Human Molecular Genetics, 2001, Vol. 10, No. 26 3083-3091
© 2001 Oxford University Press
The Leri–Weill and Turner syndrome homeobox gene SHOX encodes a cell-type specific transcriptional activator
Institute of Human Genetics, University of Heidelberg, Im Neuenheimer Feld 328, 69120 Heidelberg, Germany
Received September 18, 2001; Revised and Accepted October 30, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Functional impairment of the human homeobox gene SHOX causes short stature and Madelung deformity in Leri–Weill syndrome (LWS) and has recently been implicated in additional skeletal malformations frequently observed in Turner syndrome. To enhance our understanding of the underlying mechanism of action, we have established a cell culture model consisting of four stably transfected cell lines and analysed the functional properties of the SHOX protein on a molecular level. Results show that the SHOX-encoded protein is located exclusively within the nucleus of a variety of cell lines, including U2Os, HEK293, COS7 and NIH 3T3 cells. In contrast to this cell-type independent nuclear translocation, the transactivating potential of the SHOX protein on different luciferase reporter constructs was observed only in the osteogenic cell line U2Os. Since C-terminally truncated forms of SHOX lead to LWS and idiopathic short stature, we have compared the activity of wild-type and truncated SHOX proteins. Interestingly, C-terminally truncated SHOX proteins are inactive with regards to target gene activation. These results for the first time provide an explanation of SHOX-related phenotypes on a molecular level and suggest the existence of qualitative trait loci modulating SHOX activity in a cell-type specific manner.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Three percent of all human beings suffer from a clinically relevant form of reduced body height. Thus, growth impairment represents one of the common clinical conditions adversely affecting quality of life, and children with short stature regularly seek advice in endocrinological consultations. Since final body height is determined by the proper formation and extension of the long bones, growth retardation represents a disturbed developmental process which extends from pre- to postnatal life.
Among the genetic factors involved in short stature syndromes, the homeobox gene SHOX has offered some surprises and comprehension of its developmental function still poses a major challenge to both clinicians and molecular biologists. The pseudoautosomal SHOX gene was originally reported to be responsible for some cases of idiopathic growth retardation and SHOX haploinsufficiency for the short stature phenotype in Turner syndrome (1,2). Shortly after these initial reports, the phenotypic scope of SHOX mutations was broadened by two groups, independently describing a linkage between SHOX mutations, mesomelic short stature and Madelung deformity in the Leri–Weill syndrome (LWS) (3,4). A thorough investigation of 32 patients with LWS has demonstrated that at least half can be attributed to mutations affecting the SHOX gene (5). The majority of SHOX mutations in LWS and idiopathic short stature patients are nonsense mutations leading to a premature stop and truncation of the C-terminal portion of the SHOX protein which encodes a putative transactivating domain (1,2) most likely represented by the so called OAR motif (6,7). A study by Clement-Jones et al. (8) also revealed that the variation in phenotypic severity in LWS patients does not correlate with the nature of the underlying SHOX mutation. Additional complexity was recently added to this phenotypic heterogeneity by correlating embryonic SHOX gene expression with several other skeletal dysmorphologies observed in Turner syndrome patients (8).
In summary, SHOX confronts us with a situation in which mutations affecting a single developmental gene may lead to multiple phenotypes. Consequently, understanding SHOX gene function on a molecular level is a fundamental requirement to explain the corresponding phenotypes. RT–PCR profiling detected SHOX expression at low levels in a variety of adult organs including skeletal muscle, kidney, liver, lung and different regions of the brain (1) raising the question for the skeletal specificity of the SHOX phenotype. Since SHOX encodes a member of the paired-like homeodomain proteins and is believed to act as a DNA-binding protein, we have investigated the subcellular localization of SHOX, its DNA-binding properties and its potential to modulate target gene expression. Furthermore, we compare the wild-type with a C-terminally truncated form of the SHOX protein that is encoded by the gene mutations found in some LWS and idiopathic short stature patients. Here, we report that the SHOX protein is located exclusively within the nucleus in a variety of different cell lines, binds DNA in a sequence-specific manner and acts as transcriptional activator in osteogenic cells via a C-terminal transactivation domain.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
SHOX encodes a nuclear protein
The SHOX gene encodes a homeodomain protein of the paired-like type. As putative transcription factor this gene product is expected to act within the nucleus. Therefore, nuclear translocation is one possible level of regulation for SHOX activity. We have analysed and compared the cellular localization of the wild-type and C-terminally truncated SHOX proteins (Fig. 1A) by immunofluorescence and by expressing N-terminal GFP–SHOX fusion proteins in a variety of different cell lines. Both experimental approaches yielded identical results. We observed a discrete focal nuclear localization in all cell lines investigated (Fig. 1C). The observed focal pattern was neither cell-type specific nor did we observe any differences between the wild-type and C-terminally truncated SHOX protein (STM; Fig. 1C). Double immunoflourescence experiments using different antibodies exclude a co-localization of SHOX with nucleoli, centromers and the basic transcription factor TFIIH (data not shown). This nuclear localization was not restricted to cells of human origin, but was also observed in simian (COS7; data not shown) and murine cells (NIH 3T3).
