Further Examination of BCL-XL's Role in the Apoptotic Pathway
Abstract
Every year, one million people in the United States are diagnosed with cancer. In the fight against cancer, the search for a mechanism that promotes cancer growth is vitally important � as this provides a basis for a possible cure. One possible route is to examine the anti-apoptotic proteins that promote cell longevity; their role is vitally important to the cancer�s continued survival, and it is thus pertinent to study the factors that promote the expression of these proteins. Rhabdomyosarcomas, the subclass of cancers affecting muscle cells, are characterized by chromosomal translocations containing genes that encode for the aberrant fusion transcription factors PAX3/FKHR and PAX7/FKHR. The PAX family of genes encodes for DNA binding proteins that play important roles in cell development and differentiation. Fused with FKHR, the transactivational capabilities of the protein, and subsequently its oncogenicity, are increased. It was recently established that enhanced expression of these transcription factors and their normal counterparts stimulate transcription of BCL-XL, an anti-apoptotic protein. Our lab seeks to further understand BCL-XL�s role in the oncogenic pathway by assessing whether or not this protein exists in the same oncogenic pathway that PAX3/FKHR and PAX7/FKHR works in, and if so, what changes in the apoptotic timetable would be observed upon suppression of this pathway. Establishment of a pathway directly associated with the chromosomal translocations of rhabdomyosarcomas would serve to create a target for more efficient and informative research into the mechanisms of cancer.
Background and Significance
7% of all diagnosed pediatric cancers come from the family of soft-tissue sarcomas, the most common being rhabdomyosarcomas (RMS). RMS is found in the skeletal muscles involved in voluntary movement and symptoms include pain and malignant tumors. Normal treatments are targeted toward destroying the tumor, including chemotherapy and radiation therapy. There are two different types of RMS: alveolar and embryonal. Embryonal RMS is more common and affects younger children, but alveolar RMS affects older children and is associated with higher stage disease and poor prognosis. However, increased emphasis on treatment and mechanism for alveolar RMS has decreased the survival difference between the two types.
RMS tumor cells frequently display chromosomal translocations involving the DNA-binding domains of PAX, a member of the Paired Box transcription factor family, and the transactivation domain of the FKHR gene, a member of the forkhead/HNF-3 transcription factor family. Translocations common to alveolar RMS patients include t(2;13)(q35;q14) and variant t(1;13)(p36;q14) which result in PAX3/FKHR and PAX7/FKHR fusion genes respectively. These fusion products are better transcriptional activators than normal PAX3 and affect normal downstream targets of the PAX genes, or other possible targets, in a disruptive manner.
Recent observation of RMS in a mouse model using a deletion of PTC, a protein regulated downstream by PAX3, is evidence toward a role for PAX3 in RMS tumor development7. Other proteins that normally work with PTC are involved in normal skeletal muscle development, as is PAX3. Deletion of PAX3 in one mouse model eliminated growth of limb muscles. This defect is due to a non-migration of myogenic precursors from the lateral dermomyotome, a specific layer within the embryo8. PAX3 thus plays an important early developmental role. PAX3 is deeply associated with tumorigenesis and normal muscle development, making it a key target for mechanism-based treatment research.
One hypothesis of the function of the fusion products is that they are transcription factors with the DNA binding abilities of the PAX domains and the ehanced transcriptional activation of the FKHR domains. However, it is interesting to note that only the homeodomain recognition helix of PAX3 is necessarily required for oncogenic transformation, and not the paired-box DNA binding domain9. In vitro testing has revealed that expression of the fusion gene can transform fibroblasts and that PAX3/FKHR is indeed a greater transcriptional activator than normal PAX3. It has also become more evident that the PAX3/FKHR and PAX7/FKHR fusion products differentially induce cancer and are thus different in some basic way � the PAX3 version is associated with a poor prognosis while the PAX7 version is associated with a better prognosis12. Greater exploration of this pathway could reveal a better understanding of RMS.
