Single-Cell Sequencing for Precise Cancer Research: Progress and Prospects Xiaoyan Zhang1, Sadie L

Published OnlineFirst March 3, 2016; DOI: 10.1158/0008-5472.CAN-15-1907 Cancer Review Research

Single-Cell Sequencing for Precise Cancer Research: Progress and Prospects Xiaoyan Zhang1, Sadie L. Marjani2, Zhaoyang Hu1, Sherman M. Weissman3, Xinghua Pan1,3,4, and Shixiu Wu1

Abstract

Advances in genomic technology have enabled the faithful could potentially improve the early detection and monitoring detection and measurement of mutations and the gene expres- of rare cancer cells, such as circulating tumor cells and dissem- sion profile of cancer cells at the single-cell level. Recently, inated tumor cells, and promote the development of person- several single-cell sequencing methods have been developed alized and highly precise cancer therapy. Here, we discuss that permit the comprehensive and precise analysis of the the current methods for single cancer-cell sequencing, with a cancer-cell genome, transcriptome, and epigenome. The use of strong focus on those practically used or potentially valuable these methods to analyze cancer cells has led to a series of in cancer research, including single-cell isolation, whole unanticipated discoveries, such as the high heterogeneity and genome and transcriptome amplification, epigenome profiling, stochastic changes in cancer-cell populations, the new driver multi-dimensional sequencing, and next-generation sequenc- mutations and the complicated clonal evolution mechanisms, ing and analysis. We also examine the current applications, and the novel identification of biomarkers of variant tumors. challenges, and prospects of single cancer-cell sequencing. These methods and the knowledge gained from their utilization Cancer Res; 76(6); 1–8. 2016 AACR.

Introduction disease progression (4). However, the molecular profiling data, especially heterogeneity analysis, of these important cancer-cell Cancer is a significant cause of mortality worldwide. Cur- populations are limited. When dominant clones are killed in rently, clinical therapies do not produce complete resolution cancer therapy, other minor clones may replace them and lead to for most cancer patients due to insufficient understanding of cancer recurrence (5). Dynamic changes in cancer niches, which the molecular mechanisms for carcinogenesis and tumor pro- comprise various cell types, can affect the development, progres- gression (1). Single-cell sequencing (SCS) includes a set of sion, and metastasis of cancer (2), but the identification, charac- powerful and comprehensive techniques that can provide terization, and isolation of the distinct cancer niche components insight into these unknowns. is very difficult. Thus, a high-resolution strategy to effectively Important types of cancer cells include primary tumor cells, screen and systematically analyze the heterogeneity of these metastatic tumor cells, cancer stem cells (CSC), circulating tumor different cancer cells and cancer niche cells is desired and expected cells (CTC), and disseminated tumor cells (DTC; Fig. 1A). Primary to lead to a wave of unanticipated discoveries. tumor cells can often be described as variants of subpopulations Because mutagenesis, gene regulation, and genetic modifica- (1). Metastatic tumor cells usually include the original tumor cell tion occur intrinsically at the single-cell level, SCS will enable clones and differentially invasive and secondarily mutated cell highly precise, molecular characterization of carcinogenesis and clones. CSCs, which are often very rare, can undergo self-renewal, will hopefully resolve the important problems described above. are therapy resistant, and may cause the recurrence of cancer, This will lead to the identification of new biomarkers and drug can also be characterized into different cell subpopulations (2). targets and, in turn promote cancer diagnosis, monitoring, and CTCs and DTCs play a vital role in cancer dissemination, self- therapy (6). In this review, we summarize SCS methods recently renewal, and distant metastases (3). Cancer cells dispersed in an developed, including those that could be extremely useful, but individual or even forming a single tumor or a clone are often have not been applied in sequencing of single cancer cells. The heterogeneous, and this heterogeneity changes dynamically with clinical applications as well as the current problems and limitations of these technologies for use in the study of cancer are also 1Hangzhou Cancer Institution, Hangzhou Cancer Hospital, Hangzhou, discussed. Zhejiang Province, P. R. China. 2Department of Biology, Central Con- necticut State University, New Britain, Connecticut. 3Department of Isolation of Single Cancer Cells Genetics,Yale University School of Medicine, New Haven, Connecticut. 4Department of Biochemistry, School of Basic Medical Sciences, South- Analysis of a single cancer cell initially requires the isolation of ern Medical University, Guangzhou, Guangdong Province, P.R. China. the desired type of cells based on classic phenotypes or surface Corresponding Authors: Xinghua Pan, Yale University, 300 Cedar Street, TAC- markers, in a way that preserves the biologic integrity of the cell. S320 New Haven, CT 06520. Phone: 203-589-1087; Fax: 203-737-2286; E-mail: When isolating single cells from solid tumors, dispersion of the [email protected], and Shixiu Wu, [email protected] tumor tissue into a single-cell suspension is necessary. Mechanical doi: 10.1158/0008-5472.CAN-15-1907 separation and enzymatic treatments are often used (7, 8; Fig. 2016 American Association for Cancer Research. 1B1). Isolation of CTCs from bodily fluids involves two types of

www.aacrjournals.org OF1

Zhang et al.

A Primary tumor

Cancer cell Cancer niches Hematological tumor Solid tumor Metastatic tumor Peripheral blood types Tumor

Tumor

Epithelial cell Progenitor cells Cancer cells Disseminated Red blood cell Endothelial cell Cancer-associated ﬁbroblast Biopsy materials tumor cells White blood cell Extracellular matrix Immune inﬂammatory cells Cancer stem cells Blood platelet Circulating tumor cell

B1 Collagenases “Cloudy” interface layer “Cloudy” interface layer from bone marrow from peripheral blood Isolation of Tumor Tumor pieces single cancer cells

(B1-1) (B1-2) (B1-3)

FACS Microﬂuidics systems B2 Multispectral (i) Cancer circulating cell sorting (ii) Single cell trapping, lysis, RT and seq detectors Four Anti-EpCAM Cell Lysis buffer isolation Illumina methods trapping Laser RT seq Nano-magnetic Cell lysis reagents particle Discarded cells Magnet CTC RT

Micromanipulation system Laser capture microdissection (11) (i) Micromanipulators (i) Cell selecting (ii) Laser cutting (iii) Cells on a transfer membrane

