Germline TP53 splicing variants are uncommon, and their clinical relevance is unknown. However, splice-altering variants at exon 4–intron 4 junctions are relatively enriched in pediatric adrenocortical tumors (ACT). Nevertheless, family histories of cancer compatible with classic Li-Fraumeni syndrome are rarely seen in these patients. We used conventional and in silico assays to determine protein stability, splicing, and transcriptional activity of 10 TP53 variants at exon 4–intron 4 junctions and analyzed their clinical correlates. We reviewed public databases that report the impact of TP53 variants in human cancer and examined individual reports, focusing on family history of cancer. TP53 exon 4–intron 4 junction germline variants were identified in 9 of 75 pediatric ACTs enrolled in the International Pediatric Adrenocortical Tumor Registry and Children's Oncology Group ARAR0332 study. An additional eight independent TP53 variants involving exon 4 splicing were identified in the Pediatric Cancer Genome Project (n = 5,213). These variants resulted in improper expression due to ineffective splicing, protein instability, altered subcellular localization, and loss of function. Clinical case review of carriers of TP53 exon 4–intron 4 junction variants revealed a high incidence of pediatric ACTs and atypical tumor types not consistent with classic Li-Fraumeni syndrome. Germline variants involving TP53 exon 4–intron 4 junctions are frequent in ACT and rare in other pediatric tumors. The collective impact of these germline TP53 variants on the fidelity of splicing, protein structure, and function must be considered in evaluating cancer susceptibility.

Implications:

Taken together, the data indicate that splice variants at TP53 codon 125 and surrounding bases differentially impacted p53 gene expression and function.

Large-scale cancer genomic studies have focused on variants leading to splice-site disruption causing exon skipping, but often overlook variants having splice-creating potential (1). Mutations in the TP53 sequence causing exon skipping are significant but generally underreported. Most are incorrectly considered neutral or silent, thereby misleading their true impact on phenotypic features of individuals harboring these variants.

Pediatric adrenocortical tumor (ACT) is an uncommon malignancy and alterations in the p53 tumor suppressor signaling pathway are key to its tumorigenesis (2, 3). Approximately 50% of pediatric ACTs are associated with germline TP53 variants (2–4) and many retain significant wild-type activity (4). Furthermore, cancer histories for probands with ACT and germline TP53 variants and their carrier relatives are less pronounced than for classic families with Li-Fraumeni syndrome (LFS; refs. 4–6). Detailed analysis of TP53 mutations in pediatric ACT can detect variants whose biologic effect might obscure their pathogenicity and clinical significance.

TP53 genotyping of pediatric ACTs from the International Pediatric Adrenocortical Tumor Registry (IPACTR; ClinicalTrials.gov, ID number NCT00700414) and Children's Oncology Group (COG) ARAR0332 protocol (ClinicalTrials.gov, ID number NCT00304070) showed disruptive germline variants at exon 4 with a predominance of cases harboring the synonymous TP53 p.T125T (c.375G>A) variant. Given threonine codon degeneracy (ACU, ACC, ACA, ACG) and enrichment of missense variants at this codon and splice altering variants at adjacent intron 4 donor and acceptor sites (hereafter referred to as TP53 exon 4–intron 4 junctions) in our pediatric ACT cohorts, we investigated their functional impact on splice fidelity, protein stability and function, and influence on cancer susceptibility.

Case ascertainment

The IPACTR and COG study ARAR0332 analyzed occurrence and features of TP53 mutations in pediatric ACTs. Nine ACT patients harboring germline TP53 exon 4–intron 4 junction variants were eligible for this study. As part of the Pediatric Cancer Genome Project (PCGP), we identified 8 additional independent patients with germline or somatic mutations at these positions excluding 2 ACTs, which were also reported in IPACTR. Also, patients with cancer referred to Hospital of the University of Pennsylvania (Penn; Philadelphia, PA; n = 5), Memorial Sloan Kettering Cancer Center (MSKCC, New York, NY; n = 2), Children's Hospital of Philadelphia (CHOP, Philadelphia, PA; n = 1), and A.C. Camargo Cancer Center (ACCamargo, Sao Paulo, Brazil; n = 1) whose blood samples underwent TP53 genetic testing and germline variants at codon 125 were detected were included herein (Table 1). Age at diagnosis of primary neoplasm and family histories of cancer and pedigrees were ascertained (Supplementary Table S1A; Supplementary Fig. S1). This study was performed in compliance with the Declaration of Helsinki and was approved by the Institutional Research Ethics Board at each collaborating site and St. Jude Children's Research Hospital Institutional Review Board. All participants or caregivers provided written informed consent.

Table 1.

Cohort of patients with germline or somatic variants at TP53 codon 125 and surrounding bases included in this study.

