An N-terminal G11A Mutation in HOXD13 Causes

An N-terminal G11A mutation in HOXD13 causessynpolydactyly and interferes with Gli3R functionduring limb pre-patterning

{

Nathalie Brison1, Philippe Debeer2, Sebastian Fantini3, Christine Oley4, Vincenzo Zappavigna3,

Frank P. Luyten1 and Przemko Tylzanowski1,∗

1Laboratory of Skeletal Development and Joint Disorders, University of Leuven, Herestraat 49, O&N1 Box 813, 3000

Leuven, Belgium, 2Department of Musculoskeletal Sciences, Division of Orthopedics, University Hospital Pellenberg,

Weligerveld 1, 3202 Pellenberg, Belgium, 3Department of Animal Biology, University of Modena and Reggio Emilia,

Via Campi 213/D, 41100 Modena, Italy and 4Clinical Genetics Unit, Birmingham’s Women’s Hospital, Mindelson Way,

Edgbaston, Birmingham B15 2TG, UK

Received December 16, 2011; Revised and Accepted February 18, 2012

Synpolydactyly (SPD) is a distal limb anomaly characterized by incomplete digit separation and the presenceof supernumerary digits in the syndactylous web. This phenotype has been associated with mutations in thehomeodomain or polyalanine tract of the HOXD13 gene. We identified a novel mutation (G11A) in HOXD13that is located outside the previously known domains and affects the intracellular half life of the protein.Misexpression of HOXD13(G11A) in the developing chick limb phenocopied the human SPD phenotype.Finally, we demonstrated through in vitro studies that this mutation has a destabilizing effect on GLI3R unco-vering an unappreciated mechanism by which HOXD13 determines the patterning of the limb.

INTRODUCTION

Establishment of anterior–posterior asymmetry in the limbbud (AP axis, from thumb to little finger) requires a regulatorynetwork of signaling molecules whose expression is thoroughlycoordinated in time and space (reviewed in 1,2). A small groupof mesenchymal cells at the posterior margin of the developinglimb bud—the zone of polarizing activity (ZPA)—acts as an or-ganizer that initiates the specification of the anatomy of ourhands and feet. Digit formation largely depends on the consecu-tive cascade of events initiated by the localized expression ofSonic Hedgehog (Shh) from the ZPA (3). During the emergenceof the limb bud, the expression of other patterning genes includ-ing dHand, Twist1 and the most 5′ members of the Hoxd genecluster are activated in the posterior part of the limb bud andconfer a positive feedback loop to sustain Shh expression(4,5). Meanwhile, the transcriptional repressor form of Gli3(Gli3R) in the anterior mesenchyme restricts the expression ofdHand, 5′Hoxd genes and Shh to the posterior part of the limbbud (6,7). The integration of the resulting opposing spatial gra-dient of Shh (posterior) and Gli3R (anterior) (8) and the kinetics

of transcription of downstream target genes then further speci-fies the digit number and identity (reviewed in 1,9).

Defects in patterning genes such as GLI3 are known tocause congenital limb malformations in humans (10–12). Syn-polydactyly (SPD, MIM 186000) is an autosomal, dominantlyinherited limb malformation syndrome categorized accordingto Temtamy and McKusick as ‘type II syndactyly’ (13).SPD is manifested by fusion of the digits (syndactyly) in themesoaxial line of the hands and in the postaxial part of thefeet, particularly between the third and fourth finger andbetween the fourth and fifth toe, with a partially or completelyduplicated digit in the syndactylous web (polydactyly) (14).Additionally, it may be associated with brachydactyly, camp-todactyly or clinodactyly of the fifth fingers and variable syn-dactyly of the second to fifth toes with middle phalanxhypoplasia (14). The condition co-segregates with mutationsin the homeobox transcription factor HOXD13 which has mul-tiple functions during limb pattern formation (15,16).

The molecular mechanism of SPD has only been resolvedfor mutations residing in or near the protein’s C-terminalhomeodomain or within the N-terminal polyalanine tract

†This study was approved by the local Ethical Board of the University Hospital Leuven. The work was performed at the University of Leuven.

∗To whom correspondence should be addressed. Tel: +32 16346196; Fax: +32 16346200; Email: [email protected]

# The Author 2012. Published by Oxford University Press. All rights reserved.For Permissions, please email: [email protected]

Human Molecular Genetics, 2012, Vol. 21, No. 11 2464–2475doi:10.1093/hmg/dds060Advance Access published on February 27, 2012

by guest on October 4, 2013

http://hmg.oxfordjournals.org/

Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from













(reviewed in 17). In the former case, they result in failure toactivate target genes (18,19), whereas in the latter, theyinduce ectopic cartilage formation in the interdigital mesen-chyme (20,21). In this study, we describe a five-generationPakistani family with a complex type of SPD caused by aglycine-to-alanine substitution at the N-terminal end of theHOXD13 protein. The elucidation of the genotype–phenotypecorrelation of this novel HOXD13 mutant is of particular inter-est since (i) no missense mutations have been described thatreside at the N terminus of the protein, (ii) no N-terminal func-tional domains other than the HOXD13 polyalanine tract havebeen characterized to date and (iii) the proband is one of thefew patients homozygous for an HOXD13 mutation (19).

We show that the HOXD13(G11A) mutation does not alterthe subcellular localization, the DNA-binding activity nor thetranscription regulation function of HOXD13 in vitro, butinstead causes a significant reduction in the protein half lifewhen compared with the wild type. Our results suggest thatthe clearance of HOXD13 may not only be mediated by pro-teasomal degradation, as was postulated before (22), but alsoinvolves a significant contribution of the lysosomal pathway.

In addition, we show that chick limb misexpression ofmutated HOXD13 results in preaxial polydactyly or anteriorectopic cartilages in the autopod, recapitulating the humanSPD phenotypes. Finally, our results indicate that this pheno-type is possibly caused at the molecular level by an increase inaffinity of HOXD13(G11A) for Gli3R and directing it intopremature degradation. This results in depletion of Gli3R inthe anterior limb bud leading to the formation of extradigits. Taken together, we identify a novel functionaldomain in HOXD13 and demonstrate an unappreciated func-tion of HOXD13 in regulating digit number through its inter-action with GLI3R.

RESULTS

A G11A missense mutation at the N terminus of HOXD13causes a complex type of SPD in humans

We sequenced the exons of HOXD13 in a five-generationPakistani family with two loops of consanguinity displayingcomplex type II SPD (OMIM 186000) (Fig. 1A). Sequence

Figure 1. A novel G11A missense mutation in HOXD13 causes a complex type of SPD in humans. (A) Pedigree drawing of a five-generation Pakistani familywith a complex type of SPD. (B) Electropherograms showing the DNA sequences obtained from an unaffected control individual (on the left) and of the proband(on the right). The SPD is associated with a point mutation at position 32 of the HOXD13 coding sequence causing a G11A substitution. (C and D) Phenotype ofthe mother of the proband who is heterozygous for the G11A mutation. (E–H) The proband is homozygous for the G11A mutation and shows severe SPD of thehands and feet. See text for the clinical report.

Human Molecular Genetics, 2012, Vol. 21, No. 11 2465

analysis of the proband revealed a homozygous G-to-C muta-tion at position 32 of the coding sequence (Fig. 1B) resultingin a glycine-to-alanine conversion within an evolutionarilyconserved region of HOXD13 with a hitherto unknown func-tion. This G11A mutation has independently been identifiedby a Japanese group in a screening of patients with SPD(23). The parents of the proband and the other affectedfamily members were found to be heterozygous for theG11A mutation, while the mutation was not detected in 50 un-related unaffected controls.

