A recessive mutant caused by insertional mutagenesis in transgenic mice has been detected in which the anterior part of the forelimb is disorganized. The morphology of the thumb is always altered and sometimes the adjacent finger has an extra phalanx. This phenotype suggests that a body plan gene is affected. We have named the mutation add (anterior digit-pattern deformity). Using the cloned DNA from the flanking region of the integrated transgene, add has been mapped close to the centromere of chromosome 13. This position links add to a genetically mapped locus called extra-toes (Xt). The phenotype of the double-mutant addjXt as well as the molecular analysis suggest that add and Xt are allelic.

Although several dysmorphic mouse mutations are described (Lyon and Searle, 1989), the molecular mechanisms responsible for organizing the mammalian body are unknown. The major reason for this is the inability to define molecularly the affected gene in such a mutant. Several attempts have been started in the last few years to bridge classical genetics and molecular biology. One approach is based on the fact that certain DNA sequence motifs are conserved between species. Drosophila homologues have been screened for in the mouse genome, and, following their isolation and chromosome mapping, a comparison is made to known mouse mutations. This approach strongly suggested the molecular identification of the mouse mutations undulated using the pax gene (Balling et al. 1988). Another approach, using a combination of genetic and molecular techniques like walking and jumping, resulted in the isolation of the T gene (Herrmann et al. 1990). However, the latter approach is only feasible when several mutants of the gene are available. A third approach employed by several laboratories involves screening for developmental mutants by insertional mutagenesis in transgenic mice (Jaenisch et al. 1983; Wagner et al. 1983; Woychik et al. 1985). In such an approach, the gene becomes marked by the integrated DNA and can be cloned out of the genome. We have also tested our transgenic mouse lines (Rüther et al. 1987a,b, Dente et al. 1988) for dominant and recessive mutations. Here we describe a recessive mutation (called add) that results in a disorganization of the anterior part of the digit pattern of forelimbs. The genetic and molecular analysis suggests that add is allelic to the morphogenetic mutation Xt, first described by Johnson (1967).

Mice

Mice of line 358–3 (Dente et al. 1988), which carry the add mutation, are bred in the mouse colony at the EMBL. This mutation is kept on a mixed C57BL/6×SJL background. Mice carrying the Xt mutation were obtained from MRC Radiobiology Unit, Chilton and from the Gesellschaft für Strahlenforschung, München.

Skeleton analysis

To visualize the limb skeleton, the following procedure was used: 24h in 95% ethanol, clearance of soft tissue in 2% KOH solution and staining of bones in 1 % KOH solution containing Alizarin red. Finally the stained limbs were dehydrated by incubation in 30 %, 60 % and 87 % glycerol for several hours.

In situ hybridisation to mouse chromosomes

Concavalin A-stimulated lymphocytes of a WMP male mouse in which all the autosomes except 19 are in a form of metacentric Robertsonian translocation were cultured at 37 °C for 72 h with 5-bromodeoxyuridine added for the final six hours of culture (60 μgml-1) to ensure a chromosomal R-banding of good quality. The Sst\ fragment used as probe in Fig. 2 was labelled as cloned insert in pUC19 by nicktranslation. Hybridization, autoradiography, staining and banding was performed as described (Mattei et al. 1985).

Fig. 1.

Phenotype of the forelimb deformation. (A) Left forepaw of a homozygous mouse of line 358–3 with the most obvious phenotype. (B–H) Skeleton of forelimbs of wild-type mouse (B) and homozygous mice of line 358–3 (C–H).

Fig. 1.

Phenotype of the forelimb deformation. (A) Left forepaw of a homozygous mouse of line 358–3 with the most obvious phenotype. (B–H) Skeleton of forelimbs of wild-type mouse (B) and homozygous mice of line 358–3 (C–H).

Fig. 2.

