Research ArticleGENETICS

Germline mutation of MDM4, a major p53 regulator, in a familial syndrome of defective telomere maintenance

See allHide authors and affiliations

Science Advances  10 Apr 2020:
Vol. 6, no. 15, eaay3511
DOI: 10.1126/sciadv.aay3511
  • Fig. 1 A germline MDM4 p.T454M mutation identified in a family with bone marrow hypocellularity and short telomeres.

    (A) Pedigree of family NCI-226. Arrow indicates proband. Cancer histories include oral squamous cell carcinoma for 226-8 at age 27 years and for 226-9 at age 42 years, and melanoma at 51 years and AML at 52 years for 226-7 (see table S1 for further details). 226-5 had lung cancer at age 69 years. 226-6 had non-Hodgkin lymphoma at age 91 years. In addition, four siblings of 226-6 had cancer: one with breast, two with lung, and one with ovary or uterus (not specified). Sequencing of 226-5, 226-6, 226-7, and 226-9 was not possible because of lack of available DNA. (B and C) Lymphocyte telomere lengths (TL) of study participants. Total lymphocyte telomere lengths are shown and were measured by flow cytometry with in situ hybridization. (B) Graphical depiction of telomere length in relation to age. Four individuals had telomeres measured twice. Legend is in (C). Percentiles (%ile) are based on 400 healthy individuals (50). (C) Age at measurement(s) and telomere length in kilobases. (D) Sequence of the MDM4 RING domain (residues 436 to 490) with secondary structure residues indicated (black boxes). The P-loop motif is highlighted in gray, and the mutated residue in red. (E) The mutant RING domain retains ATP-binding capacity. Wild-type (WT) and mutant (TM) glutathione S-transferase (GST)–RING proteins, or GST alone, were incubated with 10 nM ATP and 5 μCi ATP-γ32P for 10 min at room temperature, filtered through nitrocellulose, and counted by liquid scintillation CPM, counts per minute. Results from two independent experiments. (F) The mutant MDM4 RING domain has an altered capacity to dimerize with the MDM2 RING. Two-hybrid assays were carried out as described (47). -LW, minus leucine and tryptophan; -LWHA, minus leucine, tryptophan, histidine and adenine; OD, optical density. Growth on the -LWHA medium indicates protein interaction, readily observed between MDM2 (M2-BD) and WT MDM4 (M4-AD WT) but faintly visible between MDM2 and MDM4T454M (M4-AD TM). (G) Impact of the mutation in transfected human cells. U2OS cells were transfected with an empty vector (EV) or an expression plasmid encoding a Myc-tagged MDM4 (WT or T454M) protein and then treated or not with cycloheximide (CHX) to inhibit protein synthesis, and protein extracts were immunoblotted with antibodies against Myc, p21, or actin. Bands were normalized to actin, and a value of 1 was assigned to cells transfected with the WT MDM4 expression plasmid (for Myc) or with the empty vector (for p21).

  • Fig. 2 Generation of a mouse model with a targeted Mdm4T454M mutation.

    (A) Targeting strategy. Homologous recombination in ES cells was used to target the T454M mutation at the Mdm4 locus. For the Mdm4 WT allele, exons 9 to 11 are shown [black boxes, coding sequences; white box, 3′ untranslated region (3′UTR)] and Bam HI (BH) restriction sites. Above, the targeting construct contains the following: (i) a 2.9-kb-long 5′ homology region encompassing exon 10, intron 10, and exon 11 sequences upstream the mutation; (ii) the mutation (asterisk) within exon 11; (iii) a 2.6-kb-long fragment encompassing the 3′ end of the gene and sequences immediately downstream; (iv) a neomycin selection gene (Neo) flanked by loxP sequences (gray arrowheads) and an additional BH site; (v) a 2.1-kb-long 3′ homology region containing sequences downstream Mdm4; and (vi) the Diphtheria toxin a gene (DTA) for targeting enrichment. (B to D) screening of G418-resistant ES clones as described in (A), with asterisks (*) indicating positive recombinants: (B) PCR with primers a and b; (C) sequencing after PCR with primers c and d: the sequence for codons 452 to 456 demonstrates heterozygosity at codon 454; (D) Southern blot of Bam HI–digested DNA with the 5′ (left) or 3′ (right) probe. (E) Examples of fibroblast genotyping by PCR with primers e and f. (F) The Mdm4T454M mutation does not alter Mdm4 mRNA levels. Mdm4-FL (left) and Mdm4-S (right) mRNAs were extracted from WT and Mdm4TM/TM MEFs before or after treatment for 24 hours with 10 μM Nutlin, quantified using real-time PCR, and normalized to control mRNAs, and then the value in Nutlin-treated WT MEFs was assigned a value of 1. Results from five independent experiments and >4 MEFs per genotype. ns, not significant in a Student’s t test. (G) Decreased Mdm4 protein levels in Mdm4TM/TM MEFs. Protein extracts, prepared from MEFs treated as in (F), were immunoblotted with antibodies against Mdm4 or actin. Bands were normalized to actin, and then the values in Nutlin-treated WT cells were assigned a value of 1. p53ΔP/ΔP Mdm4ΔE6/ΔE6 MEFs do not express a full-length Mdm4 protein (20): They were loaded to unambiguously identify the Mdm4(-FL) band in the other lanes.

