FocAn: automated 3D analysis of DNA repair foci in image stacks acquired by confocal fluorescence microscopy

Future Perspectives of Artificial Intelligence in Bone Marrow Dosimetry and Individualized Radioligand Therapy

Radioligand therapy is an emerging and effective treatment option for various types of malignancies, but may be intricately linked to hematological side effects such as anemia, lymphopenia or thrombocytopenia. The safety and efficacy of novel theranostic agents, targeting increasingly complex targets, can be well served by comprehensive dosimetry. However, optimization in patient management and patient selection based on risk-factors predicting adverse events and built upon reliable dose-response relations is still an open demand. In this context, artificial intelligence methods, especially machine learning and deep learning algorithms, may play a crucial role. This review provides an overview of upcoming opportunities for integrating artificial intelligence methods into the field of dosimetry in nuclear medicine by improving bone marrow and blood dosimetry accuracy, enabling early identification of potential hematological risk-factors, and allowing for adaptive treatment planning. It will further exemplify inspirational success stories from neighboring disciplines that may be translated to nuclear medicine practices, and will provide conceptual suggestions for future directions. In the future, we expect artificial intelligence-assisted (predictive) dosimetry combined with clinical parameters to pave the way towards truly personalized theranostics in radioligand therapy.

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 ∼ 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.

Multi-scale cellular imaging of DNA double strand break repair

Live-cell and high-resolution fluorescence microscopy are powerful tools to study the organization and dynamics of DNA double-strand break repair foci and specific repair proteins in single cells. This requires specific induction of DNA double-strand breaks and fluorescent markers to follow the DNA lesions in living cells. In this review, where we focused on mammalian cell studies, we discuss different methods to induce DNA double-strand breaks, how to visualize and quantify repair foci in living cells., We describe different (live-cell) imaging modalities that can reveal details of the DNA double-strand break repair process across multiple time and spatial scales. In addition, recent developments are discussed in super-resolution imaging and single-molecule tracking, and how these technologies can be applied to elucidate details on structural compositions or dynamics of DNA double-strand break repair.

Deep learning-assisted high-content screening identifies isoliquiritigenin as an inhibitor of DNA double-strand breaks for preventing doxorubicin-induced cardiotoxicity

BACKGROUND

Anthracyclines including doxorubicin are essential components of many cancer chemotherapy regimens, but their cardiotoxicity severely limits their use. New strategies for treating anthracycline-induced cardiotoxicity (AIC) are still needed. Anthracycline-induced DNA double-strand break (DSB) is the major cause of its cardiotoxicity. However, DSB-based drug screening for AIC has not been performed possibly due to the limited throughput of common assays for detecting DSB. To discover new therapeutic candidates for AIC, here we established a method to rapidly visualize and accurately evaluate the intranuclear anthracycline-induced DSB, and performed a screening for DSB inhibitors.

RESULTS

First, we constructed a cardiomyocyte cell line stably expressing EGFP-53BP1, in which the formation of EGFP-53BP1 foci faithfully marked the doxorubicin-induced DSB, providing a faster and visible approach to detecting DSB. To quantify the DSB, we used a deep learning-based image analysis method, which showed the better ability to distinguish different cell populations undergoing different treatments of doxorubicin or reference compounds, compared with the traditional threshold-based method. Subsequently, we applied the deep learning-assisted high-content screening method to 315 compounds and found three compounds (kaempferol, kaempferide, and isoliquiritigenin) that exert cardioprotective effects in vitro. Among them, the protective effect of isoliquiritigenin is accompanied by the up-regulation of HO-1, down-regulation of peroxynitrite and topo II, and the alleviation of doxorubicin-induced DSB and apoptosis. The results of animal experiments also showed that isoliquiritigenin maintained the myocardial tissue structure and cardiac function in vivo. Moreover, isoliquiritigenin did not affect the killing of HeLa and MDA-MB-436 cancer cells by doxorubicin and thus has the potential to be a lead compound to exert cardioprotective effects without affecting the antitumor effect of doxorubicin.

CONCLUSIONS

Our findings provided a new method for the drug discovery for AIC, which combines phenotypic screening with artificial intelligence. The results suggested that isoliquiritigenin as an inhibitor of DSB may be a promising drug candidate for AIC.