Mouse genome organization around the transgene integration site. (A) The structure of the normal allele is shown at the top of the figure. The 5’ integration site of the transgene copies is indicated below (only one copy is drawn). The Ssil fragment used as the probe in part B of the figure is indicated. The Hind III fragment cloned from mouse line 358–3 into the λ vector is also illustrated. (B) The polymorphism of the wild-type alleles (co) and of the affected alleles in heterozygous (he) and homozygous (ho) mice of line 358–3 is shown. For this purpose, genomic DNAs have been digested with the restriction enzyme Hindi and the Southern blots were hybridized with the nick-translated SstI fragment indicated in part A of the figure. Abbreviations: St, Ssil; He, Hindi’, Hd, Hind Ill; Nt, NotI.

Fig. 2.

Mouse genome organization around the transgene integration site. (A) The structure of the normal allele is shown at the top of the figure. The 5’ integration site of the transgene copies is indicated below (only one copy is drawn). The Ssil fragment used as the probe in part B of the figure is indicated. The Hind III fragment cloned from mouse line 358–3 into the λ vector is also illustrated. (B) The polymorphism of the wild-type alleles (co) and of the affected alleles in heterozygous (he) and homozygous (ho) mice of line 358–3 is shown. For this purpose, genomic DNAs have been digested with the restriction enzyme Hindi and the Southern blots were hybridized with the nick-translated SstI fragment indicated in part A of the figure. Abbreviations: St, Ssil; He, Hindi’, Hd, Hind Ill; Nt, NotI.

λ and cosmid libraries and DNA analysis

For the cloning of the integration site of the add mutation, DNA was isolated from homozygous mice of line 358–3 (Dente et al. 1988). This genomic DNA was completely digested with HindIII and cloned into the λ phage 1151 (Murray, 1983). The λ library was screened using as a probe the fl origin of replication which is part of the transgene. To obtain the preintegration site, a cosmid library of 129/Sv-S1CP mouse DNA was screened with the SstI DNA fragment isolated from the λ clone as shown in Fig. 2. Positive clones were characterised further as described (Rackwitz et al. 1985). Sequence analysis of preintegration site and the flankings of the transgene was performed following subcloning of suitable fragments into plasmid vectors.

Phenotypic characterization of the recessive mouse mutation add

By testing our transgenic mouse lines systematically for recessive mutations, we detected in the course of homozygote breeding in one of our lines (358–3, Dente et al. 1988) a deformation of digit 2 on the forepaw (Fig. 1A). The digit appeared bent and connected to the thumb, which is, in mice, only a rudiment. Molecular analysis of the transgene copy number revealed that all mice displaying this phenotype were homozygous for the transgene integration site. When we intercrossed homozygous animals, surprisingly we only observed a transmission of the phenotype to about 60 % of the offspring although all animals were clearly homozygous with respect to the copy number of the transgene (data not shown). To characterize the mutation in more detail, we isolated forelimbs from homozygous animals with and without visible deformations. To visualize the bony part, the soft tissue was cleared and the bones stained. Forelimbs of several dozen mice were treated in this way and have subsequently been analyzed microscopically (Fig 1BH). In all forelimbs isolated from homozygous animals of line 358–3, the thumb shows irregular growth; from slight changes in architecture (Fig. 1C) to very drastic alterations such as an elongation of the thumb by almost 100% (Fig. 1F,G). Only when the thumb is elongated to a certain extent does digit 2 become bent and connected to the thumb via soft tissue (Fig. 1EH). First appearance of this phenotype during embryonic development was obvious at day 14. Thus, digit formation and phenotypic changes are happening at the same time.

We have never found bone fusion of the thumb and digit 2, but we could detect in about 10% of the analyzed forelimbs an additional phalanx on digit 2 (Fig. 1H). Thus, the phenotype is polarized such that the structure primarily affected is the thumb, which is the most anterior part of the forelimb. The only other structure sometimes changed on the forelimb is the adjacent finger (digit 2). We have never detected any alteration on digits 3–5 of these mice. The mutated gene interferes with the organization of the anterior part of the digit pattern of forelimb, therefore we suggest add (anterior digit pattern deformity) as a name for this mutation.

