Optimized Multiplex Probe Design Enables MERFISH-Based Spatial Transcriptomics Across Multiple Species

Abstract

Deciphering where genes are expressed within the complex architecture of cells and tissues requires advanced spatial transcriptomics techniques. This study focuses on Multiplexed Error-Robust Fluorescence In Situ Hybridization (MERFISH), a technique for spatial transcriptomic. The MERFISH experiment involves probes binding to specific RNA-targeted sequences within fixed cells and tissues. I developed a pipeline to design MERFISH probes applicable to multiple species using two probe design tools: PaintSHOP and the Oligo Designer Toolsuite (ODT). Furthermore, metagenomic data from healthy lung samples were analyzed to classify and estimate the abundance of microbial species, resulting in a comprehensive microbial list for probe design. This enabled the generation of MERFISH probes that can detect bacteria, which makes the pipeline especially valuable for lung microbiome studies. Cross-species homology screening was performed to ensure probe specificity across microbial, human, and mouse genomes. Additionally, all designed target, readout, and primer sequences were evaluated for secondary structure using minimum free energy (MFE) calculations. An initial codebook was generated, followed by further structural and thermodynamic analyses and hybridization efficiency screening to refine the final codebebook. This multi-species MERFISH probe design workflow supports high-specificity spatial transcriptomic studies and opens new avenues for investigating host–microbe–xenograft interactions.

Popular Abstract

Unlocking the Secrets of Cells with Spatial Transcriptomics

Have you ever wondered how genes work inside cells and tissues—not just what they do, but where they do it? Scientists now have powerful tools to map gene activity in precise locations, thanks to a method called spatial transcriptomics.

One cutting-edge technique for spatial transcriptomics is MERFISH (Multiplexed Error-Robust Fluorescence In Situ Hybridization), which acts like a molecular barcode scanner for RNA. Unlike conventional methods that mash up cells to extract RNA, MERFISH lets researchers see exactly where thousands of different RNA molecules are inside tissues—like a high-resolution map of gene activity. To measure active genes is generally error-prone, and what makes MERFISH special is its built-in ability to detect and correct errors, allowing researchers to confidently identify RNA molecules even in complex samples.

However, here is the twist: most studies focus on just one species at a time. What if we could track RNA from multiple species interacting. This is especially useful in research studies of infections, bacterial biofilm or the tumour microenvironment, such as when bacteria invade a mouse that has been implanted with human tumor tissue (a common model for studying cancer and immune responses). That’s exactly what this study set out to do!

How Does MERFISH Work?
To visualize and measure gene expression in its native cellular context, spatial transcriptomics methods like MERFISH rely on specially designed short DNA sequences (called probes) that can stick to the RNAs I want to study. I developed a seven-step pipeline to design the probes used in MERFISH. I then used this pipeline to analyse bacterial DNA data collected from human samples to identify which microbes were present and how abundant they were. For that I used two probe design tools—PaintSHOP and Oligo Designer Toolsuite. In more detail, these probes include three components: target probes (which bind the RNA), readout probes (which help with fluorescent labeling).

An important design criterion is ensuring that the DNA probes bind only to their designated RNA targets. To achieve this, I used BLAST+ to design ultra-specific probes. Think of BLAST+ to quickly find similar things in a big list. You give it a sequence (like a recipe) and it looks through a database of sequences (like a cookbook) to find similar recipes.

I then checked for "misfolded" probes (those that might tangle up before finding their target) and calculated how well they would stick to the matching RNA. Probes that passed all these checks were included in a barcode system (called a codebook) that helps MERFISH recognize each RNA uniquely.

As part of the project, I explored to design probes that target whole bacterial families, like Prevotella, instead of individual species. I discovered that the conventional tool mostly focused on conserved RNA regions like tRNA and rRNA, and could serve as general markers in future experiments. Furthermore to dig deeper, I analyzed which genes were most common across different microbes. This helped me find gene groups (called operons) that often work together in each type of bacteria. These gene groups could be great targets for making better probes to study each bacteria species more accurately in future experiments.

Why Does This Matter?
By developing spatial transcriptomic tools that works across multiple species, scientists can now explore how bacteria interact with human cells or “behave” or “synchronize” inside lab animals. This is a big step toward understanding how microbes influence cancer, infections, and how different types of cells "communicate" within the same tissue. For example, by analyzing the genetic material of bacteria in the human body, researchers can better understand how these microbes interact with our systems—and what roles they may play in both health and disease.

In short: We are not just reading cells’ genetic messages anymore—we are mapping where those conversations happen, in stunning detail.

(Want to learn more? Think of MERFISH like a GPS for genes—pinpointing RNA locations so we can finally see the full picture of life at the cellular level!)

Master’s Degree Project in Bioinformatics 45 credits 2025
Department of Biology, Lund University

Advisor: Nils Norlin
Advisors Unit/Department (or Company or Authority)
Medicinska fakulteten
Institutionen för experimentell medicinsk vetenskap

• author: Tan Grahn, Hooi Min
• supervisor: Nils Norlin
• organization: Degree Projects in Bioinformatics
• course: BINP51 20242
• year: 2025
• type: H2 - Master's Degree (Two Years)
• subject: Biology and Life Sciences
• language: English
• id: 9215014
• date added to LUP: 2025-11-06 14:42:07
• date last changed: 2025-11-06 14:42:07