|
5 µg) were resolved on an SDS–12% polyacrylamide glycine gel. Proteins were visualized by western blot analysis (left). Equal amounts of each poly(A)+ RNA were run on a formaldehyde gel and subsequently transferred to a filter. SHOX-derived RNA was detected by hybridization of the northern blot with the SHOX cDNA as a probe (right). (C) Cellular localization of the wild-type and C-terminally truncated SHOX proteins in human and murine cells. Cells were grown on coverslips, fixed and processed for DAPI staining of nuclei (top) and immunofluorescence with FITC-conjugated anti-rabbit secondary antibodies (middle). Nuclear localization was validated by combination of DAPI and FITC (bottom) and by confocal microscopy (data not shown).Analysis of the nuclear translocation along a time course between 6 and 24 h revealed, that both wild-type and truncated proteins are transported into the nucleus with equal efficiencies. During the period of observation these proteins were not detectable within the cytosol (data not shown). These data suggest that nuclear translocation may not be the mechanism of SHOX activity regulation. Therefore, the information for the observed cell-type independent nuclear localization has to reside within the N-terminal portion of the SHOX protein.
The SHOX protein binds to DNA in a sequence-specific manner
Paired-related homeodomain proteins preferentially bind to the palindromic sequence 5'-TAAT (N)n ATTA-3' (9). It has been shown, that proteins with a Gln at position 50 of the homeodomain (Q50) bind to a P3 consensus sequence (5'-TAAT NNN ATTA-3'), named after the number of nucleotides between the palindromic boxes (10). SHOX also belongs to the Q50 subclass of paired-like homeodomain proteins and is therefore predicted to bind such a P3 element. To determine the DNA-binding sequence of the SHOX homeodomain, a selection of DNA-binding sites (SELEX) was carried out as described in Materials and Methods. The selected sequences were then tested in gel retardation assays for their affinity to the SHOX protein (data not shown). No DNA sequences other than the TAAT palindromes were isolated by this method.
Although we predominantly isolated the P3 element, we also found four other types of the TAAT palindromic sequences. In order to identify the substrate that binds to the SHOX protein with the highest affinity, a gel shift assay was carried out using increasing amounts of the SHOX protein. SHOX was able to bind all of the isolated DNAs to some extent. However, the highest affinity was seen towards P3 DNAs with electromobility shifts at a low protein concentration of 10 nM (Fig. 2A). Furthermore, the SHOX protein is able to shift the DNA probes as a homodimeric complex (Fig. 2A). Here again, mobility of P3 was shifted from a homodimeric complex at lower protein concentrations as compared to the other P elements. Dimerization of the SHOX protein was confirmed by yeast two-hybrid experiments using SHOX as both, prey and bait (Fig. 3).
|
|
In order to demonstrate the preference of SHOX for the P3 element, a competitive binding experiment was carried out using bacterially expressed and purified GST–SHOXMut protein. As shown in Figure 2B the binding of SHOX to the P3 element can be inhibited by a 50-fold excess of the same (cold) probe, but not by the addition of a 50-fold excess of cold P1/2, P2, P4 and P5. These data suggest that SHOX protein has a preferential affinity for the P3 element as compared to the other P elements.
To verify that the DNA-binding properties of eukaryotically expressed SHOX are identical to those of recombinant protein expressed in Escherichia coli, we also analysed cellular extracts from induced and uninduced T-Rex-U2Os/ST cells (Materials and Methods). A significant electrophoretic mobility shift was detected only in experiments carried out with the extract of induced cells (Fig. 2C, lane 3). To demonstrate that this gel shift was SHOX specific, a supershift experiment using rabbit anti-human SHOX-3 antibody was performed. As shown in Figure 2C (lane 5) a supershift is generated only in the presence of the extracts from the induced cells, providing clear evidence for the DNA-binding specificity of the SHOX protein.
The SHOX protein acts as a transcriptional activator
In order to investigate the effects of SHOX binding to its responsive elements (REs), we have generated a cell culture model consisting of the stably transfected cell lines U2Os-ST, U2Os-STM, HEK293-ST and HEK293-STM. In addition, several reporter plasmids containing different SHOX REs in front of an SV40 minimal promoter driving the expression of a luciferase reporter cDNA were generated. As demonstrated by western blot experiments, all stable cell lines express high amounts of either the wild-type (ST) or the C-terminally truncated SHOX protein upon induction with doxycycline (Fig. 1). Furthermore, electromobility shift analysis with nuclear extracts of all four lines showed binding of both the full-length and truncated proteins to the consensus SHOX REs (data not shown). In a first series of experiments we transiently transfected U2Os-ST cells with four reporter constructs, varying in the first nucleotide of the spacer region between the ATTA/TAAT palindrome (p3XA, p3XC, p3XG and p3XT) and induced SHOX expression with a gradient of doxycycline between 400 ng/ml and 40 µg/ml. According to these experiments, the optimal induction of SHOX activity is achieved with doxycyline concentrations of 4 µg/ml cell culture medium (data not shown) and all subsequent experiments were performed under these optimized conditions. The upregulation of all four reporter constructs observed upon induction with doxycyline (Fig. 4A) clearly define SHOX as transcriptional activator in the osteogenic cell line U2Os-ST. As shown in Figure 4A, we observed the strongest response from the p3XG reporter construct containing a G in the first position of the 3 nt spacer suggesting ATTA GGC TAAT as the preferential SHOX target sequence. Consequently, the p3XG reporter was used for subsequent transfection assays. Next, we addressed the question of cell-type specificity of the SHOX transactivating function. For this purpose, we investigated the ability of the full-length SHOX protein to stimulate reporter activity in the non-osteogenic cell line HEK293-ST. Surprisingly, we never observed a significant stimulation of reporter gene activity in this cell line (Fig. 4B). These results correlate with the low but distinct endogenous expression pattern of the SHOX gene and suggest a cell-type specific mode of SHOX activity in vitro.
