Effects of Low- Versus High-Velocity-Loss Thresholds With Similar Training Volume on Maximal Strength and Hypertrophy in Highly Trained Individuals.

AIMS: In the present intervention study, low-velocity-loss (LVL) versus high-velocity-loss (HVL) thresholds in the squat and bench press were compared for changes in muscle strength, power, and hypertrophy. METHODS: Strength-trained volunteers (7♀ and 9♂; age: 27.2 [3.4] y; height: 174.6 [8.0] cm; body mass: 75.3 [10.1] kg) were randomized into an LVL or HVL threshold group (LVL n = 3♀ + 5♂, and HVL n = 4♀ + 4♂). Training took place 3 times per week over 6 weeks (loads: ∼75%-90% of 1-repetition maximum [1RM]). The thresholds of LVLs and HVLs were set at 20% and 40% of maximal velocity, respectively, for the squat, and at 30% and 60%, respectively, for the bench press. Before and after the intervention, 1RM, leg press power, and squat jump were tested. The load (∼45% of 1RM) corresponding to 1-m/s velocity was assessed in all sessions for both exercises. In addition, the thickness of the vastus lateralis and triceps brachii and body composition (dual-energy X-ray absorptiometry [DEXA]) were measured. RESULTS: Squat and bench-press 1RM increased similarly in both groups by 7% to 11% (SD: 4%-6%, P < .05). No group differences were observed for changes in jump height, leg press power, or DEXA lean mass. However, HVL showed a small increase in muscle thickness of the vastus lateralis compared with LVL (6 ± 6% [95% CI] group difference, P < .05). CONCLUSION: For strength-trained individuals, high-volume lower-velocity-loss thresholds were as effective as higher thresholds for improvements in 1RM strength; but local hypertrophy was seemingly elicited faster with higher velocity-loss thresholds.

Deep learning-based image analysis reveals significant differences in the number and distribution of mucosal CD3 and γδ T cells between Crohn's disease and ulcerative colitis.

Colon mucosae of ulcerative colitis (UC) and Crohn's disease (CD) display differences in the number and distribution of immune cells that are difficult to assess by eye. Deep learning-based analysis on whole slide images (WSIs) allows extraction of complex quantitative data that can be used to uncover different inflammatory patterns. We aimed to explore the distribution of CD3 and γδ T cells in colon mucosal compartments in histologically inactive and active inflammatory bowel disease. By deep learning-based segmentation and cell detection on WSIs from a well-defined cohort of CD (n = 37), UC (n = 58), and healthy controls (HCs, n = 33), we quantified CD3 and γδ T cells within and beneath the epithelium and in lamina propria in proximal and distal colon mucosa, defined by the Nancy histological index. We found that inactive CD had significantly fewer intraepithelial γδ T cells than inactive UC, but higher total number of CD3 cells in all compartments than UC and HCs. Disease activity was associated with a massive loss of intraepithelial γδ T cells in UC, but not in CD. The total intraepithelial number of CD3 cells remained constant regardless of disease activity in both CD and UC. There were more mucosal CD3 and γδ T cells in proximal versus distal colon. Oral corticosteroids had an impact on γδ T cell numbers, while age, gender, and disease duration did not. Relative abundance of γδ T cells in mucosa and blood did not correlate. This study reveals significant differences in the total number of CD3 and γδ T cells in particularly the epithelial area between CD, UC, and HCs, and demonstrates useful application of deep segmentation to quantify cells in mucosal compartments.

H2G-Net: A multi-resolution refinement approach for segmentation of breast cancer region in gigapixel histopathological images.