The mother of the proband displayed a mild SPD phenotypeconsisting of bilateral camptodactyly of the fifth fingers(Fig. 1C) and a discrete shortening and external rotation ofthe fourth and the fifth toes (Fig. 1D). The proband wasborn with severe bilateral hand and foot anomalies with bilat-eral cutaneous and osseous syndactyly between the third andthe fourth fingers and absent or rudimentary nails in the syn-dactylous web and in the fifth finger of the right hand(Fig. 1E–H). Radiographs revealed a complex type II SPDwith clinical overlap with brachydactyly type A1 and B(Fig. 1E and F). The homozygous phenotype also includesbroadening of the left metacarpals, a supernumerary digit inthe syndactylous web that is fused to digit three, and bilateralosseous fusions of the distal phalanges. Furthermore, allfingers were short with small or absent middle phalangesand a rudimentary distal phalange of the first finger of theright hand. In the feet, the proband had short, mediallydeviated halluces and plantarflexion of the syndactylizedmass of the fourth and the fifth toes (Fig. 1H).

The N terminus of HOXD13 is essential to determineprotein stability in vitro, a function that is impaired by theG11A mutation

The G11A mutation maps to a region of HOXD13 with a hith-erto unknown function. To understand the molecular basis ofthe phenotype, we therefore investigated the effect of the mu-tation on various biochemical properties of the protein. Wecould not detect differences in subcellular localization, DNA-binding or transcriptional activity in immunofluorescence,electrophoretic mobility shift assays (EMSAs) or reporterassays, respectively (Supplementary Material, Fig. S1). Next,we investigated the intracellular half life of HOXD13(G11A)by arresting de novo protein synthesis using cycloheximide(CHX) and assessing the rate of degradation of wild-type ormutant HOXD13 by immunoblotting. As reported previously(18), approximately half of the initial amount of wild-typeHOXD13 was still detectable at 10 h after CHX treatment(Fig. 2A and B). However, the G11A mutation reduced thehalf life of HOXD13 by 5-fold (Fig. 2A and B). Importantly,another published HOXD13 mutation, HOXD13(G220V) (18),does not alter the stability of the protein to the same extent asthe G11A mutant.

This prompted us to investigate whether the N-terminal partis necessary and sufficient to determine the stability ofHOXD13 in vitro. We generated expression constructs encod-ing the first 78 amino acids of wild-type or mutant HOXD13(including the polyA tract and the short preceding polyS andpolyA tract) as green fluorescent protein (GFP) fusion pro-teins. The respective constructs, further referred to as

HOXD13N-term-78(WT)-GFP and HOXD13N-term-78(-G11A)-GFP, respectively (Fig. 2C, left panel), were used fortransient transfection in HEK293 cells. Twelve hours aftertransfection, the cells were exposed to CHX for 4 h, uponwhich the total cell number and the amount of GFP-positivecells was counted using fluorescent microscopy (Fig. 2D,left panel). Introduction of the G11A mutation significantlydecreased the number of GFP-positive cells when comparedwith its wild-type counterpart (Fig. 2D, left panel). Thesedata suggest that the N-terminal part of HOXD13 is aprotein region required for the correct turnover of theprotein within cells, since inserting a G11A mutation acceler-ates protein degradation.

To map further the N-terminal domain necessary and suffi-cient to exert this function, we also tested shorter GFP fusionproteins in which only the first 31 amino acids of HOXD13(excluding the polyA tract, including a short polyS stretch)were linked N-terminally to GFP [HOXD13N-term-31

(WT)-GFP and HOXD13N-term-31(G11A)-GFP in Fig. 2C,right panel]. The total cell number and the amount ofGFP-positive cells were assessed by Fluorescence-ActivatedCell Sorting (FACS) analysis after 4 h of CHX exposure.The results depicted in Figure 2D (right panel) suggest thatthe N-terminal stretch of 31 amino acids in HOXD13 isinvolved in determining protein stability in vitro. To confirmthese data, we assayed the half life of the respective GFPfusion proteins, and found that the G11A mutation negativelyinfluences the stability of the GFP fusion protein more than thewild type (Supplementary Material, Fig. S2). Taken together,we demonstrated that the integrity of the N-terminal 31amino acids of HOXD13, excluding the polyalanine tract, iscrucial to the stability of the protein, as shown by an acceler-ated degradation of its mutant (G11A) counterpart.

HOXD13 is degraded by the lysosomal pathway

Misfolded or mutated proteins are intracellularly degraded bytwo major pathways that are either lysosome- or proteasome-dependent. Since the half life of the G11A mutant protein wassignificantly shorter than the wild type, we investigated whichof the two main degradation pathways participated in thatprocess. To accomplish that, HEK293 cells transiently overex-pressing wild-type HOXD13 were treated with CHX to blockprotein translation and were simultaneously exposed to eitherlactacystin or chloroquine to block the proteasomal or thelysosomal pathway respectively. Following a 4h incubationperiod, protein levels in cell lysates were analyzed bySDS-PAGE and immunoblotting and quantified by digitalimage densitometry.

As shown in Figure 3, there was a significant decrease of thewild-type HOXD13 protein level in cells treated with CHXalone or with CHX in combination with the proteasomeblocker lactacystin. In contrast, protein levels of HOXD13 incells treated with a combination of CHX and a lysosomeblocker (chloroquine) were largely unaffected (Fig. 3), indicat-ing that wild-type HOXD13 was degraded through the lyso-somal pathway. Similar results were obtained with MG132or NH4Cl as a proteasome or lysosome blocker respectively(data not shown).

2466 Human Molecular Genetics, 2012, Vol. 21, No. 11

http://hmg.oxfordjournals.org/lookup/suppl/doi:10.1093/hmg/dds060/-/DC1


Misexpression of HOXD13(G11A) causes patterningdefects in the developing chick limb

To further explore the molecular basis of SPD induced by theHOXD13(G11A) mutation, we used the RCAS retroviralsystem to misexpress wild-type or mutant HOXD13 in thedeveloping chick limb (reviewed in 24). Replication-competentRCASBP(A) retroviral vectors expressing full-lengthHOXD13(WT) or HOXD13(G11A) were injected in ovo intothe presumptive right hind limb field of stage Hamburger Ham-ilton (HH) 10 developing chick embryos (25), thereby leavingthe left uninjected limb as a contralateral control. Injected

embryos were harvested at stage HH34 and skeletal morph-ology was assessed by Alcian blue (for cartilage) and Alizarinred staining (for mineralized tissue including bone).

In agreement with previous observations (18,26,27), misex-pression of wild-type HOXD13 induced (i) a marked shorten-ing of the cartilaginous elements of the stylopod andzeugopod, (ii) a delay in ossification and (iii) a premature ar-ticulation or fusion of the fibula to the fibulare (Fig. 4A andB). The resulting developmental defects were classified asmild, moderate or severe according to the length of the tibia,as previously described [Supplementary Material, Fig. S3Band (18)]. Of the embryos displaying limb defects (n ¼ 17),

Figure 2. Protein stability is determined by the N terminus of HOXD13 and is impaired by the G11A mutation. (A) Immunoblot of total cell extracts from COS7cells transfected with HA-tagged HOXD13(WT) or HOXD13(G11A) expression constructs, respectively. Twenty-four hours after transfection, cells were treatedwith 200 mg/ml CHX and harvested at the indicated time points. HOXD13 was visualized using an HA-tag antibody. (B) Protein levels were determined bydigital image densitometry and normalized against GAPDH, used as a loading control. Values at each time point of the graph is given as a percent of theprotein amount at time zero (100%). (C) Plasmids for expression of the first 78 (left panel) or 31 (right panel) amino acids of wild-type or mutant HOXD13as GFP fusion proteins. (D) Cell count based on immunofluorescence (left panel) or FACS analysis (right panel) of HEK293 cells transiently transfectedwith the respective HOXD13-GFP fusion constructs. The number of GFP-positive cells was normalized for transfection efficiency using co-transfection withRFP-expressing constructs. Values of untreated cells were set as 100% and compared with the cell count in samples treated with 200 mg/ml CHX for 4 h.Cells were counted by two independent investigators and statistical comparisons of three independent experiments were made using the Student’ t-test, errorbars represent SEM.



