Cambridge–MIT Institute

The zebrafish is a well-established model organism used in developmental biology. In the last decade, this technology has been extended to the generation of high-value knowledge on safety risks of novel drugs. Indeed, the larval zebrafish appear to combine advantages of whole organism phenotypic assays and those (rapid production of results with minimal resource engagement) of in vitro high-throughput screening techniques. Thus, if appropriately evaluated, it can offer undeniable advantages in drug discovery for identification of target and off-target effects. Here, we review some applications of zebrafish to identify potential safety liabilities, particularly before lead/candidate selection. For instance, zebrafish cardiovascular system can be used to reveal decreases in heart rate and atrial-ventricular dissociation, which may signal human ether-a-go-go-related gene (hERG) channel blockade. Another main area of interest is the CNS, where zebrafish behavioural assays have been and are further being developed into screening platforms for assessment of locomotor activity, convulsant and proconvulsant liability, cognitive impairment, drug dependence potential and impaired visual and auditory functions. Zebrafish also offer interesting possibilities for evaluating effects on bone density and gastrointestinal function. Furthermore, available knowledge of the renal system in larval zebrafish can allow identification of potential safety issues of drug candidates on this often neglected area in early development platforms. Although additional validation is certainly needed, the zebrafish is emerging as a versatile in vivo animal model to identify off-target effects that need investigation and further clarification early in the drug discovery process to reduce the current, high degree of attrition in development.

Real-time tool segmentation from endoscopic videos is an essential part of many computer-assisted robotic surgical systems and of critical importance in robotic surgical data science. We propose two novel deep learning architectures for automatic segmentation of non-rigid surgical instruments. Both methods take advantage of automated deep-learning-based multi-scale feature extraction while trying to maintain an accurate segmentation quality at all resolutions. The two proposed methods encode the multi-scale constraint inside the network architecture. The first proposed architecture enforces it by cascaded aggregation of predictions and the second proposed network does it by means of a holistically-nested architecture where the loss at each scale is taken into account for the optimization process. As the proposed methods are for real-time semantic labeling, both present a reduced number of parameters. We propose the use of parametric rectified linear units for semantic labeling in these small architectures to increase the regularization of the network while maintaining the segmentation accuracy. We compare the proposed architectures against state-of-the-art fully convolutional networks. We validate our methods using existing benchmark datasets, including ex vivo cases with phantom tissue and different robotic surgical instruments present in the scene. Our results show a statistically significant improved Dice Similarity Coefficient over previous instrument segmentation methods. We analyze our design choices and discuss the key drivers for improving accuracy.

More than 75% of the UK economy is based on services. Knowledge-based services generate more than five times as much value added for the UK economy as advanced manufacturing. Yet, there are persistent gaps in understanding the innovative performance of services. Using Fourth UK Community Innovation Survey (CIS4) data and the results of a detailed case study analysis, this article helps to fill this gap by analysing what innovation in services means and how it can be measured. The traditional indicators of innovation inputs (such as levels of R&D expenditures) and innovation outputs (such as the number of patents) suggest that services are less innovative than other branches of the economy. We take into account a larger spectrum of innovation indicators, both in terms of innovation inputs and outputs, to analyse whether the intensity, nature and economic impact of innovation significantly varies between the manufacturing and service sectors in UK, and between different parts of the services sector such as knowledge-intensive business services and traditional services. The results of the empirical analysis identify the ‘hidden parts’ of innovation in services, that is, the innovative activities and successful innovative outputs that are traditionally underestimated by the use of metrics based on R&D and patents. We suggest a wide range of policy measures specifically targeted at enhancing innovation in services; the UK service economy needs more focus on learning and the training of personnel, and a new balance of policy to support both R&D and non-R&D innovation activities.