SourceSexTestingProteincDNAPrimary neoplasm (age at diagnosis, years)
IPACTR WGS (G) p.T125T c.375G>A Adrenocortical tumor (8) 
COG Single gene (G) p.T125T c.375G>A Adrenocortical tumor (2) 
COG Single gene (G) p.T125T c.375G>A Adrenocortical tumor (1) 
COG Single gene (G) p.T125T c.375G>A Adrenocortical tumor (7) 
IPACTR Single gene (G) p.T125M c.374C>T Adrenocortical tumor (2) 
IPACTR Single gene (G) p.T125M c.374C>T Adrenocortical tumor (1) 
IPACTR Single gene (G) p.T125M c.374C>T Adrenocortical tumor (5) 
IPACTR Single gene (G) p.?? c.375+1G>A Adrenocortical tumor (1) 
IPACTR Single gene (G) p.?? c.376−1 G>A Adrenocortical tumor (2) 
10 PCGP WES (D) p.T125T c.375G>T High-grade glioma (16) 
11 PCGP WES (D) p.T125T c.375G>A High-grade glioma (9) 
12 PCGP NA WGS (D) p.T125T c.375G>A Wilms tumor (NA) 
13 PCGP WGS (R) p.T125T c.375G>T B-cell acute lymphoblastic leukemia (4) 
14 PCGP WGS (G) p.T125R c.374C>G Medulloblastoma (6) 
15 PCGP WGS (G) p.T125R c.374C>G B-cell acute lymphoblastic leukemia (16) 
16 PCGP WES (D) p.T125R c.374C>G High-grade glioma (8) 
17 PCGP WGS (R) p.T125M c.374C>T High-grade glioma (16) 
18 CHOP Single gene (G) p.T125T c.375G>T High-grade glioma (6) 
19 Penn Panel (G) p.T125T c.375G>T Breast cancer (23) 
20 Penn Panel (G) p.T125T c.375G>A Breast cancer (38) 
21 Penn Panel (G) p.T125T c.375G>A None (32) 
22 Penn Panel (G) p.T125M c.374C>T Breast cancer (32) 
23 Penn Panel (G) p.T125M c.374C>T Breast cancer (31) 
24 ACCamargo Single gene (G) p.T125T c.375G>A Breast cancer (44) 
25 MSKCC Panel (G) p.T125M c.374C>T Urothelial (55) 
26 MSKCC Panel (G) p.T125M c.374C>T Breast cancer (49) 
SourceSexTestingProteincDNAPrimary neoplasm (age at diagnosis, years)
IPACTR WGS (G) p.T125T c.375G>A Adrenocortical tumor (8) 
COG Single gene (G) p.T125T c.375G>A Adrenocortical tumor (2) 
COG Single gene (G) p.T125T c.375G>A Adrenocortical tumor (1) 
COG Single gene (G) p.T125T c.375G>A Adrenocortical tumor (7) 
IPACTR Single gene (G) p.T125M c.374C>T Adrenocortical tumor (2) 
IPACTR Single gene (G) p.T125M c.374C>T Adrenocortical tumor (1) 
IPACTR Single gene (G) p.T125M c.374C>T Adrenocortical tumor (5) 
IPACTR Single gene (G) p.?? c.375+1G>A Adrenocortical tumor (1) 
IPACTR Single gene (G) p.?? c.376−1 G>A Adrenocortical tumor (2) 
10 PCGP WES (D) p.T125T c.375G>T High-grade glioma (16) 
11 PCGP WES (D) p.T125T c.375G>A High-grade glioma (9) 
12 PCGP NA WGS (D) p.T125T c.375G>A Wilms tumor (NA) 
13 PCGP WGS (R) p.T125T c.375G>T B-cell acute lymphoblastic leukemia (4) 
14 PCGP WGS (G) p.T125R c.374C>G Medulloblastoma (6) 
15 PCGP WGS (G) p.T125R c.374C>G B-cell acute lymphoblastic leukemia (16) 
16 PCGP WES (D) p.T125R c.374C>G High-grade glioma (8) 
17 PCGP WGS (R) p.T125M c.374C>T High-grade glioma (16) 
18 CHOP Single gene (G) p.T125T c.375G>T High-grade glioma (6) 
19 Penn Panel (G) p.T125T c.375G>T Breast cancer (23) 
20 Penn Panel (G) p.T125T c.375G>A Breast cancer (38) 
21 Penn Panel (G) p.T125T c.375G>A None (32) 
22 Penn Panel (G) p.T125M c.374C>T Breast cancer (32) 
23 Penn Panel (G) p.T125M c.374C>T Breast cancer (31) 
24 ACCamargo Single gene (G) p.T125T c.375G>A Breast cancer (44) 
25 MSKCC Panel (G) p.T125M c.374C>T Urothelial (55) 
26 MSKCC Panel (G) p.T125M c.374C>T Breast cancer (49) 

Abbreviations: D, diagnosis; F, female; G, germline; M, male; NA, not available; WES, whole-exome sequencing; WGS, whole-genome sequencing.

Mutation screening and haplotype analysis

Genomic DNA was isolated from peripheral blood and tumor tissue obtained from children with ACTs. Complete TP53 coding region and flanking intronic sequences for each exon were determined (3, 7). For remaining cases, TP53 variants were identified by whole-genome or whole-exome sequencing or by multigene panel or single gene testing (Table 1; Supplementary Table S1A). To exclude the possibility of a founder effect for common variants, haplotype analysis was performed using polymorphic markers (VNTRp53 and p53CA; refs. 7–9). To verify whether a germline variant was inherited or de novo, genetic testing (targeted sequencing) was done for additional family members in eight index cases (Supplementary Table S1A).

Database query, annotation, and interpretation of TP53 variants

Population database examination (Supplementary Table S1B), appropriate functional annotations, and interpretation for pathogenicity classification (Supplementary Table S1C) are described in Supplementary Materials and Methods.

Stability analysis and functional in vitro assays

Functional assays to determine p53 status (Supplementary Fig. S2), stability, activity, expression, and localization of mutant proteins are described in Supplementary Materials and Methods.

In silico prediction and detection of TP53 transcriptional variants

To predict effects of TP53 exon 4–intron 4 junctions variants on splicing, we used Alamut [(Interactive Biosoftware, version 2.7), which includes splicing prediction programs SpliceSiteFinder-like, MaxEntScan, NNSPLICE, GeneSplicer, and Human Splicing Finder; RRID:SCR_005181]. We also compiled predicted or observed splicing effects on full-length transcript from the International Agency for Research on Cancer (IARC) database (10), and scores from SpliceAI (RRID:SCR_002082; ref. 11) and dbscSNV database (Supplementary Table S1D; ref. 12).