  • Fig. 3 Mdm4TM/TM fibroblasts exhibit increased p53 activity and short telomeres.

    (A) Quantification of p21 and Mdm2 mRNAs extracted from WT, Mdm4+/TM, and Mdm4TM/TM MEFs, treated or not for 24 hours with 10 μM Nutlin. mRNA levels were quantified using real-time PCR and normalized to control mRNAs, and then the value in Nutlin-treated WT MEFs was assigned a value of 1. Results from 10 independent experiments. (B) Protein extracts, prepared from p53−/−, WT, and Mdm4TM/TM MEFs treated as in (A), were immunoblotted with antibodies against Mdm2, Mdm4, p53, p21, or actin. Bands were normalized to actin, and then the values in Nutlin-treated WT MEFs were assigned a value of 1. (C) Proliferation of MEFs in a 3T3 protocol. Each point is the average value of three independent MEFs. (D) Decreased telomere length in Mdm4TM/TM MEFs, as measured by quantitative FISH with a telomeric probe. Results from two MEFs per genotype, and 68 to 75 metaphases per MEF [means + 95% confidence interval (CI) are shown in yellow]. a.u., arbitrary units. (E) Telomere-related genes down-regulated in Mdm4TM/TM MEFs. mRNAs were extracted from unstressed WT and Mdm4TM//TM MEFs, quantified using real-time PCR, and normalized to control mRNAs, and the value in WT MEFs was assigned a value of 1. Results from >3 independent experiments and two MEFs per genotype. In relevant panels: °P = 0.08, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by Student’s t (A, C at passage 7, and E) or Mann-Whitney (D) statistical tests.

  • Fig. 4 Mdm4TM/TM mice die perinatally but can be rescued by p53 loss or haploinsufficiency.

    (A) Mendelian distribution of the offspring from 8 Mdm4+/TM intercrosses. (B) Mdm4TM/TM mice die at birth. Cohort sizes are in parentheses. (C) Mdm4TM/TM neonates are smaller than their littermates and appear hypoxic. (D) Lungs from Mdm4TM/TM P0 pups are hypoplastic and sink in phosphate-buffered saline owing to a lack of air inflation. (E) Flow-FISH analysis of P0 lung cells with a telomere-specific peptide nucleic acid (PNA) probe. Top: Representative results from a WT, a Mdm4+/TM, a Mdm4TM/TM, and a G3 Terc−/− mouse are shown. Right: Green fluorescence (fluo.) with black histograms for cells without the probe (measuring cellular autofluorescence) and green histograms for cells with the probe. The shift in fluorescence intensity is smaller in Mdm4TM/TM and Terc−/− cells (c or d < a or b), indicating reduced telomere length. Left: Propidium iodide (PI) fluorescence histograms are superposed for cells with or without the probe. Below: Statistical analysis of green fluorescence shifts (see Materials and Methods). Means + 95% CI are shown; data are from two to three mice and >3800 cells per genotype. (F) Impact of decreased p53 activity on Mdm4TM/TM animals. Cohort sizes are in parentheses. (G) Examples of littermates with indicated genotypes. (H) Hind legs of mice with indicated genotypes. (I) Mdm4TM/TM p53+/− mice exhibit abnormal hemograms. Counts for white blood cells (WBC), red blood cells (RBC), and platelets (PLT) for age-matched (120 days old) animals are shown. (J) Hematoxylin and eosin staining of sternum sections from WT and Mdm4TM/TM p53+/− mice. In relevant panels: ns, not significant; *P < 0.05, ***P < 0.001, and ****P < 0.0001 by Mantel-Cox (B and F), Student’s t (C, D, G, and I), or Mann-Whitney (E) statistical tests. Photo credits: E.T. and R.D., Institut Curie (C, G, and H); R.D., Institut Curie (D).

