SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Contact NPG
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
29 July 1999, Volume 18, Number 30, Pages 4348-4356
Table of contents    Previous  Article  Next   [PDF]
PAX3 and PAX7 exhibit conserved cis-acting transcription repression domains and utilize a common gain of function mechanism in alveolar rhabdomyosarcoma
J L Bennicelli1, S Advani1, B W Schäfer2 and F G Barr1

1Department of Pathology and Laboratory Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, PA 19104, USA

2Department of Pediatrics, University of Zurich, Zurich, Switzerland

Correspondence to: J L Bennicelli, Department of Pathology and Laboratory Medicine, University of Pennsylvania School of Medicine, Room 506 Stellar Chance Laboratories, 422 Curie Blvd, Philadelphia, PA 19104, USA


The t(2;13) and t(1;13) translocations of alveolar rhabdomyosarcoma (ARMS) result in chimeric PAX3-FKHR or PAX7-FKHR transcription factors, respect-ively. In each chimera, a PAX DNA-binding domain is fused to the C-terminal FKHR transactivation domain. Previously we demonstrated that PAX3-FKHR is more potent than PAX3 because the FKHR transactivation domain is resistant to repression mediated by the PAX3 N-terminus. Here we test the hypothesis that the cis-acting repression domain is a conserved feature of PAX3 and PAX7 and that PAX7-FKHR gains function similarly. Using PAX-specific DNA-binding sites, we found that PAX7 was virtually inactive, while PAX7-FKHR exhibited activity 600-fold above background and was comparable to PAX3-FKHR. Deletion analysis showed that the transactivation domains of PAX7 and PAX7-FKHR are each more potent than either full-length protein, and resistance to cis-repression is responsible for the PAX7-FKHR gain of function. Further deletion mapping and domain swapping experiments with PAX3 and PAX7 showed that their transactivation domains exhibit subtle dose-dependent differences in potency, likely due to regions of structural divergence; while their repression domains are structurally and functionally conserved. Thus, the data support the hypothesis and demonstrate that PAX3 and PAX7 utilize a common gain of function mechanism in ARMS.


PAX3; PAX7; FKHR; transcription factor; chromosomal translocation; fusion protein


ARMS, alveolar rhabdomyosarcoma; CAT, chloramphenicol acetyltransferase; PCR, polymerase chain reaction


Chromosomal translocations are a characteristic feature of several solid tumors (Sorensen and Triche, 1996) and may contribute to tumorigenesis through altered regulation of gene expression or through creation of oncogenic, chimeric gene products. For example, the t(3;8) translocation of pleiomorphic adenoma results in promoter swapping between the developmentally regulated PLAG1 zinc finger gene on chromosome 8 and the constitutively expressed CTNNB1 gene for beta-catenin on chromosome 3 (Kas et al., 1997). The translocation leads to activation of PLAG1 and reduced expression of CTNNB1. Examples of chimeric genes encoding oncogenic fusion proteins in solid malignancies include EWS-FLI1 resulting from the t(11;22) of Ewing's sarcoma (May et al., 1993) and TLS-CHOP resulting from the t(12;16) of myxoid liposarcoma (Zinszner et al., 1994).

In a single type of malignancy, translocations sometimes involve a common gene fused to multiple different partners. Alternatively, a single gene may be fused to different partners in distinct malignancies. In Ewing's sarcoma the EWS gene may be fused to either FLI1, ETV1, or ERG; while in myxoid liposarcoma CHOP may be fused to either TLS or EWS (Ladanyi, 1995). Additionally, EWS may be fused with CHN in myxoid chondrosarcoma (Clark et al., 1996), with WT1 in desmoplastic small round cell tumor (Kim et al., 1998), or with ATF1 in malignant melanoma of soft parts (Fujimura et al., 1996).

Alveolar rhabdomyosarcoma (ARMS) is a pediatric soft tissue tumor that exhibits some of the aforementioned characteristics. In ARMS, the hallmark translocation t(2;13)(q35;q14) and the variant t(1:13)(p36;q14) result in formation of PAX3-FKHR and PAX7-FKHR chimeric transcription factors, respectively (Barr, 1997). In each chimera, the N-terminal DNA binding domain of a paired box (PAX) protein is fused to the C-terminal transactivation domain of the winged helix protein FKHR. Previously we demonstrated that PAX3-FKHR is a more potent transactivator than wild-type PAX3 (Fredericks et al., 1995). Two mechanisms were implicated in the increased activity of PAX3-FKHR including loss of cis-repression of the C-terminal transactivation domain (Bennicelli et al., 1996) and over-expression of the fusion protein due to altered transcriptional regulation (Davis and Barr, 1997). Additionally, PAX3-FKHR was found to function as an oncoprotein in vitro, causing transformation of chick embryo fibroblasts (Scheidler et al., 1996).

