U.S. patent application number 09/829251 was filed with the patent office on 2003-01-23 for methods and compositions for secretion of heterologous polypeptides.
This patent application is currently assigned to GENENTECH, INC.. Invention is credited to Simmons, Laura C., Yansura, Daniel G..
Application Number | 20030017592 09/829251 |
Document ID | / |
Family ID | 23570664 |
Filed Date | 2003-01-23 |
United States Patent
Application |
20030017592 |
Kind Code |
A1 |
Simmons, Laura C. ; et
al. |
January 23, 2003 |
Methods and compositions for secretion of heterologous
polypeptides
Abstract
The instant invention discloses the unexpected result that
mutant signal sequences with reduced translational strength
provided essentially complete processing and high levels of
expression of a polypeptide of interest as compared to wild type
signal sequences, and that many mammalian polypeptides require a
narrow range of translation levels to achieve maximum secretion. A
set of signal sequence vectors provides a range of translational
strengths for optimizing expression of a polypeptide of
interest.
Inventors: |
Simmons, Laura C.;
(Burlingame, CA) ; Yansura, Daniel G.; (Pacifica,
CA) |
Correspondence
Address: |
FLEHR HOHBACH TEST ALBRITTON & HERBERT LLP
Four Embarcadero Center - Suite 3400
San Francisco
CA
94111-4187
US
|
Assignee: |
GENENTECH, INC.
|
Family ID: |
23570664 |
Appl. No.: |
09/829251 |
Filed: |
April 9, 2001 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09829251 |
Apr 9, 2001 |
|
|
|
08397303 |
Mar 1, 1995 |
|
|
|
6242177 |
|
|
|
|
Current U.S.
Class: |
435/455 ;
435/320.1; 435/69.1 |
Current CPC
Class: |
C07K 2319/75 20130101;
C07K 2319/02 20130101; C07K 2319/61 20130101; C07K 2319/036
20130101; C12N 15/70 20130101; C12N 15/625 20130101 |
Class at
Publication: |
435/455 ;
435/69.1; 435/320.1 |
International
Class: |
C12P 021/02; C12N
015/87 |
Claims
1. A method of optimizing secretion of a heterologous polypeptide
of interest in a cell comprising comparing the levels of expression
of the polypeptide under control of a set of nucleic acid variants
of a translation initiation region, wherein the set of variants
represents a range of translational strengths, and determining the
optimal translational strength for production of mature
polypeptide, wherein the optimal translational strength is less
than the translational strength of the wild-type translation
initiation region.
2. The method of claim 1, wherein the variants comprise nucleic
acid variants of a secretion signal sequence.
3. The method of claim 2, wherein the variant secretion signal
sequences are variants of STII.
4. The method of claim 3, wherein the STII variants are the
following variants
9 5' TCTAGAGGTTGAGGTGATTTT ATG AAA AAG AAT ATC GCA TTT CTT CTT GCA
TCT ATG TTC GTT TTT TCT ATT GCT ACA AAY GCS TAT GCM 3' (SEQ ID
NO:15); 5' TCTAGAATT ATG AAA AAG AAT ATC GCA TTT CTT CTT GCA TCT
ATG TTC GTT TTT TCT ATT GCT ACA AAC GCG TAT GCM 3' (SEQ ID NO:16);
5' TCTAGAATT ATG AAG AAG AAT ATT GCG TTC CTA CTT GCC TCT ATG TTT
GTC TTT TCT ATA GCT ACA AAC GCG TAT GCM 3' (SEQ ID NO:17); 5'
TCTAGAATT ATG AAG AAG AAT ATC GCA TTT CTT CTT GCA TCT ATG TTC GTT
TTT TCT ATT GCT ACA AAC GCG TAT GCM 3' (SEQ ID NO:18); 5' TCTAGAATT
ATG AAA AAA AAC ATC GCA TTT CTT CTT GCA TCT ATG TTC GTT TTT TCT ATT
GCT ACA AAC GCG TAT GCM 3' (SEQ ID NO:19); 5' TCTAGAATT ATG AAA AAA
AAC ATT GCC TTT CTT CTT GCA TCT ATG TTC GTT TTT TCT ATT GCT ACA AAC
GCG TAT GCM 3' (SEQ ID NO:20); 5' TCTAGAATT ATG AAG AAA AAC ATC GCT
TTT CTT CTT GCA TCT ATG TTC GTT TTT TCT ATT GCT ACA AAC GCG TAT GCM
3' (SEQ ID NO:21); 5' TCTAGAATT ATG AAA AAG AAC ATA GCG TTT CTT CTT
GCA TCT ATG TTC GTT TTT TCT ATT GCT ACA AAC GCG TAT GCM 3' (SEQ ID
NO:22); and 5' TCTAGAGGTTGAGGTGATTTT ATG AAA AAA AAC ATC GCA TTT
CTT CTT GCA TCT ATG TTC GTT TTT TCT ATT GCT ACA AAC GCG TAT GCM 3'
(SEQ ID NO:23).
Description
FIELD OF THE INVENTION
[0001] This invention relates to signal sequences for the secretion
of heterologous polypeptides from bacteria.
DESCRIPTION OF BACKGROUND AND RELATED ART
[0002] Secretion of heterologous polypeptides into the periplasmic
space of E. coli and other prokaryotes or into their culture media
is subject to a variety of parameters. Typically, vectors for
secretion of a polypeptide of interest are engineered to position
DNA encoding a secretory signal sequence 5' to the DNA encoding the
polypeptide of interest. Two major recurring problems plague the
secretion of such polypeptides. First, the signal sequence is often
incompletely processed or removed, and second, the amount of
polypeptide secreted is often low or undetectable. Attempts to
overcome these problems fall into three major areas: trying several
different signal sequences, mutating the amino acid sequence of the
signal sequence, and altering the secretory pathway within the host
bacterium.
