Capturing true single-cell resolution with your spatial data

Since its inception, spatial biology has proven a revolutionary frontier in life science research. However, there is still plenty of room for growth in the systems we use to conduct spatial biology and limitations can be encountered with different systems. So, the question is, how can we capture true single-cell spatial data and could a new class of spatial technology help us get there?

We spoke with Linda Orzolek (left) to find out. Linda is a molecular biologist and the Vice President of Operations and Corporate Strategy at OMAPiX, Inc. (MD, USA), a concierge clinical research organization (CRO) that strives to educate its clientele and guide them in their research goals. Before joining OMAPiX, Inc. in 2024, Linda worked for the US Department of Defense (Washington DC, USA) and the Johns Hopkins Transcriptomics Core (MD, USA). Throughout her career, Linda has consistently prioritized education. At  Johns Hopkins University (JHU; MD, USA), she led the development and instruction of a graduate-level course on the practical applications of single-cell sequencing, as well as establishing an annual single-cell symposium. She transitioned over to industry to build OMAPiX, Inc.’s service lab, bringing her knowledge and experience from the Transcriptomics Core to the larger research community.

In this interview, we dive into how a new class of spatial technology is overcoming the limitations of other spatial methods and highlight Linda’s top tips for collecting and analyzing spatial data.

How are spatial biology technologies advancing life science research?

The introduction of spatial biology around 2021 opened a variety of opportunities in life science research. We have to acknowledge that bulk sequencing data are inherently incorrect. We are using average transcript levels to make deductions about how specific cell types are responding to stimuli. Single-cell analysis helped us take that crucial step forward in recognizing that different cell types behave differently on a transcriptional level. However, without the spatial context of those cells, the relevant conclusions are lost. A single-cell experiment may show a high population of immune cells within a tumor sample, but are those immune cells penetrating the tumor? If not, then utilizing immune response as a treatment method is not productive.

Whether we are trying to distinguish tumor variations within a sample or identify tertiary lymphoid structure (TLS) development, spatial context and distribution are crucial to downstream diagnostics and treatment. Traditional chemotherapy comes with debilitating side effects because the treatment is often systemic and indiscriminate. If we know not just the cell types that are impacted, but also understand the transcriptional changes in the precise location of disease, biomarker targets become more specific, which leads to more effective and efficient treatment without adverse reactions.

What limitations exist when using imaging- or sequencing-based methods to collect spatial data?

The most complex part of spatial data is cell boundary identification: where does one cell end and another begin? When transcripts are close to the cell membrane or the cell has an abnormal shape, transcripts can be incorrectly segmented into a neighboring cell. When this happens too often, cell type (or more specifically subtype) identification becomes incorrect and can confound downstream conclusions. With in situ methods, this becomes particularly important, since the system is directly imaging the transcript location. Coupled with non-specific signal, cross-hybridization, drop outs, overcrowding and other known limitations of particular platforms, the likelihood of miscalling a cell can be relatively high.

With sequencing-based methods, the assays utilize a grid-based labeling of cells, which of course will never match the complex geometry of our tissues. Smaller spots mean fewer cells in one place, but the data will still capture transcripts from multiple cells in one spatial target. Conversely, when a single-cell requires multiple spots, how can you identify exactly how many spots belong to only that cell? In addition, there is always the problem of migration, as sequencing based applications rely on release and capture of the probe and transcript. Slight movement of a probe or transcript can result in mis-identification of location.

Please tell us what made you interested in Trekker^® technology.

Since the inception of high-throughput spatial assays, the underlying concern about appropriately identifying a cell’s location has been a source of debate. There are many different bioinformatic approaches that try to resolve these issues, but these algorithms are only as consistent as the data.  We make assumptions about what the transcriptional profile should be and apply that to data interpretation. While iterative analysis can improve upon these outcomes, there is a bias that is applied. Trekker presented the opportunity to let the data speak to us. It takes the known limitations of other spatial applications and utilizes them for improved outcome. It is the only application with true single-cell resolution. Assays that offer 2 nm or 10 nm spots can’t be true resolution, even if they are smaller than a cell. For starters, not all cells are the same size, and none of them are going to match the perfect shape of the spatially barcoded spot on the assay slide. It’s an important semantic distinction.

How does Trekker technology overcome the limitations of other spatial methods; what makes it unique?

Trekker alleviates many of the problems of traditional spatial applications because it combines the gene-capture efficiency of single-cell methods with spatial identification. Segregating the individual nuclei means confident assignment of transcripts to a particular cell, a major concern with other spatial approaches. It also removes the potential reduced transcript capture often seen in spatial due to overcrowding, as spatial assays require hybridization at a precise location, rather than in solution. Coupling the efficiency of single-cell assays with a spatial identifier makes Trekker a unique approach.

Do you have any tips for using Trekker technology or for collecting/analyzing spatial data more broadly?

Every assay we use demands the same warnings and tips. Sample quality and preparation are the most important parts of the workflow, so:

• • Have a clear understanding of the integrity of your sample.
  • FFPE tissue does not have to be completely degraded.
  • Whenever possible, optimize sample storage and fixation.
  • Test nuclei isolation methods to find the one that is most effective for your specific sample type.
  • Overlysis means direct data loss. If particular cell types lyse more easily, there is an increased risk of loss of those nuclei in an overlysed sample, leading to bias in the data. More is better, but there is a balance.
  • More nuclei captured yields a more complete spatial picture of tissue. However, tissue that is too thick means inconsistent uptake of spatial probes, which will impact the z-axis alignment.

More about the interviewee

Linda earned her bachelors from LaSalle University (PA, USA) and her Masters in Biotechnology from Johns Hopkins University (MD, USA). She started her career in molecular biology research as an ORISE intern for the US Department of Defense, where she utilized the, then newly released, Affymetrix microarrays to understand transcriptomic changes in response to chemical weapons exposure. She parlayed that into a role at the Johns Hopkins Transcriptomics Core where she spent the next 20 years, starting as a technician running Affymetrix arrays. The Core moved into next-generation and third-generation sequencing as she moved into a lab manager position. In 2020, she took over as director of the Core and helped introduce single-cell and spatial transcriptomics to the JHU campus. She joined OMAPiX, Inc. in 2024.

The interview has no competing interests. Financial interests only in OMAPiX, Inc.

The opinions expressed in this interview are those of the interviewee and do not necessarily reflect the views of BioTechniques or Taylor & Francis Group.

This content was supported by Takara Bio.