Total RNA (2 μg) extracted from H1299 cells transfected with TP53 wild-type and variant expression constructs was DNase-treated and reverse transcribed using oligo(dT) primers and Reliance Select cDNA Synthesis kit (Bio-Rad) per manufacturer's instructions. Specific primers were designed to bind to exons 2, 6, and 11 (Supplementary Table S1E). PCR products were sequenced by Applied Biosystems 3730 xl DNA Analyzer (Thermo Fisher Scientific).

Data availability statement

The data generated in this study are available within the article and its Supplementary Data files.

Cohort selection

Pediatric patients with ACT in IPACTR (n = 54) and COG ARAR0332 (n = 21) studies identified excess cases harboring germline TP53 exon 4–intron 4 junctions variants represented by p.T125T (c.375G>A; n = 4), p.T125M (c.374C>T; n = 3), c.375+1G>A, and c.376−1G>A (Table 1). Children with germline p.T125R (c.374C>G) [n = 2; medulloblastoma and B-cell acute lymphoblastic leukemia (ALL)] and somatic variants at p.T125T (c.375G>A) [n = 2; high-grade glioma (HGG) and Wilms tumor], p.T125T (c.375G>T; n = 2; HGG and B-ALL), p.T125R (c.374C>G; n = 1; HGG), and p.T125M (c.374C>T; n = 1 HGG) were ascertained from PCGP (ref. 13; Table 1). Query of cancer registries at Penn, CHOP, MSKCC, and ACCamargo for germline codon 125 and surrounding splice sites TP53 variants identified an additional nine cases (8 adults, 1 child; Table 1; Supplementary Table S1A).

Family histories of cancer

For cases harboring germline variants (n = 20), family histories of cancer were limited because of adoption (Patient #2) or de novo mutations [Patients #8 (c.375+1G>A) and #20 (p.T125T)]. Also, 1 patient (#9) was mosaic for the c.376−1G>A variant (Supplementary Table S1A). Of 11 families with family history available, only the proband's father in family #19 met classic LFS criteria whereas families #1, 6, and 24 met Chompret criteria (14). Interestingly, six families (#18, 21, 22, 23, 25, 26) who did not fulfill classic LFS or Chompret criteria were identified because of early-onset breast cancer (#22,23) or as patients with cancer by germline and tumor panel–based genetic testing (#14, 18, 25, 26). The unaffected proband (#21) was tested as recommended by her primary physician as her deceased grandmother had ovarian cancer. In addition, our cohort included multiple individuals with tumors not considered part of the classic LFS spectrum (e.g., retinoblastoma, desmoid tumors, pheochromocytoma, and urothelial carcinoma; Supplementary Fig. S1).

Haplotype analysis of TP53 codon 125 variants

We performed segregation analysis of microsatellite markers for the p.T125T variants and observed a different haplotype in each case, excluding a common origin for these alleles (Table 1; Supplementary Table S1A). Also, a de novo p.T125T (c.375G>A) variant was documented in a 38-year-old woman with breast cancer (# Patient 20).

Survey of TP53 codon 125 and surrounding bases in public databases

A search of public databases extended the group of TP53 exon 4–intron 4 junctions variants to include p.T125A, p.T125K, p.T125M, p.T125P, p.T125R, p.T125T (c.375G>A, c.375G>C, and c.375 G>T) and two other variants at donor and acceptor splice-site regions of intron 4 (c.375+1G>A and c.376−1G>A; Fig. 1). Variants p.T125M, p.T125R, p.T125T (c.375G>A), c.375+1G>A, and c.376−1G>A occur in the general population (noncancer individuals) at relatively low frequency (Supplementary Table S1B). Except for p.T125A, all these variants have been reported as both somatic and of germline origin in patients with cancer. (Supplementary Table S1B).

Figure 1.

TP53 variants at codon 125 and intron 4 boundaries included in this study cohort. Consensus splice donor and acceptor sites are represented by letters in red. Each variant is identified at its location. Protein domains are schematically represented. Colored circles represent proband tumor type (n = 26).

Figure 1.

TP53 variants at codon 125 and intron 4 boundaries included in this study cohort. Consensus splice donor and acceptor sites are represented by letters in red. Each variant is identified at its location. Protein domains are schematically represented. Colored circles represent proband tumor type (n = 26).

Close modal

Survey of splice TP53 variants in patients with cancer

The IARC database (10) has recorded 401 (from 119 families) germline TP53 variants with altered splicing efficiency (Supplementary Fig. S3A). Of those, 121 (30%; from 47 families) are localized at exon 4–intron 4 junctions, with 56% represented by variant p.T125T (c.375G>A; Supplementary Fig. S3A). The PCGP reported 432 germline and somatic TP53 variants (n = 5,213 patients, 32 different diagnoses; ref. 13). Of those, 19 (3.7%) were predicted to disrupt splicing and 11 (58%, six different diagnoses) occurred at exon 4–intron 4 junctions (Supplementary Fig. S3B), including an osteosarcoma sample that had a fusion transcript [first 125 residues of TP53 fused to position 12,427,735 (GRCh37/hg19) in the short arm of chromosome 17] (13). Considering all germline TP53 variants identified in pediatric ACTs from IPACTR and COG studies, excluding the founder p.R337H variant (15), nucleotide variations of TP53 exon 4–intron 4 junctions were also enriched in this pediatric cancer group (Supplementary Fig. S3C).

In silico features and functional impact of splice-site variants

The interpretation of pathogenicity for p53 variants is complex and relies on frequency, in silico tools and functional in vitro studies. The clinical significance of TP53 exon 4–intron 4 junction variants have not been formally reviewed. However, ClinVar classifies each of these variants as pathogenic (P) or likely pathogenic (LP) except for p.T125P, which was classified as a variant of uncertain significance, and p.T125A that had no ClinVar entry (Supplementary Table S1C).