|
The transcription factor activity of the SHOX protein resides in its C-terminal portion
The observation that C-terminal truncations cause short stature and Madelung deformity suggests that the transactivating activity resides within the C-terminal part of the SHOX protein. To test this hypothesis we generated the two cell lines U2Os-STM (osteogenic) and 293-STM (non-osteogenic) that upon doxycycline induction express a truncated protein resembling the situation in some of the short stature and LWS patients. As shown in Figure 4B, expression of this mutated SHOX protein does not lead to a stimulation of the p3XG reporter expression in either cell line. Therefore, the C-terminally truncated SHOX protein is inactive and can no longer act as transcriptional activator. To further narrow down the position of the transactivating domain of the SHOX protein, we have generated a deletion mutant lacking only the very C-terminal portion of SHOX. Transient transfections of these constructs together with the p3XG reporter clearly showed that deletion of the C-terminal 19 amino acids (including the OAR domain) are sufficient to diminish its potential to activate reporter gene activity (Fig. 4C). Thus, the transactivating domain of SHOX is likely to reside within this region.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
SHOX is a homeobox gene encoding a paired-like homeodomain protein (1). Although its involvement in the etiology of LWS (3,4) has been demonstrated and its expression pattern in the developing bony structures of human embryos has been correlated to skeletal features of the Turner syndrome (6), very little is known about the mechanism of SHOX function on a molecular and biochemical level. In this work we have compared the subcellular localization of the SHOX protein in different cell types, investigated its DNA-binding properties and analysed the potential of different SHOX constructs to modulate transcription of reporter constructs. Furthermore, we have compared these biochemical properties of the wild-type SHOX to a C-terminally truncated form which resembles most of the SHOX mutations observed in LWS patients.
Subcellular localization of SHOX proteins
One mechanism of regulation of transcription factors is their cytosolic retention by sequestration or post-translational modification (11). To analyse this possibility, we investigated the subcellular localization of the SHOX protein. SHOX shows a strict nuclear compartmentalization in every cell-type analysed. A similar nuclear localization has been described, for example, for the aristaless-related prd-HD protein Alx-4 (12). This nuclear localization suggests that sequestration in the cytosol as described, for example, for NF-
B (13) is not the mechanism of SHOX regulation. Interestingly, we observed an exclusively nuclear localization of SHOX also in murine NIH 3T3 fibroblasts originating from an organism that does not contain a direct SHOX orthologue (6,14). Therefore, the SHOX protein carries sufficient information for nuclear translocation even in heterologous systems.
Since it has been reported that mutations in transcription factors may affect nuclear transport and thus modulate their nuclear activity (15), we have investigated the subcellular localization of a C-terminally truncated SHOX protein representing the most frequent aberrant protein in LWS patients. The observation that this C-terminal truncation has no effect on its immunolocalization clearly argues for a nuclear localization signal within the N-terminal part of the SHOX protein or within the homeodomain itself as described for the Pax3 protein (16). Furthermore, these results clearly demonstrate that the loss of function leading to short stature and LWS phenotypes is not due to instability or inadequate nuclear transport of the mutated protein.
DNA-binding properties of the SHOX protein
Paired-type homeodomains have been shown to mediate cooperative dimeric binding to sites consisting of two palindromic TAAT sequences separated by two or three less significant nucleotides, P2 and P3 sites, respectively (9). Whereas Pax-type proteins with a Ser at position 50 of the homeodomain (S50) bind slightly better to P2 sites, other paired-type proteins with a Gln at position 50 (Q50) prefer P3 sites (TAATNNNATTA). We have used the SELEX procedure (9) to investigate the DNA-binding properties of the SHOX protein and were able to demonstrate that SHOX binds preferentially as dimer to a P3 type DNA and, with somewhat lower affinity, to other palindromic DNA types. Therefore, our results regarding the DNA-binding properties of the SHOX protein are in agreement with the reports on other paired-related proteins (10,12,17–19). The dimerization of the SHOX protein upon DNA binding was also confirmed by yeast two-hybrid analysis, demonstrating that SHOX dimerization occurs in vivo. Although further experiments are necessary to delineate the regions involved in SHOX/SHOX interaction these results suggest that dimerization might play a role in modulating SHOX transcriptional functions.
Variation in space and orientation of monomeric homeodomain targets, and the different affinity of homeodomain proteins to these sites, obviously increase the variety of different binding sites and may influence tremendously the way these regulators might exert their functions, as described for the Pit-1 protein (20). The potential for heterodimerization of different homeodomain proteins and their differential expression during development may result in different regulatory interactions important for fine tuning of gene regulation. This could explain how two homeodomain proteins with an identical homeodomain, like SHOX and SHOX2 (14), exert different functions.