The PAX genes encode a family of nine developmentally regulated transcription factors characterized by the presence of a paired box DNA-binding domain similar to that encoded in the Drosophila gene paired (Tremblay and Gruss, 1994). PAX family members may also contain a partial or complete homeodomain and/or a conserved octapeptide motif. PAX3 and PAX7 comprise a subfamily based on a high degree of sequence homology and similar genomic organization. Both proteins exhibit similar DNA-binding characteristics in vitro (Schafer et al., 1994), and contain a complete homeodomain, a conserved octapeptide motif, and an extended region located N-terminal to the paired box. There is some sequence divergence in the C-terminal transactivation domains, and a unique C-terminal extension was identified in a recently cloned full-length PAX7 cDNA (Vorobyov et al., 1997).

The importance of paired box-containing genes in embryogenesis is evident in various phenotypes that result from mutations in PAX genes (Tremblay and Gruss, 1994). Inactivation of murine Pax-1, Pax-3, or Pax-6 leads to the undulated, splotch, or small eye phenotype; while inactivation of human PAX3 or PAX6 leads to Waardenburg Syndrome or aniridia, respectively. While these phenotypes result from loss of function mutations, there are two reported gain of function mutations of PAX genes associated with human malignancies. First, PAX5 expression is deregulated in lymphoplasmacytoid lymphoma by a t(9;14) translocation that juxtaposes the PAX5 gene with the immunoglobulin heavy chain locus (Busslinger et al., 1996). Secondly, we have shown that the t(2;13) translocation of ARMS results in the generation of the chimeric PAX3-FKHR protein that is a more potent transcriptional activator than wild-type PAX3 (Bennicelli et al., 1996). Here we explore the hypothesis that PAX3 and PAX7 exhibit conserved transcription repression domains and that the t(1;13) translocation of ARMS results in a gain of function mutation of PAX7.


PAX7-FKHR exhibits a gain of function relative to PAX7

To determine whether the PAX7-FKHR fusion protein of ARMS functions as a transactivator that is more potent than wild-type PAX7, we used transient cotransfection assays in NIH3T3 cells and measured the ability of each protein to activate transcription of a reporter gene located downstream of 6´PRS-9 DNA-binding sites. As shown in Figure 1, PAX7 was virtually inactive, while PAX7-FKHR showed a dose dependent increase in activity that peaked at 0.1 mug of test plasmid at a level 600-fold greater than background. Thus, the data demonstrate a gain of function for the chimeric PAX7-FKHR transcription factor relative to wild-type PAX7.

Comparison of PAX7-FKHR with PAX3-FKHR showed that the former was only slightly less active in both NIH3T3 (Figure 2a) and COP-8 cells (Figure 2b). At 0.1 mug of test plasmid, the difference in activity was less than twofold, and was most likely due to slight differences in expression levels. Immunoprecipitation of the fusion proteins from transiently transfected COP-8 cells demonstrated expression of the 97 kd PAX3-FKHR protein at a level of 1.5-fold higher than the 94 kd PAX7-FKHR protein (Figure 2c). Therefore PAX3-FKHR and PAX7-FKHR are similarly potent transactivators.

GAL4 fusion proteins mimic the activity of native proteins

To perform a structure-function analysis of PAX7, we used deletion mutants expressed as fusion proteins with the heterologous GAL4 DNA-binding domain. Figure 3 validates this approach by demonstrating that the relative activities of full-length GAL4 fusion proteins are similar to that of full-length native proteins. When tested with the reporter GAL45/E1b CAT, results obtained with GAL4-PAX7 and GAL4-PAX7-FKHR in both SJRH28 ARMS cells (Figure 3a) and NIH3T3 cells (Figure 3b) were similar to results obtained with the corresponding native proteins. Furthermore, the relative activities of GAL4-PAX3 and GAL4-PAX3-FKHR in NIH3T3 cells (Figure 3b) were also similar to that of the native proteins (Bennicelli et al., 1996).

Expression levels of GAL4 fusion proteins were determined in transiently transfected COP-8 cells by immunoprecipitation with GAL4-specific antiserum (data not shown). Simultaneous CAT assays were performed in all immunoprecipitation experiments and demonstrated that the relative transactivation potencies of the test proteins in COP-8 cells were similar to those in NIH3T3 cells. In COP-8 cells, GAL4-PAX3 and GAL4-PAX7 were overexpressed 3.3 and 1.7-fold, respectively, relative to the corresponding GAL4-PAX3-FKHR or GAL-PAX7-FKHR chimeras. These differences in expression levels result in underestimation of the magnitude of the gain of function for each chimeric transcription factor, but do not alter interpretation of the data. For GAL4-PAX3-FKHR, expression was 1.9-fold that of GAL4-PAX7-FKHR and might contribute to the observed difference in activity that was fourfold at 0.1 mug of test plasmid (Figure 3b). Thus the GAL4 fusion proteins mimicked the activity of native proteins and were expressed at roughly similar levels, validating their use for deletion analysis of wild-type PAX7.

