Note: Descriptions are shown in the official language in which they were submitted.
~3 4~~ 1
BEHRINGWERKE AKTIf_NGESELLSCHAFT HOE 86/B 017J Dr. KL/gm
Preparation of factor XIIIa by gene manipulation
Coagulation factor XIII is the final member of the "co-
agulation cascade" in the natural process of blood co-
agulation in vertE~brates. The enzymatically active form
of factor XIII, factor XIIIa, also called "activated
fibrin-stabilizinc3 factor", "fibrinoligase" or "plasma
transglutaminase" and, hereinafter, "F XIIIa", catalyzes
the fusion of fibrin units in preexistent thrombi by
intramolecular crosslinking (Lorand et al., Methods in
Enzymology 80 (1981), 333-341; Curtis et al., Annals New
York Academy of Sciences 1983, 567-576). The molecular
weight of factor XIII from plasma is about 300 kD (Loewy
et al., J. Biol. them. 236 (1961) 2634>. The molecular
weight of the active subunit F XIIIa is about 80 kD
(Bohn and Schwick,. Arzneimittelforschung 21 (1971) 1432).
During the activation of factor XIII, thrombin splits
off from the precursor a peptide which is about 4 kD in
size and has a known sequence of 36 amino acids (Takagi
and Doolittle, Biochemistry 13 (1974) 750-756). In ad-
dition, a sequence embracing four amino acids is known
(Holbrook et al., Biochem. J. 135 (1973) 901-903).
The invention relates to a process for the preparation
of F XIIIa by gens~ manipulation, to the mRNA necessary
for this, to the c;DNA obtained therefrom, to DNA struc-
tures and vectors containing all or part of this cDNA,
to cells transformed with DNA of this type, and to the
polypeptide expressed by these cells. The invention
also relates to part-sequences of the amino acid sequence
of F XIIIa, to specific antibodies obtained therewith,
to diagnostic aids and antibody columns produced from
these antibodies, and to a polypeptide obtained with the
aid of such columnis. Another aspect of the invention
relates to diagnostic aids which contain all or part of
the DNA or RNA coding for F XIIIa, and to diagnostic
methods with which body fluids and tissues are examined
r j ~ '~ ~ ~ ti
- 2 -
using diagnostic aids of this type. Further aspects of
the invention and its preferred embodiments are illus-
trated in detail hereinafter and defined in the patent
claims.
The drawings, in which the numbers coincide with those
in the examples, illustrate the invention:
Fig. 1 shows the cDNA coding for F XIIIa (the coding re-
gion being shaded) and, below this, the DNA regions of
the isolated and characterized clones.
Fig. 2 shows the construction of the expression plasmid
pFXIII-13. For clarity, in this figure the starting
plasmids pIC19H-12.1 and pIC19H-11.1, as well as the DNA
fragments located immediately below them, are represen-
ted by double lines, as is the product pFXIII-13 con-
structed from the single-stranded fragments.
Fig. 3 is a diagram of the construction of the plasmid
pTrc97A, Fig. 3a that of pFXIII-C4 from pTrc97A and
pFXIII-13, and finally Fig. 3b the construction of
pMB259 from pFXIII-13 and the known plasmids pIC20H and
pBD2.
Fig. 4 is a diagram of the construction of the plasmid
pM8240 from pFXIII-13 and the known plasmid pAAHS.
Fig. 5 shows the .construction of pZET4 from the known
plasmid pSV2dhfr .and the plasmid pSVA STOP1, Fig. 5a
shows the construction of pSVF13 from pSVA STOP1 and
pFXIII-13, Fig. 5b shows the construction of pZF13 from
pZET4 and pFXIII-13, and finally Fig. 5c shows the con-
struction of pHSF13 from pSVF13 and the known plasmid
pSPbHS9.
The amino acid sequence of F XIIIa fragments was deter-
mined for the construction of suitable probes. The cor-
responding peptide fragments were obtained by proteoly-
sis or cleavage with cyanogen bromide. Based on know-
ledge of the amino acid sequences of such fragments, two
oligonucleotide probes were synthesized, one 20mer and
one 66mer.
~~4~~r~~
- 3 -
In the 20mer probe all theoretically possible codons for
the amino acid sequence
Met-Met-Asp-Ile-Thr-Asp-Thr
were taken into account, with, in the case of the last
amino acid, the third position in the codon being omit-
ted. The 20mer probe is thus 48-fold degenerate, i.e. a
mixture of all 48 theoretically possible oligonucleo-
tides coding for the said amino acid sequence (Table 1;
Appendix)_
The 66mer probe was selected on the basis of the follow-
ing amino acid sequence
Tyr-Gly-Gln-hhe-Glu-Asp-Gly-Ile-Leu-Asp-Thr-
Cys-Leu-Tyr-'l/al-Met-Asp-Arg-Ala-Gln-Met-Asp
and with the assistance of statistical data (Lathe, J.
Molec. Biol. 183 (1985) 1-12) (Table 2, Appendix>.
These probes were used to screen a cDNA bank. The cDNA
was prepared from mRNA from a mature human placenta, the
mRNA being isolated from the latter, and the cDNA being
prepared therefrom. The cDNA was provided with EcoRI
ends and ligated into the EcoRI cleavage site of the
phage vector ~gt10. A positive clone, ~gt10-12, which
was identified with the abovementioned probe, was
further analyzed (Fig. 1). The sequencing, by methods
known per se, resulted in the DNA sequence which codes
for F XIIIa.
Rescreening of ths~ cDNA bank with this DNA sequence re
sulted in isolation of further clones which expand both
towards the 5'- and towards the 3'-end.
Fig. 1 shows i:he restriction map of the DNA sequence
which codes for F XIIIa. "N" designates the N-terminal
~'
- 4 -
end and "C" designates the C-terminal end of the coding
region, and "A(gg)" designates the poly(A) sequence of
89 bases. This sequence represents the whole of the
coding sequence of F XIIIa. Table 3 (Appendix) shows
the DNA sequence found (coding strand) and, deduced
therefrom, the amino acid sequence from the cloned cDNA
fragments from ~gt10-11 and ~gt10-12. The total length
of the cDNA is 39105 base-pairs. The N-terminal sequence
embracing 36 amino acids found by Takagi and Doolittle
(loc. cit.) is present in the sequence which was found.
This sequence is indicated in Table 3 with an unbroken
line between nucleotide positions 88 and 198. In addi-
tion to the sequence found by Takagi and Doolittle, the
cDNA codes for a valine in nucleotide positions 187-189.
The sequence embracing four amino acids found by
Holbrook et al. (loc. cit.) - Gly-Gln-Cys-Trp - is coded
for by the cDNA in positions 1021-1032. This sequence is
likewise indicated by an unbroken line. In addition,
the positions of the 20mer and 66mer oligonucleotide
probes are indicated by broken lines. The 20mer probe
hybridizes between positions 1507 and 1526, and the
66mer probe hybridizes between positions 766 and 831.
It is possible according to the invention to use the
coding cDNA for the preparation of modified genes which
code for proteins having altered biological properties.
It is possible for this purpose to undertake, in a man-
ner known per se, deletions, insertions and base-
exchanges.
It is also possible, by the choice of the host, to in-
fluence the nature of the modification to the F XIIIa.
Thus, there is no glycosylation in bacteria, while that
taking place in yeast cells differs from that in higher
eukaryotic cells_
Knowing the amino acid sequence of F XIIIa, it is pos-
sible to prepare, by conventional methods or gene
13 415 ~6
manipulation, part-sequences of amino acids which can
act as antigens for the preparation of polyclonal or
monoclonal antibodies. Such antibodies can be used not
only for diagnostic purposes but also for the prepara-
tion of antibody columns with which it is possible to
remove F XIIIa from solutions which contain this factor
in addition to other proteins.
It is also possible, by use of the cDNA or parts there-
of, straightforwardly to isolate from a genomic bank the
genomic clone which codes for F XIIIa and using which it
is possible not only to express it in eukaryotic cells
but also to gain further diagnostic information.
F XIIIa deficiencies can result in various syndromes
which, to a large extent, are attributed to the inability
to convert the precursors into the active form of the
enzyme. Knowledge of the cDNA of F XIIIa now permits
the preparation of diagnostic aids with which it is pos-
sible straightforwardly to establish whether genetic
modifications are present.
Thus, it is possible according to the invention to pre-
pare a highly pure factor XIIIa without any risk of con-
tamination by, for example, viruses or other proteins.
The dependence, which has existed to date, on human
plasma or placentae as source of raw material has thus
been overcome. In addition, the invention allows access
to valuable diagnostic aids and thus the analysis of
genetic F XIIIa defects.
The invention is illustrated in detail in the examples
which follows. Unless otherwise stated, percentages re-
late to weight where they do not relate to amounts.
Apart from those explained in the text, the following
abbreviations have been used:
~~4~J~s
- 6 _
EDTA - sodium ethylenediaminetetraacetate
SDS - sodium dode~cyl sulfate
DTT - dithiothreitol
BSA - bovine serum albumin
Examples:
1. Isolation of RNA from human placenta
RNA was obtained from a mature human placenta (by the
method of Chirgwin et al., Biochemistry 18 (1979) 5294-
5299). About 10 g of placental tissue was ground in
liquid nitrogen, suspended in 80 ml of 4 M guanidinium
thiocyanate containing 0.1 M mercaptoethanol and treated
in a homogenizer (Ultraturrax)*at 20,000 rpm for 90 sec.
