomic libraries from any cell. T~e
DNA may also be obtained by simply synthesizing the
DNA using commonly available techniques and DNA
synthesizing apparatus. Synthe~is may be
advantageous because unique restriction sites may be
introduced at the time of preparing the DNA, thereby
facilitating the use of the gene in vectors
containing restriction sites not otherwise present in
the native source. Furthermore, any desired site
modification in the DNA may be introduced by
synthesis, without the need to further modify the DNA
by mutagenesis.
In general, DNA encoding BCF may be obtained fro~
human, bovine or other sources by constructing a cDNA
library froo mRNA lgolated from bones of the
vertebrate; and screening with labeled DNA probes
encoding portions of the human or chains in order to
detect clones in the cD~A library that contain
homologou~ ~eguences; or ~y polymerase chain reaction
(PCR) amplification of the cDNA ~fro~ ~RNA) and
6ubcloning and screen~ng with l~beled DNA probe3; ~nd
then analyzing the clones by restriction enzyme
analysis and nucleic acid ~equencing so as to
.
20~ 7~66
~7-
identify full-length clones and, if full-length
clones are not present in the library, recovering
appropriate fragments from the various clones and
ligating them at restriction ~ites common to the
clones to assemble a clone encoding a full-length
molecule. Particularly preferred DNA probes are set
forth in the accompanying example~. Any sequences
missing from the library may be obtained by the 3'
extension on the complementary mRNA of synthetlc
oligodeoxynucleotides identified by screening cDNA ~n
the library ~so-called primer extension), or
homologous sequences may be 6up~1ied from known cDNAs
derived from human or bovine seguences as shown ~n
FIG. 1.
The practice of the present invention will employ,
unless otherwise indicated, conventional molecular
biology, microbiology, and recombinant DNA technigues
within the 6kill of the art. Such techniques are
explained fully in the literature. See e.g.,
Maniati~, Frit~ch ~ Sambrook, ~Molecular Cloning: A
~aboratory Manual~ (1982); ~DNA Cloning: A Practical
Approach," Volumes I and II (D.N. Glover ed. 1985);
"Oligonucleotide Synthesis~ (M.J Gait ed. 1984);
"Nucleic Acid Hybridization" (B.D. Rames ~
S.J. Higgins eds. 1985); ~Transcription And
Translation" (B.D. Hames ~ S.J. Higgins eds. 1984);
"Animal Cell Culture" (R.I. Freshney ed. 1986);
"Immobilized Cells And Enzymes~ (IRL Press, 1986);
B. Perbal, ~A Practical Gu~de ~o Molecular Cloning~
(1984).
In describing the pre6ent invention, the following
terminology will be u6ed in accordance with the
definitions 6et out below.
2~:~74~
--8--
The term "osteoinductive associated factors"
includes factors known in the art which are present
in mammalian bone or other ma~malian tissue and tend
to co-purify with BMP ~r ~XP activity. Such factors
include proteins which have been isolated from bone
having reported molecular weights of 34 RD, 24 XD,
18.5 KD, 17.5 KD, 17 XD, 16.5 RD, 14 RD (as cited in
U.S. Patent No. 4,761,471), and 6 RD (reported by
Price, P.A., et al., from ~NAS, 73, pp. 1447-1451,
1976).
A ~replicon" is any genetic element (~.~., plasmid,
chromoso~e, virus) ~ha~ f~nctions as an aut~nomous
unit of DNA replication in vivo; ~.e., capable of
replication under its own control.
A "vector" is a replicon, such as plasmid, phage or
cosmid, to which another DNA segment may be attached
so as to bring about the replication of th~ attached
segment.
A "double-stranded DNA molecule" refers to the
polymeric form of deoxyribonucleotides (adenine,
guanine, thymine, or cytosine) in its normal, double-
stranded helix. This term refers only to the primary
and secondary structure of the molecule, and does not
limit it to any particular tertiary forms. Thus,
this term includes double-stranded DNA found, inter
~li~, in linear DNA molecules (Q.g., restr~ction
fragments), viruses, plasmids, and chromosomes. In
d~scussing the structure of part~cular double-
stranded DNA molecules, ~equence~ ~ay be described
herein according to t~e ~ormal convention of g$ving
only the 6eguence in tn~ ~' to 3' dir~ction along the
nontranscribed strand of DNA (~.e., the strand having
a seguence h~mologous to the ~RNA).
A DNA "coding sequence" is that portion of a DNA
~equence, the transcript of which is translated into
a polypeptide ~n v vo when placed under the control
of appropriate regulatory ~equences. The
complementary DNA strand will be understood to be
that strand which is transcribed. The boundaries of
the coding seguence are determined by a start codon
at the 5' (amino) terminus and a translat$on stop
codon at the 3' (carboxy) terminus. A coding
seguence can include, but is not limited to,
procaryotic sequences, cDNA from eucaryotic mRNA,
genomic DNA seguences from eucaryotic (e.g.,
mammalian) DNA, and even synthetic DNA 6eguences. A
polyadenylation signal and transcription termination
seguence will usually be located 3' to the coding
seguence.
A "promoter sequence" is a DNA regulatory region
capable of binding RNA polymerase in a cell and
initiating transcription of a downstream (3'
direction) coding ~eguence. For purposes of defining
the present invention, the promoter seguence is
bounded at it~ 3' terminus by the translation ~tart
codon of a coding seguence and extend~ up~tream (S'
direction) to include the minimum number of base~ or
elements necessary to initiate transcription at
levels detectable above background. Within the
promoter sequence will be found a transcription
initiation site (conveniently defined by mapping with
nuclease Sl), as well as protein bind$ng domains
(consensus 6equences) responsible for the binding of
RNA polyme~se. Eucaryotic promoters will often, but
not alway~, contain ~TATA~ boxes and ~CAT~ boxes.
Procaryctic promoters contain Shine-Dalgarno
seguences $n addition to the -10 and -35 consensus
6eguences.
--10--
A coding sequence is "under the control" of the
promoter ~equence in a cell when RNA polymerase which
binds the promoter 6equence transcribes the coding
sequence into mRNA which is then in turn translated
S into the protein encoded by the coding sequence.
A cell has been "transformed" by exogenous DNA when
such exogenou~ DNA has been introduced into the cell
membrane. Exogenous DNA may or may not be integrated
~covalently linked) to chromosomal DNA making up the
genome of the cell. In procaryotes and yeast, for
example, the exogenous DNA may be maintained on an
episomal element such as a plasmia. ~ith respect to
eucaryotic cells, a stably transformed cell is one in
which the exogenous DNA has become integrated into a
chromosome so that it is inherited by daughter cells
through chromosome replication. This stability is
demonstrated by the ability of the eucaryotic cell to
establish cell lines or clones comprised of a
population of daughter cells contain~ng the exogenous
DNA. A "clone" is a population of cells derived from
a single cell or common ancestor by mitosis. A ~cell
line" is a clon~-of a primary cell that is capable of
stable growth in vitro for many generation6.
Two DNA sequences are "substantially homologous~ when
at least about 85S (preferably at least about 90%,
and most preferably at least about 95%) of the
nucleotides match over the defined length of the DNA
~equences. Sequences that are 6ubstant$ally
homologous can be identified in a Southern
hybridization experiment under, for ex~mple,
stringent conditions a~ defined for that particular
system. Defining appropriatQ hybridization
conditions i8 within the skill of the art. See,
e.g., Maniati~ et al., ~upra: DNA Cloning, Vols. I
II, supra; Nucleic Acid Hybridization, ~E~.
2~17~S
A "heterologous" reqion of the DNA construct is an
identifiable segment of DNA within a larger DNA
molecule that is not found in association with the
larqer molecule in nature. Thus, when the
heterologous region encodes a ~ammalian gene, the
gene will usually be flanked by DNA that does not
flank the mammalian genomic DNA in the qenome of the
source organism. Another example of a heterologous
coding sequence is a construct where the coding
sequence itself i~ not found in nature (e.g., a cDNA
where the genomic coding sequence contains introns,
or 6ynthetic 6equences having codons different than
the native gene). Allelic variations or naturally-
occurring mutational events do not give rise to a
heterologous region of DNA as defined herein.
A composition comprising ~A" (where "A" is a single
protein, DNA molecule, vector, etc.) is ~ubstantially
frQe of "B" (where "B" comprises one or more
contaminating proteins, DNA molecules, vectors, etc.)
when at least about 75% by weight of the proteins,
DNA, vectors (depending on the category of species to
which A and B belong) in the c~mposition i6 ~An.
