Machine learning-guided engineering of genetically encoded fluorescent calcium indicators.

Here we used machine learning to engineer genetically encoded fluorescent indicators, protein-based sensors critical for real-time monitoring of biological activity. We used machine learning to predict the outcomes of sensor mutagenesis by analyzing established libraries that link sensor sequences to functions. Using the GCaMP calcium indicator as a scaffold, we developed an ensemble of three regression models trained on experimentally derived GCaMP mutation libraries. The trained ensemble performed an in silico functional screen on 1,423 novel, uncharacterized GCaMP variants. As a result, we identified the ensemble-derived GCaMP (eGCaMP) variants, eGCaMP and eGCaMP+, which achieve both faster kinetics and larger ∆F/F0 responses upon stimulation than previously published fast variants. Furthermore, we identified a combinatorial mutation with extraordinary dynamic range, eGCaMP2+, which outperforms the tested sixth-, seventh- and eighth-generation GCaMPs. These findings demonstrate the value of machine learning as a tool to facilitate the efficient engineering of proteins for desired biophysical characteristics.

Far-red and sensitive sensor for monitoring real time H2O2 dynamics with subcellular resolution and in multi-parametric imaging applications.

H2O2 is a key oxidant in mammalian biology and a pleiotropic signaling molecule at the physiological level, and its excessive accumulation in conjunction with decreased cellular reduction capacity is often found to be a common pathological marker. Here, we present a red fluorescent Genetically Encoded H2O2 Indicator (GEHI) allowing versatile optogenetic dissection of redox biology. Our new GEHI, oROS-HT, is a chemigenetic sensor utilizing a HaloTag and Janelia Fluor (JF) rhodamine dye as fluorescent reporters. We developed oROS-HT through a structure-guided approach aided by classic protein structures and recent protein structure prediction tools. Optimized with JF635, oROS-HT is a sensor with 635 nm excitation and 650 nm emission peaks, allowing it to retain its brightness while monitoring intracellular H2O2 dynamics. Furthermore, it enables multi-color imaging in combination with blue-green fluorescent sensors for orthogonal analytes and low auto-fluorescence interference in biological tissues. Other advantages of oROS-HT over alternative GEHIs are its fast kinetics, oxygen-independent maturation, low pH sensitivity, lack of photo-artifact, and lack of intracellular aggregation. Here, we demonstrated efficient subcellular targeting and how oROS-HT can map inter and intracellular H2O2 diffusion at subcellular resolution. Lastly, we used oROS-HT with the green fluorescent calcium indicator Fluo-4 to investigate the transient effect of the anti-inflammatory agent auranofin on cellular redox physiology and calcium levels via multi-parametric, dual-color imaging.

Optogenetic Microwell Array Screening System: A High-Throughput Engineering Platform for Genetically Encoded Fluorescent Indicators.

Genetically encoded fluorescent indicators (GEFIs) are protein-based optogenetic tools that change their fluorescence intensity when binding specific ligands in cells and tissues. GEFI encoding DNA can be expressed in cell subtypes while monitoring cellular physiological responses. However, engineering GEFIs with physiological sensitivity and pharmacological specificity often requires iterative optimization through trial-and-error mutagenesis while assessing their biophysical function in vitro one by one. Here, the vast mutational landscape of proteins constitutes a significant obstacle that slows GEFI development, particularly for sensors that rely on mammalian host systems for testing. To overcome these obstacles, we developed a multiplexed high-throughput engineering platform called the optogenetic microwell array screening system (Opto-MASS) that functionally tests thousands of GEFI variants in parallel in mammalian cells. Opto-MASS represents the next step for engineering optogenetic tools as it can screen large variant libraries orders of magnitude faster than current methods. We showcase this system by testing over 13,000 dopamine and 21,000 opioid sensor variants. We generated a new dopamine sensor, dMASS1, with a >6-fold signal increase to 100 nM dopamine exposure compared to its parent construct. Our new opioid sensor, µMASS1, has a ∼4.6-fold signal increase over its parent scaffold's response to 500 nM DAMGO. Thus, Opto-MASS can rapidly engineer new sensors while significantly shortening the optimization time for new sensors with distinct biophysical properties.

Machine Learning Ensemble Directed Engineering of Genetically Encoded Fluorescent Calcium Indicators.