Variants p.T125A, p.T125K, p.T125M, p.T125P, and p.T125R were classified as potentially damaging to protein function by in silico prediction tools SIFT, Polyphen2, and FATHMM. In contrast, p.T125T (c.375G>A, c.375G>C, and c.375G>T) variants were predicted to be tolerated by SIFT (Supplementary Table S1C). Align-GVGD predicted that the missense variants were damaging, with p.T125P least disruptive, and confirmed by optimized pMSA/Align-GVGD (16).

A yeast-based functional assay (17) determined that missense and silent variants included herein retain residual transcriptional activity ranging from 8.3% (p.T125P) to 36.8% (p.T125A) compared with wild-type p53, and each was considered nonfunctional (Supplementary Table S1C).

In vitro human cell line–based assays and calculated relative fitness scores revealed compromised antiproliferative functionality for p.T125T, but not p.T125A, p.T125K, p.T125M, p.T125P, and p.T125R, with better performance for p.T125A and p.T125M (18). On the basis of the probability of acquiring each TP53 variant and the fitness advantage conferred by loss of function or dominant negative effect (19) as predicted by PHANTM (http://mutantp53.broadinstitute.org/), a high combined phenotype score, consistent with a pathogenic effect, was determined for p.T125P and p.T125R and was lowest for p.T125M (Supplementary Table S1C). For comparison, we included the results for the pathogenic p.R175H and the hypomorphic p.R337H variants in each of the in silico and functional assays (Supplementary Table S1C).

In vitro structural stability

The T125 amino acid position is located in the folded DNA-binding domain (DBD) of p53 and the hydroxyl group of the T125 sidechain, facilitates several favorable contacts including hydrogen bonding interactions with the backbone carbonyl of G117 and the guanidino group of R282 (Fig. 2). Thus, substitution of other amino acids at this position may destabilize the DBD fold. To test whether amino acid substitutions at T125 destabilize the DBD fold, we expressed and purified variants and determined their melting temperatures by measuring circular dichroism (CD) as a function of temperature. As hypothesized, given energetically favorable interactions made by the side chain of T125 with other residues, substitution at this position with different amino acids resulted in DBD destabilization, which likely interferes with the physiologic function of p53 (Fig. 2).

Figure 2.

Stability analysis. A, The sidechain of threonine 125 (shown in magenta) makes hydrogen bonds with the backbone carbonyl of glycine 117 and the sidechain of arginine 282, depicted as yellow dashed lines. Mutations at this residue destabilize the protein, either through steric clashes or the inability to form these favorable contacts. Carbon, nitrogen, and oxygen atoms are shown in gray, blue, and red, respectively. The coordinated zinc atom is shown as a gray sphere. The phosphate backbone of DNA is shown in orange with the bases depicted as sticks. PDB code: 1tup. B, CD melting curves of various TP53 T125 mutants show that all mutations at T125 position are destabilizing.

Figure 2.

Stability analysis. A, The sidechain of threonine 125 (shown in magenta) makes hydrogen bonds with the backbone carbonyl of glycine 117 and the sidechain of arginine 282, depicted as yellow dashed lines. Mutations at this residue destabilize the protein, either through steric clashes or the inability to form these favorable contacts. Carbon, nitrogen, and oxygen atoms are shown in gray, blue, and red, respectively. The coordinated zinc atom is shown as a gray sphere. The phosphate backbone of DNA is shown in orange with the bases depicted as sticks. PDB code: 1tup. B, CD melting curves of various TP53 T125 mutants show that all mutations at T125 position are destabilizing.

Close modal

Variant effects on splicing

To determine whether variants caused missense coding and/or erroneous splicing, we analyzed boundaries of exon 4/intron 4/exon 5 using in silico tools integrating multiple prediction algorithms for detecting splicing signals (Supplementary Table S1D). Prediction algorithms performed accurately for mutations at GT-AT sites for the c.375+1G>A and c.376−1G>A variants (Supplementary Table S1D) and revealed that substitution of nucleotide “G” for “A,” “C,” or “T” at the last base of exon 4 (Fig. 1) results in loss of the authentic 5′ splice site in favor of other cryptic splice sites (Supplementary Table S1D). Notably, in silico prediction of altered splicing for variant p.T125T (c.375G>A) was supported by experimental evidence demonstrating abnormally spliced products and loss of p53 expression in ACT from Patient #1 (Fig. 3).

Figure 3.

Experimental evidence of altered splicing for the variant p.T125T (c.375G>A) in a pediatric adrenocortical tumor (#Patient 1). A, Chromatogram of the p.T125T (c.375G>A) in tumor DNA. B, Schematic representation of the normal splicing pattern of p53 transcript (purple arrow). RNA sequencing of tumor DNA showing retention of intron 4 due to loss of donor splicing (green arrow). Sequence analysis of TP53 cDNA from the same tumor showing the use of cryptic donor site in exon 4 (orange arrow). C, WB analysis of patient #1 ACT (p.T125T) and pediatric ACTs with p.G266E and p.R337H variants. D, Loss of p53 expression (right) in patient #1 ACT (H&E, left) harboring the p.T125T (c.375G>A) variant.

Figure 3.

Experimental evidence of altered splicing for the variant p.T125T (c.375G>A) in a pediatric adrenocortical tumor (#Patient 1). A, Chromatogram of the p.T125T (c.375G>A) in tumor DNA. B, Schematic representation of the normal splicing pattern of p53 transcript (purple arrow). RNA sequencing of tumor DNA showing retention of intron 4 due to loss of donor splicing (green arrow). Sequence analysis of TP53 cDNA from the same tumor showing the use of cryptic donor site in exon 4 (orange arrow). C, WB analysis of patient #1 ACT (p.T125T) and pediatric ACTs with p.G266E and p.R337H variants. D, Loss of p53 expression (right) in patient #1 ACT (H&E, left) harboring the p.T125T (c.375G>A) variant.

