Note: Descriptions are shown in the official language in which they were submitted.
201~7~
-- 3
D e s c r i ~ t i o n
The invention concerns a method for the targetted change
of a gene within the genome of intact mammalian cells by
homologous recombination.
The investigation of the causes of different genetic
diseases and the localization of the defects at the
molecular level which produce the disease is progressing
with increasing rapidity thanks to the application of
genetic engineering which is becomming more precise and
reliable. Nevertheless, the function of many genes of
interest, particularly in mammalian cells, has not yet
been elucidated because, inter alia, no satisfactory
methodology yet exists for the mutagenesis of intact
mammalian cells. Moreover, the same methodology would be
necessary to correct genetic defects which are already
known.
Up to now, in first attempts at this in mammalian cells,
when a defective gene is present attempts have been made
to replace its function by insertion of a corresponding
functional gene at another site, however, in this
process other changes which are active over long periods ~`
may be caused at a site other than that which is to be
mutated which is why this method is ruled out for a
possible treatment of disease in mammalian cells.
Another major disadvantage of the experimental methods -
up to now for mutagenesis by homologous recombination in
mammalian cells is that the introduction of a marker -
gene is necessary to detect the resultant mutation.
These marker genes, which are foreign genes at this site
or, in general, in the genome, must also be expected to
201~7~
-- 4 --
result in undesired side effects with a high degree of
probability.
In addition, the success rate of the methods used up to
now for in vivo mutagenesis, also known as "gene
targetting" is relatively low in mammalian cells.
It is therefore one of the major aims for the further
development of genetic engineering and also the object -
of the present invention to provide a method which
enables the direct alteration of a specific part of the
mammalian cell genome in living cells without at the -~
same time involving the risk that other unintentional
mutations are caused or that foreign sequences have to
be inevitably introduced.
This object is achieved according to the present
invention by a method for the targetted change of a gene
within the genome of intact mammalian cells by
homologous recombination, in which a DNA sequence, which
is homologous to the gene to be changed but, however,
differs by at least one nucleotide from its sequence as
a result of mutation, deletion, or insertion, is
introduced into appropriate cells by microinjection and
those cells are isolated in which the resultant change
can be detected.
The method according to the present invention results
surprisingly in a particularly large proportion of
homologous recombinations at the desired sites whereby a
homologous recombination occurred in more than one in
640 cells. Even a recombination rate of 1 per 150
microinjected cells could already be achieved with the
method according to the present invention. This is an
, :~
2~7~
-- 5 --
exceptional increase in the efficiency compared to the
relation of homologous to illegitimate recombinations in
mammalian cells using previous methods. Furthermore, the
resultant change can be detected in the method according
to the present invention without having to introduce a
marker gene and without a corresponding selection by
means of this marker gene. The change can be detected
after cloning the microinjected cells by Southern
blotting or also sequencing.
In a particularly preferred embodiment of the invention
the detection of the resultant change is carried out
using the PCR (polymerase chain reaction) method (R.X.
Saiki et al., Science 239, (1988) 487-491). Using this,
it can be established whether the desired mutation has
taken place without prior cloning of the cells. In this
case the method is so sensitive that even a few DNA
molecules can be reliably analyzed without the need for
a large number of cells. The method can even be applied
in such a way that point mutations are detectable.
Two primers are used for the detection, one of which is
complementary to the position of the DNA sequence
differing from the gene. An amplification from this
primer can only occur and be detected when a
recombination has taken place.
Every natural DNA sequence which has been isolated and,
if desired, cloned or also DNA sequences produced
synthetically can be used as the DNA sequence which is
homologous to the gene to be changed but which differs
in at least one position from the sequence of this gene.
In this process one has assumed up to now that the
length of the homology is directly proportional to the
frequency of the occurrence of the gene conversion
2011'~
6 --
whereas the length of the heterology is inversely
proportional to it. It is therefore particularly
preferable to use a DNA sequence which only differs from
the genomic sequence to be changed at the required
positions but which has as large a homology as possible
from which it can be concluded that the rate of
recombination is higher the longer the DNA sequence
used.