29% showed a mild, 42% a moderate and 29% a severe pheno-type when misexpressing wild-type HOXD13 (Fig. 4A and D).Misexpression of HOXD13(G11A) induced similar skeletal ab-normalities, but the embryos (n ¼ 37) consistently showedmore severe phenotypes (Fig. 4B), with a phenotype distribu-tion of 11% for mild, 43% for moderate and 46% for severephenotypes respectively (Fig. 4B and D). In addition, asreported previously (18,26,27), misexpression ofHOXD13(WT) did not produce phenotypic changes in theautopod, where endogenous chicken Hoxd13 is normallyexpressed. Intriguingly, in contrast to the phenotypesinduced by the misexpression of wild-type HOXD13, misex-pression of HOXD13(G11A) induced anterior ectopic cartilageor digit formation in 32% of the injected embryos (Fig. 4C).

HOXD13(G11A) produces distal limb malformations bymodulating the interplay between HOXD13, dHand andGli3 prior to Shh induction in the ZPA

Thus, misexpression of HOXD13(G11A) in the developingchick limb induces phenotypes similar to those induced bythe same mutation in SPD patients (including preaxial poly-dactyly or ectopic cartilage formation). In order to facilitate

the investigation of the pathogenic mechanism, we co-infecteddeveloping chick limbs with RCASBP(B)-GFP and wild-typeor mutant RCASBP(A)-HOXD13. This strategy allowed us tostudy only the limbs that display a uniform expression patternof GFP prior to the detection of the first macroscopic signs ofthe skeletal phenotype caused by HOXD13 misexpression (seeSupplementary Material, Fig. S3A for details).

The induction of extra anterior digits are frequently asso-ciated with ectopic anterior Shh mRNA expression and/or an-teriorly extension of the dHand mRNA expression domain(28,29). Thus, following selection of co-injected embryos foruniform GFP expression in the right hind limb only (n ¼20), the effect of HOXD13(WT) or HOXD13(G11A) misex-pression on the expression pattern of Shh and dHand wasexamined by whole-mount in situ hybridization (WISH) atstage HH21. Misexpression of wild-type or mutant HOXD13did not have a detectable effect on the apparent level or loca-tion of Shh mRNA expression (i.e. no detectable Shh expres-sion in the anterior part of the limb in 100% of the analyzedembryos (n ¼ 10; Fig. 5, upper panels). As anticipated (18)however, misexpression of wild-type HOXD13 inducedectopic dHand expression in 60% of the analyzed embryos(n ¼ 5; Fig. 5, lower panels). This effect was more pronouncedin embryos injected with RCAS-HOXD13(G11A) (n ¼ 5),which, in extreme cases, resulted in expansion of the dHandexpression domain throughout the entire injected hind limb(Fig. 5, lower panels).

HOXD13(G11A) sequesters Gli3R through a directinteraction, followed by degradation of the proteincomplex

During normal limb pre-patterning, dHand is transcriptionallyrepressed by the repressor form of Gli3 (Gli3R) (29). Since the ex-pression domain of dHand was anteriorly expanded uponHOXD13 misexpression, it could be possible that HOXD13(wild type and even more so the G11A mutant) is titrating outthe available GLI3R that results in preaxial polydactyly. To testthis hypothesis, we carried out a co-immunoprecipitation(co-IP) and competition experiment of HOXD13(G11A) in thepresence of HOXD13(WT) and limiting amounts of Gli3R(Fig. 6). Briefly, HEK293 cells were co-transfected with myc-tagged HOXD13(WT), together with limiting amounts ofGli3R and increasing amounts of either wild-type or mutantHA-tagged HOXD13 (Fig. 6A). We then performed co-IP toassess the capability of the HA-tagged HOXD13 constructsto compete with the myc-tagged HOXD13(WT) for binding toGli3R. As described previously (30), wild-type HOXD13 andGli3R co-immunoprecipitated specifically and selectively whenco-expressed in HEK293 cells (Fig. 6B). We found that theHA-tagged G11A mutant is able to compete with myc-taggedHOXD13(WT) for binding to Gli3R (Fig. 6B). These results dem-onstrate that the G11A mutant binds Gli3R with a higher affinitythan wild-type HOXD13.

These results, in conjunction with the strong reduction of thehalf life of the G11A mutant, led us to speculate thatHOXD13(G11A) might sequester Gli3R into degradation. Toverify this assumption, we determined the half life of Gli3R inthe presence of wild-type or mutant HOXD13 respectively. Wecould show that HOXD13(G11A) significantly reduced the half

Figure 3. Wild-type HOXD13 is degraded by the lysosomal pathyway. Graphrepresenting protein levels of HEK293 cells transfected with wild-type ormutant HOXD13 and left untreated (CTR) or exposed to CHX or CHX com-bined with either lactacystin or chloroquine for 4 h. Protein levels were quan-tified by digital image densitometry and normalized against GAPDH. Valuesare given as a percentage of the protein amount in the untreated CTR sample(100%). Statistical comparisons were made using Student’ t-test in referenceto non-treated cells (∗∗) or to cells treated with CHX (∗). Results are represen-tative of three different experiments in two different cell types (DF1 andHEK293 cells).



Figure 4. Misexpression of HOXD13(G11A) induces patterning defects in developing chick limbs. (A and B) Dorsal views of stage HH34 hind limbs stainedwith Alcian Blue for cartilage and Alizarin Red for mineralized tissue including bone. Injected hind limbs are on the right, uninjected contralateral control limbson the left in every panel. Embryo injected with RCAS-HOXD13(WT) (A) or RCAS-HOXD13(G11A) (B) showing a severe phenotype. Red and black barsrepresent the tibia length of injected and contralateral control hind limbs, respectively. Arrow indicates bifurcation of digit 1 in the hind limb injected withmutant HOXD13. (C) Other examples of severe HOXD13(G11A) misexpressing embryos (only injected hind limbs are shown). Open arrow designates anextra digit 2. (D) Distribution of phenotypes produced by misexpression of HOXD13(WT) and HOXD13(G11A). Phenotypes were classified as mild (blue), mod-erate (red) or severe (green) according to the length of the tibia, for example, see Supplementary Material, Figure S3B.

Figure 5. HOXD13(G11A) does not upregulate Shh expression, but induces significant ectopic expression of dHand when misexpressed in the chick. WISH forShh or dHand on stage HH20-21 chick embryos injected with RCAS-HOXD13(WT) or RCAS-HOXD13(G11A). The injected right hind limb is on the right andthe contralateral control limb on the left in every panel.



life of Gli3R in vitro when compared with the wild type. Al-together, these findings indicate that HOXD13(G11A), througha direct interaction, sequesters Gli3R and that this complex is sub-sequently targeted for degradation.

DISCUSSION

The importance of the SHH/GLI3R loop during developmentof the limb in general and of the autopod in particular hasbeen extensively documented (29,31). One of the proposedfunctions of the SHH signaling cascade is the imposition ofpentadactyly on the developing autopod (6,31). The inter-action of HOXD13 with GLI3R, a component of thatpathway, has been described before (30) but the in vivo bio-logical significance of that interaction has not been fully elu-cidated. Our results show that HOXD13, through itsinteractions with GLI3R, fine tunes the number of the digits.This interaction is so strong that changes in the protein halflife of HOXD13 have a direct effect on the availability ofGLI3R. Our work corroborates the earlier work of Chenet al. (30) showing that HOXD13, also through the interactionwith GLI3R, participates in the regulation of the pentadactyly.We postulate that the HOXD13–GLI3R interaction is one ofthe primary regulatory signals delineating the autopod fieldand the number of skeletal elements within it.