Evaluation of the antibacterial activity of Elsholtzia ciliate essential oil against halitosis-related Fusobacterium nucleatum and Porphyromonas gingivalis

The broad-spectrum antimicrobial activity of Elsholtzia ciliate essential oil (ECO) has been previously reported, but its effectiveness against halitosis-causing bacteria such as Fusobacterium nucleatum and Porphyromonas gingivalis is not well understood. In this study, we investigated the bacteriostatic activity of ECO against planktonic cells and biofilms of F. nucleatum and P. gingivalis, as well as its ability to inhibit bacterial metabolism and production of volatile sulfur compounds (VSCs) at sub-lethal concentrations. Our findings revealed that ECO exhibited comparable activities to chlorhexidine against these oral bacteria. Treatment with ECO significantly reduced the production of VSCs, including hydrogen sulfide, dimethyl disulfide, and methanethiol, which are major contributors to bad breath. As the major chemical components of ECO, carvacrol, p-cymene, and phellandrene, were demonstrated in vitro inhibitory effects on F. nucleatum and P. gingivalis, and their combined use showed synergistic and additive effects, suggesting that the overall activity of ECO is derived from the cumulative or synergistic effect of multiple active components. ECO was found to have a destructive effect on the bacterial cell membrane by examining the cell morphology and permeability. Furthermore, the application of ECO induced significant changes in the bacterial composition of saliva-derived biofilm, resulting in the elimination of bacterial species that contribute to halitosis, including Fusobacterium, Porphyromonas, and Prevotella. These results provide experimental evidence for the potential clinical applications of ECOs in the prevention and treatment of halitosis.

HARLEY mitigates user bias and facilitates efficient quantification and co-localization analyses of foci in yeast fluorescence images

Quantification of cellular structures in fluorescence microscopy data is a key means of understanding cellular function. Unfortunately, numerous cellular structures present unique challenges in their ability to be unbiasedly and accurately detected and quantified. In our studies on stress granules in yeast, users displayed a striking variation of up to 3.7-fold in foci calls and were only able to replicate their results with 62-78% accuracy, when re-quantifying the same images. To facilitate consistent results we developed HARLEY (Human Augmented Recognition of LLPS Ensembles in Yeast), a customizable software for detection and quantification of stress granules in S. cerevisiae. After a brief model training on ~ 20 cells the detection and quantification of foci is fully automated and based on closed loops in intensity contours, constrained only by the a priori known size of the features of interest. Since no shape is implied, this method is not limited to round features, as is often the case with other algorithms. Candidate features are annotated with a set of geometrical and intensity-based properties to train a kernel Support Vector Machine to recognize features of interest. The trained classifier is then used to create consistent results across datasets. For less ambiguous foci datasets, a parametric selection is available. HARLEY is an intuitive tool aimed at yeast microscopy users without much technical expertise. It allows batch processing of foci detection and quantification, and the ability to run various geometry-based and pixel-based colocalization analyses to uncover trends or correlations in foci-related data. HARLEY is open source and can be downloaded from https://github.com/lnilya/harley .

A computational approach to quantifying miscounting of radiation-induced double-strand break immunofluorescent foci

Immunofluorescent tagging of DNA double-strand break (DSB) markers, such as γ-H2AX and other DSB repair proteins, are powerful tools in understanding biological consequences following irradiation. However, whilst the technique is widespread, there are many uncertainties related to its ability to resolve and reliably deduce the number of foci when counting using microscopy. We present a new tool for simulating radiation-induced foci in order to evaluate microscope performance within in silico immunofluorescent images. Simulations of the DSB distributions were generated using Monte Carlo track-structure simulation. For each DSB distribution, a corresponding DNA repair process was modelled and the un-repaired DSBs were recorded at several time points. Corresponding microscopy images for both a DSB and (γ-H2AX) fluorescent marker were generated and compared for different microscopes, radiation types and doses. Statistically significant differences in miscounting were found across most of the tested scenarios. These inconsistencies were propagated through to repair kinetics where there was a perceived change between radiation-types. These changes did not reflect the underlying repair rate and were caused by inconsistencies in foci counting. We conclude that these underlying uncertainties must be considered when analysing images of DNA damage markers to ensure differences observed are real and are not caused by non-systematic miscounting.

PyFoci is a tool that simulates distributions of fluorescently labeled DNA double-strand break marker protein foci and allows the estimation of miscounting under different radiation types, doses and microscopy settings.