In this study, we focused on the transformative potential of machine learning in the engineering of genetically encoded fluorescent indicators (GEFIs), protein-based sensing tools that are critical for real-time monitoring of biological activity. GEFIs are complex proteins with multiple dynamic states, rendering optimization by trial-and-error mutagenesis a challenging problem. We applied an alternative approach using machine learning to predict the outcomes of sensor mutagenesis by analyzing established libraries that link sensor sequences to functions. Using the GCaMP calcium indicator as a scaffold, we developed an ensemble of three regression models trained on experimentally derived GCaMP mutation libraries. We used the trained ensemble to perform an in silico functional screen on 1423 novel, uncharacterized GCaMP variants. As a result, we identified the novel ensemble-derived GCaMP (eGCaMP) variants, eGCaMP and eGCaMP+, that achieve both faster kinetics and larger fluorescent responses upon stimulation than previously published fast variants. Furthermore, we identified a combinatorial mutation with extraordinary dynamic range, eGCaMP2+, that outperforms the tested 6th, 7th, and 8th generation GCaMPs. These findings demonstrate the value of machine learning as a tool to facilitate the efficient pre-screening of mutants for functional characteristics. By leveraging the learning capabilities of our ensemble, we were able to accelerate the identification of promising mutations and reduce the experimental burden associated with trial-and-error mutagenesis. Overall, these findings have significant implications for optimizing GEFIs and other protein-based tools, demonstrating the utility of machine learning as a powerful asset in protein engineering.

Inhibition of Karyopherin ß1-Mediated Nuclear Import Disrupts Oncogenic Lineage-Defining Transcription Factor Activity in Small Cell Lung Cancer.

Genomic studies support the classification of small cell lung cancer (SCLC) into subtypes based on the expression of lineage-defining transcription factors ASCL1 and NEUROD1, which together are expressed in ∼86% of SCLC. ASCL1 and NEUROD1 activate SCLC oncogene expression, drive distinct transcriptional programs, and maintain the in vitro growth and oncogenic properties of ASCL1 or NEUROD1-expressing SCLC. ASCL1 is also required for tumor formation in SCLC mouse models. A strategy to inhibit the activity of these oncogenic drivers may therefore provide both a targeted therapy for the predominant SCLC subtypes and a tool to investigate the underlying lineage plasticity of established SCLC tumors. However, there are no known agents that inhibit ASCL1 or NEUROD1 function. In this study, we identify a novel strategy to pharmacologically target ASCL1 and NEUROD1 activity in SCLC by exploiting the nuclear localization required for the function of these transcription factors. Karyopherin ß1 (KPNB1) was identified as a nuclear import receptor for both ASCL1 and NEUROD1 in SCLC, and inhibition of KPNB1 led to impaired ASCL1 and NEUROD1 nuclear accumulation and transcriptional activity. Pharmacologic targeting of KPNB1 preferentially disrupted the growth of ASCL1+ and NEUROD1+ SCLC cells in vitro and suppressed ASCL1+ tumor growth in vivo, an effect mediated by a combination of impaired ASCL1 downstream target expression, cell-cycle activity, and proteostasis. These findings broaden the support for targeting nuclear transport as an anticancer therapeutic strategy and have implications for targeting lineage-transcription factors in tumors beyond SCLC. SIGNIFICANCE: The identification of KPNB1 as a nuclear import receptor for lineage-defining transcription factors in SCLC reveals a viable therapeutic strategy for cancer treatment.

ASCL1 represses a SOX9+ neural crest stem-like state in small cell lung cancer.

ASCL1 is a neuroendocrine lineage-specific oncogenic driver of small cell lung cancer (SCLC), highly expressed in a significant fraction of tumors. However, ∼25% of human SCLC are ASCL1-low and associated with low neuroendocrine fate and high MYC expression. Using genetically engineered mouse models (GEMMs), we show that alterations in Rb1/Trp53/Myc in the mouse lung induce an ASCL1+ state of SCLC in multiple cells of origin. Genetic depletion of ASCL1 in MYC-driven SCLC dramatically inhibits tumor initiation and progression to the NEUROD1+ subtype of SCLC. Surprisingly, ASCL1 loss promotes a SOX9+ mesenchymal/neural crest stem-like state and the emergence of osteosarcoma and chondroid tumors, whose propensity is impacted by cell of origin. ASCL1 is critical for expression of key lineage-related transcription factors NKX2-1, FOXA2, and INSM1 and represses genes involved in the Hippo/Wnt/Notch developmental pathways in vivo. Importantly, ASCL1 represses a SOX9/RUNX1/RUNX2 program in vivo and SOX9 expression in human SCLC cells, suggesting a conserved function for ASCL1. Together, in a MYC-driven SCLC model, ASCL1 promotes neuroendocrine fate and represses the emergence of a SOX9+ nonendodermal stem-like fate that resembles neural crest.

MYC Drives Temporal Evolution of Small Cell Lung Cancer Subtypes by Reprogramming Neuroendocrine Fate.

