Missense Mutations in the Homeodomain of HOXD13 Are Associated with Brachydactyly Types D and E

Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, and Department of Plastic and Reconstructive Surgery, Radcliffe Infirmary, Oxford, United Kingdom.
The American Journal of Human Genetics (Impact Factor: 10.93). 04/2003; 72(4):984-97. DOI: 10.1086/374721
Source: PubMed
ABSTRACT
HOXD13, the most 5' gene of the HOXD cluster, encodes a homeodomain transcription factor with important functions in limb patterning and growth. Heterozygous mutations of human HOXD13, encoding polyalanine expansions or frameshifts, are believed to act by dominant negative or haploinsufficiency mechanisms and are predominantly associated with synpolydactyly phenotypes. Here, we describe two mutations of HOXD13 (923C-->G encoding Ser308Cys and 940A-->C encoding Ile314Leu) that cause missense substitutions within the homeodomain. Both are associated with distinctive limb phenotypes in which brachydactyly of specific metacarpals, metatarsals, and phalangeal bones is the most constant feature, exhibiting overlap with brachydactyly types D and E. We investigated the binding of synthetic mutant proteins to double-stranded DNA targets in vitro. No consistent differences were found for the Ser308Cys mutation compared with the wild type, but the Ile314Leu mutation (which resides at the 47th position of the homeodomain) exhibited increased affinity for a target containing the core recognition sequence 5'-TTAC-3' but decreased affinity for a 5'-TTAT-3' target. Molecular modeling of the Ile314Leu mutation indicates that this mixed gain and loss of affinity may be accounted for by the relative positions of methyl groups in the amino acid side chain and target base.

Full-text

Available from: Henk Giele
Am. J. Hum. Genet. 72:984–997, 2003
984
Missense Mutations in the Homeodomain of HOXD13 Are Associated
with Brachydactyly Types D and E
David Johnson,
1,2,*
Shih-hsin Kan,
1,*
Michael Oldridge,
1
Richard C. Trembath,
4
Philippe Roche,
3
Robert M. Esnouf,
3
Henk Giele,
2
and Andrew O. M. Wilkie
1
1
Weatherall Institute of Molecular Medicine, John Radcliffe Hospital,
2
Department of Plastic and Reconstructive Surgery, Radcliffe Infirmary,
and
3
Division of Structural Biology, The Henry Wellcome Building for Genomic Medicine, Oxford, United Kingdom; and
4
Division of
Medical Genetics, Department of Genetics and Medicine, Leicester, United Kingdom
HOXD13, the most 5
gene of the HOXD cluster, encodes a homeodomain transcription factor with important
functions in limb patterning and growth. Heterozygous mutations of human HOXD13, encoding polyalanine
expansions or frameshifts, are believed to act by dominant negative or haploinsufficiency mechanisms and are
predominantly associated with synpolydactyly phenotypes. Here, we describe two mutations of HOXD13 (923CrG
encoding Ser308Cys and 940ArC encoding Ile314Leu) that cause missense substitutions within the homeodomain.
Both are associated with distinctive limb phenotypes in which brachydactyly of specific metacarpals, metatarsals,
and phalangeal bones is the most constant feature, exhibiting overlap with brachydactyly types D and E. We
investigated the binding of synthetic mutant proteins to double-stranded DNA targets in vitro. No consistent
differences were found for the Ser308Cys mutation compared with the wild type, but the Ile314Leu mutation
(which resides at the 47th position of the homeodomain) exhibited increased affinity for a target containing the
core recognition sequence 5
-TTAC-3
but decreased affinity for a 5
-TTAT-3
target. Molecular modeling of the
Ile314Leu mutation indicates that this mixed gain and loss of affinity may be accounted for by the relative positions
of methyl groups in the amino acid side chain and target base.
Introduction
The HOX genes (related to homeobox genes in the Dro-
sophila melanogaster HOM-C cluster) encode a family
of highly conserved transcription factors that play key
roles in embryonic development. In humans, there are
four HOX gene clusters (HOXA, HOXB, HOXC, and
HOXD) that are located on different chromosomes and
together comprise a total of 39 genes (Scott 1992; Kosaki
et al. 2002). The order of genes within each cluster re-
flects their temporal and spatial expression patterns dur-
ing development. In general, genes expressed early in
development and in anterior and proximal regions are
located at the 3
end of the cluster, whereas genes ex-
pressed later and in more posterior and distal regions
are located at the 5
end (reviewed by Krumlauf 1994).
HOXD13 is the most 5
member of the HOXD cluster
and, with homologues numbered D9–D12, shows greatest
similarity to the D. melanogaster Abd-B gene (Scott
1992). These so-called posterior HOXD genes and their
paralogues in the HOXA cluster play critical roles in
Received December 5, 2002; accepted for publication February 4,
2003; electronically published March 14, 2003.
Address for correspondence and reprints: Dr. A. O. M. Wilkie,
Weatherall Institute of Molecular Medicine, John Radcliffe Hospital,
Oxford OX3 9DS, United Kingdom. E-mail: awilkie@molbiol.ox.ac.uk
* These authors contributed equally to this work
2003 by The American Society of Human Genetics. All rights reserved.
0002-9297/2003/7204-0019$15.00
limb development (Dolle´ et al. 1993; reviewed by Rijli
and Chambon 1997; Za´ka´ny et al. 1997).
The homeobox region of each HOX gene encodes a
domain of 60 amino acids, termed the “homeodomain,”
which comprises a flexible N-terminal region followed
by three a-helices (I, II, and III) and is highly conserved
among eukaryotes (Bu¨ rglin 1994; Banerjee-Basu and
Baxevanis 2001; Banerjee-Basu et al. 2001). Homeo-
domains bind DNA, either alone or in heteromeric com-
plexes with partner proteins (Mann and Affolter 1998);
their corresponding DNA-binding elements and down-
stream genetic targets are gradually being elucidated
(Banerjee-Basu et al. 2001; Mann and Carroll 2002).
In the case of HOXD13, the preferred core-binding se-
quence is reported to be 5
-TTAC-3
, although 5
-TTAT-
3
is also bound by posterior HOX proteins (Shen et al.
1997b). Crystal structures of several homeodomain-
DNA complexes have been obtained (reviewed by Gehr-
ing et al. 1994; Billeter 1996). These indicate that helix
III is essential for sequence-specific DNA binding by
making several contacts in the major groove of DNA;
additional minor groove contacts are made by the N-
terminal region of the homeodomain. The contacts
made by helix III of the D. melanogaster homeodomain
protein Ultrabithorax (Ubx) (the orthologue of the sev-
enth vertebrate HOX genes; Scott 1992) to the core
DNA consensus sequence 5
-TTAT-3
are shown in figure
1 (Passner et al. 1999). A similar pattern of contacts is
observed for DNA binding by D. melanogaster Anten-
Page 1
Johnson et al.: HOXD13 Mutations in Brachydactyly 985
Figure 1 Simplified representation of binding interactions be-
tween D. melanogaster Ubx protein and a double-stranded DNA target
(Passner et al. 1999). At the top, I47, Q50, N51, and M54 are the
four amino acids—Ile, Gln, Asn, and Met, respectively (numbered
according to their position in the homeodomain)—that contact specific
bases, either through hydrogen bonds (solid lines) or hydrophobic
interactions (dashed lines). The black dot represents a water molecule.
Designation of the a and b strands follows Billeter (1996). The 5
-
TTAT-3
motif on the b-strand of the DNA target is shown in bold.
On the a-strand only, the connecting phosphate (P) groups are included
to demonstrate the position (arrow) of the ionic interaction with Arg
(R53). Key amino acids in human HOXD13 are identical to those in
Ubx, except that M54 is replaced by V54 (Val).
napedia (Antp) (the orthologue of the sixth vertebrate
HOX genes), except that the core consensus sequence
is 5
-TAAT-3
(Fraenkel and Pabo 1998).