[0003] A number of signal sequences are available for the first
approach in overcoming secretion problems. Watson (Nucleic Acids
Research 12: 5145-5164 (1984)) discloses a compilation of signal
sequences. U.S. Pat. No. 4,963,495 discloses the expression and
secretion of mature eukaryotic protein in the periplasmic space of
a host organism using a prokaryotic secretion signal sequence DNA
linked at its 3' end to the 5' end of the DNA encoding the mature
protein. In particular, the DNA encoding E. coli enterotoxin
signals, especially STII, are preferred. Chang et al. (Gene
55:189-196 (1987)) discloses the use of the STII signal sequence to
secrete hGH in E. coli. Gray et al. (Gene 39:247-245 (1985))
disclose the use of the natural signal sequence of human growth
hormone and the use of the E. coli alkaline phosphatase promoter
and signal sequence for the secretion of human growth hormone in E.
coli. Wong et al. (Gene 68:193-203 (1988)) disclose the secretion
of insulin-like growth factor 1 (IGF-1) fused to LamB and OmpF
secretion leader sequences in E. coli, and the enhancement of
processing efficiency of these signal sequences in the presence of
a prlA4 mutation. Fujimoto et al. (J. Biotech. 8:77-86 (1988))
disclose the use of four different E. coli enterotoxin signal
sequences, STI, STII, LT-A, and LT-B for the secretion of human
epidermal growth factor IGF) in E. coli. Denefle et al. (Gene 85:
499-510 (1989)) disclose the use of OmpA and PhoA signal peptides
for the secretion of mature human interleukin 1.beta..
[0004] Mutagenesis of the signal sequence has, in general, not been
especially helpful in overcoming secretion problems. For example,
Morioka-Fujimoto et al. (J. Biol. Chem. 266:1728-1732 (1991))
disclose amino acid changes in the LTA signal sequence that
increased the amount of human epidermal growth factor secreted in
E. coli. Goldstein et al. (J. Bact. 172:1225-1231 (1990)) disclose
amino acid substitution in the hydrophobic region of OmpA effected
secretion of nuclease A but not TEM .beta.-lactamase. Matteucci et
al. (Biotech. 4:51-55 (1986)) disclose mutations in the signal
sequence of human growth hormone that enhance secretion of hGH.
Lehnhardt et al. (J. Biol. Chem. 262:1716-1719 (1987) disclose the
effect of deletion mutations in OmpA signal peptide on secretion of
nuclease A and TEM .beta.-lactamase.
[0005] Finally, attempts at improving heterologous secretion in E.
coli by modulating host machinery has so far shown limited
improvement in overcoming secretion problems. For example, van Dijl
et al. (Mol. Gen. Genet. 227:40-48 (1991)) disclose the effects of
overproduction of the E. coli signal peptidase I (SPase I) on the
processing of precursors. Klein et al. (Protein Engineering
5:511-517 (1992) disclose that mutagenesis of the LamB signal
sequence had little effect on secretion of bovine somatotropin, and
that secretion properties of bovine somatotropin appear to be
determined by the mature protein rather than by changes in the
signal sequence. Perez-Perez et al. (Bio/Technology 12:179-180
(1994)) disclose that providing an E. coli host with additional
copies of prlA4 (secY allele) and secE genes, which encode the
major components of the "translocator", i.e., the molecular
apparatus that physically moves proteins across the membrane,
increased the ratio of mature to precursor hIL-6 from 1.2 to 10.8.
U.S. Pat. No. 5,232,840 discloses novel ribosome binding sites
useful in enhancing protein production in bacteria through enhanced
and/or more efficient translation. U.S. Pat. No. 5,082,783
discloses improved secretion of heterologous proteins by hosts such
as yeasts by using promoters of at most intermediate strength with
heterologous DNA secretion signal sequences. European Patent
Application No. 84308928.5, filed Dec. 19, 1984, discloses
promoter-ribosome binding site expression elements of general
utility for high level heterologous gene expression.
[0006] The instant invention discloses the unexpected result that
altered translation initiation regions with reduced translational
strength provided essentially complete processing and high levels
of secretion of a polypeptide of interest as compared to wild type
sequences, and that many mammalian polypeptides require a narrow
range of translation levels to achieve maximum secretion. A set of
vectors with variant translation initiation regions provides a
range of translational strengths for optimizing secretion of a
polypeptide of interest.
SUMMARY OF THE INVENTION
[0007] One aspect of the invention is a method of optimizing
secretion of a heterologous polypeptide of interest in a cell
comprising comparing the levels of expression of the polypeptide
under control of a set of nucleic acid variants of a translation
initiation region, wherein the set of variants represents a range
of translational strengths, and determining the optimal
translational strength for productions of mature polypeptide,
wherein the optimal translational strength is less than the
translational strength of the wild-type translation initiation
region.
[0008] In a further aspect of the invention the variants are signal
sequence variants, especially variants of the STII signal
sequence.
BRIEF DESCRIPTION OF THE DRAWINGS
[0009] FIG. 1 depicts the sequence of the PhoA promoter, Trp and
STII Shine-Dalgarno regions and STII signal sequence.
[0010] FIG. 2 is a diagram depicting relevant features of the
plasmid pLS33.
[0011] FIG. 3 is a diagram depicting construction of the library,
PSTIIBK.
[0012] FIG. 4 is a graph depicting comparison of the levels of
expression of IGF-1, as measured by the amount of IGF-1 detected in
culture supernatants, for pLS33, pSTIIBK#131, and PSTIIC.
Experiments 1 to 8 represent measurements taken on 8 separate
dates.
[0013] FIG. 5 is a diagram depicting construction of the plasmid
pSTIIC.
[0014] FIG. 6 is a diagram depicting construction of the plasmid
pSTIILys.
[0015] FIG. 7 is a diagram depicting construction of the plasmid
pPho21.
[0016] FIG. 8 is a diagram depicting construction of the plasmid
pPho31.
[0017] FIG. 9 is a diagram depicting construction of the plasmid
pPho41.
[0018] FIG. 10 is a diagram depicting construction of the plasmid
pPho51.
[0019] FIG. 11 is a diagram depicting relevant features of the
library, pSTIICBK.
[0020] FIG. 12 is a diagram depicting construction of the library,
pSTBKPhoA.
[0021] FIG. 13 is a graph depicting PhoA activity in isolates of
the pSTBKPhoA library.
[0022] FIG. 14 depicts the nucleotide sequences of the listed STII
signal sequence variants.