(ii) Manual isolation

CC C RT C=CpG Codetection: Future: GG scWGS scWES scGT-seq scRNA-seq scMT-seq scM-seq Single cell scATAC(60) sequencing (i) MDA (6, 14) (i) Physical (i) Full-length (i) Bisulﬁte-based MIDAS (15) separation-based Smart-Seq (30, 31) scRRBS (56, 57) Nuc-seq (16) Han’s method (64) PTA/STA (32) scBS (58) (ii) MALBAC (6, 18) G&T-seq (65) (ii) UMI-based … scWGBS (59) (iii) PCR (6, 17) (ii) Non-physical Islam’s UMI-WTA (34) (ii) Non-bisulﬁte-based separation-based (iii) in situ Seq … DR-seq (66) FISSEQ (35) Padlock-RCA (36) smFISH (37, 38) TIVA (39) (scWGS) (iv) Multiplexed STRT-seq (40, 41) Exome capture CEL-seq (42) (scWES) Quartz-seq (43) Fiuidigm C1 (12, 44) DropSeq (45, 46)

Library construction and/or Sequencing / and Bioinformatic analysis

OF2 Cancer Res; 76(6) March 15, 2016 Cancer Research

Single-Cell Sequencing of Cancer

methods (Fig. 1B2). The first is biomarker dependent, such as the The MDA method is most popularly applied in WGA due to its CellSearch magnetic bead system. The second uses density-gra- high fidelity and simplicity. Several updated MDA kits are suitable dient centrifugation, microfiltration, or microfluidics systems. for single-cell WGA (scWGA), such as REPLI-g Single Cell Kit, The CTC capture platform CTC-IchIP, which relies on the inherent GenomiPhi DNA Amplification Kit, and AmpliQ Genomic differences in morphology or size between leukocytes and CTCs Amplifier Kit (14). Recently, a MDA system, called the micro- (9), can be used to isolate some CTCs without known biomarkers. well displacement amplification system (MIDAS), enabled the For DTC isolation, needle aspiration is used to collect bone simultaneous amplification of thousands of single cells marrow followed by single-cell sorting (10). through the use of hundreds to thousands of nanoliter wells, Single-cell sorting methods, including marker/phenotype- and the result was more robust with improved uniformity (15). based methods for isolating single cells from bulk cell popula- Amodified MDA approach, Nuc-seq, took advantage of the fact tions [FACS, microfluidics, micromanipulation, and laser cap- that a single cell in the G2–M stage of the cell cycle has four ture microdissection (LCM); Fig. 1B2] and from rare cell popu- copies of the genome and significantly increased the genome lations (CellSearch, DEP-Array, CellCelector, MagSweeper, and coverage (16). For PCR-based scWGA methods, the formulated nanofilters), have been well summarized previously (10, 11). kits such as the Sigma GenomePlex kit and the Silicon Ampli Unbiased, marker-free decomposition of a tissue or organ, such 1 WGA Kit are often used for CNV/SV analysis; however, these as an intact tumor, into different types of single cells (12) may are generally used less than MDA methods because they may omit the FACS process but still needs dissociation of the tissues generate nonspecific artifacts, have higher error rates, and give into single cells, or localization of the single cells (13), and incomplete coverage of the genome (17). MALBAC was recently usually requires the analysis of a very large number of cells. used to amplify the DNA from single SW480 cells (a colon These types of methods would allow us to clarify the architec- cancer cell line). The results showed uniform, high sequencing ture of a whole tumor with its variants of cancer niches, such as coverage (>90%) across the human genome, with a consider- immune cells. Ultimately, the technology chosen for the iso- able increase in the detection of CNA and SV over MDA lation of single cells is dependent upon the scientific conditions performed in the same laboratory (18). of the study and sample type. Single-cell WES versus single-cell WGS The exome constitutes approximately 1% of the whole genome Single-Cell Whole Genome Amplification and contains 85% of the known disease-causing variants (19). and Sequencing and Whole Exome WES involves exome capture and enrichment after genome ampli- Sequencing fication, and it is significantly more efficient than WGS for sequencing of coding sequences. Single-cell WES (scWES) is even Until recently, most cancer SCS applications utilized whole more useful for coding mutation detection because scWES nor- genome sequencing (WGS) or whole exome sequencing (WES) to malizes the profound locus-to-locus bias introduced during fi identify speci c single-nucleotide variations/mutations (SNV), scWGA and increases sequence base coverage. Single-cell WGS copy number aberrations (CNA), and structural variants (SV). (scWGS) has been applied for CNA and SNV screening, and is This work provided insights into the driver mutation, biomarkers, particularly powerful for CNA and non-coding sequence detec- and the detailed route and mechanism of carcinogenesis and tion (20). scWES has mostly been used for SNV analysis of coding metastasis (6). sequences, but can also be used for CNA detection, although it is less efficient (21). Methods for whole genome amplification For scWES, the bait libraries used in exome capture can be A normal human cell contains only 6 pg of DNA; this amount created using established commercial kits, such as NimbleGen's is insufficient for genomic analysis (6). Therefore, whole genome SeqCap, Agilent SureSelect, Illumina TruSeq, and Nextera Exome. amplification (WGA) is needed to amplify the DNA hundreds of A comparison of these four kits with the same human tumor DNA thousand fold. Three major types of WGA methods (Fig. 1C) have sample (22) showed that the Illumina kits were the most effi- been applied with SCS: multiple displacement amplification cient—they captured more bases in coding and translated regions. (MDA), PCR, and a comprehensive method that combines dis- The Nextera Exome Enrichment Kit is a fast and efficient system placement amplification and PCR, such as the PicoPLEX single- that combines easy sample preparation with TruSeq's exon enrich- cell WGA kit and multiple annealing and looping-based ampli- ment process. It does not necessitate mechanical DNA fragmen- fication cycles (MALBAC; ref. 6). tation and so requires a small amount of input genomic DNA.

