Note: Descriptions are shown in the official language in which they were submitted.
13 3 8 3 3 8 HOE 87/F 132
A process for the preparation of foreign proteins in Streptomycetes
European Patent Application 85 105 610.1 (Filed May 8, 1985,
published November 21, 1985 as No. 0 161 629, Applicant Hoechst
Aktiengesellschaft), and South African Patent 85/3672 (issued
December 24, 1985) disclose the use of the DNA coding for the signal
peptide (prepeptide) of the a-amylase inhibitor tendamistat in order
for a Streptomycetes cell to excrete a polypeptide, in particular
tendamistat. The appropriate DNA can, in principle, be obtained from
every strain producing tendamistat, but a DNA obtained as in Example
3 of German Offenlegungsschrift 3,331,860 (published March 21, 1985,
Applicant Hoechst Aktiengesellschaft) is preferably used.
Canadian Patent Application 560,579 (Filed March 4, 1988) has already
proposed a process for the excretion of fusion proteins from
Streptomycetes, which comprises incorporating the coding sequence has
been modified where appropriate, and expressing the recombinant gene
in a Streptomycetes cell. Thus, in this case the tendamistat
structural gene is used as a "carrier" for another gene, the fusion
proteins which are obtained having the amino acid sequence of another
protein located within the tendamistat amino acid sequence.
Consequently, on chemical or enzymatic cleavage of this fusion
protein to liberate the other protein, two tendamistat part-sequences
are obtained. The said earlier application also relates to
tendamistat derivatives, which are understood to include those with a
markedly shortened amino acid chain. Derivatives of this type are
able in a reversible manner to react with the specific receptors in
the form of a competitive inhibitory mechanism.
It has now been found that foreign proteins can also be prepared in
Streptomycetes by constructing fusion protein
~ - 2 - 1338338
genes in whi ch the structural gene for t:he desired protein is
ooupled to the 3' end (of the coding strand) of the t~n~m;.~-
tat gene, ~hich has been modified where appropriate. The
modification of the tendamistat gene may comprise, in
S particular, C-terminal shortening.
The DNA coding for tendamistat is depicted in EP-A
0,161,629 (where it is DNA sequence C; Table 1 in the
annex). This structural gene contains several cleavage
sites for restriction enzymes, which can be used to modi-
fy the coded amino acid sequence. Suitable cleavage sites
are those for BstEII in the region of triplets 31 and 32,
StuI in the region of triplets 43 and 4~, and Sau3A in
the region of triplets 52 and 53. It is possible, by in-
corporation of appropriate linkers, to insert at these
sites one or more additional amino acids, to eliminate
DNA segments between these cleavage sites, or to code for
shortened amino acid sequences by incorporation of stop
codons. Furthermore, it is possible by site-specific
mutagenesis for any desired amino acids to be inserted,
replaced or eliminated. In this way proteins are obtained
which have an ~-amylase inhibitory action, as well as
proteins which do not have this activity but still react
with the corresponding receptors.
O The invention also relates to appropriate gene structures,
vectors containing these gene structures, Streptomycetes
cells transformed with these vectors, the excreted fusion
proteins, and their use for the preparation of the foreign
proteins and tendamistat derivatives. Preferred embodi-
ments of the invention are explained in detail herein-
after and defined in the patent claims.
figures 1 to ~ depict some plasmid constructions accord-
ing to the invention.
Fig. 1 shows the preparation of the hybrid plasmid pKK310
which codes for a fusion protein in which part of the
- 1338338
-- 3
tendamistat amino acid sequence is follo~ed by a bridging
member of seven amino acids and, thereafter, the amino
acid sequence of monkey proinsulin.
Fig. 2 sho~s the construction of the expression plasmid
pTF1 starting from the plasmid pKK310.
Fig. 3 sho~s the construction of the plasmid pRS10 in
~hich part of the tendamistat gene is followed by the
polylinker from pUC18, and its reconstruction into the expression
plasmid pTF10. "mcs" denotes the polylinker region
(multiple cloning site) of pUC18.
Finally, Fig. 4 sho~s the construction of the plasmid
15 pKK400 ~hich codes for a fusion protein in which the ~hole
of the amino acid sequence of tendamistat is follo~ed by
a bridging member of eleven amino acids and, thereafter,
~ the amino acid sequence of monkey proinsulin, and its
reconstruction into the expression plasmid pGF1.