[0023] FIG. 15 is a diagram depicting construction of the plasmid
pNT3PST116.
[0024] FIG. 16 is a diagram depicting construction of the plasmid
pST116Pho.
[0025] FIG. 17 is a diagram depicting relevant features of
"category A" plasmids used in the examples.
[0026] FIG. 18 is a diagram depicting relevant features of
"category B" plasmids used in the examples.
[0027] FIG. 19 is a photograph of a Coomassie blue stained
polypeptide gel depicting secretion of mature ICAM-1 in E. coli
under control of variant STII signal sequences. The TIR of relative
strength 9 was provided by the pPho31 STII variant; the TIR of
relative strength 3 was provided by the pPho41 STII variant.
Precursor and mature forms of the polypeptide are indicated in the
figure.
[0028] FIG. 20 is a photograph of a Coomassie blue stained
polypeptide gel depicting secretion of mature NT3 in E. coli under
control of variant STII signal sequences. The TIR of relative
strength 9 was provided by the pPho31 STII variant; the TIR of
relative strength 7 was provided by the pPho21 STII variant; the
TIR of relative strength 3 was provided by the pPho41 STII variant;
the TIR of relative strength 1 was provided by the pPho51 STII
variant. The mature form of the polypeptide is indicated in the
figure.
[0029] FIG. 21 is a photograph of a Coomassie blue stained
polypeptide gel depicting secretion of mature RANTES in E. coli
under control of variant STII signal sequences. Reading from left
to right in the figure, the TIRs of relative strength 9 were
provided by the pPho31 and the pSTBKPhoA#116 STII variants; the TIR
of relative strength 7 was provided by the pPho21 STII variant; the
TIR of relative strength 4 was provided by the pSTBKPhoA#81 STI
variant; the TIR of relative strength 3 was provided by the pPho41
STII variant; the TIR of relative strength 2 was provided by the
pSTBKPhoA#107 STII variant; the TIRs of relative strength 1 were
provided by the pSTBKPhoA#86 and the pPho51 STII variants. The
mature form of the polypeptide is indicated in the figure.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0030] A. DEFINITIONS
[0031] The "translation initiation region" or TIR, as used herein
refers to a region of RNA (or its coding DNA) determining the site
and efficiency of initiation of translation of a gene of interest.
(See, for example, McCarthy et al. Trends in Genetics 6:78-85
(1990).). A TIR for a particular gene can extend beyond the
ribosome binding site (rbs) to include sequences 5' and 3' to the
rbs. The rbs is defined to include, minimally, the Shine-Dalgarno
region and the start codon, plus the bases in between, but can
include the expanse of mRNA protected from ribonuclease digestion
by bound ribosomes. Thus, a TIR can include an untranslated leader
or the end of an upstream cistron, and thus a translational stop
codon.
[0032] A "secretion signal sequence" or signal sequence as used
herein refers to a sequence present at the amino terminus of a
polypeptide that directs its translocation across a membrane.
Typically, a precursor polypeptide is processed by cleavage of the
signal sequence to generate mature polypeptide.
[0033] The term "translational strength" as used herein refers to a
measurement of a secreted polypeptide in a control system wherein
one or more variants of a TIR is used to direct secretion of a
polypeptide encoded by a reporter gene and the results compared to
the wild-type TIR or some other control under the same culture and
assay conditions. For example, in these experiments translational
strength is measured by using alkaline phosphatase as the reporter
gene expressed under basal level control of the PhoA promoter,
wherein secretion of the PhoA polypeptide is directed by variants
of the STII signal sequence. The amount of mature alkaline
phosphatase present in the host is a measure of the amount of
polypeptide secreted, and can be quantitated relative to a negative
control. Without being limited to any one theory, "translational
strength" as used herein can thus include, for example, a measure
of mRNA stability, efficiency of ribosome binding to the ribosome
binding site, and mode of translocation across a membrane.
[0034] "Polypeptide" as used herein refers generally to peptides
and polypeptides having at least about two amino acids.
[0035] B. GENERAL METHODS
[0036] The instant invention demonstrates that translational
strength is a critical factor in determining whether many
heterologous polypeptides are secreted in significant quantities.
Thus, for a given TIR, a series of amino acid or nucleic acid
sequence variants can be created with a range of translational
strengths, thereby providing a convenient means by which to adjust
this factor for the optimal secretion of many different
polypeptides. The use of a reporter gene expressed under the
control of these variants, such as PhoA, provides a method to
quantitate the relative translational strengths of different
translation initiation regions. The variant or mutant TIRs can be
provided in the background of a plasmid vector, thereby providing a
set of plasmids into which a gene of interest may be inserted and
its expression measured, so as to establish an optimum range of
translational strengths for maximal expression of mature
polypeptide.
[0037] Thus, for example, signal sequences from any prokaryotic or
eukaryotic organism may be used. Preferably, the signal sequence is
STII, OmpA, PhoE, LamB, MBP, or PhoA.
[0038] Mutagenesis of the TIR is done by conventional techniques
that result in codon changes which can alter the amino acid
sequence, although silent changes in the nucleotide sequence are
preferred. Alterations in the TIR can include, for example,
alterations in the number or spacing of Shine-Dalgarno sequences,
along with alterations in the signal sequence. One preferred method
for generating mutant signal sequences is the generation of a
"codon bank" at the beginning of a coding sequence that does not
change the amino acid sequence of the signal sequence (i.e., the
changes are silent). This can be accomplished by changing the third
nucleotide position of each codon; additionally, some amino acids,
such as leucine, serine, and arginine, have multiple first and
second positions that can add complexity in making the bank. This
method of mutagenesis is described in detail in Yansura et al.
(METHODS: A Companion to Methods in Enzymol. 4:151-158 (1992)).
Basically, a DNA fragment encoding the signal sequence and the
beginning of the mature polypeptide is synthesized such that the
third (and, possibly, the first and second, as described above)
position of each of the first 6 to 12 codons is altered. The
additional nucleotides downstream of these codons provide a site
for the binding of a complementary primer used in making the bottom
strand. Treatment of the top coding strand and the bottom strand
primer with DNA polymerase I (Klenow) will result in a set of
duplex DNA fragments containing randomized codons. The primers are
designed to contain useful cloning sites that can then be used to
insert the DNA fragments in an appropriate vector, thereby allowing
amplification of the codon bank. Alternative methods include, for
example, replacement of the entire rbs with random nucleotides
(Wilson et al., BioTechniques 17:944-952 (1994)), and the use of
phage display libraries (see, for example, Barbas et al., Proc.