SHOX transactivating activity
Since the loss of function leading to short stature and LWS cannot be explained by protein instability or impaired nuclear translocation, we have investigated the functional properties of wild-type and mutated SHOX proteins. According to our results using different inducible cell lines, SHOX has the potential to act as transcriptional activator of specific target genes. Interestingly, this transactivating activity was observed only in cells of osteogenic origin, raising the possibility that additional cell-type specific cofactors are involved in SHOX function. Such cofactors have been reported to regulate the activity of several developmentally relevant transcription factors including the trunk Hox genes (21), Fushi tarazu (22) and hepatocyte nuclear factor-1 (23). The idea of cell-type specific cofactors regulating SHOX activity is especially attractive with respect to the observed phenotypic heterogeneity observed in SHOX patients, since such cofactors could also account for the modulation of residual SHOX activity. Furthermore, such cofactors could represent excellent candidates for further growth-regulating genes that warrant future investigations.
SHOX mutations observed in idiopathic short stature and LWS patients usually either represent missense mutations within the homeodomain, probably affecting DNA binding, or nonsense mutations leading to C-terminally truncated proteins (G.A.Rappold et al., manuscript submitted for publication). Therefore, we have compared the transactivating activity of wild-type and truncated SHOX proteins in different cell lines. Interestingly, a mutated SHOX protein missing 98 amino acids at the C-terminal end was completely inactive in our transactivation assay. This result is especially noteworthy, because the truncated SHOX protein used for these assays resembles the mutation originally described in a patient with idiopathic short stature (1). The inability of this mutated protein to transactivate a reporter gene may for the first time explain the growth phenotype on a molecular level. To further narrow down the transactivation domain of the SHOX protein, we have shown that a truncation of only 19 C-terminal amino acids is sufficient to abolish SHOX transactivating function. This region contains a 14 amino acid motif common to a subset of paired-related homeodomain proteins, called aristaless or OAR domain (7), and was reported to contain transactivating potential in another homeodomain protein, orthopedia (8). Having assigned the SHOX transactivating activity to the most C-terminal part of the protein is interesting for two reasons. First, from this result we can conclude that any nonsense mutation leading to a SHOX truncation upstream of the OAR domain would lead to a short stature or LWS phenotype. Secondly, the SHOX gene is expressed in two isoforms, SHOXa and SHOXb, with the latter lacking the entire C-terminal portion including the OAR domain (1). From our results, SHOXb is predicted to be inactive as a transcriptional activator. Consequently, SHOXa/SHOXb heterodimers and SHOXb homodimers are predicted to have transcriptional activation activities different from those of SHOXa homodimers. Therefore, SHOXb represents an excellent candidate for a modulator of SHOXa activity. The idea of such modulating activities of the SHOXb isoform is in agreement with reports on other developmentally important transcription factors (24). Additional attractive candidates for SHOX modulators might be encoded by the SHOX2 gene. Given an identical homoedomain of SHOX and SHOX2 (14), these proteins may compete for identical binding sites or even directly interact with each other. However, such modulating functions can only be expected in regions of overlapping SHOX and SHOX2 expression (8).
In summary, our results not only provide new and valuable guidelines for molecular SHOX diagnostics, but also for the first time provide insights into the potential mechanisms underlying the developmental defects in patients with C-terminal SHOX truncations. Furthermore, the observation that SHOX activity might be regulated in a cell-type specific manner suggests the existence of modulating qualitative trait loci that warrant further investigations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Binding site selection
The selection of DNA-binding sites for the SHOX protein was performed as described by Wilson et al. (9). The degenerate template oligonucleotide used contained a core of 20 random nucleotides flanked on each site by specific primer binding sequences: GCG GTG GCG GCC GCT CTA GAA CTA G-[N20]-GCT TAT CGA TAC CGT CGA CCT CGA C. After the fifth round of selection, the PCR products representing an enriched library of SHOX-specific targets were cloned by using the TOPOTM cloning kit (Invitrogen, CH Groningen, The Netherlands).
Antibody generation
Polyclonal rabbit anti-human SHOX-3 antibodies were generated against keyhole limpet hemocyanin (KLH)- or BSA-conjugated peptides specific for the SHOX protein (NH2-KEKREDVKSEDEDGQ-COOH). Sera were collected at 10 weeks after the second boost and SHOX proteins expressed in bacterial and eukaryotic systems were analysed by western blot analysis. SHOX-3 antibodies were affinity purified using the original peptide bound to Sepharose 4B columns (Pineda Antikörper Service, Berlin, Germany).
Electromobility shift
DNA used for mobility shift assays was generated by annealing the two complementary oligonucleotides that generated a 5' overhang on each side to permit labelling with Klenow polymerase. The sequences of the top strand of the probes are as follows (TAAT repeats in bold):
P1/2: GCGGCCGCTCTAGAACTAGCTGAGTAATTGAGCACTGTAGCTTATCGATACCGTCGACC;
P2: GCGGCCGCTCTAGAACTAGCTGAGTAATTGATTACTGTAGCTTATCGATACCGTCGACC;
P3: GCGGCCGCTCTAGAACTAGCTGAGTAATTGAATTACTGTGCTTATCGATACCGTCGACC;
P4: GCGGCCGCTCTAGAACTAGTGAGTAATTGAGATTACTGTGCTTATCGATACCGTCGACC;
P5: GCGGCCGCTCTAGAACTAGTGAGTAATTGAGAATTACTGGCTTATCGATACCGTCGACC.
