The targeted next-generation sequence revealed SMAD4, AKT1, and TP53 mutations from circulating cell-free DNA of breast cancer and its effect on protein structure – A computational approach

Abstract Breast cancer biomarkers that detect marginally advanced stages are still challenging. The detection of specific abnormalities, targeted therapy selection, prognosis, and monitoring of treatment effectiveness over time are all made possible by circulating free DNA (cfDNA) analysis. The proposed study will detect specific genetic abnormalities from the plasma cfDNA of a female breast cancer patient by sequencing a cancer-related gene panel (MGM455 – Oncotrack Ultima), including 56 theranostic genes (SNVs and small INDELs). Initially, we determined the pathogenicity of the observed mutations using PredictSNP, iStable, Align-GVGD, and ConSurf servers. As a next step, molecular dynamics (MD) was implemented to determine the functional significance of SMAD4 mutation (V465M). Lastly, the mutant gene relationships were examined using the Cytoscape plug-in GeneMANIA. Using ClueGO, we determined the gene’s functional enrichment and integrative analysis. The structural characteristics of SMAD4 V465M protein by MD simulation analysis further demonstrated that the mutation was deleterious. The simulation showed that the native structure was more significantly altered by the SMAD4 (V465M) mutation. Our findings suggest that SMAD4 V465M mutation might be significantly associated with breast cancer, and other patient-found mutations (AKT1-E17K and TP53-R175H) are synergistically involved in the process of SMAD4 translocate to nuclease, which affects the target gene translation. Therefore, this combination of gene mutations could alter the TGF-β signaling pathway in BC. We further proposed that the SMAD4 protein loss may contribute to an aggressive phenotype by inhibiting the TGF-β signaling pathway. Thus, breast cancer’s SMAD4 (V465M) mutation might increase their invasive and metastatic capabilities. Communicated by Ramaswamy H. Sarma


Introduction
Breast cancer is the deadliest and most common cancer in females accounting for about 23% of cancer cases, showing a rapid increase in its incidence.With advancements in molecular medicine and diagnosis, the mortality rate has declined gradually (Jemal et al., 2010).GLOBOCAN (2020) statistics revealed that around 9.22 million (9,227,484) newly reported cancer cases were females.Of these, 2.26 million (2,261,419) cases belong to Breast Cancer, accounting for about 24.5% of the total cancer cases in females (Sung et al., 2021).The cfDNA estimation has the potential as a 'liquid biopsy' to determine and monitor breast and other cancers, explicitly through advanced stages (Jung et al., 2010).Essentially, the term liquid biopsy implies the presence of circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) found in the blood released from the sites of tumor progression of breast cancer or other cancer patients.Determining ctDNA provides improved diagnostic prospects in detecting alterations implying cancer-specific biomarkers or cancer antigens (Schwarzenbach et al., 2011).
Studies have shown the inculcation of SNP array for genomic analysis of cfDNA to ascertain the alterations and CNVs in breast cancer patients (Shaw et al., 2012).Single nucleotide variations are common in the human genome, with about 10 million found in each individual (Cline & Karchin, 2011).The SNVs act as biological markers; their presence in the coding region or regulatory region is a directive of their role in gene expression that can be co-related with the disease gene for diagnosis (Katsonis et al., 2014).However, SNVs may not be directly involved in disease as most of them may not affect health or disease development.The NGS data analysis primarily focuses on SNVs and INDELs (Carvalho & Lupski, 2016;Kumar et al., 2022;Shigemizu et al., 2018).Recent studies and advancements in NGS paved the way for genome coverage with higher specificity of genetic heterogeneity to label for variations across the whole-genome analysis of cfDNA in cancer patients (Beck et al., 2010;Leary et al., 2012).The NGS studies on insertion-deletion INDEL in PTENa via bioinformatics analysis established their correlation as a risk factor associated with the development of breast cancer (C.Hata et al., 2019).Most studies have utilized bioinformatic algorithms to compute the pathogenicity, amino acid evolution, stability, and structural behavior of mutations responsible for different cancers (S.U. Kumar & Priya Doss, 2021;Kumar et al., 2022;Uk et al., 2022;Thirumal Kumar et al., 2021).The molecular markers EPCAM, MET, CD44 and CD47 are involved in the development of a tumor and are co-related with the TGF-b signaling pathway; through cfDNA determination and analysis, the patient-specific survival predictions can be estimated (Ascolani et al., 2015;Shimada et al., 2011).