Natl. Acad. Sci. U.S.A. 89:4457-4461 (1992); Garrard et al., Gene
128:103-109 (1993)).
[0039] Typically, the TIR variants will be provided in a plasmid
vector with appropriate elements for expression of a gene of
interest. For example, a typical construct will contain a promoter
5' to the signal sequence, a restriction enzyme recognition site 3'
to the signal sequence for insertion of a gene of interest or a
reporter gene, and a selectable marker, such as a drug resistance
marker, for selection and/or maintenance of bacteria transformed
with the resulting plasmids.
[0040] Promoters suitable for use with prokaryotic hosts include
the .beta.-lactamase and lactose promoter systems (Chang et al.,
Nature 275:617-624 (1978); and Goeddel et al., Nature 281:544-548
(1979)), alkaline phosphatase, a tryptophan (Trp) promoter system
(Goeddel, Nucleic Acids Res. 8(18):4057-4074 (1980) and EP 36,776)
and hybrid promoters such as the tac promoter (deBoer et al., Proc.
Natl. Acad. Sci. U.S.A. 80:21-25 (1983).
[0041] Suitable promoting sequences for use with yeast hosts
include the promoters for 3-phosphoglycerate kinase (Hitzeman et
al., J. Biol. Chem. 255(24):12073-80 (1980)) or other glycolytic
enzymes (Hess et al., J. Adv. Enzyme Rea. 7:149-67 (1968)); and
Holland, Biochemistry 17:4900-4907 (1978)), such as enolase,
glyceraldehyde-3-phosphate dehydrogenase, hexokinase, pyruvate
decarboxylase, phospho-fructokinase, glucose-6-phosphate isomerase,
3-phosphoglycerate mutase, pyruvate kinase, triosephosphate
isomerase, phosphoglucose isomerase, and glucokinase.
[0042] Other yeast promoters, which are inducible promoters having
the additional advantage of transcription controlled by growth
conditions, are the promoter regions for alcohol dehydrogenase 2,
isocytochrome C, acid phosphatase, degradative enzymes associated
with nitrogen metabolism, metallothionein,
glyceraldehyde-3-phosphate dehydrogenase, and enzymes responsible
for maltose and galactose utilization. Suitable vectors and
promoters for use in yeast expression are further described in
Hitzeman et al., EP 73,657A. Yeast enhancers also are
advantageously used with yeast promoters.
[0043] Any reporter gene may be used which can be quantified in
some manner. Thus, for example, alkaline phosphatase production can
be quantitated as a measure of the secreted level of the phoA gene
product. Other examples include, for example, the .beta.-lactamase
genes.
[0044] Preferably, a set of vectors is generated with a range of
translational strengths into which DNA encoding a polypeptide of
interest may be inserted. This limited set provides a comparison of
secreted levels of polypeptides. The secreted level of polypeptides
can be determined, for example, by a functional assays for the
polypeptide of interest, if available, radioimmunoassays (RIA),
enzyme-linked immunoassays (ELISA), or by PAGE and visualization of
the correct molecular weight of the polypeptide of interest.
Vectors so constructed can be used to transform an appropriate
host. Preferably, the host is a prokaryotic host. More preferably,
the host is E. coli.
[0045] Further details of the invention can be found in the
following examples, which further define the scope of the
invention. All references cited herein are expressly incorporated
by reference in their entirety.
EXAMPLES
[0046] I. Plasmid Constructs
[0047] A. Basic Plasmid Construction
[0048] All of the plasmids described in this patent application
were constructed from a basic backbone of pBR322 (Sutcliffe, Cold
Spring Harb Symp Quant Biol 43:77-90 (1978)). While the gene of
interest expressed in each case varies, the transcriptional and
translational sequences required for the expression of each gene
were provided by the PhoA promoter and the Trp Shine-Dalgarno
sequence (Chang et al., Gene 55:189-196 (1987)). Additionally, in
the cases noted, a second Shine-Dalgarno sequence, the STII
Shine-Dalgarno sequence (Picken et al., Infect Immun 42(l):269-275
(1983)), was also be present. Secretion of the polypeptide was
directed by the STII signal sequence or variants thereof (Picken et
al., Infect Immun 42(1):269-275 (1983)). The PhoA promoter, Trp and
STII Shine-Dalgarno sequences and the sequence of the wild-type
STII signal sequence are given in FIG. 1.
[0049] B. Construction of pLS33
[0050] The plasmid pLS33 was derived from phGH1 (Chang et al., Gene
55:189-196 (1987)), which was constructed for the expression of
des(1,3)-IGF-I. In the plasmid pLS33, the gene encoding this
version of insulin-like growth factor I (altered from the original
sequence (Elmblad et al., Third European Congress on Biotechnology
III. Weinheim: Verlag Chemie, pp.287-292 (1984)) by the removal of
the first three amino acids at the N-terminus) replaced the gene
encoding human growth hormone. The construction pLS33 maintained
the sequences for the PhoA promoter, Trp and STII Shine-Dalgarno
regions and the wild-type STII signal sequence described for phGH1.
However, the 3' end following the termination codon for
des(1,3)-IGF-I was altered from that described for phGH1. In the
case of pLS33, immediately downstream of the termination codon a
HindIII restriction site was engineered, followed by the methionine
start codon of the tetracycline resistance gene of pBR322
(Sutcliffe, Cold Spring Harb Symp Quant Biol 43:77-90 (1978)). A
diagram of the plasmid pLS33 is given in FIG. 2.