Close modal

The Alamut algorithm detected wild-type splice sites in reference sequences for p.T125A, p.T125K, p.T125M, p.T125P, and p.T125R. However, splice site predictions by the neural network NNSPLICE identified a defective donor site for p.T125M, p.T125P, and p.T125R variants (Supplementary Table S1D). Splice prediction using the neural network SpliceAI revealed high scores, indicating splice-altering consequences, for variants p.T125T (G>A, G>C, and G>T), c.375+1G>A and c.376−1G>A (Supplementary Table S1D). Using ANNOVAR with the “dbscSNV” database, variants p.T125M, p.T125R, p.T125T (G>A, G>C, and G>T), c.375+1G>A and c.376−1G>A were predicted to have a potential impact on splicing (Supplementary Table S1D).

To assess potential effects of variants on splicing in vivo, H1299 cells were transfected with p53 expression plasmids containing full-length human wild-type p53 cDNA containing introns 2–4 (20) and each corresponding variant (Supplementary Fig. S4). Canonical p53 sequence was exclusively observed in cells transfected with wild-type and p.T125K. Using a cryptic splice donor site in exon 4 (transcript deletes last 200 bp of exon 4: aa59–125) or total skipping of exon 4 was verified for p.T125T (G>A, G>C, G>T) and for the c.375+1G>A variant. Moreover, these two abnormal transcripts also occurred in cells transfected with p.T125A, p.T125M, p.T125P, and p.T125R. Variant c.376−1 G>A is also predicted to generate a cryptic splice acceptor in exon 5, with loss of seven residues while otherwise maintaining the reading frame (Fig. 4; ref. 21). These findings demonstrate that T125-associated variants can alter splicing in vivo and potentially disrupt p53 function.

Figure 4.

cDNA transcripts from transfected H1299 cells. Agarose gel products (top) and representative canonical and cryptic splice use (bottom) for the TP53 variants included in this study.

Figure 4.

cDNA transcripts from transfected H1299 cells. Agarose gel products (top) and representative canonical and cryptic splice use (bottom) for the TP53 variants included in this study.

Close modal

Transcriptional transactivation activity, expression, and subcellular location

To study the effects of TP53 exon 4–intron 4 junction variants on p53 transactivation function, we transiently transfected Saos-2 and H1299 cells with promoter-luciferase reporters with wild-type TP53 and each variant (Supplementary Fig. S4). Wild-type p53, but not the DBD-mutant R175H, strongly induced the promoter containing intact p53-binding consensus sites (PG13), yielding increased luciferase expression. The transactivation function of variants T125A and T125M was attenuated. All remaining variants were markedly impaired compared with wild-type p53 (Fig. 5). Western blot (WB) analysis detected full-length p53 for each missense variant, whereas the T125T variants [(c.375G>A, c.375G>C, and c.375 G>T) and c.375+1G>A and c.376−1G>A)] were expressed as a shorter form (apparent molecular weight 25 kDa), consistent with altered splicing. The short protein band (25 kDa) was also expressed by the p.T125A, p.T125M, and p.T125P variants (Fig. 5).

Figure 5.

Transcriptional activity of TP53 variants. A, Attenuated p53 luciferase activity in T125A and T125M but compromised in other variants in Saos-2 and H1299 transfected cells compared with that in the hotspot DBD-mutant R175H. B, Corresponding p53 protein expression was determined by WB analysis. Data represent two independent experiments, with each experiment including three biologic replicates. Asterisks indicate statistical significance, as determined by unpaired t test. ****, P < 0.0001 and **, P = 0.0086.

Figure 5.

Transcriptional activity of TP53 variants. A, Attenuated p53 luciferase activity in T125A and T125M but compromised in other variants in Saos-2 and H1299 transfected cells compared with that in the hotspot DBD-mutant R175H. B, Corresponding p53 protein expression was determined by WB analysis. Data represent two independent experiments, with each experiment including three biologic replicates. Asterisks indicate statistical significance, as determined by unpaired t test. ****, P < 0.0001 and **, P = 0.0086.

Close modal

Subcellular distribution of these variants in the nucleus and cytoplasm was investigated by cell fractionation and WB. Variants p.T125A, p.T125K, p.T125M, p.T125P, and p.T125R were preferentially located in the nucleus, as expected for full-length p53. However, variants p.T125T (c.375G>A, c.375G>C, and c.375 G>T), and c.375+1G>A were preferentially localized in the cytoplasm, indicating loss of an intact C-terminus and nuclear localization signals due to aberrant splicing (Fig. 6; ref. 22). Supporting these cell-based assays, p53 immunostaining of primary tumor samples harboring the same variants had a similar subcellular localization pattern (Fig. 3; Supplementary Fig. S5).

Figure 6.

Subcellular localizations of TP53 variants. WB analysis of p53, PARP, (nuclear marker), and HSP90 (cytoplasmic marker) after separation of nuclear (Nuc) and cytoplasmic (Cyt) protein fractions. Negative control was CMV only. Wild-type TP53 variant constructs are shown at the top. Peptide sizes (in kDa) are shown to the left.

Figure 6.

Subcellular localizations of TP53 variants. WB analysis of p53, PARP, (nuclear marker), and HSP90 (cytoplasmic marker) after separation of nuclear (Nuc) and cytoplasmic (Cyt) protein fractions. Negative control was CMV only. Wild-type TP53 variant constructs are shown at the top. Peptide sizes (in kDa) are shown to the left.