Huntington's disease (HD) is an autosomal dominant, neurodegenerative condition caused by a CAG trinucleotide repeat expansion that is translated into an abnormally long polyglutamine tract in the protein huntingtin. Genetic and transgenic studies suggest that the mutation causes disease predominantly via gain-of-function mechanisms. However, loss of normal huntingtin function resulting from the polyglutamine expansion might also contribute to the pathogenesis of HD. Here, we have studied the effects of huntingtin knockdown in zebrafish using morpholino antisense oligonucleotides, as its huntingtin orthologue has 70% amino acid identity with the human protein. Reduced huntingtin levels did not impact on gastrulation and early development, but caused massive apoptosis of neuronal cells by 24 hpf. This was accompanied by impaired neuronal development, resulting in small eyes and heads and enlargement of brain ventricles. Older huntingtin knockdown fish developed lower jaw abnormalities with most branchial arches missing. Molecular analysis revealed that BDNF expression was reduced by approximately 50%. Reduction of BDNF levels by injection of a BDNF morpholino resulted in phenotypes very similar to those seen in huntingtin knockdown zebrafish. The phenotypes of both huntingtin- and BDNF-knockdown zebrafish showed significant rescue when treated with exogenous BDNF protein. This underscores the physiological importance of huntingtin as a regulator of BDNF production and suggests that loss of BDNF is a major cause of the developmental abnormalities seen with huntingtin knockdown in zebrafish. Increasing BDNF expression may represent a useful strategy for Huntington's disease treatment.

Ultrasound images provide the clinician with non-invasive, low cost, and real-time images that can help them in diagnosis, planning and therapy. However, although the human eye is able to derive the meaningful information from these images, automatic processing is very difficult because of the noise and artefacts present in the image. In this work, we propose to extend the current anisotropic diffusion technique to deal with the speckle noise present in the Ultrasound images. To this end, we use a previously derived model of the noise, and we write the restoration scheme as a energy minimization constrained by the noise model and parameters. This approach leads to a new data attachment term whose optimal weight can be automatically estimated.

UNLABELLED: Salt-inducible kinase 2 (SIK2) is a multifunctional kinase of the AMPK family that plays a role in CREB1-mediated gene transcription and was recently reported to have therapeutic potential in ovarian cancer. The expression of this kinase was investigated in prostate cancer clinical specimens. Interestingly, auto-antibodies against SIK2 were increased in the plasma of patients with aggressive disease. Examination of SIK2 in prostate cancer cells found that it functions both as a positive regulator of cell-cycle progression and a negative regulator of CREB1 activity. Knockdown of SIK2 inhibited cell growth, delayed cell-cycle progression, induced cell death, and enhanced CREB1 activity. Expression of a kinase-dead mutant of SIK2 also inhibited cell growth, induced cell death, and enhanced CREB1 activity. Treatment with a small-molecule SIK2 inhibitor (ARN-3236), currently in preclinical development, also led to enhanced CREB1 activity in a dose- and time-dependent manner. Because CREB1 is a transcription factor and proto-oncogene, it was posited that the effects of SIK2 on cell proliferation and viability might be mediated by changes in gene expression. To test this, gene expression array profiling was performed and while SIK2 knockdown or overexpression of the kinase-dead mutant affected established CREB1 target genes; the overlap with transcripts regulated by forskolin (FSK), the adenylate cyclase/CREB1 pathway activator, was incomplete. IMPLICATIONS: This study demonstrates that targeting SIK2 genetically or therapeutically will have pleiotropic effects on cell-cycle progression and transcription factor activation, which should be accounted for when characterizing SIK2 inhibitors.

Glioblastoma (GBM) is characterized by heterogeneous malignant cells that are functionally integrated within the neuroglial microenvironment. In this study, we model this ecosystem by growing GBM into long-term cultured human cortical organoids that contain the major neuroglial cell types found in the cerebral cortex. Single-cell RNA sequencing analysis suggests that, compared with matched gliomasphere models, GBM cortical organoids more faithfully recapitulate the diversity and expression programs of malignant cell states found in patient tumors. Additionally, we observe widespread transfer of GBM transcripts and GFP to nonmalignant cells in the organoids. Mechanistically, this transfer involves extracellular vesicles and is biased toward defined GBM cell states and astroglia cell types. These results extend previous GBM organoid modeling efforts and suggest widespread intercellular transfer in the GBM neuroglial microenvironment. Significance: Models that recapitulate intercellular communications in GBM are limited. In this study, we leverage GBM cortical organoids to characterize widespread mRNA and GFP transfer from malignant to nonmalignant cells in the GBM neuroglial microenvironment. This transfer involves extracellular vesicles, may contribute to reprogramming the microenvironment, and may extend to other cancer types. See related commentary by Shakya et al., p. 261.

