Ex vivo radiation sensitivity assessment for individual head and neck cancer patients using deep learning-based automated nuclei and DNA damage foci detection.

Introduction: Tumor biopsy tissue response to ex vivo irradiation is potentially an interesting biomarker for in vivo tumor response, therefore, for treatment personalization. Tumor response ex vivo can be characterized by DNA damage response, expressed by the large-scale presence of DNA damage foci in tumor nuclei. Currently, characterizing tumor nuclei and DNA damage foci is a manual process that takes hours per patient and is subjective to inter-observer variability, which is not feasible in for clinical decision making. Therefore, our goal was to develop a method to automatically segment nuclei and DNA damage foci in tumor tissue samples treated with radiation ex vivo to characterize the DNA damage response, as potential biomarker for in vivo radio-sensitivity. Methods: Oral cavity tumor tissue of 21 patients was irradiated ex vivo (5 or 0 Gy), fixated 2 h post-radiation, and used to develop our method for automated nuclei and 53BP1 foci segmentation. The segmentation model used both deep learning and conventional image-analysis techniques. The training (22 %), validation (22 %), and test set (56 %) consisted of thousands of manually segmented nuclei and foci. The segmentations and number of foci per nucleus in the test set were compared to their ground truths. Results: The automatic nuclei and foci segmentations were highly accurate (Dice = 0.901 and Dice = 0.749, respectively). An excellent correlation (R2 = 0.802) was observed for the foci per nucleus that outperformed reported inter-observation variation. The analysis took âˆ¼ 8 s per image. Conclusion: This model can replace manual foci analysis for ex vivo irradiation of head-and-neck squamous cell carcinoma tissue, reduces the image-analysis time from hours to minutes, avoids the problem of inter-observer variability, enables assessment of multiple images or conditions, and provides additional information about the foci size. Thereby, it allows for reliable and rapid ex vivo radio-sensitivity assessment, as potential biomarker for response in vivo and treatment personalization.

From DNA damage to chromosome aberrations: joining the break.

Despite many years of experimental studies on radiation-induced chromosomal aberrations, and the recent progress in elucidating the molecular mechanisms of the DNA damage response, the link between DNA double-strand break repair and its expression as microscopically visible chromosomal rearrangements remains, in many ways, obscure. Some long standing controversies have partially been resolved to the satisfaction of most investigators, including the linearity of the dose-response for DNA double-strand break induction, the necessity of pairwise interaction of radiogenic damaged sites in the formation of exchange aberrations, and the importance of proximity between lesions in misrejoining. However, the contribution of different molecular DNA repair mechanisms (e.g., alternative end-joining pathways) and their impact on the kinetics of aberration formation is still unclear, as is the definition of "complex" radiogenic damaged sites - in either the chemical or spatial sense - which ostensibly lead to chromosome rearrangements. These topics have been recently debated by molecular biologists and cytogeneticists, whose opinions are summarized in this paper.

Artemis splice defects cause atypical SCID and can be restored in vitro by an antisense oligonucleotide.

Artemis deficiency is known to result in classical T-B- severe combined immunodeficiency (SCID) in case of Artemis null mutations, or Omenn's syndrome in case of hypomorphic mutations in the Artemis gene. We describe two unrelated patients with a relatively mild clinical T-B- SCID phenotype, caused by different homozygous Artemis splice-site mutations. The splice-site mutations concern either dysfunction of a 5' splice-site or an intronic point mutation creating a novel 3' splice-site, resulting in mutated Artemis protein with residual activity or low levels of wild type (WT) Artemis transcripts. During the first 10 years of life, the patients suffered from recurrent infections necessitating antibiotic prophylaxis and intravenous immunoglobulins. Both mutations resulted in increased ionizing radiation sensitivity and insufficient variable, diversity and joining (V(D)J) recombination, causing B-lymphopenia and exhaustion of the naive T-cell compartment. The patient with the novel 3' splice-site had progressive granulomatous skin lesions, which disappeared after stem cell transplantation (SCT). We showed that an alternative approach to SCT can, in principle, be used in this case; an antisense oligonucleotide (AON) covering the intronic mutation restored WT Artemis transcript levels and non-homologous end-joining pathway activity in the patient fibroblasts.