Gel shift reactions contained 15 mM HEPES pH 7.5, 60 mM NaCl, 1 mM EDTA, 0.5 mM DTT, 0.05% NP-40, 7.5% glycerol, 4 mM spermine, 4 mM spermidine, 0.25 mg/ml BSA, 0.5 µg poly(dI-dC), proteases inhibitor cocktail (Roche Diagnostics GmbH, Mannheim, Germany) and the appropiate 32P-labelled probe. Reaction mixtures were preincubated for 5 min at room temperature. After addition of the 32P-labelled probe, the mixture was incubated at room temperature for 20 min before separation on a 5% polyacrylamide gel that contained 0.25x TBE at 12 V/cm (10°C) for 2.5 h. Gels were then fixed in 10% acetic acid, 10% ethanol, dried and exposed overnight at –80°C.
Assays were performed using the GST–SHOXMut (amino acid position 1–194) fusion protein. In competition experiments, a 40-fold excess of the unlabelled competitor DNA was added during preincubation. In gel shift experiments with cellular extracts of SHOX inducible U2Os cells, 5 µg of cellular extract were used. In supershift experiments, the cellular extracts were preincubated with 1 µg of anti-human SHOX-3-specific antibody for 15 min at room temperature before adding the probe.
Western blot analysis
Western blot experiments were performed according to established protocols. Briefly, cell lysates were denatured by boiling in sample buffer (45 mM Tris–HCl pH 6.8, 10% glycerol, 3% ß-mercaptoethanol, 1% SDS, 0.01% bromophenol blue), separated by 12% SDS–PAGE and electroblotted on PVDF membranes (HybondP; Amersham Pharmacia Biotech, Uppsala, Sweden). The membranes were pre-incubated for 1 h at room temperature in PBS containing 5% skim milk. Blots were then incubated overnight at 4°C with the rabbit anti-SHOX-3 antibody at a dilution of 1:3000. After membrane washing, the peroxidase-conjugated secondary goat-anti-rabbit antibody (at a dilution of 1:15 000) was added. Proteins were visualized with the ECL detection kit (Amersham Pharmacia Biotech).
Generation of plasmids
SHOX cDNA open reading frame (ORF) was constructed by generating overlapping PCR fragments from a human bone marrow fibroblast cDNA library and from the cosmid LLN0YCO3'M'34F5 (1) with the following primers: SHOXTag forward (AAG CTT AGT CGA CGG ATC CAC CAT GGA AGA GCT CAC GGC TTT TGT ATC C)/SHOX(RI) reverse (CGC GGT GCC GAA TTC TTT CAG); SHOX(RI) forward (position 334–355 of ORF) (ACT GAA AGA ATT CGG CAC CGC G)/SHOX(PstI) reverse (position 745–719 of ORF) (CAG CTG CAG CTG AGC CTG GAC CTG TTG). The 3' end of SHOX was generated by PstI–SmaI restriction from cosmid LLN0YCO3'34F5' (254 bp fragment). The three fragments were ligated in pBluescriptSK vector (HindIII–SmaI) and transformed in XL-1 Blue competent cells generating the clone pSHOX/SK.
To generate the different full-length SHOX expression plasmids the BamHI fragment from pSHOX/SK was cloned in the BamHI site of the following plasmids: pGEX-2T (Amersham Pharmacia Biotech), pEGFP-C1 (BD Clontech, Heidelberg, Germany) and pcDNA4/TO (Invitrogen). The pCI/SHOX plasmid was constructed by inserting the SalI–NotI SHOX fragment from pSHOX/SK into the pCI vector (Promega GmbH, Mannheim, Germany).
Because of two internal EcoRI restriction sites within the SHOX ORF, the C-terminal pCI/SHOX
RI truncation clone was constructed ligating the PstI–EcoRI fragment to the SalI–PstI fragment from pSHOX/SK into the SalI–EcoRI site of pBluescript SK. This C-terminal deletion clone was digested with SalI–NotI and cloned into the respective sites of the pCI vector (Promega GmbH).
The 3' deletion clones lacking the exon VIa were generated by PCR using the SHOXTag forward primer and the SHOX mutant primer (CTA GTA CTA GTC AGC AGG CGT CTA GGT GGT TGG CTG T). The PCR product encoding a protein from position 1 to 194 was cloned in the pGEX-2T (Amersham Pharmacia Biotech) and in the pcDNA4/TO vectors (Invitrogen). Plasmid DNA for transfection experiments were isolated using plasmid isolation Midi kits (Qiagen GmbH, Hilden, Germany).