Isolation of cfDNA
Informed consent was obtained from the patient prior to the study enrollment.The institutional ethical committee of Sri Ramachandra Institute of Higher Education and Research (SRIHER), Porur, Chennai (IEC-NI/21/APR/78/80), approved the study.The peripheral blood was collected from the breast cancer patient in a cell-free DNA collection tube (VanGenes, Xiamen, China).Later, the blood was centrifuged at 1600 g for 10 min at 4 � C to partition the blood into plasma and blood cells.The upper phase was moved into a second tube to prevent blood or cell debris from contaminating the plasma layer.The plasma underwent additional centrifugation for 1 min to eliminate the cell debris.Afterward, the upper transparent layer plasma was transferred to the new tube and immediately kept at À 80 � C for cfDNA extraction.

cfDNA quantification
We used the E.Z.N.A circulating DNA kit to extract cfDNA (Omega Bio-Tek, GA, USA) from 1 ml of plasma.The amount of isolated DNA was determined using the Qubit 2.0 dsDNA high-sensitivity assay (Life Technologies, Carlsbad, CA).All procedures were carried out following the manufacturer's recommendations.

Library construction, sequencing and identification of SNVs and small INDELs
The isolated cfDNA was utilized to perform target enrichment and sequencing of a cancer-related gene panel (MGM455 -Oncotrack Ultima), including 56 theranostic genes (SNVs and INDELS).This gene panel comprises the complete coding regions of 71 genes and hotspot regions of 46 genes, which are enlisted in the Supplementary Table 1.This somatic tumor panel is being investigated to screen for somatic mutations, including 56 cancer-related genes associated with carcinogenesis, prognosis, and predictive value for chemotherapy and targeted treatment therapy in various tumor types.The sequencing of the libraries was accomplished by the Illumina sequencing method with >20000X mean coverage depth.Correspondingly, the BWA program was utilized to align the library sequences with the reference human genome (GRCh37/hg19) (Chong et al., 2003;Meyer et al., 2012).The LoFreq (version 2) variant caller was utilized to identify somatic mutations especially splice sites and nonsynonymous variants (H.Li et al., 2009;Wilm et al., 2012).Nonetheless, the annotation of mutations was achieved by using an in-house annotation pipeline (VariMAT).However, using the VEP program (McLaren et al., 2016), gene annotation of variants was accomplished against the Ensembl 90 Human Gene Model (Zerbino et al., 2018).Referring through published literature, databases and in-house propriety mutations that are clinically significant were annotated.Attributing to population databases (1000Japanese, MedVarDb, 1000 G, EVS, UK10K, and ExAC), the most frequent variants were identified and reported (Lek et al., 2016;Moayyeri et al., 2013;Auton et al., 2015;Nagasaki et al., 2015).

Pathogenicity analysis for the mutations
The mutations obtained from cfDNA were further analyzed for pathogenicity.The pathogenicity of mutations was predicted by employing the PredictSNP Server that comprehends the anticipation of multiple algorithms with established tools PhD-SNP, PolyPhen-1, PolyPhen-2, PredictSNP, SIFT, and MAPP prediction (Bendl et al., 2014).The prediction tool SIFT was utilized to examine the effect of amino acid substitution on the native form; less than a 0.05 probability score indicates deleterious mutation (Vaser et al., 2016).Nevertheless, the effects of missense mutations can be recognized by another SNP prediction tool entitled Polyphen2 (Adzhubei et al., 2010).The PhD-SNP tool, however, predicts the association of a particular mutation with the disease condition by employing the SVM method (Capriotti et al., 2006).The predictions on stability changes over mutation were gathered through the iStable server that comprises iStable, I-Mutant 2.0 SEQ, and MUpro (Chen et al., 2013).iStable is a database tool that predicts the changes in the stability of a protein in response to SNP mutations by utilizing the SVM method (Cheng et al., 2005;Gilis & Rooman, 1997).Consequently, the track of stability change determines the pertinence of stabilization or destabilization relating to the SNP variation or mutant (Guerois et al., 2002;Topham et al., 1997).Integrating the prediction tools to achieve a higher degree of coding methods and analysis is decisive.The determination of missense substitution being either deleterious or neutral can be executed by Align-GVGD tool (Tavtigian et al., 2006).The biophysical characteristics of a protein or amino acid sequences are aligned to predict and analyze the missense substitution by Align-GVGD database tool (http://agvgd.hci.utah.edu/about.php).

Evolutionary conservation analysis for mutations
The amino acid position in protein molecules is evolutionarily conserved, and identifying these conserved sequences plays a significant role in determining protein function.The web server ConSurf runs on a LINUX cluster of 2.6 GHz AMD Opteron processors that is applied to assess these evolutionarily conserved sequences and their implication for protein structure and function.The analysis by comparison assists in resolving the similarities between amino acids at specific positions in the protein (W.Li & Godzik, 2006).The ConSurf database implies two essential ML and Bayesian methods that allow evaluation between absolute conserved sequences and short-period evolutionarily conserved sequences (Mayrose, 2004).The ConSurf database was used to sequence mutant and native protein forms and compare the phylogenetic or evolutionary relationship among conserved sequences and their role in protein function.

Structural effect of mutant predicted using HOPE tool
The advancements in sequencing with NGS have revolutionized the medical field, which helps to understand the mechanism and pathway causing disease (Metzker, 2010).However, a major drawback lies in analyzing the mutations causing disease.HOPE tool is an automated program that exploits SNPs in determining how protein structure or function may be affected in response to point mutations (Camilli et al., 2011).Here, we employed the HOPE tool to determine whether the mutations alter the protein structure or functions.