Though the chromosomal translocation fusion products are highly significant genetic alterations specific to rhabdomyosarcoma, there are other fairly important mutations that are important in rhabdomyosarcomagenesis. In 50% of RMS cases examined, researchers have found mutations in the p53 gene, a common genetic lesion in human cancers. Also, mutated proteins N-myc, N-ras and K-ras have also been found to be fairly consistent oncogenic abnormalities in RMS tumors. The q13-15 region of chromosome 12 is frequently amplified in RMS, which encodes for the regulatory factos GLI, MDM2 and CDK4.
Preliminary Data
Extensive work on alveolar RMS over the years has allowed our lab to be able to localize the t(2;13) breakpoint on both chromosomes 2 and 13 that has been associated with RMS, and demonstrate that PAX-3, in combination with FKHR, produces potentially tumorigenic hybrid transcription factors that are characteristic of this cancer. Also, we have developed a molecular assay that accurately identifies these fusion transcripts. We have also demonstrated that the variant t(1;13) (PAX7/FKHR) results in a similar protein product that is activated by a different mechanism, and has its own distinct clinical phenotype. Our lab has shown in vivo and in vitro that the PAX3/FKHR protein product is a much more potent transcriptional activator than PAX-3 because of the different transactivation domains � the FKHR activation domain is insensitive to the inhibitory effects of N-terminal PAX3/PAX7 domains. Amplification of the fusion gene occurs sequentially after translocation and alters the number and structure of the genes to activate oncogenic activity by complementary mechanisms. We have demonstrated that FKHR localization is regulated by an AKT-dependent pathway, but not the fusion proteins. So, the expression of these aberrant genes leads to tumorigenesis by affecting protein function and localization � impacting control of growth, apoptosis, differentiation and motility30.
We have demonstrated that PAX3-FKHR regulates downstream effector genes involved in RMS tumorigenesis. We have found that FKHR, on its own, represents a bifunctional nuclear receptor intermediary protein that can act as either a coactivator or corepressor. Genes containing PAX-binding sites have been found to have expression levels correlated to that of PAX3 and PAX3/FKHR, thus indicating that they are part of the PAX3 regulatory mechanism. In one study, we showed that PAX3/FKHR regulates a downstream receptor involved in growth and motility signaling called MET. It is also involved with regulation of other proteins that function in tumorigenesis.
Current work has established that part of the anti-apoptotic effort of the fusion protein comes from regulation of the BCL-XL protein, which functions in normal and RMS skeletal muscle cells. We had noticed that the BCL-2 family of proteins was a common anti-apoptotic protein, and so we decided to study its expression in numerous cell lines. BCL-XL was selected as the study focus because of evidence that PAX3/FKHR could transcriptionally activate the Bcl-x promoter. Since nothing much was known about normal expression of BCL-XL, its expression was tested in two myoblast isolates and RMS cell lines individually expressing mostly PAX7, mostly PAX3 and mostly PAX3/FKHR. Normal expression of BCL-XL in the myoblasts showed that PAX3/FKHR generates the strongest BCL-XL mRNA response in a Northern blot analysis; BCL-XL expression in the RMS cell lines correlate directly. This would suggest a direct relationship in transcriptional regulation between BCL-XL and PAX3/FKHR.
Overexpression of PAX3 and PAX3/FKHR in the Rh1 cell line, where BCL-XL expression was lowest, led to heightened BCL-XL mRNA levels, after Northern blot analysis. Analysis of clones showed that the level of BCL-XL seemed to correlate with the level of PAX3 � when no PAX3 was charted, no BCL-XL was found either. Further experiments in mouse myoblasts yielded similar results. It can be understood that PAX3 proteins can directly stimulate transcription of BCL-XL, and in myoblasts as well.
