- Email: [email protected]
A functional variant in miR-605 modifies the age of onset in Li-Fraumeni syndrome
Cancer Genetics 208 (2015) 47e51
BRIEF COMMUNICATION
A functional variant in miR-605 modifies the age of onset in Li-Fraumeni syndrome a,b a,b,c, * Badr Id Said , David Malkin a
Division of Hematology/Oncology, Department of Pediatrics, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada; b Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; c Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada Li-Fraumeni syndrome (LFS) is a rare cancer predisposition syndrome, typically associated with germline mutations in the TP53 gene. Despite the high penetrance of TP53 mutations, LFS patients display striking phenotypic differences, suggesting the presence of secondary risk loci. To date, all genetic modifiers in LFS have been shown to map to either TP53 or its principal negative regulator, MDM2. Given this strong association, we set out to investigate the contribution of miR605, a recently described microRNA (miRNA) regulator of the p53-MDM2 loop. We hypothesized that, if functional, the miR-605 gene and its variant (rs2043556) could impact the cancer risk profile of TP53 mutation carriers. Consistent with this proposition, the variant G-allele of miR-605 was associated with a 10-year acceleration in the mean age of LFS tumor onset (P Z 0.04) and caused a 2.6-fold reduction in the processing levels of its host miRNA (P < 0.05). We also demonstrate that miR-605 overexpression leads to a decrease in cell proliferation, clonogenicity, and migration in two rhabdomyosaroma cell lines carrying hotspot TP53 mutations. Together, our results implicate miR-605 as a novel modifier gene of the LFS phenotype and a promising therapeutic target in TP53 mutant cancers. Keywords MicroRNA, Li-Fraumeni syndrome, single nucleotide polymorphism, p53 ª 2015 Elsevier Inc. All rights reserved.
MicroRNAs (miRNAs) are approximately 22 nucleotide noncoding RNAs that act as sequence-specific post-transcriptional regulators of gene expression, affecting >60% of human genes (1,2). In cancer, miRNAs are often aberrantly expressed as a result of a variety of mechanisms, including genomic rearrangements, mutations, and DNA copy number and epigenetic changes (3). Recently, common single nucleotide polymorphisms (SNPs) within miRNA genes have emerged as yet another potent mechanism for miRNA deregulation (4). Such functional SNPs have been associated with sporadic and familial cancers in several association studies, yet their potential contribution to Li-Fraumeni syndrome (LFS) remains unexplored. LFS is a highly penetrant autosomal dominant cancer disorder that leads to the development of a variety of earlyonset cancers, including breast cancer, brain tumors, adrenocortical carcinomas, and soft-tissue and bone sarcomas.
Received August 18, 2014; received in revised form December 22, 2014; accepted December 23, 2014. * Corresponding author. E-mail address: [email protected] 2210-7762/$ - see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.cancergen.2014.12.003
Despite compelling evidence linking LFS heterogeneity to functional differences in germline p53 mutant activity, these genotypephenotype correlations do not yet fully account for either the diversity of tumors or the disparate ages of cancer onset typically observed in LFS patients (5,6). Furthermore, functional differences in p53 activity cannot account for intrafamilial variations in LFS, given that all affected members carry the same defective copy of the TP53 gene. Instead, secondary or coexistent genetic factors that modify p53 function have been posited as responsible for an important spectrum of these clinical differences. To date, functional polymorphisms in TP53 (Arg72Pro and PIN3), MDM2 (SNP 309 [a T>G substitution at nucleotide 309 of intron one of MDM2 {rs2279744}]), as well as excess DNA copy number variations (CNVs) and accelerated telomere attrition, have all been associated with the heterogeneous tumor patterns of LFS (7e12). From this work, it has become clear that individual LFS phenotypes are defined not only by the underlying germline TP53 mutation, but also by an important interplay of different modifier genes. The influence of different combinations of these modifiers could account for the wide spectrum of different susceptibilities that characterize LFS. Therefore, developing useful biomarkers
48 for the future clinical monitoring of LFS patients will require a more thorough and complete understanding of all TP53 modifying events. Lately, several miRNAs have emerged as major players in the p53 network (13). One particular member of this family, miR-605, was found to be a p53-inducible miRNA capable of targeting MDM2, a critical cellular inhibitor of p53 (14). Interestingly, a potentially functional polymorphism (rs2043556) within the miR-605 sequence has been previously associated with cancer risk (15,16). Due to its location in the precursor structure of miR-605, this genetic variant could impact the post-transcriptional processing of miR-605. Importantly, this miRNA has been shown to be deregulated in some human tumors. It is also expressed in tissues from which common LFS tumors develop, namely, brain, skeletal muscle, and breast (14,17). Thus, SNP-mediated miR-605 deregulation could modulate LFS cancer risk by causing important transcriptomic changes in tissues typically affected by this disorder. Additional functional features of miR-605, such as its association with p53 signaling and its capacity to regulate multiple genes, render miR-605 and its variant promising candidate modifiers in LFS.