MD simulation
The SWISS PDB Viewer was used to elicit the mutations in the native protein to develop a mutant protein (Guex & Peitsch, 1997).The native form of SMAD4 protein had missing residues; hence the protein structure was constructed using the SWISS-MODEL server using the template SMAD4 protein PDB ID: 1DD1.The structure of native and mutant SMAD4 was energy minimized using GROMOS force field rooted within SWISS PDB Viewer before performing MD simulation (van Gunsteren et al., 1996).GROMACS software was utilized to study the molecular dynamics of the protein (Abraham et al., 2015).The force field for energy functions was employed through the CHARM27 all-atom-force field (Schmid et al., 2011).All the hydrogen atoms were ignored by utilizing the 'ignh' option well-found in the GROMACS plug-in.A transferable intermolecular potential with 3 points (TIP3P) water in MD was established by contemplating a cubical box.The position of the protein structure inside the cubical box was manipulated such that the distance from the edge of the cubical box measured 10 � Å toward the protein.The neutrality of the protein structure was achieved through 'genion', which allows the replacement of twelve solvents with an equal number of sodium ions.To achieve minimum energy conformation and reduce stearic clashes Energy minimization step was initiated.A force of 1000 kJ/mol/nm with 5000 steps was utilized for EM.The implications of bond length were studied through the Steepest Descent Algorithm method (Petrova & Solov'ev, 1997).
Similarly, the calculations of efficient long-range electrostatic interactions were obtained through the Particle Mesh Ewald method (Essmann et al., 1995).The Berendsen Coupling method was accustomed to maintaining a temperature of 300K inside the cubical box (Berendsen et al., 1984).The unrestrained MD was prevented by applying NPT (Number of particles, Pressure, and Temperature) that follows NVT (Number of particles, Volume, and Temperature) for a while of 500ps.To circumvent the influence of velocity on MD Parrinello-Rahman barostat pressure coupling method was applied (Parrinello & Rahman, 1981).Finally, to impel water molecules and non-water bonds, algorithms, such as SETTLE and LINCS, were employed (Hess et al., 1997;Miyamoto & Kollman, 1992).MD ran for 400 ns that were initiated using the native and mutant protein.RMSD, RMSF, hbond, Radius of gyrate, and PCA were analyzed using the plug-ins of GROMACS, and the protocols were followed from our previous research (Kumar et al., 2021;Udhaya Kumar et al., 2020).

Gene interaction network through GeneMANIA and ClueGO
GeneMANIA plug-in in Cytoscape software was utilized to determine the gene function and gene-gene interaction network of the identified mutant gene from plasma cfDNA (Franz et al., 2018).GeneMANIA is an effective tool and database for gene function prediction based on the several networks collected from various genomic/proteomic data (Mostafavi et al., 2008).A fundamentally organized GO or pathway network is provided by ClueGO, which incorporates GO and pathway analysis from KEGG and Reactome (Aoki-Kinoshita & Kanehisa, 2007;Fabregat et al., 2018).Further, ClueGO/Cluepedia was performed on the identified mutant genes from plasma cfDNA to generate thorough GO, KEGG, and Reactome pathways from the PPI network (Bindea et al., 2009(Bindea et al., , 2013)).

Mutations from cfDNA of breast cancer
We extracted the plasma DNA from the blood of a 53-yearold woman with breast cancer.The patient's plasma DNA quantity yielded more than 30 ng, which is adequate for sequencing.Utilizing the ctDNA (liquid biopsy) from the patient's plasma as the genetic material, we screened for tumor somatic mutations in the cancer-related gene panel (MGM455 -Oncotrack Ultima), including 56 cancer-related genes linked to carcinogenesis, prognosis, and predictive value for chemotherapy and targeted treatment medications in various types.An NGS-based multigene study utilizing Illumina sequencing with >20000X mean coverage depth identified three mutations: AKT1 E17K, TP53 R175H, and SMAD4 V465M.
Clinically relevant genomic mutations related to AKT1 and TP53 genes have been found in different undruggable or druggable cancers.An AKT1 mutation, the c.49G > A (p.Glu17Lys) at exon 2, with an overall depth and mutant allele percentage of 3025X and 3%, was detected in the plasma cfDNA of a breast carcinoma patient, and this is a gain of function mutation.A TP53 mutation, the c.524G > A (p.Arg175His) at exon 5, with an overall depth and mutant allele percentage of 2487X and 3.4%, and this protein impacts on loss of function.A SMAD4 mutation, the c.1393G > A (p.Val465Met) at exon 11 with an overall depth and mutant allele percentage of 1328X and 49.4%.It is still uncertain how the SMAD4 variation in this patient would affect this malignancy because its biochemical characterization remains unexplored.Therefore, our study analyzed the V465M mutation impact on the SMAD4 protein structure through MD simulation.