Two-hybrid system
For the yeast two-hybrid system the entire coding sequence of SHOX was fused either with the Gal4 DNA-binding domain in the pGBT9 vector or with the Gal4 transcription activator domain in the pGAD424 vector (BD Clontech). For a correct frame, a fragment corresponding to nucleotides 1–165 of the SHOX coding sequence was first PCR amplified from pSHOX/SK with the following forward and reverse oligonucleotides: 5'-AAA GAA TTC ATG GAA GAG CTC ACG GCT TTG TA-3' and 5'-TGG AAT CCG ACG TCC CCA GCT C-3' and cloned into EcoRI- and SmaI-digested vectors resulting in pGBT9–SHOX 1–46 and pGAD424–SHOX 1–46. The rest of the SHOX coding sequence was first isolated as a SmaI fragment from pSHOX/SK and then transferred into SmaI-digested and CIAP-treated pGBT9–SHOX 1–46 and pGAD424–SHOX 1–46 vectors. Co-transformations of the plasmids into Saccharomyces cerevisiae SFY526, and liquid assay for ß-galactosidase activity was performed as described in the Matchmaker Two-Hybrid System manual book (BD Clontech).
Generation of stable cell lines
The human T-Rex U2Os osteosarcoma cell line (R712-07; Invitrogen) and the human T-Rex HEK293 embryonic kidney cell line (R710-07; Invitrogen) were used for the construction of a cell line with a tetracycline-regulated expression of the SHOX gene. Therefore, two different constructs of SHOX [wild-type (ST) and the mutant SHOX (STM)], were cloned into the pcDNA4/TO vector and transfected into both lines using FUGENE reagent (Roche Diagnostics GmbH, Mannheim, Germany) at a ratio of 1:6.
For selection and maintenance of the stably transfected clones, T-Rex HEK293 cells were grown in DMEM supplemented with blasticidin (10 µg/ml) and zeocin (100 µg/ml), and T-Rex U2Os cells were cultured on DMEM supplemented with hygromycin (10 µg/ml) and zeocin (400 µg/ml). Induction of SHOX expression was achieved by addition of doxycycline (4 µg/ml) to the culture medium for 24–48 h.
Transfection and activity assays
SHOX activity assays were performed in T-Rex U2Os, T-Rex U2Os-ST, U2Os-STM, HEK293-ST and HEK293-STM cell lines with several reporter plasmids containing the following SHOX consensus-binding sequences in front of a SV40 minimal promoter that drives a luciferase reporter gene (pGL3-Promoter Vector; Promega GmbH):
p3XA, ATGGCTGAGTAATAGTATTACTGTC;
p3X, ATGGCTGAGTAATCGGATTACTGTC;
p3XG, ATGGCTGAGTAATGGCATTACTGTC;
p3XT, ATGGCTGAGTAATTGATATTACTGTC.
Six hours prior to transfection, cells were seeded into 24-well dishes at a density of 6–8 x 104 cells per well. Medium was changed just prior to transfection. Transfections were carried out with 100 ng/DNA construct/well using FUGENE reagent (Roche Diagnostics GmbH) at a DNA:FUGENE ratio of 1:6. All cells were co-transfected with 100 ng of a pCMVß plasmid (BD Clontech). In stable cell lines, SHOX expression was induced by the addition of doxycycline to a final concentration of 4 µg/ml culture medium, 12 h after transfection. After 48 h of SHOX induction, cells were lysed with 200 µl 1x reporter lysis buffer (Promega GmbH) per well. Twenty microlitres of these cell lysates were used to determine enzyme activities. ß-Galactosidase activity was measured by addition of 200 µl Z-buffer (100 mM sodium phosphate buffer pH 7.3, 50 mM ß-mercaptoethanol, 1 mM MgCl2, 0.7 mg/ml ONPG). The reaction mixture was incubated at 37°C for 15 min and the resulting o-nitrophenol was measured at 405 nm by using a spectrophotometer (Lucy 2; anthos Mikrosysteme GmbH, Krefeld, Germany). Luciferase activity was measured in an automated luminometer (anthos Mikrosysteme GmbH) with 50 µl cellular extract and 100 µl of luciferin (Promega GmbH). All assays were performed at least in duplicate and all luciferase measurements were corrected for transfection efficiency with the corresponding ß-galactosidase activities. Transcriptional SHOX activity was calculated from at least two independent experiments as fold induction compared to uninduced cells.
Immunofluorescence
For immunofluorescence analysis cells were plated on microscopic slides at a density of 104 cells/cm2. Twenty-four hours after induction or transfection, cells were fixed with acetone/methanol (1:1) for 2 min and washed twice with cold PBS. After blocking with PBS containing 10% skim milk for 30 min, cells were incubated for 1 h at room temperature with the rabbit anti-SHOX-3 antibody at a dilution of 1:300. For co-localization experiments the following antibodies were used: nucleolin (C23) (Santa Cruz Biotechnology, Santa Cruz, CA), HP1 against human centromers (kindly provided by Dr Wendy Bickmore, Edinburgh, UK) and TFIIH (kindly provided by Dr Sebastian Iben, Heidelberg, Germany). After three washing steps with PBS, cells were incubated with a fluorescein isothiocyanate (FITC)-conjugated goat-anti-rabbit antibody (Sigma, Deisenhofen, Germany) at a 1:10 000 dilution. For double immunoflourenscence experiments Cy-3-conjugated anti-mouse antibodies or FITC-conjugated anti-human antibodies were used. Nuclei were counterstained with DAPI dye and cells were examined by fluorescent microscopy (Leica Mikroskopie und Systeme, Wetzlar, Germany) using FITC and Cy-3 cut-off filters.