Preferably, ~A" comprises at least about 90~ by
we$ght of the A+B species in the composition, most
preferably ~t least about 99% by weight. It i8 also
preferred that a composition, which i8 6ubstantially
free of contamination, contain only a singlQ
molecular weight ~pecies having the activity or
characteristic of the species ~f interest.
As more particularly described in the following
exa~ples, human bnd bovine cDNA l$brar~es were
initially probed for sequences encoding BCF sequences
using labeled oligodeoxynucleotides whose seguences
were based on a partial am~no a~id sequence
determined from analysis of purified protein samples
2~7~
derived from bone described herein. However, it is
realized that once being provided with non-
chromosomal DMA encoding human and bovine BCF and
their leades seguences as described ~erein, one of
ordinary 6kill in the art vould recognize that other
precisely hybridiz~ng probes may be prepared from the
descri~ed sequence6 i~ order to readily obtain the
remainder of ~he desired human or ~vine gene.
The non-chromosomal DNA provided by t~e pre ent
$nvention is novel, since it is ~elieved that the
naturally-occurring human and bovi~e ~enes
(chromosomal) contain in~ron6 (transcri~ed sequences,
the corresponding amino acids of which do not appear
in the mature protein). Hence, the term "non-
chromosomal n excludes the DNA sequence~ whichnaturally occur in the chromo~omes of human or bovine
cells. The present invention ~lso en~ompasses the
non-chromosomal cDNA sequenoes deri~a~le from the DNA
sequences disclo6ed herein.
Vectors are used to amplify the DNA which encodes the
chains, either in order to p¢epare quantities of DNA
for further processing (clon~ vectnrs) or for
expression of the chains te~pn~ssion vectors).
Vectors comprise plasmids, ~in~ses (~ncluding phage),
and integr~table DNA fragme~*~ ~--, fragments that
are integratable into the h~st genome ~y
recombination. Cloning vectors need ~t contain
expression control 6eguences. However, control
seguences in ~n expression vector include a
transcriptional pr~moter, ~n option~l Dperator
sequence to control tr~nscription, a ~equence
encoding suitabl~ rRNA ribosomal bind$ng sites (for
prokaryotic expression), Dnd seguence~ which control
termination of transcripkion and translation. The
expression yector ehould pre~erably include a
?
selection gene to facilitate the stable expression of
BCF and/or to identify transformants. However, the
selection qene for maintaining expression can be
6upplied by a 6eparate vector in cotransformation
systems using eukaryotic host &ells.
Suitable vectors generally will cont~in replicon
(origins of replication, for use in non-integrative
vectors) and control 6equences which are derived from
species compatible with the intended expression host.
By the term "replicable" vector as used herein, it is
intended to encompass vectors containing such
replicons as well as vectors which are replicated by
integration into the host genome. Transformed host
cells are cells which have been transformed or
transfected with vectors containing BCF encoding DNA.
The expressed BCF will be deposited intracellularly
or secreted into either the periplasmic space or the
culture supernatant, depend$ng upon the host cell
selected and the presence of suitable proces~ing
signals in the expressed peptide, ~.g. ho~ologous or
heterologous signal sequences.
Suitable host cells are prokaryotes or eukaryotic
cells. Prokaryotes include Gram negative or Gram
positive organisms, for example ~. ÇQli or bacilli.
Eukaryotic cells include yeast, higher eukaryotic
cells such as established cell lines of mammalian
origin, or insect cells. Expression in insect cells
may be accompll6hed using host cells and in~ect
expression vectors as disclosed by luckow, Y.A., and
Summers, M.B., Biotechnolooy 6:47-55 (1976).
Expre6sion vector~ for ho~t cell~ or~narily lnclude
an origin of replication, a promoter located upstream
from the BC~ coding sequence, together with a
ribosome binding ~ite, a polyadenylation 6ite, and a
transcriptional termination se~uence. Those o~
ordinary skill will appreciate that certain of these
oeguences are not reguired for expression in certa~n
hosts. An expression vector fGr use with microbes
need only contain an origin of replicat~on recognized
by the host, a promoter whic~ will function in the
host and a selection gene.
An expression vsctor is constructed according to the
present invention so that the ~CF coding sequence is
located in the vector with the appropriate regulatory
6equences, the positioning and orientat~on of the
coding seguence with respect to the control sequences
being such that the coding sequ~n~e is transcribed
under the "cont~ol~ of the control ~eguences (~--
~
RNA polymerase which binds to the DNA molecule at thecontrol sequences transcribes the coding 6equence).
The control sequences may be ligated to the coding
sequence prior to insertion into a vector, such as
the cloning vectors described above. Alternatively,
the coding seguence can be cloned directly into an
expression vector which already contains the control
seguences and an appropriate restriction site. For
expression of ~CF in procaryotes and yeast, the
control sequences will necessarily be heterologou~ to
the coding sequence. If the host cell is a
procaryote, it is also necessary that the coding
seguence be free of introns (~.g., cDNA). If the
selected host cell is ~ mammalian oell, the control
seguences can be heterologous or homologous to the
BMP coding sequence, and the coding sequence can
either be genomic DNA containing ~ntrons or cDNA.
Either genomic or cDNA cod~ng ~quences can be
expressed in yeast.
Expression vectors must contain ~ promoter which is
recognized by the host organism. Promoters commonly
-15-
known and available which are used in recombinant DNA
construction include the ~-lactamase (penicillinase)
and lactose pr~ot~r systeme~ a tryptophan (trp)
promoter syste3 and the tac promoter. While these
are commonly used, other ~nown microbial promoters
are ~uitable.
In addition to prokaryotes, eukaryotic cells ~uch as
yeast are transformed ~ith BCF encoding vector6.
Saccharomyce$ cerevisiae, or common baker's yeast, is
the most commonly used among lower eukaryotic host
microorganisms. However, a number of other 6trains
are commonly avallable ~a useful herein. Yeast
vectors generally will contain an origin of
replication from the 2 micron yeast plasmid or an
autonomously replicating sequence (ARS), a promoter,
DNA encoding BCF, seguences for polyadenylation and
transcription termination, and a selection gene.
Suitable promoting sequences in yeast vectors include
the promoters for the glycolytic enzymes such as
enolase, 3-phosphoglycerate kinase, glyceraldehyde-3-
phosphate dehydrogenase, hexokinase, pyruvate
decarboxylase, ph~sphofructokinase, glucose-6-
phosphate isomeras2, 3-phosphoglycerate mutase,
pyruvate kinase, triosephosphate i~omerase,
phosphoglucose isomeraee, and glucokinase.
Other yeast promoters, wh~ch have the additional
advantage Or transcription controlled by growth
conditions are the pro~oter regions for alcohol
dehydrogenase 1 or 2, isocytochrom~ C, ac~d
phosphatase,-~6 wQll ~s enzy~es responsible for
maltose an~ gal~ctose utilizat$on.
Higher eukaryot~c ~ell cultures may be used, whether
from vertebrate or invertebrate cells, including
6 ~
-16-
insects, and the procedures of propagation thereof
are known. See, for example, Tissue Culture,
Academic Press, Rruse and Patterson, editors (1973).
Suitable host cells for expressing BCF in higher
eukaryotes include: ~onXey kidney CVI line
transformed by SV40 (COS-7, ATCC CRL 1651): baby
hamster kidney cells (BHX, bTCC CRL 10); chine6e
hamster ovary-cells-DHFR (described by Urlaub and
Chasin, PNAS (USA) 77: 4216 ~1980)); mouse sertoli
cells (TM4, Mather, J.P., Biol. Reprod. ~: 243-251
(1980)); monkey ~idney cells (CVI ATCC CCL 70);
african green monkey kidney cells (VER0-76, ATCC CRL-
1587); human cervical carcinoma cells (HELA, ATCC CCL
2); canine kidney cells (MDCX, ATCC CCL 34): buffalo
lS rat liver cells (BRL 3A, ATCC CRL 1442); human lung
cells (W138, ATCC CCL 75); human liver cells ~Hep G2,
HB 8065): mouse mammary tumor (MMT 060652, ATCC CCL
51); rat hepatoma cell~ (HTC, M1, 54, Baumann, M., et
al., J. Cell Biol. 85: 1-8 (1980)) and TRI cells
(Mather, J.P., et al., Annals N.Y. Acad. Sci. 383:
44-68 (1982)). Co nonly used promoters are derived
from polyoma, ~denovirus 2, and Simian Virus 40
(SV40). It will be appreciated that when expressed
in mammalian tissue, the recombinant BCF may h~e
higher molecular weight due to glycosylation. It is
therefore intended that partially or completely
glycosylated forms of BCF having ~olecular weights
greater than that provided by the amino acid backbone
are within the ~cope of this inventlon.