In a preferred embodiment of the invention somatic stem
cells are used as mammalian cells. Previous knowledge of ~
such stem cells and gene transfer into such cells is ;
summarized for example by P.M. Dexter and D. Boettinger,
Gene Transfer into Hemapoetic Cells by Retroviral
Vectors, ISI, Atlas of Science, 167-174. After
introduction of the cells into a mammal, the change
according to the present invention of the genome of
somatic stem cells enables an existing and known defect
to be rectified since the stem cells differentiate to
functional cells and defective populations in the body
are replaced successively. Blood defects such as factor
VIII deficiency (haemophilia) are examples of the
application for the correction of defective genes by
(re-)introduction of mutated stem cells according to the
present invention~ A further possibility for an
application would result if the described defects in the
insulin gene could be corrected in stem cells which
differentiate to cells of the islet of Langerhans so
that here there is a chance to combat diabetes at a
genetic level. Somatic cells are either kept in culture
on a so-called "feeder cell" layer e.g. a layer of
mitotically inactive cells incapable of division or
which have been prevented from differentiating with the -
aid of growth factors (DIA: A.G. Smith et al., Nature,
- 20117~
Vol. 336, (1988), Page 688, LIF: R.L. Williams et al.,
Nature 336 (1988), page 684) and thus enable a culture.
Applications of the method according to the present
invention which for example are directed at increasing
the secretion of growth ~actors in animals appear of
interest for agriculture and in particular for the
breeding of useful animals whereby in animal breeding
the mutated gene can then be introduced into the germ
track for instance by inject`ion of zygotes into a
fertilized egg or by use of embryonic stem cells.
A further important possibility for an application of
the method according to the present invention using
somatic or embryonic stem cells and, if desired, the
injection of zygotes into a fertilized eggj is the
opportunity which thus presents itself to investigate
the function of hitherto unknown genes in animal models.
Thus, for example, stem cells from mice, which contain a
gene mutated at a particular site can be injected into a
fertilized egg and the effects on a chimeric animal
obtained in this way can be examined. A further
exceptionally important possibility for an application
comprises the introduction of cells mutated according to
the present invention into the germ track of animals
such as e.g. mice to generate a mouse model for diseases
diagnosed in humans. By this means it would be possible
to investigate the effect of pharmaceutical agents in a
mouse model in which this mouse model would be very
meaningful for the application in man.
For the future the method according to the present
invention also appears to offer a possibility to
cultivate improved or changed e.g. resistant plants. For
this purpose the same method can, in principle, be
201~ 73~
-- 8
applied as for mammalian cells as soon as a means has
been created for the efficient introduction of DNA into
such plant cells by microinjection and for a homologous
recombination.
In summary it is possible by means of the method
according to the present invention in a hitherto
unexpected and surprising manner to carry out a change
at the target site in living cells and not at a
homologous site, and to carry out a correction in
somatic gene therapy in which no additional foreign DNA
sequences whatsoever are introduced into the cells and
no long-term defects need to be feared.
The following Examples in conjunction with the Figures
elucidate the invention further.
Fig. 1 shows a diagram of the modification of the
Hox 1.1 gene;
Fig. 2 shows a Southern analysis of the genomic DNA
amplified by PCR from microinjected 3T3- (a) and
D3-cell pools (b) and a diagram of the PCR
strategy (c):
Fig. 3 shows a Southern analysis of genomic DNA from
cloned embryonic stem cells (a, b, c) digested
by KpnI and a diagram of the mutated and the
normal Hox 1.1 allele as well as the expected
restriction fragments; and
;~
Fig. 4 shows a Southern blot of the DNA amplified by -~
PCR from tail biopses from two litters of
genetically altered mice.
20117~
E x a m p 1 e
Mutagenesis of the Hox 1.1 gene.
A Fspl fragment of the Hox 1.1 gene (nucleotide 367 to
1937, numbered from the start codon (see Fig. 1) (A.M.
Colberg-Poley et al., Nature 314 (1985), 713-717, M.
Kessel et al., Proc. Natl. Acad. Sci. USA 84 (1987)
5306-5310) was cloned in Bluescript (Stratagene) and a
20 bp oligonucleotide was inserted into the EcoRI site
of the homeobox. The sequence of the inserted
oligonucleotide was AAT TGT GAG GTA CCG CTG AC. This
construct shown in Fig. 1, line 2 was microinjected in
an injection volume of 10-11 ml by microinjection into
the nucleus of NIH-3T3 cells (200 to 600 cells) or D3
cells (T.C. Doetschmann et al., J. Embryol. exp. Morph.