Abstract Background: Quantifying circulating nucleic acids is an important new approach to cancer diagnosis/monitoring. Methods: We compared the suitability of serum versus plasma for measuring miRNAs using qRT-PCR and assessed how preanalytic variables that can affect circulating tumor DNA (ctDNA) quantification in plasma also influence miRNA levels. Results: Across 62 blood-derived specimens, plasma samples in EDTA, Streck-DNA, and Streck-RNA tubes showed significantly higher Ct values for multiple housekeeping miRNAs, compared with serum samples. For the EDTA-plasma tubes, this difference was only seen when including the high-speed centrifugation protocol used to optimize ctDNA extraction. In plasma samples derived from blood stored at room temperature for up to 14 days (conditions that typically apply to samples processed for biobanking), levels of endogenous housekeeping miRNAs gradually increased, in parallel with the hemolysis marker hsa-miR-451a, consistent with release from blood cells/platelets. It was necessary to normalize levels of the housekeeping miRNAs to those of hsa-miR-451a, to obtain the stable values needed for referencing test miRNA levels. Conclusions: Our data indicate that plasma samples prepared for ctDNA extraction are suboptimal for miRNA quantification and require the incorporation of multiple data normalization steps. For prospective studies designed to measure both miRNAs and ctDNA, the most suitable approach would be to obtain both serum (for miRNAs) and plasma (for ctDNA). If only plasma can be collected, we recommend an initial low-speed centrifugation step, followed by aliquoting the supernatant into parallel samples, one for direct miRNA quantification, and the other for a further high-speed centrifugation step to optimize ctDNA retrieval. Impact: These recommendations will help “future-proof” clinical studies in which quantification of circulating miRNAs is a component. Cancer Epidemiol Biomarkers Prev; 27(2); 208–18. ©2017 AACR.

BACKGROUND: The General Practice Assessment Questionnaire (GPAQ) has been widely used to assess patient experience in general practice in the UK since 2004. In 2013, new regulations were introduced by the General Medical Council (GMC) requiring UK doctors to undertake periodic revalidation, which includes assessment of patient experience for individual doctors. We describe the development of a new version of GPAQ - GPAQ-R which addresses the GMC's requirements for revalidation as well as additional NHS requirements for surveys that GPs may need to carry out in their own practices. METHODS: Questionnaires were given out by doctors or practice staff after routine consultations in line with the guidance given by the General Medical Council for surveys to be used for revalidation. Data analysis and practice reports were provided independently. RESULTS: Data were analysed for questionnaires from 7258 patients relating to 164 GPs in 29 general practices. Levels of missing data were generally low (typically 4.5-6%). The number of returned questionnaires required to achieve reliability of 0.7 were around 35 for individual doctor communication items and 29 for a composite score based on doctor communication items. This suggests that the responses to GPAQ-R had similar reliability to the GMC's own questionnaire and we recommend 30 completed GPAQ-R questionnaires are sufficient for revalidation purposes. However, where an initial screen raises concern, the survey might be repeated with 50 completed questionnaires in order to increase reliability. CONCLUSIONS: GPAQ-R is a development of a well-established patient experience questionnaire used in general practice in the UK since 2004. This new version can be recommended for use in order to meet the UK General Medical Council's requirements for surveys to be used in revalidation of doctors. It also meets the needs of GPs to ask about patient experience relating to aspects of practice care that are not specific to individual general practitioners (e.g. receptionists, telephone access) which meet other survey requirements of the National Health Service in England. Use of GPAQ-R has the potential to reduce the number of surveys that GPs need to carry out in their practices to meet the various regulatory requirements which they face.