|
ACKNOWLEDGEMENTS |
|---|
This study was funded by Eli Lilly and Company and the Deutsche Forschungsgesellschaft. We thank Wendy Bickmore and Sebastian Iben for antibodies and Gordon Cutler and Johannes Janssen for comments.
|
FOOTNOTES |
|---|
+ To whom correspondence should be addressed. Tel: +49 6221 565059; Fax: +49 6221 565332; Email: gudrun_rappold@med.uni-heidelberg.de The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Rao, E., Weiss, B., Fukami, M., Rump, A., Niesler, B., Mertz, A., Muroya, K., Binder, G., Kirsch, S., Winkelmann, M. et al. (1997) Pseudoautosomal deletions encompassing a novel homeobox gene cause growth failure in idiopathic short stature and Turner syndrome. Nat. Genet., 16, 54–63.[ISI][Medline]
2 Ellison, J.W., Wardak, Z., Young, M.F., Gehron Robey, P., Laig-Webster, M. and Chiong, W. (1997) PHOG, a candidate gene for involvement in the short stature of Turner syndrome. Hum. Mol. Genet., 6, 1341–1347.
3 Belin, V., Cusin, V., Viot, G., Girlich, D., Toutain, A., Moncla, A., Vekemans, M., Le Merrer, M., Munnich, A. and Cormier-Daire, V. (1998) SHOX mutations in dyschondrosteosis (Leri–Weill syndrome). Nat. Genet., 19, 67–69.[ISI][Medline]
4 Shears, D.J., Vassal, H.J., Goodman, F.R., Palmer, R.W., Reardon, W., Superti-Furga, A., Scambler, P.J. and Winter, R.M. (1998) Mutation and deletion of the pseudoautosomal gene SHOX cause Leri–Weill dyschondrosteosis. Nat. Genet., 19, 70–73.[ISI][Medline]
5 Schiller, S., Spranger, S., Schechinger, B., Fukami, M., Merker, S., Drop, S.L., Troger, J., Knoblauch, H., Kunze, J., Seidel, J. and Rappold, G.A. (2000) Phenotypic variation and genetic heterogeneity in Leri–Weill syndrome. Eur. J. Hum. Genet., 8, 54–62.[ISI][Medline]
6 Meijlink, F., Beverdam, A., Brouwer, A., Oosterveen, T.C. and Berge, D.T. (1999) Vertebrate aristaless-related genes. Int. J. Dev. Biol., 43, 651–663.[ISI][Medline]
7 Simeone, A., D’Apice, M.R., Nigro, V., Casanova, J., Graziani, F., Acampora, D. and Avantaggiato, V. (1994) Orthopedia, a novel homeobox-containing gene expressed in the developing CNS of both mouse and Drosophila. Neuron, 13, 83–101.[ISI][Medline]
8 Clement-Jones, M., Schiller, S., Rao, E., Blaschke, R.J., Zuniga, A., Zeller, R., Robson, S.C., Binder, G., Glass, I., Strachan, T., Lindsay, S. and Rappold, G.A. (2000) The short stature homeobox gene SHOX is involved in skeletal abnormalities in Turner syndrome. Hum. Mol. Genet., 9, 695–702.
9 Wilson, D., Sheng, G., Lecuit, T., Dostatni, N. and Desplan, C. (1993) Cooperative dimerization of paired class homeodomains on DNA. Genes Dev., 7, 2120–2134.
10 Wilson, D.S., Guenther, B., Desplan, C. and Kuriyan, J. (1995) High resolution crystal structure of a paired (Pax) class cooperative homeodomain dimer on DNA. Cell, 82, 709–719.[ISI][Medline]
11 Vandromme, M., Gauthier-Rouviere, C., Lamb, N. and Fernandez, A. (1996) Regulation of transcription factor localization: fine-tuning of gene expression. Trends Biochem. Sci., 21, 59–64.[ISI][Medline]
12 Hudson, R., Tanguchi-Sidle, A., Boras, K., Wiggan, O. and Hamel, P.A. (1998) Alx-4, a transcriptional activator whose expression is restricted to sites of epithelial-mesenchymal interactions. Dev. Dyn., 213, 159–169.[ISI][Medline]
13 Mercurio, F. and Manning, A.M. (1999) Multiple signals converging on NF-B. Curr. Opin. Cell Biol., 2, 226–232.
14 Blaschke, R.J., Monaghan, A.P., Schiller, S., Schechinger, B., Rao, E., Padilla-Nash, H., Ried, T. and Rappold, G.A. (1998) SHOT, a SHOX-related homeobox gene, is implicated in craniofacial, brain, heart and limb development. Proc. Natl Acad. Sci. USA, 95, 2406–2411.
15 Takebayashi, K., Chida, K., Tsukamoto, I., Morii, E., Munakata, H., Arnheiter, H., Kuroki, T., Kitamura, Y. and Nomura, S. (1996) The recessive phenotype displayed by a dominant negative microphthalmia-associated transcription factor mutant is a result of impaired nucleation potential. Mol. Cell. Biol., 3, 1203–1211.
