Xiannian Zhang

Microfluidic single-cell whole-transcriptome sequencing.

Significance RNA sequencing of single cells enables measurement of biological variation in heterogeneous cellular populations and dissection of transcriptome complexity that is masked in ensemble measurements of gene expression. The low quantity of RNA in a single cell, however, hinders efficient and consistent reverse transcription and amplification of cDNA, limiting accuracy and obscuring biological variation with high technical noise. We developed a microfluidic approach to prepare cDNA from single cells for high-throughput transcriptome sequencing. The microfluidic platform facilitates single-cell manipulation, minimizes contamination, and furthermore, provides improved detection sensitivity and measurement precision, which is necessary for differentiating biological variability from technical noise. Single-cell whole-transcriptome analysis is a powerful tool for quantifying gene expression heterogeneity in populations of cells. Many techniques have, thus, been recently developed to perform transcriptome sequencing (RNA-Seq) on individual cells. To probe subtle biological variation between samples with limiting amounts of RNA, more precise and sensitive methods are still required. We adapted a previously developed strategy for single-cell RNA-Seq that has shown promise for superior sensitivity and implemented the chemistry in a microfluidic platform for single-cell whole-transcriptome analysis. In this approach, single cells are captured and lysed in a microfluidic device, where mRNAs with poly(A) tails are reverse-transcribed into cDNA. Double-stranded cDNA is then collected and sequenced using a next generation sequencing platform. We prepared 94 libraries consisting of single mouse embryonic cells and technical replicates of extracted RNA and thoroughly characterized the performance of this technology. Microfluidic implementation increased mRNA detection sensitivity as well as improved measurement precision compared with tube-based protocols. With 0.2 M reads per cell, we were able to reconstruct a majority of the bulk transcriptome with 10 single cells. We also quantified variation between and within different types of mouse embryonic cells and found that enhanced measurement precision, detection sensitivity, and experimental throughput aided the distinction between biological variability and technical noise. With this work, we validated the advantages of an early approach to single-cell RNA-Seq and showed that the benefits of combining microfluidic technology with high-throughput sequencing will be valuable for large-scale efforts in single-cell transcriptome analysis.

Single-cell RNA-seq transcriptome analysis of linear and circular RNAs in mouse preimplantation embryos

Circular RNAs (circRNAs) are a new class of non-polyadenylated non-coding RNAs that may play important roles in many biological processes. Here we develop a single-cell universal poly(A)-independent RNA sequencing (SUPeR-seq) method to sequence both polyadenylated and non-polyadenylated RNAs from individual cells. This method exhibits robust sensitivity, precision and accuracy. We discover 2891 circRNAs and 913 novel linear transcripts in mouse preimplantation embryos and further analyze the abundance of circRNAs along development, the function of enriched genes, and sequence features of circRNAs. Our work is key to deciphering regulation mechanisms of circRNAs during mammalian early embryonic development.

Label-free chemical imaging in vivo: three-dimensional non-invasive microscopic observation of amphioxus notochord through stimulated Raman scattering (SRS)

Notochord is a rod-shaped axial supporting structure unique only to chordates. In this study, we use cephalochordate amphioxus (Branchiostroma belcheri, a living basal chordate) and zebrafish (Danio rerio), a vertebrate, as model animals and employ stimulated Raman microscopy (SRS), a newly developed label-free technique, to investigate notochord structure and chemical composition in live animals. Coherent anti-Stokes Raman scattering (CARS) images have also been acquired for comparison. Both CARS and SRS images can construct the detailed three-dimensional structure of the notochord with resolution better than 1 μm. Label-free live imaging allows us to obtain the whole animals intact internal morphology, which is difficult to obtain through other mechanical or optical sectioning methods. Intrinsically, chemical sensitive SRS images, with a simple data processing procedure, show that the amphioxus notochord is protein-rich. Our result agrees well with that drawn from the labeling and cryo-sectioning observation, verifying that the coherent Raman scattering techniques are useful to examine the structure and chemical composition of live animals without labels in a fast, simple and accurate way.