Materials and methods Study subjects The study population consists of 55 Caucasian individuals with a germline TP53 mutation. Written informed consent to participate in the research study was obtained from the individual or his or her parent (if the participant was below the age of consent). The study was approved by the Research Ethics Board at The Hospital for Sick Children (REB no. 0019910602).
miR-605 SNP genotyping The rs2043556 variant was genotyped by SNP-RFLP (restriction fragment length polymorphism). First, a 284-bp DNA fragment that contained pre-miR-605 was amplified from DNA extracted from peripheral blood lymphocytes by PCR: 50 -CGCCTCTTTTTGCTCATTCT-30 and 50 -AGAGCAGTT ACGCCACATGA-30 . In homozygote carriers of the miR-605 A-allele, a HinfI enzyme (New England Biolabs, Ipswich, MA) recognition site was created in the 284-bp PCR fragment, which resulted in two equal 142-bp fragments. In contrast, HinfI-digested PCR fragments from carriers of the A/G genotype produce three fragments: one uncut 284-bp fragment and two equal 142-bp fragments.
Cell lines The RD and Rh30 rhabdomyosarcoma cell lines (both from Dr. A. Thomas Look of Dana-Farber Cancer Institute, Boston, MA) were selected because of their mutant TP53 status (R273 C and R248 W, respectively) as well as the fact that rhabdomyosarcoma is an important component tumor of LFS. These cell lines were confirmed to be mycoplasmafree.
B. Id Said, D. Malkin
miR-605 expression vectors and RD cell nucleofection MiRNA expression vectors (pCMV-MIR) that contained a 639-bp fragment corresponding to pre-miR-605-G and premiR-605-A, along with an empty control, were purchased from OriGene (Rockville, MD). These expression plasmids were transiently transfected into RD cells by nucleofection (Amaxa Inc., Gaithersburg, MD). Total RNA was extracted 6 hours later for quantitative reverse transcription PCR (qRT-PCR).
qRT-PCR of mature and pre-miR-605 To evaluate expression levels of mature and pre-miR-605, total RNA (200 ng) from pre-miR-605-A- and pre-miR-605G-nucleofected cells was reverse transcribed into cDNA using either specific stem-loop RT primers for miR-605 and U6 (Applied Biosystems, Foster City, CA), or random hexamer primers (Bio-Rad Laboratories, Hercules, CA). Subsequently, the PCR levels of these miRNAs was quantified using the corresponding TaqMan miRNA Assays (Applied Biosystems). The amount of mature and pre-miR605 relative to the negative control was calculated using the 2-DDCt method and normalized to U6. Results represent the fold-change over control and are shown as mean SEM; each experiment was conducted at least four times (see Supplementary data).
Statistical analysis Data were analyzed using Prism 3.0 software (GraphPad Software, Inc., San Diego, CA). For functional studies, statistical analyses were performed using the unpaired twotailed Student t-test. The non-parametric Mann-Whitney test was used to compare the age of cancer onset among the different SNP genotypes.
Results The miR-605 SNP modifies the age of cancer onset in LFS patients Since miR-605 is an important regulator of the p53-MDM2 loop, and its variant is predicted by RNAfold to impact the stability of its precursor structure by DDG Z e2.6 kcal/mol (Figure S1), we analyzed the association of this polymorphism with the age of cancer onset in LFS patients who carried germline TP53 mutations (18). Our study sample consists of a total of 55 Caucasians (Table S1), 21 of whom are related to each other through 10 distinct family groupings. The majority of these families (9 of 10) are comprised of pairs, with only 1 family counting more than 2 members (n Z 3). Genotyping of blood-derived DNA samples from these 55 patients was performed using SNP-RFLP. Samples were then divided according to their miR-605 genotype (A/A or A/G), and their age at cancer onset was compared; the rare G/G genotype was not present in our study sample. Interestingly, results revealed that the G-allele was associated with a significantly earlier age of cancer onset
Functional variant in miR-605 affects Li-Fraumeni syndrome
49
(P Z 0.04). The mean and median age of cancer onset for LFS carriers of the A/G genotype (n Z 17) was 12.8 years and 4 years, respectively, compared with 22.9 years and 23.5 years for carriers of the A/A genotype (n Z 38) (Figure 1). Stratifying by tumor type revealed accelerated cancer onset in carriers of the G-allele across all of the main LFS tumor types, such as brain tumors, sarcomas, and adrenocortical carcinomas, with the exception of breast cancer (Table S2). Furthermore, even when we repeated the analysis using samples from unrelated individuals (n Z 34), we observed a similar reduction in the age of cancer onset (P Z 0.1). The mean and median age of cancer onset for carriers of the A/G genotype (n Z 9) was 14.6 years and 4 years, respectively, compared with 23.4 years and 27 years in patients with the common A/A genotype (n Z 25) (Table S3). However, the frequency of the G-allele among patients who developed more than one tumor did not differ from the frequency observed in those with the A/A genotype (Table S5). Further reinforcing our results is the lack of a significant difference in the tumor distribution across the two genotype groups (Figure S2). Thus, despite our relatively low sample size, these data demonstrate, for the first time, an association between a common SNP in a pre-miRNA gene and the age of tumor onset in LFS.