Pathogenicity analysis of mutations
Multiple consensus tools were used to determine the damaging or deleterious mutations.PredictSNP, an online tool, was utilized at first to predict the three mutations found in this patient.The results of the pathogenicity prediction analysis from the PredictSNP server and their anticipated accuracy are listed in Table 1.Further, the mutation in the protein and their impact on the protein stability were predicted using the I-stable server, and the attained results were listed in Table 1.The PredictSNP and iStable servers were combined to increase the prediction confidence level.Using Align-GVGD, mutations we identified from the plasma cfDNA were inspected for their effects on the biophysical and physiochemical changes.As a result, we anticipated that the mutation AKT1 E17K would interfere with the protein's ability to function because it falls within class C65.The other two mutations, TP53 R175H and SMAD4 V465M had a less negative impact because they belonged to classes C25 and C15 (Table 1).ConSurf calculates each amino acid's degree of evolutionary conservation and displays the distribution of modeled 3D structures' functional and structural residues.The anticipated conserved variable regions for the genes were further separated into distinct scales of nine grades using conservation scores.Conservation scores for the proteins AKT1 E17, TP53 R175, and SMAD4 V465 were 6, 9, and 8, respectively (Table 1).
Before the MD simulation, the structural impact of a point mutation on the SMAD4 protein was discovered utilizing the HOPE server.The mutant methionine residue is larger than the valine (native) residue at position 465 in the SMAD4 protein (Figure 1).The mutant residue is found in an MH2 domain of the protein that is crucial for binding to other interacting partners.

Molecular dynamics simulation
Since the simulation studies were already carried out for the proteins E17K (AKT1) and R175H (TP53), our study specifically focused on the V465M (SMAD4) mutation and their structural behavior on SMAD4 protein.The MD analysis was conducted to assess the differences and comprehend the dynamic behavior between the native and mutant SMAD4.RMSD was first plotted to evaluate the stability of the SMAD4 protein.RMSD for the backbone atoms in the SMAD4 and mutant V465M starting structures throughout the simulation were interpreted.After 120 ns, the variations in RMSD remained nearly constant, showing that the MD simulations had correctly converged throughout this time frame (Figure 2(a)).The flexibility of the mutant and SMAD4 protein structure was computed using the RMSF analysis from GROMACS.RMSF predicts each protein's residue fluctuation.Accordingly, the RMSF plot shows a higher variation in the V465M protein structure around the mutant region (residues 460-490) compared to the SMAD4 protein.Furthermore, a significant fluctuation was also seen between the region (residues 275-322); this region is considered the middle linker region of SMAD4, also known as the SMAD4 activation domain (SAD).A crucial locus regulating the transcriptional activity of the SMAD complex is SAD.Compared to the SMAD4 protein, SMAD4 with the V465M mutation exhibits lower fluctuation in the residues between 290 and 320.The V465M mutation showed structural fluctuation between the residues 275-322, �0.35 nm-0.3 nm in the mutant, whereas �0.85 nm À 0.7 nm in the native SMAD4.Similarly, a structural fluctuation was observed between the residues 465-470, 0.55 nm in mutant whereas 0.4 nm in native (Figure 2(b)).
The intramolecular h-bonds were investigated to predict the change in h-bond formation in V485M mutant SMAD4 protein in comparison to SMAD4 protein.A significant parameter of GROMACS called gmx gyrate was used to estimate the compactness of the protein in the solution.A slightly increased number of h-bonds was observed in mutant V465M (average of 172.84 H-bonds) compared to the native (average of 171.35 Hbonds) of SMAD4 protein (Figure 3(a)).The V465M mutant showed minor deviations (�1.95 nm) compared to the SMAD4 native, depicting the increased compactness.Comparatively, SMAD4 native observed a higher deviation (�1.98 nm), depicting the least compactness (Figure 3(b)).The increased h-bonds numbers can also be attributed to the lesser deviations.Additionally, the PCA was examined for the mutant structures to comprehend the dynamics and behavior of the SMAD4 protein.It is well recognized that a protein with greater ca motions and structural changes is less stable.PCA revealed a significant change in the mobility of the proteins, which supports the instability of the mutant structure.The SMAD4 native showed changes in atomic motions ranging from �À 7.5 to 10 nm on the x-axis and �À 7.5 to 6 nm on the y-axis.In contrast, mutant V465M SMAD4 observed changes in atomic motions ranging from �À 15 to 2.5 nm on the x-axis and �À 7.5 to 8 nm on the y-axis (Figure 3(c)).Further, the mutant showed lower motion correlation and a divergent cluster direction from the SMAD4 native, which led to an unstable mutant structure and disturbed function of SMAD4.