Analysis of the Bcl-x promoter found PAX3 DNA-binding target sequences and ATTA boxes that would mediate binding of the PAX3 homeodomain. Experiments using Bcl-x promoter deletions generated in front of a luciferase reporter gene and increasing amounts of PAX3 and PAX3/FKHR yielded increasing luciferase activity. Elimination experiments showed that -163 to -25 on the Bcl-x promoter was the probable location of important binding sites for PAX3 and PAX3/FKHR. To further refine the search, more restriction fragments were taken and run with PAX3 extracts to generate bands. Using knowledge of the ATTA-box location, we used oligonucleotides related to the ATTA-box constructs and a mutated version as a control in electrophoretic mobility shift assays using PAX3 and its antibody. This returned to us a location of -67 to -18 as the more precise location. Similar assays using the same oligonucleotides with competitor promoters showed that PAX3/FKHR binds with higher affinity to the Bcl-x promoter.
Overexpression of BCL-XL, using retroviral gene transfer, protected against apoptosis in cells that had PAX3 downregulated. Immunofluorescent stains confirmed overexpression of BCL-XL in the right places, and incubation with antisense oligonucleotides against PAX3 and PAX7 was done, leading to heightened survival for the cell lines.
This study identified a novel protein as a target gene downstream, regulated by PAX3/FKHR. We were able to show that it was a direct regulation of transcription and that PAX3/FKHR can bind the Bcl-x promoter with higher affinity. We were also able to show that BCL-XL can protect RMS cells from apoptosis in the absence of PAX3 � signifying the importance of BCL-XL�s function in RMS and possible treatment mechanisms.
Specific Aim
The PAX family of genes, which are associated with DNA binding, have increased transactivational capabilities and oncogenicity in combination with the FKHR gene. These fusion transcription factors characterize the subclass of cancers known as rhabdomyosarcomas (RMS) and are encoded for by translocations of chromosomes 2 and 13, and 1 and 13. It has recently been suggested that these proteins have important roles in the pathogenesis of RMS and lie in a specific pathway directly associated with ani-apoptotic functions. Recent work has shown that enhanced expression of PAX3 and its oncogenic counterpart PAX3/FKHR stimulates transcription of BCL-XL, a protein that protects against apoptosis in normal and cancerous cells. Also, ectopic overexpression of any of these proteins can reverse apoptosis induced by downregulation of PAX proteins. An important finding was that these proteins can transcriptionally activate the Bcl-x promoter. In these studies, we aim to assess the capabilities of the related PAX7 and PAX7/FKHR fusion protein in transcriptional activation of the Bcl-x promoter.
1) Assessing the transactivation capabilities of PAX7 and fusion protein on the Bcl-x promoter. It was found that PAX5, a protein from the same family as PAX3 and PAX7, could not activate the Bcl-x promoter, as PAX3 had done. It was recently suggested that PAX3 only needed the homeobox domain for oncogenic activation; this could partially explain why PAX5 could not bind: it did not contain the full homeodomain. PAX7 does contain a full homeobox domain, and if the explanation is correct, it should be able to activate the Bcl-x promoter. These studies will attempt to assess whether or not PAX7 is capable of activating the promoter, and if so, if the homeodomain is important to activation, and thus to the oncogenic mechanism.
Experimental Design
To find out if PAX7 and PAX7/FKHR can transcriptionally activate the Bcl-x promoter, we must find out if these proteins are related at all to transcriptional regulation of BCL-XL. From previous data, we can conclude that PAX7 induces little BCL-XL response; however, it would behoove us to double check. Using B6M and A33 myoblast cells, Rh1 cells that express PAX7, and a cell line that expresses PAX7/FKHR, Northern blots and immunoblots will be prepared for analysis.
Northern blot analysis: 1) Five micrograms of total RNA loaded in each lane, separated on 1% agarose-formaldehyde gels and blotted on nitrocellulose. Lane 1 through 4 will be growth and fusion medium runs of the myoblast isolates, and Lane 5 and 6 will be the Rh1 cells and the PAX7/FKHR expressed cell line. 2) Using a 474 base pair human cDNA BamHI/PvuII BCL-XL restriction fragment as a probe with a primer as a label, hybridization and washing will be done with SSC and SDS, before being visualized with autoradiography. 3) As a control, all lanes are run with a probe against actin.