Over the past decades, histopathological cancer diagnostics has become more complex, and the increasing number of biopsies is a challenge for most pathology laboratories. Thus, development of automatic methods for evaluation of histopathological cancer sections would be of value. In this study, we used 624 whole slide images (WSIs) of breast cancer from a Norwegian cohort. We propose a cascaded convolutional neural network design, called H2G-Net, for segmentation of breast cancer region from gigapixel histopathological images. The design involves a detection stage using a patch-wise method, and a refinement stage using a convolutional autoencoder. To validate the design, we conducted an ablation study to assess the impact of selected components in the pipeline on tumor segmentation. Guiding segmentation, using hierarchical sampling and deep heatmap refinement, proved to be beneficial when segmenting the histopathological images. We found a significant improvement when using a refinement network for post-processing the generated tumor segmentation heatmaps. The overall best design achieved a Dice similarity coefficient of 0.933±0.069 on an independent test set of 90 WSIs. The design outperformed single-resolution approaches, such as cluster-guided, patch-wise high-resolution classification using MobileNetV2 (0.872±0.092) and a low-resolution U-Net (0.874±0.128). In addition, the design performed consistently on WSIs across all histological grades and segmentation on a representative × 400 WSI took ~ 58 s, using only the central processing unit. The findings demonstrate the potential of utilizing a refinement network to improve patch-wise predictions. The solution is efficient and does not require overlapping patch inference or ensembling. Furthermore, we showed that deep neural networks can be trained using a random sampling scheme that balances on multiple different labels simultaneously, without the need of storing patches on disk. Future work should involve more efficient patch generation and sampling, as well as improved clustering.

Visual and digital assessment of Ki-67 in breast cancer tissue - a comparison of methods.

BACKGROUND: In breast cancer (BC) Ki-67 cut-off levels, counting methods and inter- and intraobserver variation are still unresolved. To reduce inter-laboratory differences, it has been proposed that cut-off levels for Ki-67 should be determined based on the in-house median of 500 counted tumour cell nuclei. Digital image analysis (DIA) has been proposed as a means to standardize assessment of Ki-67 staining in tumour tissue. In this study we compared digital and visual assessment (VA) of Ki-67 protein expression levels in full-face sections from a consecutive series of BCs. The aim was to identify the number of tumour cells necessary to count in order to reflect the growth potential of a given tumour in both methods, as measured by tumour grade, mitotic count and patient outcome. METHODS: A series of whole sections from 248 invasive carcinomas of no special type were immunohistochemically stained for Ki-67 and then assessed by VA and DIA. Five 100-cell increments were counted in hot spot areas using both VA and DIA. The median numbers of Ki-67 positive tumour cells were used to calculate cut-off levels for Low, Intermediate and High Ki-67 protein expression in both methods. RESULTS: We found that the percentage of Ki-67 positive tumour cells was higher in DIA compared to VA (medians after 500 tumour cells counted were 22.3% for VA and 30% for DIA). While the median Ki-67% values remained largely unchanged across the 100-cell increments for VA, median values were highest in the first 1-200 cells counted using DIA. We also found that the DIA100 High group identified the largest proportion of histopathological grade 3 tumours 70/101 (69.3%). CONCLUSIONS: We show that assessment of Ki-67 in breast tumours using DIA identifies a greater proportion of cases with high Ki-67 levels compared to VA of the same tumours. Furthermore, we show that diagnostic cut-off levels should be calibrated appropriately on the introduction of new methodology.

Code-Free Development and Deployment of Deep Segmentation Models for Digital Pathology.

Application of deep learning on histopathological whole slide images (WSIs) holds promise of improving diagnostic efficiency and reproducibility but is largely dependent on the ability to write computer code or purchase commercial solutions. We present a code-free pipeline utilizing free-to-use, open-source software (QuPath, DeepMIB, and FastPathology) for creating and deploying deep learning-based segmentation models for computational pathology. We demonstrate the pipeline on a use case of separating epithelium from stroma in colonic mucosa. A dataset of 251 annotated WSIs, comprising 140 hematoxylin-eosin (HE)-stained and 111 CD3 immunostained colon biopsy WSIs, were developed through active learning using the pipeline. On a hold-out test set of 36 HE and 21 CD3-stained WSIs a mean intersection over union score of 95.5 and 95.3% was achieved on epithelium segmentation. We demonstrate pathologist-level segmentation accuracy and clinical acceptable runtime performance and show that pathologists without programming experience can create near state-of-the-art segmentation solutions for histopathological WSIs using only free-to-use software. The study further demonstrates the strength of open-source solutions in its ability to create generalizable, open pipelines, of which trained models and predictions can seamlessly be exported in open formats and thereby used in external solutions. All scripts, trained models, a video tutorial, and the full dataset of 251 WSIs with ~31 k epithelium annotations are made openly available at https://github.com/andreped/NoCodeSeg to accelerate research in the field.

