A Deep Learning Model for Cancer Type Prediction Sets a New Standard.

SUMMARY: Classifying tumor types using machine learning approaches is not always trivial, particularly for challenging cases such as cancers of unknown primary. In this issue of Cancer Discovery, Darmofal and colleagues describe a new tool that uses information from a clinical sequencing panel to diagnose tumor type, and show that the model is particularly robust. See related article by Darmofal et al., p. 1064 (1).

Left atrial strain predicts improvement in left atrial functions of severe rheumatic mitral stenoses undergoing successful percutaneous transmitral commissurotomy.

AIMS: Chronic rheumatic heart disease (RHD) is prevalent in India. The mitral valve in isolation or combination with the aortic or tricuspid valve is involved in 31.6% and 52.8% of chronic RHD patients, respectively. The left atrium (LA) functions as a reservoir during the cardiac cycle. Therefore, the LA enlargement leads to longitudinal lengthening, measured as a positive strain, permitting the measurement of the longitudinal strain of LA. This study aimed to assess the LA functions using peak atrial longitudinal strain (PALS) in patients with severe rheumatic mitral stenoses (MS) in sinus rhythm who underwent successful percutaneous transvenous mitral commissurotomy (PTMC). MATERIAL AND METHODS: We recruited 56 patients with severe rheumatic MS for the study, of which 06 PTMC procedures were considered unsuccessful. So, 50 patients of chronic severe rheumatic MS in sinus rhythm undergoing PTMC were enrolled in a tertiary care centre of the Armed Forces from August 2017 to May 2019. Patients included in the study were not consecutive, and patients with atrial fibrillation (AF) were excluded. RESULTS: PALS improved post-PTMC (P < .001) in this study, effectively concluding that PALS is impaired in patients with severe symptomatic MS and is acutely enhanced after treatment. CONCLUSIONS: PALS is a good indicator of LA function and may predict the success of PTMC on the rheumatic mitral valve.

Transcription factor antagonism regulates heterogeneity in embryonic stem cell states.

Gene expression heterogeneity underlies cell states and contributes to developmental robustness. While heterogeneity can arise from stochastic transcriptional processes, the extent to which it is regulated is unclear. Here, we characterize the regulatory program underlying heterogeneity in murine embryonic stem cell (mESC) states. We identify differentially active and transcribed enhancers (DATEs) across states. DATEs regulate differentially expressed genes and are distinguished by co-binding of transcription factors Klf4 and Zfp281. In contrast to other factors that interact in a positive feedback network stabilizing mESC cell-type identity, Klf4 and Zfp281 drive opposing transcriptional and chromatin programs. Abrogation of factor binding to DATEs dampens variation in gene expression, and factor loss alters kinetics of switching between states. These results show antagonism between factors at enhancers results in gene expression heterogeneity and formation of cell states, with implications for the generation of diverse cell types during development.

Developmental Deconvolution for Classification of Cancer Origin.

Cancer is partly a developmental disease, with malignancies named based on cell or tissue of origin. However, a systematic atlas of tumor origins is lacking. Here we map the single-cell organogenesis of 56 developmental trajectories to the transcriptomes of over 10,000 tumors across 33 cancer types. We deconvolute tumor transcriptomes into signals for individual developmental trajectories. Using these signals as inputs, we construct a developmental multilayer perceptron (D-MLP) classifier that outputs cancer origin. D-MLP (ROC-AUC: 0.974 for top prediction) outperforms benchmark classifiers. We analyze tumors from patients with cancer of unknown primary (CUP), selecting the most difficult cases in which extensive multimodal workup yielded no definitive tumor type. Interestingly, CUPs form groups distinguished by developmental trajectories, and classification reveals diagnosis for patient tumors. Our results provide an atlas of tumor developmental origins, provide a tool for diagnostic pathology, and suggest developmental classification may be a useful approach for patient tumors. SIGNIFICANCE: Here we map the developmental trajectories of tumors. We deconvolute tumor transcriptomes into signals for mammalian developmental programs and use this information to construct a deep learning classifier that outputs tumor type. We apply the classifier to CUP and reveal the developmental origins of patient tumors. See related commentary by Wang, p. 2498. This article is highlighted in the In This Issue feature, p. 2483.