Close modal

We report an excess of missense and synonymous variants at the 3′ end of TP53 exon 4 and adjacent intron 4 donor and acceptor splice site motifs in pediatric ACTs (2–6, 23, 24). In 5213 newly diagnosed pediatric malignancies reported to PCGP (13), excluding ACTs, there were two cases of germline variants in this TP53 region versus nine of 75 (12%, P < 0.00001) consecutive ACT cases in IPACTR or COG ARAR0332 studies. Enrichment of exon 4–intron 4 junction variants in pediatric ACTs, rarely observed in other pediatric/adult cancer types (25–28), suggest a tissue-selective role for these variants in tumorigenesis (19). Usually, the frequency of germline TP53 variants in pediatric ACT is highest among human tumor subtypes. In pediatric ACT cases unselected for family history of cancer or histology, 50% had germline TP53 variants versus 1% to 4% for other pediatric tumor subtypes, except for pediatric choroid plexus tumors (29). This suggests that the developing adrenal cortex is prone to p53-driven tumorigenesis in general, and exon 4 and surrounding splice-sites motifs, in particular, is a “hot spot” for TP53 mutagenesis.

We showed the p.T125T (c.375G>A) variant to be both inherited and de novo. Two probands (one ACT, one breast cancer) had a distinct de novo variant and one child carried a mosaic TP53 c.376−1G>A variant (10% of mutated alleles in blood mononuclear cells), which generated a non-consensus splice acceptor site, resulting in an abnormal in-frame deletion transcript (p.Y126_K132del). The frequency of de novo mutations (25%) was almost twice that in a larger series (30, 31) but might be biased given our study ascertainment criteria. Nonetheless, de novo and mosaic TP53 variants are relatively common and can indicate the first germline TP53 mutation in a family. Therefore, TP53 testing is strongly recommended for children with ACT or individuals with early onset breast cancer independent of family histories of cancer (32, 33).

The biological impact of TP53 exon 4–intron 4 junction variants was analyzed by in silico predictions, structural analyses, and functional assays in yeast and human cancer cell lines (16–19). Our studies demonstrated compromised stability, structure, and expression patterns of these variants. Using in silico tools and cell-based splicing assays, we demonstrated that except for p.T125K, all variants analyzed have compromised splicing. A short form of p53 consistent with splice alterations was observed for variants p.T125T (c.375G>A, c.375G>C, and c.375 G>T) and two other variants at the corresponding splice site (c.375+1G>A, c.376−1G>A). In addition to canonical full-length cDNA, aberrant p53 transcripts were observed in cells transfected with p.T125A, p.T125M, p.T125P, and p.T125R, which are predicted to compromise p53 activity. The combination of defects in the missense DBD variants on splicing and protein structure and function could modulate its pathogenicity and consequently its cancer profile in carriers.

Variant p.T125T (c.375G>A) can have TP53 activity and segregate in a family presenting with an atypical spectrum of tumors at relatively late onset (33). Other cancer-prone families harboring p.T125T are reported, and family history of cancer was not typical of classic LFS, including affected relatives in the segregated lineage exhibiting late onset of cancers (34–37).

The p.T125A and p.T125M variants performed near normal in transactivation assays (17) and dominant negative and growth suppression studies (18). The p.T125A is not reported as a somatic or germline variant in cancer cases or our ACT cases, suggesting no clinical relevance. ClinVar does not provide a definitive classification for the T125M variant. The p.T125M families included in this study did not fit classic clinical criteria for LFS (14) and affected members exhibited a wide range of tumor types, most not included in LFS core cancers.

In conclusion, we reveal an association between disruptive TP53 exon 4 sequence variants and pediatric ACTs, and their effect on splicing and protein structure and function. Furthermore, germline alterations at exon 4–intron 4 junctions represent a relative hotspot for mutations in human cancer representing approximately 30% of all reported splice variants in the IARC database. Although these variants are associated with cancer predisposition, penetrance is incomplete with atypical tumor types and late onset. The effect of exon 4 splice altering variants on the expression of p53 isoforms, which have been reported to play a role in human cancers, could also impact tumor phenotype (38). Additional variables influencing cancer susceptibility due to exon 4–intron 4 junction variants such as tissue-specific context, genetic modifiers (9, 39), and environmental factors must also be considered.

K. Breen reports an immediate family member is on the scientific advisory board of Emendo Biotherapeutics, Karyopharm Therapeutics, Imago BioSciences, and DarwinHealth; is co-founder of Isabl Technologies; and has equity interest in Imago BioSciences, Emendo Biotherapeutics, and Isabl Technologies. No disclosures were reported by the other authors.

E.M. Pinto: Conceptualization, resources, data curation, formal analysis, supervision, validation, investigation, visualization, methodology, writing–original draft, writing–review and editing. K.N. Maxwell: Resources, formal analysis, writing–review and editing. H. Halalsheh: Writing–review and editing. A. Phillips: Data curation, writing–review and editing. J. Powers: Resources, data curation, writing–review and editing. S. MacFarland: Resources, data curation, writing–review and editing. M.F. Walsh: Resources, writing–review and editing. K. Breen: Resources, data curation, writing–review and editing. M.N. Formiga: Resources, writing–review and editing. R. Kriwacki: Data curation, writing–review and editing. K.E. Nichols: Data curation, writing–review and editing. R. Mostafavi: Data curation, writing–review and editing. J. Wang: Data curation, writing–review and editing. M.R. Clay: Data curation, writing–review and editing. C. Rodriguez-Galindo: Resources, writing–review and editing. R.C. Ribeiro: Resources, data curation, writing–review and editing. G.P. Zambetti: Conceptualization, resources, formal analysis, funding acquisition, writing–review and editing.

We thank the patients and their family members for indirect participation in this study. We also thank Vani Shanker for editing the article.