The lysate was centrifuged at 7,000 rpm for 15 min.
(Sorvall*GSA rotor) and 2 ml of 1 M acetic acid and 60
ml of abs. ethanol. were added to the supernatant,
which was allowed to precipitate at -20°C overnight.
After sedimentation at 6,000 rpm and -10°C for 10 min,
the nucleic acids were completely dissolved in 40 ml
of 7.5 M guanidinium hydrochloride (pH 7.0) and pre-
cipitated with a mixture of 1 ml of 1 M acetic acid and
20 ml of abs. ethanol. To remove the DNA the precipi-
tation was repeated once more with half the volumes.
The RNA was dissolved in 12 ml of H20, precipitated with
a mixture of 1.2 ml of 4 M potassium acetate and 24 ml
of abs_ ethanol, sedimented and finally redissolved in
10 ml of H20 (1 ml per g tissue).
Isolation of placental mRNA containing poly(A)
To isolate mRNA containing poly(A), the placental RNA
was fractionated by oligo(dT)-cellulose chromatography
(Aviv and Leder, Proc. Natl. Acad. Sci. USA 69 (1973)
1408-1412) in 2 ml Pasteur pipettes in LiCI. About 5 mg
of placental RNA were applied to the column in buffer 1
(500 mM LiCI, 20 mM tris (pH 7.5), 1 mM EDTA, 0.1% SDS).
* denotes trade mark
i 0
- 7 -
Whereas the poly(~A)+ RNA was bound to the oligo(dT)-
cellulose, the poly(A) RNA could be eluted again.
After a wash with buffer 2 (100 mM l_iCl, 29 mM tris
(pH 7.5), 1 mM EDTA, 0.1% SDS), the poly(A)+ RNA
(placental mRNA) was eluted from the column with buffer
3 (5 mM tris (pH 7.5), 1 mM EDTA, 0.05% SOS).
For further purification, the poly(A)+ RNA was adjus-
ted to buffer 1 and rechromatographed on oligo(dT)-
cellulose. After this second purification step, the
yield of placental poly(A)+ RNA was about 4% of the
RNA used.
Synthesis of cDNA from human placenta (placental cDNA)
and double-stranded cDNA (dsDNA)
Before the cDNA synthesis, a check that the placental
mRNA containing poly(A) was intact was carried out in a
1.5% agarose gel.
Then 4 ug of placental mRNA were dissolved in 65.5 ul of
H20, denatured at 70°C for 10 min, and cooled again
in ice. The cDNA was synthesized in a 100 ul mixture
after addition of 20 ul of RT1 buffer (250 mM tris
(pH 8.2) at 42°C, 250 mM KCI, 30 mM MgCl2), 2.5 ul
of 20 mM dNTP (i.e. all four deoxynucleoside tri-
phosphates), 1 ul of 1 ug/ml oligo(dT), 1 ul of 1 M DTT,
2 ul of RNAsin and 8 ul of reverse transcriptase (24 U/
ul) at 42°C for 90 min.
Double-stranded cDNA (dsDNA) was synthesized by the
method of Gubler and Hoffmann (Gene 25 (1983) 263-269).
The synthesis was carried out immediately after the cDNA
synthesis by add ition of 305.5 ul of H20, 80 ul of RT2
buffer (100 mM tris (pH 7.5>, 25 mM MgCl2, 500 mM KCI,
50 mM DTT, 250 ug/ml BSA), 2 ul of RNase H (2 U/ul),
2.5 ul of E. coli DNA ligase (5 U/ul), 5 ul of 15 mM (3-
NAD, and 5 ul of DNA polymerase I (5 U/ul) and
_ g _
incubation at 15°I: for 5 h. The reaction was stopped
by heat inactivation (70°C, 30 min).
After addition of 55 ul of 250 uM dNTP, 55 ul of 10 mM
tris (pH 7.5), 10 mM MgCl2, 10 ug/ml BSA, 3 ul of T4
DNA polymerase I (1 U/ul), 2 ul RNase H (2 U/ul) and 2
ul of RNase A (2 ilg/ml), the reaction mixture was in-
cubated at 37°C for a further 30 min in order to cor-
rect faulty syntheses of the polymerase I on the second
DNA strand ("repair reaction").
Ligation of EcoRI linkers to the dsDNA, and opening of
the linkers
To set up a placental cDNA bank, the dsDNA was provided
with EcoRI ends in order to be able to ligate it in the
EcoRI cleavage site of the phage vector J~gt10 (T. Maniatis
et al. (1982), Molecular Cloning, A Laboratory Manual,
Cold Spring Harbor). For this purpose the dsDNA was
a> treated with EcoRI methylase in order to protect
internal i:coRl cleavage sites of the dsDNA, and
b) provided with EcoRI linkers which
c) were then opened with EcoRI.
Re a):
The methylase reaction of the dsDNA was carried out im-
mediately after the repair reaction by addition of 25
ul of 500 mM EDTA (pH 8.0), 60 ul of methylase buffer
(100 mM NaOAc (pH 5.2), 2 mg of S-adenosyl-L-methionine)
and 2 ul of EcoRI methylase (20 U/ul) by incubation at
37°C for 30 min.
The reaction mixture was extracted with phenol, and the
dsDNA was precipitated with 60 ul of 4 M NaOAc and 1300
ul of ethanol. The dsDNA was washed twice with 70% eth-
anol, extracted once by shaking with ether, and dried.
io
- 9 -
Re b):
The EcoRI-methylal:ed dsDNA was dissolved in 88 ul of
H20 and, after addition of 10 ul of ligase buffer (500
mM tris (pH 7.4), 100 mM MgCl2, 100 mM DTT, 10 mM
spermidine, 10 mM ATP, 1 mg/ml 85A) and 1 ul of T4 DNA
ligase (10 U/ul), was ligated with 1 ul of EcoRI linkers
(0.5 ug/ul) (pGGAATTCC and pAGAATTCT) at 15°C over-
night.
Re c):
The volume of the ligase mixture was made up to 120 ul
with 6 ul of H20, 12 ul of 10x EcoRI buffer and 2 ul
of EcoRI (120 U/ul.). The EcoRI digestion was carried
out at 37°C for 2 h.
Removal of unbound linkers via a potassium acetate grad-
ient and size-selE~ction of the dsDNA
To remove all the unbound EcoRI linkers from the dsDNA,
the EcoRI reaction mixture was applied in toto to a pot-
assium acetate gradient (5-20% KOAc, 1 mM EDTA, 1 ul/ml
ethidium bromide) which was centrifuged (Beckman*SN 65/
rotor) at 50,000 rpm and 20°C for 3 h. The gradient
was fractionated from below in such a way that the vol-
ume of the first five fractions was 500 ul, and that of
all the remainder was 100 ul. The fractions were pre-
cipitated with 0.01 volume of acrylamide (2 mg/ml) and
2.5 volumes of ethanol, washed once with 70% strength
ethanol and dried, and each was taken up in 5 ul of
H20.
To determine the size of the dsDNA, 1 ul of each frac-
tion was analyzed in a 1.5% agarose gel. In addition,
the quantity of dsDNA was determined on 1 ~tl of each
fraction.
* denotes trade mark
~34~~ t
- 10 -
Fractions contain=ing dsDNA with over 1000 by were com-
bined, and the sample was concentrated until the final
concentration was 27 ug/ml.
Incorporation of the dsDNA into the phage vector agt10
and in vitro packaging reaction
The incorporation of the dsDNA into the EcoRI cleavage
site of the phage vector agt10 (Vector Cloning Systems,
San Diego, CA) was carried out in a 4 ul ligase mixture:
2 ul of dsDNA, 1 yl of ~gt10 x EcoRI (1 ug/ml), 0.4 ul
of ligase buffer, 0.5 ul of H20 and 0.1 ul of T4 DNA
ligase. The mixture was incubated at 15°C for 4 h.
To establish the placental cDNA bank in the phage vector
~gt10, the ligase mixture was then subjected to an in
vitro packaging reaction with the a-lysogenic cell ex-
tracts E. coli NS 428 and NS 433 at room temperature for
2 h (Vector Cloning Systems, San Diego, CA; Enquist and
Sternberg, Methods in Enzymology 68, (1979), 281-298).
The reaction was stopped with 500 ul of suspension med-
ium (SM: 0.1 M NaCI, 8 mM MgS04, 50 mM tris (pH 7.5),
0.01y gelatine) and 2 drops of chloroform.
Determination of the titer and analysis of the placental
..nue h~..I.
The number of plaque-forming units (PFU) of the placen-
tal cDNA bank was determined using competent cells of
the E. coli K 12 strain C600 HFL: it was 1 x 106 PFU.
About 80% of the phages contained DNA inserts Larger
than 1000 base-pairs.
Oligonucleotide probes for screening the placental cDNA
bank
Two oligonucleotide probes (20mer probe and bbmer probe)
were synthesized for analysis of the placental cDNA
- 11 -
bank. Their sequences were deduced from the amino acid
primary sequence of several proteolytic and BrCN frag-
ments of F XIIIa. In some cases overlapping, and hence
longer, amino acid sequences were found, and these per-
mitted the synthesis of a very long probe, namely the
bbmer probe.