Chemoproteomics reveals baicalin activates hepatic CPT1 to ameliorate diet-induced obesity and hepatic steatosis

Significance Baicalin is a major flavonoid component from the herbal medicine Scutellaria baicalensis that has been shown to have an antisteatosis effect. Through quantitative chemoproteomic profiling, we discovered that baicalin acts as a natural allosteric activator of carnitine palmitoyltransferase 1 (CPT1), the rate-limiting enzyme of fatty acid β-oxidation (FAO). By directly binding to CPT1 and activating its activity to accelerate fatty acid degradation, baicalin can significantly ameliorate symptoms associated with hepatic steatosis and reduce diet-induced obesity (DIO). Our study provides an example of a natural product agonist for CPT1. The results provide mechanistic insights to explain the bioactivity of baicalin in reducing lipid accumulation and introduce exciting opportunities for developing novel flavonoid-based FAO activators for pharmacologically treating DIO and associated metabolic disorders. Obesity and related metabolic diseases are becoming worldwide epidemics that lead to increased death rates and heavy health care costs. Effective treatment options have not been found yet. Here, based on the observation that baicalin, a flavonoid from the herbal medicine Scutellaria baicalensis, has unique antisteatosis activity, we performed quantitative chemoproteomic profiling and identified carnitine palmitoyltransferase 1 (CPT1), the controlling enzyme for fatty acid oxidation, as the key target of baicalin. The flavonoid directly activated hepatic CPT1 with isoform selectivity to accelerate the lipid influx into mitochondria for oxidation. Chronic treatment of baicalin ameliorated diet-induced obesity (DIO) and hepatic steatosis and led to systemic improvement of other metabolic disorders. Disruption of the predicted binding site of baicalin on CPT1 completely abolished the beneficial effect of the flavonoid. Our discovery of baicalin as an allosteric CPT1 activator opens new opportunities for pharmacological treatment of DIO and associated sequelae.

Comparative analysis of droplet-based ultra-high-throughput single-cell RNA-seq systems

Since its debut in 2009, single-cell RNA-seq has been a major propeller behind biomedical research progress. Developmental biology and stem cell studies especially benefit from the ability to profile single cells. While most studies still focus on individual tissues or organs, recent development of ultra-high-throughput single-cell RNA-seq has demonstrated potential power to depict more complexed system or even the entire body. Though multiple ultra-high-throughput single-cell RNA-seq systems have acquired attention, systematic comparison of these systems is yet available. Here we focus on three prevalent droplet-based ultra-high-throughput single-cell RNA-seq systems, inDrop, Drop-seq, and 10X Genomics Chromium. While each system is capable of profiling single-cell transcriptome, detailed comparison revealed distinguishing features and suitable application scenario for each system.

Terminal transfer amplification and sequencing for high-efficiency and low-bias copy number profiling of fragmented DNA samples