[0051] C. Construction of pSTIIBK
[0052] A plasmid library containing a variable codon bank of the
STII signal sequence (pSTIIBK) was constructed to screen for
improved nucleotide sequences of this signal. The vector fragment
for the construction of PSTIIBK was created by isolating the
largest fragment when pLS33 was digested with XbaI and BstEII. This
vector fragment contains the sequences that encode the PhoA
promoter, Trp Shine-Dalgarno sequence and amino acids 16-67 of
des(1,3)-IGF-I. The coding region for amino acids 3-15 of
des(1,3)-IGF-I was provided by isolating the DraIII-BstEII fragment
(approximately 45 bp) from another IGF-I expression plasmid,
pLS331amB. The variations in the nucleotide sequence for the STII
signal were derived from the two strands of synthetic DNA listed
below:
1 5'- GCATGTCTAGAATT ATG AAR AAR AAY ATH GCN TTY CTN CTN GCN TCN
ATG TTY (SEQ ID NO:1) GTN TTY TCN ATH GCT ACA AAC GCG TAT GCC ACTCT
- 3' 3'- CGA TGT TTG CGC ATA CGG TGAGACACGCCACGACTT - 5' (SEQ ID
NO:2) R: A, G Y: T, C H: A, T, C N: G, A, T, C
[0053] These two strands of synthetic DNA were annealed and treated
with DNA Polymerase I (Klenow Fragment) to form duplex DNA of
approximately 101 bp. This duplex DNA was then digested with XbaI
and DraIII to generate the fragment of approximately 82 bp encoding
the STII signal sequence with variable codons and the first two
amino acids of des(1,3)-IGF-I. These fragments were then ligated
together as shown in FIG. 3 to construct the library, pSTIIBK.
[0054] D. Selection of DSTIIBK#131
[0055] The plasmid library containing a variable codon bank of the
STII signal sequence (PSTIIBK) was screened for improved growth of
transformants and increased secretion of IGF-1. Basically, plasmids
were transformed into host strain 27C7 (see below) and screened for
enhanced ability to grow in a low phosphate medium (see Chang et
al., supra) plus carbenicillin (50 .mu.g/ml) based on OD.sub.600
measurements of cell density. Candidate colonies were tested for
increased levels of IGF-1 secretion as follows. Colonies were
inoculated into 3-5 ml LB plus carbenicillin (50 .mu.g/ml) and
grown at 37.degree. C. with shaking for about 5-15 hours. Cultures
were diluted 1:100 into 1-3 ml low phosphate medium plus
Carbenicillin (50 .mu.g/ml) and induced for 24 hours shaking at
37.degree. C. The induced cultures were centrifuged in
microcentrifuge tubes for 5 minutes. Supernatants were diluted into
IGF RIA diluent and stored at -20.degree. C. until assayed. The
amount of IGF-1 secreted into the medium was measured by a
radioimmunoassay.
[0056] The level of expression of IGF-1, as measured by the amount
of IGF-1 detected in culture supernatants, was compared for pLS33,
pSTIIBK#131, and PSTIIC, in FIG. 4. The variant #131 consistently
improved IGF-1 expression over the "original" or wild-type STII
signal sequence. PSTIIC showed some slight improvement in
expression over the wild-type sequence. pSTIIBK#131 differed from
the wild-type STII in 12 codons and in the deletion of one
Shine-Dalgarno sequence. PSTIIC was constructed as described below
as a control plasmid having only one Shine-Dalgarno sequence and
three codon changes near the extreme 3' end of the signal.
[0057] E. Construction of pSTIIC
[0058] In pSTIIC the STII Shine-Dalgarno sequence was removed from
the plasmid pLS33. In addition, by incorporating silent mutations
near the 3' end of the STII signal, an MluI site was engineered
into PSTIIC. The identical fragments described for the construction
of PSTIIBK (the vector from pLS33 and the approximately 45 bp
DraIII-BstEII fragment from pLS331amB) were used for the
construction of this plasmid. However, the synthetic DNA differed
from that described above for the construction of PSTIIBK. For the
construction of pSTIIC, the synthetic DNA coding for the STII
signal sequence and the first two amino acids of des(1,3)-IGF-I was
as follows:
2 5'- CTAGAATT ATG AAA AAG AAT ATC GCA TTT CTT CTT GCA TCT ATG TTC
GTT (SEQ ID NO:3) 3'- TTAA TAC TTT TTC TTA TAG CGT AAA GAA GAA CGT
AGA TAC AAG CAA (SEQ ID NO:4) .sub.----MluI.sub.---- TTT TCT ATT
GCT ACA AAC GCG TAT CCC ACTCT - 3' AAA AGA TAA CGA TGT TTG CGC ATA
CGG TG - 5'
[0059] These fragments were ligated together as illustrated in FIG.
5 to construct the plasmid PSTIIC.
[0060] F. Construction of pSTIILys
[0061] The plasmid pSTIILys contained an STII signal sequence that
differs from the signal sequence of PSTIIC by only one nucleotide
change at the position of the second codon. This signal sequence
was constructed from synthetic DNA and placed in a pBR322-based
vector for the expression of the polypeptide RANTES (Schall et al.,
J Immunol 141(3):1018-1025 (1988)). The XbaI-MluI vector fragment
for this construction was isolated from the plasmid pBK131Ran (a
derivative of the plasmid pSTIIBK#131 with the gene encoding RANTES
replacing the gene encoding des(1,3)-IGF-I). This vector contained
the PhoA promoter, Trp Shine-Dalgarno sequence, the last three
amino acids of the STIIC signal sequence and the gene encoding the
polypeptide RANTES. As illustrated in FIG. 6, this fragment was
then ligated with the following strands of synthetic DNA to
construct the plasmid pSTIILys (SEQ ID NO:3):
3 5'- CTAGAATT ATG AAG AAG AAT ATC GCA TTT CTT CTT GCA TCT ATG TTC
GTT (SEQ ID NO:5) 3'- TTAA TAC TTC TTC TTA TAG CGT AAA GAA GAA CGT
AGA TAG AAG CAA (SEQ ID NO:6) TTT TCT ATT GCT ACA AA - 3' AAA AGA
TAA CGA TGT TTG CGC - 5'
[0062] G. Construction of Alkaline Phosphatase Plasmids
[0063] In order to determine a quantitative TIR value for each of
the STII signal sequences described, the alkaline phosphatase gene
of E. coli was used as a reporter gene. In each of these
constructions, the PhoA gene was placed downstream of the PhoA
promoter, Trp Shine-Dalgarno sequence and a version of the STII
signal sequence. The plasmids pPho21, pPho31, pPho41 and pPho51
contained the signal sequences derived from PSTIIC, pLS33,
pSTIIBK#131 and pSTIILys, respectively. In the case of pPho31, the
construction also contained the STII Shine-Dalgarno region.
