Deep sequencing reveals cell-type-specific patterns of single-cell transcriptome variation.

BACKGROUND: Differentiation of metazoan cells requires execution of different gene expression programs but recent single-cell transcriptome profiling has revealed considerable variation within cells of seeming identical phenotype. This brings into question the relationship between transcriptome states and cell phenotypes. Additionally, single-cell transcriptomics presents unique analysis challenges that need to be addressed to answer this question. RESULTS: We present high quality deep read-depth single-cell RNA sequencing for 91 cells from five mouse tissues and 18 cells from two rat tissues, along with 30 control samples of bulk RNA diluted to single-cell levels. We find that transcriptomes differ globally across tissues with regard to the number of genes expressed, the average expression patterns, and within-cell-type variation patterns. We develop methods to filter genes for reliable quantification and to calibrate biological variation. All cell types include genes with high variability in expression, in a tissue-specific manner. We also find evidence that single-cell variability of neuronal genes in mice is correlated with that in rats consistent with the hypothesis that levels of variation may be conserved. CONCLUSIONS: Single-cell RNA-sequencing data provide a unique view of transcriptome function; however, careful analysis is required in order to use single-cell RNA-sequencing measurements for this purpose. Technical variation must be considered in single-cell RNA-sequencing studies of expression variation. For a subset of genes, biological variability within each cell type appears to be regulated in order to perform dynamic functions, rather than solely molecular noise.

Divergence of RNA localization between rat and mouse neurons reveals the potential for rapid brain evolution.

BACKGROUND: Neurons display a highly polarized architecture. Their ability to modify their features under intracellular and extracellular stimuli, known as synaptic plasticity, is a key component of the neurochemical basis of learning and memory. A key feature of synaptic plasticity involves the delivery of mRNAs to distinct sub-cellular domains where they are locally translated. Regulatory coordination of these spatio-temporal events is critical for synaptogenesis and synaptic plasticity as defects in these processes can lead to neurological diseases. In this work, using microdissected dendrites from primary cultures of hippocampal neurons of two mouse strains (C57BL/6 and Balb/c) and one rat strain (Sprague-Dawley), we investigate via microarrays, subcellular localization of mRNAs in dendrites of neurons to assay the evolutionary differences in subcellular dendritic transcripts localization. RESULTS: Our microarray analysis highlighted significantly greater evolutionary diversification of RNA localization in the dendritic transcriptomes (81% gene identity difference among the top 5% highly expressed genes) compared to the transcriptomes of 11 different central nervous system (CNS) and non-CNS tissues (average of 44% gene identity difference among the top 5% highly expressed genes). Differentially localized genes include many genes involved in CNS function. CONCLUSIONS: Species differences in sub-cellular localization may reflect non-functional neutral drift. However, the functional categories of mRNA showing differential localization suggest that at least part of the divergence may reflect activity-dependent functional differences of neurons, mediated by species-specific RNA subcellular localization mechanisms.

Transcriptome in vivo analysis (TIVA) of spatially defined single cells in live tissue.

Transcriptome profiling of single cells resident in their natural microenvironment depends upon RNA capture methods that are both noninvasive and spatially precise. We engineered a transcriptome in vivo analysis (TIVA) tag, which upon photoactivation enables mRNA capture from single cells in live tissue. Using the TIVA tag in combination with RNA sequencing (RNA-seq), we analyzed transcriptome variance among single neurons in culture and in mouse and human tissue in vivo. Our data showed that the tissue microenvironment shapes the transcriptomic landscape of individual cells. The TIVA methodology is, to our knowledge, the first noninvasive approach for capturing mRNA from live single cells in their natural microenvironment.

Subcellular RNA sequencing reveals broad presence of cytoplasmic intron-sequence retaining transcripts in mouse and rat neurons.

Recent findings have revealed the complexity of the transcriptional landscape in mammalian cells. One recently described class of novel transcripts are the Cytoplasmic Intron-sequence Retaining Transcripts (CIRTs), hypothesized to confer post-transcriptional regulatory function. For instance, the neuronal CIRT KCNMA1i16 contributes to the firing properties of hippocampal neurons. Intronic sub-sequence retention within IL1-ß mRNA in anucleate platelets has been implicated in activity-dependent splicing and translation. In a recent study, we showed CIRTs harbor functional SINE ID elements which are hypothesized to mediate dendritic localization in neurons. Based on these studies and others, we hypothesized that CIRTs may be present in a broad set of transcripts and comprise novel signals for post-transcriptional regulation. We carried out a transcriptome-wide survey of CIRTs by sequencing micro-dissected subcellular RNA fractions. We sequenced two batches of 150-300 individually dissected dendrites from primary cultures of hippocampal neurons in rat and three batches from mouse hippocampal neurons. After statistical processing to minimize artifacts, we found a broad prevalence of CIRTs in the neurons in both species (44-60% of the expressed transcripts). The sequence patterns, including stereotypical length, biased inclusion of specific introns, and intron-intron junctions, suggested CIRT-specific nuclear processing. Our analysis also suggested that these cytoplasmic intron-sequence retaining transcripts may serve as a primary transcript for ncRNAs. Our results show that retaining intronic sequences is not isolated to a few loci but may be a genome-wide phenomenon for embedding functional signals within certain mRNA. The results hypothesize a novel source of cis-sequences for post-transcriptional regulation. Our results hypothesize two potentially novel splicing pathways: one, within the nucleus for CIRT biogenesis; and another, within the cytoplasm for removing CIRT sequences before translation. We also speculate that release of CIRT sequences prior to translation may form RNA-based signals within the cell potentially comprising a novel class of signaling pathways.