Since its invention, next generation sequencing (NGS) has greatly facilitated biomedical research and clinical diagnosis (Sikkema-Raddatz et al., 2013). Continuous dropping of the cost further accelerated the adaptation of sequencing as a standard analytical tool, from identification of drug candidates (Walker et al., 2015) to deciphering the complex biological systems (McConnell et al., 2013). However, progress in sample preparation technology has not been able to catch the speed of sequencing method evolution. For most NGS platforms, samples with DNA fragments to be sequenced need to be firstly converted into a ‘library’ in which each molecule can be further amplified into clones and then be sequenced. Library preparation is a critical step that generates short DNA fragments with certain adapters, and sometimes with barcodes, at both ends. While library construction from bulk genomic DNA samples is a routine procedure, traditional protocols become challenging when the starting material is limited. Materials from many research and diagnostic fields such as archaeology and preimplantation genetic diagnosis (PGD) (Treff et al., 2013) require DNA amplification before library construction. Various related methods have been developed for whole genome amplification from minute amount of DNA, or even single cells. Methods involving random primers, such as MDA, DOP-PCR, or MALBAC, would get high yields yet still produce nonspecific amplification products (Marcy et al., 2007), incomplete coverage along genome and shortened DNA length (Fu et al., 2015). A recently development method, LIANTI (Chen et al., 2017), employed linear amplification and showed improved amplification evenness and less errors, compared to those exponential amplification protocols. However, all these methods require high quality starting material, and DNA from many types of samples such as chromatin immunoprecipitation (ChIP) products, formalinfixed paraffin-embedded (FFPE) tissues, or ancient remains is highly fragmented. Short DNA could not be amplified efficiently by these methods. Adapter-involving PCR strategy becomes more suitable for short-fragment samples by adding sequencing adapters directly to the ends of nucleotide strands through template-switching primers coupled with ligation. Various methods, such as LM-PCR (Dai et al., 2000), LinDA (Shankaranarayanan and Mendoza-Parra, 2011) and others (Liu et al., 2008) have been developed but they all facing the inevitable sample loss majorly because of incomplete ligation and repetitive purification. Hence handing small amount of starting DNA material is still challenging. FFPE samples are always problematic for sequencing library preparation. Fixation, paraffin embedding, and archival storage conditions all contribute to fragmentation and other chemical damages of DNA. For FFPE samples, high efficient and short-length-specific library preparation methods are in great demand. In this paper, we present a versatile method for amplification and sequencing of minute amount of short-length DNA fragments. Our method, named Terminal Transfer Amplification and Sequencing (TTAS), relied on two rounds of tailing and amplification processes. TTAS is free of ligation step, independent of specific template DNA sequence, and compatible with prevalent sequencing library preparation protocols. We performed TTAS on fragmented genomic DNA from mouse embryo stem cells (mESCs) and HeLa cells and verified its amplification uniformity and efficiency. TTAS further demonstrated its capability of amplifying FFPE tissue-derived DNA from slices as small as 1 mm. For high throughput sequencing, it is always a challenge if the input material is short DNA since such short DNA is not suitable for prevalent protocols such as high efficient ligation or Tn5 tagmentation. We hence developed a novel method that employed a terminal elongation strategy to add universal tails at both ends of DNA for future PCR amplification (Fig. 1A). Since short DNA molecules generated by chemical degradation (eg., ancient DNA) or physical fragmentation (eg., sonication) may retain phosphate groups at 3′-terminus that prevented extension (Liu et al., 2008), template DNA requires dephosphorylation by phosphatase. We then used

Phenotype classification of single cells using SRS microscopy, RNA sequencing, and microfluidics(Conference Presentation)

Phenotype classification of single cells reveals biological variation that is masked in ensemble measurement. This heterogeneity is found in gene and protein expression as well as in cell morphology. Many techniques are available to probe phenotypic heterogeneity at the single cell level, for example quantitative imaging and single-cell RNA sequencing, but it is difficult to perform multiple assays on the same single cell. In order to directly track correlation between morphology and gene expression at the single cell level, we developed a microfluidic platform for quantitative coherent Raman imaging and immediate RNA sequencing (RNA-Seq) of single cells. With this device we actively sort and trap cells for analysis with stimulated Raman scattering microscopy (SRS). The cells are then processed in parallel pipelines for lysis, and preparation of cDNA for high-throughput transcriptome sequencing. SRS microscopy offers three-dimensional imaging with chemical specificity for quantitative analysis of protein and lipid distribution in single cells. Meanwhile, the microfluidic platform facilitates single-cell manipulation, minimizes contamination, and furthermore, provides improved RNA-Seq detection sensitivity and measurement precision, which is necessary for differentiating biological variability from technical noise. By combining coherent Raman microscopy with RNA sequencing, we can better understand the relationship between cellular morphology and gene expression at the single-cell level.

Microfluidics facilitated genome sequencing for limited number of cells

We develop a microfluidics-based single cell RNA-Seq transcriptome analysis technology to perform the library-prep reaction steps at nanoliter range within sealed chambers on-chip, eliminating potential contaminations and sophisticated manual handlings.