The miR-605 SNP affects the processing efficiency of pre-miR-605 To determine the functional consequences of the miR-605 SNP, miRNA plasmids (pCMV-MIR) containing pre-miR-605G, pre-miR-605-A, and a control vector were nucleofected into the rhabdomyosarcoma cell line (RD); these plasmids were sequence confirmed (Figure 2A). Relative to the control cells, significantly higher levels of mature miR-605 (a 2.6-fold difference) were detected in RD cells transfected with premiR-605-A compared with the pre-miR-605-G-transfected cells (Figure 2B). Interestingly, qRT-PCR measurements of the precursor form of miR-605 (pre-miR-605) in these same samples revealed no such expression difference, indicating that the G-allele of this SNP impacts the efficiency of miR605 processing from its precursor to mature form (Figure 2C). By calculating the expression ratio of mature miR-605 to that of pre-miR-605 in each transfected sample, we determined the processing efficiency of the pre-miR-605G plasmid to be 65% lower than that of the pre-miR-605-A plasmid (Figure 2D). Since miR-605 had been shown in a previous study to exhibit potent tumor suppressive functions in TP53 wild-type cells, but not in isogenic cell lines lacking the TP53 gene (14), it was also important to investigate the tumor suppressive role of miR-605 in the context of TP53 mutations. Overexpression of miR-605 resulted in a significant reduction in cell viability, clonogenicity, and cell migration in two TP53 mutant cell types (Figure S3). Thus, our data confirms that SNP-mediated deregulation of miR-605 is a plausible cancer modifier in TP53 mutant cells.
Discussion
Figure 1 (A) Age of cancer onset in TP53 mutation carriers according to miR-605 genotype. Germline DNA samples from 55 Caucasian TP53 mutation carriers who had written informed consent to participate in the research study were genotyped for the miR-605 variant (rs2043556) using SNP-RFLP. Randomly chosen samples were further genotyped by direct sequencing in order to confirm the RFLP data. Samples were then classified according to their rs2043556 genotype (A/A or A/G), and their age at cancer onset was compared; the low frequency G/G genotype was not detected. Using the non-parametric MannWhitney test, TP53 mutation carriers with the A/G genotype developed cancer at significantly earlier ages compared with individuals with the A/A genotype (P Z 0.04). (B) Percentage of unaffected TP53 mutation carriers for each miR-605 genotype as a function of age of cancer onset.
To our knowledge, this is the first report describing an association between a common SNP in a pre-miRNA gene and the age of tumor onset in LFS. Functionally, the G-allele of miR605 appeared to cause a defect in the processing efficiency of its host miRNA. These findings are consistent with a recent report in which mature miR-605 expression levels were shown to be attenuated in patients with colorectal cancers who carried the G/G genotype (15). Given that the unwinding of premiRNAs is an important miRNA processing step, less stable RNA duplexes, such as the pre-miR-605-A variant (DDG Z 2.6 kcal/mol), would be predicted to be processed faster and more efficiently by Dicer as compared with the more stable pre-miR-605-G variant. This difference can be partly explained by the presence of the stronger G:C pairing in pre-miR-605-G compared with the less stable G:U pairing in pre-miR-605-A. This mechanism could potentially explain the observed overexpression of mature miR-605 from the A-allele as well as the lack of such difference in the precursor levels of miR-605 between these two variants. Thus, our work establishes cytoplasmic miRNA processing by Dicer as a potential mechanism for miR-605 SNP-mediated dysfunction. Because each miRNA is thought to regulate >100 different gene products, SNP-mediated deregulation of miR605 can be predicted to have widespread biological effects. Specifically, transcripts normally regulated by miR-605, including MDM2, could be less efficiently regulated as a result of lower miR-605 expression. In LFS patients who carry a mutant TP53 gene, higher levels of MDM2 have been
50
B. Id Said, D. Malkin
Figure 2 (A) Sanger sequencing of the pre-miR-605-A and pre-miR-605-G vectors. miRNA expression vectors (pCMV-MIR) containing a 639-bp fragment corresponding to pre-miR-605-G and pre-miR-605-A were sequenced in order to confirm the correct DNA sequence. (B) Impact of the miR-605 SNP on mature miR-605 levels. The aforementioned miRNA plasmids were transiently transfected into RD cells by nucleofection. Compared with the empty vector, mature miR-605 levels were up-regulated 137.1-fold in RD cells transfected with pre-miR-605-A vector compared with a 52.4-fold up-regulation in RD cells transfected with pre-miR-605-G. Results are normalized to the negative control and represent mean SEM (n Z 4). *P < 0.05. (C) Impact of the miR-605 SNP on precursor miR-605 levels. No significant difference in pre-miR-605 levels was detected in both pre-miR-605-A and pre-miR-605-G transfected cells. (D) Processing efficiency of pre-miR-605 according to genotype. The ratio of mature miR-605 over precursor miR-605 levels was significantly reduced in the pre-miR-605-G transfected cells versus the pre-miR-605-A group (P < 0.05).