[0064] H. Construction of pPho21
[0065] The vector fragment for the construction of pPho21 was
created by digesting pBR322 with EcoRI and BamHI and isolating the
largest fragment. The PhoA promoter, Trp Shine-Dalgarno sequence
and STIIC signal sequence (amino acids 1-20) were provided by
isolating the approximately 484 bp fragment of pCN131Tsc following
digestion with EcoRI and MluI. An identical fragment of
approximately 484 bp could have also been generated from PSTIIC, a
plasmid which has been described previously. The PhoA gene fragment
(approximately 1430 bp) encoding amino acids 24-450 of alkaline
phosphatase was generated from the plasmid pb0525 following
digestion with Bsp1286 and BamHI (Inouye et al., J. Bacteriol
146(2):668-675 (1981)). This Bsp1286-BamHI fragment also contains
approximately 142 bp of SV40 DNA (Fiers et al., Nature 273:113-120
(1978)) following the termination codon of alkaline phosphatase.
Synthetic DNA was used to link the STII signal sequence with the
PhoA gene. The sequence of this DNA encoding the last three amino
acids of the STIIC signal sequence and amino acids 1-23 of alkaline
phosphatase was as follows:
4 5'-
CGCGTATGCCCGGACACCAGAAATGCCTGTTCTGGAAAACCGGGCTGCTCAGGGCGATATT- ACTG
(SEQ ID NO:7) 3'- ATACGGGCCTGTGGTCTTTACGGACAAGACCTTTTGGCC-
CGACGAGTCCCGCTATAATGAC (SEQ ID NO:8) CACCCGGCGGTGCT - 3' GTGGGCCGCC
- 5'
[0066] In order to facilitate the construction of this plasmid, the
synthetic DNA was preligated to the EcoRI-MluI fragment of
pCN131Tsc. This preligation generated a new fragment of about 575
bp. As illustrated in FIG. 7, the fragment generated from the
preligation was then ligated together with the other fragments
described to construct pPho21.
[0067] I. Construction of pPho31
[0068] The vector fragment for the construction of this plasmid was
the identical vector described for pPho21. The PhoA promoter, Trp
Shine-Dalgarno sequence, STII Shine-Dalgarno sequence and STII
signal sequence (amino acids 1-20) were generated from pJAL55. The
necessary fragment (approximately 496 bp) from pJAL55 was isolated
following digestion with EcoRI and MluI. This EcoRI-MluI fragment
only differed from the same region of pLS33 by an engineered MluI
site starting at amino acid 20 of the STII signal sequence (as
described for pSTIIC). The last three amino acids of the STIIC
signal sequence and the sequence encoding the PhoA gene were
provided by digesting the plasmid pPho21 with MluI and BamHI and
isolating the approximately 1505 bp fragment. These fragments were
ligated together as shown in FIG. 8 to yield pPho31.
[0069] J. Construction of pPho41
[0070] The vector fragment for the construction of this plasmid was
the identical vector described for pPho21. The PhoA promoter, Trp
Shine-Dalgarno sequence and STII signal sequence with pSTIIBK#131
codons (amino acids 1-20) were provided by isolating the
approximately 484 bp EcoRI-MluI fragment of pNGF131. An identical
fragment could have also been generated from pSTIIBK#131. The last
three amino acids of the STIIC signal sequence and the sequence
encoding the PhoA gene were provided by digesting the plasmid
pPho21 with MluI and BamHI and isolating the approximately 1505 bp
fragment. As illustrated in FIG. 9, these three fragments were then
ligated together to construct pPho41.
[0071] K. Construction of pPho51
[0072] The vector fragment for the construction of pPho51 was
generated by digesting the plasmid pLS18 with XbaI-BamHI and
isolating the largest fragment. The plasmid pLS18 is a derivative
of phGH1 (Chang et al., Gene 55:189-196 (1987)) and an identical
vector would have been generated had phGH1 been used in place of
pLS18. This XbaI-BamHI vector contains the PhoA promoter and the
Trp Shine-Dalgarno sequence. The STII signal sequence (amino acids
1-20) with pSTIILys codons was provided by isolating the
approximately 67 bp fragment generated when pSTIILys was digested
with XbaI and MluI. The last three amino acids of the STIIC signal
sequence and the sequence encoding the PhoA gene were provided by
digesting the plasmid pPho21 with MluI and BamHI and isolating the
approximately 1505 bp fragment. A diagram for the construction of
pPho51 is given in FIG. 10.
[0073] L. Construction of pSTIICBK
[0074] A second variable codon library of the STII signal sequence,
pSTIICBK, was constructed. This second codon library was designed
only to focus on the codons closest to the met initiation codon of
the STII signal sequence. As illustrated in FIG. 11, PSTIICBK was a
pBR322-based plasmid containing the gene encoding the polypeptide
RANTES (Schall et al., J Immunol 141(3):1018-1025 (1988)) under the
control of the PhoA promoter and the Trp Shine-Dalgarno sequence.
In this plasmid, secretion of RANTES is directed by an STII signal
sequence codon library derived from the following two strands of
synthetic DNA:
5 5'- GCATGTCTAGAATT ATG AAR AAR AAY ATH GCN TTT CTT CTT GCA TCT
ATG TTC (SEQ ID NO:9) GTT TTT TCT ATT GCT ACA AAC GCG TAT GCC - 3'
3'- AGA TAA CGA TGT TTG CGC ATA CGG TGA - 5' (SEQ ID NO:10) R: A, G
Y: T, C H: A, T, C N: G, A, T, C
[0075] These two strands of synthetic DNA were annealed and treated
with DNA Polymerase I (Klenow Fragment) to form duplex DNA of
approximately 86 bp. This duplex DNA was then digested with XbaI
and MluI to generate a fragment of approximately 67 bp encoding the
first 20 amino acids of the STII signal sequence with variable
codons at positions 2-6.