AID expression in B-cell lymphomas causes accumulation of genomic uracil and a distinct AID mutational signature.

The most common mutations in cancer are C to T transitions, but their origin has remained elusive. Recently, mutational signatures of APOBEC-family cytosine deaminases were identified in many common cancers, suggesting off-target deamination of cytosine to uracil as a common mutagenic mechanism. Here we present evidence from mass spectrometric quantitation of deoxyuridine in DNA that shows significantly higher genomic uracil content in B-cell lymphoma cell lines compared to non-lymphoma cancer cell lines and normal circulating lymphocytes. The genomic uracil levels were highly correlated with AID mRNA and protein expression, but not with expression of other APOBECs. Accordingly, AID knockdown significantly reduced genomic uracil content. B-cells stimulated to express endogenous AID and undergo class switch recombination displayed a several-fold increase in total genomic uracil, indicating that B cells may undergo widespread cytosine deamination after stimulation. In line with this, we found that clustered mutations (kataegis) in lymphoma and chronic lymphocytic leukemia predominantly carry AID-hotspot mutational signatures. Moreover, we observed an inverse correlation of genomic uracil with uracil excision activity and expression of the uracil-DNA glycosylases UNG and SMUG1. In conclusion, AID-induced mutagenic U:G mismatches in DNA may be a fundamental and common cause of mutations in B-cell malignancies.

Error-free versus mutagenic processing of genomic uracil--relevance to cancer.

Genomic uracil is normally processed essentially error-free by base excision repair (BER), with mismatch repair (MMR) as an apparent backup for U:G mismatches. Nuclear uracil-DNA glycosylase UNG2 is the major enzyme initiating BER of uracil of U:A pairs as well as U:G mismatches. Deficiency in UNG2 results in several-fold increases in genomic uracil in mammalian cells. Thus, the alternative uracil-removing glycosylases, SMUG1, TDG and MBD4 cannot efficiently complement UNG2-deficiency. A major function of SMUG1 is probably to remove 5-hydroxymethyluracil from DNA with general back-up for UNG2 as a minor function. TDG and MBD4 remove deamination products U or T mismatched to G in CpG/mCpG contexts, but may have equally or more important functions in development, epigenetics and gene regulation. Genomic uracil was previously thought to arise only from spontaneous cytosine deamination and incorporation of dUMP, generating U:G mismatches and U:A pairs, respectively. However, the identification of activation-induced cytidine deaminase (AID) and other APOBEC family members as DNA-cytosine deaminases has spurred renewed interest in the processing of genomic uracil. Importantly, AID triggers the adaptive immune response involving error-prone processing of U:G mismatches, but also contributes to B-cell lymphomagenesis. Furthermore, mutational signatures in a substantial fraction of other human cancers are consistent with APOBEC-induced mutagenesis, with U:G mismatches as prime suspects. Mutations can be caused by replicative polymerases copying uracil in U:G mismatches, or by translesion polymerases that insert incorrect bases opposite abasic sites after uracil-removal. In addition, kataegis, localized hypermutations in one strand in the vicinity of genomic rearrangements, requires APOBEC protein, UNG2 and translesion polymerase REV1. What mechanisms govern error-free versus error prone processing of uracil in DNA remains unclear. In conclusion, genomic uracil is an essential intermediate in adaptive immunity and innate antiviral responses, but may also be a fundamental cause of a wide range of malignancies.