Immunoblot analysis: 1) Lysis of Rh1 and PAX7/FKHR expressed cells and extraction of protein, 100 micrograms separated on 12% SDS-polyacrylamide gel and transferred to a polyvinylidene difluoride membrane. 2) Incubation with rabbit anti-BCL-X antiserum with a anti-rabbit IgG as detection system.
This would not only show PAX7 activity, but a comparison with PAX7/FKHR activity. We would expect that the fusion protein would induce greater amounts of BCL-XL response than PAX7 by itself. The PAX7 solo response is less but still exists as compared to the myoblast isolates, which are the �normal� cells.
We would be interested to know if PAX7 does somehow have a similar relationship albeit weaker with BCL-XL. So, we would run another Northern blot analysis using ectopic expression of PAX7 and PAX7/FKHR in RD cells (expressing PAX3). Since there is high expression of BCL-XL in the presence of PAX3, we must be careful to not include any data bias by accidentally including PAX3.
Northern blot analysis: Using the earlier procedure, the RD cell line and several RD clones will be used for the RNA samples. Probes used will be the 474 base pair human cDNA BamHI/PvuII BCL-XL restriction fragment, and PAX7 and PAX7/FKHR identifying probes, with labels. The procedure will then be repeated, using mouse fibroblast and myoblast cells, and also PAX7 and PAX7/FKHR expressed clone lines.
The results from this experiment are highly dependent on results from the first. If there has been any interaction, this experiment would prove that there is a direct interaction in transcriptional regulation. However, it is also possible that this experiment will not show any correlations between levels and thus PAX7 should not play a role in this mechanism. If it does, the fact that it might bind to the Bcl-x promoter would mean that the increased levels of BCL-XL are indeed caused by direct transcriptional activation.
We must then search the murine Bcl-x promoter for potential PAX7 DNA-binding target sequences as well as ATTA boxes. Using selected target sequences, we can compile deletion constructs to use with a luciferase reporter gene in several cotransfection assays. These constructs should vary in size and both encompass and exclude the selected sequence areas, using selected restriction enzymes. First, we will use the whole sequences to decide if promoter binding occurs. If it does, then it will be useful to proceed with more specific deletion constructs. Through Swiss-PROT, we know where the homeobox domain is, at 217-276. The paired-box domain is located at 34-16138. If the surmise that the homeobox domain is the important domain for binding is correct, than results should show promoter binding in cotransfection assays containing just that sequence.
Cotransfection assays: 1) PCR-generated deletion constructs cloned using mouse genomic DNA subcloned into vector containing the luciferase gene. Complete sequences for PAX7 and PAX7/FKHR cloned into separate retroviral vectors. 2) Transfection done by standard CaPO4 method, using 2 mg of the luciferase reporter construct and increasing amounts of the PAX7 expressing constructs to transfect into 10T1/2 mouse fibroblasts of C2Cl2 mouse myoblasts. 3) Luciferase assays (visible as a glow) done 48 hours later using commercially available kits. 4) Assays repeated using smaller PAX7 and PAX7/FKHR sequences.
If promoter binding does occur, then smaller sequences of PAX7 can be used, namely the homeobox and paired-box domains discussed earlier. This whole experiment would reveal whether or not PAX7 and PAX7/FKHR, which induces a different clinical phenotype of RMS than PAX3/FKHR, can work through the same mechanism for protection against apoptosis. However, a secondary goal would be to establish if the homeobox domain is important in the binding of the Bcl-x promoter, so another set of transfection assays should be completed, using other PAX proteins containing the homeodomain.