Figure 1. The skeleton diagram of this review. A, an assemblage of the distinct cell types that constitute cancer niches (cancer microenvironment), primary tumor (including hematologic tumor and solid tumor), metastatic tumor and the peripheral blood of a cancer patient. B, methods for isolation of single (cancer) cells. B1, preparation of single-cell suspension. B1-1, solid tumor single-cell suspension by cutting biopsy material or surgical tumor into pieces, which are then dissociated by collagenases or papain dissociation system, etc. B1-2, single-cell suspension of the bone marrow "cloudy" interface layer for tumor of the blood system. B1-3, single- cell suspension of the peripheral blood "cloudy" interface layer for CTCs, which mainly are of solid tumors. B2, four popular methods for single (cancer) cell isolation: FACS, microfluidics systems, micromanipulation system, and LCM (27). C, single-cell sequencing. Amplification methods for SCS of whole genome (scWGS), whole exome (scWES), whole transcriptome (scRNA-seq), whole DNA methylation (scM-seq), and whole assay for transposase-accessible chromatin (scATAC). A method for codetection of the whole genome and whole transcriptome from a single cell is also listed (scGT-seq). A putative method for non–bisufite-based scM-seq (maybe less damaging and with better coverage) and a method for scMT-seq (codetection of DNA methylome and transcriptome of a single cell) are proposed (no reports seen yet). After amplification, seq library for the genome-wide material or a panel of targets (not shown) is prepared and next-generation sequencing, such as Illumina HiSeq/Miseq platform or Thermo Fisher Scientific Ion Torrent Sequencer, is applied. Finally, bioinformatics analysis is conducted, and important biologic information is revealed.

www.aacrjournals.org Cancer Res; 76(6) March 15, 2016 OF3

Zhang et al.

However, it has been found to introduce bias when the target to promote high-throughput RNA-seq such as STRT-seq regions have a high GC content (22). (40, 41), CEL-seq (42), and Quartz-seq (43) involved the Currently, massively parallel sequencing is predominately per- pooling/multiplexing of single cells each barcoded at an early formed on two platforms, Illumina's HiSeq/MiSeq platform and stage of scWTA (29). In recent years, novel instruments with Thermo Fisher Scientific's Ion Torrent sequencer. HiSeq is widely automatic operation in combination with cell barcoding, UMI, used because it can produce the largest amount of sequencing and nanoliter-size reaction volume stretch the multiplexing reads at relatively low cost (23). On the other hand, Ion Torrent ability to parallel analysis of 96 or hundreds (Fluidigm C1; provides easy target-sequencing and is the gold standard for refs. 12, 44) to thousands of single cells at very low cost per cell. mutation detection in clinical genetic testing. The sequencing The latter is represented by the recently reported droplet-based depth required depends on the goals of the specific experiment. sequencing, or Drop-seq (45, 46). A shallow RNA-seq could For example, CNA or SV analysis in scWGS requires relatively light faithfully represent the transcriptomic character of single cells sequencing (as little as 0.1–2 fold depth), whereas SNV in scWGS/ and detect the cellular heterogeneity and activated signaling scWES needs deeper sequencing (typically, 30–50 fold; ref. 24). In pathways; as low as 50,000 reads per cell is sufficient for addition, the Pacific Biosciences' PacBio RS II and the Oxford unsupervised cell-type classification (44). Therefore, these Nanopore's MiniON technologies provide super length, single- highly multiplex methodologies allow for super–high-through- molecule real-time (SMRT) sequencing, which is extremely useful put and financially efficient analysis of many cancer cells, in de novo and full-length sequencing and may lead to novel potentially revealing inherent dynamic architecture, heteroge- discovery of haplotypes, isoforms, etc., in cancer genomics. Once neous molecular mechanisms and biomarkers. the sequencing is complete, the raw sequencing reads are aligned qRT-PCR is used to quantify a panel of genes in single cells to the reference genome and then mapped, followed by down- and may have even higher sensitivity for these selected targets stream bioinformatic analysis (20). compared with scRNA-seq (47, 48); sequencing of a panel of PCR-selected target genes is another choice in expression Single-Cell Whole Transcriptome profiling of single cells. fi Finally, the scRNA-seq data have to be interpreted by appro- Ampli cation and scRNA-Seq priate computational and statistical methods, which has been Single-cell RNA sequencing (scRNA-seq) data reveal not only a challenge and the focus of much attention (49). the quantitative and global expression profile, but exon splicing fi and allele-speci c expression with appropriate sequencing depth Single-Cell Epigenomic Sequencing and analysis (25, 26). scRNA-seq analysis provides high resolution of the molecular phenotype and plays an important role in Epigenetics is defined as the sum of heritable changes to the revealing the functional heterogeneity or diversity of a population DNA that do not cause direct alteration of the primary DNA of cells, marker-based types, and uncovering the molecular char- sequence. Such changes include DNA methylation, histone acters, specific signals, and pathways underlying the process of modification, chromatin-binding structural and regulatory pro- cancer. scRNA-seq data can also be used to detect mutations (27). teins, spatial structures of the chromosomes and spatial inter- Generally, a human cell contains <1 pg mRNA, and more than actions of components forming transcriptional complexes. 85% of the transcripts are present at less than 100 copies (28). Epigenomic analysis reveals different aspects of gene regula- Single-cell whole transcriptome amplification (scWTA) amplifies tion, including gene interactions at different levels. A great transcript abundance allowing for characterization by sequenc- number of diseases are attributed to epigenetic abnormalities ing. A number of approaches have been developed for scWTA (Fig. (50). Epigenetic alterations are reversible and dictate gene 1C; ref. 29). Three of the scWTA methods have been able to expression and are responsible for cellular heterogeneity amplify full-length transcripts (6): Smart-Seq or Smart-Seq2 (51). Therefore, single-cell epigenetics will have an important (commercial kit: SMARTer Ultra Low RNA Kit; refs. 30, 31), role in cancer studies and clinical investigation. Phi29-transcriptome amplification (PTA) and semi-random primed PCR-based transcriptome amplification (STA; ref. 32). Single-cell DNA methylation sequencing However, at the single-cell level, full-length sequences are rarely DNA methylation predominately refers to the addition of the obtained for most transcripts. In addition, most scRNA-seq meth- methyl group to the cytosine (5-methylcytosine, 5mC) of a CpG ods suffer from amplification bias (33). Islam and colleagues (34) dinucleotide in CpG islands (CGI) by DNA methyltransferases established a unique molecular identification (UMI) technique to (52), although the role of C methylation outside of CGIs and non- directly count the original molecules in single cells with WTA. CpG methylation is also recognized (53) as well as 5-hydroxy- UMI greatly improved the accuracy, quantification, and digital methylcytosine (54). Specific tumor-suppressor genes are usually feature of WTA and was highly efficient. Furthermore, in order to significantly hypermethylated at their promoter regions or asso- retain the spatial and temporal information of RNAs in cells, ciated CGIs during carcinogenesis (55). Meanwhile, the hypo- several new RNA sequencing methods have been developed, methylation of repetitive DNA sequences across the genome and including fluorescent in situ RNA sequencing (FISSEQ; ref. 35), of specific oncogenes is the other side of cancer epigenomic Padlock-RCA (36), single molecule fluorescent in situ hybridiza- regulation, and overall dominates cancer genome (55). tion (smFISH; refs. 37, 38), and single-cell transcriptome in vivo Many DNA methylation detection methods require a large analysis (TIVA; ref. 39). number of cells. Thus, the epigenetic profiles at present are The analysis of thousands or tens of thousand individual averaged, and the cellular heterogeneity of the population escapes cells promises to decode the architecture of a complicated evaluation. Since 2013, three single-cell DNA methylation system (12, 13). This is particularly true for cancer studies due sequencing (scM-seq) methods have been reported (Fig. 1C). to the highly heterogeneous nature of cancers. The early effort Single-cell reduced representation bisulfite sequencing (scRRBS)