[0076] M. Construction of pSTBKPhoA
[0077] To increase the number of STII signal sequences available
with differing relative TIR strengths, a convenient method of
screening the codon library of PSTIICBK was required. The plasmid
pSTBKPhoA was constructed as a solution to this problem. In the
plasmid pSTBKPhoA, the STII codon library of PSTIICBK was inserted
upstream of the PhoA gene and downstream of the PhoA promoter and
the Trp Shine-Dalgarno sequence. PhoA activity thus provided a
means by which to discriminate between different versions of the
STII signal sequences.
[0078] The vector fragment for this construction was created by
isolating the largest fragment when p131TGF was digested with XbaI
and BamHI. An identical vector could have also been generated from
phGH1 (Chang et al., Gene 55:189-196 (1987)). This vector contained
the PhoA promoter and the Trp Shine-Dalgarno sequence. The codon
library of the STII signal sequence was provided by isolating the
approximately 67 bp fragment generated from pSTIICBK following
digestion with XbaI and MluI. The last three amino acids of the
STIIC signal sequence and the sequence encoding the PhoA gene were
provided by digesting pPho21 with MluI and BamHI and isolating the
approximately 1505 bp fragment. As illustrated in FIG. 12, the
fragments were then ligated together to construct pSTBKPhoA.
[0079] N. Selection of pSTBKPhoA #81. 86, 107. 116
[0080] The plasmids pSTBKPhoA #81, 86, 107, 116 were selected from
the codon library of pSTBKPhoA based on their basal level PhoA
activity (FIG. 13). As listed in FIG. 14, each had a different
nucleotide sequence encoding the STII signal sequence.
[0081] O. Construction of pST116Pho
[0082] This version of the STII signal sequence, ST116, combined
the double Shine-Dalgarno sequence described by Chang et al. (Gene
55:189-196 (1987)) with the codons of the selected STII sequence
pSTBKPhoA #116. This signal sequence was initially constructed in a
plasmid designed for the secretion of the pro region of NT3
(pNT3PST116) and then was transferred into a plasmid containing the
PhoA gene to obtain a relative TIR measurement (pST116Pho).
[0083] P. Construction of DNT3PST116
[0084] The vector for this construction was generated by digesting
the plasmid pLS18 with XbaI and BamHI and isolating the largest
fragment. The plasmid pLS18 was a derivative of phGH1 (Chang et
al., Gene 55:189-196 (1987)) and an identical vector could have
been generated from phGH1. This XbaI-BamHI vector contained the
PhoA promoter and the Trp Shine-Dalgarno sequence. A fragment
(approximately 682 bp) containing the last three amino acids of the
STII signal sequence and the coding region for amino acids 19-138
of proNT3 (Jones et al., Proc Natl Acad Sci 87:8060-8064 (1990))
was generated from the plasmid pNT3P following digestion with MluI
and BamHI. The plasmid pNT3P was a pBR322-based plasmid containing
the PhoA promoter, STIIBK#131 version of the STII signal sequence
and the coding region for amino acids 19-138 of proNT3. The strands
of synthetic DNA listed below provided the sequence for the STII
Shine-Dalgarno sequence and the first 20 amino acids of the STII
signal sequence:
6 5'- CTAGAGGTTGAGGTGATTTT ATG AAA AAA AAC ATC GCA TTT CTT CTT GCA
TCT (SEQ ID NO:11) 3'- TCCAACTCCACTAAAA TAC TTT TTT TTG TAG CGT AAA
GAA GAA CGT AGA (SEQ ID NO:12) ATG TTC GTT TTT TCT ATT GCT ACA AA -
3' TAC AAG CAA AAA AGA TAA CGA TGT TTG CGC - 5'
[0085] These fragments were then ligated together as shown in FIG.
15 to construct pNT3PST116.
[0086] Q. Construction of ST116Pho
[0087] The vector for the construction of this plasmid was the
identical vector described for the construction of pNT3PST116. The
STII Shine-Dalgarno sequence and the first 20 amino acids of the
STII signal sequence (pSTBKPhoA#116 codons) were generated by
isolating the approximately 79 bp fragment from pNT3PST116
following digestion with XbaI and MluI. The last three amino acids
of the STIIC signal sequence and the sequence encoding the PhoA
gene were isolated from pSTBKPhoA#116 following digestion with MluI
and BamHI (approximately 1505 bp fragment). As illustrated in FIG.
16, ligation of these three fragments resulted in the construction
of pST116Pho.
[0088] II. Alkaline Phosphatase Assay
[0089] In these experiments the altered TIR constructs utilizing
the phoA reporter gene were assayed for relative translational
strengths by a modification of the method of Amenura et al. (J.
Bacteriol. 152:692-701, 1982). Basically, the method used was as
follows. Plasmids carrying altered sequences, whether in the TIR,
the Shine-Dalgarno region, the nucleotide sequence between the
Shine Dalgarno region and the start codon of the signal sequence,
or the signal sequence itself, whether amino acid sequence variants
or nucleotide sequence variants, were used to transform E. coli
strain 27C7 (ATCC 55,244) although any PhoA.sup.- strain of E. coli
could be used. Transformant colonies were inoculated into
Luria-Bertani medium (LB) plus carbenicillin (50 .mu.g/ml, Sigma,
Inc.). Cultures were grown at 37.degree. C. with shaking for 4-8
hr. The equivalent of OD.sub.600 of each culture was centrifuged,
then resuspended in 1 ml strict AP media (0.4% glucose, 20 mM
NH4Cl, 1.6 mM MgSO.sub.4, 50 mM KCl, 20 mM NaCl, 120 mM
triethanolamine, pH 7.4) plus carbenicillin (50 .mu./ml). The
mixtures were then immediately placed at -20.degree. C. overnight.
After thawing, 1 drop toluene was added to 1 ml of thawed culture.
After vortexing, the mixtures were transferred to 16.times.125 mm
test tubes and aerated on a wheel at 37.degree. C. for 1 hr. 40
.mu.l of each toluene treated culture was then added to 1 ml 1 M
Tris-HCl pH 8 plus 1 mM PNPP (disodium 4-nitrophenyl phosphate
hexahydrate) and left at room temperature for 1 hr. The reactions
were stopped by adding 100 ml 1 M sodium phosphate pH 6.5. The
OD.sub.410 was measured within 30 minutes. Enzyme activity was
calculated as micromoles of p-nitrophenol liberated per minute per
one OD.sub.600 equivalent of cells.