To date, intragenic mutations in only two HOX genes
have been confirmed to cause human disorders. Mu-
tations in HOXD13 (located on chromosome 2q31;
Ensembl) were first identified in synpolydactyly (SPD
[MIM 186000]) (Muragaki et al. 1996), and mutations
in HOXA13 cause hand-foot-uterus syndrome (HFUS
[MIM 140000]) (Mortlock and Innis 1997). SPD is a
rare autosomal dominant limb disorder with incomplete
penetrance and variable expressivity (Sayli et al. 1995),
typically comprising 3/4 finger syndactyly and 4/5 toe
syndactyly with partial or complete digit duplication
within the syndactylous web (Muragaki et al. 1996;
Goodman et al. 1997; reviewed by Goodman 2002). The
causative mutations are heterozygous expansions of a 15-
residue polyalanine tract encoded by an imperfect tri-
nucleotide repeat in exon 1 of HOXD13. Longer ex-
pansions are associated with greater penetrance and
phenotypic severity (Goodman et al. 1997). Analysis of
an equivalent mouse model, Hoxd13
spdh
, suggests a dom-
inant negative mechanism of action by interfering with
the function of the remaining wild-type HOXD13 pro-
tein and other 5
HOXD proteins (Bruneau et al. 2001).
The mutant protein has pleiotropic actions, which affect
patterning, joint formation, and chrondrocyte differen-
tiation during development (Albrecht et al. 2002).
A distinct phenotype, including a novel foot malfor-
mation with partial duplication of the base of the second
metatarsal in the first web space and a broad hallux, was
described in three families with frameshifting deletions
of HOXD13 (Goodman et al. 1998; Calabrese et al.
2000). These mutations are predicted to lead to a trun-
cated protein missing part or all of the homeodomain;
hence, the phenotype may be caused by haploinsuffi-
ciency. A family segregating a missense mutation in the
homeodomain of HOXD13, Arg298Trp, also exhibits
this phenotype (Debeer et al. 2002). Finally, heterozy-
gous deletions including HOXD13 have been described
in split hand–split foot malformation, but the critical
region for this phenotype probably lies outside the
HOXD cluster (reviewed by Goodman 2002). Although
brachydactyly (shortening of the digits) is prominent in
both the mouse mutants (Dolle´ et al. 1993; Bruneau et
al. 2001) and in rare human SPD homozygotes (Mura-
gaki et al. 1996), only minor manifestations have pre-
viously been associated with heterozygous HOXD13
mutations (fifth-finger clinodactyly, hypoplasia or ab-
sence of the middle phalanges of toes 2–5, and occasional
shortening of the first metacarpal; reviewed by Goodman
2002).
In view of the diversity of limb malformations asso-
ciated with HOXD13 mutations, we performed a mu-
tation screen of HOXD13 in 128 consecutive patients
with unselected congenital limb abnormalities who re-
quired reconstructive surgery. In two subjects, we
identified a novel mutation in the homeodomain of
HOXD13, Ile314Leu. Detailed characterization of the
phenotype associated with these mutations delineated a
distinctive polydactyly/brachydactyly pattern, which led
us to investigate a family reported elsewhere to have
brachydactyly type E (Brailsford 1945; Oude Luttikhuis
et al. 1996). A different missense mutation in the homeo-
domain, Ser308Cys, was found to segregate in the latter
family. We have investigated the effect of these muta-
tions on binding to double-stranded (ds) oligonucleo-
tides in vitro and report that the Ile314Leu mutation
exhibits mixed gain and loss of affinity to different
targets.
Patients and Methods
Patients
Approval for the study was obtained from the Central
Oxford Research Ethics Committee. Venous blood sam-
ples were obtained at the time of surgery from a con-
secutive series of 128 patients with congenital limb ab-
normalities presenting to the Department of Plastic and
Reconstructive Surgery, Oxford; from selected relatives;
and from an additional family reported elsewhere as
being affected with brachydactyly type E (Brailsford 1945;
Oude Luttikhuis et al. 1996). Additional buccal samples
from relatives were obtained using Cytosoft cytology
brushes (Gentra Systems). Genomic DNA was isolated
either from venous blood samples by phenol/chloroform
extraction or from buccal cells using the Puregene DNA
Page 2
986 Am. J. Hum. Genet. 72:984–997, 2003
Isolation Kit (Gentra Systems). Metacarpophalangeal
profiles (MCPP) were analyzed from direct measure-
ments taken from upper limb radiographs in families A
and B using ANTRO software (Garn et al. 1972; Poz-
nanski et al. 1972; Armour et al. 2000).
PCR Amplification and Analysis
Oligonucleotides were designed from HOXD13 ge-
nomic and cDNA sequences (GenBank accession num-
bers AF005219, AF005220, AC009336, and NM_
000523). DNA numbering starts from the first nucleotide
of the ATG start codon. Primers for genomic PCR were
synthesized by Sigma-Genosys. Exon 1 was amplified us-
ing two overlapping primer pairs: 1aF 5
-ATGAGCCGC-
GCCGGGAGCTGGGAC-3
and 1aR 5
-CGAGGCGTG-
CGGCGATGACTTGAGCG-3
, 1bF 5
-TACCACTTC-
GGCAACGGCTACTACAGCTGC-3
and 1bR 5
-GCA-
CAACTCCCACTCCCAAGTAGGGG-3
, which gener-
ated fragments of 474 bp and 501 bp, respectively. Exon
2 was amplified using the primer pair 2F 5
-CTAGGTG-
CTCCGAATATCCCAGCCT-3
and 2R 5
-AAGCTGT-
CTGTGGCCAACCTGGA-3
, which generated a frag-
ment of 333 bp. Reactions took place in a volume of
25 ml that contained 40 ng genomic DNA, 1.5 mM
MgCl
2
, 120 mM dNTPs, 0.4 mM primers, 1 # GeneAmp
PCR Buffer II, 0.5 U AmpliTaq Gold DNA polymerase
(Applied Biosystems), and 0.05 U Pwo polymerase
(Roche). Ten percent DMSO was included in the reac-
tion mix for exon 1 amplifications. Thermocycling was
performed on an MJ Research PTC-200 and consisted
of 94C for 10 min, followed by 35 cycles of 94Cfor
45 s, 62C annealing for 45 s, 72C for 30 s, and a final
step of 72C for 4 min, except that an annealing tem-
perature of 56C was used for the 1aF-1aR primer pair.
The samples were heteroduplexed by heating to 95C,
followed by slow cooling, and analyzed by denaturing
high-performance liquid chromatography (DHPLC) on
the WAVE DNA Fragment Analysis System (Transgeno-
mic) at the following temperatures: 1aF-1aR at 64C,
67C, and 70C; 1bF-1bR at 62C and 64C; and 2F-
2R at 60C.
Microsatellite markers were purchased from Research
Genetics. The Ge´ne´thon panel of markers used for hap-
lotype analysis was D2S335, D2S326, D2S2307,
D2S2188, D2S2257, D2S2314, D2S148, D2S2173,
D2S385, and D2S324 (Dib et al. 1996). An additional
microsatellite marker, HOXD8 (Sarfarazi et al. 1995),
was employed. Microsatellite analysis was performed
by blot hybridization of polyacrylamide gels with
[a-
32
P]dCTP-labeled 5
-(CA)
10
-3
.
DNA Sequencing and Mutation Conformation
PCR products to be sequenced were gel-purified by use
of the QIAquick gel extraction kit (Qiagen). Cycle se-
quencing was performed with primers used for the orig-
inal PCR by use of the BigDye Terminator Cycle Sequenc-
ing Kit (Applied Biosystems) and the ABI 3100 Sequencer.
All mutations were confirmed by an independent
method in all available family members. The 923CrG
mutation was confirmed by loss of a DdeI restriction
site. The 940ArC mutation was confirmed by blot hy-
bridization, using the allele-specific oligonucleotide (ASO)
5
-AGTGACCCTTTGGTTTC-3
(the position of the
mutation is underlined). For each mutation, the entire
HOXD13 open reading frame was sequenced in an af-
fected individual (IV-2 from family B; III-2 from family
C) and was otherwise normal.