reportedly associated with younger ages of tumor onset. Functionally, these events have been linked to deregulated MDM2 and/or p53 function. It is possible that LFS patients who carry copies of the G-allele produce lesser amounts of miR-605, and thus have higher levels of its target transcripts. As a result of elevated MDM2 levels, these cells would be predicted, as has been shown in some studies, to have an impaired p53 response to DNA damage and other forms of genetic insults (19). Because LFS patients already lack a functional copy of p53, further impairment of the p53 pathway could render them at an even greater risk of developing cancer. Thus, from this work, we add yet more evidence that regulators of the p53 pathway are promising genetic modifiers in LFS. In the future, better risk stratification strategies for LFS patients might enable an improvement in clinical outcomes through earlier screening and more timely therapeutic interventions. In conclusion, our results identified a miR-605 SNP as a potential modifying event in TP53 mutation carriers. Future work should focus on elucidating the functional consequences of the miR-605 SNP on p53 signaling, while also pursuing a more general understanding of miR-605 tumor suppression. Finally, the role of miR-605 in suppressing MDM2 expression, and its association with cancer risk in
LFS, suggests that modulating miR-605 might represent a novel therapeutic strategy for the successful treatment of both TP53 wild-type and mutated tumors.
Acknowledgments We thank Ana Novokmet for careful annotation of the TP53 status and clinical demographic parameters of the individuals in this study, and Harriet Druker and Laura Zahavich for preand post-counseling families for the initial TP53 mutation analysis. Grant support was provided by the Canadian Institutes for Health Research (no. MOP-300105).
Supplementary data Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.cancergen.2014.12.003
References 1. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 2004;116:281e297.
Functional variant in miR-605 affects Li-Fraumeni syndrome
51
2. Friedman RC, Farh KK, Burge CB, et al. Most mammalian mRNAs are conserved targets of microRNAs. Genome Res 2009;19:92e105. 3. Deng S, Calin GA, Croce CM, et al. Mechanisms of microRNA deregulation in human cancer. Cell Cycle 2008;7:2643e2646. 4. Ryan BM, Robles AI, Harris CC. Genetic variation in microRNA networks: the implications for cancer research. Nat Rev Cancer 2010;10:389e402. 5. Olivier M, Goldgar DE, Sodha N, et al. Li-Fraumeni and related syndromes: correlation between tumor type, family structure, and TP53 genotype. Cancer Res 2003;63:6643e6650. R, Baert-Desurmont S, et al. Molecular €e 6. Bougeard G, Sesbou basis of the Li-Fraumeni syndrome: an update from the French LFS families. J Med Genet 2008;45:535e538. 7. Tabori U, Nanda S, Druker H, et al. Younger age of cancer initiation is associated with shorter telomere length in LiFraumeni syndrome. Cancer Res 2007;67:1415e1418. 8. Marcel V, Palmero EI, Falagan-Lotsch P, et al. TP53 PIN3 and MDM2 SNP309 polymorphisms as genetic modifiers in the LiFraumeni syndrome: impact on age at first diagnosis. J Med Genet 2009;46:766e772. 9. Ruijs MW, Schmidt MK, Nevanlinna H, et al. The singlenucleotide polymorphism 309 in the MDM2 gene contributes to the Li-Fraumeni syndrome and related phenotypes. Eur J Hum Genet 2007;15:110e114. 10. Bougeard G, Baert-Desurmont S, Tournier I, et al. Impact of the MDM2 SNP309 and p53 Arg72Pro polymorphism on age of tumour onset in Li-Fraumeni syndrome. J Med Genet 2006;43: 531e533.
11. Shlien A, Tabori U, Marshall CR, et al. Excessive genomic DNA copy number variation in the Li-Fraumeni cancer predisposition syndrome. Proc Natl Acad Sci U S A 2008;105: 11264e11269. 12. Aury-Landas J, Bougeard G, Castel H, et al. Germline copy number variation of genes involved in chromatin remodelling in families suggestive of Li-Fraumeni syndrome with brain tumours. Eur J Hum Genet 2013;21:1369e1376. 13. Hermeking H. MicroRNAs in the p53 network: micromanagement of tumour suppression. Nat Rev Cancer 2012;12:613e626. 14. Xiao J, Lin H, Luo X, et al. miR-605 joins p53 network to form a p53:miR-605:Mdm2 positive feedback loop in response to stress. EMBO J 2011;30:524e532. 15. Zhang MW, Jin MJ, Yu YX, et al. Associations of lifestyle-related factors, hsa-miR-149 and hsa-miR-605 gene polymorphisms with gastrointestinal cancer risk. Mol Carcinog 2012;51(Suppl 1): E21eE31. 16. Hu Y, Yu CY, Wang JL, et al. MicroRNA sequence polymorphisms and the risk of different types of cancer. Sci Rep 2014;4:3648. 17. Dahiya N, Sherman-Baust CA, Wang TL, et al. MicroRNA expression and identification of putative miRNA targets in ovarian cancer. PLoS One 2008;3:e2436. 18. Hofacker IL. Vienna RNA secondary structure server. Nucleic Acids Res 2003;31:3429e3431. 19. Bond GL, Hu W, Bond EE, et al. A single nucleotide polymorphism in the MDM2 promoter attenuates the p53 tumor suppressor pathway and accelerates tumor formation in humans. Cell 2004;119:591e602.