The figures are not dra~n true to scale.
The fusion proteins obtained according to the invention,
~hich are exported from the cell, have the advantage that
25 they can readily be isolated from the culture filtrate.
O The isolation can be carried out in a manner known per
se, advantageously by adsorption or ion exchange chroma-
tography and/or gel filtration.
30 The desired foreign protein (fusion partner) is liberated
by enzymatic or chemical cleavage like~ise in a manner
known per se.
In this connection, the type of cleavage depends, in par-
35 ticular, on the amino acid sequence of the desired pro-
tein. It will be expedient in many cases to incorporate
a connecting member or bridging member in the cleavage
site bet~een the tendamistat sequence and the amino acid
` _ 4 _ I 33 83 3~
~
sequence of the desired protein. If the desired protein
contains, for example, no methionine, the connecting mem-
ber can denote Met, whereupon chemical cleavage with
cyanogen chloride or bromide is carried out. If the con-
necting member has a carboxyl-terminal cysteine, or if
the connecting member represents Cys, it is possible for
enzymatic cysteine-specific cleavage, or chemical cleav-
age, for example after specific S-cyanylation, to follow.
If the bridging member has a carboxyl-terminaL tryptophan,
or if the connecting member represents Trp, chemical
cleavage with N-bromosuccinimide can be carried out.
Desired proteins which do not contain Asp-Pro in their
a0ino acid sequence and are sufficiently stable to acid
can, as fusion proteins having this bridging member, be
cleaved proteolytically in a manner known per se. This
results in proteins which contain N-terminal prolineand
C-terminal aspartic acid. Thus, it is also possible in
this way to synthesize modified proteins.
The Asp-Pro bond can be made even more labile to acid if
this bridging member denotes (Asp)n-Pro or is Glu-~Asp)n-
Pro, n denoting 1 to 3.
Examples of enzymatic cleavages have also been disclosed,
O it also being possible to use modified enzymes having
ioproved specificity (cf. C.S. Craik et al., Science 228
(1985) 291-297). If the desired eukaryotic peptide is
proinsulin, it is expedient to choose a peptide sequence
in which an amino acid which can be eliminated by trypsin
(Arg, Lys) is bonded to the N-terminal amino acid (Phe)
of the proinsulin, for example Ala-Ser-Met-Thr-Arg, since
it is then possible to carry out the arginine-specific
cleavage using the protease trypsin.
If the desired protein does not contain the amino acid
sequence
- _ 5 _ 1338338
Ile-Glu-Gly-Arg,
the fusion protein having the corresponding bridging member can be
cleaved with factor Xa (European Patent Application published under
No. EP-A 0,025,190 on March 18, 1981 and European Patent Application
published under No. EP-A 0,161,973 on November 21, 1985.
The isolation of the cleavage products depends on the properties of
these proteins. Concerning the isolation of tendamistat and its
derivatives, reference may be made to the literature cited in German
Offenlegungsschrift 3,331,860 (published March 21, 1985, Applicant
Hoechst Aktiengesellschaft).
The invention is explained in detail in the examples which follow.
Unless other~ise indicated, percentage data relate to weight. The
figures having the same numbers as the examples relate thereto.
Ex~mple 1
The starting material used is the plasmid pKAI650, ~hich is described
in European Patent Application 86 113 627.3 (Filed October 2, 1986,
published April 15, 1987 as No. 0 218 204, Applicant Hoechst
Aktiengesellschaft). This plasmid can be obtained from the plasmid
pKAIl, which is described in German Offenlegungsschrift 3,331,860
(published March 21, 1985, Applicant Hoechst Aktiengesellschaft), by
isolation of the 650 bp HincII/SstI fragment and cloning into the
plasmid pUCl9 which has been opened with these enzymes. The unique
HindIII cleavage site in this plasmid is removed (by cutting with
this enzyme, filling in the protruding ends, and ligation) to result
in the plasmid pKAI650a (1).
