Recommendations of scRNA-seq Differential Gene Expression Analysis Based on Comprehensive Benchmarking.

To guide analysts to select the right tool and parameters in differential gene expression analyses of single-cell RNA sequencing (scRNA-seq) data, we developed a novel simulator that recapitulates the data characteristics of real scRNA-seq datasets while accounting for all the relevant sources of variation in a multi-subject, multi-condition scRNA-seq experiment: the cell-to-cell variation within a subject, the variation across subjects, the variability across cell types, the mean/variance relationship of gene expression across genes, library size effects, group effects, and covariate effects. By applying it to benchmark 12 differential gene expression analysis methods (including cell-level and pseudo-bulk methods) on simulated multi-condition, multi-subject data of the 10x Genomics platform, we demonstrated that methods originating from the negative binomial mixed model such as glmmTMB and NEBULA-HL outperformed other methods. Utilizing NEBULA-HL in a statistical analysis pipeline for single-cell analysis will enable scientists to better understand the cell-type-specific transcriptomic response to disease or treatment effects and to discover new drug targets. Further, application to two real datasets showed the outperformance of our differential expression (DE) pipeline, with unified findings of differentially expressed genes (DEG) and a pseudo-time trajectory transcriptomic result. In the end, we made recommendations for filtering strategies of cells and genes based on simulation results to achieve optimal experimental goals.

Cell-Specific Transcriptional Responses to Heat Shock in the Mouse Utricle Epithelium.

Sensory epithelia of the inner ear contain mechanosensory hair cells (HCs) and glia-like supporting cells (SCs), both of which are required for hearing and balance functions. Each of these cell types has unique responses to ototoxic and cytoprotective stimuli. Non-lethal heat stress in the mammalian utricle induces heat shock proteins (HSPs) and protects against ototoxic drug-induced hair cell death. Induction of HSPs in the utricle demonstrates cell-type specificity at the protein level, with HSP70 induction occurring primarily in SCs, while HSP32 (also known as heme oxygenase 1, HMOX1) is induced primarily in resident macrophages. Neither of these HSPs are robustly induced in HCs, suggesting that HCs may have little capacity for induction of stress-induced protective responses. To determine the transcriptional responses to heat shock of these different cell types, we performed cell-type-specific transcriptional profiling using the RiboTag method, which allows for immunoprecipitation (IP) of actively translating mRNAs from specific cell types. RNA-Seq differential gene expression analyses demonstrated that the RiboTag method identified known cell type-specific markers as well as new markers for HCs and SCs. Gene expression differences suggest that HCs and SCs exhibit differential transcriptional heat shock responses. The chaperonin family member Cct8 was significantly enriched only in heat-shocked HCs, while Hspa1l (HSP70 family), and Hspb1 and Cryab (HSP27 and HSP20 families, respectively) were enriched only in SCs. Together our data indicate that HCs exhibit a limited but unique heat shock response, and SCs exhibit a broader and more robust transcriptional response to protective heat stress.

The Inner Ear Heat Shock Transcriptional Signature Identifies Compounds That Protect Against Aminoglycoside Ototoxicity.

Mechanosensory hair cells of the inner ear transduce auditory and vestibular sensory input. Hair cells are susceptible to death from a variety of stressors, including treatment with therapeutic drugs that have ototoxic side effects. There is a need for co-therapies to mitigate drug-induced ototoxicity, and we showed previously that induction of heat shock proteins (HSPs) protects against hair cell death and hearing loss caused by aminoglycoside antibiotics in mouse. Here, we utilized the library of integrated cellular signatures (LINCS) to identify perturbagens that induce transcriptional profiles similar to that of heat shock. Massively parallel sequencing of RNA (RNA-Seq) of heat shocked and control mouse utricles provided a heat shock gene expression signature that was used in conjunction with LINCS to identify candidate perturbagens, several of which were known to protect the inner ear. Our data indicate that LINCS is a useful tool to screen for compounds that generate specific gene expression signatures in the inner ear. Forty-two LINCS-identified perturbagens were tested for otoprotection in zebrafish, and three of these were protective. These compounds also induced the heat shock gene expression signature in mouse utricles, and one compound protected against aminoglycoside-induced hair cell death in whole organ cultures of utricles from adult mice.

A kinase inhibitor library screen identifies novel enzymes involved in ototoxic damage to the murine organ of Corti.

Ototoxicity is a significant side effect of a number of drugs, including the aminoglycoside antibiotics and platinum-based chemotherapeutic agents that are used to treat life-threatening illnesses. Although much progress has been made, the mechanisms that lead to ototoxic loss of inner ear sensory hair cells (HCs) remains incompletely understood. Given the critical role of protein phosphorylation in intracellular processes, including both damage and survival signaling, we screened a library of kinase inhibitors targeting members of all the major families in the kinome. Micro-explants from the organ of Corti of mice in which only the sensory cells express GFP were exposed to 200 µM of the ototoxic aminoglycoside gentamicin with or without three dosages of each kinase inhibitor. The loss of sensory cells was compared to that seen with gentamicin alone, or without treatment. Of the 160 inhibitors, 15 exhibited a statistically significant protective effect, while 3 significantly enhanced HC loss. The results confirm some previous studies of kinase involvement in HC damage and survival, and also highlight several novel potential kinase pathway contributions to ototoxicity.