The PAX3 and PAX7 C-terminal transactivation domains exhibit subtle functional differences

PAX3 and PAX7 have highly homologous N-terminal DNA-binding domains but exhibit sequence differences in their C-termini. Furthermore, when the two sequences are aligned, PAX7 diverges 5 amino acids from the end of PAX3 and extends an additional 47 amino acids. Since PAX7 was essentially inactive in our assay system, while PAX3 exhibited weak activity, we compared their respective C-terminal transactivation domains to determine if they function differently. As shown in Figure 4a, the transactivation domains of PAX3 and PAX7 were comparably potent, but exhibited distinct transactivation profiles. The relative potencies of the PAX3 and PAX7 C-termini varied with the amount of transfected expression plasmid. At 0.1 mug of plasmid they were similar, but at 1 mug of test construct the activity of GAL4-cterPAX7 decreased and was about 11-fold less than GAL4-cterPAX3. These results may not be explained by differences in expression levels, since GAL4-cterPAX7 was expressed at 1.7-fold the level of GAL4-cterPAX3 in COP-8 cells (data not shown). Thus the PAX3 and PAX7 transactivation domains exhibit dose-dependent differences in potency.

Deletion analysis of the PAX7 C-terminus (amino acids 275 - 520) was performed to map the regions important for transcriptional function. As shown (Figure 4b), deletion of the region encoded by exon 8 (amino acids 386 - 520) resulted in a 77% decrease in activity at 0.1 mug test plasmid; while this region alone exhibited 44% of full activity. In a previous study of the PAX3 C-terminus, deletion of the region encoded by PAX3 exon 8 resulted in a 96% decrease in activity at 0.5 mug test plasmid; while this region alone exhibited only 11% of full activity (Bennicelli et al., 1995). These results demonstrate that for both PAX3 and PAX7, the region encoded by exon 8 is required, but not sufficient for full transactivation. When compared to their respective C-terminal transactivation domains, however, the fragment of PAX7 encoded by exon 8 is more active than the corresponding fragment of PAX3.

To determine if the extended C-terminal region of PAX7 is involved in the observed differences in PAX3 and PAX7 activity, we performed a deletion analysis of this region. As shown, deletion of amino acids 464 - 520 did not appreciably alter the activity of full-length GAL4-PAX7 (Figures 3b and 6a) nor GAL4-cterPAX7 (Figures 4b and 6a). It did, however, contribute to the ability of the fragment encoded by exon 8 to activate transcription (Figure 4b). Whereas the exon 8 fragment exhibited 44% activity, the shorter fragment 386 - 563 exhibited only 24% activity. Thus the C-terminal extension of PAX7 contributes to its transactivation function.

PAX7 contains an N-terminal cis-acting repression domain

PAX3-FKHR gains function relative to PAX3 via resistance to cis-repression (3). To test the hypothesis that PAX3 and PAX7 utilize similar gain of function mechanisms in ARMS, we looked for evidence of transcriptional repression mediated by the PAX7 N-terminus. GAL4 fusions of the full-length proteins or their respective C-terminal transactivation domains were tested in NIH3T3 cells. Expression levels of the proteins were measured in COP-8 cells and found to differ by 2.3-fold or less (data not shown). As shown in Figure 5a, the C-termini of PAX7 and PAX7-FKHR each contained a potent transactivation domain that was more potent than either full-length protein. At 0.1 mug test plasmid (Figure 5b), the activity of GAL4-PAX7-FKHR was 17-fold less than its C-terminus; whereas GAL4-PAX7 was inactive. These results demonstrate that PAX7-FKHR gains function through a mechanism similar to PAX3-FKHR. The N-terminus of PAX7 contains a repression domain that abrogates transactivation by the C-terminus of PAX7, but only partially inhibits transactivation by the C-terminus of PAX7-FKHR.

