Ataxia-Telangiectasia Mutated (ATM) gene ranks among the most common gene affected by germline variants without hot-spots. Germline homozygous pathogenic mutations are responsible for Ataxia-Telangiectasia (A-T) syndrome (Savitsky et al., 1995), an autosomal recessive inherited disease characterized by progressive cerebellar ataxia, immunodeficiency, radiosensitivity, and high predisposition for different type of cancer (Choi et al., 2016). Heterozygous carriers of these pathogenic germline mutations are usually asymptomatic and are considered healthy carriers, but a correlation between single copy ATM mutations and increased cancer risk, sensitivity to ionizing radiation and oxidative stress have been found (Stagni et al., 2014). In addition, next generation sequencing (NGS) of various solid and hematologic cancers revealed that somatic ATM mutations are among the most commonly observed (Cerami et al, 2012), and that these variants are more heterogeneous and with much less evident functional impact then the ones found in A-T patients, who carry almost exclusively nonsense truncating mutations. Despite advances in high-throughput sequencing, which have generated detailed catalogues of genetic variations in both cancer patients and in the general population, our ability to predict the biological consequences of individual ATM variant remains poor, limiting the clinical impact of sequence information. This problem is supported by the large number of ATM missense variants with unclear consequence on protein function, the variants of uncertain significance (VUS), and by the necessity of functional characterization of these variants. (Federici and Soddu, 2020). ATM is involved in important cell pathways, such as the recognition of DNA double strand breaks, the DNA damage response (DDR), and the regulation of mitochondrial homeostasis. Recently, we demonstrated that ATM is required for the centrosomal localization of p53 during mitosis (p53-MCL). This mechanism is impaired in cells carrying ATM pathogenic variants thus, its evaluation identifies pathogenic nonsense and truncating variants. According to this, in my host laboratory, a functional test that can discriminate between homozygous and heterozygous carriers of pathogenic ATM mutations and healthy individuals carrying the wild-type ATM gene has been developed (Prodosmo et al, 2013). We asked whether the p53-MCL test is also able to distinguish ATM missense pathogenic variants from benign polymorphisms and, therefore, can be used to predict pathogenicity of ATM VUS and classify them. Preliminary data obtained on PBMCs from patients carrying ATM pathogenic missense variants, breast cancer patients, carrying missense VUS, and healthy donors with different ATM polymorphisms, showed that p53-MCL test results indicate that all pathogenic variants alter p53-MCL, while all polymorphisms do not modify p53-MCL. These data indicate that p53-MCL test can functionally discriminate between loss-of-function mutations and benign variants of ATM. The goal of this project is to genetically validate the p53-MCL test for ATM missense variants by the generation of specific CRISPR/Cas9 knock-in cell clones to predict the pathogenicity of ATM VUS. To this aim, we selected twelve ATM variants to generate CRISPR/Cas9 KI cell clones in non-transformed RPE-1 and MCF10A cells each carrying a different ATM variant by co-transfecting the cells with a vector expressing a specific guideRNA (gRNA) sequence and the endonuclease Cas9, together with their matched single strand oligodeoxynucleotides (ssODN) as donor template to promote the homologous directed repair. We initially tried to genetically modify the RPE-1 cells with CRISPR construct carrying the c.2413C>T pathogenic variants. Despite several attempts, we failed to obtain clones. We also failed in our try to generate RPE-1 clones containing two other variants (c.2572T>C and c.7157C>A), so we decided to change our model for the CRISPR/Cas9-mediated genome editing experiments, switching to MCF10A cells. In these cells, we initially got two clones carrying c.2572T>C variant, a VUS which has been found in breast cancer patients (Prodosmo et al., 2013) but recently reclassified as benign/likely benign (ClinVar). Because the functional profiles of these two cell populations were comparable to the wild type, we decided to re-sequence their DNA at consecutive cell passages, and we found out that both cell populations lost the c.2572T>C variant, suggesting that the presence of this variant was detrimental for cell fitness. On these bases, we changed our strategy from analyzing single cell clones to bulk cell populations at different time points to assess which of our variants of interest were retained over time and which were not. Our hypothesis is that if a variant has a benign phenotype and does not affect ATM protein functions, it will be retained over time, while a pathogenic variant that affects ATM features, will be lost in few passages within a mixed population.