The 20mer probe is a conventional DNA probe in which all
the theoretically possible codons for the amino acid se-
quence Met-Met-Asp-Ile-Thr-Asp-Thr are taken into ac-
count (in the case of the terminal amino acid, Thr, the
third position in the codon, called the "wobble" posi-
tion, was omitted,; see Table 1). The 20mer probe is
thus 48-fold degenerate, i.e. a mixture of all the 48
theoretically possible coding oligonucleotides for the
said amino acid sequence.
The manner of construction and the use of the 66mer
probe essentially followed the rules of Lathe, J. Mol.
Biol. 183 (1985) '1-12. In order to construct the 66mer
probe two 39mer probes were synthesized (39mer A with
the sequence:
S' TATGGCCAGTTTGAGGATGGCATCCTGGACACCTGTCTG 3'; and
39mer B with the sequence:
5' GTCCATCTGGGCCCGGTCCATCACATACAGACAGGTGTC 3'). 39mer A
and 39mer B have <~ complementary sequence comprising 12
bases so that hybridization of the two sequences results
in long free 5' ends.
The two 39mer probes were (as was the 20mer probe) labeled
at the 5' end with T4 polynucleotide kinase in the pres-
ence of (Y-32P)-A'fP (about 1 ug of DNA, (Y-32P)-ATP:
3000 Ci/mmol, 10 yCi/ul, with 6 ul/40 ul reaction mix-
ture being used). The 20mer probe had a specific acti-
vity of 1 x 108 Bq/ug or 1.5 x 106 Bq/pmol. The two
39mer probes were heated at 95°C for 5 min., mixed and
slowly cooled to 4°C in a cold room, and thus hybrid-
ized. Then about 1 ug of the hybridized 39mer probes
- 12 -
was treated with DNA polymerase I, Klenow fragment, with
the addition of (cx-32P)-dATP (3000 Ci/mmol, 10 uCi/ul,
4 ul/50 ul reaction mixture) (5'->3' filling-in reac-
tion). The filled-in 66mer probe had a specific activi-
ty of 1.5 x 10$ Bq/ug. The DNA probes were stored at
-20oC; the 20mer probe was used immediately for analy-
sis (screening), while the 66mer probe had previously
been heated at 95°C for 5 min and then rapidly cooled
in an ice bath.
Since the 66mer probe had been produced by hybridization
of two 39mers followed by a 5'->3' filling-in reaction,
it is possible to carry out various experiments. On the
one hand, it is possible to hybridize cDNA banks with
the denatured 66mer DNA probe and, on the other hand, it
is then possible to hybridize positive clones with the
39mer A and B probes individually.
It is highly probable that the clones which hybridize
both with the long probe and with both short 39mer probes
A and B have the desired sequence. Thus, the method of
constructing a long, complex oligonucleotide probe and
of "rescreening" the clones using the partially com-
plementary short DNA probes, which has been described,
represents an enhancement of specificity. In addition,
it is possible w ith the long oligonucleotide and its
partially complementary short partoligonucleotides to
screen genomic banks with enhanced specificity. Another
advantage of the said method is that the synthesis of
shorter oligonucleotides can be carried out more easily
and with higher yields and accuracy. The sequence of
two enzymatic reactions, namely 1) T4 polynucleotide
kinase for the (Y.-32P>-ATP labeling, and 2) DNA poly-
merase filling-in reaction with addition of (a-32P)-dNTP,
means that it is possible to obtain higher specific
activities (at least1x108 Bq/ug DNA).
f3 x+15 iri
- 13 -
Screening of the placental cDNA with F XIIIa-specific
oligonucleotides
x 105 PFU of the placental cDNA bank were examined
5 for cDNA sequences coding for F XIIIa using the 20mer
probe and the 66mer probe. This entailed 3 x 104 PFU
being plated out with cells of the E. coli K 12 strain
C 600 HFL in soft agar in 13.5 cm Petri dishes and in-
cubated at 37°C for b h. Any lysis which had taken
place by this timed was still incomplete. The plates
were incubated in a refrigerator overnight, and the
phages were transferred to nitrocellulose filters
(Sc.hleicher & Schi~ll, BA 85, Ref. No. 401124) (dupli-
cates). The nitrocellulose filters and Petri dishes
were marked with <~n injection needle in order to allow
subsequent allocation. The Petri dishes were stored in
a cold room during the processing of the nitrocellulose
filters. The DNA on the nitrocellulose filters was de-
natured by placinc3 the filters for 5 min. on filter
paper (Whatman M 3) impregnated with 1.5 M NaCI, 0.5 M
NaOH. The filters were then renatured in the same way
using 1.5 M NaCI, 0.5 M tris (pH 8.0) and washed with
2x SSPE (0.36 M N<~Cl, 16 mM NaOH, 20 mM NaH2P04, 2 mM
EDTA). The filters were then dried in vacuo at 80°C
for 2 h. The fili:ers were washed in 3x SSC, 0.1% SDS
(20x SSC - 3 M Na(:l, 0.3 M Na citrate) at 65°C for 4 h,
arid prehybridized at 65°C for 4 h (prehybridization
solution: 0.6 M NaCI, 0.06 M tris (pH 8.3), 6 mM EDTA,
0.2% non-ionic synthetic sucrose polymer ( O Ficoll),
0.2% polyvinylpyrrolidone 40, 0.2% BSA, 0.1% SDS, 50
ug/ml denatured herring sperm DNA). The filters were
incubated overnight with the addition of 100,000-
200,000 Bq of the Labeled oligonucleotide per ml of
hybridization solution (as prehybridization solution
but without herring sperm DNA) in beakers or in sealed
polyethylene films, shaking gently. The hybridization
temperature for the 20mer probe and for the 39mer
probes was 42°C, and that,for the 66mer probe was 47°C.
i3 ~ ~ 5 1~
- 14 -
The nitrocellulose filters were washed with 6x SSC,
0.05 M sodium pyrophosphate at room temperature for
1 h and at the particular hybridization temperature for
a further hour. The filters were dried and autoradio-
graphed overnight.. Signals occurring on both duplicate
X-ray films were allocated to the Petri dish, and the
region (about SO plaques) was punched out with the wide
end of a Pasteur pipette, and the phages were re-
suspended in 1 ml of SM buffer. Positive phages were
singled out over i:hree rounds until a single clone was
obtained.
A total of 5 x 10~' PFU of the placental cDNA bank was
gxamined in several passages. 17 signals were iden-
tified on duplicai:e filters. Further screening under
more stringent conditions resulted in 7 signals still
being positive. Of these 7 PFU only one PFU showed a
positive signal both after hybridization with the 20mer
and 66mer probes and with the 39mer probes A and B.
This clone - called ~gt10-12 hereinafter - has a se-
quence of 1704 base-pairs coding for F XIIIa and having
an internal EcoRI cleavage site. Southern blot analysis
shows that the smaller EcoRI fragment of 540 base-pairs
hybridizes with the 20mer DNA probe, and the larger
fragment of 1164 base-pairs hybridizes with the 66mer
DNA probe.
On rescreening, it emerged from the Southern blot that
there is more reaction with the 39mer probe A than with
the 39mer probe B.. Sequence analysis of the clone J~gt10-
12 showed subsequently that, over the entire length of
the 66mer probe, l:here are only seven mismatches to the
sequence found for' F XIIIa (Table 2). The seven mis-
matches are distributed as follows: there are three in
the 39mer A probe and five mismatches in the 39mer B probe
(one mismatch occurs in the overlapping region, and thus
is common to both;l. The five mismatches in the 39mer B
probe are clustered, which is possibly the reason for the
'~3 ~1~ ~~O
- 15 -
weaker hybridization signals in the case of the 39mer B
probe.
Screening of the placental cDNA with nick-translated
EcoRI fragments
The two subcloned EcoRI fragments, which were 540 base-
pairs and 1164 base-pairs in length, were cloned into the
EcoRI cleavage site of the commercially available vector
pUCB (1164 by = pUCB-12.1 and 540 by - pUC8-12.2) and
were isolated therefrom preparatively with EcoRI,
and nick-translated in the presence of (a-32P)-dNTP. The
specific activity of both fragments was 1x108 Bq/ug DNA.
Using the (32P)-labeled fragments in several passages,
about 1 x 106 recombinant phages of the placental cDNA
bank were examined (hybridization temperature 65°C) and
thus 13 hybridizing phages were identified. The phages
were singled out over three rounds until a single homo-
geneous phage preparation was obtained. 20 ml lysates of
each phage were set up, and the DNA was extracted. The
DNA was digested with EcoRI and fractionated on a 1X
agarose gel. The gel was subjected to the Southern blot
technique, and the nitrocellulose filter was hybridized
with the labeled 540 by EcoRI fragment_ Eleven phages
showed hybridization signals. The nitrocellulose filter
was boiled and hybridized with the nick-translated 1164
by EcoRI fragment. Nine phages showed hybridization
signals.
It was possible to identify, on the basis of the size of
the hybridizing fragments, clones which, in comparison
with ~gt10-12, expand both towards the 5' end such as
agt10-20 and towards the 3' end such as ~gt10-11. It was
possible by use of these clones to determine the com-
plete F XIIIa cDNA sequence. It was possible to combine
part-sequences whiich were present by use of internal re-
striction sites or by hybridization of overlapping se-
quences. The complete cDNA sequence can be ligated into
X34 i~~~
- 16 -
expression vectors and expressed in suitable prokaryotic
or eukaryotic systems.