A number of ~rocaryotic expression vectors are ~nown
in the art. See, e.g., U.S. Patent Nos. 4,440,859:
4,436,815; 4,431,740; 4,431,739; 4,~28,941;
4,425,437; 4,418,149; 4,411,994; 4,366,246;
4,342,832; ~ç~ Q U.R. Pub. Nos. GB 2,121,054;
GB 2,008,123; GB 2,007,675; and European Pub.
2Q~746~
-17-
No. 103,395. Preferred procaryotic expression
systems are in E. ÇQli- Other preferred expression
vectors are those for use in eucaryotic systems. An
exemplary eucaryotic expression system is that
employing vaccinia virus, which i well-~nown in the
art. See. e.g., Macket et ~1. (1984) J. virQl.
49:857; "DNA Cloning," Vol. II, pp. 191-211, ~upra;
PCT PUb. No. WO 86/07593. Yeast exprecsion vectors
are known in the art. See, ~.g., U.S. Patent
Nos. 4,446,235; 4,443,539: 4,430,428: see also
European Pub. Nos. 103,409; 100,561: 96,491. Another
expression system is vector pHSl, which transforms
Chinese hamster ovary cells. The use of the vector
is described in PCT Pub. No. WO 87/02062, the
disclosure of which is incorporated herein by
reference.
Mammalian tissue can be cotransformed with DNA
encoding a 6electable marker such as dihydrofolate
reductase (DHFR) or thymidine kinase ~nd DNA encoding
BCF. If wild type DHFR protein i5 employed, it is
preferable to select a host cell which is deficient
in DHFR, thus permitting the use of the DHFR coding
~eguence a~ marker for successful transfection in
hgt~ medium, which lacks hypoxanthine, glycine, and
2S thymidine. An appropriate host cell in t~is case is
the Chinese hamster ovary (CHO) cell line deficient
in DHFR activity, prepared and propagated as
described by Urlaub and Chasin, 1980, roc. Nat.
Acad. Sci. ~SA) 77: ~216.
Recently, expression vectors derived ~rom Bacclovirus
for use in insect cells have become ~nown in the art.
Dependinq on the expression system ~nd host
selected, BCF i6 produced by growing host cells
transformed by an expression vector described above
20~7466
-18-
under conditions whereby the BCP protein is
expressed. The enzyme protein is then isolated from
the host cells and purified~ If the expression
system secretes the enzyoe i~to growth media, the
protein can be purified directly ~rom cell-~ree
media. If the BCF protein i~ not secreted, it $s
isolated from cell lysates. the selection of the
approprlate growth conditions and recovery methods
are within the skill of the art.
The recombinantly made BCF is recovered from
transformed cell~ in accordance with p~ 6e ~nown
procedures. Preferably, an expression vector will be
used which provides for secretion of BCF from the
transformed cells, thus the cells may be separated
by centrifugation. The BCF generally is purified by
general protein purification techniques, including,
but not limited to, size exclusion, ion-exchange
chromatography, HPLC, and the like.
Once a coding seguence for BCF has been prepared or
isolated, it can be cloned into any suitable vector
or replicon and thereby ~ainta~ned in a composition
which is ~ubstantially free of vectors that do not
contain a BCF coding sequence ~ ., free of other
library clones). Numerous cloning vectors are known
to those of skill in the art. Examples of
recombinant DNA vectors for cloning and host cells
which they can tranQform $nclude the various
bacteriophage lambda vectors ~E. ÇQ10 ~ pBR322
(E. ÇQl~), pACYC177 (~. ÇQlO , pKT230 (gram-negat$ve
bacterla), pGV1106 (gra~-negative bacteria), pLaFRl
(gram-negative bacter$a), pME290 ~non-E. ÇQli qra~-
negative ~acterla), pflV14 ~. ÇQl~ and Baclllus
subtilis), pBD9 (BacilluQ), pIJ61 (Streptomyces),
pUC6 (Streptomyces), actinophage, ~C31
(Streptomyces), YIp5 (Saccharomyces), YCpl9
--19--
(Saccharomyces), and bovine papillo~a virus
(mammalian cells). See generallv, DNA Cloning:
Vols. I ~ upra: T. Maniatis et ~1-, EYE~;
B. Perbal, supra.
It i8 further intended that calcification-initiating
fragments of BCF are within the scope of the present
invention. Such active fragments may be prGduced,
for example, ~y pepsin digestion o~ BCF. The active
fragments may be identified by the ln ViYo and/or ~n
y~5~Q assays described hereinbelow.
Alternatively the BCF may be made by conventional
peptide synthesis using the principles of the
Merrifield synthesis and preferably using commercial
automatic apparatus designed to employ the methods of
~5 the Nerrifield synthesis. Peptides prepared using
the Merrifield synthesis may be purified using
conventional a~finity chromatography, gel filtration
and/or RP-~PLC.
FIG. 1 shows the aligned nucleotide and deduced amino
acid ~equences for both bovine and human BCF for
maximum amino acid seguence identity. Amino acids
not conserved in both species are boxed. The
put~tive initiation codon i6 located at position -~7
followed by a ~tretch of amino acids sbowing strong
hydrophobicity characteristics of signal peptides. A
putati~e signal peptide cle~vage ~ite iB indicated by
the arrow. The putative ~ature proteins begin at
positi~n l(Gln), and contain 183 amino ~cids, of
which 96.2~ ~re identical. The derived molecular
weight of human BCF i8 21,967 and for bovine BCF iB
21,984. The underlined ~equences are those from
which the oligonucleotide probes are deriv~d.
2 0 ~ 6 ~
-20-
Substantially pure BCF, higher molecular glycosylated
forms thereof, or active fragments thereof, or the
nontoxic salts thereof, combined with a
pharmaceutically acceptable carrier to form a
pharmaceutical composition, may be administered to
mammals, including humans, e~ther intravenously,
subcutaneously, percutaneously, intramuscularly or
orally.
Such proteins are often a~ministered in the form of
pharmaceutically acceptable nontoxic salts, ~uch as
acid addition salt6 or metal complexes, .g., with
zinc, iron or the li~e twhich are considered as salts
for purposes of this appli~ation). Illustrative of
such acid addition salts are hydrochlor~de,
hydrobromide, sulphate, phosphate, maleate, acetate,
citrate, benzoate, succin~te, malate, ascorbate,
tartrate and the like. If the active ingrediQnt is
to be administered in tablet form, the tablet may
contain a binder, such a~ tragacanth, corn starch or
gelatin; a disintegrating agent, such as alginic
acid; and a lubricant, such as magnesium stearate.
If administration in liguid form i~ desired,
sweetening and/or flavoring may be used, and
intravenous administration in isotonic saline,
phosphate buffer solutions or the like may be
effected.
Pharmaceutical compositions will usually contain an
effective amount of BCF in con~unction with a
conventional, pharmaceutically acceptable carrier.
Ihe dosage w~ll vary depending upon the specific
purpose for which the protein i8 being administered,
and dosage levels in the range of ~bout 0.1 pg to
about 100 milligrams per Kg. of body weight may be
used.
2~-~ 7~66
-21-
Implants of recombinant BCF, when mixed with matrix
Gla protein (MGP) (see Example 7), will initiate
calcification. The BCF may ~e either the human or
bovine form or mixtures thereof. Similarly, the MGP
may be any mammalian form thereof, preferably human,
bovine or mixtures thereof. ~atrix Gla protein (MGP)
may be isolated in the course of preparation vf bone
msrphogenetic protein (BMP) from de~ineralized
gelatinized bovine cortical bone ~y the methods of
Urist et ~1. [See: Price et al., proc. Natl. Acad.
Sci. USA 73:1447, 1976; Urist, M.R., Huo, Y.g.
Brownell, A.G., Hohl, W.M. Buyske, J., Lietze, A~o
Tempst, P., Humkapillar, M., and ~e~ange, R.J.:
Purification of bovine bone morphogenetic protein by
hydroxyapatite chromatography. Proc. Natl. Acad.