A deep learning model (FociRad) for automated detection of γ-H2AX foci and radiation dose estimation

DNA double-strand breaks (DSBs) are the most lethal form of damage to cells from irradiation. γ-H2AX (phosphorylated form of H2AX histone variant) has become one of the most reliable and sensitive biomarkers of DNA DSBs. However, the γ-H2AX foci assay still has limitations in the time consumed for manual scoring and possible variability between scorers. This study proposed a novel automated foci scoring method using a deep convolutional neural network based on a You-Only-Look-Once (YOLO) algorithm to quantify γ-H2AX foci in peripheral blood samples. FociRad, a two-stage deep learning approach, consisted of mononuclear cell (MNC) and γ-H2AX foci detections. Whole blood samples were irradiated with X-rays from a 6 MV linear accelerator at 1, 2, 4 or 6 Gy. Images were captured using confocal microscopy. Then, dose-response calibration curves were established and implemented with unseen dataset. The results of the FociRad model were comparable with manual scoring. MNC detection yielded 96.6% accuracy, 96.7% sensitivity and 96.5% specificity. γ-H2AX foci detection showed very good F1 scores (> 0.9). Implementation of calibration curve in the range of 0-4 Gy gave mean absolute difference of estimated doses less than 1 Gy compared to actual doses. In addition, the evaluation times of FociRad were very short (< 0.5 min per 100 images), while the time for manual scoring increased with the number of foci. In conclusion, FociRad was the first automated foci scoring method to use a YOLO algorithm with high detection performance and fast evaluation time, which opens the door for large-scale applications in radiation triage.

DeepFoci: Deep learning-based algorithm for fast automatic analysis of DNA double-strand break ionizing radiation-induced foci

DNA double-strand breaks (DSBs), marked by ionizing radiation-induced (repair) foci (IRIFs), are the most serious DNA lesions and are dangerous to human health. IRIF quantification based on confocal microscopy represents the most sensitive and gold-standard method in radiation biodosimetry and allows research on DSB induction and repair at the molecular and single-cell levels. In this study, we introduce DeepFoci - a deep learning-based fully automatic method for IRIF counting and morphometric analysis. DeepFoci is designed to work with 3D multichannel data (trained for 53BP1 and γH2AX) and uses U-Net for nucleus segmentation and IRIF detection, together with maximally stable extremal region-based IRIF segmentation. The proposed method was trained and tested on challenging datasets consisting of mixtures of nonirradiated and irradiated cells of different types and IRIF characteristics - permanent cell lines (NHDFs, U-87) and primary cell cultures prepared from tumors and adjacent normal tissues of head and neck cancer patients. The cells were dosed with 0.5-8 Gy γ-rays and fixed at multiple (0-24 h) postirradiation times. Under all circumstances, DeepFoci quantified the number of IRIFs with the highest accuracy among current advanced algorithms. Moreover, while the detection error of DeepFoci remained comparable to the variability between two experienced experts, the software maintained its sensitivity and fidelity across dramatically different IRIF counts per nucleus. In addition, information was extracted on IRIF 3D morphometric features and repair protein colocalization within IRIFs. This approach allowed multiparameter IRIF categorization of single- or multichannel data, thereby refining the analysis of DSB repair processes and classification of patient tumors, with the potential to identify specific cell subclones. The developed software improves IRIF quantification for various practical applications (radiotherapy monitoring, biodosimetry, etc.) and opens the door to advanced DSB focus analysis and, in turn, a better understanding of (radiation-induced) DNA damage and repair.

DNA damage and repair in peripheral blood mononuclear cells after internal ex vivo irradiation of patient blood with 131I

Aim

The aim of this study was to provide a systematic approach to characterize DNA damage induction and repair in isolated peripheral blood mononuclear cells (PBMCs) after internal ex vivo irradiation with [¹³¹I]NaI. In this approach, we tried to mimic ex vivo the irradiation of patient blood in the first hours after radioiodine therapy.

Material and methods

Blood of 33 patients of two centres was collected immediately before radioiodine therapy of differentiated thyroid cancer (DTC) and split into two samples. One sample served as non-irradiated control. The second sample was exposed to ionizing radiation by adding 1 ml of [¹³¹I]NaI solution to 7 ml of blood, followed by incubation at 37 °C for 1 h. PBMCs of both samples were isolated, split in three parts each and (i) fixed in 70% ethanol and stored at − 20 °C directly (0 h) after irradiation, (ii) after 4 h and (iii) 24 h after irradiation and culture in RPMI medium. After immunofluorescence staining microscopically visible co-localizing γ-H2AX + 53BP1 foci were scored in 100 cells per sample as biomarkers for radiation-induced double-strand breaks (DSBs).