DNA-Protein Binding Studies
The full-length wild-type HOXD13 cDNA open read-
ing frame was obtained by RT-PCR from RNA extracted
from a normal fibroblast cell line using the Expand Long
Template PCR System (Roche) with primer pair cDNA-
F(5
-ATGGACGGGCTGCGGGCAGAC-3
) and cDNA-
R(5
-TCAGGAGACAGTATCTTTGAGC-3
) and cloned
into pGEM-T easy vector (Promega). This was then sub-
cloned into pRD67:HA vector (Davey et al. 1997), after
introducing EcoRI and XhoI restriction sites at either
end of the product. Mutant bases were introduced into
wild-type HOXD13 clones using the QuikChange site-
directed mutagenesis kit (Stratagene), with primers:
S308C-F, 5
-GCTACGAACCTATGTGAGAGACAA-
GTG-3
; I314L-F, 5
-GAGAGACAAGTGACCCTTTG-
GTTTCAGAACCG-3
; R320A-F, 5
-GTTTCAGAAC-
CGAGC
AGTGAAGGACAAG-3
, and their respective
complementary primers S308C-R, I314L-R, and
R320A-R (mutated residues are underlined). The integ-
rity of the HOXD13 sequence was confirmed in all mu-
tant constructs.
Wild-type and mutated proteins were expressed using
the TNT SP6 quick-coupled in vitro transcription/trans-
lation system (Promega) in the presence of [
35
S]-methio-
nine (Redivue Pro-mix L-[
35
S] in vitro cell labeling mix;
Amersham Bioscience). Equalization (to 20%) of con-
centrations of wild-type and mutant protein used in
binding assays was ensured by counting of
35
S-labeled
proteins on a PhosphorImager (Molecular Dynamics).
Electrophoretic mobility shift assays (EMSA) were used
to detect complex formation between labeled oligonu-
cleotide and HOXD13 proteins. The HPLC-purified
oligonucleotides (ThermoHybaid): 5
-GGGATCTGAC-
AGTTTTAC
GACAGATCT-3
(b-strand) and 5
-GGAG-
ATCTGTCGTAA
AACTGTCAGATC-3
(a-strand)
contain the core recognition sequence (underlined) 5
-
TTAC-3
(Shen et al. 1997a). These oligonucleotides
were annealed and radiolabeled by incubation with [a-
32
P]-dCTP, in the presence of Klenow DNA polymerase
(Amersham Bioscience). Other oligonucleotide pairs were
synthesized in which the underlined b-strand sequence
was varied to 5
-TTAT-3
,5
-TTAA-3
,5
-TTAG-3
, and
Page 3
Johnson et al.: HOXD13 Mutations in Brachydactyly 987
5
-TTAU-3
, and the complementary sequence was in-
corporated into the a-strand.
The binding conditions used were similar to those de-
scribed by Shen et al. (1997a). Binding was carried out
in a total volume of 20 ml, using 2 ml of equalized protein
and 0.1 pmole
32
P-labeled ds probe, incubated on ice for
30 min in buffer (final concentration: 75 mM NaCl, 1
mM EDTA, 1 mM DTT, 10 mM Tris-HCl [pH 7.5], 6%
glycerol, 2 mg BSA, 20 ng poly[dI-dC], 0.2 mg single-
stranded salmon sperm DNA). Cold competition involved
addition of 100-fold excess of unlabeled dsDNA oligo-
nucleotides. Samples were separated on a nondenaturing
6% acrylamide gel in 0.5#TBE at 220 V for 2 h at 4C.
The gels were dried and exposed in a PhosphorImager.
Structural Modeling
The D. melanogaster Ubx/Extradenticle/DNA complex
(Protein Data Bank [PDB] ID: 1B8I) (Passner et al. 1999)
and Antp/DNA complex (PDB ID: 9ANT) (Fraenkel and
Pabo 1998) were used as templates for the modeling
experiments. Residues Ile147 in 1B8I and Ile47 in 9ANT
are the equivalents of HOXD13 Ile314. The equivalents
of the 3
-thymidine (Thy) in 5
-TTAT-3
are Thy26 and
Thy221 in 1B8I and 9ANT, respectively. For each initial
template (WT-IT), three “mutants” (Mut-LT, IlerLeu;
Mut-IC, ThyrCyt [cytosine]; and Mut-LC, IlerLeu and
ThyrCyt) were generated in silico with CHARMM
(Brooks et al. 1983). The backbone conformation was
not altered in any case. All calculations were performed
in vacuo, with the solvent being approximated using a
linear distance-dependent dielectric constant. However,
the hydrogen bonding to a water molecule by residues
Gln150, Asn151, and Thy26 of 1B8I (Gln50, Asn51,
and Thy221 in 9ANT) was kept. Electrostatic interac-
tions beyond 10 A
˚
were not taken into account. The
energy of structures was minimized by steps12 # 1,000
of conjugate gradient minimization. To study the local
influence of mutations, only Ile/Leu147 and Thy/Cyt26
(1B8I numbering), or their equivalents in the 9ANT
model, were allowed to move during the minimization.
Harmonic constraints were applied to allow smooth
minimization. Harmonic forces were decreased every
1,000 steps, from 5,000 to 0 kcal/mol#A
˚
2
. Interaction
energies between Ile/Leu and Thy/Cyt pairs were cal-
culated with CHARMM. Molecular models were gen-
erated with BobScript (Esnouf 1999) and rendered with
Raster 3D (Merrit and Bacon 1997).
Results
Identification of Mutations in HOXD13
The mutation screen of HOXD13 in patients with con-
genital limb abnormalities revealed four probands (of 128
studied) who were heterozygous for pathogenic altera-
tions. One had a novel 21-bp duplication (160_180dup),
corresponding to a polyalanine expansion (A54_A60
dup) similar to those described elsewhere (Muragaki et
al. 1996; Goodman et al. 1997; reviewed by Goodman
2002); another had a novel splice-site mutation (752-
2delA), associated with a partial duplication of the sec-
ond metatarsal within the first web space (Kan et al., in
press).
The two remaining probands, both of whom had sig-
nificant family histories (families A and B; fig. 2), harbored
an identical heterozygous transversion in HOXD13
(940ArC), predicting the amino acid substitution
Ile314Leu, which locates in helix III at the 47th residue
of the conserved homeodomain (fig. 3A). The mutation
was confirmed by ASO blot hybridization and identified
in all 18 clinically affected family members (10 from
family A; 8 from family B) from whom DNA was avail-
able but in none of 10 unaffected members at 50% prior
risk (fig. 3B). Genotyping, using 11 microsatellite mark-
ers spanning HOXD13 (over a distance of 6.2 Mb),
revealed sharing of a haplotype of six consecutive mark-
ers (D2S2314–D2S324) between affected individuals in
the uppermost generations (II-2, -4, and -8 in family A;
I-2 in family B; data not shown). Given that these fam-
ilies live 90 km away from each other in southern Eng-
land, it is likely that the two families are related through
a common affected ancestor. A founder effect is also
suggested by the identification of the same mutation in
a third British family (Caronia et al. 2003).
Clinical Phenotype
Clinical examination was performed on 16 affected in-
dividuals (8 each from families A and B) as well as 10
unaffected individuals at 50% prior risk (8 from family
A; 2 from family B). Clinical photographs of both hands
of one additional affected individual in family A were also
available. Radiographs of the hands suitable for MCPP
(excluding postoperative cases) were available on 15
hands of 8 affected subjects. Apart from moderate gen-
eralized brachydactyly, we found four phenotypic patterns
associated with the mutation: first, severe middle-finger
metacarpal brachydactyly (fig. 4A); second, severe little-
finger distal phalanx hypoplasia/aplasia (fig. 4B); third, a
combination of the above; and fourth, ring-finger lateral
phalangeal duplication accompanied by 3/4 syndactyly
and/or additional features (fig. 4C). Five of 17 (29%) of
individuals differed in phenotype between their hands (fig.