Molecular mechanisms of DNA double strand break repair.

DNA double-strand breaks (DSBs) are major threats to the genomic integrity of cells. If not taken care of properly, they can cause chromosome fragmentation, loss and translocation, possibly resulting in carcinogenesis. Upon DSB formation, cell-cycle checkpoints are triggered and multiple DSB repair pathways can be activated. Recent research on the Nijmegen breakage syndrome, which predisposes patients to cancer, suggests a direct link between activation of cell-cycle checkpoints and DSB repair. Furthermore, the biochemical activities of proteins involved in the two major DSB repair pathways, homologous recombination and DNA end-joining, are now beginning to emerge. This review discusses these new findings and their implications for the mechanisms of DSB repair.

Rejoining of DNA by the RAG1 and RAG2 proteins.

Assembly of immunoglobulin and T cell receptor genes from separate gene segments [V(D)J recombination] begins with DNA double-strand breakage by the RAG1 and RAG2 proteins, acting at a pair of recombination signal sequences (RSSs). Here, the RAG proteins are shown to reverse the cleavage reaction by joining an RSS to a broken coding sequence end. These "hybrid joints" have also been found in lymphoid cells, even when the normal pathway of DNA double-strand break repair is inactive, and can now be explained by this activity of the RAG proteins.

Similarities between initiation of V(D)J recombination and retroviral integration.

In the first step of V(D)J recombination, the RAG1 and RAG2 proteins cleave DNA between a signal sequence and the adjacent coding sequence, generating a blunt signal end and a coding end with a closed hairpin structure. These hairpins are intermediates leading to the formation of assembled antigen receptor genes. It is shown here that the hairpins are formed by a chemical mechanism of direct trans-esterification, very similar to the early steps of transpositional recombination and retroviral integration. A minor variation in the reaction is sufficient to divert the process from transposition to hairpin formation.

Specificity in V(D)J recombination: new lessons from biochemistry and genetics.

Recent in vitro work on V(D)J recombination has helped to clarify its mechanism. The first stage of the reaction, which can be reproduced with the purified RAG1 and RAG2 proteins, is a site-specific cleavage that generates the same broken DNA species found in vivo. The cleavage reaction is closely related to known types of transpositional recombination, such as that of HIV integrase. All the site specificity of V(D)J recombination, including the 12/23 rule, is determined by the RAG proteins. The later steps largely overlap with the repair of radiation-induced DNA double-strand breaks, as indicated by the identity of several newly characterized factors involved in repair. These developments open the way for a thorough biochemical study of V(D)J recombination.

DNA-binding and strand-annealing activities of human Mre11: implications for its roles in DNA double-strand break repair pathways.

DNA double-strand breaks (DSBs) in eukaryotic cells can be repaired by non-homologous end-joining or homologous recombination. The complex containing the Mre11, Rad50 and Nbs1 proteins has been implicated in both DSB repair pathways, even though they are mechanistically different. To get a better understanding of the properties of the human Mre11 (hMre11) protein, we investigated some of its biochemical activities. We found that hMre11 binds both double- and single-stranded (ss)DNA, with a preference for ssDNA. hMre11 does not require DNA ends for efficient binding. Interestingly, hMre11 mediates the annealing of complementary ssDNA molecules. In contrast to the annealing activity of the homologous recombination protein hRad52, the activity of hMre11 is abrogated by the ssDNA binding protein hRPA. We discuss the possible implications of the results for the role(s) of hMre11 in both DSB repair pathways.

Targeted inhibition of metastatic melanoma through interference with Pin1-FOXM1 signaling.