GeneMANIA and ClueGO
The three mutant genes identified from the study, SMAD4, TP53, and AKT1, were imported into GeneMANIA to anticipate their activities and interactions due to the unknown roles, co-relation of genes, and co-expressed genes between them.As seen in Figure 4(a), several genes, comprising SMAD2, PHLPP1, SMAD1, FOXO3, MDM2, MYB, SMAD3, RGCC, PHLPP2, APPL1, TP53BP1, SIN3A, FOXO4, COP1, RPS6KB2, DLX1, DRAP1, MTOR, AKT2, and TGIF2 were co-localized, coexpressed, genetically and physically interacted, and shared domains and pathways with SMAD4, TP53, and AKT1.As a result, it is possible to hypothesize that SMAD4, TP53, and AKT1 may have functional connections.SMAD4, TP53, and AKT1, together with the other genes specified in Table 2, may be crucial regulators of SMAD binding, cellular response to transforming growth factor beta stimulus, cell cycle arrest, transcription regulator complex, regulation of cell cycle G1/S phase transition, and positive regulation of transcription by RNA polymerase II, according to the top 20 GO annotated functions in line with the GeneMANIA network.
The ClueGO/CluePedia plug-in from Cytoscape was used to examine the functional enrichment of the three mutant genes (SMAD4, AKT1, and TP53) identified from the breast cancer patient as an input.ClueGO assists in visualizing the biological terms and pathways of these genes cluster based on the Benjamini-Hochberg correction, p 0.05, and kappa score 0.4 as thresholds (Figure 4 today).Improving the treatment strategies has shown a significant advantage against breast cancer; however, until now, targeted therapies against HER2 over-expression were predominantly applied (Slamon et al., 2001).Nevertheless, not only HER2 amplifications but other genetic alterations have also been associated with inducing Breast cancer (Curtis et al., 2012;Stephens et al., 2012).CfDNA analysis reported three pathogenic mutations in the present study, SMAD4 (V465M), AKT1 (E17K), and TP53 (R175H), through gene panel sequencing.
The AKT1 gene is essential in cellular processes, notably proliferation, growth, apoptosis, and angiogenesis (Landgraf et al., 2008).AKT1 belongs to the serine-threonine kinase family, also known as Protein kinase B, and represents a predominant signaling molecule in the PI3K cascade pathway (Carpten et al., 2007).Ordinarily, the kinase domain of PI3K activates AKT1 by phosphorylation via the binding of PIP3 or PIP2 to the pleckstrin homology domain embedded in the protein (Vivanco & Sawyers, 2002).The genetic alteration of the AKT1 gene induces a gain of function mutation that   develops a non-functional protein with disrupted cellular processes in an individual that ultimately develops breast, ovarian, colorectal, or leukemic cancer (Carpten et al., 2007;M. S. Kim et al., 2008).The most common genomic alteration in the AKT pleckstrin homology domain is in amino acid position 17, substituted from glutamine to lysine (Yesil€ oz et al., 2017).The prevalence of AKT1 E17K mutation frequency is estimated to be 1.4% to 12.5%, with a mean frequency of 3.1% (Bleeker et al., 2008, p. 1;Carpten et al., 2007;M. S. Kim et al., 2008;Stemke-Hale et al., 2008;Stephens et al., 2012;Troxell, 2012).
The AKT1 E17K mutant manifests the association of AKT1 with the plasma membrane and uninterrupted activation of the enzyme that results in an elevated level of the protein involved in cellular processes, including proliferation and growth factors (Vivanco & Sawyers, 2002).A study conducted on 600 breast cancer patients was analyzed for AKT1 E17K by examining their cfDNA using exome sequencing, beads, amplification, magnetic digital PCR, and emulsions.The data suggested that the prevalence of the mutation was 6.3%, independent of patient age and menopause status.However, the frequency of mutation may vary, mainly the lowest mutation frequency was seen with grade 3 (1.9%)disease than with grade 2 (6%) and grade 1 (11.1%)disease suggesting an inverse relationship and its direct association with mortality (Rudolph et al., 2016).Another study by AACR Project GENIE framework utilized 44 AKT1-mutant patients and 87 AKT1-native to treat them with CDK 4/6 inhibitor with a median DOT of 5.3 months.Correspondingly, mTOR inhibitor such as everolimus was utilized on 49 AKT1-mutant and 97 AKT1-native patients.The study showed that the AKT1 E17K mutation obstructs proliferation, growth, and angiogenesis, although it promotes apoptosis of breast cancer cells.The aforementioned is associated with the disrupted PI3K/AKT/mTOR signaling pathway (Smyth et al., 2020).The inhibitors of AKT in response to cancer are challenging since AKT1 E17K is exceptional in individuals with tumor cell lineages.In a study designed by Astrazeneca, wherein 58 patients with advanced breast cancer were treated with AZD5363 which acts as an ATP-competitive AKT kinase inhibitor.The study involved 52 patients with AKT1 E17K mutation, of which 18, 20, and 20 reported with gynecological cancer, ER þ breast cancer, and other solid tumors.The progression-free survival was estimated for each group as 6.6 months (95% CI, 1.5 to 8.3 months) considering gynecological cancer, 5.5 months (95% CI, 2.9 to 6.9 months) in the case of ER þ breast cancer, and 4.2 months (95% CI, 2.1 to 12.8 months) in other solid tumor patients.The reducing levels of cfDNA with advancing PFS determine the efficacy of AZD5363 in inhibiting AKT1 E17K.The study justifies the relevance of AKT1 E17K in therapeutic targeting against breast cancer (Hyman et al., 2017).In another study with 35 patients with cancer, 86% were females between 32-73 years.About 18 patients had breast cancer, 15 with HR þ or ERRB2; gynecologic cancers were observed in 11 patients, whereas three patients suffered from the triple-negative disease.The treatment was assessed by providing capivasertib (AZD5363) to the patients through oral administration at a concentration of 480 mg.The treatment continued for 28-day cycles afore to reaching toxic effects and was administered twice only for four days a week.The study reached a clinically significant ORR with an overall PFS rate of 6 months for about 50% of the patients.However, grade 3-related adverse effects were manifested in capivasertib-treated patients suffering from AKT1 E17K mutant that included rash in about four patients; hyperglycemia was seen in 8 patients, whereas one patient showed hyperglycemia in grade 4 disease (Kalinsky et al., 2021).