OF4 Cancer Res; 76(6) March 15, 2016 Cancer Research

Single-Cell Sequencing of Cancer

developed by Guo and colleagues (56, 57) enabled the detection sibility track, and found their dynamic association with cell type– of 0.5 to 1.5 million CpG sites (CpGs) within the genome of a specific trans-factors and cis-elements. The single-cell level of single cell, which on average was 40% of the CpGs that bulk RRBS detection of chromosome organization in the three-dimensional can detect, corresponding to 10% of all CpGs in a mouse genome. space of the nucleus showed cell-to-cell variability in genome- Smallwood and colleagues (58) reported a single-cell genome- wide chromosome conformation capture (Hi-C), while the indi- wide bisulfite sequencing method (scBS), covering 3.7 million vidual chromosomes maintain domain organization at the mega- CpGs (range, 1.8–7.7 million) corresponding to 17.7% of all base scale (61). This method filled the gaps between genomics and CpGs in a mouse genome. scBS used a post-bisulfite adaptor microscopy analyses of chromosomes, enabling the study of gene tagging (PBAT) protocol, avoiding damage of the two ligated regulation of cancer based on chromatin organization. A few other adapters during the bisulfite treatment and improved PCR ampli- tools for epigenetic analyses of single-cell chromatin have also fication. Farlik and colleagues (59) further extended the scBS been reported, such as the single chromatin molecule analysis at method, named scWGBS, with direct library construction after the nanoscale (SCAN) platform (62), and methods using anti- bisulfite conversion, low-coverage sequencing (capturing 1.54 body-based individual intact chromatin fiber labeling, capture million CpGs with 4.6 million reads on average), and a supervised and image analysis (63). bioinformatics strategy using in silico merging of the data from In summary, single-cell epigenetic assays are very promising multiple single cells to increase the overall coverage. tools that can assess the cellular variation of the "regulome" Overall, these approaches provide single-nucleotide resolution associated with the complicated phenotypes of human can- of CpG methylation patterns, representing an exciting start to cers. Further progress in the methodology will promote such scM-seq and showing its significance in biologic analysis. How- studies. ever, the consistency of CpG, CGI, or promoter detection across multiple single cells represents only less than a few percent of the Single-Cell Genome and Transcriptome sequences of the human genome, which is too low to enable an unsupervised clustering of single cells for a heterogeneous Codetection and Sequencing population like cancer. This probably is attributable to the nature The integrated and simultaneous measurement of the genome of the bisulfite-conversion–based procedure, and the stochastic and transcriptome, and probably epigenome of a single cell, can sequencing of the genomic sequences, especially in scBS and lead to a multidimensional analysis of heterogeneity, stratifica- scWGBS. The scBS covers CpGs and CGIs better than scWGBS tion, and phenotype regulation associated with cancer. Currently, and scRRBS, with better consistency across multiple cells, but with three integrated single-cell genome and transcriptome codetec- the cost of deep sequencing of the whole genome. scRRBS is the tion and sequencing (scGT-seq) methods have been reported most cost efficient choice for consistent detection of CGIs across a (Fig. 1C). panel of single cells. The supervised analysis, using in silico In 2014, Han and colleagues (64) first developed a micro- merging of multiple single cells that are known to be in the same fluidics-facilitated scGT-seq approach that used an optimized subpopulation, or aggregating cohorts of sequence blocks with lysis buffer to separate the cytoplasmic content and the intact presumably similar functions, surely increased the CpGs and nucleus of a single cell. After separation, the genomic DNA and probably also CGIs coverage, and consistent detection among mRNA were MDA amplified separately. The power of this the single cells in parallel. However, for scM-seq application in technique is in its controlled separation strategy that can be cancer, we will not know in advance any subpopulation structure coupled with any kind of existing scWGS and scWTS protocol to of the single cells analyzed, and we primarily require an unsu- obtain the genomic and transcriptomic information from the pervised clustering analysis. Therefore, a scM-seq method robust same cell. In the future, this method could potentially be used enough to cover sufficient CGIs or CpGs consistently across many to analyze the correlation between DNA mutation (or meth- single cells is still needed in spite of the great value of these ylation or chromatin) and transcription. Recently, a similar available methods. Further development could involve a sub- physical separation-based method, named G&T-seq (genome stantially revised bisulfite-sequencing method, or probably a non- and transcriptome sequencing), was reported (65). After cell BS–based method. Other extensions of scM-seq, such as detection lysis, G&T-seq used a biotinylated oligo-dT primer to separate of DNA hydroxymethylation (5hmC; ref. 54), will also be an polyadenylated (polyA) RNA from gDNA and then (polyA) interesting direction of development. mRNA was amplified by Smart-Seq2, while the gDNA was amplified by MDA or PicoPlex before sequencing (65). In Single-cell chromatin structure sequencing 2015, Dey and colleagues (66) reported a quasilinear ampli- The dynamics of chromatin architecture regulates cellular pro- fication-based method, termed gDNA-mRNA sequencing (DR- cesses, including carcinogenesis and cancer evolution. Buenrostro Seq). DR-Seq is convenient in the sense that it can quantify and colleagues (60) adapted transposase-accessible chromatin scGT without physically separating the cytoplasm and nuclear analysis for single cells on a microfluidics platform (scATAC-Seq). contents before amplification and uses a one-tube reaction Here, a nucleus is incubated with Tn5 transposase, which enables during cell lysis and amplification to minimize the loss adaptor incorporation and PCR amplification of the open chro- of mRNA or gDNA (66), although the gDNA may not be fit matin regions for sequencing at single nucleotide–pair resolution. for DNA methylation profiling.Further,DR-seqdevelopeda Aggregate analysis of 254 individual GM12878 lymphoblastoid computational strategy to determine CNVs by eliminating the cells recapitulated the patterns of accessible chromatin observed interference of the mRNA reads. Undoubtedly, simultaneous in bulk cells (60). Although the consistent sequence coverage is analysis of genomics and/or epigenomics (CpG methylation or not that high—about 9.4% of promoters are represented in a chromatin) and transcriptomics in single cells will lead to single cell—scATAC-seq confidently detected the variability of a single-cell omics, providing new insights into the molecular set of open chromatin peaks identified using the aggregate acces- understanding of cancer.

www.aacrjournals.org Cancer Res; 76(6) March 15, 2016 OF5

Zhang et al.