DNA sequence analysis
The phage clone ~gt10-12 was multiplied, and the DNA was
extracted. The two EcoRI fragments were isolated and
cloned into the EcoRI site of the plasmid vector pUC8.
pUC8-12.1 has the 1164 base-pair fragment, and pUC8-12.2
has the 540 base-pair fragment_ In order to isolate the
entire fragment comprising 1704 base-pairs, 7~gt10-12 was
partially digestecl with EcoRI, and the 1704 base-pair
band was isolated and cloned into the EcoRI site of
pIC19H (Marsh et al., Gene 32 (1984) 481-485). The re-
suiting plasmid is called pIC19H-12.
It was possible by cloning Sau 3A, AIuI and TaqI sub-
fragments of the clones pUCB-12.1, pUC8-12.2 and
pIC19H-12 into pUC plasmids and M13 phages, followed by
sequencing of the relevant regions using the enzymatic
dideoxy method of Sanger and the chemical method of
Maxam and Gilbert, to determine the sequence of the 1704
by fragment (Tables 3). The sequence shows only one open
reading frame and codes for the first 542 amino acids of
the factor XIIIa molecule.
Restriction analysis of the clone agt10-12 or pICl9H-12
and of the clones agt10-11 and ~gt10-20 was carried out
both by suitable single and multiple digestions and by
partial digestion of (32P)-labeled DNA fragments by
the method of Smith and eirnstiel (Smith, H.O. and
Birnstiel, M_L.,Nucleic Acids Res. 3 (1976) 2387-2398)
(Figure 1).
The clone ~gt10-11 has a fragment which is 2432 by in
size and has an internal EcoRI cleavage site. This
fragment overlaps by 237 by at the 3' end the cDNA frag-
ment from agt10-12:, and comprises the remaining 570 by
~
'7 I ~
,! '~ i J
- 17 -
of the coding sequence plus 1625 by of the non-coding
region including a poly(A) sequence of 89 bases. The
clone ~gt10-20 wil:h a cDNA fragment about 700 by in size
also has at the 5' end b bases more (GAG GAA ...) than
agt10-12.
2. Preparation of a clone which can be expressed and
contains the entire cDNA coding for F XIIIa
The starting clonEas ~gt10-11 and 1gt10-12 were used to
obtain a plasmid which contains the entire coding region
of the F XIIIa cDNA. With the aid of partial EcoRI di-
gestion, the insertion, comprising 1704 base-pairs, of
~gt10-12 was cloned into the EcoRI site of pIC19H (Marsh
et al., loc. cit.). The resulting plasmid pIC19H-12.1
was used subsequently (see Fig. 2). The clone ~gt10-11
has an insertion 2432 base-pairs in size and has an in-
ternal EcoRI cleavage site. The left ("5'-terminal")
EcoRI fragment, which is 1224 base-pairs in size and em-
braces the C-terminal 570 base-pairs of the coding se-
quence plus 654 base-pairs of the 3' non-coding region,
was likewise cloned into the EcoRI cleavage site of
pIC19H. The resulting plasmid pIC19H-11.1 was used sub-
sequently (Fig. 2). The plasmids pIC19H-12.1 and
pICl9H-11.1 have the coding region for F XIIIa in the
same orientation in the vector and have an overlapping
region which comprises 237 base-pairs. This overlapping
region was used to construct from the part-clones
pIC19H-12.1 and pIC19H-11.1 a clone which embraces the
entire coding region. This entailed preparation, from
the two plasmids mentioned, of partially single-stranded
heteroduplex molecules by hybridization in vitro (Fig.
2) and transformation of the reaction mixture into
E. coli. By utilization of the repair mechanisms of the
bacterium (3'-~5' exonuclease activity, 5'~3' polymeri
zation activity of the enzyme DNA polymerase I), a plas
mid with the entire coding region was obtained.
i
- 18 -
Specifically, this entailed 1 ug of the DNA of each of
r
the plasmids pIC19H-12.1 (CIaI-digested) and pIC19H-11.1
(BamHI-digested) being mixed and precipitated with eth-
anol. The DNA was dried in vacuo, taken up in 20 ul of
H20 and, after addition of S ul of 1 N NaOH, incubated
at room temperature for 10 minutes. The following were
then added in the sequence indicated: 200 ul of H20,
25 ul of 1 M tris.HCl (pH 8.0) and 50 ul of 0.1 N HCI.
The reaction mixture was incubated at 65oC for 3
hours, precipitated with ethanol, and dried. The DNA
was resuspended in 20 ul of H20 and transformed by
known methods into E. coli, and the cells were plated
onto LB-amp plates and incubated overnight. Of the
total of 96 ampicillin-resistant clones, 24 clones were
worked up by the alkali method (Birnboim and Doly, Nucl.
Acid Res. 7 (1979) 1513-1523), and the plasmids were
characterized by restriction endonucleolysis with
HindIII, EcoRI, BamHI and PvuII. Six plasmids showed
the expected restriction pattern. One of these plas-
mids, pFXIII-13 (Fig. 2), was characterized in detail.
This entailed the StuI (position 1225)-PvuII (position
1870) fragment, which embraces the 237 base-pair over-
lapping region, being sequenced, and the DNA sequence was
confirmed as correct. The plasmid pFXIII-13 comprises
2693 base-pairs of FXIIIa cDNA, of which 78 by are
of the 5' non-translated region, 2196 by are the entire
coding region, and 419 by are of the 3' non-translated
region. pFXIII-13 was the starting plasmid for all sub-
sequent expression experiments.
3. Expression of biologically active factor XIIIa in
E. coli
a) Construction of F XIIIa expression plasmids
pFXIII-13 has the F XIIIa cDNA insert in the correct or-
ientation with respect to the lac promoter and in the
correct reading frame with respect to the lacZ
;3~1~~~,
- 19 -
a-peptide. pFXII:f-13 is thus able to induce the syn-
thesis of a F XIIIa fusion protein in E. coli. The mol-
ecular weight of this protein comprises the 732 amino
acids of natural F XIIIa together with 16 vector-coded
amino acids plus 28 amino acids specified by the 5' non-
coding region. These additional 44 amino acids are lo-
cated at the N-terminal end of the fusion protein. The
expected molecular weight of this protein is 85,250 D
(Tab. 4).
An expression plasmid which uses in place of the lac
promoter a more efficient trpllac hybrid promoter was
subsequently constructed. For this, pKK233-2 (Amann and
Brosius, Gene 40 (1985), 183-190) was cut with EcoRI and
treated with Ba131 (Fig. 3). The DNA was then cut with
PvuII, and the fragment 2800 base-pairs in size was pur-
ified on a PAA gel. This fragment was religated in the
presence of a BgI:II linker (5'-CAGATCTG). The resulting
plasmid pTrc89-1 '(Fig. 3) was cut with NcoI and HindIII,
and the synthetic linker
5' CATGGAATTCGA 3'
3' CTTAAI,;CTTCGA 5'
was incubated together with the 2800 base-pair fragment
in a ligase mixture. The resulting plasmid, pTrc96A
(Fig. 3), was cut with EcoRI and HindIII, and the frag-
ment 2800 base-pairs in size was gel-purified and ligated
with the EcoRI-HindIII linker which is 55 base-pairs in
size from pUC18. The resulting plasmid is pTrc97A (Fig.
3). In contrast to pKK233-2, pTrc97A has, downstream of
the Ncol site which occurs only once in the plasmid, the
polylinker from pUC18 and thus has numerous cloning
sites.
The FXIIIa cONA cloned into pFXIII-13 has a Pstl site
21 base-pairs 5' away from the ATG initiation codon
(position 61). The next PstI site is located in the 3'
i~ 415 16
- 20 -
untranslated region (position 2398). The PstI fragment
which is 2337 base-pairs long was isolated from pFXIII-
13 and ligated into the PstI site of pTrc97A. The res-
ulting plasmid with the PstI fragment in the desired
orientation is pF:XIII-C4 (Fig. 3a). The expected FXIII
molecule specified by pFXIII-C4 has 22 additional N-
terminal amino acids, 15 of them being vector-coded and
7 being specified by the 5' non-coding region of the
F XIII cDNA. The expected molecular weight of this pro-
tein is 82,830 D (Tab. 4).
Use of thrombin to cleave off the 37 amino-terminal
amino acids converts the F XIIIa into the active form.
Since such a F XIIIa molecule which has already been ac-
tivated is of therapeutic interest, an attempt was made
to express in E. coli a F XIIIa that can be shortened by
thrombin cleavage. In order to obtain a high yield the
cloning was carried out in such a way that the shortened
F XIIIa is expressed in the form of a hybrid protein,
fused to an E. coli S-galactosidase fragment. The SmaI-
HindIII fragment which is about 2700 by in size from
pFXIII-13 was isollated and ligated into pBD2IC20H which
had been hydrolyzed with SmaI and HindIII. In the new
plasmid pMB259 (Fig. 3b), the coding region of the cDNA
for F XIIIa from amino acid Pro37 to Met732 is located
in the reading frame applying to the 375 aminoterminal
amino acids of the S-galactosidase, and it has the throm-
bin cleavage site Arg3g/Gly3g so that it is possible to
obtain activated F XIIIa by thrombin cleavage from the
synthesized fusion proteins.