Melanoma is the most lethal form of skin cancer and successful treatment of metastatic melanoma remains challenging. BRAF/MEK inhibitors only show a temporary benefit due to rapid occurrence of resistance, whereas immunotherapy is mainly effective in selected subsets of patients. Thus, there is a need to identify new targets to improve treatment of metastatic melanoma. To this extent, we searched for markers that are elevated in melanoma and are under regulation of potentially druggable enzymes. Here, we show that the pro-proliferative transcription factor FOXM1 is elevated and activated in malignant melanoma. FOXM1 activity correlated with expression of the enzyme Pin1, which we found to be indicative of a poor prognosis. In functional experiments, Pin1 proved to be a main regulator of FOXM1 activity through MEK-dependent physical regulation during the cell cycle. The Pin1-FOXM1 interaction was enhanced by BRAF(V600E), the driver oncogene in the majority of melanomas, and in extrapolation of the correlation data, interference with\ Pin1 in BRAF(V600E)-driven metastatic melanoma cells impaired both FOXM1 activity and cell survival. Importantly, cell-permeable Pin1-FOXM1-blocking peptides repressed the proliferation of melanoma cells in freshly isolated human metastatic melanoma ex vivo and in three-dimensional-cultured patient-derived melanoids. When combined with the BRAF(V600E)-inhibitor PLX4032 a robust repression in melanoid viability was obtained, establishing preclinical value of patient-derived melanoids for prognostic use of drug sensitivity and further underscoring the beneficial effect of Pin1-FOXM1 inhibitory peptides as anti-melanoma drugs. These proof-of-concept results provide a starting point for development of therapeutic Pin1-FOXM1 inhibitors to target metastatic melanoma.

Decreased immunoglobulin class switching in Nijmegen Breakage syndrome due to the DNA repair defect.

Nijmegen breakage syndrome (NBS) is a rare chromosomal-instability syndrome associated with defective DNA repair. Approximately 90% of NBS patients are homozygous for a truncating mutation of the NBS1 gene. As development of the immune system relies on recombination, which involves repair of DNA breaks, one might predict that mutations in the NBS1 gene could cause immunodeficiency. We immunologically investigated the world's largest series of NBS patients (n = 74), confirmed immunodeficiency, and found a discrepancy between relatively normal IgM concentrations, and decreased IgG and IgA concentrations. In addition, a significant relation between low IgA and low IgG levels was found. These data are compatible with a defective class switching in NBS and can be explained by a role of the NBS1 protein in DNA repair, signal transduction, cell cycle regulation or apoptosis.

Non-homologous end-joining, a sticky affair.

Rejoining of broken chromosomes is crucial for cell survival and prevention of malignant transformation. Most mammalian cells rely primarily on the non-homologous end-joining pathway of DNA double-strand break (DSB) repair to accomplish this task. This review focuses both on the core non-homologous end-joining machinery, which consists of DNA-dependent protein kinase and the ligase IV/XRCC4 complex, and on accessory factors that facilitate rejoining of a subset of the DSBs. We discuss how the ATM protein kinase and the Mre11/Rad50/Nbs1 complex might function in DSB repair and what role ionizing radiation-induced foci may play in this process.

The RAG1 and RAG2 proteins establish the 12/23 rule in V(D)J recombination.

V(D)J recombination requires a pair of signal sequences with spacer lengths of 12 and 23 base pairs. Cleavage by the RAG1 AND RAG2 proteins was previously shown to demand only a single signal sequence. Here, we established conditions where 12- and 23-spacer signal sequences are both necessary for cleavage. Coupled cutting at both sites requires only the RAG1 and RAG2 proteins, but depends on the metal ion. In Mn2+, a single signal sequence supports efficient double strand cleavage, but cutting in Mg2+ requires two signal sequences and is best with the canonical 12/23 pair. Thus, the RAG proteins determine both aspects of the specificity of V(D)J recombination, the recognition of a single signal sequence and the correct 12/23 coupling in a pair of signals.