Preliminary Applications Have Led to The limitations of the above-discussed SCS methods pose fi Exciting Discoveries other key challenges. The rst limitation is the low coverage, which is a common scenario when scWGA, scWTA, scM, or sc Recently, a series of exciting studies utilizing SCS in some major chromatin profile amplicons are sequenced. Aside from current cancers were reported, such as breast cancer (16, 47), colon cancer DNA methylation methods that suffer from low coverage, even (18, 67), muscle-invasive bladder cancer (68), intestinal cancer the commonly used full-length RNA amplification kit, Smart- (69), lung adenocarcinoma cancer (27), kidney tumor (70), and Seq, when applied to a large number of single cells, generally, acute myeloid leukemia (71, 72). These studies made unprece- gives mostly fragments of sequences that are far from full- dented progress in understanding carcinogenesis, progression, length. More critically, with the Smart-Seq method, only metastasis, and drug resistance. 10% efficiency of detection can be obtained for any sequence Interestingly, some genes discovered by SCS exhibited a high of an expressed transcript. This is due to the damage, loss, and frequency of mutations at the single-cell level, but low preva- low efficiency during the single-cell processes, and also the lence in the tumor mass (16, 67). Some genes exhibited specific locus-to-locus bias and allele bias in the amplification. There- alternative splicing of exon isoforms in different single-cell fore, the development of more sensitive methods that increase transcriptomes (30). Allele-specific expression occurs in cancer the overall coverage, and the consistency of the coverage (73), but has not been reported with SCS data, probably due to between single cells should be the initial objective for research- technical limitations. These are critical in cancer progression ers. The second limitation is the core technologies available for and potentially can be used as candidate markers for cancers. the comprehensive molecular analysis of a single cell. For More in-depth single-cell analysis for carcinogenesis explo- example, up to now, the technology for simultaneously char- ration will be useful to acquire necessary knowledge of different acterizing a cell's genome, transcriptome, and DNA methylome rare and dynamic cancer cells (CTCs, DTCs, and other tumor profile is at the infant stage, and few technologies have the cells) and metastatic processes. A comprehensive summary of ability to robustly detect the whole proteome of a single cell reported analysis of CTCs, including SCS-based CTCs studies (77). This ability would enable the synchronized delineation of was published recently (74). Remarkably, a recent SCS analysis heterogeneity at multidimensional levels and reveal complex of CTCs revealed heterogeneous signaling pathways and het- relationships. The third challenge is temporal and spatial erogeneous drug-resistance mechanisms in advanced prostate measurement of the molecular profile in a single cell. In situ cancer (75). These results could significantly improve cancer sequencing and real-time sequencing, as well as in vivo analysis therapeutic efficacy and promote CTCs to become a noninva- of the DNA and RNA from single cells, have been developed, sive tool for cancer staging, monitoring and precise therapy. but these methods need enhanced sensitivity, coverage, and DTCs are another important cell type in carcinogenesis explo- accessibility, and a reduction in cost (78, 79). The ideal in vivo ration. SCS uncovered new and subtle genetic changes in DTCs technology for single-cell analysis would provide real-time by comparing them with the primary tumor and normal tissue dynamics, rather than a snapshot profile as demonstrated with (76). SCS identified aneuploidy, chromosomal rearrangements, other in vitro technologies. To this end,progresshasbeenmade and point mutations where the occurrence influenced the evolu- with just a limited number of transcripts (80). The fourth tion of the cancer (16). challenge involves data analysis. Most algorithms currently Validation and characterization of the heterogeneity of can- used for SCS analysis were originally designed for bulk cell cer is a powerful application of SCS. With SCS, some cancers samples. These analysis methods do not take into account the were found to be of biclonal origin (such as colon cancer; inherent properties of a single cell or any amplification- or ref. 67) or monoclonal-to-biclonal origin (such as muscle- sequencing-introduced biases, noise, incomplete coverage, or invasive bladder cancer; ref. 68), while others consisted of a errors, which have to be considered before a true conclusion is monoclonal population of cells (such as kidney tumor; ref. 70). drawn. Therefore, enhanced algorithms are needed to meet the Additionally, the deep SCS of numerous individual cells was new era of SCS (20). able to trace the lineages of a cancer (acute myeloid leukemia; In summary, the described SCS techniques are new powerful ref. 71). Non-genome scale methods, such as targeted PCR or approaches that provide more precise genomic analysis of cancer. target sequencing of a panel of markers of mutation, expres- The initial application of some of these methods has brought sion, or epigenomics for a number of samples will be a logical forth exciting discoveries about carcinogenesis, and clinicians follow up to the genome-wide scanning analyses and may be have initiated the use of these tools for cancer detection, diag- more practically useful in clinical practice, but are beyond the nostics, and therapeutic targeting. With further technical improve- discussion of this current review. ments, SCS methods will provide unprecedented insights into the characteristics of various cancers with the potential to uncover Challenges and Prospects their etiology, new markers and drug targets, and ultimately SCS promises an unprecedented ability to lead to more effi- reduce the morbidity of this plight on humanity. cient, precise, and successful cancer therapies. However, the After this paper was revised, new technologies for single-cell technology still faces several challenges. chromatin profiling were published: single-cell chromatin immu- The major application drawback of SCS in clinics is the high noprecipitation sequencing (scChIP-seq; ref. 81) and scDNase- cost of sequencing and the platforms appropriate for different seq (82). These epigenomic parameters (including scChIP-seq, clinical requirements such as different cancers and different scDNase-seq, and scATAC-seq) can also be codetected with the purposes—screening, detection, monitoring, or personal therapy transcriptome in single cells. guidance. Luckily, the dramatic progress of technology and mar- ketplace competition have driven down the cost of sequencing, Disclosure of Potential Conflicts of Interest and more kits are now available. No potential conflicts of interest were disclosed.