The expression vector pBD2IC20H had been constructed by
ligating the polyl'~inker region, which comprises 58bp,
of the plasmid pIC20H (Marsh et al., loc. cit.) as BamHI-
HindIII fragment in pBD2 (Broker, Gene Anal. Techn. 3
(1986) 53-57) which has the lac promoter.
"34i~~~
- z1 -
b) Expression
It was found that E. coli cells of the strain D29A1
which are transformed with pFXIII-13, with pFXIII-C4 or
with pMB259 are able to synthesize the expected F XIIIa
proteins. The expression of F XIII by the plasmids
pFXIII-13, pMB 259 and pFXIII-C4 can be induced with
IPTG. Comparison of protein extracts after IPTG in-
duction of E. coli D29A1 (pFXIII-13) and D29A1 (pFXIII-
C4) on PAA gels stained with Coomassie blue showed the
expected molecular weights of the F XIII fusion pro-
teins. The estimated expression of the "44aa FXIII
fusion protein" is about 5 times that of the "22aa FXIII
fusion protein".
It was also found that the F XIIIa molecules specified
by pFXIII-73 and pFXIII-C4 have biological activity. In
contrast, no F XIII activity was found in E. coli D29A1
control extracts. The activity found in the clot stabi-
lity assay (Karges, in Bergmeier, Methods of Enzymatic
Analysis, Volume '_i, Enzymes 3: Peptidases, Proteinases
and their Inhibitors, pages 400-410) is S ug/l for
E. coli D29A1 (pF);III-13>, based on the E. coli culture
(0D250=1-5), and is 15 ug/l for D29A1 (pFXIII-C4>.
The amount of factor XIII found in the F XIIIa-specific
ELISA is 1.5 mg/l, based on the E. coli culture, for
pFXIII-13 and is 3 mg/l for pFXIII-C4. The discrepancy
between the amounts of F XIII measured in the biological
assay and in the ELISA derives from the fact that the
major part (> 90X) of F XIII in the E. coli cell is in
the form of an insoluble precipitate which is bio-
logically inactive and dissolves only in 7 M urea. The
soluble fraction of the F XIII molecules present in E.
coli extracts shows in the Ouchterlony test
(Ouchterlony, Progr. Allergy 5 (1958) 1) a precipitation
curve which is substantially identical to that of F XIII
isolated from placenta.
~~ 4~~ ~~
- 22 -
The expression of eukaryotic proteins in E. coli in the
form of insoluble protein aggregates has already been
described for several proteins. These proteins can be
dissolved out of such aggregates using a chaotropic
agent and can be converted by suitable renaturing con-
ditions into their biologically active form.
4. Expression of F XIII in yeasts
The synthesis of biologically active F XIII obtained by
gene manipulation from yeasts was achieved by incorpor-
sting the cDNA coding for F XIII into expression vec-
tors which are able to replicate autonomously in yeasts.
It was possible to isolate F XIII-active protein from
extracts of the recombinant clones.
The conditions for growing yeasts and the molecular bio
logical methods are described in Dillon et al., Recombi
nant DNA Methodology, John Wiley & Sons, New York (1985)
and in Maniatis et al., Loc. cit..
The F XIII cDNA was isolated from the vector pFXIII-13
as a HindIII fragment about 2700 by in size, and was
cloned into the HindIII site of the vector pAAHS (Ammerer,
Meth. Enzymol. 10'1 (1983), 192-201). Thus, the F XIII
cDNA in the resulting plasmid pMB240 (Fig. 4) is under
the control of then strong ADHI promotor which contains
the gene expression signals of alcohol dehydrogenase.
The plasmid pMB240 was transformed into baker's yeast,
Saccharomyces cere~visiae, strain Leu 2-3, Pep 4-3, by
the method of Itoh et al. (J. Bacteriol. 153 (1983),
163-168) and Leu+ transformants were selected on YNB
minimal medium. One colony of transformed yeast cells
was used to inoculate a liquid culture with YNB medium.
After growth at 30°C for two days, transfer into com-
plex YPB medium was carried out and, after a further
three days, the cells were removed by centrifugation and
were disrupted with a glass bead mill in isotonic saline
1:~~~5~6
- 23 -
solution containing 100 mM sodium citrate (pH 7.2).
The cell extract was subjected to high-speed centrifu-
gation in a Sorval,l high-speed centrifuge, in a SS34
rotor, at 20,000 rpm and 4°C for 1 hour.
The cell-free supernatant of S. cerevisiae (pMB240> was
analyzed by the Western blot method. In addition to a
band which can be detected in the same position as
F XIII from placenta, a protein of about 116,000 D re-
acted specifically with the anti-F XIII serum used.
This proves that part of the F XIII formed in yeasts is
glycosylated. The glycosylation of proteins is often a
factor prolonging the half-life of proteins, especially
plasma proteins. In addition, carbohydrate side-chains
may increase the activity or extend the duration of the
action of plasma proteins, for example antithrombin III.
A F XIII which is expressed in yeast and which, in con-
trast to F XIII obtained from placenta, is glycosylated
or has undergone iposttranslatimal modificaiton in some
other way can have, owing to an increased activity,
advantages over F XIII from placenta.
The cell-free supernatant was examined for F XIII by an
ELISA, and the F XIII concentration was found to be 150
ng/ml, based on the yeast culture. The biological acti-
vity of F XIII was determined by the method of Karges,
loc. cit., and confirmed the concentrations measured in
the ELISA. It was possible to rule out a non-specific
F XIII-like activity by yeast proteins because the bio-
logical activity of the F XIII obtained from baker's
yeast could be specifically inhibited by anti-F XIII
antibodies.
5. Expression of F XIIIa in animal cells
a) Construction of expression vectors for animal cells
'.i
33 415 ~6
- 24 -
The synthesis of this vector is in the Append-
dix). Apart from this plasmid, use was made of the vec-
tors pZET4 (see below) and pSP6HS9, which has the Droso-
phila heat shock protein 70 promotor (Wurm et al., Proc.
Natl. Acad. Sci. ~JSA 83 (1986) 5414-5418>, for the
expression of F XIIIa in animal cells.
pZET4 (Fig. 5): the plasmid pSVA STOP1 was cut with
BamHI, and the vector fragment which is 2.6 kb in size
and has the SV40 early promotor was isolated. The
BgIII-BamHI fragment 0.85 kb in size from pSV2dhfr (Lee
et al., Nature 294 (1981) 228-232) was ligated into the
vector which had been pretreated in this way, which re-
suited in the plasmid pZET4. Located on the 0.85 kb
fragment from pSV2dhfr are mRNA splice sites from exon-
intron joins, and the polyadenylation site of the gene
for the t antigen from the SV40 DNA.
b) Construction of F XIIIa expression vectors for animal
cells
pSVF13 (Fig. 5a): the expression vector pSVA STOP1 was
cut with HindIII and XbaI. A HindIII-XbaI fragment,
about 2.7 kb in size and having the F XIIIa cDNA, from
pFXIII-13, was ligated into the vector which had been
treated in this way. The F XIIIa transcription unit on
pSVF13 has no mRNA splice sites.
pZF13 (Fig. 5b): a HindIII fragment about 2.7 kb in
size and containing the F XIIIa cDNA was isolated from
the plasmid pFXIII-13. The resulting 5' protruding end
was eliminated by filling in the complementary strand
with DNA polymerase I (Klenow fragment). The expression
vector pZET4 was linearized by cutting at the unique
XbaI site. The resulting 5' protruding end was likewise
eliminated by filling in the complementary strand with
DMA polymerase I (Klenow fragment). Ligation of the
filled-in vector with the filled-in F XIIIa cDNA fragment
~~ 4 ~ ~ ~6
- 25 -
results in the F ;KIIIa expression plasmid pZF13. As in
pSVF13, the F XIIIa cDNA is under the transcriptional
control of the SV~+0 early promotor but is provided with
mRNA splice sites.
S
pHSF13 (Fig. 5c): the plasmid pSVF13 was partially di-
gested with EcoRI,, and a fragment about 2.9 kb in size
and having the F ;KIIIa cDNA followed by the SV40 poly-
adenylation site 'for early transcripts was isolated.
This fragment was ligated into the unique EcoRI site of
the plasmid pSP6H:i9 downstream of the heat shock protein
70 promotor.
c) DHFR expression vectors for the cotransfection of
CHO (Chinese hamster ovary) dhfr cells
Either the DHFR vector pSV2dhfr (Lee et al., loc. cit.)
or the promotorless DHFR plasmid pSVOAdhfr (German Patent
Application P 36 24 453.8, see Appendix) was used for the
cotransfection w ith the described F XIIIa expression
vectors in CHO dhfr cells. Both vectors have the DHFR
cDNA from the mouse.
d) Vector conferring 6418 resistance for the cotrans-
fection of BHK (baby hamster kidney) cells
The F XIIIa expression vectors which have been described
were cotransfected with the vector pRMH140 (Hudziak et
al., Cell 31 (1982), 137-146) in BHK cells.
e) Expression of F XIIIa in CHO cells
Cotransfection of the plasmid pZF13 with the DHFR vector
pSV2dhfr, and of the expression plasmid pSVF13 together
with the DHFR veci:or pSVOAdhfr, was carried out using the
calcium phosphate precipitation method (Graham and van
der Eb, Virology 52 (1973), 456-467) in CHO dhfr cells.