Il gene Ataxia-Telangiectasia Mutated (ATM) è tra i geni più frequentemente colpiti da varianti germinali, senza specifici hot-spot. Le mutazioni patogenetiche della linea germinale, in omozigosi sono responsabili della Ataxia-Telangiectasia (A-T) (Savitsky et al., 1995), una malattia autosomica recessiva caratterizzata da progressiva atassia cerebellare, immunodeficienza, radiosensibilità e un'elevata predisposizione a diversi tipi di tumore (Choi et al., 2016). In eterozigosi, queste mutazioni patogenetiche generalmente non sono associate a sintomi e gli individui che le hanno sono considerati portatori sani, ma è stata comunque riscontrata una correlazione tra la presenza di una singola copia mutata di ATM e un aumento del rischio di cancro, della sensibilità alle radiazioni ionizzanti e allo stress ossidativo (Stagni et al., 2014). In aggiunta, il sequenziamento (NGS) di diversi tumori solidi ed ematologici ha rivelato che le mutazioni somatiche di ATM sono tra le più comunemente osservate (Cerami et al., 2012) e che queste varianti sono più eterogenee e con un impatto funzionale molto meno evidente rispetto a quelle trovate nei pazienti con A-T, che presentano quasi esclusivamente mutazioni troncanti. Nonostante i progressi nelle tecniche di sequenziamento, che hanno permesso di generare collezioni dettagliate delle varianti genetiche, sia nei pazienti oncologici che nella popolazione generale, la nostra capacità di prevedere le conseguenze biologiche delle singole varianti di ATM rimane limitata, riducendo l’impatto clinico di queste tecnologie. Questo problema è supportato dal gran numero di varianti missenso ritrovate in ATM che hanno conseguenze poco chiare sulla funzione della proteina, le varianti con significato incerto (Variant of Uncertain Significance, VUS) e dalla necessità di caratterizzazione funzionale di queste ultime (Federici e Soddu, 2020). ATM è coinvolto in importanti processi cellulari, come il riconoscimento delle rotture del doppio filamento di DNA, la risposta al danno al DNA e la regolazione dell’omeostasi mitocondriale. Recentemente, abbiamo dimostrato che ATM è necessario per la localizzazione centrosomale di p53 durante la mitosi (p53 mitotic centrosome localization, p53-MCL). Questo meccanismo è compromesso nelle cellule che presentano varianti patogenetiche di ATM, e quindi la sua valutazione ne permette l’identificazione. In base a questo, nel laboratorio dove ho svolto il mio Dottorato di Ricerca è stato sviluppato un test funzionale in grado di distinguere tra portatori omozigoti ed eterozigoti di mutazioni patogenetiche di ATM e individui sani che possiedono il gene ATM wild-type (Prodosmo et al., 2013). Ci siamo chiesti se il test p53-MCL fosse in grado di distinguere anche le varianti missenso patogenetiche di ATM dai polimorfismi e, quindi, potesse essere utilizzato per prevedere la patogenicità delle VUS di ATM e classificarle. Dati preliminari ottenuti su cellule di pazienti con varianti missenso patogenetiche di ATM, pazienti con carcinoma mammario portatori di VUS e donatori sani con diversi polimorfismi di ATM hanno mostrato che i risultati del test p53-MCL indicano che tutte le varianti patogenetiche alterano la localizzazione di p53 ai centrosomi, mentre tutti i polimorfismi non la modificano. Questi dati suggeriscono che il test p53-MCL è in grado di discriminare funzionalmente tra mutazioni con perdita di funzione e varianti benigne di ATM. L’obiettivo di questo progetto è quello di validare geneticamente il test p53-MCL per le varianti missenso di ATM mediante la generazione di specifici cloni cellulari modificati tramite tecnica CRISPR/Cas9 con singole varianti di ATM, al fine di prevedere la patogenicità delle VUS di ATM. A questo scopo, abbiamo selezionato dodici varianti di ATM per generare questi cloni nelle linee cellulari non trasformate, RPE-1 e MCF10A, ciascuno contenente una variante diversa di ATM, co-trasfettando le cellule con un vettore che esprime una specifica sequenza di RNA guida (guideRNA, gRNA) e l'endonucleasi Cas9, insieme ai rispettivi oligonucleotidi a singolo filamento (ssODN) come templato per promuovere la ricombinazione omologa. Inizialmente, abbiamo tentato di modificare geneticamente le cellule RPE-1 con il costrutto CRISPR contenente la variante patogenetica c.2413C>T. Nonostante numerosi tentativi, non siamo riusciti a ottenere nessun clone. Abbiamo inoltre fallito nel tentativo di generare cloni di cellule RPE-1 contenenti altre due varianti (c.2572T>C e c.7157C>A), quindi abbiamo deciso di cambiare il nostro modello sperimentale per l’editing genomico CRISPR/Cas9, iniziando a utilizzare le cellule MCF10A. In queste cellule, inizialmente abbiamo ottenuto due cloni contenenti la variante c.2572T>C, una VUS, trovata in pazienti con carcinoma mammario (Prodosmo et al., 2013) ma recentemente riclassificata come benign/likely benign (ClinVar). Poiché i profili funzionali di questi due cloni cellulari erano comparabili a quello delle cellule wild-type, abbiamo deciso di risequenziare il loro DNA a passaggi cellulari consecutivi e abbiamo scoperto che entrambe le popolazioni avevano perso la variante c.2572T>C e che quindi, probabilmente, la presenza di questa variante era dannosa per la crescita cellulare. Sulla base di queste osservazioni, abbiamo modificato la nostra strategia, passando dall'analisi di cloni cellulari singoli all'analisi di popolazioni cellulari a tempi successivi, per valutare quali delle varianti di interesse venissero mantenute nel tempo e quali no. La nostra ipotesi è che se una variante ha un fenotipo benigno e non altera le funzioni della proteina ATM, verrà mantenuta nel tempo, mentre una variante patogenetica che compromette le funzioni di ATM verrà persa dopo pochi passaggi all'interno di una popolazione mista.