2). Most affected individuals had little-finger distal pha-
langeal hypoplasia/aplasia, either alone or in association
with other abnormalities, but there were three hands with
middle-finger metacarpal brachydactyly but relatively
normal 5th digit distal phalanges (figs. 4A and 5A). There
was mild ring-finger clinodactyly in 12 hands, and 6 hands
showed mild to moderate fifth-finger camptodactyly,
usually in association with the little-finger distal phalanx
hypoplasia.
Page 4
988 Am. J. Hum. Genet. 72:984–997, 2003
Figure 2 Pedigrees of families A, B, and C. Squares represent males; circles represent females; patterns of manifestation (hands only) are
categorized as shown in the key at the bottom. The arrows identify the probands, and asterisks indicate individuals who underwent radiographic
examination.
Of 20 feet from molecularly confirmed cases that we
examined clinically and radiographically, 8 showed mild
clinical abnormalities. Seven had external evidence of
fifth-toe distal phalangeal hypoplasia/aplasia with absent
or rudimentary nail complexes, which was confirmed
radiographically. One individual had radiographically
confirmed unilateral fourth ray metatarsal brachydac-
tyly with symphalangism of the middle and distal pha-
langes of the little toe (fig. 4D). Radiographs obtained
on the remaining 12 clinically normal feet identified 3
with symphalangism of the middle and distal phalanges
of the little toe and one further case of little-toe distal
phalanx hypoplasia. Duplication of the second metatarsal
within the first web space was not observed in any case.
All affected individuals had normal stature.
The presence of brachydactyly of specific metacarpal
or metatarsal bones has not been described in association
with other HOXD13 mutations, but it is the cardinal
Page 5
Figure 3 Missense mutations in the homeodomain of HOXD13. A, above, Schematic structure of HOXD13 showing portions encoded
by each exon, the position of the splice site (sp), characteristic motifs (polyalanine tract and homeodomain), and the positions of different types
of mutation (red dots identify missense mutations reported here). Below, Amino acid sequence conservation in the second half of the homeo-
domain. The sequence of human HOXD13 is compared with its paralogues, other members of the HOXD group, and the D. melanogaster
HOM-C members Abd-B, Ubx, and Antp. The two amino acid substitutions in HOXD13 described in this report are shown above the wild-
type sequence in red. Levels of conservation are indicated as follows: black, fully conserved; yellow, moderately conserved; white, poorly
conserved. B, Identification of 940ArC mutation in families A and B. The upper panel shows DNA sequence chromatograms from a normal
individual (above) and affected individual IV-2 from family B (below). The lower panels show confirmation of the mutation by ASO blot
hybridization. C, Identification of 923CrG mutation in family C. The upper panel shows DNA sequence chromatograms from a normal
individual (above) and affected individual III-2 (below). The lower panel shows confirmation of the mutation by loss of a DdeI restriction site
(arrowhead).
Page 6
Figure 4 Clinical and radiographic appearance and MCPPs of limbs in individuals heterozygous for the Ile314Leu mutation in HOXD13. A, Individual III-13 from family A. Severe middle-
finger metacarpal brachydactyly is present bilaterally and confirmed on the radiograph. Note also the mild clinodactyly of the ring finger. One other hand shows a similar MCPP. B, Individual IV-3
from family B (age 5 years). Absence of the distal phalanx of the little finger is confirmed on the radiograph. Two other hands show similar MCPPs. C, Left and middle, individual II-2 from family
B. Note asymmetrical abnormality of the ring fingers: on the right hand there is lateral duplication of the middle phalanx, and on the left hand, there is mild clinodactyly. In addition, the hands show
features illustrated in panels A and B. C, Right, The right hand of proband IV-2 in family B (age 1 year) showing lateral duplication of the middle and distal phalanges of the ring finger. D, Foot
radiograph of individual III-5 from family B. Note the unilateral shortening of the left fourth metatarsal.
Page 7
Johnson et al.: HOXD13 Mutations in Brachydactyly 991
Figure 5 Combinations of individual abnormalities in hands from subjects with the Ile314Leu mutation (A) and the Ser308Cys mutation
(B). In A, figures in parentheses refer to additional cases classified clinically but not radiographically.
feature of brachydactyly type E (BDE [MIM 113300]).
Therefore, we sequenced HOXD13 in a family (family
C; fig. 2) previously classified with brachydactyly type E
(Brailsford 1945; Oude Luttikhuis et al. 1996). We iden-
tified a novel heterozygous missense mutation (923CrG)
corresponding to the amino acid substitution Ser308Cys,
located at the 41st position of the homeodomain (fig.
3A), which segregated in concordance with the pheno-
type in nine affected individuals (six by direct testing
and three intermediate relatives) and one unaffected in-
dividual at 50% prior risk (fig. 3C).
The detailed pedigree and phenotype in family C has
been reported elsewhere (Brailsford 1945; Oude Lutti-
khuis et al. 1996). The characteristic features are short-
ening of one or more of the metacarpals or metatarsals
or of both, often occurring asymmetrically, together with
either shortening or elongation of specific distal phalanges
(notably the first and fifth) and carpal-bone fusion. Pub-
lished details of the hand phenotype are summarized in
figure 5B and document wide intrafamilial variation in
these features, which overlap those described in brachy-
dactyly type D as well as type E (see “Discussion” section).
DNA Binding Assays
The distinct phenotype in these three families, notably
including metacarpal and metatarsal brachydactyly, sug-
gested that the mutations act by a mechanism distinct
from that described in other HOXD13 mutations. No-
tably, the isoleucine at the 47th position of the homeo-
domain equivalent to Ile314 in HOXD13 makes hydro-
phobic contacts with DNA in several available crystal
structures, including DNA-bound Ubx/Extradenticle
(Passner et al. 1999) (fig. 1), Antennapedia (Fraenkel
and Pabo 1998), and HOXB1-PBX1 homeodomain
complexes (Piper et al. 1999). We speculated that the
HOXD13 missense mutations might affect DNA binding
and tested this using EMSA.
As the DNA contact made by I47 in Ubx is to the last
base in the 5
-TTAT-3
core sequence (fig. 1), we first tested
the binding of wild-type and mutant proteins to ds ol-
igonucleotides that contained 5
-TTAT-3
,5
-TTAC-3
,5
-
TTAA-3
,5
-TTAG-3
, and 5
-TTAU-3
(b-strand) se-
quences. Significant binding was observed only to the
5
-TTAT-3
and 5
-TTAC-3
sequences (data not shown).
As a negative control, we created an artificial mutation,
Arg320Ala. Arg320 lies at the 53rd position of the ho-
meodomain, a highly conserved residue important for
contacting with the phosphate backbone of DNA (fig.
1); it is a common site for homeobox gene mutations
with demonstrated loss of function (Dattani et al. 1998;
Percin et al. 2000; D’Elia et al. 2001). This residue was
mutated to alanine, a neutral amino acid with a high
propensity to promote a-helix formation. As expected,
no DNA binding was obtained using the Arg320Ala mu-
tant protein.
We then compared the affinities of the wild-type and
mutant proteins for the 5
-TTAT-3
and 5
-TTAC-3
target
sequences in more detail. The wild-type protein had
similar binding affinities for each target sequence: 50%
or more of this binding could be competed off with
excess unlabeled competitor oligonucleotide. By con-
trast, the Ile314Leu mutant protein consistently exhib-
ited greater binding (average 2.4-fold) to 5
-TTAC-3
but
only 0.23-fold binding to 5
-TTAT-3
, when compared
with wild-type protein. These results were in keeping
with experiments involving competition assays with a
100-fold excess of unlabeled ds oligonucleotides (fig. 6).
In parallel studies on the mutant Ser308Cys protein,
although there was a modest increase in binding to 5
-
TTAC-3
compared with wild-type protein (average of
Page 8
992 Am. J. Hum. Genet. 72:984–997, 2003
Figure 6 In vitro binding of wild-type (WT) and mutant (I314L, S308C, and R320A) HOXD13 proteins to synthetic ds oligonucleotides.