The PAX3 and PAX7 repression domains exhibit conserved structure and function

To test the hypothesis that the transcription repression domain is a conserved feature in PAX3 and PAX7, we directly compared N-terminal deletion mutants of each protein fused to the heterologous GAL4 DNA-binding domain. The ability of each construct to be expressed was confirmed by in vitro translation (data not shown). As shown in Figure 6a, the C-termini of both PAX3 and PAX7 exhibited potent activity. Full-length GAL4-PAX3 exhibited weak activity that was fourfold greater than background; while full-length GAL4-PAX7 was inactive. Deletion of either the N-terminal extension and first alpha-helix of the paired box (amino acids 1 - 108) or the homeodomain from full-length GAL4-PAX3 resulted in sixfold or tenfold increases in activity respectively; while deletion of both regions resulted in a 63-fold increase in activity (Figure 6a). For full-length GAL4-PAX7, deletion of either the N-terminal extension and first alpha-helix of the paired box (amino acids 1 - 108) or the homeodomain resulted in transactivation that was sixfold or threefold greater than background, respectively; while deletion of both regions resulted in a 146-fold increase over background. Deletion of the C-terminal extension from the PAX7 homeodomain deletion mutant resulted in only a modest increase in activity (Figure 6a). For both GAL4-PAX3 and GAL4-PAX7, the second and third alpha-helices of the paired box and the conserved octapeptide region did not significantly contribute to cis-repression. Thus, the cis-acting transcription repression domains of PAX3 and PAX7 map to similar regions in their respective N-termini.

To determine whether the repression domains of PAX3 and PAX7 are interchangeable, we created GAL4-PAX3/PAX7 and GAL4-PAX7/PAX3 chimeras and measured their activity in NIH3T3 cells. Each chimera contained the complete DNA-binding domain of one protein fused to the C-terminus of the opposite protein. The fusion points were located immediately C-terminal to each homeodomain. As shown in Figure 6b, the PAX3 N-terminus inhibited the PAX7 transactivation domain more effectively than its own transactivation domain. At 0.01 mug test plasmid where the activities of the PAX3 and PAX7 C-termini were equal, GAL4-PAX3 exhibited 3% activity relative to GAL4-cterPAX3; while GAL4-PAX3/PAX7 exhibited 0.8% activity relative to GAL4-cterPAX7. These results suggest that the activity of the PAX7 C-terminus may be more easily repressed, perhaps indicating why PAX7 is less active than PAX3. On the other hand, there was no significant difference in the ability of the PAX7 N-terminus to inhibit either the PAX3 or PAX7 transactivation domain. Despite these differences, however, each repression domain was able to effectively inhibit each transactivation domain, demonstrating conservation of function between the PAX3 and PAX7 N-termini.


These data demonstrate that the N-terminal, cis-acting transcription repression domain is a conserved feature in PAX3 and PAX7 and that these proteins utilize a common gain of function mechanism in ARMS. Resistance to repression is accomplished through creation of a fusion protein containing the N-terminus of PAX3 or PAX7 and the resistant C-terminal transactivation domain of FKHR. Thus PAX3-FKHR and PAX7-FKHR are potent transactivators; while PAX3 and PAX7 exhibit little or no activity. We mapped interchangeable repression domains to the same regions in PAX3 and PAX7, including the N-terminal extension and first alpha-helix of the paired box (region 1, Figure 6c) and the homeodomain (region 2, Figure 6c). Conservation of structure and function in these domains was predicted, since the proteins exhibit 93% and 97% sequence identity in regions 1 and 2, respectively.

The mechanism by which the PAX3 and PAX7 N-termini repress transactivation is unknown. A recent study, however, suggests that binding of a co-repressor to the N-terminus may inhibit transactivation mediated by the C-terminus. The authors (Wiggan et al., 1998) showed that nonphosphorylated pRB binds to the PAX3 homeodomain and inhibits PAX3-dependent transactivation from reporter constructs containing either PAX-specific DNA-binding sites or the c-met or myosin light chain promoters. These results suggest that pRB modulates PAX3 activity during development in the myogenic lineage and support a model of transcriptional repression involving protein-protein interactions.

While the aforementioned study involves region 2 of the PAX3 repression domain, it is likely that region 1 also interacts with negative regulators. In this paper, we demonstrated that regions 1 and 2 independently inhibit transactivation mediated by the C-terminus in both PAX3 and PAX7. Therefore we propose a model in which the differential abilities of developmentally regulated co-repressors to interact with regions 1 and/or 2 confer flexibility in the function of PAX3 or PAX7.

Conservation of N-terminal domain structure and function suggests that PAX3 and PAX7 have overlapping function. In fact, they exhibit overlapping expression patterns in the murine embryo (Tremblay and Gruss, 1994) and interact with the same experimental DNA-binding sites in vitro (Schafer et al., 1994). Our data demonstrate, however, that PAX3 and PAX7 are not functionally identical. For example, their C-terminal domains exhibited similar potency, but different dose-dependent transactivation profiles when tested in the absence of cis-repression. The PAX7 transactivation domain was more susceptible than the PAX3 domain to repression mediated by the PAX3 N-terminus. And finally, the extended C-terminal region of PAX7 was shown to be part of the transactivation domain, contributing significantly to the activity of the protein fragment encoded by exon 8. Thus we demonstrated that PAX3 and PAX7 exhibit conservation in N-terminal, transcription repression function and subtle differences in C-terminal transactivation function.