BRIEF COMMUNICATION
A functional variant in miR-605 modifies the age of onset in Li-Fraumeni syndrome a,b a,b,c, * Badr Id Said , David Malkin a
Division of Hematology/Oncology, Department of Pediatrics, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada; b Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; c Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada Li-Fraumeni syndrome (LFS) is a rare cancer predisposition syndrome, typically associated with germline mutations in the TP53 gene. Despite the high penetrance of TP53 mutations, LFS patients display striking phenotypic differences, suggesting the presence of secondary risk loci. To date, all genetic modifiers in LFS have been shown to map to either TP53 or its principal negative regulator, MDM2. Given this strong association, we set out to investigate the contribution of miR605, a recently described microRNA (miRNA) regulator of the p53-MDM2 loop. We hypothesized that, if functional, the miR-605 gene and its variant (rs2043556) could impact the cancer risk profile of TP53 mutation carriers. Consistent with this proposition, the variant G-allele of miR-605 was associated with a 10-year acceleration in the mean age of LFS tumor onset (P Z 0.04) and caused a 2.6-fold reduction in the processing levels of its host miRNA (P < 0.05). We also demonstrate that miR-605 overexpression leads to a decrease in cell proliferation, clonogenicity, and migration in two rhabdomyosaroma cell lines carrying hotspot TP53 mutations. Together, our results implicate miR-605 as a novel modifier gene of the LFS phenotype and a promising therapeutic target in TP53 mutant cancers. Keywords MicroRNA, Li-Fraumeni syndrome, single nucleotide polymorphism, p53 ª 2015 Elsevier Inc. All rights reserved.
MicroRNAs (miRNAs) are approximately 22 nucleotide noncoding RNAs that act as sequence-specific post-transcriptional regulators of gene expression, affecting >60% of human genes (1,2). In cancer, miRNAs are often aberrantly expressed as a result of a variety of mechanisms, including genomic rearrangements, mutations, and DNA copy number and epigenetic changes (3). Recently, common single nucleotide polymorphisms (SNPs) within miRNA genes have emerged as yet another potent mechanism for miRNA deregulation (4). Such functional SNPs have been associated with sporadic and familial cancers in several association studies, yet their potential contribution to Li-Fraumeni syndrome (LFS) remains unexplored. LFS is a highly penetrant autosomal dominant cancer disorder that leads to the development of a variety of earlyonset cancers, including breast cancer, brain tumors, adrenocortical carcinomas, and soft-tissue and bone sarcomas.
Received August 18, 2014; received in revised form December 22, 2014; accepted December 23, 2014. * Corresponding author. E-mail address: [email protected] 2210-7762/$ - see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.cancergen.2014.12.003
Despite compelling evidence linking LFS heterogeneity to functional differences in germline p53 mutant activity, these genotypephenotype correlations do not yet fully account for either the diversity of tumors or the disparate ages of cancer onset typically observed in LFS patients (5,6). Furthermore, functional differences in p53 activity cannot account for intrafamilial variations in LFS, given that all affected members carry the same defective copy of the TP53 gene. Instead, secondary or coexistent genetic factors that modify p53 function have been posited as responsible for an important spectrum of these clinical differences. To date, functional polymorphisms in TP53 (Arg72Pro and PIN3), MDM2 (SNP 309 [a T>G substitution at nucleotide 309 of intron one of MDM2 {rs2279744}]), as well as excess DNA copy number variations (CNVs) and accelerated telomere attrition, have all been associated with the heterogeneous tumor patterns of LFS (7e12). From this work, it has become clear that individual LFS phenotypes are defined not only by the underlying germline TP53 mutation, but also by an important interplay of different modifier genes. The influence of different combinations of these modifiers could account for the wide spectrum of different susceptibilities that characterize LFS. Therefore, developing useful biomarkers
48 for the future clinical monitoring of LFS patients will require a more thorough and complete understanding of all TP53 modifying events. Lately, several miRNAs have emerged as major players in the p53 network (13). One particular member of this family, miR-605, was found to be a p53-inducible miRNA capable of targeting MDM2, a critical cellular inhibitor of p53 (14). Interestingly, a potentially functional polymorphism (rs2043556) within the miR-605 sequence has been previously associated with cancer risk (15,16). Due to its location in the precursor structure of miR-605, this genetic variant could impact the post-transcriptional processing of miR-605. Importantly, this miRNA has been shown to be deregulated in some human tumors. It is also expressed in tissues from which common LFS tumors develop, namely, brain, skeletal muscle, and breast (14,17). Thus, SNP-mediated miR-605 deregulation could modulate LFS cancer risk by causing important transcriptomic changes in tissues typically affected by this disorder. Additional functional features of miR-605, such as its association with p53 signaling and its capacity to regulate multiple genes, render miR-605 and its variant promising candidate modifiers in LFS.