Implementation and Clinical Adoption of Precision Oncology Workflows Across a Healthcare Network.

BACKGROUND: Precision oncology relies on molecular diagnostics, and the value-proposition of modern healthcare networks promises a higher standard of care across partner sites. We present the results of a clinical pilot to standardize precision oncology workflows. METHODS: Workflows are defined as the development, roll-out, and updating of disease-specific molecular order sets. We tracked the timeline, composition, and effort of consensus meetings to define the combination of molecular tests. To assess clinical impact, we examined order set adoption over a two-year period (before and after roll-out) across all gastrointestinal and hepatopancreatobiliary (GI) malignancies, and by provider location within the network. RESULTS: Development of 12 disease center-specific order sets took ~9 months, and the average number of tests per indication changed from 2.9 to 2.8 (P = .74). After roll-out, we identified significant increases in requests for GI patients (17%; P < .001), compliance with testing recommendations (9%; P < .001), and the fraction of "abnormal" results (6%; P < .001). Of 1088 GI patients, only 3 received targeted agents based on findings derived from non-recommended orders (1 before and 2 after roll-out); indicating that our practice did not negatively affect patient treatments. Preliminary analysis showed 99% compliance by providers in network sites, confirming the adoption of the order sets across the network. CONCLUSION: Our study details the effort of establishing precision oncology workflows, the adoption pattern, and the absence of harm from the reduction of non-recommended orders. Establishing a modifiable communication tool for molecular testing is an essential component to optimize patient care via precision oncology.

Diagnostic Value of MAML2 Rearrangements in Mucoepidermoid Carcinoma.

Mucoepidermoid carcinoma (MEC) is often seen in salivary glands and can harbor MAML2 translocations (MAML2+). The translocation status has diagnostic utility as an objective confirmation of the MEC diagnosis, for example, when distinction from the more aggressive adenosquamous carcinoma (ASC) is not straightforward. To assess the diagnostic relevance of MAML2, we examined our 5-year experience in prospective testing of 8106 solid tumors using RNA-seq panel testing in combinations with a two-round Delphi-based scenario survey. The prevalence of MAML2+ across all tumors was 0.28% (n = 23/8106) and the majority of MAML2+ cases were found in head and neck tumors (78.3%), where the overall prevalence was 5.9% (n = 18/307). The sensitivity of MAML2 for MEC was 60% and most cases (80%) were submitted for diagnostic confirmation; in 24% of cases, the MAML2 results changed the working diagnosis. An independent survey of 15 experts showed relative importance indexes of 0.8 and 0.65 for "confirmatory MAML2 testing" in suspected MEC and ASC, respectively. Real-world evidence confirmed that the added value of MAML2 is a composite of an imperfect confirmation test for MEC and a highly specific exclusion tool for the diagnosis of ASC. Real-world evidence can help move a rare molecular-genetic biomarker from an emerging tool to the clinic.

Acquired resistance to PRMT5 inhibition induces concomitant collateral sensitivity to paclitaxel.

Epigenetic regulators play key roles in cancer and are increasingly being targeted for treatment. However, for many, little is known about mechanisms of resistance to the inhibition of these regulators. We have generated a model of resistance to inhibitors of protein arginine methyltransferase 5 (PRMT5). This study was conducted in KrasG12D;Tp53-null lung adenocarcinoma (LUAD) cell lines. Resistance to PRMT5 inhibitors (PRMT5i) arose rapidly, and barcoding experiments showed that this resulted from a drug-induced transcriptional state switch, not selection of a preexisting population. This resistant state is both stable and conserved across variants arising from distinct LUAD lines. Moreover, it brought with it vulnerabilities to other chemotherapeutics, especially the taxane paclitaxel. This paclitaxel sensitivity depended on the presence of stathmin 2 (STMN2), a microtubule regulator that is specifically expressed in the resistant state. Remarkably, STMN2 was also essential for resistance to PRMT5 inhibition. Thus, a single gene is required for both acquisition of resistance to PRMT5i and collateral sensitivity to paclitaxel in our LUAD cells. Accordingly, the combination of PRMT5i and paclitaxel yielded potent and synergistic killing of the murine LUAD cells. Importantly, the synergy between PRMT5i and paclitaxel also extended to human cancer cell lines. Finally, analysis of The Cancer Genome Atlas patient data showed that high STMN2 levels correlate with complete regression of tumors in response to taxane treatment. Collectively, this study reveals a recurring mechanism of PRMT5i resistance in LUAD and identifies collateral sensitivities that have potential clinical relevance.