87 (1985) 27-45) (50 to 200 cells) at a concentration of
300 ng/ml which corresponds to 5 molecules of DNA per
cell. After the microinjection the cells were allowed to
grow for 4 to 7 days, trypsinized and applied again onto
the same plate. The cells were collected again after 4
to 7 days. Half of the cells were used for DNA analysis,
the remainder was stored in liquid nitrogen. The genomic
DNA or a control plasmid (1 ~g or 15 pg) were subjected
to 30 PCR cycles using heat-stabilized taq polymerase
(Saiki, R.K. et al., Science 239, 487-491 (1988), ~ ~-
Letson, A. and Liskay, R.M., Genetics, 117, 1987, 759
769) in a 50 ~1 reaction mixture containing 10 %
dimethylsulphoxide, 67 mM Tris-HCl (pH 8.8), 16.6 mM -~
NH4SO4, 6.7 mM MgC12, 10 mM 2-mercaptoethanol, 170 ~g/ml
bovine serum albumin (BSA), 450 ~M of each of the four -~
deoxyribonucleoside triphosphates (dATP, dCTP, dTTP,
dGTP) and 0.5 ~M of two primers. The first primer was a ~
synthetic oligonucleotide with the sequence TTC CGC ATC ~-
TCA CCC TGG AT, which is specific for a sequence in the ~
20117~
-- 10 --
;~ .
first exon of the Hox 1.1 gene and which binds on the
5' side of the Fspl cleavage site, and the second primer
has an identical DNA sequence to the above-mentioned
synthetic oligonucleotide. The samples were overlayered
with paraffin and heated in each cycle to 91.5C for one
minute in order to denature the DNA, cooled to 55C for
one minute in order to allow the primer to bind and
heated to 70C for 6 minutes to activate the polymerase.
Since the normal Hox 1.1 allele contains no binding site
for the second primer and the microinjected fragment
contains no binding site for the first primer, the
sought-after fragment can only be amplified by the
polymerase chain reaction in cells mutated according to
~1 the present invention. Aliquots of the reaction mixture
(10 ~1) were subjected to electrophoresis on a 1 %
agarose gel, transferred onto Gene Screen Plus (DuPont)
and examined in 50 % formamide, 1 M NaCl and 1 % SDS
with a hybridization probe which was labelled with 32p
to an activity of lxlO9 cpm/~g using a Multiprime kit
(Amersham) and used with 5x105 cpm/ml. The filters were
washed in 2xSSC, then in 2xSSC-1 % SDS and finally in
O.lxSSC (lxSSC: 8.765 g/l NaCl, 4.41 g/l sodium citrate,
pH 7.0) at 65C. Fig. 2 shows an autoradiogram of the
dried filters; part a after 2 hours and part b after 5
hours exposure. In this connection, Fig. 2a shows a
Southern analysis of genomic DNA amplified by PCR from
microinjected 3T3 pools (tracks 1 to 3) and of the
plasmid pH l.1/O-BF in track 4 as control. This plasmid
contains the entire Hox 1.1 gene and the inserted -~
oligonucleotide. In part b of the diagram a Southern
analysis of genomic DNA from microinjected D3 cell pools
amplified by PCR is shown. Fig. 2c shows a diagram of
the plasmid pH1.1/0-BF. A 3643 base pair long 5-BamHI-
FspI fragment of the Hox 1.1 gene (nucleotide - 1711 to
+ 1932 in relation to the translation start codon) was
cloned in Bluescript (Stratagene) and the intervening
. ~:
~'
. . . .
~ ,, ~ : - ~: , : . .
201~7g~
oligonucleotide was cloned in the EcoRI site of the
homeobox.
E x a m p 1 e 2
Further investigation of the stem cells
Two cell pools of the embryonic stem (ES) cells D3 from
Example 1 (G.R. Dressler and P. Gru~, Trans Genet. 4,
214-219 (1988), T.C. Doetschmann et al., J. Embryol.
Exp. Morph. 87 (1985), 27-45) were used for the further
analysis. Single cell clones from these pools were
picked out by drawing up individual cells with glass
capillaries (M.A. Rudnicki ~ M.W. McBurney in
Teratocarcinomas and Embryonic Stem Cells, (ed. E.J.
Robertson) 19-49 (Publisher, Town, 1987). Clonal cell
lines were prepared from these cells by plating first on
plates with 96 wells, then on plates with 24 wells of
1.5 cm diameter and then in 6 cm tissue culture plates.