Materials and methods Study subjects The study population consists of 55 Caucasian individuals with a germline TP53 mutation. Written informed consent to participate in the research study was obtained from the individual or his or her parent (if the participant was below the age of consent). The study was approved by the Research Ethics Board at The Hospital for Sick Children (REB no. 0019910602).
miR-605 SNP genotyping The rs2043556 variant was genotyped by SNP-RFLP (restriction fragment length polymorphism). First, a 284-bp DNA fragment that contained pre-miR-605 was amplified from DNA extracted from peripheral blood lymphocytes by PCR: 50 -CGCCTCTTTTTGCTCATTCT-30 and 50 -AGAGCAGTT ACGCCACATGA-30 . In homozygote carriers of the miR-605 A-allele, a HinfI enzyme (New England Biolabs, Ipswich, MA) recognition site was created in the 284-bp PCR fragment, which resulted in two equal 142-bp fragments. In contrast, HinfI-digested PCR fragments from carriers of the A/G genotype produce three fragments: one uncut 284-bp fragment and two equal 142-bp fragments.
Cell lines The RD and Rh30 rhabdomyosarcoma cell lines (both from Dr. A. Thomas Look of Dana-Farber Cancer Institute, Boston, MA) were selected because of their mutant TP53 status (R273 C and R248 W, respectively) as well as the fact that rhabdomyosarcoma is an important component tumor of LFS. These cell lines were confirmed to be mycoplasmafree.
B. Id Said, D. Malkin
miR-605 expression vectors and RD cell nucleofection MiRNA expression vectors (pCMV-MIR) that contained a 639-bp fragment corresponding to pre-miR-605-G and premiR-605-A, along with an empty control, were purchased from OriGene (Rockville, MD). These expression plasmids were transiently transfected into RD cells by nucleofection (Amaxa Inc., Gaithersburg, MD). Total RNA was extracted 6 hours later for quantitative reverse transcription PCR (qRT-PCR).
qRT-PCR of mature and pre-miR-605 To evaluate expression levels of mature and pre-miR-605, total RNA (200 ng) from pre-miR-605-A- and pre-miR-605G-nucleofected cells was reverse transcribed into cDNA using either specific stem-loop RT primers for miR-605 and U6 (Applied Biosystems, Foster City, CA), or random hexamer primers (Bio-Rad Laboratories, Hercules, CA). Subsequently, the PCR levels of these miRNAs was quantified using the corresponding TaqMan miRNA Assays (Applied Biosystems). The amount of mature and pre-miR605 relative to the negative control was calculated using the 2-DDCt method and normalized to U6. Results represent the fold-change over control and are shown as mean SEM; each experiment was conducted at least four times (see Supplementary data).
Statistical analysis Data were analyzed using Prism 3.0 software (GraphPad Software, Inc., San Diego, CA). For functional studies, statistical analyses were performed using the unpaired twotailed Student t-test. The non-parametric Mann-Whitney test was used to compare the age of cancer onset among the different SNP genotypes.
Results The miR-605 SNP modifies the age of cancer onset in LFS patients Since miR-605 is an important regulator of the p53-MDM2 loop, and its variant is predicted by RNAfold to impact the stability of its precursor structure by DDG Z e2.6 kcal/mol (Figure S1), we analyzed the association of this polymorphism with the age of cancer onset in LFS patients who carried germline TP53 mutations (18). Our study sample consists of a total of 55 Caucasians (Table S1), 21 of whom are related to each other through 10 distinct family groupings. The majority of these families (9 of 10) are comprised of pairs, with only 1 family counting more than 2 members (n Z 3). Genotyping of blood-derived DNA samples from these 55 patients was performed using SNP-RFLP. Samples were then divided according to their miR-605 genotype (A/A or A/G), and their age at cancer onset was compared; the rare G/G genotype was not present in our study sample. Interestingly, results revealed that the G-allele was associated with a significantly earlier age of cancer onset
Functional variant in miR-605 affects Li-Fraumeni syndrome
49
(P Z 0.04). The mean and median age of cancer onset for LFS carriers of the A/G genotype (n Z 17) was 12.8 years and 4 years, respectively, compared with 22.9 years and 23.5 years for carriers of the A/A genotype (n Z 38) (Figure 1). Stratifying by tumor type revealed accelerated cancer onset in carriers of the G-allele across all of the main LFS tumor types, such as brain tumors, sarcomas, and adrenocortical carcinomas, with the exception of breast cancer (Table S2). Furthermore, even when we repeated the analysis using samples from unrelated individuals (n Z 34), we observed a similar reduction in the age of cancer onset (P Z 0.1). The mean and median age of cancer onset for carriers of the A/G genotype (n Z 9) was 14.6 years and 4 years, respectively, compared with 23.4 years and 27 years in patients with the common A/A genotype (n Z 25) (Table S3). However, the frequency of the G-allele among patients who developed more than one tumor did not differ from the frequency observed in those with the A/A genotype (Table S5). Further reinforcing our results is the lack of a significant difference in the tumor distribution across the two genotype groups (Figure S2). Thus, despite our relatively low sample size, these data demonstrate, for the first time, an association between a common SNP in a pre-miRNA gene and the age of tumor onset in LFS.