OF6 Cancer Res; 76(6) March 15, 2016 Cancer Research

Single-Cell Sequencing of Cancer

Acknowledgments for Natural Sciences (No. LZ15H220001), and Zhejiang Science and The authors thank Hongjin Wu and Ling Yang for their suggestions in single- Technology Planning Project of Health and Medicine (No. 2015PYA009), cell RNA-seq and bioinformatic analysis, respectively. and the US National Institutes of Health Grant 1P01GM099130-01 and R01DK100858. Grant Support This work was supported by the National Natural Science Found- Received July 13, 2015; revised October 17, 2015; accepted October 22, 2015; ation of China (No. 81402529), the Zhejiang Provincial Foundation published OnlineFirst March 3, 2016.

References 1. Saunders NA, Simpson F, Thompson EW, Hill MM, Endo-Munoz L, Leggatt 22. Chilamakuri CSR, Lorenz S, Madoui M-A, Vodak D, Sun J, Hovig E, et al. G, et al. Role of intratumoural heterogeneity in cancer drug resistance: Performance comparison of four exome capture systems for deep sequenc- molecular and clinical perspectives: Contribution of intratumoural het- ing. BMC Genomics 2014;15:449. erogeneity to drug resistance. EMBO Mol Med 2012;4:675–84. 23. Navin N, Hicks J. Future medical applications of single-cell sequencing in 2. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell cancer. Genome Med 2011;3:31. 2011;144:646–74. 24. Sims D, Sudbery I, Ilott NE, Heger A, Ponting CP. Sequencing depth and 3. Kang Y, Pantel K. Tumor cell dissemination: emerging biological insights coverage: key considerations in genomic analyses. Nat Rev Genet from animal models and cancer patients. Cancer Cell 2013;23:573–81. 2014;15:121–32. 4. Gerlinger M, Rowan AJ, Horswell S, Larkin J, Endesfelder D, Gronroos E, 25. Marinov GK, Williams BA, McCue K, Schroth GP, Gertz J, Myers RM, et al. et al. Intratumor heterogeneity and branched evolution revealed by multi- From single-cell to cell-pool transcriptomes: stochasticity in gene expres- region sequencing. N Engl J Med 2012;366:883–92. sion and RNA splicing. Genome Res 2014;24:496–510. 5. Wang E, Zou J, Zaman N, Beitel LK, Trifiro M, Paliouras M. Cancer systems 26. Borel C, Ferreira PG, Santoni F, Delaneau O, Fort A, Popadin KY, et al. biology in the genome sequencing era: part 1, dissecting and modeling of Biased allelic expression in human primary fibroblast single cells. Am J tumor clones and their networks. Semin Cancer Biol 2013;23:279–85. Hum Genet 2015;96:70–80. 6. Van Loo P, Voet T. Single cell analysis of cancer genomes. Curr Opin Genet 27. Kim K-T, Lee HW, Lee H-O, Kim SC, Seo YJ, Chung W, et al. Single-cell Dev 2014;24:82–91. mRNA sequencing identifies subclonal heterogeneity in anti-cancer drug 7. Dalerba P, Kalisky T, Sahoo D, Rajendran PS, Rothenberg ME, Leyrat AA, responses of lung adenocarcinoma cells. Genome Biol 2015;16:127. et al. Single-cell dissection of transcriptional heterogeneity in human colon 28. Macaulay IC, Voet T. Single cell genomics: advances and future perspec- tumors. Nat Biotechnol 2011;29:1120–7. tives. PLoS Genet 2014;10:e1004126. 8. Zeisel A, Munoz-Manchado~ AB, Codeluppi S, Lonnerberg€ P, La 29. Liu N, Liu L, Pan X. Single-cell analysis of the transcriptome and its Manno G, Jureus A, et al. Brain structure. Cell types in the mouse application in the characterization of stem cells and early embryos. Cell cortex and hippocampus revealed by single-cell RNA-seq. Science Mol Life Sci CMLS 2014;71:2707–15. 2015;347:1138–42. 30. Ramskold D, Luo S, Wang Y-C, Li R, Deng Q, Faridani OR, et al. Full-Length 9. Ozkumur E, Shah AM, Ciciliano JC, Emmink BL, Miyamoto DT, Brachtel E, mRNA-Seq from single cell levels of RNA and individual circulating tumor et al. Inertial focusing for tumor antigen–dependent and –independent cells. Nat Biotechnol 2012;30:777–82. sorting of rare circulating tumor cells. Sci Transl Med 2013;5:179ra47. 31. Picelli S, Bjorklund€ ÅK, Faridani OR, Sagasser S, Winberg G, Sandberg R. 10. Navin NE. Cancer genomics: one cell at a time. Genome Biol 2014;15:452. Smart-seq2 for sensitive full-length transcriptome profiling in single cells. 