2 1~9 of (1) DNA puritied by Cstl gradient centrifugat ion
are completely digested, in a 50 yl reaction mixture,
uith StuI for 2 hours as stated by the manufacturer, and
the en~yme is removed by phenol extraction. The lineari~ed
DNA is pretipitated ~ith ethanol, redissolved and intro-
duced into a ligation mixture to ~hich is added, as
additional reactant, 0.1 ~9 of the chemically synthesi~ed
double-stranded oligonucleotide ~2) ~hich has been
. _ - 6 - 1338338
phosphorylated at the 5' end
HindIII
5' C C C A A G C T T G G G 3' (2)
3' G G G T T C G A A C C C S'
Transformation of the ligation mixture into E. coli
JM 109 is followed by isolation of those clones ~hich
harbor the recombinant plasmid pKK3a (3). The isolated
plasmid DNA has a cleavage site for the restriction enzyme
HindIII, ~hich permits characterization by restriction
analysis. pKK3a (3) is 12 base-pairs larger than pKAI650a
C~ and has a nucleotide sequence which extends the amino
acid sequence by 4 amino acids, as follo~s
42 43 (44)
- Glu Gly Pro Ser Leu Gly Leu
5' GAA GG C CCA AGC TTG GG C CTG 3'
3' CTT CC G GGT TCG AAC CC G GAC 5'
HindIII
The other starting material used is the plasmid pYE24 (4).
This plasmid is obtained by opening the vector pUC8 ~ith
EcoRI and HindIII, and ligating into this linearized plas-
mid the gene for monkey proinsulin (Table 2; cf. ~etekam
et al., Gene 19 (1982) 179-183).
2 ~9 of the plasmid pYE24 (4) are reacted with the re-
striction enzymes EcoRI and HindIII, and the gene formonkey proinsulin is isolated by electroelution and,
after purification and concentration by ethanol precipi-
tation, it is ligated ~ith the synthetic DNA linker (5)
5' AGC TTG ATG GCG 3'
3' AC TAC CGC TTA A 5' (5)
(HindIII) (EcoRI)
- 1338338
-- 7
The ligation product (6) is now inserted into the plasmid
pKK3a (3) ~hich has been opened with HindIII, resulting
in the plasmid pKK31 (7). This construction results in
the following bridging member being do~nstream of the
codon for amino acid Gly 43 of the tendamistat gene:
43 B 1
Gly Pro Ser Leu Met Ala Asn Ser Phe
GGC CCA AGC TTG ATG GCG AAT TCT TTT
CCG GGT TCG AAC TAC CGC TTA AGA AAA
HindIII EcoRI
"~ 1" here, and in Table Z, designates the start of the
B chain of monkey proinsulin.
In plasmid (7) the proinsulin sequence is located within
- the tendamistat gene. To reconstruct this plasmid into a
plasmid according to the invention, (7) is digested with
SphI and SalI, and the fragment (8) is isolated. The
vector pUC19 is opened ~ith SphI and SalI, and the
linearized plasmid is ligated ~ith the fragment (8). The
resulting plasmid pKK310 (9) codes for a fusion protein
in ~hich the shortened tendamistat sequence and the linker
uhich is depicted above are follo~ed only by the proinsu-
lin sequence.
G The entire construction is depicted in Figure 1.
Example 2
To reconstruct the plasmid pKK310 (9) in an expression plas-
mid, (9) is reacted with SstI and SphI, and the fragment
(10) is isolated.
The commercially available expression vector pIJ702 (11)
(obtainable from the John Innes Foundation, Nor~ich,
England) is opened ~ith SphI and SstI, and the linearized
plasmid (12) is ligated ~ith the fragment (10). After
transformation of the strain sereptomyces lividans TK Z4
- 1338338
_ - 8
~John Innes Foundation), the desired clones are identified
by selection for resistance to thiostreptone. The plas-
mid DNA from thiostreptone-resistant clones is isolated
and examined by restriction analysis. Plasmids having
the desired orientation of the gene are called pTF1 t13).
Clones which contain this recombinant plasmid secrete a
protein of molecular weight 16 kD into the culture medium.
This protein shows a positive "immunoblotting" reaction
with insulin antibodies (cf. Example 5).
The construction of pTF1 (13) is depicted in Figure 2.
