As senior science writer, Ruairi pens and edits scientific news, articles and features, with a focus on the complexities and curiosities of the brain and emerging informatics technologies. Ruairi also drives Technology Networks' search engine optimization (SEO) and editorial AI strategy and created the site’s podcast, Opinionated Science, in 2020. Ruairi has a Master’s degree in Clinical Neurosciences from the University of Cambridge.
Spatial transcriptomics is a term describing techniques that add information about location to gene expression data. Spatial data gives valuable context to the flood of genomic information that has emerged since the human genome was sequenced.
In this infographic, we explore the milestones that led to the spatial transcriptomics techniques of today, beginning decades before the genomic revolution.
Download this infographic to discover:
Four key techniques powering spatial transcriptomics
How key molecular approaches such as in situ hybridization were developed
50 years of genomic advances that have led to today’s techniques
Spatial transcriptomics is a term describing techniques that add information about
location to gene expression data. Spatial data gives valuable context to the flood of
genomic information that has emerged since the human genome was sequenced.
In this infographic, we explore the milestones that led to the spatial transcriptomics
techniques of today, beginning decades before the genomic revolution.
In situ hybridization
In situ hybridization, where nucleic acids are
tagged to reveal their location, is perhaps spatial
transcriptomics’ most important precursor. The
first studies to tag both DNA and RNA were both
published in 1969, when immature egg cells of
the African clawed frog (Xenopus laevis) were
tagged with a radioisotope.
ISH based on radioisotopes remains one of the
most sensitive methods of detecting nucleic
acids, but exposures can take weeks, rendering
them impractical. In the early 1980s, researchers
first developed the technology to directly
label nucleic acids with fluorescent reporter
fluorochromes. This became the basis of a
technique that would endure for decades.
At the end of the 1980s, the next step in ISH
was reported – whole-mount ISH (WMISH). This
technique can detect mRNA in intact tissue
fragments or embryos. This allowed researchers
to view the cellular distribution of gene
expression across the embryo. The mid-1990s
brought higher throughput WMISH screens that
looked at the expression of multiple genes, and
gene atlases that mapped expression data using
automated techniques were developed from the
late 1990s onwards.
One way of visualizing transcripts quantitatively
is to record signals from FISH assays. The first
study to show that this could be used to count
individual molecules of mRNA was published
in 1998. This initial study used the approach of
incubating transcripts with probes containing
multiple fluorophores, which produced a
muddied signal. Later efforts used multiple
probes to address this issue.
In situ sequencing
In situ sequencing methods read nucleic acid
information without removing it from the tissue.
A landmark ISS paper was published in 2013. A
ligase enzyme is used to link a known-sequence
primer to a probe. Non-matching probes are
washed away, and the linked DNA
that remains is amplified by a technique called
rolling-circle amplification and then sequenced.
One of the first commercial applications of
spatial transcriptomics originates in 2016.
While seqFISH is a single-cell method, this
paper technique works on the level of whole
tissue. Samples are directly bound to reverse
transcriptase primers attached to barcodes.
These codes assign each primer a unique spatial
position. RNA sequencing is then used to reveal
the transcripts involved and is paired with the
To enable RNA analysis, the tissue of interest
must be isolated. LCM is the most common
method used for tissue microdissection. Two
types of LCM are used today - ultraviolet (UV) and
infrared (IR). In both methods, lasers are used to
precisely cut tissue. The first use of a UV laser in
LCM was recorded in 1976. IR LCM followed in
1996. Some modern systems incorporate both
forms of LCM, with UV used to slice tissue, and IR
used to fuse the sample to a thermoplastic film,
which allows for easy removal.
Gene Enhancer Traps
In the 1980s, researchers did not have access
to reference genomes that record loci for model
organisms. Even then, efforts to link DNA spatial
location to expression patterns were underway.
One such technology is the enhancer trap
screen. This involves inserting a reporter gene
into a genome, paired with a promoter that drives
expression. Reporter gene expression is taken as
a proxy for expression of genes of interest near the
insertion point. The first enhancer gene trap for
the model organism Drosophila melanogaster was
conducted in 1987.
Single-cell cDNA amplification
Single-cell analysis produces minuscule
sections of RNA, which must be amplified to
be detected. This amplification is achieved
by either polymerase chain reaction (PCR) or
linear amplification of complementary cDNA.
The first study to attempt this amplification was
published in 1989.
Prior to 2008, microarrays had been the tool
of choice for transcriptomic analysis. This
technique requires species-specific probes,
produces data that isn’t easily quantifiable and
has issues with incomplete hybridization. That
all changed with the first publication using
RNA-seq in 2008. This technique piggybacks
on other advances in gene sequencing
and involves analyzing the cDNA sequence
complementary to the RNA of interest.
This technology, which made quantifiable
transcriptomics at scale possible, would alter
the course of the field.
With massive transcriptomic analysis techniques
now available, advances started coming at pace.
In 2014, seqFISH was published. This approach
enables RNAs (as well as DNAs and proteins) to
be visualized in single cells with their location
preserved. SeqFISH uses rounds of hybridization
involving thousands of molecules. In each
round, new provides are added to the molecules,
recorded and stripped off, allowing a picture of
complementary sequences to be built up over
time. seqFISH does not require any preselection
of genes and has no amplification step, which can
bias techniques like PCR.
Method of the Year
Spatial transcriptomics has continued to
advance. Efforts to marry the precision of
single-cell RNA sequencing with spatial
transcriptomics data are underway. Spatial
transcriptomics was given Nature Methods ’s
coveted Method of the Year award for
2020, cementing its position as a powerful new
tool in molecular biologists’ arsenals.
There are four different ways that spatial information about mRNA transcripts can be recorded. These are:
Imaging Methods Sequencing Methods
In situ hybridization
In situ sequencing
1. Hybridize labelled probes to
1. Rolling circle amplification of
2. Short, labelled probes
are hybridized to determine
1-2nt of transcripts’
1. Hybridize probes
to target mRNAs
1. Start with a spatially
barcoded probe array
2. Barcode locations imaged via
in situ sequencing
3. Image fluorophore locations
3. Image fluorophore of each
2. Image region of interest (ROI)
sequencing of captured
3. Overlay sample on array.
Ligate mRNA to probes
4. Repeat n times
to generate a genespecific
4. Repeat as
3. Free probes from
ROI with UV. Capture
sequencing of captured
Transcript imaging directly from intact tissue via hybridization
to labeled probes
Imaging of short probes, which read along an amplified transcript
Labeling using spatially barcoded probe arrays followed by NGS
Microdissection of cells or tissues followed by NGS
of their transcriptomes
Here’s a timeline that follows the development of these major spatial technologies.