Lineages of embryonic stem cells show non-Markovian state transitions.

Pluripotent embryonic stem cells (ESCs) constitute the cell types of the adult vertebrate through a series of developmental state transitions. These states can be defined by expression levels of marker genes, such as Nanog and Sox2. In culture, ESCs reversibly transition between states. However, whether ESCs retain memory of their previous states or transition in a memoryless (Markovian) process remains relatively unknown. Here, we show some highly dynamic lineages of ESCs do not exhibit the Markovian property: their previous states and kin relations influence future choices. Unexpectedly, the distribution of lineages across their composition between states is constant over time, contrasting with the predictions of a Markov model. Additionally, highly dynamic ESC lineages show skewed cell fate distributions after retinoic acid differentiation. Together, these data suggest ESC lineage is an important variable governing future cell states, with implications for stem cell function and development.

MicroRNAs organize intrinsic variation into stem cell states.

Pluripotent embryonic stem cells (ESCs) contain the potential to form a diverse array of cells with distinct gene expression states, namely the cells of the adult vertebrate. Classically, diversity has been attributed to cells sensing their position with respect to external morphogen gradients. However, an alternative is that diversity arises in part from cooption of fluctuations in the gene regulatory network. Here we find ESCs exhibit intrinsic heterogeneity in the absence of external gradients by forming interconverting cell states. States vary in developmental gene expression programs and display distinct activity of microRNAs (miRNAs). Notably, miRNAs act on neighborhoods of pluripotency genes to increase variation of target genes and cell states. Loss of miRNAs that vary across states reduces target variation and delays state transitions, suggesting variable miRNAs organize and propagate variation to promote state transitions. Together these findings provide insight into how a gene regulatory network can coopt variation intrinsic to cell systems to form robust gene expression states. Interactions between intrinsic heterogeneity and environmental signals may help achieve developmental outcomes.

RNAi screens of lysosomal trafficking.

Here, we describe the general principles of RNA interference screens to study lysosomal functions in mammalian cells. Lysosomes occupy a central position in the biology of numerous processes such as degradation, microbial killing, and immunological antigen presentation to T cells. Selection of a screening system, conducting pooled versus arrayed screens, and appropriate steps in assay development, validation, and verification of novel gene candidates are all discussed. We focus on our experience in developing an arrayed short hairpin RNA screen to identify novel lysosomal trafficking proteins involved in vesicle and cargo trafficking and illustrate how such a trafficking library can be applied to screens involving lysosomes.

MHC class II presentation is controlled by the lysosomal small GTPase, Arl8b.

Dendritic cells (DCs) are specialized APCs with the ability to prime naive T cells. DCs first sample Ags from the environment and then orchestrate their processing and loading onto MHC class II (MHC II) Ag-presenting molecules in lysosomes. Once MHC II molecules have bound a peptide, the MHC II-peptide complex is delivered to the cell surface for presentation to CD4(+) T cells. Regulation of Ag uptake via macropinocytosis and phagocytosis has been extensively studied, as well as trafficking in early endocytic vesicles notably regulated by the small GTPase Rab5 and its effectors. However, little is known about the regulators of Ag delivery from early endosomes to lysosomal compartments where the proper pH, proteases, MHC II, invariant chain, and HLA-DM reside, awaiting exogenous Ags for loading. In this article, we report the crucial role of the small GTPase ADP-ribosylation factor-like 8b (Arl8b) in MHC II presentation in DCs. We show for the first time, to our knowledge, that Arl8b localizes to MHC II compartments in DCs and regulates formation of MHC II-peptide complexes. Arl8b-silenced DCs display a defect in MHC II-Ag complex formation and its delivery to the cell surface during infection resulting in a defect in T cell recognition. Our results highlight the role of Arl8b as a trafficking regulator of the late stage of complex formation and MHC II presentation in DCs.