Molecular characterization of add

To define the add mutation molecularly, we cloned the flanking sequences of the transgene from a λ library of homozygous add mice. As shown for one flanking side (here defined as 5’, Fig. 2), we could identify unique DNA sequences that can be used as a probe to demonstrate the restriction fragment length polymorphism of the affected add allele. Using this probe we have screened a cosmid library to obtain DNA sequences of the preintegration site. Sequencing of both transgenic flanking sites and that of preintegration reveals that the transgene has replaced only one base pair in the course of integration (Fig. 3). In addition, we have found that some unknown mouse DNA has cointegrated with several copies of the transgene.

Fig. 3.

Sequence of the integration site. At the top of the figure part of the DNA sequence of the wild-type chromosome 13 is presented. Several copies of the transgene DNA are integrated by replacing one base pair as shown in this diagram. The small letters are indicating unknown mouse sequences that cointegrated with the transgene.

Fig. 3.

Sequence of the integration site. At the top of the figure part of the DNA sequence of the wild-type chromosome 13 is presented. Several copies of the transgene DNA are integrated by replacing one base pair as shown in this diagram. The small letters are indicating unknown mouse sequences that cointegrated with the transgene.

To map the chromosomal localization of add, about one hundred chromosomal metaphases were hybridized with the same probe as used to detect the polymorphism (see legend Fig. 2). In the metaphase cells examined, 106 silver grains were associated with chromosomes and 48 of these (45.3%) were located on chromosome 13 (Fig. 4). The distribution of grains on this chromosome was not random: 85 % of them mapped to the [A1–A3] region with a maximum in the 13A2 band. Thus, the chromosomal localization of the add mutation is most likely the A2 band of chromosome 13. This position suggests a linkage between add and a genetically mapped locus called extra-toes (Xt), a morphogenetic mutation on the proximal part of chromosome 13 (Johnson, 1967; Lyon et al. 1967).

Fig. 4.

Localization of the add gene to mouse chromosome 13 by in situ hybridization. (A) The partial WMP mouse metaphase shows the specific site of hybridization to the proximal part of chromosome 13 (the arrow indicates the silver grains). (B) The chromosomes of the metaphase shown in A were subsequently identified by their characteristic R-banding. (C) Diagram of WMP mouse Rb (13; 15) chromosome, indicating the distribution of labelled sites.

Fig. 4.

Localization of the add gene to mouse chromosome 13 by in situ hybridization. (A) The partial WMP mouse metaphase shows the specific site of hybridization to the proximal part of chromosome 13 (the arrow indicates the silver grains). (B) The chromosomes of the metaphase shown in A were subsequently identified by their characteristic R-banding. (C) Diagram of WMP mouse Rb (13; 15) chromosome, indicating the distribution of labelled sites.

Phenotypic characterization of add/Xt double mutant

Although the Xt mutation shows differences compared to add (extra-toes on the hindlimb, dominant mutation), we wondered whether add and Xt were allelic. First, the phenotype on the forelimb is almost identical and, second, in the course of breeding several hundred add mice we found mice with an extra-toe on one hindlimb with a frequency of less than 1 %. Therefore, we mated homozygous add mice (males and females) with heterozygous Xt mice (Xt is homozygous lethal). 50 % of the offspring of such a mating should be normal (add/+) or with extra-toes (add/Xt). Indeed, 50% of offspring were normal, but the other 50% showed, besides the extra-toes, an enhanced alteration of fore- and hindlimbs. Skeletal analysis demonstrated the presence of seven digits on both fore- and hindlimbs (Fig. 5). Sometimes two of the additional digits were fused (Fig. 5B) but could still be counted as individual digits. To exclude any specific genetic background for the appearance of seven digits, we have backcrossed the double mutants. Mating of add/Xt with wild-type mice (C57BL/6×SJL F1) resulted in 50% normal mice (all positive for the transgene, when tested by Southern blot) and 50% mice with only one extra-toe per hindlimb (all negative for the transgene). We have tested several of the add/Xt double mutants for other skeletal changes but none could be identified. All the add/Xt or Xt/add mice were fertile and seem to have a normal lifespan (the oldest is now more than twelve months).