Results

Thirty-two of 33 blood samples could be analysed. The mean absorbed dose to the blood in all irradiated samples was 50.1 ± 2.3 mGy. For all time points (0 h, 4 h, 24 h), the average number of γ-H2AX + 53BP1 foci per cell was significantly different when compared to baseline and the other time points. The average number of radiation-induced foci (RIF) per cell after irradiation was 0.72 ± 0.16 at t = 0 h, 0.26 ± 0.09 at t = 4 h and 0.04 ± 0.09 at t = 24 h. A monoexponential fit of the mean values of the three time points provided a decay rate of 0.25 ± 0.05 h⁻¹, which is in good agreement with data obtained from external irradiation with γ- or X-rays.

Conclusion

This study provides novel data about the ex vivo DSB repair in internally irradiated PBMCs of patients before radionuclide therapy. Our findings show, in a large patient sample, that efficient repair occurs after internal irradiation with 50 mGy absorbed dose, and that the induction and repair rate after ¹³¹I exposure is comparable to that of external irradiation with γ- or X-rays.

Opposite effects of the triple target (DNA-PK/PI3K/mTOR) inhibitor PI-103 on the radiation sensitivity of glioblastoma cell lines proficient and deficient in DNA-PKcs

BACKGROUND

Radiotherapy is routinely used to combat glioblastoma (GBM). However, the treatment efficacy is often limited by the radioresistance of GBM cells.

METHODS

Two GBM lines MO59K and MO59J, differing in intrinsic radiosensitivity and mutational status of DNA-PK and ATM, were analyzed regarding their response to DNA-PK/PI3K/mTOR inhibition by PI-103 in combination with radiation. To this end we assessed colony-forming ability, induction and repair of DNA damage by γH2AX and 53BP1, expression of marker proteins, including those belonging to NHEJ and HR repair pathways, degree of apoptosis, autophagy, and cell cycle alterations.

RESULTS

We found that PI-103 radiosensitized MO59K cells but, surprisingly, it induced radiation resistance in MO59J cells. Treatment of MO59K cells with PI-103 lead to protraction of the DNA damage repair as compared to drug-free irradiated cells. In PI-103-treated and irradiated MO59J cells the foci numbers of both proteins was higher than in the drug-free samples, but a large portion of DNA damage was quickly repaired. Another cell line-specific difference includes diminished expression of p53 in MO59J cells, which was further reduced by PI-103. Additionally, PI-103-treated MO59K cells exhibited an increased expression of the apoptosis marker cleaved PARP and increased subG1 fraction. Moreover, irradiation induced a strong G2 arrest in MO59J cells (~ 80% vs. ~ 50% in MO59K), which was, however, partially reduced in the presence of PI-103. In contrast, treatment with PI-103 increased the G2 fraction in irradiated MO59K cells.

CONCLUSIONS

The triple-target inhibitor PI-103 exerted radiosensitization on MO59K cells, but, unexpectedly, caused radioresistance in the MO59J line, lacking DNA-PK. The difference is most likely due to low expression of the DNA-PK substrate p53 in MO59J cells, which was further reduced by PI-103. This led to less apoptosis as compared to drug-free MO59J cells and enhanced survival via partially abolished cell-cycle arrest. The findings suggest that the lack of DNA-PK-dependent NHEJ in MO59J line might be compensated by DNA-PK independent DSB repair via a yet unknown mechanism.

Fully automated counting of DNA damage foci in tumor cell culture: A matter of cell separation

Analysis and quantification of residual, unrepaired DNA double-strand breaks by detecting damage-associated γH2AX or 53BP1 foci is a promising approach to evaluate radiosensitivity or radiosensitization in tumor cells. Manual foci quantification by eye is well-established but unsatisfactory due to inconsistent foci numbers between different observers, lack of information about foci size and intensity and the time-consuming scoring process. Therefore, automated foci counting is an important goal. Several software solutions for automated foci counting in separately acquired fluorescence microscopy images have been established. The AKLIDES NUK technology by Medipan combines automated microscopy and image processing/ counting, enabling affordable high throughput foci analysis as a routine application. Using this machine, automated foci counting is well established for lymphocytes but has not yet been reported for adherent tumor cells with their irregularly shaped nuclei and heterogeneous foci textures. Here we aimed to use the AKLIDES NUK system for adherent tumor cells growing in clusters. We identified cell separation as a critical step to ensure fast and reliable automated nuclei detection. We validated our protocol for the fully automated quantification of (i) the IR-dose dependent increase and (ii) the ATM as well as PARP inhibitor-induced radiosensitization. Collectively, with this protocol the AKLIDES NUK system facilitates cost effective, fast and high throughput quantitative fluorescence microscopic analysis of DNA damage induced foci such as γH2AX and 53BP1 in adherent tumor cells.