Use of fetal echocardiography for characterization of fetal cardiac structure in women with normal pregnancies and gestational diabetes mellitus.

OBJECTIVES: To assess fetal cardiac structure and function and to evaluate the efficacy of routine fetal echocardiography for detection of fetal cardiac abnormalities in women with normal pregnancies and those with gestational diabetes mellitus. METHODS: In this prospective study, we studied fetal cardiac structure and function in 294 uncomplicated singleton pregnancies and 302 pregnancies complicated by gestational diabetes. Fetal echocardiography (2-dimensional sonography and pulsed wave Doppler imaging) was used to assess functional parameters and to detect any cardiac structural abnormality. Data from clinical and echocardiographic evaluations of neonates at birth and 3 months were obtained to confirm the antenatal findings. RESULTS: The mean maternal age ± SD was 28.9 ± 5.0 years in the diabetes group and was comparable to that of women with normal pregnancies. The mean hemoglobin A1c value was 6.3%, and the mean body mass index was 28.0 kg/m(2). The systolic function as assessed by the ejection fraction increased significantly in the diabetes group compared to the normal pregnancy group independent of glycemic control (P < .001). The pulsed wave parameters (early diastolic peak flow velocity and early-to-late diastolic peak flow velocity ratio) were significantly different between the groups (P < .001). The interventricular septum and fetal ventricular wall thicknesses were significantly increased in the presence of gestational diabetes (P < .001). No major fetal cardiac structural anomaly was detected in either group. On follow-up after delivery, all neonates were assessed clinically and by transthoracic echocardiography to rule out congenital defects. CONCLUSIONS: In our study, significant increases in the interventricular septum and ventricular wall thicknesses were detected in the presence of gestational diabetes mellitus. Interestingly, none of the neonates of pregnant women with gestational diabetes were found to have echocardiographic evidence of congenital heart disease.

Arf-like GTPase Arl8b regulates lytic granule polarization and natural killer cell-mediated cytotoxicity.

Natural killer (NK) lymphocytes contain lysosome-related organelles (LROs), known as lytic granules, which upon formation of immune synapse with the target cell, polarize toward the immune synapse to deliver their contents to the target cell membrane. Here, we identify a small GTP-binding protein, ADP-ribosylation factor-like 8b (Arl8b), as a critical factor required for NK cell-mediated cytotoxicity. Our findings indicate that Arl8b drives the polarization of lytic granules and microtubule-organizing centers (MTOCs) toward the immune synapse between effector NK lymphocytes and target cells. Using a glutathione S-transferase pull-down approach, we identify kinesin family member 5B (KIF5B; the heavy chain of kinesin-1) as an interaction partner of Arl8b from NK cell lysates. Previous studies showed that interaction between kinesin-1 and Arl8b is mediated by SifA and kinesin-interacting protein (SKIP) and the tripartite complex drives the anterograde movement of lysosomes. Silencing of both KIF5B and SKIP in NK cells, similar to Arl8b, led to failure of MTOC-lytic granule polarization to the immune synapse, suggesting that Arl8b and kinesin-1 together control this critical step in NK cell cytotoxicity.

Arl13b regulates endocytic recycling traffic.

Intracellular recycling pathways play critical roles in internalizing membrane and fluid phase cargo and in balancing the inflow and outflow of membrane and cell surface molecules. To identify proteins involved in the regulation of endocytic recycling, we used an shRNA trafficking library and screened for changes in the surface expression of CD1a antigen-presenting molecules that follow an endocytic recycling route. We found that silencing of the ADP-ribosylation factor (Arf)-like small GTPase Arl13b led to a decrease in CD1a surface expression, diminished CD1a function, and delayed CD1a recycling, suggesting that Arl13b is involved in the regulation of endocytic recycling traffic. Arl13b appears to be required for the major route of endocytic trafficking, causing clustering of early endosomes and leading to the accumulation of endocytic cargo. Moreover, Arl13b colocalized with markers of the endocytic recycling pathway followed by CD1a, namely Arf6 and Rab22a. We also detected an interaction between Arl13b and the actin cytoskeleton. Arl13b was previously implicated in cilia formation and function. Our present results indicate a previously unidentified role for Arl13b in endocytic recycling traffic and suggest a link between Arl13b function and the actin cytoskeleton.