Established methods for imaging the living mammalian brain have, to date, taken optical properties of the tissue as fixed; we here demonstrate that it is possible to modify the optical properties of the brain itself to significantly enhance at-depth imaging while preserving native physiology. Using a small amount of any of several biocompatible materials to raise the refractive index of solutions superfusing the brain prior to imaging, we could increase several-fold the signals from the deepest cells normally visible and, under both one-photon and two-photon imaging, visualize cells previously too dim to see. The enhancement was observed for both anatomical and functional fluorescent reporters across a broad range of emission wavelengths. Importantly, visual tuning properties of cortical neurons in awake mice, and electrophysiological properties of neurons assessed ex vivo, were not altered by this procedure.

Portland cement concrete faces challenges related to durability, strength, and repair costs, and it also contributes to CO₂ emissions. Although sodium silicate and biopolymeric microcapsules have already been reported in the literature, this study focuses on the autonomous self-healing of the cementitious matrix by evaluating the effect of different sodium silicate concentrations encapsulated within biopolymeric microcapsules: MC.A (0%), MC.SS10 (10%), and MC.SS20 (20%). It prepares cement pastes with approximately 16% (v/v) of microcapsules. The characterization of microcapsules employs optical microscopy, particle size distribution analysis, and SEM-EDS, confirming their spherical morphology and the presence of sodium silicate in the core. The characterization of cement pastes and specimens employs rheology, calorimetry, compressive strength tests, water absorption, and SEM-EDS microstructural analysis. The results indicate that incorporating microcapsules does not compromise the fresh or hardened properties of the pastes. Calorimetry shows reductions in maximum heat flow of 54%, 17%, and 16% for MC.A, MC.SS10, and MC.SS20, respectively, compared to the reference sample (REF). The compressive strength test reveals that, except the MC.A, the inclusion of microcapsules does not negatively affect axial compressive strength over time. In pre-cracked specimens, especially MC.SS10 and MC.SS20, strength recovery is comparable and, in some cases, superior to the control samples. Although the effect is more pronounced at early ages and immediately after crack formation, the microcapsules promote a more efficient strength recovery process than systems without their inclusion. This behavior highlights the healing potential of microcapsules, establishing them as a promising alternative for strengthening cracked cementitious materials. • Biopolymeric microcapsules with sodium silicate enhanced autonomous self-healing in cement pastes. • Calorimetry showed reduced heat flow and modified hydration due to microcapsule incorporation. • Sodium silicate microcapsules preserved compressive strength and improved crack recovery. • SEM-EDS confirmed silicate release and healing product formation in cracked regions. • Integrated tests confirmed strength recovery without mechanical performance loss.

In the real world, it is extremely difficult to avoid errors; for instance, a doctor may misdiagnose patients. In other words, databases are never free from data entry or other related errors, and many kinds of mistakes are unavoidable in real world data sets. In existing approaches to pattern recognition, handling noisy data in the learning process always produces better generalization performance than if the noise were ignored. In this article, a novel and adaptive weighting mechanism for noise learning tasks is proposed, especially for boosting learning approaches, preventing the algorithm from concentrating on unreasonably noisy learning samples. Several experiments on UC Irvine Machine Learning Repository and a facial expression data set demonstrate the effectiveness of our method.