Cotransfection assays: Repeat previous procedure, using PAX4 and PAX6 complete sequences, and homeobox and paired-box sequences. Since it has not been established to work with FKHR in any oncogenic mechanism, the combined sequence is not needed.
1. �Who Gets Cancer?� American Cancer Society. 2003.
2. �Soft-Tissue Sarcomas.� Memorial Sloan-Kettering Cancer Center. 2003.
3. Anderson, J. et al. �Genes, Chromosomes, and Rhabdomyosarcoma.� 1999. Genes, Chromosomes & Cancer. 26: 275-285.
4. Barr, F.G. et al. �Rearrangement of the PAX3 paired box gene in the paediatric solid tumour alveolar rhabdomyosarcoma.� 1993. Nat. Genet. 3: 113?117.
5. Kenet, G. et al. �Chromosomal translocation (1;13) in a case of alveolar rhabdomyosarcoma.� 1998. J. Pediatric. Haem. Oncol. 20: 86�87.
6. Merlino, G. et al. �Rhabdomyosarcoma - Working Out The Pathways.� 1999. Oncogene. 18(38): 5340-5348.
7. Hahn, H. et al. �Rhabdomyosarcomas and radiation hypersensitivity in a mouse model of Gorlin syndrome.� 1998. Nat. Med. 4: 619�622.
8. Bober, E. et al. �Pax-3 is required for the development of limb muscles: a possible role for the migration of dermomyotomal muscle progenitor cells.� 1994. Development 120: 603-612.
9. Lam, P. et al. �The oncogenic potential of the Pax3-FKHR fusion protein requires the Pax3 homeodomain recognition helix but not the Pax3 paired-box DNA binding domain.� 1999. Mol. Cell. Biol. 19:594-601.
10. Scheidler, S. et al. �The hybrid PAX3-FKHR fusion protein of alveolar rhabdomyosarcoma transforms fibroblasts in culture.� 1996. Proc. Natl. Acad. Sci. USA 93: 9805�9809.
11. Bennicelli, J. L. et al. �Wild-type PAX3 protein and the PAX3-FKHR fusion protein of alveolar rhabdomyosarcoma contain potent, structurally distinct transcriptional activation domains.� 1995. Oncogene 11: 119�130.
12. Kelly, K.M. �Common and variant gene fusions predict distinct clinical phenotypes in rhabdomyosarcoma.� 1997. J. Clin. Oncol. 15: 1831�1836.
13. Stratton, M.R. et al. �Mutation of the p53 gene in human soft tissue sarcomas: association with abnormalities of the RB1 gene.� 1990. Oncogene. 5: 1297-1301.
14. Wexler, L.H. et al. �Pediatric soft tissue sarcomas.� 1994. CA. Cancer J. Clin. 44: 211-247.
15. Khan, J. et al. �Gene expression profiling of alveolar rhabdomyosarcoma with cDNA microarrays.� 1998. Cancer Res. 58: 5009-5013.
16. Khatib, Z.A. �Coamplification of the CDK4 gene with MDM2 and GLI in human sarcomas.� 1993. Cancer Res. 53: 5535-5541.
17. Barr, F.G. et al. �Localization of the rhabdomyosarcoma t(2;13) breakpoint on a physical map of chromosome 13. � 1991. Genomics. 11(4):941-7.
18. Barr, F.G. et al. �Localization of the t(2;13) breakpoint of alveolar rhabdomyosarcoma on a physical map of chromosome 2. � 1992. Genomics. 13(4):1150-6.
19. Biegel, J.A. et al. �Detection of the t(2;13)(q35;q14) and PAX3-FKHR fusion in alveolar rhabdomyosarcoma by fluorescence in situ hybridization.� 1995. Genes, Chromosomes & Cancer. 12(3):186-92.
20. Galili, N. et al. �Fusion of a fork head domain gene to PAX3 in the solid tumour alveolar rhabdomyosarcoma.� 1993. Nature Genetics. 5(3):230-5.