Fig. 5.

Phenotype of fore- and hindlimb alterations. Skeletons of (A) fore- and (B) hindlimbs isolated from a Xt/add double mutant are shown.

Fig. 5.

Phenotype of fore- and hindlimb alterations. Skeletons of (A) fore- and (B) hindlimbs isolated from a Xt/add double mutant are shown.

Molecular analysis of the Xt locus

Assuming that Xt and add are allelic, we were expecting to find restriction fragment length polymorphisms using probes either 5’ or 3’ of the transgene integration site. However, probing of DNA isolated from Xt heterozygous animals showed only the wild-type DNA pattern (Fig. 6). By testing DNA isolated from the add/Xt double mutant, we obtained the first indication of a deletion on the Xt chromosome (Fig. 6) as only the add restriction fragment length polymorphism was detectable. For further analysis we have chosen DNA isolated from Xt homozygous embryos. Although we have tested unique DNA probes isolated from cosmid clones that cover about 80 kilobases of DNA (≈kb 5’ and ≈30 kb 3’ of the transgene integration site), none hybridized to DNA from homozygous Xt mice (data not shown). Thus, the length of the deletion on the Xt chromosome is at least 80 kilobases.

Fig. 6.

Restriction fragment length polymorphism of genomic DNA isolated from different mouse genotypes. Genomic DNAs have been digested with BglII and the Southern blots were hybridized with the labelled H2-SstI probe (see Materials and methods).

Fig. 6.

Restriction fragment length polymorphism of genomic DNA isolated from different mouse genotypes. Genomic DNAs have been digested with BglII and the Southern blots were hybridized with the labelled H2-SstI probe (see Materials and methods).

In this report, we have described a recessive morphogenetic mouse mutation (called add) which always affects the organization of the most anterior part of the digit pattern of forelimbs. Occasionally, with a frequency of less than 1 % in addition to the forelimb alteration, we have found an extra-toe on the hindlimb. These findings, plus the fact that add could be mapped to the proximal part of chromosome 13 (13 A2) by in situ hybridization, suggested that add might be allelic to a genetically mapped mutation called extra-toes (Xt). The forelimb deformation in Xt mice (Johnson, 1967) is very similar to that of the add mutant, however, the Xt mice have one or more extra digits on the anterior part of the hindlimbs. Since Xt is a dominant mutation, every mouse has limb alterations on fore- and hindlimbs. Homozygous Xt mice die at birth or in utero (Johnson, 1967).

Our mating studies of double mutant add/Xt or Xt/add mice have already provided some evidence that add and Xt might be allelic. Each of these double mutant mice displayed a limb phenotype (seven digits on fore- and hindlimbs) which is reminiscent of Xt homozygous mice. As described by Johnson (1967), homozygous Xt embryos at about day 15 of embryonal development show seven digits or sometimes even more on fore- and hindlimbs. In addition, the homozygote Xt mice show multiple abnormalities in the skeleton e.g. vertebrae and thorax. A severe malformation of the brain, central nervous system and sense organs is also reported (Johnson, 1967). We could never detect any of these changes in the add/Xt mice.

Additional evidence is provided by molecular analysis of add. and Xt. We could characterize add as a mutation caused by the integration of several transgene copies by replacing one basepair of genomic sequence on chromosome 13. Using DNA probes specific for sequences 5’ and 3’ of the integration site, Xt was characterized as a deletion of at least 80 kilobases. Although we cannot demonstrate that both mutations are affecting the same transcript, a highly likely interpretation is that these mutations are allelic.

Assuming that both mutations are allelic, how can we explain that the deletion in Xt is causing a dominant, and the transgene insertion a recessive, mutation? In principle there are at least two possible explanations. First, one can assume that the presumptive protein encoded by the Xt/add gene has at least two functional domains. In the deletion only one domain is affected and therefore a normal functional protein complex is no longer possible, the phenotype being already present in the heterozygote. The insertion would knock out the gene and a phenotype is only visible in homozygous mice. A second possibility is that the Xt deletion is removing most of the gene, resulting in a 50% reduction of a protein product in the heterozygous mice, of which the precise concentration for its normal function is crucial (haploinsufficiency). Likewise, the insertion mutation add only reduces the expression of one allele but does not abolish it and therefore no phenotype can be detected in add/ + mice. However, in add/add mice, the reduction of protein is similar to Xt/ + mice and therefore the phenotype becomes obvious.