Integration of human immunodeficiency virus types 1 and 2 DNA in vitro by cytoplasmic extracts of Moloney murine leukemia virus-infected mouse NIH 3T3 cells.

An essential step in the life cycle of the human immunodeficiency virus (HIV) is integration of a DNA copy of the viral RNA into the genome of the infected cell. We show here that this step can be faithfully accomplished in vitro by the enzymatic machinery of another retrovirus, Moloney murine leukemia virus (MoMLV). Mini-HIV substrates, which are linearized plasmids with long terminal repeat sequences at their ends, were incubated with cytoplasmic extracts of MoMLV-infected NIH 3T3 cells and target DNA. The MoMLV integration apparatus carried out integration of the mini-HIV substrates correctly; the terminal nucleotides of the viral substrate were removed, and a 4-base-pair duplication of the target DNA flanked the inserted viral DNA (C. Shoemaker, S. P. Goff, E. Gilboa, M. Paskind, S. W. Mitra, and D. Baltimore, Proc. Natl. Acad. Sci. USA 77:3932-3936, 1980). Our experiments show that the substrate sequence requirements for integration in vitro were limited to a few nucleotides, as the similarity between HIV and MoMLV long terminal repeat ends is minimal.

Chromosomal stability and the DNA double-stranded break connection.

Genome stability is of primary importance for the survival and proper functioning of all organisms. Double-stranded breaks in DNA are important threats to genome integrity because they can result in chromosomal aberrations that can affect, simultaneously, many genes, and lead to cell malfunctioning and cell death. These detrimental consequences are counteracted by two mechanistically distinct pathways of double-stranded break repair: homologous recombination and non-homologous end-joining. Recently, unexpected links between these double-stranded break-repair systems, and several human genome instability and cancer predisposition syndromes, have emerged. Now, interactions between both double-stranded break-repair pathways and other cellular processes, such as cell-cycle regulation and replication, are being unveiled.

DNA binding properties of the integrase proteins of human immunodeficiency viruses types 1 and 2.

Integration of retroviral DNA into the host chromosome requires the integrase protein (IN). We overexpressed the IN proteins of human immunodeficiency viruses types 1 and 2 (HIV-1 and HIV-2) in E. coli and purified them. Both proteins were found to specifically cut two nucleotides off the ends of linear viral DNA, and to integrate viral DNA into target DNA. This demonstrates that HIV IN is the only protein required for integration of HIV DNA. Although the two types of IN proteins have only 53% amino acid sequence similarity, they act with equal efficiency on both type 1 and type 2 viral DNA. Binding of IN to DNA was tested: purified IN does not bind very specifically to viral DNA ends. Nevertheless, only viral DNA ends are cleaved and integrated. We interpret this as follows: in vitro quick aspecific binding to DNA is followed by slow specific cutting and integration. IN can not find viral DNA ends in the presence of an excess of aspecific DNA; in vivo this is not required since the IN protein is in constant proximity of viral DNA in the viral core particle.

Human immunodeficiency virus integrase protein requires a subterminal position of its viral DNA recognition sequence for efficient cleavage.

Retroviral integration requires cis-acting sequences at the termini of linear double-stranded viral DNA and a product of the retroviral pol gene, the integrase protein (IN). IN is required and sufficient for generation of recessed 3' termini of the viral DNA (the first step in proviral integration) and for integration of the recessed DNA species in vitro. Human immunodeficiency virus type 1 (HIV-1) IN, expressed in Escherichia coli, was purified to near homogeneity. The substrate sequence requirements for specific cleavage and integration of retroviral DNA were studied in a physical assay, using purified IN and short duplex oligonucleotides that correspond to the termini of HIV DNA. A few point mutations around the IN cleavage site substantially reduced cleavage; most other mutations did not have a drastic effect, suggesting that the sequence requirements are limited. The terminal 15 bp of the retroviral DNA were demonstrated to be sufficient for recognition by IN. Efficient specific cutting of the retroviral DNA by IN required that the cleavage site, the phosphodiester bond at the 3' side of a conserved CA-3' dinucleotide, be located two nucleotides away from the end of the viral DNA; however, low-efficiency cutting was observed when the cleavage site was located one, three, four, or five nucleotides away from the terminus of the double-stranded viral DNA. Increased cleavage by IN was detected when the nucleotides 3' of the CA-3' dinucleotide were present as single-stranded DNA. IN was found to have a strong preference for promoting integration into double-stranded rather than single-stranded DNA.