Role of homocysteine & MTHFR C677T gene polymorphism as risk factors for coronary artery disease in young Indians.

BACKGROUND & OBJECTIVES: Hyperhomocysteinaemia (HCA) either due to mutation of MTHFR gene or deficiency of vitamin B 12 and folic acid, has been reported as a risk factor for coronary artery disease (CAD). The present study was aimed to determine plasma homocysteine (Hcy) levels and to evaluate MTHFR C677T gene polymorphism as risk factors for CAD, and to study the role of Hcy in conjunction with a few other risk factors of CAD in young Indians. The effect of vitamin B12 and folic acid supplements on the raised plasma Hcy levels in patients of CAD was also assessed. METHODS: The present study included 199 consecutive angiography confirmed CAD patients, <45 yr of age, without any other known pro- coagulant state and 200 age- and sex-matched healthy controls. Fasting blood samples were collected in EDTA and plasma Hcy was estimated by ELISA test and the MTHFR C677T polymorphism detection was carried out by PCR-RFLP method. RESULTS: Significant difference (P<0.001) was found between mean fasting levels of plasma Hcy in cases (22.14 ± 10.62 µmol/l) and controls (17.38 ± 8.46 µmol/l) with an Odds ratio as 1.93 (95% CI, 1.27-2.94). Levels of cholesterol, LDL, and triglycerides were significantly (P<0.001) higher in cases compared with controls. INTERPRETATION & CONCLUSIONS: Our study showed significant correlation between hyperhomocysteinaemia and coronary artery disease. Multivariate analysis by logistic regression of the various risk factors of CAD, found high levels of Hcy, cholesterol, LDL and low levels of HDL and smoking as independent predictors of CAD when all other factors were controlled. Significant post-treatment decrease found in HCA was easily modifiable by vitamin intervention irrespective to their CT or TT genotype of C677T MTHFR gene. Further studies to look at the plasma levels of folate and cobalamines and their association with Hcy are required to be done.

Lysosomal trafficking, antigen presentation, and microbial killing are controlled by the Arf-like GTPase Arl8b.

Antigen presentation and microbial killing are critical arms of host defense that depend upon cargo trafficking into lysosomes. Yet, the molecular regulators of traffic into lysosomes are only partly understood. Here, using a lysosome-dependent immunological screen of a trafficking shRNA library, we identified the Arf-like GTPase Arl8b as a critical regulator of cargo delivery to lysosomes. Homotypic fusion and vacuole protein sorting (HOPS) complex members were identified as effectors of Arl8b and were dependent on Arl8b for recruitment to lysosomes, suggesting that Arl8b-HOPS plays a general role in directing traffic to lysosomes. Moreover, the formation of CD1 antigen-presenting complexes in lysosomes, their delivery to the plasma membrane, and phagosome-lysosome fusion were all markedly impaired in Arl8b silenced cells resulting in corresponding defects in T cell activation and microbial killing. Together, these results define Arl8b as a key regulator of lysosomal cellular and immunological functions.

Regulation of CD1 antigen-presenting complex stability.

For major histocompatibility complex class I and II molecules, the binding of specific peptide antigens is essential for assembly and trafficking and is at the center of their quality control mechanism. However, the role of lipid antigen binding in stabilization and quality control of CD1 heavy chain (HC).beta(2)-microglobulin (beta(2)m) complexes is unclear. Furthermore, the distinct trafficking and loading routes of CD1 proteins take them from mildly acidic pH in early endososmal compartments (pH 6.0) to markedly acidic pH in lysosomes (pH 5.0) and back to neutral pH of the cell surface (pH 7.4). Here, we present evidence that the stability of each CD1 HC.beta(2)m complex is determined by the distinct pH optima identical to that of the intracellular compartments in which each CD1 isoform resides. Although stable at acidic endosomal pH, complexes are only stable at cell surface pH 7.4 when bound to specific lipid antigens. The proposed model outlines a quality control program that allows lipid exchange at low endosomal pH without dissociation of the CD1 HC.beta(2)m complex and then stabilizes the antigen-loaded complex at neutral pH at the cell surface.