DNA was isolated from half of these cell lines; the
other half was stored in liquid nitrogen. During the ~
entire cloning procedure the ES cells were kept in ~ ~-
culture on so-called "feeder cells". Aliquots (10 ~g) of
the genomic DNA were digested with KpnI, subjected to -~
electrophoresis on 0.8 % agarose gels and transferred to
Gene Screen Plus. The filters were hybridized with -~
32P-labelled probes as described in Example 1 and washed
as described there. Since the inserted oligonucleotide
has a KpnI restriction site, a new 5.4 kb band was
expected to occurr in homologous recombinant clones in
addition to the 14 kb band of the normal allele after
hybridization with a probe (probe 1) specific for the
first exon in Fig. 2d. The occurrence of this band is
shown in track 1 of Fig. 2a. The ES cell DNA was
contaminated with feeder cell DNA because the ES cells
.~ ' ~'.
20117~
- 12 -
were prevented from forming large colonies during the
sub-cloning. Therefore the 14 kb Hox 1.1 fragment not
only represents the wild-type allele from ES cells but
also wild-type alleles from the feeder cells. Altogether
161 individual clones were analyzed by Southern blotting
from pool 1 and 165 clones from pool 7. In order to
confirm and to enlarge on the Southern analysis clones
from both pools were cultured until large ES colonies
formed on the feeder cells. The relative amounts of ES
cell DNA were then very much larger as can be seen from
Fig. 3b and c. A Southern blot of DNA digested with KpnI
from one of the clonal lines derived from pool 1 was
hybridized with probe 1 (Fig. 3b, line 1) and probe 2
(Fig. 3b, track 2). Details of the probes are shown in
Fig. 3c. The relative intensities of the 5.4 kb and the
8.6 kb bands correspond to the 5' and 3' fragments (see
Fig. 3d) in comparison with that of the 14 kb fragment
which shows that at least 50 % of the ES cells contain
the mutated allele. DNA from a clonal line from pool 7
was digested with StuI/KpnI (Fig. 3c, track 1) and KpnI
(Fig. 3c, track 2) and Southern blots were hybridized in
succession with probe 3 (see Fig. 3d). The 600 bp and
4.27 kb fragments (Fig. 3c, track 1) which form are
derived from mutated and normal alleles (see Fig. 3 for
further details). The mutated and the normal allele are
present in the same amounts in the StuI/KpnI and the
KpnI digestion mixture (Fig. 3c, tracks 1 and 2) which
shows that the ES cells from pool 7 are clonal. An
analysis of the karyotype of the clone derived from pool
1 showed that 70 % of the cells contained 40
chromosomes. It is therefore assumed that the low
frequency of occurrence of the mutated allele in this
cell line is a consequence of contamination by non~
mutated ES cells and does not indicate a chromosomal ~
instability. It was estimated that the ratio of ~ ;
homologous to illegitimate recombinations is 1:30 since
. . .
: :~
20117~
.:
~ 13 -
.~
~j about 20 % of the microinjected cells were stably
,,r, transformed after nuclear microinjection (M.R. Capecchi,
Cell 22 (1980) 479-488). The homologous recombination
took place in 1 out of 150 microinjected cells. This is
surprisingly high compared to comparable investigations
(K.R. Thomas & M.R. Capecchi, Cell 51 (1987) 503-512,
0. Smithies et al., Nature 317 (1985) 230-234, K.R.
Thomas et al., Cell 44 (1986), 419-428).
~ E x a m p 1 e 3
: ~i
The embryonic stem cell lines obtained in Example 1 and
shown in Fig. 3a and 3b (litter 1: tracks 1 to 4 of Fig.
4, litter 2: tracks 5 to 7 of Fig. 4) were introduced
into blastocytes of C57BL/6 mice and two litters were
obtained with a total number of 4 chimeric animals. In
order to confirm that the chimeric animals contain the
~ mutated allele a biopsy was taken from the tails of the
;~ animals and 100 ng of genomic DNA from these biopsies
was subjected to 30 PCR cycles as described in Example
1. The tracks 2, 5, 6 and 7 of Fig. 4 show an
amplification of the 1.1 kb fragment which exhibits the
mutated allele (see Fig. 1 and Fig. 2).
. '.~
....
. . ~
::.
.
:
~ '' '.'
;
: :
'.' , ~
"' :~
-`. ` ,