Quantitative biology of single neurons.

The building blocks of complex biological systems are single cells. Fundamental insights gained from single-cell analysis promise to provide the framework for understanding normal biological systems development as well as the limits on systems/cellular ability to respond to disease. The interplay of cells to create functional systems is not well understood. Until recently, the study of single cells has concentrated primarily on morphological and physiological characterization. With the application of new highly sensitive molecular and genomic technologies, the quantitative biochemistry of single cells is now accessible.

Heterochronic evolution reveals modular timing changes in budding yeast transcriptomes.

BACKGROUND: Gene expression is a dynamic trait, and the evolution of gene regulation can dramatically alter the timing of gene expression without greatly affecting mean expression levels. Moreover, modules of co-regulated genes may exhibit coordinated shifts in expression timing patterns during evolutionary divergence. Here, we examined transcriptome evolution in the dynamical context of the budding yeast cell-division cycle, to investigate the extent of divergence in expression timing and the regulatory architecture underlying timing evolution. RESULTS: Using a custom microarray platform, we obtained 378 measurements for 6,263 genes over 18 timepoints of the cell-division cycle in nine strains of S. cerevisiae and one strain of S. paradoxus. Most genes show significant divergence in expression dynamics at all scales of transcriptome organization, suggesting broad potential for timing changes. A model test comparing expression level evolution versus timing evolution revealed a better fit with timing evolution for 82% of genes. Analysis of shared patterns of timing evolution suggests the existence of seven dynamically-autonomous modules, each of which shows coherent evolutionary timing changes. Analysis of transcription factors associated with these gene modules suggests a modular pleiotropic source of divergence in expression timing. CONCLUSIONS: We propose that transcriptome evolution may generally entail changes in timing (heterochrony) rather than changes in levels (heterometry) of expression. Evolution of gene expression dynamics may involve modular changes in timing control mediated by module-specific transcription factors. We hypothesize that genome-wide gene regulation may utilize a general architecture comprised of multiple semi-autonomous event timelines, whose superposition could produce combinatorial complexity in timing control patterns.

Allopatric divergence, secondary contact, and genetic isolation in wild yeast populations.

In plants and animals, new biological species clearly have arisen as a byproduct of genetic divergence in allopatry. However, our understanding of the processes that generate new microbial species remains limited [1] despite the large contribution of microbes to the world's biodiversity. A recent hypothesis claims that microbes lack biogeographical divergence because their population sizes are large and their migration rates are presumably high [2, 3]. In recapitulating the classic microbial-ecology dictum that "everything is everywhere, and the environment selects"[4, 5], this hypothesis casts doubt on whether geographic divergence promotes speciation in microbes. To date, its predictions have been tested primarily with data from eubacteria and archaebacteria [6-8]. However, this hypothesis's most important implication is in sexual eukaryotic microbes, where migration and genetic admixture are specifically predicted to inhibit allopatric divergence and speciation [9]. Here, we use nuclear-sequence data from globally distributed natural populations of the yeast Saccharomyces paradoxus to investigate the role of geography in generating diversity in sexual eukaryotic microbes. We show that these populations have undergone allopatric divergence and then secondary contact without genetic admixture. Our data thus support the occurrence of evolutionary processes necessary for allopatric speciation in sexual microbes.

Mate choice assays and mating propensity differences in natural yeast populations.

In sexual microbes, mating occurs by fusion of individual cells. This complete fitness investment suggests that cell behaviour could potentially mediate prezygotic isolation between microbial species, a topic about which very little is known. To investigate this possibility, we conducted individual cell mate choice trials and mass-culture mating propensity assays with isolates from sympatric natural populations of the closely related yeasts Saccharomyces cerevisiae and Saccharomyces paradoxus. Although we found no evidence for active species recognition in mate choice, we observed a marked difference in mating propensity between these two species. We briefly discuss the possibility that this mating propensity difference may contribute to reproductive isolation between S. cerevisiae and S. paradoxus in nature.

Distribution and sequence analysis of a novel Ty3-like element in natural Saccharomyces paradoxus isolates.

Little is known about the transposable elements of species closely related to Saccharomyces cerevisiae. We present a novel transposable element in Saccharomyces paradoxus, a close congener of S. cerevisiae. Sequence analysis of this element, designated Ty3-1p, indicates that it is a homologue of the S. cerevisiae Ty3 element. Ty3-1p shares 82% nucleotide identity with an S. cerevisiae Ty3 element and appears to be structured identically to Ty3, containing two overlapping open reading frames, six retroviral-like domains, a J domain, and flanking sigma-like elements. A sigma element from Ty3-1p is 75% identical to a Ty3 sigma element. There is no evidence of horizontal transfer of Ty3 in Saccharomyces sensu stricto. We assess the distributions of Ty3p and Ty3 element insertions in natural population samples of S. paradoxus and S. cerevisiae. The S. paradoxus population sample exhibits Ty3p insertions present at a variety of sites at low frequency; this suggests that Ty3p elements are active in the sampled population. The S. cerevisiae population sample exhibits a uniform Ty3 hybridization profile in which all element insertions appear to be fixed. We comment on the possible causes of these contrasting observed distributions (GenBank Accession Nos AY198186 and AY198187).