~Çi. 81:371-375, 1984 and Urist, M.R., Chang, J.J.,
Lietze, A., Huo, Y.K., Brownell, A.G., and DeLange,
R.J.: Methods of preparation and bioassay of bone
morphogenetic protein and polypeptide fragments. In:
Barnes, D., and S~rbaska, D.A. (eds.): Methods in
Enzvmolooy, vol. ~46. New York, Academic Press,
1987, pp. 294-312]. A preparation containing HGP is
first separated from other bone matrix prote$n by
hollow fiber ultrafiltration through a 10 K pore-size
filter. Under dissociated conditions in 6M urea and
0.02 M edetic acid (EDTA), the MGP assumes an
elongated structure in which proteins with 14 to 15 R
molecular weiqht (Mr) pass through ~ 10 R filter.
The MGP iB further purified by ion exchange
chro~atography (Berg, R.A. In: Methods in
Enzymology, 1982, vol. 82:372-398).
Furthermore, to initiate calcificatlon BCF and ~GP
may be mixed with any combination of one or more
other proteins, particularly, with one or more other
proteins derived from bone. Such mixtures may not
~7~
-22-
only initiate calcification, but may 8160 induce
cartilage formation and bone growth.
Implants of mixture~ of 8C~ and MGP induce
calcification in the quadriceys compart~ent. The BCF
cDNA may al80 be utili2ed in a diagnostic test for
identifying subjects having defective ~CF-genes,
defective BCF or autoantibodies directed against BCF:
or to detect levels of BCF, ~hich ~ay be an
indication of osteoporosis.
Preparations of BCF may be as6ayed ~n yivo according
to the method ~es~ri~ed by ~rist et ~1., Methods in
Enzymolo~y (D. Barnes and D.A. Sirbaska, Eds.),
vol. 146, pp. 294-312, Academic Press, N.Y. (1987),
and in Yitro by the ~ethod of Sato and Urist, Clin.
Orthop., 183:180-187 (1984) as modified by Kawamura
and Urist, ev. Biol., 130:435-442 (19~8), all of
which are incorporated by ref~ence herein.
It is preferred ~hat the BCF ~e admixed with matrix
Gla protein ~MGP) t~ for~ a de~lvery system
comprising the6e two proteins. The amount of MGP in
the composition is not believed ~o ~e critical and,
for convenience, equal porti~ Df BCF and MGP may be
used in dosages $n the range o~ about 0.1 ~g
(combined weight of BCF and ~ to 100 mg/Kg. body
weight.
The BCF and MGP ~ay be implant~a as a t~me-rolease
composition encapsulated, for instance, in liposomes
or other time-selea~e m~hranes, natur~l or
~ynthetic, which are a~sorbable by the host ~ub~oct.
m e purification protocol~, described in detail
below, allow for the f~r~t time the purification of
native BCF $n su~ficient quantity and at a high
enough purity to permit accurate amino acid
~17~
-23-
sequencing. The ~mino acid sequences derived from
the purified BCF ~llow for the design of probes to
aid in the isolation of native BCF nucleic ac$d
6equence, or the design of synthetic nucleic acid
seguences encoding the amino acid seguence of BCF.
Specific anti-sera or monoclonal antibodies
(described below) can be made to a synthet~c or
recombinant BCF peptide havinq the sequence or
fragments of the seguence of amino acid residues,
~uch as those shown in Figures lA or lB. An example
is the tryptic fragment shown in FIG. 2, and
antibodies thereto can be used to immunoprecipitate
any BCF present in a selected tissue, cell extract,
or body fluid. Purified BCF from this source can
then be sequenced and used as a basis for designing
specific probes as described above. Antibodies to
other regions that diverge from ~nown BCF can also be
used. Also useful as antigens are purified native or
recombinant BCF.
As mentioned above, a DNA sequence encoding BCF can
be prepared synthetically rather than cloned. The
DNA sequence can be designed with the appropriate
codons for the BCF amino acid sequence. In general,
one will select preferred codons for the intended
host if the sequence will be used for expression.
The complete sequence i~ assembled from overlapping
oligonucleotides prepared by ~tandard methods ~nd
asse~bled into a complete coding seguence. ~Ç,
~.g., Edge (1981) Nature ~92:7S6; Nambair, et ~1.
(1984) Science ~ 1299; Jay et ~1. (1984) ~. Biol.
Ç~ L:6311.
Synthetic DNA sequences allow convenient construction
of genes which will express BCF analogs or "muteins~.
Alternatively, DNA encoding muteins can be made by
&
-24-
site-directed mutagenesis of native BCF genes or
cDNAs, and muteins can be made directly using
conventional polypeptide synt~esi~. ~uteins altered,
for example, by the substitutio~ ~f acidic residues
~e.g., Glu or Asp) could have reduced activity toward
membrane-bound or complex substrates or have anti-
~ense therapeutic uses for overproduction of BCF.
Site-directed mutagenesis i8 conducted using a primer
synthetic oligonucleotide complementary to a sin~le
stranded phage DNA to be mutagen~zed except for
limited mismatching, representing ~he desired
mutation. ~riefly, the synthetic oligonucleotide is
used as a primer to direct synthesis of a strand
complementary to the phage, and the resulting double-
stranded DNA is transformed into a phage-supporting
host bacterium. Cultures of the transformed bacteria
are plated in top agar, permitting plaque formation
from single cells which harbor the phage.
Theoretically, 50% of the new plagues will contain
the phage hav$ng, as a single strand, the mutated
form: 50% will have the original 6equence. The
resulting plagues are hybridized ~ith kinased
synthetic primer at a temperature which permits
hybridization of an exact match, but at which the
mismatches with the original strand are sufficient to
prevent hybridization. Plague~ which hybridize with
the probe are then picked, cultured, and the DNA
recovered.
Native, recombinant or synthetic B~F peptide- (~ull
length or ~ubunits) can be u~ed to produce both
polyclonal and monoclonal hntibodies. IY polyclonal
antibodies ~re desired, purified 8CF peptide is used
to i~munize a selected mammal (e.~., mouse, rabbit,
goat, horse, etc.) and serum from the immunized
2 ~
-25-
animal later collected and treated according to known
procedures. Compositions containing polyclonal
antibodies to a variety of anti~en~ in addition to
BCF can be made ~ubstantially free of antibodies
which are not anti-BCF by ~mmunoaffinity
chromatography.
Monoclonal anti-BCF antibodies can also be readily
produced by one skilled in the art form the
disclosure herein. The general me~hodology for
making monoclonal antibodies by hybridomas is well
known. I~mortal, antibody-producing cell lines can
also be created by techniques ot~er t~an fusion, such
as direct transformation of B lymphocytes w~tb
oncogenic DNA, or transfection with Epstein-Barr
virus. See, e.g., M. Schreier et ~1-, "~ybridoma
Techniques" (1980); Bammerling et ~1., "Honoclonal
Antibodies And T-cell Hybridomas" (1981): Xennett ç~
~1., "Monoclonal Antibodies" ~1980): Ç~ al80 U.S.
Patent Nos. 4,341,761: 4,399,121: 4,427,783:
4,444,887; 4,451,570: 4,466,917; 4,472,500:
4,491,632: 4,493,890.
Panels of monoclonal ant$bodies produced against BCF
peptides can be ~creened for various propertie~:
i.~., isotype, epitope, affinity, etc. Of particular
interest are monoclonal antibodies that ne~tralize
the activity of BCF. Such monoclonals can be readily
identified in ~CF activity assays. High affinity
antibodies are also useful ~n immunoaffinity
purification of native or recombinant BCF.
Antibodies to ~CF forms de~ribed herein (both
polyclonal and monoclonal~ ~ay b4 u~ed to inhibit or
to reverse arterial calc~ficatio~. An ~ppropriate
therapeutic method would be to treat the patient with
an e~fective dose of anti-BCF antibodies through a
~7~
-26-
conventional intravenous route. In the treatment of
local, acute inflammation, treatment with anti-BCF
antibody wo~ld ~e indicated, perhaps by intramuscular
injection. The~e compssitions may also be useful in
targeting ~a~ious forms of tumors, since tumors are
known to sQ~eti~s calcify, suggesting the presence
of BCF. 2CF antagonists, such as BCF muteins, could
also be used in place of an~ibodies.