The importance of HOXD13 in limb development has beenillustrated before since distinct mutations in the HOXD13 geneare associated with skeletal malformations in both humans andmice (15,16 and reviewed in 17). All SPD-causing HOXD13mutations identified so far have been mapped to theN-terminal polyalanine tract or to the C-terminal

homeodomain, causing typical or atypical features of SPD, re-spectively (reviewed in 17). The molecular basis of these phe-notypes has recently been investigated (19–21). There arehowever many intriguing aspects of HOXD13 biology to beunraveled, including the characterization of other functionaldomains for instance through the analysis of (missense) muta-tions outside of the polyalanine tract or the homeobox.

A complex type of SPD in humans is associated with anovel G11A missense mutation at the N terminus ofHOXD13

We present a novel case of SPD linked to a G11A missensemutation in a previously uncharacterized region ofHOXD13. Taking into consideration the classification ofSPD phenotypes published by Malik et al. (14), we concludethat affected family members present some of the canonicalfeatures of SPD, in particular the severe phenotype of thehomozygous proband including the cutaneous and osseousfusions in hands and feet and the involvement of metacarpalsand metatarsals. In addition to the classical features of SPD,heterozygous individuals showed some minor variants fre-quently associated with SPD, including camptodactyly or clin-odactyly of the fifth fingers, and brachydactyly of the fourthand the fifth toes. Finally, there were also signs of overlapwith Brachydactyly type B since the proband displayed rudi-mentary or hypoplastic distal phalanges and absent or under-developed nails.

In general, the SPD phenotypes of the affected familymembers correlate with the traits of the severe end of thephenotypic spectrum, with the homozygous phenotype being

Figure 6. HOXD13(G11A) induces preaxial polydactyly by negatively influencing anterior Gli3R function. (A) Experimental setup of co-IP of Gli3R andHOXD13. HEK293 cells were co-transfected with myc-tagged HOXD13(WT), together with limiting amounts of Gli3R and either wild-type or mutantHA-tagged HOXD13. (B) Co-IP was performed using an antibody against the Xpress-tag of Gli3R. Arrowhead indicates the band corresponding withHOXD13 (visualized using an HA-tag antibody). (2), untransfected control. (C) The in vitro half life of Gli3R is compromised in the presence ofHOXD13(G11A). Graph representing quantification of Gli3R protein levels (normalized against GAPDH) in COS7 cells transfected with HA-taggedHOXD13(WT) or (G11A) in combination with limiting amounts of Gli3R. Cells were treated with CHX and harvested at the indicated time points. Cell extractswere analyzed by immunoblotting using an HA-tag antibody (for HOXD13) and an affinity purified Gli3 antibody. (D) Model of the mechanism by whichHOXD13(G11A) induces polydactyly in the chick. In wild-type limbs (left panel), a balanced mutual genetic antagonism between Gli3R (anterior) anddHand and the 5′Hoxd genes (including Hoxd13) results in posterior restriction of their expression domain, with subsequent induction of Shh expression inthe ZPA. Upon HOXD13(G11A) misexpression, the short-lived mutant protein binds Gli3R with a higher affinity than HOXD13(WT), thereby promotingGli3R degradation. The enhanced clearance of Gli3R from the anterior limb bud then releases its inhibitory action followed by induction of preaxial polydactyly.


the most prominent one. Moreover, these phenotypes signifi-cantly overlap with those induced by deregulation of theSHH signaling loop during limb development (reviewed in32,33). The finding that HOXD13 directly interacts withGLI3 opens a possibility that HOXD13 mutations might actby interfering with the SHH signaling cascade at the level ofGLI3.

HOXD13 protein stability is determined by its N terminusand is negatively affected by the G11A mutation

Proteins, in particular those with regulatory activity (asopposed to the structural ones), are degraded upon completionof their function. For instance, transmembrane receptors,ligands or transcription factors are degraded, sometimesrapidly, in order to fine-tune their regulatory activities. There-fore, it is of great interest to identify and characterize mechan-isms regulating the intracellular pool size of the proteins. Oneof the best described cases is the intracellular processing anddegradation of beta-catenin, a transcriptional co-activator ofWnt signaling (34). In the case of HOX transcription factors,and in particular HOXD13, very little data considering theintracellular processing and the protein domains involved inthat process is currently available (22,35). Molecular analysisof the HOXD13(G11A) mutation, located in an uncharacter-ized region of the protein, thus provides a tool to gaininsight into functional domains of HOXD13 other than thehomeodomain or the polyalanine tract.

We were unable to detect any difference in DNA-bindingcapacity or transcriptional activity between wild-type andmutant HOXD13. However, we discovered that the G11A mu-tation causes a dramatic reduction of the half life of HOXD13,unlike any other published HOXD13 mutations (18). By deter-mining the half life of HOXD13-GFP fusion proteins, we alsodemonstrated that the N-terminal 31 amino acids of HOXD13,excluding the polyalanine tract, are necessary and sufficient todetermine the protein stability in vitro. A significant aspect ofthis function is most likely to be mediated by the first stretchof amino acids since we could show that the G11A mutationdestabilizes the rest of the protein significantly more than itswild-type counterpart. Thus, in addition to the previously char-acterized polyalanine tract and the homeodomain, these find-ings reveal the existence of a novel functional domainwithin the N terminus of HOXD13 that, in particular whenmutated, dramatically decreases the intracellular half life ofthe protein.

The HOXD13(G11A) mutation could affect protein stabilityin at least two ways. It could create a novel protease targetsite and further studies will be required to determine the exacttarget sequence. Alternatively, it is possible that the G11A mu-tation alters the structural conformation of HOXD13 leading toaggregation and accelerated degradation of the protein. It hasbeen reported that HOXD13 mutants, such as HOXD13(+14Ala), carrying a 14-residue polyalanine tract expansion muta-tion, and HOXD13(G220V) are prone to form cytoplasmicaggregates in addition to its normal nuclear localization(18,22). We could not find a difference in subcellular localiza-tion between wild-type and the HOXD13(G11A) mutant. It isunclear whether these results imply that the mutant is notretained in cytosolic inclusions or whether it does form

cytoplasmic aggregates, but at low production rates so thatthey do not overload the protein’s degradation machinery.The latter mechanism has been shown by Albrecht et al. (22)for HOXD13 polyalanine tract expansion mutations. Interest-ingly, this group postulates that polyalanine tract expansionmutants, as many other misfolded proteins, form cytoplasmicaggregates that are mainly degraded by the proteasome.

It is the interconnection of three main pathways that is es-sential for the management of misfolded proteins and subse-quent cell viability (36): (i) the molecular chaperonemachinery—including heat shock proteins (hsp)—assists inproper protein folding, (ii) the proteasome pathway degradesmisfolded proteins and (iii) the lysosome pathway deliversprotein aggregates to autophagy for clearance (36). It shouldbe noted however that in fact these pathways are not mutuallyexclusive (36,37), and several lines of evidence point toward acombined mechanism of action. Indeed, Albrecht et al. (22)and Kopito (38) showed that the amorphous aggregates aredifferent from classical protein aggregates that are well-structured or associated with aggresome formation within theso-called microtubule-organizing center. Once aggregated,misfolded proteins are not efficiently degraded by proteasomes(36). This, in accordance with our data that clearly show thatHOXD13 is degraded through the lysosomal pathway, couldtherefore indicate that both proteasomal or lysosomal degrad-ation is involved in HOXD13 degradation. Moreover, itremains to be resolved whether the G11A mutation, likeother Hoxd13 mutants (39), exerts a dominant-negativeeffect over the wild-type protein, confirming again-of-function pathogenic mechanism.