Circular RNAs (circRNAs) are covalently closed back-splicing products involved in the regulation of different cellular processes, and their dysregulation has been frequently observed in cancer cells. CircCDYL, a circRNA derived from the back-splicing of CDYL exon 4, has an emerging role in breast cancer (BC) biology. In this study, we investigated the role of circCDYL in modulating alternative splicing (AS) and isoform switching in MCF-7 BC cells. The circRNA profiling in MCF-7 showed circCDYL as the most abundant circRNA, with an expression increasing upon Estrogen Receptor α (ERα) silencing. RNA-Sequencing analysis of circCDYL knock-down cells revealed significant alterations in the splicing pattern, with over 2900 AS events significantly affected. Through RNA immunoprecipitation and RNA pull-down assays, we found evidence of an association between circCDYL and the splicing factor hnRNPL. To explore the consequences of this association, we compared the RNA-Sequencing of hnRNPL-silenced cells, unraveling 96 overlapping AS events accompanied by a switching usage of 223 isoforms, including those of CDYL. The self-loop regulation of circCDYL on its host gene was confirmed by isoform-specific qRT-PCR, observing that it was primarily dependent on an alternative promoter usage, rather than an AS regulation. Accordingly, epigenetic changes at CDYL alternative promoters were confirmed in circCDYL and hnRNPL knockdown cells. The confirmation of a chromatin occupancy of hnRNPL and ERα at CDYL-regulated promoters supported the role of these proteins in CDYL regulation. Our results support a synergic activity of circCDYL and hnRNPL in the regulation of AS and promoter usage in BC cells.

Abstract Global loss of genomic DNA methylation is associated with the progression of cancer. This genome-wide hypomethylation is thought to result from increased rates of genome replication in rapidly dividing cancer cells. We leveraged this aspect of cancer biology to address the challenge of early disease detection in cell-free DNA from a cohort of lung cancer patients. We combined Tagomics’ Activace platform, which enriches unmethylated CpG sites of the genome for sequencing, with genome-wide, fragmentomic analysis on this hypomethylated DNA fraction. Using a single sample input (>5 ng of cell free DNA) Activace enables a multiomic (epigenomic and fragmentomic) read-out with only 70 M sequencing reads. We tested the utility of this approach on a small cohort of 36 patients, who presented at the clinic with symptoms of lung cancer, had a blood sample taken and went on to receive a diagnosis using gold-standard testing (imaging, pathology). The cohort contained 14 cancer-free patients, 12 patients with an early-stage (Stage I & II) diagnosis of non-small cell lung cancer and 10 patients with a late-stage (Stage III & IV) diagnosis of non-small cell lung cancer. We sequenced the cfDNA obtained from the plasma of these patients; and enriched and sequenced the unmethylated DNA fraction from the same patients, using Tagomics’ Activace platform. We compared the fragmentomics profiles of both the enriched, unmethylated DNA and the whole cfDNA and evaluated the impact of the enrichment on our ability to classify the cohort. Using a leave-one-out approach, a logistic regression classifier, trained on the end motif profiles of the Activace dataset (enriched for unmethylated CpG sites), we correctly classified all but two of the early-stage lung cancer cases (91% sensitivity, 100% specificity). By contrast, using an analogous approach in the WGS dataset (whole cfDNA) eight lung cancer cases (seven early-stage and one late-stage) were misclassified (64% sensitivity, 100% specificity). Multiomics tools, such as the Tagomics’ Activace platform enable information-rich insight into disease biology and will likely outperform single analyte-based tests, in the future. The enrichment of DNA containing unmethylated CpG sites improves sensitivity of fragmentomic analysis, relative to analogous analysis of the whole genome, likely due to the enrichment of tumour-derived, hypomethylated DNA fragments in patients with cancer. Citation Format: Valentina Miano, Paulina Siejka-Zielińska, Calum Mould, Sean Knight, Seamus Grundy, Robert K. Neely. Enzymatic enrichment of unmethylated cell-free DNA improves the sensitivity of fragmentomic analysis for the diagnosis of patients with lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3209.