The determination of the appropriate treatment
regi~en (~.e~, dosage, freguency of administration,
systemic YS. lor~l, etc.) is within the ~kill of the
art. For administration, the antibodies will be
formulated in a unit dosage in;ectable form
(solution, suspension, emulsion, etc.) in association
with a pharmaceutically acceptable parenteral
vehicle. Such vehicles are usually nontox$c and
non~herapeutic. Examples of such vehicles are water,
saline, Ringer'~ ~olution, dextrose ~olution, and
Hank's solution. Nonagueo~s vehicles ~uch as fixed
oils and ethyl oleate may also be used. A preferred
vehicle is ~% ~w/w) human albumin in saline. The
vehicle ~ay contain minor amounts of additives, ~uch
as substances that enhance isotonicity and chemical
stability, ~.g., buffers and preservatives. The
antibody is typically formulated in such vehicle~ at
concentrations of abcut 1 ~g~ml to 10 mg/ml.
Anti-9CF antibodies will also be useful ln d~agnoctic
applications. The present invention contemplate~ ~
method, particularly a diagnostic method, ln which a
~ample from a human (or other mammal) iB provided,
and the amount of BCF i8 quantitatively measured in
an assay. For example, employing anti-BCF
antibodies in a quantitative ~mmunoassay could be
used to detect genetic deficiency in BCF. Antibody
specific for BCF could be formulated into any
~7~6
conventional immunoassay format: e.g., homogeneous or
heterogeneous, radioimmunoassay or ELISA. The
various formats are well known to tho~e skilled in
the art. See, e.a., "Immunoassayh A Practical
Guide" (D.W. Chan and M.T. Perlstein, eds. 1987) the
di~clo~ure of which i6 incorporated herein by
reference.
In general, recombinant production of B~F can provide
co~positions of that BCF ~ubstantially free of other
proteins having osteoinductive associated functions.
The ability to obtain high levels of purity i8 a
result of recombinant expression systems which can
produce ~CF in substantial quantities vis-a-vis ~n
vivo sources. Thus, by applying conventional
techniques to recombinant cultures, BCF compositions
can be produced that are substantially ~ore pure than
the compositions available from bone sources.
Purified BCF will be particularly useful as a tool
in the design and screening of calcification
inhibitors. First, milligram amounts of the material
are obtainable according to the present invention.
Milligram amounts are capable of crystallization to
permit three dimensional studies using ~-ray
diffraction and computer analysis. Thi~ may permit
deduction concerning the shape of the mDlecule, thus
defining proper shapes for substances u~able as
inhibitors of the activity normally exhibited by BCF.
Generally, antagonists have been ~peptides" whose
interactions with a factor which i~ inhibited are
stabilized by ~odification of the ~esidues~
participating in the peptide bond so as to enhance
the ability of the "peptide" to interact
specifically with converting factor. ~hus the
peptide bond ~oins specifically chosen carboxylic
acids and amines (not necessarily amino ~cids).
- 2 ~ ~ r~
-28-
These "peptides~ are configured in a three
dimensional array 60 ~ to complement the contour~ of
*he intended target, converting enzyme. A similar
lock and ~ey spatial arrangement may result from
~olecules designed c~mplementary to the surface
contours of the BCF of t~e invention. It i8
understood tbat "surface~ includes convolutions which
may face inward, and specifically includes the active
site. Furthermore, "complementary" is understood to
mean that, in addition to ~patial conformations which
~fit", interactions between the protein and the
molecule which matches its surface contour~ are
attractive and positive. T~ese interactions may be
hydrogen bonding, ionic, or hydrophobic affinity.
Accordingly, the invention contemplates peptide
antagonists or agonists (2-15 amino acid6) to BCF
which are characterized by three dimensional contours
complementary to the three dimens~onal contours on
the surface of recombinant BCF. By peptide in this
context i6 meant that the antagonist or agonist
contains carboxylic acid amide bonds correspondinq to
one less than the number of resldues. The carboxylic
acid and amine participants need not be ~-amino
acids.
Second, even without the assistance of a three
dimensional structure determination, purified BCF of
the invention is of significance as a reagent in
screening BCF inhibitor~ ~n vitro as an 8~ hQQ
approach to evaluation. lmpure BCF preparation6
currently available yield confusing data due to the
impact of the impurities on the test results. For
example, cont2~inants ~hich turn out to be themselves
inhibitors, activators, or substrates for BCF may
interfere with the evaluation. Thu8, a substantial
3S improvement in current screening techn$ques for BCF
6 ~
-29-
inhibitors would be effected by the availability of
the purified BCF protein.
It will be understood that this description and
disclosure of the invention i8 intended to cover all
S changes and m~difications of the invention which are
within the spirit and scope of the invention. It is
within the knowledge of the art to insert, delete or
substitute amino acids within the amino acid sequence
o~ a BCF without fiubstantially affecting the
calcification and bone growth inducing activ~ty of
the molecule. The invention i~ expressly ~tat~d to
be broad enough to include intentional deletions,
add$tions or substitutions. Furthermore, lt is
recognized that one skilled in the art could
lS recombinantly produce such modified proteins.
Native, reco~inant or synthetic BCF peptides (full
length or subunits~ can be further used to produce
both polyclonal and monoclonal antibodies. If
polyclonal ~ntibodies are desired, purified BCF is
used to i~munize a selected mammal ~e.g., mouse,
rabbit, goat, horse, etc.) and serum from the
immunized animal later collected and treated
according to known procedures. Compos$tions
containing polyclonal antibodies to a variety of
ant$gens in addition to BCF can be made ~ubstantially
free of antibodies which are not anti-BCF by
immunoaffinity chromatography.
Monoclonal ant~-BCF antlbodies can also be readily
produce~ by one skilled ln t~e art from the
disclosure herein. The general methodology for
~ak$nq monoclonal antibodies by hybridomas i8 well
known. Immortal, antibody-producing cell lines can
also be created by techn~ques other than fusion, such
as direct transformation of B lymphocytes with
2 ~
-30-
oncogenic DNA, or transfection with Epstein-Barr
virus. See, ~.q., M. Schreier et al., ~Hybridoma
Techniques" (1980); Hammerling et ~1-, "Monoclonal
Antibodies And T-cell Hybridomas" (1981); Kennett et
~1., ~Monoclonal Antibodies" (1980); see ~l~Q U.S.
Patent Nos. 4,341,761; 4,399,121; 4,427,783;
4,444,887, 4,451,570; 4,466,917; 4,472,500;
4,491,632; 4,493,890.
Panels of monoclonal antibodies produced against BCF
peptides can be screened for various properties;
.e., isotype, epitope, affinity, etc. Of particular
interest are monoclonal antibodie~ that neutralize
the activity of BCF. Such monoclonals can be readily
identified in BCF activity assays. High affinity
antibodies are also useful in immunoaffinity
purification of native or recombinant BCF.
Anti-BCF antibodies will also be useful in diagnostic
application~. For example, bone isolated ~rom
osteoporosis patients may show that it i6 deficient
in BCF. Thus, the present $nvention contemplates a
method, particularly a diagnostic method, in which a
bone 6ample from a human tor other mammal) is
provided, and the amount of BCF i~ guantitatively
measured in an assay. Antibody ~pecific ~or BCF
could be formulated into any conventional immunoassay
format; e.g., homogeneous or heterogeneous,
radioimmunoassay or ELISA. Tho variou~ formats are
well known to tho~e ~killed in tho art. Seo, ~.g.,
~ unoassay: A ~ractical Gu~de~ (D.W. Chan and
M.T. Perl~tein, eds. 1987) the di6closure of wh~ch iB
incorporated herein by reference. Quantitative
assays other than immunoassays could al~o bo u~ed to
measure the relative levels of BCF compared to a
s$andard or prior observed BCF level in a patient.
,
~7~
-31-
rhe following examples are provided by way of
illu~tration but are not intended to limit the
invention in any way.
EXAMPLE 1
~equence Analysis of BCF
The 22K proteins of interest, partially puri~ied from
human and bovine sources as descr~bed by Urist, et
~1-, Proc. Nat. Acad. Sci. ~a, ~1, 3?1-375 (1984),
were further purified to homogene~ty by preparative
gel electrophoresis and electroelution (M.W.
Hunkapiller, E. ~ujan, F. Ostrander and L.E. Hood,
Methods in EnzYmolooy, ~1: 227-236 (1983)). This
purification showed that the initial partially
purified sample6 contained, in addition to the 22R
BCF, other mammalian proteins at 34K, l9K, 14R and
6X. After precipitation with acetone (W. H.