21. Barr, F.G. et al. �Rearrangement of the PAX3 paired box gene in the paediatric solid tumour alveolar rhabdomyosarcoma. � 1993. Nature Genetics. 3(2):113-7.
22. Edwards, R.H. et al. �Detection of gene fusions in rhabdomyosarcoma by reverse transcriptase-polymerase chain reaction assay of archival samples.� 1997. Diagnostic Molecular Pathology. 6(2):91-7.
23. Davis R.J. et al. �Fusion of PAX7 to FKHR by the variant t(1;13)(p36;q14) translocation in alveolar rhabdomyosarcoma. � 1994. Cancer Research. 54(11):2869-72.
24. Davis, R.J. et al. �Fusion genes resulting from alternative chromosomal translocations are overexpressed by gene-specific mechanisms in alveolar rhabdomyosarcoma.� 1997. Proc. Natl. Acad. Sci. 94(15):8047-51.
25. Kelly, K.M. et al. �Common and variant gene fusions predict distinct clinical phenotypes in rhabdomyosarcoma.� 1997. Journal of Clinical Oncology. 15(5):1831-6.
26. Fredericks, W. J. et al. �The PAX3-FKHR fusion protein created by the t(2;13) translocation in alveolar rhabdomyosarcomas is a more potent transcriptional activator than PAX3.� 1995. Molecular & Cellular Biology. 15(3):1522-35.
27. Bennicelli, J.L. et al. �Wild type PAX3 protein and the PAX3-FKHR fusion protein of alveolar rhabdomyosarcoma contain potent, structurally distinct transcriptional activation domains.� 1995. Oncogene. 11(1):119-30.
28. Scheidler, S. et al. �The hybrid PAX3-FKHR fusion protein of alveolar rhabdomyosarcoma transforms fibroblasts in culture.� 1996. Proc. Natl. Acad. Sci. 93(18):9805-9.
29. Bennicelli, J. L. et al. �Mechanism for transcriptional gain of function resulting from chromosomal translocation in alveolar rhabdomyosarcoma.� 1996. Proc. Natl. Acad. Sci. 93(11):5455-9.
30. Barr, F. G. Gene fusions involving PAX and FOX family members in alveolar rhabdomyosarcoma. 2001. Oncogene. 20(40):5736-46.
31. Barr, F.G. et al. �In vivo amplification of the PAX3-FKHR and PAX7-FKHR fusion genes in alveolar rhabdomyosarcoma.� 1996. Human Molecular Genetics. 5(1):15-21.
32. Del Peso, L. et al. �Regulation of the forkhead transcription factor FKHR, but not the PAX3-FKHR fusion protein, by the serine/threonine kinase Akt.� 1999. Oncogene. 18(51):7328-33.
33. Tand, E.D. et al. �Negative regulation of the forkhead transcription factor FKHR by Akt.� 1999. Journal of Biological Chemistry. 274(24):16741-6.
34. Zhao, H. H. et al. �Forkhead homologue in rhabdomyosarcoma functions as a bifunctional nuclear receptor-interacting protein with both coactivator and corepressor functions.� 2001. Journal of Biological Chemistry. 276(30):27907-12.
35. Barber et al. �Identification of Target Genes Regulated by PAX3 and PAX3�FKHR in Embryogenesis and Alveolar Rhabdomyosarcoma.� 2002. Genomics. 79(3):278-284.
36. Ginsberg, J.P. et al. �Up-regulation of MET but not neural cell adhesion molecule expression by the PAX3-FKHR fusion protein in alveolar rhabdomyosarcoma.� 1998. Cancer Research. 58(16):3542-6.
37. Margue. C.M. et al. �Transcriptional modulation of the anti-apoptotic protein BCL-XL by the paired box transcription factors PAX3 and PAX3/FKHR.� Oncogene. 19(25):2921-9, 2000.
38. PAX7 Protein Structure. SwissPROT. 2003.