Misexpression of HOXD13(G11A) induces patterningdefects in the chick limb by disturbing the anterior Gli3Rfunction

Human limb malformations caused by HOX gene mutationshave been studied in the chick model system by misexpressingthe mutant genes from replication-competent RCAS retro-viruses (18,27,40). To investigate the in vivo effect of theG11A mutation during chick limb development, we imple-mented an adapted, dual RCAS injection strategy in whichwe combined two RCAS viruses expressing distinct types ofenvelope glycoproteins in order to bypass receptor interfer-ence (24) and superinfect chicken cells simultaneously withboth HOXD13 and GFP. This enables the selection ofembryos that express the transgenes uniformly and exclusivelythroughout the right hind limb, thereby enriching the pool ofembryos representing the severe end of the phenotypic spec-trum. This technique thus allows a significant reduction inintra-assay variability and thereby opens new avenues forthe future study of the in vivo effects of limb patterninggenes with improved accuracy.

It should be noted that RCAS-based HOXD13 misexpressionin the chick reproduces some aspects of human SPD (18,26,27).Misexpression of wild-type HOXD13 has been demonstrated tocause limb defects mainly at the level of the stylopod, zeugopodand proximal autopod, where endogenous chicken Hoxd13 isnot expressed (18,26,27). Indeed, HOXD13(WT) misexpressingembryos displayed a delay in ossification, reductions in size ofthe zeugopodal and stylopodal skeletal elements, sometimes


associated with a premature articulation or fusion of the fibulato the fibulare, which corroborates previously published data(18,27). These findings are in contrast with the polydactyly phe-notypes of wild-type Hoxd13 misexpression in mice (41). Theabsence of any phenotypic changes in the distal autopod inthe chick are presumably correlated with the level ofHOXD13(WT) misexpression.

When misexpressing HOXD13(G11A), the same phenotypicclasses as upon misexpression of wild-type HOXD13 wereobserved [mild, moderate or severe, Supplementary Material,Fig. S3B and (18)], but with a higher incidence ofmoderate-to-severe phenotypes. Interestingly, the G11A muta-tion induced anterior ectopic cartilage or digit formation inone-third of the injected embryos. These findings suggestthat the G11A mutation might act through a gain-of-functionmechanism, since it also affects the distal autopod in contrastto the wild-type or other mutants (18,26,27). This particularmutation is thus powerful enough to phenocopy typicalhuman SPD phenotypes in the chick, providing a novel toolto study the molecular basis of the disorder in detail.

It has been generally accepted that ectopic anterior Shh signal-ing is a major cause of preaxial polydactyly in mice, chicks andhumans (42). However, deregulation of genes involved in pre-patterning of the limb prior to ZPA establishment and Shh induc-tion, including an imbalance of the mutual antagonism of Gli3R(anterior) and dHand (posterior), has also been found to beimplicated in ectopic digit development (29). To explore the mo-lecular mechanism by which the G11A mutation induces the pre-axial ectopic digit or cartilage formation, we examined theexpression pattern of both dHand and Shh using WISH onstage HH21 embryos. As anticipated (18), misexpression ofwild-type HOXD13 resulted in ectopic expression of dHand.This effect was even more pronounced in HOXD13(G11A) mis-expressing embryos which, in extreme cases, resulted in a com-plete failure to posteriorly restrict dHand leading to expansion ofits expression domain throughout the entire injected hind limb.These results suggest that HOXD13(G11A), in contrast to thewild type, induces preaxial polydactyly by upregulating dHandexpression above a certain threshold in the anterior part of thelimb bud. These phenotypes are in agreement withloss-of-function mutations in Lst (Alx42/2) (43) or Xt(Gli32/2) mice (44), as well as with gain-of-function experi-ments with dHand (28) or other 5′Hoxd genes (4,45). In fact,these phenotypes are attributed to ectopic Shh activity in the an-terior mesenchyme. However, in our experiments, we were notable to detect ectopic Shh mRNA expression in both wild-typeor mutant injected limbs. The same observations were madeby Fernandez-Teran et al. (28) in embryos misexpressingdHand. It could be argued that the Shh pathway is activatedeither indirectly, without induction of Shh itself, or transientlyand at levels below the detection limit of WISH (28). Thelatter effect has also been observed in other recent studies (46)as well as in the Gli3Xt/Xt/Hoxd11-132/2 mouse, in whichectopic activation of Shh due to a combined loss of Gli3 andHoxd11-13 function is usually small, transitory and sometimesdifficult to detect (6). In addition, ectopic Shh activation inGli32/2 or in Tg-Hoxd12;Gli3+/2 limbs is also thought to rep-resent a secondary ‘readout’ of Hedgehog pathway activation viareduced Gli3R, rather than a direct cause for the polydactylyphenotypes (30).

We favor the hypothesis that the occurrence of preaxial poly-dactyly in HOXD13(G11A) misexpressing embryos occurs priorto Shh signaling and involves deregulation of the Gli3 repressorform. The mutual antagonistic action of Gli3R (anterior) anddHand and the 5′Hoxd genes (posterior) is indispensable forpre-patterning of the limb along the AP-axis (29). Gli3R isrequired for posterior restriction of both dHand and 5′Hoxdgenes, and is thought to carry out a protective functionagainst the prevalence of posterior Hox genes (15). Further-more, Gli3R has been shown to physically interact withHoxd12 and Hoxd13 (30). In turn, dHand directly binds toHoxd13, and both proteins induce Shh expression by bindingits upstream cis-regulatory region (47,48). Consequently, wehypothesize that the mechanism by which short-livedHOXD13(G11A) mutant decreases the intracellular levels ofGLI3R is through a direct interaction resulting in prematuredegradation of GLI3R. To verify this assumption, we carriedout a co-IP competition experiment of HOXD13(G11A) inthe presence of HOXD13(WT) and limiting amounts ofGli3R, and found that the mutant indeed has a higher affinityfor Gli3R than the wild type. Furthermore, we could showthat HOXD13(G11A) reduces the in vitro half life of Gli3Rmore than that of the wild type. These findings open the possi-bility that HOXD13, through a direct interaction, sequestersGli3R into protein heterodimers that are subsequently targetedfor degradation. Since the G11A mutant has a higher affinityfor Gli3R than its wild-type counterpart, this property couldexplain the dominant effect of the mutation and the heterzygouspatient phenotypes.

Hence, we propose a molecular explanation for the pheno-type induced by the HOXD13(G11A) mutation wheremutated HOXD13 depletes the anterior levels of Gli3R bydirecting it into degradation (Fig. 6D). A premature decreasein GLI3R consequently results in the ectopic anterior induc-tion of dHand expression (Fig. 6D). Accordingly, theabsence of ectopic digit formation when misexpressingHOXD13(WT) is likely attributable to a threshold effect, bywhich sufficient amounts of Gli3R remain to posteriorly re-strict dHand expression to some extent. It should be notedhowever, that the enhanced binding of Gli3R byHOXD13(G11A) may also have an effect on Gli3R functionbesides affecting its stability. Both mechanisms are not neces-sarily mutually exclusive, and they could therefore both con-tribute to the G11A phenotype.

Taken together, our results suggest that the SPD in patientsharboring an HOXD13(G11A) mutation is caused by depletionof GLI3R levels during early limb development through intracel-lular protein degradation. Furthermore, these findings argue infavor of a more prominent role of HOXD13 during pre-patterning of the limb, co-operating at the same level of dHandin the antagonism of anterior Gli3R function and establishmentof anterior–posterior polarity in the developing limb bud.