Clearance of rapid adenosine release is regulated by nucleoside transporters and metabolism.

Adenosine is a neuromodulator that regulates neurotransmission in the brain and central nervous system. Recently, spontaneous adenosine release that is cleared in 3-4 sec was discovered in mouse spinal cord slices and anesthetized rat brains. Here, we examined the clearance of spontaneous adenosine in the rat caudate-putamen and exogenously applied adenosine in caudate brain slices. The V max for clearance of exogenously applied adenosine in brain slices was 1.4 ± 0.1 µmol/L/sec. In vivo, the equilibrative nucleoside transport 1 (ENT1) inhibitor, S-(4-nitrobenzyl)-6-thioinosine (NBTI) (1 mg/kg, i.p.) significantly increased the duration of adenosine, while the ENT1/2 inhibitor, dipyridamole (10 mg/kg, i.p.), did not affect duration. 5-(3-Bromophenyl)-7-[6-(4-morpholinyl)-3-pyrido[2,3-d]byrimidin-4-amine dihydrochloride (ABT-702), an adenosine kinase inhibitor (5 mg/kg, i.p.), increased the duration of spontaneous adenosine release. The adenosine deaminase inhibitor, erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA) (10 mg/kg, i.p.), also increased the duration in vivo. Similarly, NBTI (10 µmol/L), ABT-702 (100 nmol/L), or EHNA (20 µmol/L) also decreased the clearance rate of exogenously applied adenosine in brain slices. The increases in duration for blocking ENT1, adenosine kinase, or adenosine deaminase individually were similar, about 0.4 sec in vivo; thus, the removal of adenosine on a rapid time scale occurs through three mechanisms that have comparable effects. A cocktail of ABT-702, NBTI, and EHNA significantly increased the duration by 0.7 sec, so the mechanisms are not additive and there may be additional mechanisms clearing adenosine on a rapid time scale. The presence of multiple mechanisms for adenosine clearance on a time scale of seconds demonstrates that adenosine is tightly regulated in the extracellular space.

Characterization of spontaneous, transient adenosine release in the caudate-putamen and prefrontal cortex.

Adenosine is a neuroprotective agent that inhibits neuronal activity and modulates neurotransmission. Previous research has shown adenosine gradually accumulates during pathologies such as stroke and regulates neurotransmission on the minute-to-hour time scale. Our lab developed a method using carbon-fiber microelectrodes to directly measure adenosine changes on a sub-second time scale with fast-scan cyclic voltammetry (FSCV). Recently, adenosine release lasting a couple of seconds has been found in murine spinal cord slices. In this study, we characterized spontaneous, transient adenosine release in vivo, in the caudate-putamen and prefrontal cortex of anesthetized rats. The average concentration of adenosine release was 0.17±0.01 µM in the caudate and 0.19±0.01 µM in the prefrontal cortex, although the range was large, from 0.04 to 3.2 µM. The average duration of spontaneous adenosine release was 2.9±0.1 seconds and 2.8±0.1 seconds in the caudate and prefrontal cortex, respectively. The concentration and number of transients detected do not change over a four hour period, suggesting spontaneous events are not caused by electrode implantation. The frequency of adenosine transients was higher in the prefrontal cortex than the caudate-putamen and was modulated by A1 receptors. The A1 antagonist DPCPX (8-cyclopentyl-1,3-dipropylxanthine, 6 mg/kg i.p.) increased the frequency of spontaneous adenosine release, while the A1 agonist CPA (N(6)-cyclopentyladenosine, 1 mg/kg i.p.) decreased the frequency. These findings are a paradigm shift for understanding the time course of adenosine signaling, demonstrating that there is a rapid mode of adenosine signaling that could cause transient, local neuromodulation.

Sound preconditioning therapy inhibits ototoxic hearing loss in mice.

Therapeutic drugs with ototoxic side effects cause significant hearing loss for thousands of patients annually. Two major classes of ototoxic drugs are cisplatin and the aminoglycoside antibiotics, both of which are toxic to mechanosensory hair cells, the receptor cells of the inner ear. A critical need exists for therapies that protect the inner ear without inhibiting the therapeutic efficacy of these drugs. The induction of heat shock proteins (HSPs) inhibits both aminoglycoside- and cisplatin-induced hair cell death and hearing loss. We hypothesized that exposure to sound that is titrated to stress the inner ear without causing permanent damage would induce HSPs in the cochlea and inhibit ototoxic drug­induced hearing loss. We developed a sound exposure protocol that induces HSPs without causing permanent hearing loss. We used this protocol in conjunction with a newly developed mouse model of cisplatin ototoxicity and found that preconditioning mouse inner ears with sound has a robust protective effect against cisplatin-induced hearing loss and hair cell death. Sound therapy also provided protection against aminoglycoside-induced hearing loss. These data indicate that sound preconditioning protects against both classes of ototoxic drugs, and they suggest that sound therapy holds promise for preventing hearing loss in patients receiving these drugs.