Notably, another breast cancer-inducing alteration is a TP53-derived mutation.In our research, we found a TP53 R175H mutation in the plasma cfDNA of a female breast cancer patient.The TP53 gene encodes for a p53 protein that is referred to as the 'guardian of the cell'.The p53 protein plays a crucial role in DNA repair, apoptosis, and cell-cycle arrest in response to DNA damage or alteration.The disruption in DNA repair or apoptosis caused by altered p53 protein acquires mutations that ultimately develop cancer hallmarks (Lane, 1992).The protein is subdivided into domains the N-terminus transactivation domain, a domain that binds specifically to the DNA sequences referred to as the DNA binding domain; the C-terminus domain, and regulatory domains for monitoring gene expression (Guimaraes & Hainaut, 2002).About 41.8% of cancer patients have shown alterations in the TP53 gene, as suggested by the TCGA program.Nonetheless, the most prevalent is the missense mutations that have been predominantly encountered in the protein's DNA binding domain (96-293 AA) (Bouaoun et al., 2016).Several hotspot regions have been evaluated for mutations in the TP53 gene associated with breast cancers; the TP53 R175H hotspot mutation has the highest prevalence as per the TCGA database.The R175H hotspot mutation at the DNA binding domain interface alters the protein structure, consequently interrupting its stability (Cho et al., 1994).The somatic mutation interacted with several transcription factors that induce or repress targeted gene expression (Stein et al., 2019).Stabilizing the p53 protein in the case of R175H mutation may provide a basis for reactivation of the protein to its normal form and recapitulate its functions (Ang et al., 2006).A study based on the principle of restoration of the p53 protein utilized a drug PRIMA À 1 that improves protein folding in R175H mutation.The result of establishing the usage of the PRIMA-1 drug was an induction of apoptosis to suppress tumor growth by preserving the DNA binding ability of p53.
Additionally, pretreating with cyclohexamide and then administering PRIMA-1 improved the 4-fold efficacy of the drug (Bykov et al., 2002).A recent study on zinc metallochaperones (ZMC) has provided a new approach to treating p53 mutant breast cancers.The ZMC enables p53 mutant protein folding by delivering the zinc ions into the cytoplasm from outside of the cell.Consequently, the ZMC exhibit an advantageous function in restoring the p53 mutant protein enabling its normal function and inducing apoptosis to deteriorate tumor growth (Blanden et al., 2015;Blanden et al., 2015).
Extraordinarily, the instability of the genome due to loss of heterozygosity has been established to screen for breast cancer.At the same time, most alterations are detected in the 'q-arm' region of chromosome 18 (Callahan et al., 1993).One representative tumor suppressor gene in the 18q21 region is the SMAD4 gene, a downstream mediator of the TGF-b signaling pathway (Yokota et al., 1997).The ligandbinding activity may either positively or negatively regulate the gene expression of target genes; however, SMAD4, a signal transducer protein of the TGF-b family, acts as a corepressor of the genes expressed by triggering ERa activity (Smith & O'Malley, 2004;Wu et al., 2003).The involvement of SMAD proteins in signaling responses is persuaded by Activin, TGF-b, and BMP 2/4 pathways (de Caestecker et al., 1997).The transforming growth factor-b family is a serine/threonine kinase receptor that binds TGF-b signaling protein extracellularly and builds a cascade of the mechanism inside the cell to activate target genes for expression (Massagu� e, 1996).The point mutations in the SMAD4 carboxy-terminal end disrupted the protein structure and affected SMAD proteins' phosphorylation and complex formation capability (A.Hata et al., 1998).Intriguingly, the N-terminal domains (NTD) of SMAD proteins bind to DNA; missense mutations are seen in the NTD domain of SMAD to intercept the protein-DNA interaction ensuing ERa targeted transcriptional activity (Dennler et al., 1998;J. Kim et al., 1997;ten Dijke et al., 2000).The disruption of the healthy signaling pathway induces tumorigenesis resulting from impaired TbRII receptors or alterations in the SMAD gene specifically affecting SMAD4 (B€ ottinger et al., 1997;Schutte et al., 1996).The alterations or deletions of SMAD4 are associated with advanced breast cancer; a significant decrease in SMAD4 protein concentration advanced the disease (breast cancer) from grade 1 to grade 3 (Stuelten et al., 2006, p. 4).A recent study confirmed the quintessential role of SMAD4 protein depletion in contributing to the activation of cytokine IL-11.The activation of IL-11 is mediated by SMAD4-dependent activator protein-1 in breast cancer.The enhanced IL-11 is implicated in osteolytic bone metastasis suggesting the switch in the role of Smads from tumor suppressors to prometastatic mediators (Kang et al., 2005).Since no earlier study investigated the structural insights of SMAD4 V465M mutation, we used MD simulation to analyze the structural impact of the SMAD4 mutation.We discovered significant variation in the mutant structure compared to the SMAD4 native.Based on prior studies, our findings suggested that V465M mutation may cause the SMAD4 protein to truncate or inactive, leading to loss of SMAD4 function by blocking TGF's tumor suppressor.Thus, the mutant SMAD4 in breast cancer may increase their migratory and invasive capabilities.