Xonigs~erg and L. Henderson, Methods in Enzymology,
~: 254-259 ~1983)) and quantitation by amino acid
analysis (B~A. Bidlingmeyer, S.A. ~ohen and T.~.
Tarvin, JournaL Qf Chromat~araphv, 336: 93-104
(lg84)), the material was reduced under denaturing
conditions with 2-mercaptoethanol and cysts~ne
residues were derivatized with 4-vinyl-pyridine (H.
Friedman, L.G. Rrull and J.F. Cavins, Journal of
Bioloaical Chem~6trv, ~ 3868-3871 11970)). After
exhaustive dialysi6 to remove the ~enaturant, proteln
recovery was a~sessed by a repetition o~ ~mino acld
analysi6. The protein~ were digested with TPCR-
trypsin ln the pre~ence o~ 2H ~re~ to gensrat~
unblocked peptide fragment~ ~u$tab1e for ~equ~nce
analys$~ (G. Allen, Sequencina o~ Proteins ~nd
Peptides, page~ 51-62 (1981), Elsevier/North Nolland
Publishing Company, Amsterdam, Holland). Product~ of
the digestion were resolved by reverse-phas0 high
performance liquid chromatography using gradient~ of
2 ~
-32-
~cetonitrile or acetcnitrile/isopropanol in aqueous
trifluoroacetic acid tJ.E. Shlvely, Met~ods of
Protein Microcharacterization, pages 41-87 (1986),
Humana Press, Cl~fton, New Jer6ey). Peptide
fractions were sub~ec~ed to automated E~m~n
degradation using an kpplied Biosystems ~70A protein
6eauencer (M.W. Hunkapiller, ~.M. ~ewick, W.J. Dreyer
and L.E. Hood, Methods in Enzvmolooy, ~1: 399-413
~1983)). The phenylthiohydantoin amino ~cid
derivatives were identified by chromatography on an
Applied Biosystems 120A PIH analyzer (M.W.
Hunkapiller, Applied Biosvstems, User Bulletin Number
14 (1~85), Applied Biosystems, Foster City,
California). The hBCF sequence determined by this
method is confirmed by the ~equence deduced from the
human cDNA in FIG. 1.
EXAMPLE 2
RNA I601ation
mRNA was isolated ~rom fresh 7-month old calf bones
(obtained from Rancho Veal Meat Packers, Petaluma,
CA) or from human osteosarcoma cells.
Calf femur mids~afts were scraped free of connective
tissue and ~arrow, broken into cQarse fragments, and
frozen at -80'C. ~uman osteosarcoma cells were also
frozen ~t -80'C. RNA was i601ated ~ro~ both the
frozen ti~sues by the guanidinium thiocyanate/C6Cl
method (Maniat~s, T., Fr$tscb, E.F. and 8ambrooX, J.
~olecular Clonina: A Laboratorv Manual (Cold Spring
Harbor Lab., Cold Spring Harb~r, NY ~1982); Freeman,
G.J., Clayberger, ~., DeKruyf~, R., Rosenblum, D.S.
and Cantor, H. Proc Natl. Ac~d. Sci.: ~SA, 80: ~094-
4098 (1983)). An Qsterizer was used to pulverize
the bovine tissue directly in the guanidinium
thiocyanate extraction solution. Poly~A)+ RNA was
.. : ' ' .
~01746S
purified by a 6ingle fractionation over oligo~dT?-
Cellulose ~Maniatis, ~ al., supra).
Construction of the cDNA Libraries
Fir~t strand cDNA was synthesized from bovine matrix
or human osteosarcoma poly(A)+ RNA using conditions
similar to Okayama and Berg (Okayama, H. and Berg, P.
Holec. and Cell Biol. l: 4094-4098 (1983)). About
5~g of poly(A)+ RNA in 20~1 5mM tris-hydrochloride
(pH 7.5) was heated to 65-C for 3 min, then quick
cooled on wet ice and immediately ad~usted (at room
temperature) to contain 50mM Tris-hydrochloride (pH
8.3 at 42-C), 8mM ~gC12, 30mM XCl, lOmM
dithiothreitol, 2mM each of dATP, dGTP, dTTP ~nd
t~-32P]dCTP(-300cpm/pmol), 60 units RNasin, and 2.5
~g oligo (dT)12-18 (total volume 40 ml). The
reaction was initiated by the addition of 50-60 units
of cloned moloney murine leukemia virus reverse
transcriptase and continued for 60 min at 42-C.
Double-stranded [ds) cDNA synthesis and EcoRI linker
addition were performed by two different methods.
Initially, the second cDNA strand was synthesized by
the method of Wickens et ~1. (Wickens, H.P., Buell,
G.N. and Schi~ke, R.T. J. Biol. Chem. ~ 2483-2495
(1978)). The hairpin loop was removed from the ds
cDNA by SI nuclease, methylated ~ith EcoRI methylase,
made blunt-ended with T4 DNA polymeraQe, ligated to
phosphorylated EcoRI llnkers and flnally digested
with Eco~I (Maniatis, et ~1., su~ra). Later, the
~econd cDNA strand was synthesized by the method o~
Gubler and Hoffman (Gubler, U. and Hoffman, B.J. Gene
2S: 263-269 (1983)) as modified by Aruffo and Seed
(Aruffo, A. and Seed, B. Proc. Natl. Acad. 8ci.: USA
74: 8573-8577 (1987)). The ds cDNA was then ligated
to asymmetrically (hemi) pho~phorylated EcoRI
adapters (see oligonucleotide synthesis) as described
2 Q ~
-34-
by Aruffo and Seed, supra, phosphorylated with T4
polynucleotide kinase (Maniati6, et ~1., supra),
adjusted to 0.5M NaCl, 25mM EDTA ~nd heated at 75-C
for 15 ~in to inactivate the polynucleotide kinase.
The ds cDNA prepared by both procedures was separated
from unligated linkers/adapters ~y chromatography on
Biogel A-15m and recovered by ethanol precipitation.
cDNA was ligated to ~ZAP arms (Str?tagene) with T4
DNA ligase (New England ~iolabs) as described by
supplier, but included 15% polyethylene glycol (PEG)
8000 ~Sigma), a modification described by Pfeiffer
and Zimmerman (Pheiffer, B.H. and Zim~er~an, S.B.
Nucl. Acids Res. 11: 7853-7871 (1983)). The ligated
DNA was recovered by centrifugation (12,000 xg),
washed with chloroform, dried, resuspended in 4~1
water and incubated with an in vitro packaging
extract (Stratagene) according to supplier.
Reco~b~nant phage were propagated in ~. Qli BB4
(Stratagene).
EXAMPLE 3
~xn~hçsls of Oliaonucleot$des
Oligonucleotides were synthesized by the
phosphoramidite method with an Applied Biosystems
(Foster City, CA) model 380A synthesizer, pur$fied by
polyacrylamine gel electrophoresis and desalted on a
Waters SEP-PAK (C18) cartridge. A 10-mer
oligonucleot$de (5'CCGAATTCGG3') was synthesized and
u6ed as the EcoRI l$nker for cDNA library
construction. Prior to ligation, the linker was
phosphorylated with T4 polynucleotide k$nase
(Naniati~, T., Frit~ch, E.F. ~nd Sambrook, J.,
Molecular Clon$ng: A Laboratory ~anual (Cold Sprinq
Harbor Lab., Cold Spring Harbor, NY, ~982)). A 14-
mer oligonucleotide (5'CCTGTAGATCTCCG3') and a 18-mer
oligonucleotide (5'AATTCGGAGATCTACAGG3') were
2Q~7~6
-35-
synthesized and used as the EcoRI adapters. The 14-
mer was phosphorylated (Maniatis, ç~ ~1., upra) and
subsequently heated to 95'C for 15 min to inactivate
polynucleotide kinase, prior to annealing with the
18-mer. These asymmetrically phosphorylated adapters
also contained an internal ~glll restriction enzyme
site. Based on the amino acid ~e~uence of the human
~CF trypt~c fragment, a two-iold degenerate 45-mer
oligonucleotide probe was designed (FIG. 2, probe A)
following the rules of ~athe (~a~e, R.~, Mol.
Biol. 183: 1-12 (1985)). The two oligonucleotide
probes (A and B) were synthesized based on the ~m~no
acid seguence of a purified ~ryptlc peptide of the
human bone calcificati~ factor (hBCF) (FIG. 2~.