Conservation of structure (Burri et al., 1989) and function in the N-termini of paired box-containing genes has been reported by others. For example, viability of Drosophila paired-embryos can be rescued by ectopic expression of either Drosophila gooseberry or murine Pax-3 (Xue and Noll, 1996). The potential role of domain conservation in proteins that regulate complex processes such as development is addressed in a gene network model (Frigerio et al., 1986). In this model, genes encoding proteins with multiple conserved domains arise by duplication and recombination of primordial genes corresponding to the conserved domains. Evolution of these genes results in diversification and the formation of networks with members linked by different combinations of functional domains. According to this model, PAX3 and PAX7 would belong to a developmental gene network and encode proteins that perform overlapping, but distinct functions in the regulation of embryogenesis.

An alternative, but complementary hypothesis is also supported by our data. Based on studies of the Drosophila genes paired, gooseberry, and gooseberry neuro, it was proposed that the acquisition of distinct cis-regulatory elements affecting gene expression is an evolutionary mechanism important for functional diversification of genes that arose from duplication events (Li and Noll, 1994). Therefore we suggest a unified model in which the observed combination of functional similarities and differences between PAX3 and PAX7 is explained by the gene network and the altered gene regulation models of diversification.

Although PAX3-FKHR and PAX7-FKHR were functionally equivalent in our assays, the subtypes of ARMS expressing these proteins differ in frequency, clinical presentation, and prognosis. In a study of tumors histologically diagnosed as ARMS, 75% exhibited the t(2;13) translocation; while only 10% exhibited the t(1;13) translocation (Barr et al., 1995). PAX7-FKHR expressing tumors occurred in younger patients, were more likely to occur in an extremity, were more likely to be localized lesions, and had a better prognosis when compared to PAX3-FKHR expressing tumors (Kelly et al., 1997). These differences suggest that the expression and/or function of PAX3-FKHR and PAX7-FKHR differ in vivo. In fact, we reported that distinct mechanisms account for over-expression of each fusion protein relative to the corresponding wild-type PAX protein (Davis and Barr, 1997). PAX3-FKHR over-expression was gene copy number-independent and involved increased transcription, while PAX7-FKHR over-expression resulted from gene amplification.

Besides differences in gene expression, other factors may contribute to the differential function of PAX3-FKHR and PAX7-FKHR. For example, non-conserved regions that are located between the paired box and homeodomain of both proteins might result in subtle structural differences in their DNA-binding domains, and consequent activation of distinct target genes. Alternatively, recruitment of distinct co-activators may lead to differential activation of the same target genes. In addition, temporal and spatial differences in PAX3 and PAX7 expression during development may influence phenotype by changing the environment in which the fusion proteins are expressed. PAX3 is expressed in the dorsal part of the somite and in skeletal muscle progenitors that migrate to the limb buds. Its expression precedes PAX7 whose expression occurs later and persists longer (Tremblay and Gruss, 1994). Therefore, different myogenic precursors may be involved in the two ARMS subtypes; or conversely, the timing of the gene fusion event in a single cell type may be critical in influencing phenotype.

In conclusion, two different translocations may lead to formation of ARMS. Each results in expression of a chimeric protein consisting of the N-terminal DNA-binding and overlapping cis-repression domains of either PAX3 or PAX7 and the C-terminal transactivation domain of FKHR. Thus PAX3 and PAX7 exhibit conserved domains in their respective N-termini and utilize a common gain of function mechanism in ARMS. Future studies will address potential differences in expression and transcriptional function of PAX3-FKHR and PAX7-FKHR in vivo that lead to distinct clinical phenotypes in ARMS.

Materials and methods

Cell lines

NIH3T3 murine fibroblasts (American Type Culture Collection), SJRH28 human ARMS cells (courtesy of Dr E Douglass), and COP-8 polyoma virus-transformed murine fibroblasts (courtesy of Dr Ales Cvekl) were maintained by weekly passage in Dulbecco's Modified Eagle Medium with high glucose containing 10% fetal bovine serum (FBS) and 100 mug/ml gentamycin. Media and serum were obtained from GIBCO, Life Technologies, Inc.

Expression plasmids: native proteins

All plasmids were prepared by alkaline lysis and resin purification (Qiagen). pcD3-PAX7 contains the complete PAX7 ORF (521 codons) and 43 bp of 3'UTR in a 1.7 kb insert cloned into the HindIII and ApaI sites of pcDNA3 (Invitrogen). Nucleotides 1 - 1517 of the insert were derived from a previously published PAX7 cDNA clone (Schafer et al., 1994); while the remainder of the sequence was derived from a genomic clone containing PAX7 exon 8. pcD3-PAX7-FKHR contains the complete human PAX7-FKHR ORF (831 codons) and 1175 bp of 3'UTR in a 3.8 kb insert cloned into the NotI and XbaI sites of pcDNA3. The 5' (Schafer et al., 1994) and 3' (Bennicelli et al., 1995) regions were obtained from cDNA clones; while a 330 bp ApaLI-NcoI fragment spanning the translocation breakpoint was amplified from a t(1;13)-positive tumor specimen by RT - PCR as described (Davis et al., 1994). Plasmids pcX-1 and pcD3M8, containing the complete human PAX3 and PAX3-FKHR ORFs, respectively, have been described previously (Bennicelli et al., 1996).