In addition, the GeneMANIA network determined the inter-relationship between the three mutant genes, namely, SMAD4, AKT1, and TP53.Furthermore, these genes are engaged in vital functions such as SMAD binding, cellular response to transforming growth factor beta stimulus, cell cycle arrest, transcription regulator complex, and cell cycle G1/S phase transition regulation.We presented the signaling pathway activated by these genes in normal cells and how it would be disrupted by these mutations based on the interaction and inter-relationship between the genes obtained from the GeneMANIA network.
Figure 5(a) illustrates the signaling pathways activated by SMAD4, AKT1, and TP53.TGF-b ligand activation and binding to type II and I (T-bR II & T-bR I) receptors generate the classical TGF-b/SMAD4 signal, which then phosphorylates SMAD2/3.Phosphorylated SMAD2/3 form a heterodimeric complex with SMAD4 and translocate to the nucleus, where they directly bind to SBE and regulate target gene transcription with the help of transcriptional factors.These target genes are primarily responsible for growth arrest and apoptosis.I-SMAD, such as SMAD6 and SMAD7, primarily operate to block receptor-mediated R-SMAD phosphorylation, preventing complex formation with co-SMAD (Hata & Chen, 2016;Miyazono, 2000;Vander Ark et al., 2018).The PI3K/AKT pathway inhibits T-bR I mediated SMAD3 phosphorylation through its downstream molecule mTOR.AKT can directly phosphorylate FOXO and maintain it in the cytoplasm, preventing it from binding to the promoters of p27 and p21 and suppressing TGF-b/SMAD4 cytostatic signals (Zhang, 2009).PTEN, a tumor suppressor, can inhibit PI3K and thereby reduce AKT activity (Georgescu, 2010).In positive feedback, AKT promotes G1 phase cell cycle progression.AKT stimulates cyclinD1 translation by indirectly activating mTOR.The PI3K/AKT/mTOR pathways regulate cell growth and survival (Vadlakonda et al., 2013).MDM2 and p53 combine to establish a negative feedback loop in which p53 promotes MDM2 expression, which then encourages p53 degradation and inhibits cellular p53 function (Pant et al., 2013).
Figure 5(b) illustrates the signaling pathways disturbed by mutations in SMAD4, AKT1, and TP53.Several TGF-b pathway steps have been reported to be affected by p53 mutations, including the repression of the TGF-bRII gene, the delay or reduction of smad2 phosphorylation by TGF-bRI, the interference with SMAD2/3 and SMAD4 interaction, and the inhibition of SMAD translocation to the nucleus (Kalo et al., 2007).By interfering with the TGF pathway, the SMAD-dependent transcription of target genes was inhibited.Additionally, p53 mutation can lead to loss of DNA-binding ability, which prevents coupling with Smads at gene-specific promoters, reducing transcriptional activation of these genes.Signifying cross-talk between the mutant p53 and TGFb1 signaling pathway has been reported (Cordenonsi et al., 2007;Kalo et al., 2007).AKT1 (E17K) gain-of-function mutations cause aberrant signaling through the PI3K-AKT pathway, altering p27 de-localization and boosting cell proliferation, migration, and invasion to exhibit oncogenic activity (De Marco et al., 2015).SMAD4 acts at the G1/S checkpoint to keep cells in the G1 phase, resulting in cell cycle arrest (Zhao et al., 2018).Mutations in SMAD4 cause the protein to become inactive or truncated, or they can cause it to be degraded by the ubiquitin-proteasome pathway.Loss of SMAD4 promotes a more aggressive phenotype by blocking TGF's tumor suppressor actions without altering tumor responsiveness (Ahmed et al., 2017).SMAD4 loss increases tumor growth by increasing angiogenesis and inflammation, which are mediated by increased VEGF and TGF expression, respectively (Malkoski & Wang, 2012).
Here, we found that these three mutant genes from our study are involved in the negative regulation of macroautophagy based on the ClueGO/CluePedia plug-in Cytoscape.As suggested earlier, oncogenic class I PI3K/AKT signaling negatively affects autophagy (Petiot et al., 2000).Uncontrolled cell development typically results from a lack of apoptosis; in this context, cytotoxic stimuli can activate the class II PI3K/Beclin 1 Pathway, block the mTOR pathway, or the class I PI3K/AKT route to cause autophagic cell death in order to eliminate tumor cells.Autophagy's ability to kill oneself aids in the suppression of tumors.(Mizushima & Levine, 2020).The association between autophagy and cancer is challenging.For instance, signaling of mTOR inhibition, which in turn might trigger autophagy, is connected to the tumor suppressors PTEN and p53 (Botti et al., 2006).However, the tumor-suppressing action of mTOR may be influenced by the regulation of cell growth and proliferation through p70s6 kinase and 4E-BP1, two of its downstream translation effectors (Fingar & Blenis, 2004).Following that KEGG/Reactome pathways analysis, these genes were mainly engaged in the regulation of TP53, longevity regulating pathway, central carbon metabolism in cancer, intrinsic pathway for apoptosis, activation of BH3-only proteins, and many cancer pathways such as colorectal, pancreatic, endometrial, glioma, prostate, melanoma, Chronic myeloid leukemia, small cell lung cancer, non-small cell lung cancer.