~x~NpLE 4
Screenina of the cDNA Librar~es
a. Human osteosarcoma l~braries.
Approximately 300,000 recQmbinant phage were plated
(SO,OOO phage/137mm dia. platet ~n ~. ~gli BB4, grown
for S-6 h ~t 37-C, transferred to nitrocellulose
filters (Hillipore, HATF 137)j prooessed accord~ng to
Benton and Davis (Benton, W.D. and Davis, R.~.,
~ç~çn~ 180 (1977)1 and soreened ~ith
oligonucleotide probe ~. Eighteen putative hBCF cDNA
clones from 300,000 recombinants were identified.
Southern blot analyses (described below) of the cDNA
inserts ~ere performed u~ing probe A and also probe B
(FIG~ 2), a fully de~enerate 18-~er oligonucleotide
oontained withln probe A. Most of tho clone~
hybr~dized to both probes. The probe wa8 ~nd-labeled
with T~ polynucleotide ~a~e ~nd 1~32-P]ATP
(Maniatis, et ~1., supra) to a speclfic activity of
1-2x108 cpm/~g. The Pilter~ were prehybrid~zed for
1-2 h at 37-C in 20% (vol/vol) formamide, 5xSSC
21~7~6
-36-
(lxSSC = 0.15 ~ sodium chloride/O.l~M sodium citrate,
pH 7), 5x Denhardt'~ solution (lx Denhardt's solution
- O.02~ polyvinylpyrrolidone/0.02% Ficoll/0.02%
bovine serum ~lbumin), 10% dextran sulfate, 50mM
~odium phosphate pH 6.8, 1 ~M sodiu~ pyrophosphate,
0.1% NaDcdS04 a~d 50pg/ml denatured salmon sperm DNA.
Labeled probe was added to a concentration o~
106cpm/~1 and hybridization was continued overnight
at 37-C with gentle 6haking. The filter~ were washed
twice (20 mln/wash) in 2X5SC, 0.1% NaDodS04 at 55-C
and expo~ed to Rodak XAR-2 film with ~ Dupont
~ightning Plu~ intensifying ~creen overnight at
-80-C. Areas of plagues giving signals on duplicate
filters were pic~ed, replated and rescreened as above
until pure plagues were obtained.
Two of the double positive clones (Ost 3-7 and 0st
3-17, FIG. 3) were sequenced and shown to contain
identical overlapping sequences as well as a region
encoding the tryptic fragment (FIG. 1 underlined and
FIG. 2). 0st 3-17 contains the complete mature
protein cod~ng seguence, but not the complete signal
peptide, as ~6 evident from the bBCF cDNA shown in
FIG. 1.
An additional 300,000 reco~binant phages fro~ two
different osteosarcoma cDNA libraries were later
plated and screened a8 above but ~ith the following
changes: (1) The hybridization mix contained 40%
~ormamide, 5xSSC, 5X Denhardt'~ solution, 10% PEG
8000, 50mM sodium phosphate pH 6.8, 0.5% NaDodS04 and
50~g/ml denatured; (2) the filters washed at 65-C ln
2xSSC, 0.1% NaDodS04; and (3) the probe wa~ ~ 240bp.
DNA ~ragment obta$ned by digesting cDNA clone 0st 3-
7 with ~glII and A~p718 (probe C, FIG 3). The probe
was purified and labeled by the oligo~primer method
(Feinberg, A.P. and Vogelste~n, ~., Anal. Biochem.
-37-
137: 266-267 (1984)) to a specific activity of >1 x
109 cpm/~g. Approximately 20 clones from each
library gave ~trong hybridization signal6 and
restriction enzyme analysis of these clones
identified several that were longer than 0st 3-17.
One of these, O~t 1-7 ~FIG. 3~, was ~eguenced and
~udged to be full length based on its homology to t~e
bovine BCF cDNA clone, described below.
b. 80vine bone matrix cDNA library.
Approximately 300,000 recombinants from the bovine
bone matrix cDNA library were screened with probe C
(FIG. 3), a 240b.p. BglII - Asp 718 gNA fragment from
0st 3-7, under the conditions described for probe A
except that formamide was om~tted fro~ the
hybridization solution. The filters were washed at
55-C in 2xSSC, 0.1% NaDodS04. Twenty-four positive
plaques were ldentified. Clone bbl.l-7 (FIG. 4),
which conta$ned the largest insert, was sequenced and
~hown to contain sequences homologous to hBCP. The
deduced amino acid seguence of the bBCF cDNA
indicates that amino acid -4 is not the initiation
codon due to the presence of a Val at this position.
The most likely initiation codon i~ located at a-ino
acid position -17 which is 28 nucleotides beyond the
5' end of the hBCF 0st 3-17 clone. The Het at
position -17 iB also preceded by an acceptable
ribosome binding s1te.
~g
Subclonina. Sequenc~ng and Analv~i~
Recomb~nant pla6~ids were relea~ed in the 81uescr~pt
SK(-)vector from ~ZAP by the M13 rescue/exci~ion
protocol described by the supplier (Stratagene). The
plasmids were propagated in E. çQl~ BB4, and plasmid
r~
--38--
DNA was isolated by the alkaline lysis method
(Maniat$s, çt al., SUDra). cDNA inserts were excised
with either EcoRI or BqlII restriction enzymes
~Boehringer-Mannhe$m), pur~fied by polyacrylamide gel
electrophoresis (MAniatis, ~ ~1., supra) and passage
over an Elutip-d colu~n (Schle$cher and Schuell) and
subcloned into the ~13 seguencing vector~ (Yanisch-
Perron, C., Viera, J. ~nd Messing, J., Gene 33: 103-
119 (198g)). DNA sequencing was performed by the
dideoxy chain termination method (Sanger, F. Nicklen,
S. and Coulson, A.R., ~roc. Natl. Acad. Sci. USA 74:
5463-67 (1977)) using M13 primers as well as ~pecific
internal primers. Ambiguous reg$ons were resolved 7-
deaza-2-deoxyguanidine-triphosphate (Barr, P.J.,
Thayer, R.M., Laybourn, P., Najarian, R.C., Seela,
F., and ~olan, D., ~otechniques 4: 428-32 (1986~)-
and sequenase (U.S. Biochemicals).
. Northern Blot
Poly(A)~ RNA was fractionated on a 1.4S agarose gel
in the presence of formaldehydé (Lehrach, H.,
Diamond, D., Wozney, J.~. and Boedtker, H.,
~iQQhemistrv ~: 4743-51 (1977)) ~nd directly
transferred to nitrocellulose according to Thomas
(Thomas, P., Proc. Natl. Acad. Sci. USA 77: 5201-5
(1980)). Filters were hybridized with probe C as
previously descr$bed (EXAHPL~ ~, Screening of the
cDNA ~ibrarie~) in the ~0% formamide containing
hybridizat$on solution and were washed at 55-C in
2xSSC, 0.1% NaDodS04 And O.lxSSC, 0.1% NaDodS04 with
autoradiography following each set of washings.
RNA transfer blot analysi~ demonstrate~ the presence
of two mRNA forms of ~0.9 and 1.8kB in human
osteosarcoma tissue. The two forms were also
observed in human placenta but were absent in a human
- 2 0 ~
-39-
liver cell line, HEPG2. The two forms were al~o
observed in bovine bone matrix cells ~ith the larger
form predominating. A trace of the large 6pecies
could also be seen in boviDe bone marrow. The two
mRNAs are ~ost likely generated by dif$erential
polyadenylation at the 2 site~ (AATAAA) found in the
3' untranslated region.
b. Southern Blot
1. Genomic
lO~g o~ geno~ic DNA ~lontech~ was digested with
EcoRI, fracti~nated on 0.7S ~garose gels ~nd
transferred to nitrocellulose according to Maniatis,
et al., supra. Hybridization and wa~hing conditions
were identical to those described in a. above.
Genomic DNA transfer bl~ analysis suggest6 that
hBCF and bBCF are elngle copy ~or low copy) genes due
to the few bands seen in ~he EcoRI digest.
2. cDNA Clon
DNA from c~NA clone~ were ~igested with EcoRI or
BglII, fractionated on 1.0% agarose gels, transferred
to nitrocellulo6e (Maniatis~ ç~ u~ra) and
hybridized with probe A as previously described or
with probe B in a tetramethylammonium chlorid-
contain~ng hybridization solution under conditions
described by Wood (Wood, W.I., Git~chier, J., LaskQy,
L. and Law~, R., p~gc. Natl. A~ad. Sci. ~BA 82: 1585-
88 (1985~).