11. Saliba A-E, Westermann AJ, Gorski SA, Vogel J. Single-cell RNA-seq: Nat Methods 2013;10:1096–8. advances and future challenges. Nucleic Acids Res 2014;42:8845–60. 32. Pan X, Durrett RE, Zhu H, Tanaka Y, Li Y, Zi X, et al. Two methods for full- 12. Jaitin DA, Kenigsberg E, Keren-Shaul H, Elefant N, Paul F, Zaretsky I, et al. length RNA sequencing for low quantities of cells and single cells. Proc Natl Massively parallel single cell RNA-Seq for marker-free decomposition of Acad Sci 2013;110:594–9. tissues into cell types. Science 2014;343:776–9. 33. Pan X. Single cell analysis: from technology to biology and medicine. Single 13. Achim K, Pettit J-B, Saraiva LR, Gavriouchkina D, Larsson T, Arendt D, et al. Cell Biol 2014;3:pii:106. High-throughput spatial mapping of single-cell RNA-seq data to tissue of 34. Islam S, Zeisel A, Joost S, La Manno G, Zajac P, Kasper M, et al. Quantitative origin. Nat Biotechnol 2015;33:503–9. single-cell RNA-seq with unique molecular identifiers. Nat Methods 14. Sørensen KM, Jespersgaard C, Vuust J, Hougaard D, Nørgaard-Pedersen 2014;11:163–6. B, Andersen PS. Whole genome amplification on DNA from filter paper 35. Lee JH, Daugharthy ER, Scheiman J, Kalhor R, Yang JL, Ferrante TC, et al. blood spot samples: an evaluation of selected systems. Genet Test Highly multiplexed subcellular RNA sequencing in situ. Science 2014;343: 2007;11:65–71. 1360–3. 15. Gole J, Gore A, Richards A, Chiu Y-J, Fung H-L, Bushman D, et al. Massively 36. Ke R, Mignardi M, Pacureanu A, Svedlund J, Botling J, W€ahlby C, et al. In parallel polymerase cloning and genome sequencing of single cells using situ sequencing for RNA analysis in preserved tissue and cells. Nat Methods nanoliter microwells. Nat Biotechnol 2013;31:1126–32. 2013;10:857–60. 16. Wang Y, Waters J, Leung ML, Unruh A, Roh W, Shi X, et al. Clonal evolution 37. Lubeck E, Cai L. Single-cell systems biology by super-resolution imaging in breast cancer revealed by single nucleus genome sequencing. Nature and combinatorial labeling. Nat Methods 2012;9:743–8. 2014;512:155–60. 38. Lubeck E, Coskun AF, Zhiyentayev T, Ahmad M, Cai L. Single cell in situ 17. Barker DL, Hansen MST, Faruqi AF, Giannola D, Irsula OR, Lasken RS, et al. RNA profiling by sequential hybridization. Nat Methods 2014;11:360–1. Two methods of whole-genome amplification enable accurate genotyping 39. Lovatt D, Ruble BK, Lee J, Dueck H, Kim TK, Fisher S, et al. Transcriptome In across a 2320-SNP linkage panel. Genome Res 2004;14:901–7. Vivo Analysis (TIVA) of spatially defined single cells in intact live mouse 18. Zong C, Lu S, Chapman AR, Xie XS. Genome-wide detection of single- and human brain tissue. Nat Methods 2014;11:190–6. nucleotide and copy-number variations of a single human cell. Science 40. Islam S, Kj€allquist U, Moliner A, Zajac P, Fan J-B, Lonnerberg€ P, et al. 2012;338:1622–6. Characterization of the single-cell transcriptional landscape by highly 19. Ng SB, Turner EH, Robertson PD, Flygare SD, Bigham AW, Lee C, et al. multiplex RNA-seq. Genome Res 2011;21:1160–7. Targeted capture and massively parallel sequencing of twelve human 41. Islam S, Kj€allquist U, Moliner A, Zajac P, Fan J-B, Lonnerberg€ P, et al. Highly exomes. Nature 2009;461:272–6. multiplexed and strand-specific single-cell RNA 50 end sequencing. Nat 20. Ning L, Liu G, Li G, Hou Y, Tong Y, He J. Current challenges in the Protoc 2012;7:813–28. bioinformatics of single cell genomics. Front Oncol 2014;4:7. 42. Hashimshony T, Wagner F, Sher N, Yanai I. CEL-seq: single-cell RNA-seq by 21. Jiang Y, Oldridge DA, Diskin SJ, Zhang NR. CODEX: a normalization and multiplexed linear amplification. Cell Rep 2012;2:666–73. copy number variation detection method for whole exome sequencing. 43. Sasagawa Y, Nikaido I, Hayashi T, Danno H, Uno KD, Imai T, et al. Quartz- Nucleic Acids Res 2015;43:e39. Seq: a highly reproducible and sensitive single-cell RNA sequencing