MATERIALS AND METHODS

Patient evaluation and mutation analysis

The family of Pakistani origin was identified following referralto the Clinical Genetics Unit, Birmingham Women’s Hospital,UK. The proband and his mother were clinically examined,




and radiographs of both hands and feet were obtained. Venousblood samples for genomic DNA extraction were obtainedfrom the proband, both parents and other family members,as well as 50 unaffected controls, with their informedconsent and approval from the local ethical committee. Toscreen for mutations in the HOXD13 gene, the coding regionwas amplified by PCR in four segments as described in (49).Cycle sequencing of the amplified fragments was performedon an ABI 377 automated sequencer (Applied Biosystems)either directly (Applied Biosystems Prism Dye TerminatorKit) or after cloning the PCR fragments into pCR-Script(Stratagene).

Expression constructs

The Xpress-tagged pcDNA3.1-Gli3R expression construct waskindly provided by S. Mackem (NIH, USA). pSG5-HOXD13(+14Ala) was described before (18). pcDNA3.1-HOXD13(WT) was generated by cloning the human HOXD13coding sequence into the EcoRI/BamHI sites of pBS-SK+vector (Stratagene) and subsequent subcloning into the HindIII/XbaI sites of pcDNA3.1(+) (Invitrogen). HA-taggedHOXD13(WT) expression constructs were produced by subclon-ing the open reading frame (ORF) of wild-type HOXD13 into theBamHI/XbaI sites of pcDNA3.1-HA(C) (Invitrogen). The G11Amutation was introduced by PCR mutagenesis using the primers5′-ATGCGGATCCATGAGCCGCGCCGGGAGCTGGGACATGGACGCGCTGCGGGCAGA-3′ and 5′-CCATGTACTTCTCCACGGGAAAGCCTCCCAGCGAGGCGTGC-3′. Theamplified fragment was sequence-verified and substituted forthe corresponding wild-type BamHI/NotI within pcDNA3.1-HOXD13(WT) and pcDNA3.1-HA-HOXD13(WT). Myc-HIS-tagged pcDNA3.1-HOXD13(WT) was generated by amplifyingthe ORF of HOXD13(WT) using primers 5′-ATATATGGATCCATGAGCCGCGCCGG-3′ and 5′-TATATATCTAGAGGAGACAGTA-TCTTTGAGCTTGGAG-3′. The ampliconwas then cloned into the BamHI/XbaI sites of pcDNA3.1-myc-HIS(A) (Invitrogen). Plasmids expressing GFP fusion pro-teins were obtained by cloning the N-terminal 78 or 31 aminoacids of HOXD13 translationally in frame with the eGFPcoding sequence in the pEGFP-N1 vector (Clontech). For this,the N-terminal fragments of the wild-type or mutant HOXD13coding sequence were amplified by PCR using 5′-TATATAAGCTTATGAGCCGCGCCGGG-3′ as a forward primer andreverse primers 5′-ATATGGATCCACCGATGAGGAGGAGGAAGAGGC-3′ and 5′- TATATGGATCCCCGG-GGTACGCAAAG-3′ for HOXD13N-term-31 and HOXD13N-term-78, respect-ively. The respective amplicons were subsequently cloned intothe HindIII/BamHI sites of pEGFP-N1 to generate pEGFP-HOXD13N-term-31 and pEGFP-HOXD13N-term-78, respectively.An RCASBP(A) and RCASBP(B) avian retroviral vector forthe expression of GFP was kindly provided by M. Logan(London, UK). RCAS constructs expressing humanHOXD13(WT) and HOXD13(G11A) were generated by amplify-ing the ORF of full-length wild-type or mutant HOXD13 by PCRusing primers 5′-TATATAGACGTGCAGCCGACCACCATGAGCCGCGCCGGGAGCTG-3′ and 5′-TCTCAG-ATAGGTTCGTAGCAGCCG-3′. The amplicons were cloned into theBmgBI site of the shuttle vector pSLAX13-HOXD13(WT),

followed by subcloning of the entire ClaI-adapter fragment intothe ClaI-linearized retroviral pRCAS-BP(A) vector (50).

Antibodies and reagents

Commercially available antibodies were mouse monoclonalHA tag antibody (Eurogentec), rabbit polyclonal HOXD13antibody (kind gift of S. Mackem, NIH, USA), rabbit poly-clonal 6xHIS tag antibody (Abcam), c-myc monoclonalmouse antibody (Santa Cruz Biotechnology), affinity purifiedrabbit polyclonal Gli3 antibody (gift of S. Mackem), rabbitIgG Trueblot (eBioscience), mouse monoclonal GAPDH anti-body (Ambion), rabbit polyclonal GFP antibody (gift ofM. Quattrocelli, Leuven, Belgium), secondary HRP-conju-gated goat anti-mouse (Jackson), Alexa Fluor 488-coupleddonkey anti-mouse (Molecular Probes). Following reagentswere used: CHX (Sigma), NH4Cl (Merck), lactacystin(Sigma), chloroquine (Sigma), MG132 (Merck).

Cell culture and transfections

For overexpression studies, immortalized chicken embryonicfibroblast (DF1) cells, COS7 cells or HEK293 cells were cul-tured in Dulbecco’s Modified Eagle’s Medium (DMEM) sup-plemented with glucose (4.5 g/l, Gibco-Invitrogen) with 10%(v/v) fetal bovine serum (Gibco-Invitrogen), antibiotics/antimy-cotics (100 U/ml penicillin, 100 mg/ml streptomycin, 0.25 mg/ml amphotericin B; Gibco-Invitrogen) and 1% (v/v) sodiumpyruvate (Gibco-Invitrogen). All cultures were maintained at378C, 5% CO2 and 95% humidity. Transient transfectionswere carried out using Lipofectamine LTX with Plus Reagentaccording to the manufacturer’s instructions (Invitrogen).

Microinjection and phenotypic analysis of developingchick embryos

Virus production, concentration and titration of the viralsupernatants was performed as described elsewhere (18).Microinjection of stage HH10 chick embryos and phenotypicanalysis (skeletal staining) was performed as outlined before(18,50). For the dual RCAS injection, RCASBP(B)-GFP wasmixed with either wild-type or mutant RCASBP(A)-HOXD13at a 3:7 ratio and subsequently injected as described previously(18,50). WISH using digoxigenin-labeled anti-sense RNAprobes was performed according to Salsi et al. (51). Followingantisense probes were used: Shh (kind gift of M. Logan,London, UK) and dHand. Tibia measurements, phenotypeclassification and statistics were performed as previouslypublished (18).

Protein stability assay, protein degradation assay,immunofluorescence and FACS

The protein stability assay was performed according to Fantiniet al. (18). To determine the subcellular localization, cellswere fixed in methanol and analyzed by immunofluorescenceaccording to standard protocols. To determine the intracellulardegradation pathway, HEK293 or DF1 cells were transientlytransfected with wild-type HOXD13 expressing constructs.Twelve hours after transfection, the cells were trypsinized


and replated in a 24-well plate or 8-well CultureSlide (BDBiosciences) to compensate for differences in transfection ef-ficiencies. When attached, cells were incubated with 200 mg/mlCHX in the presence of 25 mM lactacystin, 20 mM chloroquine,1 mM NH4Cl or 25 mM MG132, respectively. Following 4 h oftreatment, cells were lysed in 40 ml of 1× lithium dodecylsulfate (LDS) sample buffer (Invitrogen), sonicated and ana-lyzed by immunoblotting. To verify the effect of the N ter-minus on protein stability, two 10-cm dishes of HEK293cells were transfected with wild-type or mutant pEGFP-HOXD13N-term-31 or pEGFP-HOXD13N-term-78 respectively.Cells were co-transfected with a plasmid encoding red fluores-cent protein (kind gift of S. Wilson) to normalize for the trans-fection efficiency. Twelve hours after transfection, cells weretrypsinized and replated into a six-well plate, followed byCHX treatment for 8h at a final concentration of 200 mg/ml.The total cell number and the amount of GFP-positive cellswere counted by two independent investigators using fluores-cent microscopy and normalized to the amount ofRFP-positive cells to compensate for differences in transfec-tion efficiency. Alternatively, the ratio of GPF-positive cellsto the total cell count was established by FACS analysis onan FACSCanto cytometer using the BD FACSDiva 5.0 soft-ware (both from BD Biosciences). Briefly, following trypsini-zation and centrifugation at 1050 rpm for 10′, cells wereresuspended and fixed in 2% paraformaldehyde in PBS priorto FACS analysis.