Multiple microRNAs may regulate the DNA repair enzyme uracil-DNA glycosylase.

Human nuclear uracil-DNA glycosylase UNG2 is essential for post-replicative repair of uracil in DNA, and UNG2 protein and mRNA levels rapidly decline in G2/M phase. Previous work has demonstrated regulation of UNG2 at the transcriptional level, as well as by protein phosphorylation and ubiquitylation. UNG2 mRNA, encoded by the UNG gene, contains a long 3'untranslated region (3'UTR) of previously unknown function. Here, we demonstrate that several conserved regions in the 3'UTR are potential seed sites for microRNAs (miRNAs), such as miR-16, miR-34c, and miR-199a. Our results show that these miRNAs down-regulate UNG activity, UNG mRNA, and UNG protein levels. Down-regulation was dependent on the 3'UTR, indicating that the miRNAs directly target the conserved seed sites in the 3'UTR. These results add miRNAs as a new modality to UNG's increasing list of complex regulatory mechanisms.

UNG-initiated base excision repair is the major repair route for 5-fluorouracil in DNA, but 5-fluorouracil cytotoxicity depends mainly on RNA incorporation.

Cytotoxicity of 5-fluorouracil (FU) and 5-fluoro-2'-deoxyuridine (FdUrd) due to DNA fragmentation during DNA repair has been proposed as an alternative to effects from thymidylate synthase (TS) inhibition or RNA incorporation. The goal of the present study was to investigate the relative contribution of the proposed mechanisms for cytotoxicity of 5-fluoropyrimidines. We demonstrate that in human cancer cells, base excision repair (BER) initiated by the uracil-DNA glycosylase UNG is the major route for FU-DNA repair in vitro and in vivo. SMUG1, TDG and MBD4 contributed modestly in vitro and not detectably in vivo. Contribution from mismatch repair was limited to FU:G contexts at best. Surprisingly, knockdown of individual uracil-DNA glycosylases or MSH2 did not affect sensitivity to FU or FdUrd. Inhibitors of common steps of BER or DNA damage signalling affected sensitivity to FdUrd and HmdUrd, but not to FU. In support of predominantly RNA-mediated cytotoxicity, FU-treated cells accumulated ~3000- to 15 000-fold more FU in RNA than in DNA. Moreover, FU-cytotoxicity was partially reversed by ribonucleosides, but not deoxyribonucleosides and FU displayed modest TS-inhibition compared to FdUrd. In conclusion, UNG-initiated BER is the major route for FU-DNA repair, but cytotoxicity of FU is predominantly RNA-mediated, while DNA-mediated effects are limited to FdUrd.

Uracil-DNA glycosylase in base excision repair and adaptive immunity: species differences between man and mouse.

Genomic uracil is a DNA lesion but also an essential key intermediate in adaptive immunity. In B cells, activation-induced cytidine deaminase deaminates cytosine to uracil (U:G mispairs) in Ig genes to initiate antibody maturation. Uracil-DNA glycosylases (UDGs) such as uracil N-glycosylase (UNG), single strand-selective monofunctional uracil-DNA glycosylase 1 (SMUG1), and thymine-DNA glycosylase remove uracil from DNA. Gene-targeted mouse models are extensively used to investigate the role of these enzymes in DNA repair and Ig diversification. However, possible species differences in uracil processing in humans and mice are yet not established. To address this, we analyzed UDG activities and quantities in human and mouse cell lines and in splenic B cells from Ung(+/+) and Ung(-/-) backcrossed mice. Interestingly, human cells displayed ∼15-fold higher total uracil excision capacity due to higher levels of UNG. In contrast, SMUG1 activity was ∼8-fold higher in mouse cells, constituting ∼50% of the total U:G excision activity compared with less than 1% in human cells. In activated B cells, both UNG and SMUG1 activities were at levels comparable with those measured for mouse cell lines. Moreover, SMUG1 activity per cell was not down-regulated after activation. We therefore suggest that SMUG1 may work as a weak backup activity for UNG2 during class switch recombination in Ung(-/-) mice. Our results reveal significant species differences in genomic uracil processing. These findings should be taken into account when mouse models are used in studies of uracil DNA repair and adaptive immunity.