The miR-605 SNP affects the processing efficiency of pre-miR-605 To determine the functional consequences of the miR-605 SNP, miRNA plasmids (pCMV-MIR) containing pre-miR-605G, pre-miR-605-A, and a control vector were nucleofected into the rhabdomyosarcoma cell line (RD); these plasmids were sequence confirmed (Figure 2A). Relative to the control cells, significantly higher levels of mature miR-605 (a 2.6-fold difference) were detected in RD cells transfected with premiR-605-A compared with the pre-miR-605-G-transfected cells (Figure 2B). Interestingly, qRT-PCR measurements of the precursor form of miR-605 (pre-miR-605) in these same samples revealed no such expression difference, indicating that the G-allele of this SNP impacts the efficiency of miR605 processing from its precursor to mature form (Figure 2C). By calculating the expression ratio of mature miR-605 to that of pre-miR-605 in each transfected sample, we determined the processing efficiency of the pre-miR-605G plasmid to be 65% lower than that of the pre-miR-605-A plasmid (Figure 2D). Since miR-605 had been shown in a previous study to exhibit potent tumor suppressive functions in TP53 wild-type cells, but not in isogenic cell lines lacking the TP53 gene (14), it was also important to investigate the tumor suppressive role of miR-605 in the context of TP53 mutations. Overexpression of miR-605 resulted in a significant reduction in cell viability, clonogenicity, and cell migration in two TP53 mutant cell types (Figure S3). Thus, our data confirms that SNP-mediated deregulation of miR-605 is a plausible cancer modifier in TP53 mutant cells.
Discussion
Figure 1 (A) Age of cancer onset in TP53 mutation carriers according to miR-605 genotype. Germline DNA samples from 55 Caucasian TP53 mutation carriers who had written informed consent to participate in the research study were genotyped for the miR-605 variant (rs2043556) using SNP-RFLP. Randomly chosen samples were further genotyped by direct sequencing in order to confirm the RFLP data. Samples were then classified according to their rs2043556 genotype (A/A or A/G), and their age at cancer onset was compared; the low frequency G/G genotype was not detected. Using the non-parametric MannWhitney test, TP53 mutation carriers with the A/G genotype developed cancer at significantly earlier ages compared with individuals with the A/A genotype (P Z 0.04). (B) Percentage of unaffected TP53 mutation carriers for each miR-605 genotype as a function of age of cancer onset.
To our knowledge, this is the first report describing an association between a common SNP in a pre-miRNA gene and the age of tumor onset in LFS. Functionally, the G-allele of miR605 appeared to cause a defect in the processing efficiency of its host miRNA. These findings are consistent with a recent report in which mature miR-605 expression levels were shown to be attenuated in patients with colorectal cancers who carried the G/G genotype (15). Given that the unwinding of premiRNAs is an important miRNA processing step, less stable RNA duplexes, such as the pre-miR-605-A variant (DDG Z 2.6 kcal/mol), would be predicted to be processed faster and more efficiently by Dicer as compared with the more stable pre-miR-605-G variant. This difference can be partly explained by the presence of the stronger G:C pairing in pre-miR-605-G compared with the less stable G:U pairing in pre-miR-605-A. This mechanism could potentially explain the observed overexpression of mature miR-605 from the A-allele as well as the lack of such difference in the precursor levels of miR-605 between these two variants. Thus, our work establishes cytoplasmic miRNA processing by Dicer as a potential mechanism for miR-605 SNP-mediated dysfunction. Because each miRNA is thought to regulate >100 different gene products, SNP-mediated deregulation of miR605 can be predicted to have widespread biological effects. Specifically, transcripts normally regulated by miR-605, including MDM2, could be less efficiently regulated as a result of lower miR-605 expression. In LFS patients who carry a mutant TP53 gene, higher levels of MDM2 have been
50
B. Id Said, D. Malkin
Figure 2 (A) Sanger sequencing of the pre-miR-605-A and pre-miR-605-G vectors. miRNA expression vectors (pCMV-MIR) containing a 639-bp fragment corresponding to pre-miR-605-G and pre-miR-605-A were sequenced in order to confirm the correct DNA sequence. (B) Impact of the miR-605 SNP on mature miR-605 levels. The aforementioned miRNA plasmids were transiently transfected into RD cells by nucleofection. Compared with the empty vector, mature miR-605 levels were up-regulated 137.1-fold in RD cells transfected with pre-miR-605-A vector compared with a 52.4-fold up-regulation in RD cells transfected with pre-miR-605-G. Results are normalized to the negative control and represent mean SEM (n Z 4). *P < 0.05. (C) Impact of the miR-605 SNP on precursor miR-605 levels. No significant difference in pre-miR-605 levels was detected in both pre-miR-605-A and pre-miR-605-G transfected cells. (D) Processing efficiency of pre-miR-605 according to genotype. The ratio of mature miR-605 over precursor miR-605 levels was significantly reduced in the pre-miR-605-G transfected cells versus the pre-miR-605-A group (P < 0.05).