This work was supported by the American Lebanese Syrian Associated Charities (ALSAC), Speer Charitable Trust, and Cancer Center Support Grant CA21765. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Jung
H
,
Lee
D
,
Lee
J
,
Park
D
,
Kim
YJ
,
Park
WY
, et al
Intron retention is a widespread mechanism of tumor-suppressor inactivation
.
Nat Genet
2015
;
47
:
1242
8
.
2.
Pinto
EM
,
Chen
X
,
Easton
J
,
Finkelstein
D
,
Liu
Z
,
Pounds
S
, et al
Genomic landscape of paediatric adrenocortical tumours
.
Nat Commun
2015
;
6
:
6302
.
3.
Pinto
EM
,
Rodriguez-Galindo
C
,
Pounds
SB
,
Wang
L
,
Clay
MR
,
Neale
G
, et al
Identification of clinical and biologic correlates associated with outcome in children with adrenocortical tumors without germline TP53 mutations: a St. Jude Adrenocortical Tumor Registry and Children's Oncology Group study
.
J Clin Oncol
2017
;
35
:
3956
63
.
4.
Wasserman
JD
,
Novokmet
A
,
Eichler-Jonsson
C
,
Ribeiro
RC
,
Rodriguez-Galindo
C
,
Zambetti
GP
, et al
Prevalence and functional consequence of TP53 mutations in pediatric adrenocortical carcinoma: a Children's Oncology Group study
.
J Clin Oncol
2015
;
33
:
602
9
.
5.
Chompret
A
,
Brugières
L
,
Ronsin
M
,
Gardes
M
,
Dessarps-Freichey
F
,
Abel
A
, et al
P53 germline mutations in childhood cancers and cancer risk for carrier individuals
.
Br J Cancer
2000
;
82
:
1932
7
.
6.
Varley
JM
,
McGown
G
,
Thorncroft
M
,
James
LA
,
Margison
GP
,
Forster
G
, et al
Are there low-penetrance TP53 alleles? Evidence from childhood adrenocortical tumors
.
Am J Hum Genet
1999
;
65
:
995
1006
.
7.
Pinto
EM
,
Ribeiro
RC
,
Li
J
,
Taja-Chayeb
L
,
Carrasco
LF
,
de Lourdes Peña-Torres
M
, et al
An identical, complex TP53 mutation arising independently in two unrelated families with diverse cancer profiles: the complexity of interpreting cancer risk in carriers
.
Oncogene
2012
;
1
:
e1
.
8.
Powers
J
,
Pinto
EM
,
Barnoud
T
,
Leung
JC
,
Martynyuk
T
,
Kossenkov
AV
, et al
A rare TP53 mutation predominan in Ashkenazi Jews confers risk of multiple cancers
.
Cancer Res
2020
;
80
:
3732
44
.
9.
Pinto
EM
,
Figueiredo
BC
,
Chen
W
,
Galvao
HCR
,
Formiga
MN
,
Fragoso
MCBV
, et al
XAF1 as a modifier of p53 function and cancer susceptibility
.
Sci Adv
2020
;
6
:
eaba3231
.
10.
Bouaoun
L
,
Sonkin
D
,
Ardin
M
,
Hollstein
M
,
Byrnes
G
,
Zavadil
J
, et al
TP53 variations in human cancers: new lessons from the IARC TP53 database and genomics data
.
Hum Mutat
2016
;
37
:
865
76
.
11.
Jaganathan
K
,
Kyriazopoulou Panagiotopoulou
S
,
McRae
JF
,
Darbandi
SF
,
Knowles
D
,
Li
YI
, et al
Predicting splicing from primary sequence with deep learning
.
Cell
2019
;
176
:
535
48
.
12.
Jian
X
,
Boerwinkle
E
,
Liu
X
. 
In silico prediction of splice-altering single nucleotide variants in the human genome
.
Nucleic Acids Res
2014
;
42
:
13534
44
.
13.
Zhou
X
,
Wang
J
,
Patel
J
,
Valentine
M
,
Shao
Y
,
Newman
S
, et al
Exploration of coding and non-coding variants in cancer using GenomePaint
.
Cancer Cell
2021
;
11
:
83
95
.
14.
McBride
KA
,
Ballinger
ML
,
Killick
E
,
Kirk
J
,
Tattersall
MHN
,
Eeles
RA
, et al
Li-Fraumeni syndrome: cancer risk assessment and clinical management
.
Nat Rev Clin Oncol
2014
;
11
:
260
71
.
15.
Pinto
EM
,
Zambetti
GP
. 
What 20 years of research has taught us about the TP53 p. R337H mutation
.
Cancer
2020
;
126
:
4678
86
.
16.
Fortuno
C
,
James
PA
,
Young
EL
,
Feng
B
,
Olivier
M
,
Pesaran
T
, et al
Improved, ACMG-compliant, in silico prediction of pathogenicity for missense substitutions encoded by TP53 variants
.
Hum Mutat
2018
;
39
:
1061
9
.
17.
Kato
S
,
Han
SY
,
Liu
W
,
Otsuka
K
,
Shibata
H
,
Kanamaru
R
, et al
Understanding the function-structure and function-mutation relationships of p53 tumor suppressor protein by high-resolution missense mutation analysis
.
Proc Natl Acad Sci U S A
2003
;
100
:
8424
9
.
18.
Kotler
E
,
Shani
O
,
Goldfeld
G
,
Lotan-Pompan
M
,
Tarcic
O
,
Gershoni
A
, et al
A systematic p53 mutation library links differential functional impact to cancer mutation pattern and evolutionary conservation
.
Mol Cell
2018
;
71
:
178
90
.
19.
Giacomelli
AO
,
Yang
X
,
Lintner
RE
,
McFarland
JM
,
Duby
M
,
Kim
J
, et al