C-v Example 3
15 The plasmid pKK3a (3), on the one hand, and the vector
- pUC18, on the other hand, are each opened with HindIII,
and are ligated together. The ligation mixture is used
to transform the E. coli strain JM 109, which indicates
successful cloning in the presence of isopropyl-B-thio-
20 galactopyranoside (IPTG) and 5-bromo-4-chloro-3-indolyl-B-
D-galactopyranoside (X-Gal) by the formation of colorless
colonies. The resultant recombinant plasmid pRS1 (14)
is isolated in a manner known per se. Digestion of 1 ~9
of the plasmid with the restriction enzyme SstI, followed
25 by religation results in deletion of the pUC18 portion
apart from the polylinker sequence (mcs) and the remainder
of the tendamistat gene. The plasmid pRS10 (15) is ob-
tained.
30 The plasmid (15) is, owing to its polylinker portion,
sui~able for cloning any desired structural genes, re-
sulting in plasmids which code for the corresponding
fusion proteins with the shortened tendamistat sequence.
35 ~hen pRS10 (15) is digested with SphI and SstI, and the
smaller fragment is isolated, the latter can be ligated
into the expression vector pIJ702 in analogy to Example 2.
In this way the expression vector pTF10 (16) is obtained,
1~38338
_ _ 9 _
and this likewise, by reason of its polylinker portion,
allows versatile constructions.
The construction of pTF10 (16) is depicted in Figure 3.
S
Example 4
The plasmid prE24 (4) is opened ~ith EcoRI, and the linker
5' AAT TCA AGC TTG 3'
3' GT TCG AAC TTA A 5'
(EcoRI) Hind I II (EcoRI)
C~
is inserted, resulting in the plasmid pYE241. Cutting
with HindIII, and ligation into pKK3a (3) cut ~ith HindIII
results in the plasmid pKK32, in analogy to Example 1.
e The latter codes for a fusion protein in which the ten-
damistat sequence is linked to the proinsulin sequence
by the foLlo~ing bridging member:
43
Gly Pro Ser Leu Asn Phe Ala Arg
GGC CCA AGC TTG AAT TCT GCA AGA TTT
CCG GGT TCG AAC TTA AGA CGT TCT AAA
In analogy to Example 1, pKK32 is cut ~ith SphI and SstI,
and the fragment which is approximately 650 bp in size
is cloned into pUC19, which has been opened ~ith these
enzymes. The resulting plasmid pKK320 corresponds to
plasmid pKK310 ~9) apart from the abovementioned bridging
member (in which the sequence introduced by the linker
is emphasized by emboldening).
In analogy to Example 2, the SstI-SphI fragment having
the recombinant gene from pKK320 is cloned into pIJ702,
resulting in the expression plasmid pTF2. A fusion pro-
tein of 16 kD is expressed and secreted in S. lividans
TK 24, and the protein reacts uith insulin antibodies
- 10 -
(cf. Example 5). 1338338
~ecause of the similarity of the construction of pTF2 to
that of pTF1 (13), Figures 1 and 2, no depiction in a
dra~ing has been given.
ExampLe 5
pKK310 (9) is partially digested with EcoRI so that only
one of the two EcoRI cleavage sites is opened. After the
protruding ends have been filled in using Kleno~ polymer-
ase, the plasmid is religated, and the result is checked
C by restriction analysis. The desired plasmid, in which
~ the EcoRI site located at the end of the proinsulin gene
has been eliminated, is called pKK310a (17). Thus, the
latter no~ contains a unique restrict;on site for EcoRI
in the linker region bet~een the shortened tendamistat
gene and the proinsulin gene.
To construct the plasmid which codes for a fusion protein
having the complete tendamistat sequence, a unique cleav-
age site for KpnI is introduced, in the region of the
codons for amino acids 68/69, into the DNA sequence cod-
ing for tendamistat (Table 1). This entails the isolated
DNA from pKAI650a (1) being digested ~ith SstI and SphI,
O and the fragment which is 650 bp in size being cloned
into the phage M13mp18 RF DNA, ~hich has likewise been
digested ~ith these t~o enzymes, and the single-stranded
DNA being prepared by knovn methods. 1 ~9 of this ssDNA
is used together ~ith 0.1 ~9 of the mutagenic "primer"
5' C GAG GTA CCG GGC GT 3'
in site-directed mutagenesis (M.J. Zoller and J. Smith,
Nucleic Acid Res. 10 (1982) 6487-6500).