Above, representative gel shift assays employing
32
P-labeled 5
-TTAT-3
probe (left) and 5
-TTAC-3
probe (right) in the absence () or presence
of unlabeled competitor oligonucleotides. The arrow shows the position of the bound oligonucleotide/protein complexes. Below, quantitation
(mean SEM) of total counts from 4–5 experiments. Note the difference in scale of absolute counts on the Y-axis.
1.4-fold), this was not consistently replicated in inde-
pendent experiments (fig. 6).
Molecular Modeling
To evaluate the mechanism of the enhanced binding
of the Ile314Leu mutant protein to the 5
-TTAC-3
oli-
gonucleotide, we modeled the interaction on the basis of
the crystal structures of D. melanogaster Ubx and Antp
homeodomains bound to DNA (fig. 7A). We chose these
structures because they are the most closely related pro-
teins for which the three-dimensional structure has been
solved. Ubx interacts with a 5
-TTAT-3
sequence, whereas
the Antp homeodomain interacts with 5
-TAAT-3
in the
respective crystal structures. To study the role of residue
314 in HOXD13 (Ile or Leu) and the interacting base
(Thy or Cyt), four models were generated, in silico, from
each crystal structure to mimic the different possible in-
teractions (WT-IT, Ile/Thy; Mut-LT, Leu/Thy; Mut-IC,
Ile/Cyt; and Mut-LC, Leu/Cyt). Visual inspection of the
models showed that the orientation of the side chain of
residue 314 differs in the different mutants. For instance,
in the Ile314Leu model (Mut-LT), the side chain of the
leucine residue rotates to avoid steric clashes with the
5-methyl group of Thy (not shown), whereas in the dou-
ble-mutant model (Mut-LC), the leucine can move to-
ward the cytosine, since the latter lacks the 5-methyl
group (fig. 7A). In an attempt to estimate the differences
among the models, the interaction energy among each
of the four combinations of amino acid and base was
computed (fig. 7B). Although the energy differences
Page 9
Johnson et al.: HOXD13 Mutations in Brachydactyly 993
Figure 7 Interaction between HOXD13 and DNA sequences. A, Models based on the Ubx/DNA complex, highlighting the interaction
between Ile and thymine (Thy) in the wild type (left panel) and between Leu and cytosine (Cyt) in the “double mutant” (right panel). The cyan
spiral indicates the main chain of homeodomain helix III; atoms (except H) in the position 47 side chain and DNA are color coded as follows:
C p gray; O p red; N p dark blue; P p orange. B, Total (dashed line) and van der Waals (solid line) interaction energies (in kcal/mol)
between Ile and Thy (WT-IT, wild type), Leu and Thy (Mut-LT, IlerLeu mutant), Ile and Cyt (Mut-IC, ThyrCyt mutant), and Leu and Cyt
(Mut-LC, double mutant). Models were built using the Ubx (left) and Antp (right) homeodomain/DNA complexes.
among the four models are small, they are consistent
with the DNA-binding results. The WT-IT and Mut-IC
models have comparable interaction energies, in agree-
ment with the similar affinity of HOXD13 for 5
-TTAT-
3
and 5
-TTAC-3
. The higher energy of interaction be-
tween Leu and Thy in the Mut-LT model, as compared
with Leu and Cyt in the Mut-LC model, is consistent
with the preference of the Ile314Leu mutant for 5
-
TTAC-3
. The major differences are found in the van der
Waals energy term, suggesting that the lower affinity of
the Ile314Leu mutant for 5
-TTAT-3
is mainly due to
steric hindrance.
Discussion
Pathogenicity of HOXD13 Mutations
Using WAVE DHPLC, we identified mutations of
HOXD13 in 4 of 128 (3.1%) consecutive samples tested
from patients with a wide range of congenital limb ab-
normalities that required surgery. This indicates that
HOXD13 mutations contribute a small but significant
percentage to the overall burden of congenital limb ab-
normalities. In this report, we focus on two missense
mutations of HOXD13: Ile314Leu, detected in two pa-
Page 10
994 Am. J. Hum. Genet. 72:984–997, 2003
tients in the initial screen, and Ser308Cys, detected in a
published family following the insight that HOXD13
mutations may present with metacarpal or metatarsal
brachydactyly.
Biochemically, isoleucine and leucine are both non-
polar amino acids and share many similar characteristics
(Grantham 1974), so it is important to consider the pos-
sibility that the Ile314Leu substitution could be only a
silent polymorphism. Indeed, in a tabulation of the rela-
tive frequency of pathogenic substitutions in humans, as
compared with tolerated substitutions in orthologous pro-
teins between species, Miller and Kumar (2001) found
that for the isoleucine-to-leucine change, most substi-
tutions are tolerated. Nevertheless, there is compelling
evidence that Ile314Leu in HOXD13 is pathogenic: first,
the mutation segregates perfectly with the phenotype in
18 affected and 10 unaffected family members at 50%
prior risk (two-point LOD , using a pene-
score p 7.2
trance in heterozygotes of 0.99); second, the substitution
was not present in 168 chromosomes from unrelated
normal individuals; third, the isoleucine residue is highly
conserved in HOX proteins (fig. 3A); fourth, structural
information from NMR and x-ray crystallographic data
support isoleucine 47 as one of four key residues in helix
III (I47, Q50, N51, and M54) that are responsible for
sequence-specific DNA contacts (fig. 1; reviewed by Bil-
leter 1996); and fifth, functional studies indicate that the
substitution affects DNA-binding affinity (fig. 6).
The Ser308Cys substitution occurs at position 41, be-
tween helix II and helix III of the homeodomain. Bio-
chemically, this substitution is of intermediate severity
(Grantham 1974), but it occurs in a less-conserved re-
gion of the protein, represented by a threonine residue
in Ubx and Antp (fig. 3A). In both structures, this residue
interacts via a water molecule with the oxygen atom of
phosphates of A and T (underlined) in 5
-T
T
/
A
AT-3
, mak-
ing the effect of substitution difficult to predict (Fraenkel
and Pabo 1998; Passner et al. 1999). No definite func-
tional alteration could be established from the DNA-bind-
ing studies (fig. 6), but there are several lines of evidence
that Ser308Cys is also pathogenic: first, it segregates con-
cordantly with the phenotype in nine affected and one
unaffected individual; second, the nucleotide substitution
was not detectable in 170 chromosomes from unrelated
normal individuals; third, cysteine has been recorded in
only two (of
11,000) homeobox sequences, those of D.
melanogaster zerknu¨ llt and mouse Pem (Homeodomain
Resource; Banerjee-Basu et al. 2001); and fourth, pre-
vious functional analysis of a mutation occurring at the
identical position of the homeodomain but involving a
different substitution (Thr178Met in NKX2-5, which
causes human congenital heart disease) showed reduced
DNA binding and transcriptional activation compared
with wild type (Zhu et al. 2000).
HOXD13 Mutations Are Associated with
Brachydactyly Types D and E
The missense mutation Ile314Leu, identified in two
families, exhibits a novel phenotype; as in the case of
SPD (Goodman et al. 1997), the variability in expressivity
is striking, both between family members and in opposite
limbs. The most distinctive manifestations of this mu-
tation, seen in the upper limbs, were isolated middle-
finger metacarpal brachydactyly and lateral duplication
of the distal elements of the ring finger. However, these
features were observed in only a proportion of patients
(fig. 5A) and were sometimes asymmetrical (fig. 4C); the
lateral duplications did not involve more proximal pha-
langes or metacarpal elements, in contrast to more severe
cases of SPD. Only a minority of individuals exhibited
3/4 syndactyly overlapping the typical appearance of SPD.
The occurrence of metacarpal and metatarsal brachy-
dactyly led us to analyze a family published elsewhere
(family C) segregating this phenotype through five gen-
erations (Brailsford 1945; Oude Luttikhuis et al. 1996).