  • Fig. 5 Mdm4+/TM mice are hypersensitive to increased p53 activity.

    (A) Impact of increased p53 activity on Mdm4+/TM animals. Cohort sizes are in parentheses. (B) Footpads from Mdm4+/TM mice appear normal (top) or exhibit a subtle increase in pigmentation (bottom). (C) Mdm4+/TM p53+/Δ31 mice exhibit strong skin hyperpigmentation. (D) Mdm4+/TM p53+/Δ31 mice are smaller than age-matched WT mice. (E to G) Mdm4+/TM p53+/Δ31 mice exhibit heart hypertrophy (E) as well as thymic (F) and testicular (G) hypoplasia. (H) Mdm4+/TM p53+/Δ31 mice exhibit abnormal hemograms. Counts for white blood cells, red blood cells, and platelets for five age-matched (P18) animals per genotype are shown. (I) Hematoxylin and eosin staining of sternum sections from mice of the indicated genotypes. (J) Flow-FISH analysis of P18 bone marrow cells with a telomere-specific PNA probe. Top: Representative results for a WT, a Mdm4+/TM with normal skin pigmentation (nsp), a Mdm4+/TM with increased footpad skin pigmentation (isp), a p53+/Δ31, and a Mdm4+/TM p53+/Δ31 mouse are shown; black histograms, cells without the probe; green histograms, cells with the probe. The smallest shift in fluorescence intensity (e) was observed with Mdm4+/TM p53+/Δ31 cells. Bottom: Statistical analysis of green fluorescence shifts. Means + 95% CI are shown; data are from >1500 cells per genotype. In relevant panels: ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 by Mantel-Cox (A), Student’s t (D and E to H), or Mann-Whitney (J) statistical tests. Photo credits: R.D. and P.L., Institut Curie (B); E.T. and R.D., Institut Curie (C and D).

  • Fig. 6 TP53 or MDM4 polymorphisms might alter the impact of the MDM4T454M mutation among NCI-226 family members.

    (A) Genotyping of polymorphisms that may affect the p53 pathway. The SNPs rs1800371 and rs1042522 modify the p53 protein sequence (28,29), whereas rs17878362 and rs17880560 are singlets (A1) or doublets (A2) of G-rich sequences in noncoding regions of TP53 that affect p53 expression (30). SNPs rs117039649 and rs2279744, in the MDM2 promoter, affect MDM2 mRNA levels (3,31). Three SNPs are at the MDM4 locus: rs4245739 in the 3′UTR region affects MDM4 protein levels (4), whereas rs11801299 and rs1380576 were associated with an increased risk of developing retinoblastoma (32), a cancer type with frequent MDM4 alterations (51). Polymorphisms that differ among family members are in bold, with the allele (or haplotype) associated with increased p53 activity in green (because it may synergize with the effects of the MDM4T454M mutation). Alleles (or haplotypes) for which there is evidence of decreased p53 activity, or for which the effect is uncertain, are highlighted in red or blue, respectively. Please note that the clinical effects of the TP53 rs1042522 SNP have recently been contested (33), so that all alleles for this SNP were labeled in blue. MAF, minor allele frequency reported for all gnomAD populations combined. https://gnomad.broadinstitute.org (52). (B) Comparative analysis of primary fibroblasts from family members 226-4 and 226-8. p21 and RTEL1 mRNAs, extracted from cells from relatives NCI 226-4 and NCI 226-8 or two unrelated patients with DC carrying a TINF2 or a TERT mutation, were quantified using real-time PCR, normalized to control mRNAs, and then expressed relative to the mean values in TINF2 and TERT mutant cells. ns, not significant, **P < 0.01 and ***P < 0.001 in a Student’s t test.

Supplementary Materials

  • Supplementary Materials

    Germline mutation of MDM4, a major p53 regulator, in a familial syndrome of defective telomere maintenance

    Eléonore Toufektchan, Vincent Lejour, Romane Durand, Neelam Giri, Irena Draskovic, Boris Bardot, Pierre Laplante, Sara Jaber, Blanche P. Alter, José-Arturo Londono-Vallejo, Sharon A. Savage, Franck Toledo

    Download Supplement

    This PDF file includes:

    • Figs. S1 to S5
    • Table S1 to S5
    • References

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article