The RF DNA is isolated from the isolated M13 clones having
the mutated gene, which can be selected by the additional
- 11 - 1338338
KpnI cleavage site, and the base exchange (C for G at the
third position in the codon for Arg68) is confirmed by
sequencing. Thus, the nucleotide exchange brings about
no change in the amino acid sequence but does introduce
S the desired new unique cleavage site into the tendamistat
structural gene.
66 67 68 69 70
His Ala Arg Tyr Leu
10CAC GCC CGC TAC CTC
GTC CGG GCC ATC GAG
KpnI
C~ The mutated sequence is, after SstI-Sphl digestion, cloned
15out of the M13mp18 RF DNA into the plasmid pUC19, result-
ing in the plasmid pKAI651 (18).
To check, the 650 bp SStI-SphI insert from (18) is incor-
porated, as in Example 2, into the plasmid pIJ702, re-
sulting in the plasmid pAX651. After this plasmid hasbeen transformed into Streptomyces lividans TK 24, the
expression rates for tendamistat ~hich are obtained are
the same as for the plasmid pAX650 having the unmodified
tendamistat gene (German Offenlegungsschrift 3,536,182,
Fig. 3).
o
To prepare a plasmid, according to the invention, for a
fusion protein having the entire amino acid sequence of
tendamistat, the plasmid pKAI651 (18) is no~ digested
~ith SphI and KpnI, and the small fragment is ligated
~ith the linker (19)
69 70 71 72 73 74
(Tyr)Leu Ala Arg Cys Leu Phe Asn Ala Met Ala Thr Gly 3'
5' CTC GCT CGC TGC CTT TTC AAT GCG ATG GCC ACC GGG
3' C ATG GAG CGA GCG ACG GAA AAG TTA CGC TAC CGG TGG CCC TTA A 5'
(KpnI) (19) (EcoRI)
~ - 12 - 1338338
and the plasmid pKK310a (17) ~hich has been opened ~ith
Sphl and EcoRI.
~he ligation mixture is used to transform E. coli JM 109,
the plasmid DNA is isolated, and the correct fusion is
verified by DNA sequencing. ~he plasmid having the
correct sequence is called pKK400 (20).
~he linker (19) codes not only for the remaining amino
acids of tendamistat but also for the portion of a spacer
~hich separates the tendamistat and proinsulin genes from
one another, and overall embraces, ~ith the 5' end of the
gene as shovn in Table 2, the codons for the follo~ing
11 amino acids:
Phe-Asn-Ala-Met-Ala-Thr-Gly-Asn-Ser-Ala-Arg
Thus, the fusion protein contains in this spacer, inter
alia, the amino acids methionine and arginine, ~hich per-
mit cleavage vith cyanogen halide or trypsin.
The insert of about 1090 bp is isolated from the plasmidpKK400 (20) by double-digestion ~ith SstI and SphI, and
the DNA is ligated into the plasmid plJ~02 (12) vhich has
been opened uith the same enzymes. The result is the
plasmid pGF1 (21). The ligation mixture is transformed
into S. lividans TK 24, and the plasmid DNA is isolated
from thiostreptone-resistant transformants ~hich have
tendamistat activity. AlL positive clones contain the
pGF1 Sstl-Sphl insert ~hich is 1090 bp in si~e.
The construction of pGF1 (21) is depicted in Figure 4
The tendamistat activity is determined by the plate assay which is
described in Example 3 in European Patent Application 85 105 610.1 (Filed
May 8, 1985, published November 21, 1985 as No. 0 161 629, Applicant
Hoechst Aktiengesellschaft) and in Example 2 in European Patent
Application 86 113 627.3 (Filed October 2, 1986, published April 15, 1987
as No. 0 218 204, Applicant Hoechst Aktiengesellschaft).