This entailed 20 ug of the particular F XIIIa expression
plasmid (pZF13 or pSVF13) being mixed and coprecipitated
- 26 -
with 5 ug of the DHFR vectors (pSV2dhfr or pSVOAdhfr).
The coprecipitate was used for transfection as described
above (0.5 x 106 cells in a 25 cm2 culture bottle).
After 3 days, the cells were trypsinized, transferred
into three 60 mm Petri dishes and mixed with selection
medium (containing no glycine, hypoxanthine or thymi-
dine). The only cells which survive under these con-
ditions are those which have undergone stable trans-
fection with the DHFR gene. Colonies of transfected
cells become visible on the Petri dishes after 1-3
weeks. The following transfection rates were achieved
by this method:
pSV2dhfr 5 x 10 5/plate
pSVOAdhfr 1 x 10 5/plate
Individual clones were isolated and multiplied in a me-
dium containing no glycine, hypoxanthine or thymidine.
A specific ELISA with a lower detection limit of about
3 ng/ml was used to detect F XIIIa in culture super-
natants and cell Lysates of the individual clones. Cul-
ture supernatants were used as such in the ELISA. The
cell lysates were prepared as follows:
Confluent cells in 25 cm2 culture bottles were washed
twice in 40 mM tris.HCl (pH 7.4), 1 mM EDTA, 150 mM
NaCI, taken up in 150 ul of 0.25 M tris.HCl (pH 7.8),
5 mM DTT, 2% glycerol, 0.2% detergent (Triton*X100), and
lyzed by freezing and thawing three times.
Insoluble constituents of the cells were removed by cen-
trifugation. The lysate was diluted 1:2.5 for use in
the ELISA. The detection method which has been des-
cribed was applied to 17 clones for the combination of
the plasmids pZF13/pSV2dhfr, and one clone expressing
F XIIIa (CHO 59-5-~C7) was found. After cotransfection
with the plasmids pSVF13/pSVOAdhfr and analysis of 12
clones, a further F XIIIa-expressing clone (CHO b0-3-C1)
was detected. With both the positive clones it was pos-
sible to detect F XIIIa in the medium and in the lysate
in the same relative amounts. In order to determine
* denotes trade mark
~~ 4~5 i~
- 27 -
quantitatively the expression rate of the lines producing
F XIIIa the following standard procedure was carried out:
0.5 x 106 cells were plated out in 5 ml medium in 25 cm2
culture bottles. The medium was changed after 24 hours
(5 ml). Another 24 hours later the medium was removed,
the cell count was determined, and lysates were prepared.
For all the expression rates (n9/106 cells/ 24 h) stated
hereinafter, the cell count per 25 cm2 bottle at the end
of the test was 1 + 0.25 x 106 cells. The table which
follows shows the expression rate of the basic clones
tested in the manner described:
extracellular intracellular
(ng/106 cells/ (ng/106 cells/
24h) 24 h)
CHO 59-5-C7 12 9
CHO 60-3-C1 17 13
On SDS electrophoresis followed by F XIIIa-specific
immunoblotting of lysates and supernatants of the clone
CHO 59-5-C7 and the clone CHO 60-3-C1, in each case one
band with the molE~cular weight of the protein isolated
from human placenl:a showed a reaction.
The clone CHO 59-.'i-C7 was exposed to increasing con-
centrations of mei_hotrexate (Mtx) for gene amplifica-
tion. Starting with a concentration of 10 nM Mtx and a
4-transfer adaptation time, the Mtx concentration in the
medium was increased to 50 nM. The following expression
rates were determined in the standard procedure:
Mtx (nM) extracellular intracellular
(ng/106 cells/24 h) (n9/106 cells/24 h)
0 12 9
10 19 16
50 38 66
i5 4'~5'~~
_ 28 _
f) Expression of f= XIIIa in BHK cells
20 ug of each of the F XIIIa expression plasmids pZF13,
pSVF13 and pHSF13 were cotransfected with 5 ug of the
plasmid pRMH140 which codes for 6418 resistance, by the
calcium phosphate precipitation method described in ex-
ample 5e) in BHK cells. After 3 days, the cells were
trypsinized, tran:>ferred into three 60 mm Petri dishes
and mixed with selection medium containing 400 ug/ml
6418. After 12 days about 200-300 6418 resistant colo-
nies had grown in each Petri dish. The total number of
clones was trypsinized and subjected to transfers as
combined clone (C(:) in 25 cm2 culture bottles (5 ml of
medium).
In the case of cells transfected with pZF13 and pSVF13,
where an 80-100% confluence had been reached F XIIIa was
determined in the medium and in the relevant lysate (see
example 5e))using a specific ELISA. Combined clones
which had been transfected with pHSF13 and had likewise
reached 80-100% confluence were mixed with fresh medium
which had been preheated to 42°C and were incubated at
42°C for one hour.. After another replacement of the
medium with fresh medium equilibrated at 37°C, the
cells were maintained at 37°C for 24 hours. The
medium and lysate<.~ were then examined for their content
of F XIIIa as described above. The table which follows
summarizes the cellular distribution of F XIIIa for the
various combined clones.
extracellular intracellular
(ng) (ng)
BHK-MK1 (pZF13) 65 22
BHK-MK2 (pZF13) 80 20
BHK-MK3 (pSVF13) 85 18
BHK-MK5 (pHSF13) 170 11
i
29
The expression rates relating to the F XIIIa present in
the medium were determined for the individual combined
clones by the standard procedure described in example
5e). For all the expression rates (ng/106 cells/ 24 h)
stated hereinafter, the cell count per 25 cm2 bottle at
the end of the test was 4.5 + 0.5 x 106 cells.
extracellular
(ng/106 cells/24 h>
BHK-MK1 (pZF13) 3.4
BHK-MK2 (pZF13) 5.6
BHK-MKS (pHSF13) 3.8
BHK-MKb (pHSF13) 3.8
Since the transfected BHK lines described hitherto have
been mixed populations including cells which were not
producing or differed in their expression rates, it was
subsequently attempted to isolate genetically uniform
cell lines with high expression rates by singling out
clones. For this purpose, cells from the particular
combined clone were placed on microtiter plates in a
concentration of 1 cell/well, 2 cells/well or 4 cells/
well. Supernatants from wells in which only one clone
had grown were analyzed by the F XIIIa-specific ELISA.
The clones with the highest expression rates were multi-
plied in 25 cm2 culture bottles, and their expression
rates were investigated by the standard procedure des-
cribed above. The table which follows shows the expres-
sion rate of clones obtained by singling out BHK-MK1
(pZF13) in the manner described.
~3 41 ~ 16
- 30 -
extracellular
(ng/106 cells/24 h)
BHK MK1 (pZF13) 3.4
BHK MK1-A12 ( " ) 8
BHK MK1-E2 ( " ) 25
BHK MK1-F12 ( " ) 14
BHK MK1-C1 ( " ) 22
It was also shown,, taking the example of the BHK cell
line MK1-E2, that the F XIIIa molecules synthesized by
these cells have biological activity. 108 cells of
the BHK MK1-E2 line and of the non-transfected BHK line
which was used (negative control) were taken up in 1.5
ml of 0.25 M tris.HCl (pH 7.8) containing 2% glycerol.
The cells were lyzed by freezing and thawing three
times, followed by treatment in a Dounce homogenizer.
After removal of insoluble constituents by centrifugation,
the lysate was used in the biological assay (see example
3). The followinc3 F XIIIa activities were found:
BHK MK1-E2 0.06 units/108 cells
BHK (not transfected) 0 "
Table 1
20mer probe, 48-fold degenerate
Amino acid sequence Met Met Asp Ile Thr Asp Thr
DNA probe ATG ATG GAT ATT ACT GAT AC
C C A C
A C
G
~ '~ 4 ~1 ~ .~ ~
- 31
-
w
",
!V
N ~
m
t
t
it
N ~ c
p
t
a,. s
v v a
ac~csa V
:
v
~
"J V
V V f.~'~
L~'~
it?
t
. "err
v w
v ar's'e
w
'
~ w ~
.~. t'~ ~
c v
t~
.r ~ i
t
s
is
,e~ ~
~ ~ w
~ v
.. ;
w c~
w
w.
c ~
c c
, t~
> >
t t~
~ r-
'-
w
~~$ wv
v
..
~ w
cv
a
+ ", ~w ~
~ ~
~'
c. V 1~
w
w
' ~wss~
d = c
ss c
cs : w
s c wr
~'
r
Q 3
cue!
'
' '~
c~!