Co-immunoprecipitation

Co-IP was performed by co-transfecting a 10-cm dish ofHEK293 cells with 50 ng (limiting amounts) of Gli3R,200 ng of myc-HIS-tagged HOXD13(WT) and increasingamounts (50, 200 or 500 ng) of HA-tagged wild-type ormutant HOXD13. The DNA content was adjusted to 1 mgusing empty pcDNA3.1 vector (Invitrogen). Twenty-fourhours after transfection, cells were lysed in 400 ml of IPbuffer [10 mM HEPES (pH 7.5), 1 mM EDTA, 250 mM NaCl,0.5% NP-40, protease inhibitors] on ice with trituration. Celllysates were centrifuged at 5000 rpm for 5′, and 600 ml oflysate was immunoprecipitated with 2 mg of anti-Xpress tagantibody with gentle rotation at 48C. Following 3 h of incuba-tion, immunoprecipitates were coupled overnight to Protein GSepharose beads with gentle rotation at 48C. Beads werewashed with IP buffer and protein solubilization was per-formed by adding a ‘bead volume’ of 2× LDS samplebuffer (Invitrogen) supplemented with 0.1 M DTT. Immuno-precipitates were analyzed by immunoblotting on a 12%Bis-Tris gel.

Electrophoretic mobility shift assay

Full-length HOXD13(WT), HOXD13(G11A) andHOXD13(I47L) were synthesized in vitro using theTNT-coupled transcription/translation system (Promega) andused in EMSAs as described in (27) together with the follow-ing oligonucleotide probes: 5′-GCCGTCCGTT-TTATTGGGGACACA-3′ and 5′-GCCGTCCGTTTTACGAGGGACACA-3′. The total amount of reticulocyte lysate in each

binding reaction was adjusted to normalize for the translatedprotein content.

Luciferase reporter assays

Luciferase reporter assays using pT81(TTAT)6 reporter con-structs were performed as described elsewhere (18).

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG online.

FUNDING

This work was partially funded by the Research FoundationFlanders (FWO) (grant number G.0199.07N to P.D. andP.T.). P.D. is a senior clinical investigator of the FWO. N.B.is a PhD fellow from the Research Foundation Flanders(FWO) and recipient of an EMBO short-term Fellowship.This work was also supported by grants from the Italian Asso-ciation for Cancer Research (AIRC) and from the FondazioneCassa di Risparmio di Modena (CRMO) (to V.Z.).

Conflict of Interest statement. None declared.

REFERENCES

1. Zeller, R. (2010) The temporal dynamics of vertebrate limb development,teratogenesis and evolution. Curr. Opin. Genet. Dev., 20, 384–390.

2. Duboc, V. and Logan, M.P. (2009) Building limb morphology throughintegration of signalling modules. Curr. Opin. Genet. Dev., 19, 497–503.

3. Riddle, R.D., Johnson, R.L., Laufer, E. and Tabin, C. (1993) Sonichedgehog mediates the polarizing activity of the ZPA. Cell, 75,1401–1416.

4. Knezevic, V., Hutson, R., DeSanto, R., Schughart, K., Mahon, K., Chiang,C. and Mackem, S. (1997) Role of the Hoxd-12 transactivator inpatterning and potential links to growth regulation. Dev. Biol., 186,A143–A143.

5. Zakany, J., Kmita, M. and Duboule, D. (2004) A dual role for Hox genesin limb anterior-posterior asymmetry. Science, 304, 1669–1672.

6. Sheth, R., Bastida, M.F. and Ros, M. (2007) Hoxd and Gli3 interactionsmodulate digit number in the amniote limb. Dev. Biol., 310, 430–441.

7. Wang, C., Pan, Y. and Wang, B. (2007) A hypermorphic mouse Gli3allele results in a polydactylous limb phenotype. Dev. Dyn., 236,769–776.

8. Wang, B., Fallon, J.F. and Beachy, P.A. (2000) Hedgehog-regulatedprocessing of Gli3 produces an anterior/posterior repressor gradient in thedeveloping vertebrate limb. Cell, 100, 423–434.

9. Benazet, J.D., Bischofberger, M., Tiecke, E., Goncalves, A., Martin, J.F.,Zuniga, A., Naef, F. and Zeller, R. (2009) A self-regulatory system ofinterlinked signaling feedback loops controls mouse limb patterning.Science, 323, 1050–1053.

10. Vortkamp, A., Gessler, M. and Grzeschik, K.H. (1991) GLI3 zinc-fingergene interrupted by translocations in Greig syndrome families. Nature,352, 539–540.

11. Biesecker, L.G. (1997) Strike three for GLI3. Nat. Genet., 17, 259–260.12. Cao, D., Jin, C., Ren, M., Lin, C., Zhang, X. and Zhao, N. (2009) The

expression of Gli3, regulated by HOXD13, may play a role in idiopathiccongenital talipes equinovarus. BMC Musculoskelet. Disord., 10, 142.

13. Temtamy, S.A. and McKusick, V.A. (1978) The genetics of handmalformations. Birth Defects Orig. Artic Ser., 14, i-619.

14. Malik, S. and Grzeschik, K.H. (2008) Synpolydactyly: clinical andmolecular advances. Clin. Genet., 73, 113–120.



15. Zakany, J. and Duboule, D. (2007) The role of Hox genes duringvertebrate limb development. Curr. Opin. Genet. Dev., 17, 359–366.

16. Kmita, M., Tarchini, B., Zakany, J., Logan, M., Tabin, C.J. and Duboule,D. (2005) Early developmental arrest of mammalian limbs lacking HoxA/HoxD gene function. Nature, 435, 1113–1116.

17. Brison, N., Tylzanowski, P. and Debeer, P. (2011) Limb skeletalmalformations—What the HOX is going on? Eur. J. Med. Genet., 55, 1–7.

18. Fantini, S., Vaccari, G., Brison, N., Debeer, P., Tylzanowski, P. andZappavigna, V. (2009) A G220V substitution within the N-terminaltranscription regulating domain of HOXD13 causes a variantsynpolydactyly phenotype. Hum. Mol. Genet., 18, 847–860.

19. Goodman, F.R. (2002) Limb malformations and the human HOX genes.Am. J. Med. Gen., 112, 256–265.

20. Kuss, P., Villavicencio-Lorini, P., Witte, F., Klose, J., Albrecht, A.N.,Seemann, P., Hecht, J. and Mundlos, S. (2009) Mutant Hoxd13 inducesextra digits in a mouse model of synpolydactyly directly and bydecreasing retinoic acid synthesis. J. Clin. Invest., 119, 146–156.

21. Villavicencio-Lorini, P., Kuss, P., Friedrich, J., Haupt, J., Farooq, M.,Turkmen, S., Duboule, D., Hecht, J. and Mundlos, S. (2010) Homeoboxgenes d11-d13 and a13 control mouse autopod cortical bone and jointformation. J. Clin. Invest., 120, 1994–2004.