Expression plasmids: GAL4 fusion proteins

Analyses of C-terminal transactivation and N-terminal DNA-binding/repression domains were performed using GAL4 fusion proteins. PAX7 test regions were amplified from pcD3-PAX7 by PCR with recombinant Pfu polymerase (Stratagene) and fused in-frame to codons 1 - 147 of the heterologous GAL4 DNA-binding domain in the expression vector pcD3.GAL4 (Bennicelli et al., 1996). Primer sequences are described in Table 1. Full-length and nested 5' deletions of PAX7 were prepared using reverse primer P7R521 and forward primers P7F1, P7F108, P7F215, or P7F275. An internal deletion of the PAX7 homeodomain was made by amplification of codons 1 - 216/278 - 280 with primer pair P7F1/P7R216 and codons 215 - 216/278 - 521 with primer pair P7F278/P7R521. The overlapping products were annealed and re-amplified using primer pair P7F1/P7R521. Deletion of exon 8 from C-terminal PAX7 was performed using primer pair P7F275/P7R385. Additionally, small fragments of the 3' region encoding the transactivation domain were prepared using primer pairs P7F275/P7R385 and P7F385/P7R521. All amplified PAX7 fragments were digested with XbaI and either KpnI (fragments amplified with P7F275) or BamHI (all other fragments) and subcloned into pcD3.GAL4. Fragments amplified using P7R521 contain 21 bp of 3' untranslated region. Deletion of codons 464 - 521 from full-length or 3' PAX7 fragments was performed by digesting plasmid constructs with SfiI and XbaI, blunting the ends, and religating the plasmids.

Chimeric PAX3/PAX7 or PAX7/PAX3 proteins were prepared by PCR amplification of half-products using consensus primers located immediately 3' to the homeodomain-encoding region of each gene. The PAX3 and PAX7 N-termini were amplified using primer pairs PAX3 -6F (3)/HDCONR and P7F1/HDCONR, respectively. The C-termini were amplified using primer pairs HDCONF/PAX3 R1 (Bennicelli et al., 1995) and HDCONF/P7R521. Half products were mixed, annealed, and reamplified using the flanking primers. Chimeric PCR products were digested with BamHI and XbaI and subcloned into pcD3.GAL4.

Full-length PAX7-FKHR and a fragment encoding its C-terminal transactivation domain were PCR amplified from pcD3-PAX7-FKHR using primer pairs P7F1/5' PAX3-3' FKHR R1 (Bennicelli et al., 1995) and P7F275/5' PAX3-3' FKHR R1 respectively. Deletion of the PAX3 homeodomain was performed by amplification of codons 1 - 218/280 - 282 with primer pair PAX3-6F (Bennicelli et al., 1996) /P3R218 and codons 216 - 218/280 - 481 with primer pair P3F280/PAX3 R1 (Bennicelli et al., 1995), followed by annealing of the products and reamplification using the flanking primers. These amplified fragments were digested and subcloned into pcD3.GAL4 as described above. GAL4 fusion constructs expressing PAX3, C-terminal PAX3, and PAX3-FKHR were previously described (Bennicelli et al., 1996).

Reporter plasmids

E1b CAT (Lillie and Green, 1989) contains the adenovirus minimal promoter E1b TATA, located upstream of the reporter gene chloramphenicol acetyltransferase (CAT). Plasmids GAL45/E1b CAT (Lillie and Green, 1989) and 6´PRS-9/E1b CAT (Bennicelli et al., 1996) contain five DNA binding sites for the yeast transactivator GAL4 or six DNA binding sites for PAX3/PAX7 (Chalepakis et al., 1991), respectively.

Transfections and CAT assays

Transient transfections were performed by the calcium phosphate coprecipitation method as described (Bennicelli et al., 1995). Transfected cells (100 mm dishes) were lysed in 100 mul 250 mM Tris, pH 7.5. Expression of CAT was tested by measuring acetylation of [14C]chloramphenicol using phosphorimage analysis (Molecular Dynamics) of thin layer chromatography assays (Gorman et al., 1982). The amount of each lysate tested was normalized for transfection efficiency that was determined by measuring expression of placental alkaline phosphate from the cotransfected plasmid pSV2APAP (Henthorn et al., 1988). When necessary, simultaneous CAT reactions were performed with different amounts of lysate so that all data points could be measured within the linear range of the enzymatic assay (<50% acetylation). Experiments were performed twice, and results of single representative experiments are shown. To facilitate uniformity in graphical presentation, relative activity was calculated individually for each experiment by artibrary definition of 1 unit.