The fusion proteln coded for by pGFl can be expressed in
- - 13 - 1338338
a known manner. ~hen the transformed strain S. lividans
TK 24 is incubated in shaken flasks at 28C for 4 days, and
the mycelium is removed from the culture solution by cen-
trifugation, the fusion protein can be detected in the
clear solution as follows:
10 to 100 ~l of solution are mixed with 20 to 200 ~l of
15% strength trichloroacetic acid, and the precipitated
protein is concentrated by centrifugation, washed and
taken up in SDS-containing sample buffer (U. Laemmli,
Nature 227 (1970) 680-685). After incubation at 90C for
2 minutesthe sample is separated electrophoretically on a
C 10to 17% strength SDS polyacrylamide gel. A proteinof molecular weight 19 kD is obtained, that is to say in
the expected molecular weight range for the fusion pro-
tein composed of tendamistat and proinsulin. The fusion
protein reacts both with antibodies against tendamistat
and with antibodies against insulin.
Tab~e 1 - 14 - 1338338
DNA sequence (coding strand) and amino acid sequence of
tendamistat
5 ' - GAC ACG ACC GTC TCC GAG CCC GCA CCC TCC TGC GTG
NH - Asp Thr Thr Val Ser Glu Pro Al a Pro Ser Cys Val
ACG CTC TAC CAG AGC TGG CGG TAC TCA CAG GCC GAC
Thr Leu Tyr Gln Ser Trp Arg Tyr Ser Gln Ala Asp
AAC GGC TGT GCC GAG ACG GTG ACC GTG AAG GTC GTC
C Asn Gly Cys Ala Glu Thr Val Thr Val Lys Val Val
15 TAC GAG GAC GAC ACC GAA GGC CTG TGC TAC GCC GTC
Tyr Glu Asp Asp Thr Glu Gly Leu Cys Tyr Ala Val
GCA CCG GGC CAG ATC ACC gCC GTC GGC GAC GGC TAC
Ala Pro Gly Gln Ile Thr Thr Val Gly Asp Gly Tyr
ATC GGC TCG CAC GGC CAC GCG CGC TAC CTG GCT CGC
Ile Gly Ser His Gly His Ala Arg Tyr Leu Ala Arg
TGC CTT TAG-3 '
25 Cys Leu Stp
- ~ - 15 - 1338338
Tab~e 2
5 ' AAT TCT GCA AGA
3 ' GA CGT TCT
( Asn ) Ser Al a Arg
(EcoRI )
B l
TTT GTG AAC CAG CAC CTG TGC GGC TCC CAC CTA GTG GAA GCT CTC
AAA CAC TTG GTC GTG GAC ACG CCG AGG GTG GAT CAC CTT CGA GAG
10 Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu
TAC CTG GTG TGC GGG GAG CGA GGC TTC TTC TAC ACA CCC AAG ACC
ATG GAC CAC ACG CCC CTC GCT CCG AAG AAG ATG TGT GGG TTC TGG
Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr
C~
C 1
2 o CGC CGG GAG GCA GAG GAC CCT CAG GTG GGG CAG GTG GAG CTG GGC
GCG GCC CTC CGT CTC CTG GGA GTC CAC CCC GTC CAC CTC GAC CCG
Arg Arg Glu Ala Glu Asp Pro Gln Val Gly Gln Val Glu Leu Gly
Z 5 GGG GGC CCT GGC GCA GGC AGC CTG CAG CCC TTG GCG CTG GAG GGG
CCC CCG GGA CCG CGT CCG TCG GAC GTC GGG AAC CGC GAC CTC CCC
Gly Gly Pro Gly Ala Gly Ser Leu Gln Pro Leu Ala Leu Glu Gly
A l
TCC CTG CAG AAG CGC GGC ATC GTG GAG CAG TGC TGC ACC AGC ATC
AGG GAC GTC TTC GCG CCG TAG CAC CTC GTC ACG ACG TGG TCG TAG
Ser Leu Gln Lys Arg Gly Ile Val Glu Gln Cys Cys Thr Ser Ile
.
~' TGC TCC CTC TAC CAG CTG GAG AAC TAC TGC AAC TAA TAG TCG ACC
ACG AGG GAG ATG GTC GAC CTC TTG ATG ACG TTG ATT ATC AGC TGG
40 Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn SalI
TGC AGC CA 3 '
ACG TCG GTT CGA 5 '
PstI (HindI I I )
B 1, C 1 and A 1 designate the starts of the B, C and A
chains of monkey proinsulin