0
s
c to
v
.r w
w
d ~ '~ c
i' s
o p " ccc
v
~
,.. ~
~v~
r.
aw
~ ~ ~
~ c~
c a. ~
' a~i
~
c g ~ a
3
,s ~
w w w
s,
~ ~
c
'
'G L
E w O
v
r v ~ ~'
d7 d c
N
C r ff ,p
~ ~ N
. Q N pyn
'3 ~~~ ~fi
- 32 -
TABLE 3
1
GAGGAAGTCCCCGAGGCGCACAGAGCAAGCCCACGCGAGGGCACCTCTGGAGGGGAGCGCCTGCAGGACCTTGTAAAGT
C
81
AAAA MET S~R GLU THR SER ARG THR ALA PH~ GLY GLY ARG ARG ALA VAL PRO PRO ASN
ASN
ATG TCA GAA ACT TCC AGG ACC GCC TTT GGA GGC AGA AGA GCA GTT CCA CCC AAT AAC
1~2
SER ASN ALA ALA GLU ASP ASP LEU PRO THR VAL GLU LEU GLN GLY VAL YAL PRO ARG
GLY
TCT AAT GCA GCG GAA GAT GAC CTG CCC ACA GTG GAG CTT CAG GGC GTG GTG CCC CGG
GGC
202
VAL ASN LEU GLN GLU PHE LEU ASN VAL THR S~R VAL HIS LEU PHE LYS GLU ARG TRP
ASP
GTC AAC CTG CAA GAG TTT CTT AAT GTC ACG AGC GTT CAC CTG TTC AAG GAG AGA TGG
GAC
262
THR ASN LYS VAL ASP H1S HIS THR ASP LYS TYR GLU ASN ASN LYS LEU ILE YAL ARG
ARG
ACT AAC AAG GTG GAC CAC CAC ACT GAC AAG TAT GAA AAC AAC AAG CTG ATT GTC CGC
AGA
322
GLY GLN SER PH~ TYR VAL GLN ILE ASP L~U SER ARG PRO TYR ASP PRO ARG ARG ASP
L~U
GGG CAG TCT TTC TAT GTG CAG ATT GAC CTC AGT CGT CCA TAT GAC CCC AGA AGG GAT
CTC
382
PHE ARG VAL GLU TYR VAL ILE GLY ARG TYR PRO GLN GLU ASN LYS GLY THR TYR ILE
PRO
TTC AGG GTG GAA TAC GTC ATT 6GT CGC TAC CCA CAG GAG AAC AAG GGA ACC TAC ATC
CCA
442
VAL PRO IL~ VAL SER GLU L~U GLN SER GLY LYS TRP GLY ALA LYS IL~ VAL MET ARG
GLU
GTG CCT ATA GTC TCA GAG TTA CAA AGT GGA AAG TGG GGG GCC AAG ATT GTC ATG AGA
GAG
502
ASP ARG SER VAL ARG LEU S~R ILE GLN SER SER PRO LYS CYS IL~ VAL GLY LYS PH~
ARG
GAC AGG TCT GTG CGG CTG TCC ATC CAG TCT TCC CCC AAA TGT ATT GTG GGG AAA TTC
CGC
562
MET TYR VAL ALA VAL TRP THR PRO TYR GLY VAL LEU ARG THR SER ARG ASN PRO GLU
THR
ATG TAT GTT GCT GTC TGG ACT CCC TAT GGC GTA CTT CGA ACC AGT CGA AAC CCA GAA
ACA
622
ASP THR TYR IL~ L~U PHE ASN PRO TRP CYS GLU ASP ASP ALA VAL TYR LEU ASP ASN
GLU
GAC ACG TAC ATT CTC TTC AAT CCT TGG TGT GAA GAT GAT GCT GTG TAT CTG GAC AAT
GAG
~~~15~s
- 33 -
682
LYS GLU ARG GLU GLU TYR VAL l~U ASN ASP IL~ GLY VAL IL~ PHE TYR GLY GLU YAL
ASN
AAA GAA AGA GAA GAG TAT GTC CTG AAT GAC ATC GGG GTA ATT TTT TAT GGA GAG GTC
AAT
742
ASP IL~ LYS THR ARG S~R TRP SER TYR GLY GLN PHE GLU ASP GLY ILE L~U ASP THR
CYS
GAC ATC AAG ACC AGA AGC TGG AGC LAT GGT CAG TTT GAA GAT GGC ATC CTG~AC ACT TGC
802
LU VALMETASP ARG GLN ASP LEU GLY ARG ASN ILE VAL
TYR ALA MT SR GLY PRO LYS
CTG GTGATGGAS A~ CAA G~ CTC GGA AGA AAT ATC GTC
TAT GSA A~ TCT GGG CCC AAA
66 mer
862
SR VALGLYSER ALA VAL ALA LYS ASP GLU VAL YAL SR
ARG MET ASN ASP GLY LU GLY
AGC GTGGGGTCT GCA GTG GCC AAA GAC GAA GTC GTT TCC
CGT A'fG AAT GAT GGT CTC GGA
922
TRP ASP ASN ILE TYR ALA TYR ~Y VAL PRO PRO SER ALA TRP THR GLY S~R VAL ASP ILE
TGG GAC AAT ATC TAT GCC TAT GGC GTC CCC CCA TCG GCC TGG ACT GGA AGC GTT GAC
ATT
982
LEU LEU GLU TYR ARG S~R S~R GLU ASN PRO VAL ARG TYR GLY GLN CYS TRP VAL PH~
ALA
CTA TTG GAA TAC CGG AGC TCT GAG AAT CCA GTC CGG TAT GGC CAA TGC TGG GTT TTT
GCT
1042
GLY VAL PH~ ASN THR PHE LEU ARG CYS L~U GLY ILE PRO ALA ARG IL~ YAL THR ASN
TYR
GGT GTC TTT AAC ACA TTT TTA CGA TGC CTT GGA ATA CCA GCA AGA ATT GTT ACC AAT
TAT
1102
PHE SER ALA HIS ASP ASN ASP ALA ASN L~U GLN MET ASP IL~ PHE LEU GLU GLU ASP
GLY
TTC TCT GCC CAT GAT AAT GAT GCC AAT TTG CAA ATG GAC ATC TTC CTG GAA GAA GAT
GGG
1162
ASN VAL ASN SER LYS LEU THR LYS ASP SER VAL TRP ASN TYR HIS CYS TRP ASN GLU
ALA
AAC GTG AAT TCC AAA CTC ACC AAG GAT TCA GTG TGG AAC TAC CAC TGC TGG AAT GAA
GCA
1222
TRP MET THR ARG PRO ASP L~U PRO VAL GLY PH~ GLY GLY TRP GLN ALA VAL ASP SER
THR
TGG ATG ACA AGG CCT GAC CTT CCT GTT GGA TTT GGA GGC TGG CAA GCT GTG GAC AGC
ACC
1282
PRO GLN GLU ASN S~R ASP GLY MET TYR ARG CYS GLY PRO ALA SER VAL GLN ALA 1LE
LYS
CCC CAG GAA AAT AGC GAT GGC ATG TAT CGG TGT GGC CCC GCC TCG 6TT CAA GCC ATC
AAG
1342
HIS GLY HIS VAL CYS PHE GLN PHE ASP ALA PRO PHE VAL PHE ALA GLU VAL ASN SER
ASP
CAC GGC CAT GTC TGC TTC CAA TTT GAT GCA CCT TTT GTT TTT GCA GAG GTC AAC AGC
GAC
_34- "3 ~~5 ~6
1402
LEU ILE TYR ILE THR ALA LYS LYS ASP GLY THR HIS VAL VAL GLU ASN VAL ASP ALA
THR
CTC ATT TAC ATT ACA GCT AAG AAA GAT GGC ACT CAT GTG GTG GAA AAT GTG GAT GCC
ACC
1462
HIS IL~ GLY LYS LEU ILE VAL THR LYS GLN ILE GLY GLY ASP GLY MET MET ASP ILE
THR
CAC ATT GGG AAA TTA ATT GTG ACC AAA CAA ATT GGA GGA GAT GGC(ATG ATG GAT ATT
ACT
1522
ASP THR TYR LYS PHE GLN GLU GLY GLN GLU GLU GLU ARG LEU ALA L~U GLU THR ALA
L~U
GAT ACT TAC AAA TTC CAA GAA GGT CAA GAA GAA GAG AGA TTG GCC CTA GAA ACT GCC
CTG
20 meter
1582
MET TYR GLY ALA LYS LYS PRO LEU ASN THR GLU GLY YAL MET LYS S~R ARG S~R ASN
VAL
ATG TAC GGA GCT AAA AAG CCC CTC AAC ACA GAA GGT GTC ATG AAA TCA AGG TCC AAC
GTT
1642
ASP MET ASP PHE GLU VAL GLU ASN ALA VAL L~U GLY LYS ASP PHE LYS LEU S~R IL~
THR
GAC ATG GAC TTT GAA GTG GAA AAT GCT GTG CTG GGA AAA GAC TTC AAG CTC TCC ATC
ACC
1702
PHE ARG ASN ASN SER HIS ASN ARG TYR THR IL~ THR ALA TYR L~U S~R ALA ASN ILE
THR
TTC CGG AAC AAC AGC CAC AAC CGT TAC ACC ATC ACA GCT TAT CTC TCA GCC AAC ATC
ACC
1762
PHE TYR THR GLY VAL PRO LYS ALA GLU PH~ LYS LYS GLU THR PH~ ASP YAL THR LEU
GLU
TTC TAC ACC GGG GTC CCG AAG GCA GAG TTC AAG AAG GAG ACG TTC GAC GTG ACG CTG
GAG
1822
PRO LEU SER PH~ LYS LYS GLU ALA VAL LEU ILE GLN ALA GLY GLU TYR MET GLY GLN
LEU
CCC TTG TCC TTC AAG AAA GAG GCG GTG CTG ATC CAA GCC GGC GAG TAC ATG GGT CAG
CTG
1882
L~U GLU GLN ALA S~R L~U HIS PH~ PH~ VAL THR ALA ARG ILE ASN GLU THR ARG ASP
VAL
CTG GAA CAA GCG TCC CTG CAC TTC TTT GTC ACA GCT CGC ATC AAT GAG ACC AGG GAT
GTT