Conclusion
We reported three pathogenic mutations in a female breast cancer patient, such as SMAD4 V465M, AKT1 E17K, and TP53 R175H, through gene panel sequencing and in silico pathogenicity analysis.Systematic structural analysis using MD simulation reveals that the SMAD4 V465M mutation was highly deleterious.Our findings showed that SMAD4 V465M mutation act synergistically with AKT1 E17K and TP53 R175H mutation, which might significantly alter the TGF-b signaling pathway and increase breast cancer progression.We also suggested that SMAD4 V465M mutation may inactivate and decrease the SMAD4 protein expression resulting in an invasive and metastatic behavior via inhibiting the TGF-b signaling pathway.Furthermore, we describe that a collection of mutations discovered in cfDNA by NGS can aid in determining the origin of cancer, target therapy, and monitoring breast cancer development & treatment effectiveness.

Figure 1 .
Figure 1.Structural representation of valine (V) substituted with Methionine (M) at codon 465 (amino acid changes) in the SMAD4 protein using the HOPE server.The mutant methionine residue is larger than the residue of the native valine.

Figure 2 .
Figure 2. The black color represents the SMAD4 native; the turquoise color represents the mutant-V465M SMAD4.(a) RMSD plot of the native and mutant V465M of SMAD4 protein.Time (ns) is shown on the X-axis, while RMSD (nm) is depicted on the Y-axis.(b) RMSF plot of the native and mutant V465M of SMAD4.The residue is shown on the X-axis, while RMSF (nm) is depicted on the Y-axis.

Figure 3 .
Figure 3.The black color represents the SMAD4 native; the turquoise color represents the mutant-V465M SMAD4.(a) h-bonds between the protein and protein in the native and mutant V465M of SMAD4.Time (ps) is shown on the X-axis, while the number of h-bonds is depicted on the Y-axis.(b) Rg plot of the native and mutant V465M of SMAD4.Time (ps) is shown on the X-axis, while Rg (nm) is depicted on the Y-axis.(c) 2D projection map of the ca-atom conformational space and atomic number monitoring Native and mutant V465M SMAD4 structures in the crucial subspace along eigenvectors 1 and 2.

Figure 4 .
Figure 4. (a) Cytoscape and GeneMANIA interaction of SMAD4, TP53, and AKT1 genes with physical interactions, co-expression, predicted interaction, co-localization, genetic interactions, pathway, and shared protein domains.The blue color node represents the (SMAD4, TP53, and AKT1) query genes, whereas the pinkish orange color node represents the genes related to the query genes.The circle size represents the bond's strength with the gene under evaluation.The pinkishorange color line represents physical interactions.The purple line represents co-expression, the orange line represents predicted interaction, the dark blue line represents co-localization, the green line represents genetic interactions, and the sky-blue line represents the pathway.The mustard yellow line represents shared protein domains.(b) The functional enrichment network of the three genes (SMAD4, AKT1, and TP53) using Cytoscape's ClueGO/CluePedia plug-in.The green ellipse nodes represent the KEGG/Reactome pathways, and the octagon node represents the GO annotation.The node is interconnected with edges based on the interrelationship with each other.

Figure 5 .
Figure 5. (a) A schematic representation of the signaling pathways that SMAD4, AKT1 activate, and TP53.(b) A schematic representation of the signaling pathways that are disturbed by mutations in SMAD4, AKT1, and TP53.

Table 2 .
The gene cluster network of SMAD4, AKT1, and TP53 genes exhibiting top 20 GO annotated functions predicted using Cytoscape Plugin GeneMANIA.