Small cell lung cancer (SCLC) is a neuroendocrine tumor treated clinically as a single disease with poor outcomes. Distinct SCLC molecular subtypes have been defined based on expression of ASCL1, NEUROD1, POU2F3, or YAP1. Here, we use mouse and human models with a time-series single-cell transcriptome analysis to reveal that MYC drives dynamic evolution of SCLC subtypes. In neuroendocrine cells, MYC activates Notch to dedifferentiate tumor cells, promoting a temporal shift in SCLC from ASCL1+ to NEUROD1+ to YAP1+ states. MYC alternatively promotes POU2F3+ tumors from a distinct cell type. Human SCLC exhibits intratumoral subtype heterogeneity, suggesting that this dynamic evolution occurs in patient tumors. These findings suggest that genetics, cell of origin, and tumor cell plasticity determine SCLC subtype.

New Approaches to SCLC Therapy: From the Laboratory to the Clinic.

The outcomes of patients with SCLC have not yet been substantially impacted by the revolution in precision oncology, primarily owing to a paucity of genetic alterations in actionable driver oncogenes. Nevertheless, systemic therapies that include immunotherapy are beginning to show promise in the clinic. Although, these results are encouraging, many patients do not respond to, or rapidly recur after, current regimens, necessitating alternative or complementary therapeutic strategies. In this review, we discuss ongoing investigations into the pathobiology of this recalcitrant cancer and the therapeutic vulnerabilities that are exposed by the disease state. Included within this discussion, is a snapshot of the current biomarker and clinical trial landscapes for SCLC. Finally, we identify key knowledge gaps that should be addressed to advance the field in pursuit of reduced SCLC mortality. This review largely summarizes work presented at the Third Biennial International Association for the Study of Lung Cancer SCLC Meeting.

MYC-Driven Small-Cell Lung Cancer is Metabolically Distinct and Vulnerable to Arginine Depletion.

PURPOSE: Small-cell lung cancer (SCLC) has been treated clinically as a homogeneous disease, but recent discoveries suggest that SCLC is heterogeneous. Whether metabolic differences exist among SCLC subtypes is largely unexplored. In this study, we aimed to determine whether metabolic vulnerabilities exist between SCLC subtypes that can be therapeutically exploited. EXPERIMENTAL DESIGN: We performed steady state metabolomics on tumors isolated from distinct genetically engineered mouse models (GEMM) representing the MYC- and MYCL-driven subtypes of SCLC. Using genetic and pharmacologic approaches, we validated our findings in chemo-naïve and -resistant human SCLC cell lines, multiple GEMMs, four human cell line xenografts, and four newly derived PDX models. RESULTS: We discover that SCLC subtypes driven by different MYC family members have distinct metabolic profiles. MYC-driven SCLC preferentially depends on arginine-regulated pathways including polyamine biosynthesis and mTOR pathway activation. Chemo-resistant SCLC cells exhibit increased MYC expression and similar metabolic liabilities as chemo-naïve MYC-driven cells. Arginine depletion with pegylated arginine deiminase (ADI-PEG 20) dramatically suppresses tumor growth and promotes survival of mice specifically with MYC-driven tumors, including in GEMMs, human cell line xenografts, and a patient-derived xenograft from a relapsed patient. Finally, ADI-PEG 20 is significantly more effective than the standard-of-care chemotherapy. CONCLUSIONS: These data identify metabolic heterogeneity within SCLC and suggest arginine deprivation as a subtype-specific therapeutic vulnerability for MYC-driven SCLC.

The Lineage-Defining Transcription Factors SOX2 and NKX2-1 Determine Lung Cancer Cell Fate and Shape the Tumor Immune Microenvironment.

The major types of non-small-cell lung cancer (NSCLC)-squamous cell carcinoma and adenocarcinoma-have distinct immune microenvironments. We developed a genetic model of squamous NSCLC on the basis of overexpression of the transcription factor Sox2, which specifies lung basal cell fate, and loss of the tumor suppressor Lkb1 (SL mice). SL tumors recapitulated gene-expression and immune-infiltrate features of human squamous NSCLC; such features included enrichment of tumor-associated neutrophils (TANs) and decreased expression of NKX2-1, a transcriptional regulator that specifies alveolar cell fate. In Kras-driven adenocarcinomas, mis-expression of Sox2 or loss of Nkx2-1 led to TAN recruitment. TAN recruitment involved SOX2-mediated production of the chemokine CXCL5. Deletion of Nkx2-1 in SL mice (SNL) revealed that NKX2-1 suppresses SOX2-driven squamous tumorigenesis by repressing adeno-to-squamous transdifferentiation. Depletion of TANs in SNL mice reduced squamous tumors, suggesting that TANs foster squamous cell fate. Thus, lineage-defining transcription factors determine the tumor immune microenvironment, which in turn might impact the nature of the tumor.