1942
LEU ALA LYS GLN LYS S~R THR YAL LEU THR ILE PRO GLU ILE ILE IL~ LYS VAL ARG
GLY
CTG GCC AAG CAA AAG TCC ACC GTG CTA ACC ATC CCT GAG ATC ATC ATC AAG GTC CGT
GGC
2002
THR GLN VAL VAL GLY SER ASP MET THR VAL THR VAL GLN PH~ THR ASN PRO LEU LYS
GLU
AC1 CAG GTA GTT GGT TCT GAC ATG ACT GTG ACA GTT CAG TTT ACC AAT CCT TTA AAA
GAA
2062
THR L~U ARG ASN VAL TRP VAL HIS LEU ASP GLY PRO GLY VAL THR ARG PRO MET LYS
LYS
ACC CTG CGA AAT GTC TGG GTA CAC CTG GAT GGT CCT GGA GTA ACA AGA CCA ATG AAG
AAG
_35- X341516
2122
M~T PHE ARG GLU IL~ ARG PRO ASN S~R THR YAL GLN TRP GLU GLU YAL CYS ARG PRO
TRP
ATG TTC CGT GAA ATC CGG CCC AAC TCC ACC GTG CAG TGG GAA GAA GTG TGC CGG CCC
TGG
2182
VAL S~R GLY HIS ARG LYS L~U IL~ ALA S~R M~T S~R S~R ASP S~R L~U ARG HIS VAL
TYR
GTC TCT GGG CAT CGG AAG CTG ATA GCC AGC ATG AGC AGT GAC TCC CTG AGA CAT GTG
TAT
2242
GLY GLU LEU ASP VAL GLN IL~ GL1V ARG ARG PRO S~R M~T SSS
ATGCACAGGAAGCTGAGATGAAC
GGC GAG CTG GAC GTG CAG ATT CAA AGA CGA CCT TCC ATG TGA
2307
CCTGGCATTTGGCCTCTTGTAGTCTTGGCTAAGGAAATTCTAACGCAAAAATAGCTCTTGCTTTGACTTAGGTGTGAAG
A
2387
CCCAGACAGGACTGCAGAGGGCCCCAGAGTGGAGATCCCACATATTTCAAAAACATACTTTTCCAaacCCAGGCTATTC
G
2467
6CAAGGaAGTTaGTTTTTAATCTCTCCACCTTCCAAAGAGTGCTAAGCATTAGCTTTaATTAAGCTCTCATAGCTCATA
A
2547
GAGTAACAGTCATCATTTATCATCACAAATGGCTACATCTCCAAATATCAGTGGGCTCTCTTACCAGGGAGATTTGCTC
A
2627
ATACCTGGCCTCATTTAAAACAAGACTTCAGATTCCCCACTCAGCCTTTTGGGAATAATAGCACATGATTTGGGCTCTA
G
2707
AATTCCAGTCCCCTTTCTCGGGGTCAGGTTCTACCCTCCATGTGAGAATATTTTTCCCAGGACTAGAGCACAACATAAT
T
2787
TTTATTTTTGGCAAAGCCAGAAAAAGATCTTTCATTTTGCACCTGCAGCCAAGCAAATGCCTGCCAAATTTTAGATTTA
C
2867
CTTGTTAGAAGAGGTGGCCCCATATTAACAAATTGCATTTGTGGGAAACTTAACCACCTACAAGGAGATAAGAAAGCAG
G
2947
TGCAACACTCAAGTCTATTGAATAATGTAGTTTTGTGATGCATTTTaTAGAATGTGTCACACTGTGGCCTGATCAGCAG
G
3027
AGCCAATATCCCTTACTTTAACCCTTTCTGGGATGCAATACTAGGAAGTAAAGTGAAGAATTTATCTCTTTAGTTAGTG
A
3107
TTATATTTCACCCATCTCTCAGGAATCATCTCCTTTGCAGAATGATGCAGGTTCAGGTCCCCTTTCAGAGATATAATAA
G
3187
CCCAACAAGTTGAAGAAGCTGGCGGATCTAGTGACCAGATATATAGAAGGACTGCAGCCACTGATTCTCTCTTGTCCTT
C
-as- . ~3 41~ i6
.J7
ACATCACCATTTTGAGACCTCAGCTTGGCACTCAGGTGCTGAAGGGTAATATGGACTCAGCCTTGCAAATAGCCAGTGC
T
3347
AGTTCTGACCCAACCACAGAGGATGCTGACATCATTTGTATTATGTTCCAAGGCTACTACAGAGAAGGCTGCCTGCTAT
G
3427
TATTTGCAAGGCTGATTTATGGTCAGAATTTCCCTCTGATATGTCTAGGGTGTGATTTAGGTCAGTAGACTGTGATTCT
T
3507
AGCAAAAAATGAACAGTGATAAGTATACTGGGGGCAAAATCAGAATGGAATGCTCTGGTCTATATAACCACATTTCTGA
G
3587
CCTTTGAGACTGTTCCTGAGCCTTCAGCACTAACCTATGAGGGTGAGCTGGTCCCGTCTATATATACATCATACTTAAC
T
3fi67
TTACTAAGTAATCTCACAGCATTTGCCAAGTCTCCCAATATCCAATTTTAAAATGAAATGCATTTTGCTAGACAGTTAA
A
3747
CTGGCTTAACTTAGTATATTATTATTAATTACAATGTAATAGAAGCTTAAAATAAAGTTAAACTGATTATAAAAAAAAA
A
3827
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
~3 ~~~ do
- 37 -
w
0
N C9
v~ N
C9
T~ 4
U
U
tn N
4 ~ ~ 4 ~
m
s~a U it U
tOn N e~:i N N eN Qi
t' 'c~ ~. t~ ~o
~ ~ t~.~ C9
~ ao
~-O~ 4 ~~~a 4 i~r
U U U U 4 U
r
d U d U O~U
v~ N cn N w U
''' 1~.1 t~ U
r-1
r, ~r
x x
.T~~. N m
x
a 4 a 4 w c~
v~ H cn N 4 V
~ 1
~ .
~ . ~ . at
~-Or~~ ..~~4N ~Ur
UNU UNU ~N4
V
..1 ,.
~ ~ ~ N ~ N
t
G4 p,
V
.a
N
.r
U
U c
~r
t 1
H E
H H ~ t0
W ~ ~ II
4 a~6
- 38 -
Appendix
Synthesis of Vector pSVA STOP 1
a) Construction of an expression vector for animal cells
The plasmid pSV2dhfr (Lee et al., loc. cit.) was cut
with HindIII and EcoRI, and the 2.65 kb vector fragment
which has the SV4~D early promotor was isolated. A 67 by
HindIII-EcoRI fragment from pUC12 STOP (broker and
Amann, Appl. Microbiol. Biotechnol. 23 (1986) 294-296)
was ligated into the vector which had been pretreated in
this way, which results in the plasmid pSV2 STOP. On the
67 by fragment from pUCl2 STOP there are translation
stop codons in all three reading frames. pSV2 STOP was
linearized with SacI, and the resulting 3' protruding
end was removed using the 3'-~5' exonuclease activity of
DNA polymerase I. Then digestion with EcoRI was carried
out. After ligation with an EcoRI-HpaI fragment 133 by
in size from p8B3 (e. 8ourachot et al., EMBO J. 1 (1982)
895-900), which has the SV40 polyadenylation signal for
early transcripts, it was possible to obtain the expres-
sion vector pSVA STOP1.
The polyadenylation site can also be isolated from the
vector pIG6 (Bourachot et al., loc. cit.). It is pos-
sible in exactly the same way to isolate from the SV40
gene the 133bp BamHI-HpaI fragment, to fill in the
BamHI cleavage site, and to attach an EcoRI linker.
pSVA STOP1 thus has, between the SV40 early promotor and
the SV40 polyadenylation signal for early transcripts, a
cloning polylinke~r with three unique restriction sites
(HindIII-SaII-XbaI) and a sequence with translation
stops in all three reading frames.
J
~3 4~~ ~~
- 39 -
c) Construction of DHFR expression vectors for co-
transfection
The starting point for the DHFR vectors used for the co-
y transfection was the plasmid pMTVdhfr (Lee et al., loc.
cit.). pMTVdhfr was cut with BgIII, and the protruding
5' ends of the DNA were filled in using DNA polymerase
I (Klenow fragment). After digestion with EcoRI, a
fragment 4.47 kb in size was isolated and ligated with a
133 by EcoRI-HpaI fragment from pBB3 (Bourachot et al.,
Loc. cit.). The new plasmid pMTVAdhfr has the mouse
DHFR cDNA flanked by MMTV-LTR and the SV40 polyadenylation
site for early transcripts.
pSVOAdhfr was obtained from pMTVAdhfr by deletion of a
HindIII fragment which is 1450 by in size and has the
MMTV-LTR.
Neither pMTVAdhfr nor pSVOAdhfr have mRNA splice sites.