Direct interaction between XRCC1 and UNG2 facilitates rapid repair of uracil in DNA by XRCC1 complexes.

Uracil-DNA glycosylase, UNG2, interacts with PCNA and initiates post-replicative base excision repair (BER) of uracil in DNA. The DNA repair protein XRCC1 also co-localizes and physically interacts with PCNA. However, little is known about whether UNG2 and XRCC1 directly interact and participate in a same complex for repair of uracil in replication foci. Here, we examine localization pattern of these proteins in live and fixed cells and show that UNG2 and XRCC1 are likely in a common complex in replication foci. Using pull-down experiments we demonstrate that UNG2 directly interacts with the nuclear localization signal-region (NLS) of XRCC1. Western blot and functional analysis of immunoprecipitates from whole cell extracts prepared from S-phase enriched cells demonstrate the presence of XRCC1 complexes that contain UNG2 in addition to separate XRCC1 and UNG2 associated complexes with distinct repair features. XRCC1 complexes performed complete repair of uracil with higher efficacy than UNG2 complexes. Based on these results, we propose a model for a functional role of XRCC1 in replication associated BER of uracil.

Uracil in DNA and its processing by different DNA glycosylases.

Uracil in DNA may result from incorporation of dUMP during replication and from spontaneous or enzymatic deamination of cytosine, resulting in U:A pairs or U:G mismatches, respectively. Uracil generated by activation-induced cytosine deaminase (AID) in B cells is a normal intermediate in adaptive immunity. Five mammalian uracil-DNA glycosylases have been identified; these are mitochondrial UNG1 and nuclear UNG2, both encoded by the UNG gene, and the nuclear proteins SMUG1, TDG and MBD4. Nuclear UNG2 is apparently the sole contributor to the post-replicative repair of U:A lesions and to the removal of uracil from U:G contexts in immunoglobulin genes as part of somatic hypermutation and class-switch recombination processes in adaptive immunity. All uracil-DNA glycosylases apparently contribute to U:G repair in other cells, but they are likely to have different relative significance in proliferating and non-proliferating cells, and in different phases of the cell cycle. There are also some indications that there may be species differences in the function of the uracil-DNA glycosylases.

Uracil-DNA glycosylases SMUG1 and UNG2 coordinate the initial steps of base excision repair by distinct mechanisms.

DNA glycosylases UNG and SMUG1 excise uracil from DNA and belong to the same protein superfamily. Vertebrates contain both SMUG1 and UNG, but their distinct roles in base excision repair (BER) of deaminated cytosine (U:G) are still not fully defined. Here we have examined the ability of human SMUG1 and UNG2 (nuclear UNG) to initiate and coordinate repair of U:G mismatches. When expressed in Escherichia coli cells, human UNG2 initiates complete repair of deaminated cytosine, while SMUG1 inhibits cell proliferation. In vitro, we show that SMUG1 binds tightly to AP-sites and inhibits AP-site cleavage by AP-endonucleases. Furthermore, a specific motif important for the AP-site product binding has been identified. Mutations in this motif increase catalytic turnover due to reduced product binding. In contrast, the highly efficient UNG2 lacks product-binding capacity and stimulates AP-site cleavage by APE1, facilitating the two first steps in BER. In summary, this work reveals that SMUG1 and UNG2 coordinate the initial steps of BER by distinct mechanisms. UNG2 is apparently adapted to rapid and highly coordinated repair of uracil (U:G and U:A) in replicating DNA, while the less efficient SMUG1 may be more important in repair of deaminated cytosine (U:G) in non-replicating chromatin.