Identification of amino acids in HIV-2 integrase involved in site-specific hydrolysis and alcoholysis of viral DNA termini.

The human immunodeficiency virus integrase (HIV IN) protein cleaves two nucleotides off the 3' end of viral DNA and subsequently integrates the viral DNA into target DNA. IN exposes a specific phosphodiester bond near the viral DNA end to nucleophilic attack by water or other nucleophiles, such as glycerol or the 3' hydroxyl group of the viral DNA molecule itself. Wild-type IN has a preference for water as the nucleophile; we here describe a class of IN mutants that preferentially use the 3' hydroxyl group of viral DNA as nucleophile. The amino acids that are altered in this class of mutants map near the putative active-site residues Asp-116 and Glu-152. These results support a model in which multiple amino acid side-chains are involved in presentation of the (soluble) nucleophile. IN is probably active as an oligomeric complex, in which the subunits have non-equivalent roles; we here report that nucleophile selection is determined by the subunit that supplies the active site.

Complementation between HIV integrase proteins mutated in different domains.

HIV integrase (IN) cleaves two nucleotides off the 3' end of viral DNA and integrates viral DNA into target DNA. Previously, three functional domains in the HIV IN protein have been identified: (i) the central catalytic domain, (ii) the C-terminal DNA binding domain, and (iii) the N-terminal region, which is also necessary for activity. We have now investigated whether IN proteins mutated in different domains can complement each other. Mutant D116I does not contain an intact active site, but does bind DNA, whereas the C-terminal deletion mutant C delta 73 does not bind DNA, but does have an intact active site. Neither mutant protein mediates site-specific cleavage or integration. However, a mixture of both proteins is active, suggesting that IN functions as an oligomer, and that two subunits can have different functions; one subunit binds the (viral) DNA and another subunit provides the active site. We found three classes of mutants, corresponding to the three domains mentioned above. Mutants from different classes, but not from the same class, can complement each other. However, complementation is most efficient when the N- and C-termini are present on the same molecule.

Mutational analysis of the integrase protein of human immunodeficiency virus type 2.

Purified integrase protein (IN) can nick linear viral DNA at a specific site near the ends and integrate nicked viral DNA into target DNA. We have made a series of 43 site-directed point mutants of human immunodeficiency virus type 2 IN and assayed purified mutant proteins for the following activities: site-specific cleavage of viral DNA (donor cut), integration (strand transfer), and disintegration. In general, the different activities were similarly affected by the mutations. We found three mutations that (almost) totally abolished IN function: Asp-64-->Val, Asp-116-->Ile, and Glu-152-->Leu, whereas 25 mutations did not affect IN function. A few mutations affected the different activities differentially. Near the amino terminus a zinc finger-like sequence motif His-Xaa3-His-Xaa20-30-Cys-Xaa2-Cys is present in all retroviral IN proteins. Two mutations in this region (His-12-->Leu and Cys-40-->Ser) strongly inhibited donor cut but had less effect on strand transfer. The central region of IN is most highly conserved between retroviral INs. Three mutants in this region (Asn-117-->Ile, Asn-120-->Leu, and Lys-159-->Val) were inhibited in strand transfer but were inhibited less strongly in donor cut. Mutation of Asn-120 (to glycine, leucine, or glutamate) resulted in changes in integration-site preference, suggesting that Asn-120 is involved in interactions with target DNA. We did not find a mutant in which one activity was lost and the others were unaffected, supporting the notion that IN has only one active site for the catalysis of donor cut and strand transfer.