Calconi, G. (2025). Evaluation of individual ATM variants to validate p53-MCL as a functional test to predict pathogenicity of ATM VUS.
Evaluation of individual ATM variants to validate p53-MCL as a functional test to predict pathogenicity of ATM VUS
Giulia Calconi
2025-07-01
Abstract
Ataxia-Telangiectasia Mutated (ATM) gene ranks among the most common gene affected by germline variants without hot-spots. Germline homozygous pathogenic mutations are responsible for Ataxia-Telangiectasia (A-T) syndrome (Savitsky et al., 1995), an autosomal recessive inherited disease characterized by progressive cerebellar ataxia, immunodeficiency, radiosensitivity, and high predisposition for different type of cancer (Choi et al., 2016). Heterozygous carriers of these pathogenic germline mutations are usually asymptomatic and are considered healthy carriers, but a correlation between single copy ATM mutations and increased cancer risk, sensitivity to ionizing radiation and oxidative stress have been found (Stagni et al., 2014). In addition, next generation sequencing (NGS) of various solid and hematologic cancers revealed that somatic ATM mutations are among the most commonly observed (Cerami et al, 2012), and that these variants are more heterogeneous and with much less evident functional impact then the ones found in A-T patients, who carry almost exclusively nonsense truncating mutations. Despite advances in high-throughput sequencing, which have generated detailed catalogues of genetic variations in both cancer patients and in the general population, our ability to predict the biological consequences of individual ATM variant remains poor, limiting the clinical impact of sequence information. This problem is supported by the large number of ATM missense variants with unclear consequence on protein function, the variants of uncertain significance (VUS), and by the necessity of functional characterization of these variants. (Federici and Soddu, 2020). ATM is involved in important cell pathways, such as the recognition of DNA double strand breaks, the DNA damage response (DDR), and the regulation of mitochondrial homeostasis. Recently, we demonstrated that ATM is required for the centrosomal localization of p53 during mitosis (p53-MCL). This mechanism is impaired in cells carrying ATM pathogenic variants thus, its evaluation identifies pathogenic nonsense and truncating variants. According to this, in my host laboratory, a functional test that can discriminate between homozygous and heterozygous carriers of pathogenic ATM mutations and healthy individuals carrying the wild-type ATM gene has been developed (Prodosmo et al, 2013). We asked whether the p53-MCL test is also able to distinguish ATM missense pathogenic variants from benign polymorphisms and, therefore, can be used to predict pathogenicity of ATM VUS and classify them. Preliminary data obtained on PBMCs from patients carrying ATM pathogenic missense variants, breast cancer patients, carrying missense VUS, and healthy donors with different ATM polymorphisms, showed that p53-MCL test results indicate that all pathogenic variants alter p53-MCL, while all polymorphisms do not modify p53-MCL. These data indicate that p53-MCL test can functionally discriminate between loss-of-function mutations and benign variants of ATM. The goal of this project is to genetically validate the p53-MCL test for ATM missense variants by the generation of specific CRISPR/Cas9 knock-in cell clones to predict the pathogenicity of ATM VUS. To this aim, we selected twelve ATM variants to generate CRISPR/Cas9 KI cell clones in non-transformed RPE-1 and MCF10A cells each carrying a different ATM variant by co-transfecting the cells with a vector expressing a specific guideRNA (gRNA) sequence and the endonuclease Cas9, together with their matched single strand oligodeoxynucleotides (ssODN) as donor template to promote the homologous directed repair. We initially tried to genetically modify the RPE-1 cells with CRISPR construct carrying the c.2413C>T pathogenic variants. Despite several attempts, we failed to obtain clones. We also failed in our try to generate RPE-1 clones containing two other variants (c.2572T>C and c.7157C>A), so we decided to change our model for the CRISPR/Cas9-mediated genome editing experiments, switching to MCF10A cells. In these cells, we initially got two clones carrying c.2572T>C variant, a VUS which has been found in breast cancer patients (Prodosmo et al., 2013) but recently reclassified as benign/likely benign (ClinVar). Because the functional profiles of these two cell populations were comparable to the wild type, we decided to re-sequence their DNA at consecutive cell passages, and we found out that both cell populations lost the c.2572T>C variant, suggesting that the presence of this variant was detrimental for cell fitness. On these bases, we changed our strategy from analyzing single cell clones to bulk cell populations at different time points to assess which of our variants of interest were retained over time and which were not. Our hypothesis is that if a variant has a benign phenotype and does not affect ATM protein functions, it will be retained over time, while a pathogenic variant that affects ATM features, will be lost in few passages within a mixed population.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