Recent cell type methylome atlases revealed thousands of differentially methylated regions, 97% of which are unmethylated and typically overlap cell type-specific enhancers, as biomarkers that could be used to quantify cell-specific cfDNA. 9,10These developmentally linked loci appear to be universal across individuals regardless of age, sex, or ethnicity and are

Functional screening through systematic deletion, editing or addition of libraries of genes is a powerful approach for discovering gene functions and developing improved molecular tools. However, due to the need for high throughput, such campaigns are typically conducted in vitro, leading to many discoveries, especially tools and therapeutics, which fail to translate in vivo. Tissue context, cellular physiology, and systemic regulation shape both tool performance and gene function in ways that simplified culture systems cannot predict. Pooled in vivo screening methods have the potential to enable screening within living animals while preserving the physiological context, but current approaches using viral vectors face three critical limitations: multi-transgene insertions per cell confound genotype-phenotype association, viral packaging constrains transgene size, and cell-type tropism restricts and biases targeting. Here, we introduce a zebrafish library transgenesis method that overcomes these limitations through delayed site-specific mosaic integration. We exploit a temporal delay between library microinjection with PhiC31 mRNA, and library integration, to allows the library to spread episomally throughout the developing embryo before integration begins. This produces mosaic animals where each cell independently integrates one randomly-selected library member, enforced by a single genomic AttP landing site. We demonstrate delivery of multi-kilobase transgenes with high library coverage of 1,378-1,989 unique integrants per animal, and single-transgene-per-cell in ~99% of brain cells. This method provides a platform for direct in vivo screening of large transgene libraries with single-transgene precision, with potential applications in both biological discovery and tool development.

Abstract Genome-wide methylation patterns are a powerful biomarker. The detection of aberrant methylation has many oncological applications, including using tumor DNA for cancer detection and therapy selection. In liquid biopsies, the ability to detect small amounts of circulating tumor DNA (ctDNA) is crucial to enable earlier diagnosis, monitor minimal residual disease, and assess treatment response. This study evaluated Interlace, an enzymatic technology that enriches unmethylated DNA molecules, to detect ctDNA. This methodology uses an engineered methyltransferase enzyme to target and tag unmethylated CpG sites across the genome. Libraries are then prepared from these tagged regions and sequenced while simultaneously allowing unenriched product to be utilized for downstream molecular applications such as hybrid capture. Interlace was run in parallel with Labcorp Plasma Complete (LPC), a 521-gene next-generation sequencing (NGS) hybrid capture assay that detects genetic alterations in cell free DNA (cfDNA) from the plasma of cancer patients. LPC reports single nucleotide variants (SNVs), insertions and deletions (indels), translocations, copy number amplifications, microsatellite instability (MSI), and blood tumor mutation burden (bTMB). To evaluate the analytical performance of the two tests in parallel, a sample cohort (n=45) including cfDNA reference samples, noncancerous wild-type controls, cancer cell lines, and cfDNA from cancer patients was processed. A head-to-head comparison of LPC results for samples processed with and without the Interlace assay demonstrated 96.7% positive percent agreement and 99.5% positive predictive value. Interlace was able to detect tumor DNA down to 0.1% tumor content in a dilution series created by mixing sheared tumor and wild-type cell line DNA, representative of ctDNA, at target concentrations from 0% to 10%. Using identified differentially methylated regions as input into unsupervised clustering or a proprietary machine learning model, Interlace was able to differentiate blinded cancer cfDNA from cancer cell lines from wild-type samples. Multiomics platforms that evaluate multiple analytes from a single sample, as described here, are powerful tools to provide a more comprehensive view of patient samples. The combination of LPC for somatic variant reporting with Interlace methylation-based ctDNA detection may allow for increased sensitivity in reporting circulating tumor DNA in plasma samples from cancer patients. Citation Format: Kimberly A. Holden, Cynthia Maddox, Calum Mould, Nana Mensah, Luca Tosti, Iwo Pieniak, Paulina Siejka-Zielinska, Stephen Evans, Debora Lucarelli, Robert Neely, Anthony Smith, Kenneth C. Valkenburg, Marcia Eisenberg, Brian Caveney, Eric Severson, Taylor J. Jensen, Shakti Ramkissoon, Jonathan Williams. Enhancing the detection of circulating tumor DNA by combining Interlace™ global methylation discovery with next-generation sequencing test Labcorp Plasma Complete™ [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3212.