The identification of a heterozygous Ser308Cys muta-
tion in family C confirms that HOXD13 mutations are
a cause of metacarpal and metatarsal brachydactyly. The
phenotype of this latter mutation was characterized by
selective shortening of metacarpals (rays 3, 4, and 5)
and metatarsals (rays 1 and 4) as well as shortening of
the distal phalanx of the thumb. No family member had
syndactyly, but foot radiographs of several affected in-
dividuals show an abnormally broad first metatarsal,
sometimes with a bony spur somewhat reminiscent of the
abnormality described by Goodman et al. (1998), pos-
tulated to be associated with haploinsufficiency.
The most commonly adopted classification of the non-
syndromic brachydactylies is that devised by Bell (1951).
Type D brachydactyly (BDD [MIM 113200] is defined as
a short, broad distal phalanx in the thumb, and BDE is
defined as comprising one or more shortened metacar-
pals and metatarsals. Although both types of brach-
ydactyly may be a feature of several syndromes (Rubin-
stein-Taybi and Saethre-Chotzen syndromes in type D;
Turner syndrome, mutations in the a-subunit of Gs pro-
tein, and deletions of 2q37 in type E), no locus has been
defined for nonsyndromic BDD or BDE. Although not
breeding true for the pure phenotypes, families A–C all
exhibit features of BDE; in family C, this is combined
in some individuals with BDD. The present study iden-
tifies HOXD13 mutations as a cause of both BDD and
BDE, and further mutation analysis of patients with these
diagnoses may help to refine the clinical classification of
these disorders.
Functional Effects of Homeodomain Substitutions
EMSA showed that Ile314Leu confers an increased
binding affinity for the target sequence 5
-TTAC-3
but
Page 11
Johnson et al.: HOXD13 Mutations in Brachydactyly 995
a decreased affinity for 5
-TTAT-3
compared with wild-
type protein, suggesting that the mutation has mixed gain-
and loss-of-affinity effects. Mutations of the homeodo-
main that increase binding affinity are unusual, but an
example reported elsewhere was the Pro148His substi-
tution of MSX2 (located in the N-terminal arm of the
homeodomain), which exhibited this effect in a similar
assay (Ma et al. 1996). Isoleucine and leucine are iso-
mers; that is, they only differ by the position of a methyl
group that is linked to the first (Ile: Cb) or to the second
(Leu: Cg) carbon atom of the side chain. As a conse-
quence, the Ile314Leu mutation may create steric hin-
drance between the leucine and the 5-methyl group of
the thymine at the 3
end of the 5
-TTAT-3
core, ex-
plaining the reduced binding affinity of the mutant pro-
tein for this sequence. By contrast, in the presence of
cytosine, which lacks the 5-methyl group, we propose
that the Ile314Leu substitution may complement the lack
of methyl in the base, resulting in an increased affinity
for 5
-TTAC-3
(fig. 7). On a cautionary note, in vivo
activity is likely to require the binding of partner proteins
(such as MEIS1; Shen et al. 1997a) to HOXD13, which
might modulate the relative binding affinities of the tran-
scription complex to different DNA targets.
Although EMSA analysis of the Ser308Cys substi-
tution did not identify consistent differences from wild
type, our results show that this mutation does not simply
abolish DNA binding (in contrast to the Arg320Ala con-
trol). The substitution lies between helices II and III and
involves relatively little change in molecular volume, com-
patible with a subtle effect on DNA binding. Further work
will be required to determine the pathophysiological
mechanism of this mutation.
Three Distinct Classes of Mutation Exist in HOXD13
Three distinct classes of mutations have now been
described in HOXD13: polyalanine tract expansions,
truncations, and specific amino acid substitutions. Each
is likely to exert its effect through different mechanisms,
consistent with differences in the phenotypic outcome.
Accumulating evidence suggests that polyalanine expan-
sion mutations may exert a dominant negative effect over
wild-type protein (Bruneau et al. 2001). The mutations
which truncate the HOXD13 protein are likely to cause
loss of function (haploinsufficiency), whereas we have
presented evidence that the Ile314Leu substitution in the
homeodomain exerts both gain and loss of DNA-binding
affinity to different targets. This was also speculated to
be the pathological mechanism of missense substitutions
of two key base-contacting residues (Q50 and N51; see
fig. 1) in HOXA13, both of which were associated with
severe or variant HFUS phenotypes; however, no func-
tional data were presented in these cases (Goodman et
al. 2000; Innis et al. 2002).
Our conclusion is that a leucine at position 47 of the
homeodomain causes an altered repertoire of DNA bind-
ing rather than simple loss of function, which implies that
in a different molecular and developmental context, a
homeobox gene encoding Leu47 could act in a physio-
logical manner. It is therefore interesting that several
homeodomain-containing proteins from the fungal
pathogen Ustilago maydis indeed utilize a leucine at this
position (Kronstad and Leong 1990; Schulz et al. 1990;
Banerjee-Basu et al. 2001).
Acknowledgments
We thank the families who took part in the study for their
help and cooperation. We thank Navaratnam Elanko, for help
with DNA extraction; Kevin Clark, for DNA sequencing; Anne
Goriely and Christina Tufarelli, for technical advice; and Alas-
dair Hunter, for access to the ANTRO software. This study
was funded by the Wellcome Trust (to D.J. and A.O.M.W.),
the Ministry of Education in Taiwan (to S-h.K.), the Overseas
Research Students Awards Scheme (to S-h.K.), CNRS France
(to P.R.), and MRC-UK (to P.R. and R.M.E.).
Electronic-Database Information
Accession numbers and URLs for data presented herein are
as follows:
Ensembl, http://www.ensembl.org/Homo_sapiens/mapview
?chrp2 (for physical map of 2q31 region)
GenBank, http://www.ncbi.nlm.nih.gov/GenBank/ (for
HOXD13 [accession numbers. AF005219, AF005220,
AC009336, and NM_000523])
Homeodomain Resource, http://research.nhgri.nih.gov
/homeodomain/ (for homeodomain sequences and DNA bind-
ing sites)
Online Mendelian Inheritance in Man (OMIM), http://www
.ncbi.nlm.nih.gov/Omim/(for SPD, HFUS, BDD, and BDE)
Protein Data Bank, http://www.pdb.org/ (for Antp [ID: 9ANT]
and Ubx [ID: 1B8I] homeodomain-DNA structures)
References
Albrecht AN, Schwabe GC, Stricker S, Bo¨ddrich A, Wanker
EE, Mundlos S (2002) The synpolydactyly homolog (spdh)
mutation in the mouse: a defect in patterning and growth
of limb cartilage elements. Mech Dev 112:53–67
Armour CM, Bulman DE, Hunter AGW (2000) Clinical and
radiological assessment of a family with mild brachydactyly
type A1: the usefulness of metacarpophalangeal profiles. J
Med Genet 37:292–296
Banerjee-Basu S, Baxevanis AD (2001) Molecular evolution of
the homeodomain family of transcription factors. Nucleic
Acids Res 29:3258–3269
Banerjee-Basu S, Sink DW, Baxevanis AD (2001) The Homeo-
domain Resource: sequences, structures, DNA binding sites
and genomic information. Nucleic Acids Res 29:291–293
Bell J (1951) On brachydactyly and symphalangism. In: The
Page 12
996 Am. J. Hum. Genet. 72:984–997, 2003
treasury of human inheritance. Vol 5. Cambridge University
Press, Cambridge, pp 1–31
Billeter M (1996) Homeodomain-type DNA recognition. Prog
Biophys Mol Biol 66:211–225
Brailsford JF (1945) Familial brachydactyly. Br J Radiol 18:
167–172
Brooks BR, Bruccoleri RE, Olafson BD, States DJ, Swami-
nathan S, Karplus M (1983) CHARMM: a program for
macromolecular energy, minimization, and dynamics cal-
culations. J Comput Chem 4:187–217
Bruneau S, Johnson KR, Yamamoto M, Kuroiwa A, Duboule
D (2001) The mouse Hoxd13
spdh
mutation, a polyalanine
expansion similar to human type II synpolydactyly (SPD),
disrupts the function but not the expression of other Hoxd
genes. Dev Biol 237:345–353
Bu¨ rglin TR (1994) A comprehensive classification of homeo-
box genes. In: Duboule D (ed) Guidebook to the homeobox
genes. Oxford University Press, Oxford, pp 27–71
Calabrese O, Bigoni S, Gualandi F, Trabanelli C, Camera G,
Calzolari E (2000) A new mutation in HOXD13 associated
with foot pre-postaxial polydactyly. Eur J Hum Genet 8 S1:
140
Caronia G, Goodman FR, McKeown CME, Scambler PJ, Zap-
pavigna V (2003) An I47L substitution in the HOXD13 hom-
eodomain causes a novel human limb malformation by pro-
ducing a selective loss of function. Development 130:1701–
1712
Dattani MT, Martinez-Barbera J-P, Thomas PQ, Brickman JM,
Gupta R, Ma˚rtensson I-L, Toresson H, Fox M, Wales JKH,
Hindmarsh PC, Krauss S, Beddington RSP, Robinson ICAF
(1998) Mutations in the homeobox gene HESX1/Hesx1 as-
sociated with septo-optic dysplasia in human and mouse.