16 Dingwall, C. and Laskey, R.A. (1991) Nuclear targeting sequences—a consensus? Trends Biochem. Sci., 12, 478–481.
17 Cai, R.L. (1998) Human CART1, a paired-class homeodomain protein, activates transcription through palindromic binding sites. Biochem. Biophys. Res. Commun., 250, 305–311.[ISI][Medline]
18 Qu, S., Tucker, S.C., Zhao, Q., deCrombrugghe, B. and Wisdom, R (1999) Physical and genetic interactions between Alx4 and Cart1. Development, 126, 359–369.[Abstract]
19 Sheng, G., Thouvenot, E., Schmucker, D., Wilson, D.S. and Desplan, C. (1997) Direct regulation of rhodopsin 1 by Pax-6/eyeless in Drosophila: evidence for a conserved function in photoreceptors. Genes Dev., 11, 1122–1131.
20 Scully, K.M., Jacobson, E.M., Jepsen, K., Lunyak, V., Viadiu, H., Carriere, C., Rose, D.W., Hooshmand, F., Aggarwal, A.K. and Rosenfeld, M.G. (2000) Allosteric effects of Pit-1 DNA sites on long-term repression in cell type specification. Science, 290, 1127–1131.
21 Ryoo, H.D. and Mann, R.S. (1999) The control of trunk Hox specificity and activity by Extradenticle. Genes Dev., 13, 1704–1716.
22 Yu, Y., Li, W., Su, K., Yussa, M., Han, W., Perrimon, N. and Pick, L. (1997) The nuclear hormone receptor Ftz-F1 is a cofactor for the Drosophila homeodomain protein Ftz. Nature, 385, 552–555.[Medline]
23 Hansen, L.P. and Crabtree, G.R. (1993) Regulation of the HNF-1 homeodomain proteins by DCoH. Curr. Opin. Genet. Dev., 3, 246–253.[Medline]
24 Lopez, A.J. (1995) Developmental role of transcription factor isoforms generated by alternative splicing. Dev. Biol., 172, 396–411.[ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Marchini, B. Hacker, T. Marttila, V. Hesse, J. Emons, B. Weiss, M. Karperien, and G. Rappold BNP is a transcriptional target of the short stature homeobox gene SHOX Hum. Mol. Genet., December 15, 2007; 16(24): 3081 - 3087. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-Y. Chu and A. Ohtoshi Cloning and Functional Analysis of Hypothalamic Homeobox Gene Bsx1a and Its Isoform, Bsx1b Mol. Cell. Biol., May 15, 2007; 27(10): 3743 - 3749. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Rappold, W. F Blum, E. P Shavrikova, B. J Crowe, R. Roeth, C. A Quigley, J. L Ross, and B. Niesler Genotypes and phenotypes in children with short stature: clinical indicators of SHOX haploinsufficiency J. Med. Genet., May 1, 2007; 44(5): 306 - 313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Marchini, T. Marttila, A. Winter, S. Caldeira, I. Malanchi, R. J. Blaschke, B. Hacker, E. Rao, M. Karperien, J. M. Wit, et al. The Short Stature Homeodomain Protein SHOX Induces Cellular Growth Arrest and Apoptosis and Is Expressed in Human Growth Plate Chondrocytes J. Biol. Chem., August 27, 2004; 279(35): 37103 - 37114. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. F. Munns, H. R. Haase, L. M. Crowther, M. T. Hayes, R. Blaschke, G. Rappold, I. A. Glass, and J. A. Batch Expression of SHOX in Human Fetal and Childhood Growth Plate J. Clin. Endocrinol. Metab., August 1, 2004; 89(8): 4130 - 4135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Sabherwal, K. U. Schneider, R. J. Blaschke, A. Marchini, and G. Rappold Impairment of SHOX nuclear localization as a cause for Leri-Weill syndrome J. Cell Sci., June 15, 2004; 117(14): 3041 - 3048. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N Sabherwal, R J Blaschke, A Marchini, D Heine-Suner, J Rosell, J Ferragut, W F Blum, and G Rappold A novel point mutation A170P in the SHOX gene defines impaired nuclear translocation as a molecular cause for Leri-Weill dyschondrosteosis and Langer dysplasia J. Med. Genet., June 1, 2004; 41(6): e83 - e83. [Full Text] [PDF]
|
|
|
|
|
|
R. J. Blaschke, C. Topfer, A. Marchini, H. Steinbeisser, J. W. G. Janssen, and G. A. Rappold Transcriptional and Translational Regulation of the Leri-Weill and Turner Syndrome Homeobox Gene SHOX J. Biol. Chem., November 28, 2003; 278(48): 47820 - 47826. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Binder, M. B. Ranke, and D. D. Martin Auxology Is a Valuable Instrument for the Clinical Diagnosis of SHOX Haploinsufficiency in School-Age Children with Unexplained Short Stature J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4891 - 4896. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L Stuppia, G Calabrese, V Gatta, S Pintor, E Morizio, D Fantasia, P Guanciali Franchi, M M Rinaldi, G Scarano, D Concolino, et al. SHOX mutations detected by FISH and direct sequencing in patients with short stature J. Med. Genet., February 1, 2003; 40(2): e11 - 11. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||