Fluorescent In Situ Sequencing
The groundwork for FISSEQ was laid in 1999 by George Church at Harvard Medical School with the publication of the first spatially localized massively parallel PCR amplification of DNA templates¹. Church conducted PCR within a thin polyacrylamide film which limits diffusion of DNA molecules so that amplification products remain localized to the template molecule. They coined the term “polony”, a portmanteau of “polymerase” and “colony”, to describe these DNA amplicons. They also demonstrated solid-phase polony PCR by covalently linking one of the PCR primers into the gel. Church concluded the discussion of this landmark paper by highlighting that polony slides can be used for genomic sequencing by performing solid-phase fluorescent DNA extension in situ within the polony gel. They described this technique as fluorescent in situ sequencing (FISSEQ). They also note, presciently, that polony slides generated from cDNA could be used to measure whole genome mRNA expression from single cells.
In 2002, digital genotyping and haplotyping of polymerase colonies was demonstrated², and the FISSEQ method was subsequently published in 2003 ³ for de novo sequencing of DNA.
In the same year, FISSEQ was used for single molecule profiling of alternative pre-mRNA splicing, the first demonstration of RNA FISSEQ4. In 2005, FISSEQ was used to re-sequence the genome of an evolved strain of Escherichia coli. Instead of using polonies, FISSEQ was performed on beads, clonally amplified using emulsion PCR and then embedded in a thin acrylamide film. That paper reported two important innovations: automated sequencing and the SBL chemistry.
FISSEQ and polony technology continued to be developed in the Church lab throughout the late 2000s. Using this technology, three human genomes were sequenced in 2010 at an average reagent cost of only $4,400 each5. The continued development and use of this technology, both in the Department of Genetics at Harvard Medical School and at the Wyss Institute for Biologically Inspired Engineering, which was formed in 2010, was instrumental to the development of RNA FISSEQ inside intact cells and tissues.
The expression of many RNA transcripts, both coding and non-coding, is spatially regulated. Transcripts may be localized to a particular membrane, or genes may be expressed in a cell- or tissue-specific manner. Tissues exhibit highly heterogeneous gene expression patterns; heterogeneity in gene expression across a tissue can represent important functional differences or disease states. In addition, non-coding RNA, RNA splice isoforms, post-transcriptionally edited RNA, and parental allele-specific RNA can all be expressed in a tissue-specific manner.
While it is becoming easier to discover such variation, existing RNA localization methods are limited to a handful of genes per specimen, making it costly and laborious to localize RNA transcriptome-wide. Recognizing the potential for multiplex in situ measurement of gene expression, Nature Methods named in situ sequencing one of the technologies to watch in 2014.¹
Fluorescent in situ sequencing (FISSEQ) was originally proposed in 2003. Subsequently, the Church lab and Synthetic Biology Platform of the Wyss Institute developed methods to sequence DNA amplicons on a solid substrate for genome and transcriptome sequencing, resulting in open source DNA sequencing chemistry and hardware platforms. To extend these methods into fixed biological samples, our researchers developed novel sequencing library construction methods.
We developed a method to convert RNA into cDNA in situ within the cells, which is immobilized and amplified locally in a three dimensional space. The single molecule DNA amplicons are then cross-linked to form a stable matrix, resistant to chemical and thermal perturbation. We have formed these cDNA amplicon matrices in diverse sample types including human iPS cells, fibroblasts, brain and embryo tissue sections, as well as whole embryos. Using a variety of fluorescent sequencing chemistries, including those developed by the Wyss Institute, the sequence of each cDNA amplicon is optically determined, and the expression pattern of each gene is revealed within the natural context of the sample.
This technology represents a high-throughput fluorescent in situ RNA sequencing method (FISSEQ). Our technology allows one to visualize the whole transcriptome with a single nucleotide resolution in individual cells and tissues in situ and identify the functional variant clusters based on the expression pattern. In addition to clustering gene expression patterns based on their spatial organization, our platform enables detection of DNA or RNA based sequence barcodes expressed in multiple cell types under multiple experimental conditions in parallel in situ. FISSEQ enables study of the functional organization of tissues and organs, as well as the relationship between genotype and phenotype after genome editing
Using 30-base reads from 8742 genes in situ, we examined RNA expression and localization in human primary fibroblasts with a simulated wound-healing assay, detecting 12 differentially regulated genes. In February 2014, these results and the FISSEQ method were published in Science.² In February 2015, the FISSEQ method was published in detail in Nature Protocols, extending access to this technology to researchers worldwide.³
We are continuing to develop FISSEQ by increasing the number of molecules sequenced per cell, optimizing sequencing library construction, and applying new microscopy and sequencing modalities. For example, we are combining FISSEQ with a cutting edge super-resolution microscopy technique called Expansion Microscopy (ExM), which will give unprecedented nanoscale resolution of RNA localization inside cells.⁴ ExM also enables incorporation of genomic and proteomic in situ visualization methods into RNA FISSEQ, providing a “panomic” view of biological specimens.
One application of FISSEQ currently in progress is to map the single-cell transcriptomics and neural connectivity of the brain. Measurement of the “connectome,” or connections of neurons in the brain, is traditionally done by electron microscopy, which is not scalable to the whole human brain, and does not provide any information about the biology of the neurons themselves. Our strategy, which we refer to as Rosetta Brain, is to use FISSEQ of molecularly barcoded brains to read out both the connections between neurons and the molecular phenotype of the individual neurons.⁵ We believe this strategy will provide unique insights into the how the organization of the human brain gives rise to cognition.
Another application of FISSEQ is to advance cancer biology. Tumors are internally heterogeneous in genotype due to somatic evolution, and are also influenced by the presence of nearby healthy cells, immune infiltration and inflammation, and cell-cell signaling by both nanoscale interactions between cells and the presence of secreted factors. Moreover, heterogeneity leads to profound differences response to treatment and outcomes.⁶ We are using FISSEQ of both in vitro culture systems and tumor sections to measure and understand these various forms of spatial organization.
¹ Nawy, T. “In situ sequencing.” (2014) Nature Methods 11: 29.
² Lee, J., et al. “Highly Multiplexed Subcellular RNA Sequencing in Situ.” (2014) Science DOI: 10.1126/science.1250212.
³ Lee, J., et al. “Fluorescent in situ sequencing (FISSEQ) of RNA for gene expression profiling in intact cells and tissues.” (2015) Nature Protocols DOI: 10.1038/nprot.2014.
⁴ Chen, F., Tillberg, P.W., and E.S. Boyden. (2015) Expansion Microscopy, Science DOI:10.1126/science.1260088.
⁵ Marblestone, A.H., et al. “Rosetta Brains: A Strategy for Molecularly-Annotated Connectomics.” (2014) Neurons and Cognition arXiv:1404.5103.
⁶ Junttila, M.R. and F.J. de Sauvage. “Influence of tumour micro-environment heterogeneity on therapeutic response.” (2013) Nature doi:10.1038/nature12626.