reportedly associated with younger ages of tumor onset. Functionally, these events have been linked to deregulated MDM2 and/or p53 function. It is possible that LFS patients who carry copies of the G-allele produce lesser amounts of miR-605, and thus have higher levels of its target transcripts. As a result of elevated MDM2 levels, these cells would be predicted, as has been shown in some studies, to have an impaired p53 response to DNA damage and other forms of genetic insults (19). Because LFS patients already lack a functional copy of p53, further impairment of the p53 pathway could render them at an even greater risk of developing cancer. Thus, from this work, we add yet more evidence that regulators of the p53 pathway are promising genetic modifiers in LFS. In the future, better risk stratification strategies for LFS patients might enable an improvement in clinical outcomes through earlier screening and more timely therapeutic interventions. In conclusion, our results identified a miR-605 SNP as a potential modifying event in TP53 mutation carriers. Future work should focus on elucidating the functional consequences of the miR-605 SNP on p53 signaling, while also pursuing a more general understanding of miR-605 tumor suppression. Finally, the role of miR-605 in suppressing MDM2 expression, and its association with cancer risk in
LFS, suggests that modulating miR-605 might represent a novel therapeutic strategy for the successful treatment of both TP53 wild-type and mutated tumors.
Acknowledgments We thank Ana Novokmet for careful annotation of the TP53 status and clinical demographic parameters of the individuals in this study, and Harriet Druker and Laura Zahavich for preand post-counseling families for the initial TP53 mutation analysis. Grant support was provided by the Canadian Institutes for Health Research (no. MOP-300105).
Supplementary data Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.cancergen.2014.12.003
References 1. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 2004;116:281e297.
Functional variant in miR-605 affects Li-Fraumeni syndrome
51
2. Friedman RC, Farh KK, Burge CB, et al. Most mammalian mRNAs are conserved targets of microRNAs. Genome Res 2009;19:92e105. 3. Deng S, Calin GA, Croce CM, et al. Mechanisms of microRNA deregulation in human cancer. Cell Cycle 2008;7:2643e2646. 4. Ryan BM, Robles AI, Harris CC. Genetic variation in microRNA networks: the implications for cancer research. Nat Rev Cancer 2010;10:389e402. 5. Olivier M, Goldgar DE, Sodha N, et al. Li-Fraumeni and related syndromes: correlation between tumor type, family structure, and TP53 genotype. Cancer Res 2003;63:6643e6650. R, Baert-Desurmont S, et al. Molecular €e 6. Bougeard G, Sesbou basis of the Li-Fraumeni syndrome: an update from the French LFS families. J Med Genet 2008;45:535e538. 7. Tabori U, Nanda S, Druker H, et al. Younger age of cancer initiation is associated with shorter telomere length in LiFraumeni syndrome. Cancer Res 2007;67:1415e1418. 8. Marcel V, Palmero EI, Falagan-Lotsch P, et al. TP53 PIN3 and MDM2 SNP309 polymorphisms as genetic modifiers in the LiFraumeni syndrome: impact on age at first diagnosis. J Med Genet 2009;46:766e772. 9. Ruijs MW, Schmidt MK, Nevanlinna H, et al. The singlenucleotide polymorphism 309 in the MDM2 gene contributes to the Li-Fraumeni syndrome and related phenotypes. Eur J Hum Genet 2007;15:110e114. 10. Bougeard G, Baert-Desurmont S, Tournier I, et al. Impact of the MDM2 SNP309 and p53 Arg72Pro polymorphism on age of tumour onset in Li-Fraumeni syndrome. J Med Genet 2006;43: 531e533.
11. Shlien A, Tabori U, Marshall CR, et al. Excessive genomic DNA copy number variation in the Li-Fraumeni cancer predisposition syndrome. Proc Natl Acad Sci U S A 2008;105: 11264e11269. 12. Aury-Landas J, Bougeard G, Castel H, et al. Germline copy number variation of genes involved in chromatin remodelling in families suggestive of Li-Fraumeni syndrome with brain tumours. Eur J Hum Genet 2013;21:1369e1376. 13. Hermeking H. MicroRNAs in the p53 network: micromanagement of tumour suppression. Nat Rev Cancer 2012;12:613e626. 14. Xiao J, Lin H, Luo X, et al. miR-605 joins p53 network to form a p53:miR-605:Mdm2 positive feedback loop in response to stress. EMBO J 2011;30:524e532. 15. Zhang MW, Jin MJ, Yu YX, et al. Associations of lifestyle-related factors, hsa-miR-149 and hsa-miR-605 gene polymorphisms with gastrointestinal cancer risk. Mol Carcinog 2012;51(Suppl 1): E21eE31. 16. Hu Y, Yu CY, Wang JL, et al. MicroRNA sequence polymorphisms and the risk of different types of cancer. Sci Rep 2014;4:3648. 17. Dahiya N, Sherman-Baust CA, Wang TL, et al. MicroRNA expression and identification of putative miRNA targets in ovarian cancer. PLoS One 2008;3:e2436. 18. Hofacker IL. Vienna RNA secondary structure server. Nucleic Acids Res 2003;31:3429e3431. 19. Bond GL, Hu W, Bond EE, et al. A single nucleotide polymorphism in the MDM2 promoter attenuates the p53 tumor suppressor pathway and accelerates tumor formation in humans. Cell 2004;119:591e602.