Nat Genet 19:125–133
Davey RA, Hamson CA, Healey JJ, Cunningham JM (1997)
In vitro binding of purified murine ecotropic retrovirus en-
velope surface protein to its receptor, MCAT-1. J Virol 71:
8096–8102
Debeer P, Bacchelli C, Scambler PJ, De Smet L, Fryns J-P,
Goodman FR (2002) Severe digital abnormalities in a pa-
tient heterozygous for both a novel missense mutation in
HOXD13 and a polyalanine tract expansion in HOXA13.
J Med Genet 39:852–856
D’Elia AV, Tell G, Paron I, Pellizzari L, Lonigro R, Damante
G (2001) Missense mutations of human homeoboxes: a re-
view. Hum Mutat 18:361–374
Dib C, Faure S, Fizames C, Samson D, Drouot N, Vignal A,
Millaseau P, Marc S, Hazan J, Seboun E, Lathrop M, Gyapay
G, Morissette J, Weissenbach J (1996) A comprehensive ge-
netic map of the human genome based on 5,264 microsa-
tellites. Nature 380:152–154
Dolle´ P, Dierich A, LeMeur M, Schimmang T, Schuhbaur B,
Chambon P, Duboule D (1993) Disruption of the Hoxd-13
gene induces localized heterochrony leading to mice with
neotenic limbs. Cell 75:431–441
Esnouf RM (1999) Further additions to MolScript version 1.4,
including reading and contouring of electron-density maps.
Acta Crystallogr D Biol Crystallogr 55:938–940
Fraenkel E, Pabo CO (1998) Comparison of X-ray and NMR
structures for the Antennapedia homeodomain-DNA com-
plex. Nat Struct Biol 5:692–697
Garn SM, Hertzog KP, Poznanski AK, Nagy JM (1972) Me-
tacarpophalangeal length in the evaluation of skeletal mal-
formation. Radiology 105:375–381
Gehring WJ, Qian YQ, Billeter M, Furukubo-Tokunaga K,
Schier AF, Resendez-Perez D, Affolter M, Otting G, Wu¨th-
rick K (1994) Homeodomain-DNA recognition. Cell 78:
211–223
Goodman F, Giovannucci-Uzielli ML, Hall C, Reardon W,
Winter R, Scambler PJ (1998) Deletions in HOXD13 seg-
regate with an identical, novel foot malformation in two
unrelated families. Am J Hum Genet 63:992–1000
Goodman FR (2002) Limb malformations and the human
HOX genes. Am J Med Genet 112:256–265
Goodman FR, Bacchelli C, Brady AF, Brueton LA, Fryns J-P,
Mortlock DP, Innis JW, Holmes LB, Donnenfeld AE, Fein-
gold M, Beemer FA, Hennekam RCM, Scambler PJ (2000)
Novel HOXA13 mutations and the phenotypic spectrum of
hand-foot-genital syndrome. Am J Hum Genet 67:197–202
Goodman FR, Mundlos S, Muragaki Y, Donnai D, Giovan-
nucci-Uzielli ML, Lapi E, Majewski F, McGaughran J,
McKeown C, Reardon W, Upton J, Winter RM, Olsen BR,
Scambler PJ (1997) Synpolydactyly phenotypes correlate with
size of expansions in HOXD13 polyalanine tract. Proc Natl
Acad Sci USA 94:7458–7463
Grantham R (1974) Amino acid difference formula to help
explain protein evolution. Science 185:862–864
Innis JW, Goodman FR, Bacchelli C, Williams TM, Mortlock
DP, Sateesh P, Scambler PJ, McKinnon W, Guttmacher AE
(2002) A HOXA13 allele with a missense mutation in the
homeobox and a dinucleotide deletion in the promoter un-
derlies Guttmacher syndrome. Hum Mutat 19:573–574
Kan S-h, Johnson D, Giele H Wilkie AOM. An acceptor splice
site mutation in HOXD13 results in variable hand, but con-
sistent foot malformations. Am J Med Genet (in press)
Kosaki K, Kosaki R, Suzuki T, Yoshihashi H, Takahashi T,
Sasaki K, Tomita M, McGinnis W, Matsuo N (2002) Com-
plete mutation analysis panel of the 39 human HOX genes.
Teratology 65:50–62
Kronstad JW, Leong SA (1990) The b-mating type locus of
Ustilago maydis contains variable and constant regions. Genes
Dev 4:1384–1395
Krumlauf R (1994) HOX genes in vertebrate development.
Cell 78:191–201
Ma L, Golden S, Wu L, Maxson R (1996) The molecular basis
of Boston-type craniosynostosis: the Pro148rHis mutation
in the N terminal arm of the MSX2 homeodomain stabilizes
DNA binding without altering nucleotide sequence prefer-
ences. Hum Mol Genet 5:1915–1920
Mann RS, Affolter M (1998) HOX proteins meet more part-
ners. Curr Opin Genet Dev 8:423–429
Mann RS, Carroll SB (2002) Molecular mechanisms of selector
gene function and evolution. Curr Opin Genet Dev 12:592–
600
Merritt EA, Bacon DJ (1997) Raster 3D: photorealistic molec-
ular graphics. Methods Enzymol 277:505–524
Miller MP, Kumar S (2001) Understanding human disease mu-
tations through the use of interspecific genetic variation.
Hum Mol Genet 10:2319–2328
Mortlock DP, Innis JW (1997) Mutation of HOXA13 in hand-
foot-genital syndrome. Nat Genet 15:179–180
Page 13
Johnson et al.: HOXD13 Mutations in Brachydactyly 997
Muragaki Y, Mundlos S, Upton J, Olsen BR (1996) Altered
growth and branching patterns in synpolydactyly caused by
mutations in HOXD13. Science 272:548–551
Oude Luttikhuis MEM, Williams DK, Trembath RC (1996)
Isolated autosomal dominant type E brachydactyly: exclu-
sion of linkage to candidate regions 2q37 and 20q13. J Med
Genet 33:873–876
Passner JM, Ryoo HD, Shen L, Mann RS, Aggarwal AK (1999)
Structure of a DNA-bound Ultrabithorax-Extradenticle hom-
eodomain complex. Nature 397:714–719
Percin EF, Ploder LA, Yu JJ, Arici K, Horsford DJ, Rutherford
A, Bapat B, Cox DW, Duncan AMV, Kalnins VI, Kocak-
Altintas A, Sowden JC, Traboulsi E, Sarfarazi M, McInnes
RR (2000) Human microphthalmia associated with muta-
tions in the retinal homeobox gene CHX10. Nat Genet 25:
397–401
Piper DE, Batchelor AH, Chang C-P, Cleary ML, Wolberger
C (1999) Structure of a HoxB1-Pbx1 heterodimer bound to
DNA: role of the hexapeptide and a fourth homeodomain
helix in complex formation. Cell 96:587–597
Poznanski AK, Garn SM, Nagy JM, Gall JC Jr (1972) Meta-
carpophalangeal pattern profiles in the evaluation of skeletal
malformations. Radiology 104:1–11
Rijli FM, Chambon P (1997) Genetic interactions of Hox genes
in limb development: learning from compound mutants. Curr
Opin Genet Dev 7:481–487
Sarfarazi M, Akarsu AN, Sayli BS (1995) Localization of the
syndactyly type II (synpolydactyly) locus to 2q31 region and
identification of tight linkage to HOXD8 intragenic marker.