[0090] The results are summarized in Table 1.
7TABLE 1 Determination of TIR Relative Strength: Use of PhoA as a
Reporter Gene Standard Relative TIR PhoA Activity.sup.1 Deviation
Strength pBR322 0.0279 0.0069 -- pPho51.sup.2 0.0858 0.0165 1
pSTBKPhoA#86 0.1125 0.0246 1 pSTBKPhoA#107 0.1510 0.0267 2
pPho41.sup.3 0.1986 0.0556 3 pSTBKPhoA#81 0.2796 0.0813 4
pPho21.sup.4 0.4174 0.1145 7 pSTBKPhoA#116 0.5314 0.1478 9
pPho31.sup.5 0.5396 0.0869 9 pST116Pho 0.7760 0.1272 13
.sup.1micromoles of p-nitrophenol/min/O.D..sub.600 cells .sup.2same
STII variant as pSTIILys .sup.3same STII variant as pSTIIBK#131
.sup.4same STII variant as pSTIIC .sup.5wild-type STII + MluI site,
last codon GCC.
[0091] III. Secretion of Heterologous Polypeptide Examples
[0092] The plasmids used in these examples were all very similar in
design as described above. Rather than describe in detail each
construction, the expression plasmids are described here in general
terms. Although a different polypeptide of interest was expressed
in each example, the only significant variation between these
constructions was the nucleotide sequence following the 3' end of
each coding region. Thus, for descriptive purposes, these plasmids
were loosely grouped into the following two categories based on
their 3' sequence:
[0093] Category A: Within about 25 bp 3' to the termination codon
of each gene of interest began the sequence encoding the
transcriptional terminator described by Scholtissek and Grosse
(Nucleic Acids Res. 15(7):3185 (1987)) followed by the tetracycline
resistance gene of pBR322 (Sutcliffe, Cold Spring Harb Symp Quant
Biol 43:77-90 (1978)). Examples in this category included plasmids
designed for the secretion of mature NGF (Ullrich et al., Nature
303:821-825 (1983)), mature TGF-.beta.1 (Derynck et al., Nature
316:701-705 (1985)) and domains 1 and 2 of ICAM-1 (Staunton et al.,
Cell 52:925-933 (1988)). A schematic representation of these
plasmids is given in FIG. 17.
[0094] Category B: Examples in this category included plasmids
designed for the secretion of mature VEGF (Leung et al., Science
246:1306-1309 (1989)), mature NT3 (Jones et al., Proc. Natl. Acad.
Sci. U.S.A. 87:8060-8064 (1990), RANTES (Schall et al., J Immunol
141(3):1018-1025 (1988)), and PhoA. The termination codon in each
of these plasmids is followed in the 3' direction by a segment of
untranslated DNA (VEGF: approximately 43 bp; mature NT3:
approximately 134 bp; RANTES: approximately 7 bp; PhoA:
approximately 142 bp). Following this 3' untranslated region, the
sequence of pBR322 was re-initiated beginning with either the
HindIII site (as in the mature NT3 secretion plasmid) or the BamHI
site (PhoA, VEGF, RANTES secretion plasmids). A schematic
representation of the plasmids included in this category is
illustrated in FIG. 18.
[0095] These plasmids were used to transform the host E. coli
strain 27C7. Transformant colonies were inoculated into 3-5 ml
LB+carbenicillin (50 .mu.g/ml). The cultures were grown at
37.degree. C. with shaking for 3-8 hours. The cultures were then
diluted 1:100 into 3 ml low phosphate medium (Chang et al., supra)
and grown for about 20 hours with shaking at 37.degree. C. For each
culture, an 0.5 OD.sub.600 aliquot was centrifuged in a microfuge
tube.
[0096] Each 0.5 OD.sub.600 pellet was then prepared for gel
analysis as follows. Each pellet was resuspended in 50 .mu.l TE (10
mM Tris pH7.6, 1 mM EDTA). After the addition of 10 .mu.l 10% SDS,
5 .mu.l reducing agent (1M dithiothreitol or 1M
.beta.-mercaptoethanol), the samples were heated at about
90.degree. C. for 2 minutes and then vortexed. Samples were allowed
to cool to room temperature, after which 500 .mu.l acetone was
added. The sample were vortexed and then left at room temperature
for about 15 minutes. Samples were centrifuged for 5 minutes. The
supernatants were discarded, and the pellets resuspended in 20
.mu.l water, 5 .mu.l reducing agent, 25 .mu.l NOVEX 2.times. sample
buffer. Samples were heated at about 90.degree. C. for 3-5 minutes,
then vortexed. After centrifugation for 5 minutes supernatants were
transferred to clean tubes and the pellets discarded. 5-10 .mu.l of
each sample was loaded onto 10 well, 1,0 mm NOVEX manufactured gel
(San Diego Calif.) and electrophoresed for 1.5-2 hr at 120 volts.
Gels were stained with Coomassie blue to visualize polypetide
(FIGS. 19-20).
[0097] To provide further quantitation of the results, some gels
were analyzed by densitometry. These results are displayed in Table
2 below. Both the polypeptide gels and the densitometry results
indicate that the heterologous polypeptides tested were
consistently secret more efficiently when an STII variant of
reduced translation strength was used to direct secretion of that
polypeptide.
8TABLE 2 Examples of Improved Polypeptide Secretion By TIR
Modification: Densitometer Scans of Polypeptide Gels Amount
Secreted TIR (Relative (% total host Polypeptide Strength)
polypeptide) VEGF 9 0.6 3 5.9 NGF 9 1.6 7 1.8 4 5.7 1 5.5 RANTES 9
0.3 9 0.2 7 0.4 4 3.9 3 3.6 2 3.5 1* 1.6 1 1.7 TGF-.beta.1 7 1.7 3
9.2 *pSTBKPhoA#86 signal sequence
[0098]
Sequence CWU 1
1
* * * * *