Immunoprecipitations and in vitro translations

In vitro translations were performed using the TNT T7-Coupled Reticulocyte Lysate System (Promega). Immunoprecipitations were performed as described (Bennicelli et al., 1995). After transient transfection with a cocktail of plasmids for 48 h, cells were 35S-labeled for 2 h. Labeled cells from a single dish were processed for both CAT analysis and immunoprecipitation. Aliquots (1/4 dish) of cell lysate were immunoprecipitated with polyclonal anti-FKHR antibody (Fredericks et al., 1995), polyclonal anti-GAL4 antibody (Upstate Biotechnology, Inc.), or normal rabbit serum. Portions of the immunoprecipitated samples, normalized for transfection efficiency, were resolved by SDS - PAGE.


We would like to thank Paul Pelavin and Rob McDonald for their technical assistance. These studies were supported by National Institutes of Health grant CA64202.


Barr FG. (1997). Curr. Top. Microbiol. Immunol. 220, 113-129. MEDLINE

Barr FG, Chatten J, D'Cruz CM, Wilson AE, Nauta LE, Nycum LM, Biegel JA and Womer RB. (1995). Jama 273, 553-557. MEDLINE

Bennicelli JL, Edwards RH and Barr FG. (1996). Proc. Natl. Acad. Sci. USA 93, 5455-5459. MEDLINE

Bennicelli JL, Fredericks WJ, Wilson RB, Rauscher III FJ and Barr FG. (1995). Oncogene 11, 119-130. MEDLINE

Burri M, Tromvoukis Y, Bopp D, Frigerio G and Noll M. (1989). EMBO J. 8, 1183-1190. MEDLINE

Busslinger M, Klix N, Pfeffer P, Graininger PG and Kozmik Z. (1996). Proc. Natl. Acad. Sci. USA 93, 6129-6134. Article MEDLINE

Chalepakis G, Fritsch R, Fickenscher H, Deutsch U, Goulding M and Gruss P. (1991). Cell 66, 873-884. MEDLINE

Clark J, Benjamin H, Gill S, Sidhar S, Goodwin G, Crew J, Gusterson BA, Shipley J and Cooper CS. (1996). Oncogene 12, 229-235. MEDLINE

Davis RJ and Barr FG. (1997). Proc. Natl. Acad. Sci. USA 94, 8047-8051. MEDLINE

Davis RJ, D'Cruz CM, Lovell MA, Biegel JA and Barr FG. (1994). Cancer Res. 54, 2869-2872. MEDLINE

Fredericks WJ, Galili N, Mukhopadhyay S, Rovera G, Bennicelli J, Barr FG and Rauscher III FJ. (1995). Mol. Cell. Biol. 15, 1522-1535. MEDLINE

Frigerio G, Burri M, Bopp D, Baumgartner S and Noll M. (1986). Cell 47, 735-746. MEDLINE

Fujimura Y, Ohno T, Siddique H, Lee L, Rao VN and Reddy ES. (1996). Oncogene 12, 159-167. MEDLINE

Gorman CM, Moffat LF and Howard BH. (1982). Mol. Cell. Biol. 2, 1044-1051. MEDLINE

Henthorn P, Zervos P, Raducha M, Harris H and Kadesch T. (1988). Proc. Natl. Acad. Sci. USA 85, 6342-6346. MEDLINE

Kaz K, Voz ML, Roijer E, Astrom AK, Meyen E, Stenman G and Van de Ven WJ. (1997). Nat. Genet. 15, 170-174. MEDLINE

Kelly KM, Womer RB, Sorensen PH, Xiong QB and Barr FG. (1997). J. Clin. Oncol. 15, 1831-1836. MEDLINE

Kim J, Lee K and Pelletier J. (1998). Oncogene 16, 1973-1979. MEDLINE

Ladanyi M. (1995). Diagn. Mol. Pathol. 4, 162-173. MEDLINE

Li X and Noll M. (1994). Nature 367, 83-87. MEDLINE

Lillie JW and Green MR. (1989). Nature 338, 39-44. MEDLINE

May WA, Lessnick SL, Braun BS, Klemsz M, Lewis BC, Lunsford LB, Hromas R and Denny CT. (1993). Mol. Cell. Biol. 13, 7393-7398. MEDLINE

Schäfer BW, Czerny T, Bernasconi M, Genini M and Busslinger M. (1994). Nucleic. Acids. Res. 22, 4574-4582. MEDLINE

Scheidler S, Fredericks WJ, Rauscher III FJ, Barr FG and Vogt PK. (1996). Proc. Natl. Acad. Sci. USA 93, 9805-9809. MEDLINE

Sorensen PH and Triche TJ. (1996). Semin. Cancer Biol. 7, 3-14. Article MEDLINE

Tremblay P and Gruss P. (1994). Pharmacol. Ther. 61, 205-226. MEDLINE

Vorobyov E, Mertsalov I, Dockhorn-Dworniczak B, Dworniczak B and Horst J. (1997). Genomics 45, 168-174. MEDLINE

Wiggan O, Taniguchi-Sidle A and Hamel PA. (1988). Oncogene 16, 227-236.