;` 2~7~6~
-40-
ExAMpLE 6
Expression of hBCF in Yeast
Yeast expression vectors were constructed for
intracellular production or secretion of hBCF from
the ADH2/GAPDH regulatable promoter. Since the first
cDNA clone did not contain DNA encoding a signal
peptide, methionine-4, was used as the N-terminal
amino acid for these constructions. Su~sequent cDNA
analyses, and identification o~ a classical ~ignal
pept~de, and a ~ignal peptidase cleavage ~te,
allowed construction of a yeast ~-factor/hBCF fusion
with glutamine ~1 as the N-terminal amino acid of
recombinant hBCF. For clonlng into expression
vectors, natural Nco-l, Bgl-l and Spe-1 sites were
used together with the synthetic oligonucleotide
adapters shown (FIG. 4). Since Spe-l and Xba-l give
the same restriction enzyme overhang, the initial
construction w~s simply an in~ertion of the Nco-l/
SPE-l hBCF geno into Nco-l/Xba-l digested
pBSlOOhaFGF, a vector contain~ng ADH2/GAPDH promoter
and GAPDH terminator elements flanking a synthetic
human acidic fibroblast growth factor (FGF) gene.
The haFGF gene contains an Nco-l and unique Xba-l
sites. The resulting plasmid, pBSlOOhBCF, was used
for further constructions. Thus, the Nco-l/Sal-l
fragment containing the hBCF gene was cloned together
with BAMHl~Nco-l fragments encoding the GAPDH and
ADH2/~APDH promoters lnto BamHl/Sal-l dig~sted
pBS24.1. The resulting plasmids; pBS24A/GhBCF~Q ~-4
to 183) and pBS24 GAPhBCF (-3 to 183), were u~ed to
direct intracellul~r expression of hBCF. For
~ecretion, Bgl-l/S~ ragments were exci~ed and
cloned into pBS24.1 together with BamHl/Xba-l
~ragment3 encoding the ADH2/GAPDH promoter, th~ yeast
n-factor secretory ~$gnal/leader sequence, Dnd the
~ynthstic linker~ shown $n FIG. 4 (boxed), ~or
~7~:6
-41-
expression of hBCF(-4 to 183) and hBCF(l-lB3). Yeast
cells transformed with expression plasmids encoding
ADH2/GAPDH promoter-hBCF(-4-183) and GAPDH promoter-
hBCF(-4-183) fusions, as analyzed by SDS-PAGE and
Coomass~e blue 6taining of total proteins, 6howed no
expresgion a8 compared wit~ control yeast cells.
Therefore, to study secretion ~ystems, the pBS24
plasmids containing ~-factor leader-hBCF (-3-183) and
hBCF (1-183) fusions were constructed. In each case,
transcription was driven by the ADH2/GAPDH promoter.
Yeast 6train ABllO was transformed with the yeast
expression plasmids, and yeast supernatants were
analyzed by precipitation with 10% trichloroacetic
acid and separation by SDS-PAGE. ~i~h level6 of
expression of the approximately 22kD product was
observed by Coomassie blue ~taining, when compared
with control yeast ceils transformed with the parent
yeast vector pBS24. The transformant ABllO (pBS24.1
22kQ) is deposited under accession number ATCC 20948.
The active lot of yeast cell~ comprised a pool of
two lot~, KQ-2 and KQ-3, from which the recombinant
BCF waæ isolated and purified as follows.
Lot K0-2
The cells were removed by centrifugation from the
fermentation medium, and the medium concentrated
using a YM10 Amicon ~piral cartridge. The
concentrate was diafiltered $nto water, tben 20 ~M
Tris-Cl, 1 ~M EDTA, 3 M NaCl, pH 7.5. This was
passed over a column of prep grado Superose 12 ~t
4-C, then at room temperature. The 22XBCF did not
stlck to either column. The ~lowthrou~h wa~ purified
by adsorption onto SuRerose 12 ~R 10/30, and eluted
with the same buffer but with 1 ~ NaCl.
-42- 2 ~
Lot K0-3
The cells were ~emoved and the medium concentrated as
described ~b~ve. The concentrate was diafiltered
against water, then 20 mM Tri~-Cl, 1 mM EDTA, pH 7.5.
This was loaded onto a Mono-Q HR 10/10 column,
washed, and eluted with a gradient to 0.5 M ~aCl.
The 22KBCF-cDntaining fractions were identified by
gel electrophoresis, pooled, ad~usted to 3 M NaCl
with solid NaCl, and loaded onto a Superose 12 HR
10/30 column. The 22K~CF was eluted with buffer
containing 1 H NaCl.
L~t X0-2/3
~he Superose eluates (from Lots KQ-2 and XQ-3) were
pooled, concentrated using YM 10 membrane in Amicon
6tirred cell, dialyzed versus water, and lyophilized.
Alternatively, the recombinant BCF may be isolated
and purified as follows.
The media is removed from the cells and concentrated
approximately ten fold. The pH is adjusted to 7.5,
the concentrate $~ diluted to a conductivity below
5 mS/cm, then applied to Fast Flow Q ion-exchange
resin pre-eguilibrated with 50mM Tris/Hcl, lmM EDTA,
lmM PNSF, pH 7.5. The column i~ washed with 1 column
volume of the above buffer and eluted u~ing a 0-lM
~aCl ~alt gradient in the above buffer. ~he 22R~CF
i~ eluted at a conductivity o~ 20-30 mSfcm, which i8
confirmed using SDS-PAGE. The 22KBCF conta$ning
fractions are pooled, t~e pH maintained at 7.5 ~nd
made 4M with sespect to Urea, concentrated, and run
over ~ S-100 ~izing column in 4N Urea, lOOmN
Tri6/HCl, lmM EDTA, lmM PMSF at pH 7.5. The 22RBCF
containing fractions are ident~ied by SDS-PAGE,
20~6~
-43-
pooled, concentrated, and dialysed aga~nst lOmM
ammonium bicar~onate pH 7.8. The 22X~CF is then
lyophilized and may be stored dry at 4-C.
EXa~PLE 7
A ~ample of purified recombinant human ~CF(rhBCF)
expressed as in Example 6, 1 ~g tlot XQ 2/3), was
dissolved in water containing 5 mg of buman fibrin.
The rBCF-fibrin composite was lyophilized and
implanted in the mouse thigh muscle pouch. In
another mouse, bovine matrix Gla protein was
dissolved in 6M urea containing 1 mg of human rhBCF
and dialyzed again~t water. The precipitate and
6upernatant were lyophilized to prepare a co~posite
of MGP and hBCF protein~. The composite was
implanted in the guadriceps pouch. For controls, the
contralateral thiqhs were implanted with bovine
matrix Gla protein plus albumin in one nouse: in the
other mouse the control con~i~ted of 5 ng of human
fibrin and 1.0 mg o~ albumin.
Microradiographs of the r~CF-matrix Gla protein
composite showed areas of calcified tissue.
Histological section6 6howed small round cells,
multinucleated cells, macrophages with hyperplasia
and hypertrophy of mesenchymal type cell6. There
were plates of calcified ground 6ubstance but no
cartilage or bone cell~. FIG. S is a photomicrograph
showing hyperplasi~ and hypertrophy of connective
tissue cells on the surface of a composite i~plant
~indicatad by P) of reco~binant protein hBCF ~ot KQ
2/3, 1 ~gt ~nd 1 ~g of human ~atrix Gla protein.
Calcificatlon of the implanted proteins i8 indicated
by CP in the guadriceps pouch of a ~ouse on day 21.
The fibrous connective tissue envelope i~ indicated
by F.
FIG. 6 is a photomicrograph showing islands of
calcified protein (CP) induced by a composite of
recombinant hBCF and biological human matrix Gla
protein. Note the large foam cell (arrow). The
entire implant is enclosed in a fibrous capsule (F)
by day 21.
FIG. 7 is a photomicrograph showing ~ co~posite of
recombinant h~CF and ~atrix Gla protein on day 21.
Note the highly vascular interior of the l~plant
including small round cells, multinucleated cells,
macrophages, calcifying intercellular substance (CP)
and large foam cells (arrow). All of the implanted
protein and reactive tissue was enclosed in a fibrous
envelope (F) on day 21.