Characterization of certain phenotypes in Drosophila melanogaster provides a molecular explanation for dominant mutations. One of the most extensive studies has’ been performed with the act 88 F gene, which encodes actin III of the indirect flight muscles. A dominant mutation called KM88 is described, which abolishes mRNA and protein (Okomoto et al. 1986). This reduction of about 50% leads to a dominant phenotype due to a filament imbalance in the myofibrils (Beall et al. 1989). For the same gene, negative dominant mutations have also been described in which a truncated actin protein (due to either a deletion or a nonsense mutation) affects the assembly of myofibrils (Karlik et al. 1984; Okomoto et al. 1986).

One observation supports the speculation regarding a negative influence on a gene product concentration in the add mice. The phenotypical changes show a strong correlation with a certain genotype, which can actually be quantified, when we simplify the analysis and count only the number of digits: add/add, only one thumb is changed; Xt/+, in addition one extra-toe; add/Xt, two extra-toes and Xt/Xt, up to four extra-toes. It is tempting to speculate about the nature of this gene product and it is also unclear why Xt is homozygous-lethal. Homozygous Xt mice might either die because the gene product is essential in other processes or there might be other gene(s) within the deletion that are recessive mutated, not involved in limb formation, but essential for normal development.

At present we are attempting to detect transcript alterations in the Xt and/or add mice using DNA probes derived from the genomic region deleted in the Xt mice. In parallel, we have begun inducing further mutations on chromosome 13 by chemical mutagenesis of wild-type males and test mating with homozygous add females. By using chlorambucil as a reagent there is approximately a 1 % possibility of finding more mutations at this locus (Russell et al. 1989). With the help of different deletion end points, it should be possible to narrow down the region encoding this gene involved in the pattern formation of the digits. The ultimate proof of allelism of add and Xt will then be possible by performing a phenotype rescue experiment.

We thank Carol Murphy for critical reading of the manuscript and Helen Fry for typing. We also thank Erwin Wagner in whose laboratory the experiments were first started, Anne-Marie Frischauf for the 129 mouse cosmid library and Catherine Woodroofe for technical assistance. This work was partially supported by the ‘Deutsche Forschungsgemeinschaft’ (Schwerpunkt Musterbildung).