Mutational processes shape the landscape of TP53 mutations in human cancer
.
Nat Genet
2018
;
50
:
1381
7
.
20.
Tuck
SP
,
Crawford
L
. 
Characterization of the human p53 gene promoter
.
Mol Cell Biol
1999
;
9
:
2163
72
.
21.
Smeby
J
,
Sveen
A
,
Eilertsen
IA
,
Danielsen
SA
,
Hoff
AM
,
Eide
PW
, et al
Transcriptional and functional consequences of TP53 splice mutations in colorectal cancer
.
Oncogenesis
2019
;
8
:
35
.
22.
Shaulsky
G
,
Goldfinger
N
,
Ben-Ze'ev
A
,
Rotter
V
. 
Nuclear accumulation of p53 protein is mediated by several nuclear localization signals and plays a role in tumorigenesis
.
Mol Cell Biol
1990
;
10
:
6565
77
.
23.
Ribeiro
RC
,
Pinto
EM
,
Zambetti
GP
. 
Familial predisposition to adrenocortical tumors: clinical and biological features and management strategies
.
Best Pract Res Clin Endocrinol Metab
2010
;
24
:
477
90
.
24.
Rodriguez-Galindo
C
,
Krailo
MD
,
Pinto
EM
. 
Treatment of pediatric adrenocortical carcinoma with surgery, retroperitoneal lymph node dissection, and chemotherapy: The Children's Oncology Group ARAR0332 Protocol
.
J Clin Oncol
2021
;
39
:
2463
73
.
25.
Pondrom
M
,
Bougeard
G
,
Karanian
M
,
Bonneau‐Lagacherie
J
,
Boulanger
C
,
Boutroux
H
, et al
Rhabdomyosarcoma associated with germline TP53 alteration in children and adolescents: the French experience
.
Ped Blood Cancer
2020
;
67
:
e28486
.
26.
Mirabello
L
,
Yeager
M
,
Mai
PL
. 
Germline TP53 variants and susceptibility to osteosarcoma
.
J Natl Cancer Inst
2015
;
107
:
djv101
.
27.
Qian
M
,
Cao
X
,
Devidas
M
. 
TP53 germline variations influence the predisposition and prognosis of B-cell acute lymphoblastic leukemia in children
.
J Clin Oncol
2018
;
36
:
591
9
.
28.
Orr
BA
,
Clay
MR
,
Pinto
EM
,
Kesserwan
C
. 
An update on the central nervous system manifestations of Li-Fraumeni syndrome
.
Acta Neuropathol
2020
;
139
:
669
87
.
29.
Liu
APY
,
Wu
G
,
Orr
BA
,
Lin
T
,
Ashford
JM
,
Bass
JK
, et al
Outcome and molecular analysis of young children with choroid plexus carcinoma treated with non-myeloablative therapy: results from the SJYC07 trial
.
Neurooncol Adv
2021
;
3
:
vdaa168
.
30.
Renaux-Petel
M
,
Charbonnier
F
,
Théry
J-C
,
Fermey
P
,
Lienard
G
,
Bou
J
, et al
Contribution of de novo and mosaic TP53 mutations to Li-Fraumeni syndrome
.
J Med Genet
2018
;
55
:
173
80
.
31.
Gonzalez
KD
,
Buzin
CH
,
Noltner
KA
,
Gu
D
,
Li
W
,
Malkin
D
, et al
High frequency of de novo mutations in Li-Fraumeni syndrome
.
J Med Genet
2009
;
46
:
689
93
.
32.
Grill
S
,
Ramser
J
,
Hellebrand
H
,
Pfarr
N
,
Boxberg
M
,
Brambs
C
, et al
TP53 germline mutations in the context of families with hereditary breast and ovarian cancer: a clinical challenge
.
Arch Gynecol Obstet
2021
;
303
:
1557
67
.
33.
Varley
JM
,
Chapman
P
,
McGown
G
,
Thorncroft
M
,
White
GRM
,
Greaves
MJ
, et al
Genetic and functional studies of a germline TP53 splicing mutation in a Li-Fraumeni-like family
.
Oncogene
1998
;
16
:
3291
8
.
34.
Varley
JM
,
Attwooll
C
,
White
G
,
McGown
G
,
Thorncroft
M
,
Kelsey
AM
, et al
Characterization of germline TP53 splicing mutations and their genetic and functional analysis
.
Oncogene
2001
;
20
:
2647
54
.
35.
Bougeard
G
,
Sesboue
R
,
Baert-Desurmont
S
,
Vasseur
S
,
Martin
C
,
Tinat
J
, et al
Molecular basis of the Li-Fraumeni syndrome: an update from the French LFS families
.
J Med Genet
2008
;
45
:
535
8
.
36.
Warneford
SG
,
Witton
LJ
,
Townsend
ML
,
Rowe
PB
,
Reddel
RR
,
Dalla-Pozza
L
, et al
Germ-line splicing mutation of the p53 gene in a cancer-prone family
.
Cell Growth Differ
1992
;
3
:
839
46
.
37.
Ruijs
MWG
,
Verhoef
S
,
Rookus
MA
,
Pruntel
R
,
van der Hout
AH
,
Hogervorst
FBL
, et al
TP53 germline mutation testing in 180 families suspected of Li-Fraumeni syndrome: mutation detection rate and relative frequency of cancers in different familial phenotypes
.
J Med Genet
2010
;
47
:
421
8
.
38.
Hayman
L
,
Chaudhry
WR
,
Revin
VV
,
Zhelev
N
,
Bourdon
JC
. 
What is the potential of p53 isoforms as a predictive biomarker in the treatment of cancer?
Expert Rev Mol Diagn
2019
;
19
:
149
59
.
39.
Houlston
RS
,
Tomlinson
IPM
. 
Modifiers genes in humans: strategies for identification
.
Eur J Hum Genet
1998
;
6
:
80
8
.