22. Albrecht, A.N., Kornak, U., Boddrich, A., Suring, K., Robinson, P.N.,Stiege, A.C., Lurz, R., Stricker, S., Wanker, E.E. and Mundlos, S. (2004)A molecular pathogenesis for transcription factor associated poly-alaninetract expansions. Hum. Mol. Genet., 13, 2351–2359.

23. Nakano, K., Sakai, N., Yamazaki, Y., Watanabe, H., Yamada, N., Sezaki,K., Susami, T., Tokunaga, K., Takato, T. and Uchinuma, E. (2007) Novelmutations of the HOXD13 gene in hand and foot malformations. Int.Surg., 92, 287–295.

24. Gordon, C.T., Rodda, F.A. and Farlie, P.G. (2009) The RCAS retroviralexpression system in the study of skeletal development. Dev. Dyn., 238,797–811.

25. Hamburger, V. and Hamilton, H.L. (1992) A series of normal stages in thedevelopment of the chick embryo. 1951. Dev. Dyn., 195, 231–272.

26. Goff, D.J. and Tabin, C.J. (1997) Analysis of Hoxd-13 and Hoxd-11misexpression in chick limb buds reveals that Hox genes affect both bonecondensation and growth. Development, 124, 627–636.

27. Caronia, G., Goodman, F.R., McKeown, C.M., Scambler, P.J. andZappavigna, V. (2003) An I47L substitution in the HOXD13homeodomain causes a novel human limb malformation by producing aselective loss of function. Development, 130, 1701–1712.

28. Fernandez-Teran, M., Piedra, M.E., Kathiriya, I.S., Srivastava, D.,Rodriguez-Rey, J.C. and Ros, M.A. (2000) Role of dHAND in theanterior–posterior polarization of the limb bud: implications for the Sonichedgehog pathway. Development, 127, 2133–2142.

29. te Welscher, P., Fernandez-Teran, M., Ros, M.A. and Zeller, R. (2002)Mutual genetic antagonism involving GLI3 and dHAND prepatterns thevertebrate limb bud mesenchyme prior to SHH signaling. Genes Dev., 16,421–426.

30. Chen, Y., Knezevic, V., Ervin, V., Hutson, R., Ward, Y. and Mackem, S.(2004) Direct interaction with Hoxd proteins reverses Gli3-repressorfunction to promote digit formation downstream of Shh. Development,131, 2339–2347.

31. Litingtung, Y., Dahn, R.D., Li, Y., Fallon, J.F. and Chiang, C. (2002) Shhand Gli3 are dispensable for limb skeleton formation but regulate digitnumber and identity. Nature, 418, 979–983.

32. Biesecker, L.G. (2011) Polydactyly: how many disorders and how manygenes? 2010 update. Dev. Dyn., 240, 931–942.

33. Zeller, R., Lopez-Rios, J. and Zuniga, A. (2009) Vertebrate limb buddevelopment: moving towards integrative analysis of organogenesis. Nat.Rev. Genet., 10, 845–858.

34. Geetha-Loganathan, P., Nimmagadda, S. and Scaal, M. (2008) Wntsignaling in limb organogenesis. Organogenesis, 4, 109–115.

35. Gabellini, D., Colaluca, I.N., Vodermaier, H.C., Biamonti, G., Giacca, M.,Falaschi, A., Riva, S. and Peverali, F.A. (2003) Early mitotic degradationof the homeoprotein HOXC10 is potentially linked to cell cycleprogression. EMBO J., 22, 3715–3724.

36. Yao, T.P. (2010) The role of ubiquitin in autophagy-dependent proteinaggregate processing. Genes Cancer, 1, 779–786.

37. Qiao, L. and Zhang, J. (2009) Inhibition of lysosomal functions reducesproteasomal activity. Neurosci. Lett., 456, 15–19.

38. Kopito, R.R. (2000) Aggresomes, inclusion bodies and protein aggregation.Trends Cell Biol., 10, 524–530.

39. Bruneau, S., Johnson, K.R., Yamamoto, M., Kuroiwa, A. and Duboule, D.(2001) The mouse Hoxd13(spdh) mutation, a polyalanine expansionsimilar to human type II synpolydactyly (SPD), disrupts the function butnot the expression of other Hoxd genes. Dev. Biol., 237, 345–353.

40. Yokouchi, Y., Nakazato, S., Yamamoto, M., Goto, Y., Kameda, T., Iba,H. and Kuroiwa, A. (1995) Misexpression of Hoxa-13 induces cartilagehomeotic transformation and changes cell adhesiveness in chick limbbuds. Genes Dev.., 9, 2509–2522.

41. Williams, M.E., Lehoczky, J.A. and Innis, J.W. (2006) A group 13homeodomain is neither necessary nor sufficient for posterior prevalencein the mouse limb. Dev. Biol., 297, 493–507.

42. Towers, M., Mahood, R., Yin, Y. and Tickle, C. (2008) Integration ofgrowth and specification in chick wing digit-patterning. Nature, 452,882–886.

43. Qu, S., Niswender, K.D., Ji, Q., van der, M.R., Keeney, D., Magnuson,M.A. and Wisdom, R. (1997) Polydactyly and ectopic ZPA formation inAlx-4 mutant mice. Development, 124, 3999–4008.

44. Masuya, H., Sagai, T., Wakana, S., Moriwaki, K. and Shiroishi, T. (1995)A duplicated zone of polarizing activity in polydactylous mouse mutants.Genes Dev., 9, 1645–1653.

45. Mackem, S. and Knezevic, V. (1999) Do 5′Hoxd genes play a role ininitiating or maintaining A-P polarizing signals in the limb? Cell Tissue

Res., 296, 27–31.46. Wyngaarden, L.A., Delgado-Olguin, P., Su, I.H., Bruneau, B.G. and

Hopyan, S. (2011) Ezh2 regulates anteroposterior axis specification andproximodistal axis elongation in the developing limb. Development, 138,3759–3767.

47. Galli, A., Robay, D., Osterwalder, M., Bao, X., Benazet, J.D., Tariq, M.,Paro, R., Mackem, S. and Zeller, R. (2010) Distinct roles of Hand2 ininitiating polarity and posterior Shh expression during the onset of mouselimb bud development. PLoS Genet., 6, e1000901.

48. Capellini, T.D., Di Giacomo, G., Salsi, V., Brendolan, A., Ferretti, E.,Srivastava, D., Zappavigna, V. and Selleri, L. (2006) Pbx1/Pbx2requirement for distal limb patterning is mediated by the hierarchicalcontrol of Hox gene spatial distribution and Shh expression. Development,133, 2263–2273.

49. Goodman, F.R., Mundlos, S., Muragaki, Y., Donnai, D., Giovannucci-Uzielli, M.L., Lapi, E., Majewski, F., McGaughran, J., McKeown, C.,Reardon, W. et al. (1997) Synpolydactyly phenotypes correlate with sizeof expansions in HOXD13 polyalanine tract. Proc. Natl Acad. Sci. USA,94, 7458–7463.

50. Logan, M. and Tabin, C. (1998) Targeted gene misexpression in chicklimb buds using avian replication-competent retroviruses. Methods, 14,407–420.

51. Salsi, V., Vigano, M.A., Cocchiarella, F., Mantovani, R. and Zappavigna,V. (2008) Hoxd13 binds in vivo and regulates the expression of genesacting in key pathways for early limb and skeletal patterning. Dev. Biol.,317, 497–507.


An N-terminal G11A Mutation in HOXD13 Causes

Documents

Transcript of An N-terminal G11A Mutation in HOXD13 Causes