Balling
,
R.
,
Deutsch
,
U.
and
Gruss
,
P.
(
1988
).
undulated, a mutation affecting the development of the mouse skeleton, has a point mutation in the paired box of Pax 1
.
Cell
55
,
531
535
.
Beall
,
C. J.
,
Sepanski
,
M. A.
and
Fyrberg
,
E. A.
(
1989
).
Genetic dissection of Drosophila myofibril formation: effects of actin and myosin heavy chain null alleles
.
Genes and Develop.
3
,
131
140
.
Dente
,
L.
,
Rüther
,
U.
,
Tripodi
,
M.
,
Wagner
,
E. F.
and
Córtese
,
R.
(
1988
).
Expression of human al-acid glycoprotein genes in cultured cells and in transgenic mice
.
Genes and Develop.
2
,
259
266
.
Herrmann
,
B. G.
,
Labeit
,
S.
,
Poustka
,
A.
,
King
,
T. R.
and
Lehrach
,
H.
(
1990
).
Cloning of the T gene required in mesoderm formation in the mouse
.
Nature
343
,
617
622
.
Jaenisch
,
R.
,
Harbers
,
K.
,
Schnieke
,
A.
,
Lôhler
,
J.
,
Chumakor
,
I.
,
Jâhner
,
D.
,
Grotkopp
,
D.
and
Hoffmann
,
E.
(
1983
).
Germline integration of moloney murine leukemia virus at the Movl3 locus leads to recessive lethal mutation and early embryonic death
.
Cell
32
,
209
216
.
Johnson
,
D. R.
(
1967
).
Extra-toes: a new mutant gene causing multiple abnormalities in the mouse
.
J. Embryol. exp. Morph.
17
,
543
581
.
Karlik
,
C. C.
,
Coutu
,
M. D.
and
Fyrberg
,
E. A.
(
1984
).
A nonsense mutation within the Act88F actin gene discrupts myofibril formation in Drosophila indirect flight muscles
.
Cell
38
,
711
719
.
Lyon
,
M. F.
,
Morris
,
T.
,
Searle
,
A. G.
and
Butler
,
J.
(
1967
).
Occurrences and linkage relations of the mutant ‘extra-toes’ in the mouse
.
Genet. Res.
9
,
383
385
.
Lyon
,
M. F.
and
Searle
,
A. G
(eds.) (
1989
).
Genetic Variants and Strains of the Laboratory Mouse.
Gustav Fischer Verlag: Stuttgart
,
New York
.
Mattei
,
M. G.
,
Philip
,
N.
,
Passage
,
E.
,
Moisan
,
J. P.
,
Mandel
,
J. L.
and
Mattei
,
J. F.
(
1985
).
DNA probe localization at 18pll3 band by in situ hybridization and identification of a small supernumerary chromosome
.
Human Genetics
69
,
268
271
.
Murray
,
N. E.
(
1983
). In
Lambda II
(ed.
R. W.
Hendrix
,
J. W.
Roberts
,
F. W.
Stahl
,
R. A.
Weisberg
), pp.
395
398
.
Cold Spring Harbor Laboratory
:
New York
.
Okomoto
,
H.
,
Hiromi
,
Y.
,
Ishikawa
,
E.
,
Yamado
,
T.
,
Isoda
,
K.
,
Maekawa
,
H.
and
Hotte
,
Y.
(
1986
).
Molecular characterization of mutant actin genes which induce heat-shock proteins in Drosophila flight muscles
.
EMBO J.
5
,
589
596
.
Rackwitz
,
H. R.
,
Zehetner
,
G.
,
Murialdo
,
H.
,
Delius
,
H.
,
Chai
,
J. H.
,
Poustka
,
A.
,
Frischauf
,
A.
and
Lehrach
,
H.
(
1985
).
Analysis of cosmids using linearization by phage lambda terminase
.
Gene
40
,
259
266
.
Rüther
,
U.
,
Garber
,
C.
,
Komitowski
,
D.
,
Müller
,
R.
and
Wagner
,
E. F.
(
1987a
).
Deregulated c-fos expression interferes with normal bone development in transgenic mice
.
Nature
325
,
412
416
.
Rüther
,
U.
,
Tripodi
,
M.
,
Córtese
,
R.
and
Wagner
,
E. F.
(
1987b
).
The human alpha-l-antitrypsin gene is efficiently expressed from two tissue-specific promotors in transgenic mice
.
Nucl. Acids Res.
15
,
7519
7529
.
Russel
,
L. B.
,
Hunsicker
,
P. R.
,
Cacheiro
,
N. L. A.
,
Bangham
,
J. W.
,
Russel
,
W. L.
and
Shelby
,
M. D.
(
1989
).
Chlorambucil effectively induces deletion mutations in mouse germ cells
.
Proc, natn. Acad. Sci. U.S.A.
86
,
3704
3708
.
Wagner
,
E. F.
,
Covarrubias
,
L.
,
Stewart
,
T. A.
and
Mintz
,
B.
(
1983
).
Prenatal lethalities in mice homozygous for human growth hormone gene sequences integrated in the germ line
.
Cell
35
,
647
655
.
Woychik
,
R. P.
,
Stewart
,
T. A.
,
Davis
,
L. G.
,
D’eustachio
,
P.
and
Leder
,
P.
(
1985
).
An inherited limb deformity created by insertional mutagenesis in a transgenic mouse
.
Nature
318
,
36
40
.