www.aacrjournals.org Cancer Res; 76(6) March 15, 2016 OF7

Zhang et al.

method, reveals non-genetic gene-expression heterogeneity. Genome Biol 63. Oana H, Nishikawa K, Matsuhara H, Yamamoto A, Yamamoto 2013;14:R31. TG, Haraguchi T, et al. Non-destructive handling of individual 44. Pollen AA, Nowakowski TJ, Shuga J, Wang X, Leyrat AA, Lui JH, et al. Low- chromatin fibers isolated from single cells in a microfluidic device coverage single-cell mRNA sequencing reveals cellular heterogeneity and utilizing an optically driven microtool.LabChip2014;14: activated signaling pathways in developing cerebral cortex. Nat Biotechnol 696–704. 2014;32:1053–8. 64. Han L, Zi X, Garmire LX, Wu Y, Weissman SM, Pan X, et al. Co-detection 45. Macosko EZ, Basu A, Satija R, Nemesh J, Shekhar K, Goldman M, et al. and sequencing of genes and transcripts from the same single cells Highly Parallel genome-wide expression profiling of individual cells using facilitated by a microfluidics platform. Sci Rep 2014;4:6485. nanoliter droplets. Cell 2015;161:1202–14. 65. Macaulay IC, Haerty W, Kumar P, Li YI, Hu TX, Teng MJ, et al. G&T-seq: 46. Klein AM, Mazutis L, Akartuna I, Tallapragada N, Veres A, Li V, et al. Droplet parallel sequencing of single-cell genomes and transcriptomes. Nat Meth- barcoding for single-cell transcriptomics applied to embryonic stem cells. ods 2015;12:519–22. Cell 2015;161:1187–201. 66. Dey SS, Kester L, Spanjaard B, Bienko M, van Oudenaarden A. Integrated 47. Eirew P, Steif A, Khattra J, Ha G, Yap D, Farahani H, et al. Dynamics of genome and transcriptome sequencing of the same cell. Nat Biotechnol genomic clones in breast cancer patient xenografts at single-cell resolution. 2015;33:285–9. Nature 2015;518:422–6. 67. Yu C, Yu J, Yao X, Wu WK, Lu Y, Tang S, et al. Discovery of biclonal origin 48. Lawson DA, Bhakta NR, Kessenbrock K, Prummel KD, Yu Y, Takai K, et al. and a novel oncogene SLC12A5 in colon cancer by single-cell sequencing. Single-cell analysis reveals a stem-cell program in human metastatic breast Cell Res 2014;24:701–12. cancer cells. Nature 2015;526:131–5. 68. Li Y, Xu X, Song L, Hou Y, Li Z, Tsang S, et al. Single-cell sequencing analysis 49. Stegle O, Teichmann SA, Marioni JC. Computational and analytical chal- characterizes common and cell-lineage-specific mutations in a muscle- lenges in single-cell transcriptomics. Nat Rev Genet 2015;16:133–45. invasive bladder cancer. GigaScience 2012;1:12. 50. Hyun B-R, McElwee JL, Soloway PD. Single molecule and single cell 69. Grun€ D, Lyubimova A, Kester L, Wiebrands K, Basak O, Sasaki N, et al. epigenomics. Methods 2015;72:41–50. Single-cell messenger RNA sequencing reveals rare intestinal cell types. 51. Kantlehner M, Kirchner R, Hartmann P, Ellwart JW, Alunni-Fabbroni M, Nature 2015;525:251–5. Schumacher A. A high-throughput DNA methylation analysis of a single 70. Xu X, Hou Y, Yin X, Bao L, Tang A, Song L, et al. Single-cell exome cell. Nucleic Acids Res 2011;39:e44. sequencing reveals single-nucleotide mutation characteristics of a kidney 52. Ziller MJ, Gu H, Muller€ F, Donaghey J, Tsai LT-Y, Kohlbacher O, et al. tumor. Cell 2012;148:886–95. Charting a dynamic DNA methylation landscape of the human genome. 71. Hughes AEO, Magrini V, Demeter R, Miller CA, Fulton R, Fulton LL, et al. Nature 2013;500:477–81. Clonal architecture of secondary acute myeloid leukemia defined by single- 53. Schultz MD, He Y, Whitaker JW, Hariharan M, Mukamel EA, Leung D, et al. cell sequencing. PLoS Genet 2014;10:e1004462. Human body epigenome maps reveal noncanonical DNA methylation 72. Paguirigan AL, Smith J, Meshinchi S, Carroll M, Maley C, Radich JP. Single- variation. Nature 2015;523:212–6. cell genotyping demonstrates complex clonal diversity in acute myeloid 54. Uribe-Lewis S, Stark R, Carroll T, Dunning MJ, Bachman M, Ito leukemia. Sci Transl Med 2015;7:281re2. Y, et al. 5-Hydroxymethylcytosine marks promoters in colon 73. Mayba O, Gilbert HN, Liu J, Haverty PM, Jhunjhunwala S, Jiang Z, et al. that resist DNA hypermethylation in cancer. Genome Biol. 2015; MBASED: allele-specific expression detection in cancer tissues and cell 16:69. lines. Genome Biol 2014;15:405. 55. Hovestadt V, Jones DTW, Picelli S, Wang W, Kool M, Northcott PA, et al. 74. Alix-Panabieres C, Pantel K. Challenges in circulating tumour cell research. Decoding the regulatory landscape of medulloblastoma using DNA meth- Nat Rev Cancer 2014;14:623–31. ylation sequencing. Nature 2014;510:537–41. 75. Miyamoto DT, Zheng Y, Wittner BS, Lee RJ, Zhu H, Broderick KT, et al. RNA- 56.GuoH,ZhuP,WuX,LiX,WenL,TangF.Single-cellmethylome Seq of single prostate CTCs implicates noncanonical Wnt signaling in landscapes of mouse embryonic stem cells and early embryos analyzed antiandrogen resistance. Science 2015;349:1351–6. using reduced representation bisulfite sequencing. Genome Res 2013; 76. Møller EK, Kumar P, Voet T, Peterson A, Van Loo P, Mathiesen RR, et al. 23:2126–35. Next-generation sequencing of disseminated tumor cells. Front Oncol 57. Guo H, Zhu P, Guo F, Li X, Wu X, Fan X, et al. Profiling DNA methylome 2013;3:320. landscapes of mammalian cells with single-cell reduced-representation 77. Shi Q, Qin L, Wei W, Geng F, Fan R, ShikShin Y, et al. Single-cell proteomic bisulfite sequencing. Nat Protoc 2015;10:645–59. chip for profiling intracellular signaling pathways in single tumor cells. 58. Smallwood SA, Lee HJ, Angermueller C, Krueger F, Saadeh H, Peat J, et al. Proc Natl Acad Sci U S A 2012;109:419–24. Single-cell genome-wide bisulfite sequencing for assessing epigenetic het- 78. Byrd TF, Hoang LT, Kim EG, Pfister ME, Werner EM, Arndt SE, et al. The erogeneity. Nat Methods 2014;11:817–20. microfluidic multitrap nanophysiometer for hematologic cancer cell char- 59. Farlik M, Sheffield NC, Nuzzo A, Datlinger P, Schonegger€ A, Klugh- acterization reveals temporal sensitivity of the calcein-AM efflux assay. Sci ammer J, et al. Single-cell DNA methylome sequencing and bioinfor- Rep 2014;4:5117. matic inference of epigenomic cell-state dynamics. Cell Rep 2015;10: 79. Nguyen TA, Yin T-I, Reyes D, Urban GA. Microfluidic chip with integrated 1386–97. electrical cell-impedance sensing for monitoring single cancer cell migra- 60. Buenrostro JD, Wu B, Litzenburger UM, Ruff D, Gonzales ML, Snyder MP, tion in three-dimensional matrixes. Anal Chem 2013;85:11068–76. et al. Single-cell chromatin accessibility reveals principles of regulatory 80. Buxbaum AR, Yoon YJ, Singer RH, Park HY. Single-molecule insights into variation. Nature 2015;523:486–90. mRNA dynamics in neurons. Trends Cell Biol 2015;468–75. 61. Nagano T, Lubling Y, Stevens TJ, Schoenfelder S, Yaffe E, Dean W, et al. 81. Rotem A, Ram O, Shoresh N, Sperling RA, Goren A, Weitz DA, et al. Single- Single-cell Hi-C reveals cell-to-cell variability in chromosome structure. cell ChIP-seq reveals cell subpopulations defined by chromatin state. Nat Nature 2013;502:59–64. Biotechnol 2015;33:1165–72. 62. Benitez JJ, Topolancik J, Tian HC, Wallin CB, Latulippe DR, Szeto K, et al. 82. Jin W, Tang Q, Wan M, Cui K, Zhang Y, Ren G, et al. Genome-wide detection Micro fluidic extraction, stretching and analysis of human chromosomal of DNase I hypersensitive sites in single cells and FFPE tissue samples. DNA from single cells. Lab Chip 2012;12:4848–54. Nature 2015;528:142–6.

OF8 Cancer Res; 76(6) March 15, 2016 Cancer Research

Single-Cell Sequencing for Precise Cancer Research: Progress and Prospects

Xiaoyan Zhang, Sadie L. Marjani, Zhaoyang Hu, et al.

Cancer Res Published OnlineFirst March 3, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-15-1907

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2016/03/03/0008-5472.CAN-15-1907. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.