Hum Mol Genet 4:1453–1458
Sayli BS, Akarsu AN, Sayli U, Akhan O, Ceylaner S, Sarfarazi
M (1995) A large Turkish kindred with syndactyly type II
(synpolydactyly). 1. Field investigation, clinical and pedigree
data. J Med Genet 32:421–434
Schulz B, Banuett F, Dahl M, Schlesinger R, Scha¨fer W, Martin
T, Herskowitz I, Kahmann R (1990) The b alleles of U.
maydis, whose combinations program pathogenic develop-
ment, code for polypeptides containing a homeodomain-re-
lated motif. Cell 60:295–306
Scott MP (1992) Vertebrate homeobox gene nomenclature. Cell
71:551–553
Shen W-F, Montgomery JC, Rozenfeld S, Moskow JJ, Lawrence
HJ, Buchberg AM, Largman C (1997a) AbdB-like Hox pro-
teins stabilize DNA binding by the Meis1 homeodomain
proteins. Mol Cell Biol 17:6448–6458
Shen W-F, Rozenfeld S, Lawrence HJ, Largman C (1997b) The
Abd-B-like Hox homeodomain proteins can be subdivided
by the ability to form complexes with Pbx1a on a novel
DNA target. J Biol Chem 272:8198–8206
Za´ka´ny J, Fromental-Ramain C, Warot X, Duboule D (1997)
Regulation of number and size of digits by posterior Hox
genes: a dose-dependent mechanism with potential evolution-
ary implications. Proc Natl Acad Sci USA 94:13695–13700
Zhu W, Shiojima I, Hiroi Y, Zou Y, Akazawa H, Mizukami
M, Toko H, Yazaki Y, Nagai R, Komuro I (2000) Functional
analyses of three Csx/Nkx-2.5 mutations that cause human
congenital heart disease. J Biol Chem 275:35291–35296
Page 14
  • Source
    • "Three LOF-alleles, R31G, R31Q, and R31W, cause a SPD phenotype, with incomplete penetrance in the case of R31G and R31W [Debeer et al., 2002; Dai et al., 2014]. The four remaining missense mutations are GOF-alleles and do not cause SPD, but Brachydactyly type E and D (S41C and I47L), Syndactyly Type V (Q50R), and a complex brachydactyly-syndactyly phenotype (Q50K) [Caronia et al., 2003; Johnson et al., 2003; Zhao et al., 2007; Ibrahim et al., 2013]. "
    [Show abstract] [Hide abstract] ABSTRACT: Synpolydactyly (SPD) is a rare congenital limb disorder characterized by syndactyly between the third and fourth fingers and an additional digit in the syndactylous web. In most cases SPD is caused by heterozygous mutations in HOXD13 resulting in the expansion of a N-terminal polyalanine tract. If homozygous, the mutation results in severe shortening of all metacarpals and phalanges with a morphological transformation of metacarpals to carpals. Here, we describe a novel homozygous missense mutation in a family with unaffected consanguineous parents and severe brachydactyly and metacarpal-to-carpal transformation in the affected child. We performed whole exome sequencing on the index patient, followed by Sanger sequencing of parents and patient to investigate cosegregation. The DNA-binding ability of the mutant protein was tested with electrophoretic mobility shift assays. We demonstrate that the c.938C>G (p.313T>R) mutation in the DNA-binding domain of HOXD13 prevents binding to DNA in vitro. Our results show to our knowledge for the first time that a missense mutation in HOXD13 underlies severe brachydactyly with metacarpal-to-carpal transformation. The mutation is non-penetrant in heterozygous carriers. In conjunction with the literature we propose the possibility that the metacarpal-to-carpal transformation results from a homozygous loss of functional HOXD13 protein in humans in combination with an accumulation of non-functional HOXD13 that might be able to interact with other transcription factors in the developing limb. © 2015 Wiley Periodicals, Inc.
    Full-text · Article · Nov 2015 · American Journal of Medical Genetics Part A
  • Source
    • "BDC (MIM 113100) can present with a range of anomalies (from shortening to hyperphalangy of the middle phalanges of the index, middle, and little fingers) and is caused by mutations affecting growth/differentiation factor 5 (GDF5) [14]. Two mutations of HOXD13 are associated with distinctive limb phenotypes in which brachydactylies of specific metacarpals, metatarsals, and phalangeal bones are the most constant features and exhibit overlap with brachydactyly types D (MIM 113200) and E (MIM 113300) [15]. BDA2 (MIM 112600) was described first by Mohr and Wiredt in a large Norwegian kindred of Danish descent [16]. "
    [Show abstract] [Hide abstract] ABSTRACT: Brachydactyly type A2 (BDA2, MIM 112600) is characterized by the deviation and shortening of the middle phalange of the index finger and the second toe. Using genome-wide linkage analysis in a Chinese BDA2 family, we mapped the maximum candidate interval of BDA2 to a ∼1.5 Mb region between D20S194 and D20S115 within chromosome 20p12.3 and found that the pairwise logarithm of the odds score was highest for marker D20S156 (Zmax = 6.09 at θ = 0). Based on functional and positional perspectives, the bone morphogenetic protein 2 (BMP2) gene was identified as the causal gene for BDA2 in this region, even though no point mutation was detected in BMP2. Through further investigation, we identified a 4,671 bp (Chr20: 6,809,218-6,813,888) genomic duplication downstream of BMP2. This duplication was located within the linked region, co-segregated with the BDA2 phenotype in this family, and was not found in the unaffected family members and the unrelated control individuals. Compared with the previously reported duplications, the duplication in this family has a different breakpoint flanked by the microhomologous sequence GATCA and a slightly different length. Some other microhomologous nucleotides were also found in the duplicated region. In summary, our findings support the conclusions that BMP2 is the causing gene for BDA2, that the genomic location corresponding to the duplication region is prone to structural changes associated with malformation of the digits, and that this tendency is probably caused by the abundance of microhomologous sequences in the region.
    Full-text · Article · Apr 2014 · PLoS ONE
  • Source
    • "Zhao et al., 2007). For those mutations, a mixed gain-and loss-offunction mechanism has been proposed as an explanation for the complex overlapping limb malformation phenotypes (Johnson et al., 2003; Zhao et al., 2007). "
    [Show abstract] [Hide abstract] ABSTRACT: Synpolydactyly (SPD, OMIM 186000) is a rare congenital limb disorder characterized by syndactyly between the third and fourth fingers and between the fourth and fifth toes, with partial or complete digit duplication in the syndactylous web. The majority of these anomalies co-segregate with mutations in the HOXD13 gene, a homeobox transcription factor crucial for distal limb development. Different classes of HOXD13 mutations are involved in the pathogenesis of synpolydactyly, but an unequivocal genotype-phenotype correlation cannot always be achieved due to the clinical heterogeneity and reduced penetrance of SPD. All mutations identified so far mapped to the N-terminal polyalanine tract or to the C-terminal homeodomain of HOXD13, causing typical or atypical features of SPD respectively. However, mutations outside of these domains cause a broad variety of clinical features that complicate the differential diagnosis. The existing animal models that are currently used to study HOXD13 (mal)function are therefore instrumental in unraveling potential genotype-phenotype correlations. Both mouse- and chick-based approaches allow the in vivo study of the pathogenic mechanism by which HOXD13 mutations cause SPD phenotypes as well as help in identifying the transcriptional targets. Developmental Dynamics, 2013. © 2013 Wiley Periodicals, Inc.
    Full-text · Article · Feb 2014 · Developmental Dynamics
Show more