Xue L and Noll M. (1996). EMBO J. 15, 3722-3731. MEDLINE

Zinszner H, Albalat R and Ron D. (1994). Genes Dev. 8, 2513-2526. MEDLINE


Figure 1 PAX7-FKHR gain of function. CAT assay: NIH3T3 cells were transiently transfected for 48 h with 0, 0.01, 0.1 or 1 mug pcD3-PAX7 or pcD3-PAX7-FKHR; 5 mug 6´PRS-9/E1b CAT; and 5 mug pSV2APAP. Results are shown as the mean of triplicate data points ±s.d. (n=3)

Figure 2 Potent transactivation by PAX3-FKHR and PAX7-FKHR. CAT assays: NIH3T3 cells (a; n=3) or COP-8 cells (b; n=1) were transiently transfected for 48 h with 0, 0.01, 0.1 or 1 mug pcD3-PAX3-FKHR or pcD3-PAX7-FKHR; 5 mug 6´PRS-9/E1b CAT; and 5 mug pSV2APAP. (c) Aliquots of the same COP-8 cell lysates used for the CAT assay in b were immunoprecipitated with 4 mul of either polyclonal anti-FKHR (alpha-FKHR) antibody or normal rabbit serum (NRS). Portions of the immunoprecipitated samples, normalized for transfection efficiency, were resolved by SDS - PAGE. Lane 1, molecular weight markers; horizontal arrows, bands corresponding to PAX3-FKHR (upper) and PAX7-FKHR (lower); none, lysate of cells transfected with 0 mug test plasmid; PAX3-FKHR, PAX7-FKHR, lysates of cells transfected with 1 mug test plasmid

Figure 3 Activities of GAL4 fusion test proteins. CAT assays: cells were transiently transfected for 48 h with 0, 0.01, 0.1 or 1 mug expression plasmid, 5 mug reporter plasmid; and 5 mug PSV2APAP. (a) GAL4-fusion proteins tested with GAL45/E1b CAT reporter in SJRH28 ARMS cells (n=3); (b) GAL4-fusion proteins tested with GAL45/E1b CAT reporter in NIH3T3 cells (n=3)

Figure 4 Comparison of PAX3 and PAX7 C-terminal transactivation domains. CAT assays: NIH3T3 cells were transiently transfected for 48 h with 0, 0.01, 0.1 or 1 mug expression plasmid; 5 mug GAL45/E1b CAT; 5 mug pSV2APAP. (a) GAL4-cterPAX3 vs GAL4-cterPAX7 (n=3); (b) Deletion analysis of GAL4-cterPAX7 (n=3). GAL4, heterologous yeast GAL4 DNA-binding domain; PB, paired box; HD, homeodomain; vertical arrows, amino acid residues

Figure 5 N-terminal cis-repression domain of PAX7. CAT assays: NIH3T3 cells were transiently transfected for 48 h with various amounts of expression plasmid; 5 mug GAL45/E1b CAT; and 5 mug pSV2APAP. (a) Dose response using 0, 0.01, 0.1 or 1 mug of full-length GAL4 fusion proteins or GAL4-C-terminal fragments (n=1); (b) Single dose of 0.1 mug test plasmid (n=3)

Figure 6 Conserved structure and function of PAX3 and PAX7 cis-repression domains. CAT assays: NIH3T3 cells were transiently transfected for 48 h with various amounts of expression plasmid; 5 mug GAL45/E1b CAT; 5 mug pSV2APAP. (a) Deletion analysis of N-terminal PAX3 and PAX7 using 0.1 mug test plasmid (n=3). GAL4, heterologous yeast GAL4 DNA-binding domain; PB, paired box; HD, homeodomain; arrows, amino acid residues; hatched bars, PAX7; stippled bars, PAX3; none, no expression construct. (b) Dose response using 0, 0.01, 0.1 or 1 mug of full-length GAL4-fusion proteins; GAL4-C-terminal fragments; or GAL4-fusion chimeric PAX3/PAX7 or PAX7/PAX3 proteins (n=3). (c) Map of PAX7 showing regions 1 and 2 of the cis-repression domain


 PAX3 and PAX7 PCR primers

Received 15 October 1998; revised 9 February 1999; accepted 9 March 1999
29 July 1999, Volume 18, Number 30, Pages 4348-4356
Table of contents    Previous  Article  Next    [PDF]