Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust

Harry M. T. Choi, Maayan Schwarzkopf, Mark E. Fornace, Aneesh Acharya, Georgios Artavanis, Johannes Stegmaier, Alexandre Cunha, Niles A. PiercebioRxiv 2018 · DOI

TL;DR

This paper introduces the third generation of in situ hybridization chain reaction (HCR v3.0), a powerful technique for visualizing mRNA within cells and tissues. The key innovation is "automatic background suppression" using novel "split-initiator probes," which dramatically improves signal quality and simplifies the experimental workflow, making it more robust and accessible for researchers.

What Problem Does This Paper Address?

Imaging mRNA expression within biological samples is crucial for understanding gene function and cellular processes. Traditional methods like in situ hybridization (ISH) have faced significant challenges, particularly in complex samples like whole-mount embryos or thick tissues. These challenges include:

  • Low signal-to-background ratio: Autofluorescence from tissues can overwhelm the signal from the target mRNA.
  • Difficulty with multiplexing: Simultaneously detecting multiple different mRNAs is complex and time-consuming.
  • Need for probe optimization: Ensuring probes bind specifically to their intended targets often requires extensive, laborious testing and optimization, especially for new genes or organisms.
  • Quantitative limitations: Accurately measuring the amount of mRNA present has been difficult.

The previous generation of HCR (v2.0) offered significant improvements in multiplexing, sensitivity, and resolution but still suffered from potential background noise if probes bound non-specifically, necessitating probe optimization. This paper aims to overcome these limitations by developing a more robust and user-friendly HCR system.

Methods

The core innovation in HCR v3.0 is the development of split-initiator probes. In previous HCR versions (v2.0), each DNA probe designed to bind a specific mRNA carried the entire HCR initiator sequence. If such a probe bound non-specifically to cellular components, it could still trigger the HCR amplification process, leading to unwanted background signal.

HCR v3.0 addresses this by splitting the HCR initiator sequence into two halves. Each half is attached to a separate probe (P1 and P2). These two probes are designed to bind to adjacent sites on the target mRNA. Only when both P1 and P2 bind specifically to their respective sites on the target mRNA do the two halves of the initiator come into close proximity, allowing them to cooperatively initiate the HCR amplification cascade. If either P1 or P2 binds non-specifically, or if they bind to different locations, the initiator is not formed, and no amplification occurs. This "automatic background suppression" is the key to the improved robustness and ease of use.

The HCR amplification itself involves two types of DNA hairpins (H1 and H2) that are metastable. When the HCR initiator is present, these hairpins self-assemble into long, fluorescently labeled polymers that are tethered to the target mRNA. This isothermal, enzyme-free process amplifies the signal significantly.

The paper also details the application of HCR v3.0 in three main imaging modes:

  • Multiplexed in situ hybridization: Imaging multiple mRNAs simultaneously in whole-mount embryos and tissue sections.
  • qHCR imaging: Quantitative HCR imaging for analog mRNA relative quantitation with subcellular resolution within an anatomical context.
  • dHCR imaging: Digital HCR imaging for absolute quantitation of individual mRNA molecules, even in challenging, autofluorescent samples.
  • qHCR flow cytometry: High-throughput analog mRNA quantitation in single cells.

Datasets

The paper does not explicitly mention training or evaluation on specific, pre-existing public datasets in the traditional machine learning sense. Instead, the "datasets" are biological samples and the mRNA targets within them. The experiments were performed on:

  • Whole-mount vertebrate embryos (specifically chicken embryos at stage HH 10 and HH 11).
  • Cultured human cells (HEK cells).
  • Bacterial cells (E. coli).
  • Thick tissue sections (mentioned as a capability, with reference to previous work on adult mouse brain sections).

The paper introduces the concept of using large, unoptimized probe sets for new targets and organisms, which is a departure from previous methods requiring extensive optimization. The validation of HCR v3.0 relies on the performance metrics (signal-to-background, colocalization) obtained from these biological samples.

Key Insights and Ideas

  • Automatic Background Suppression is a Game Changer: The most significant insight is that splitting the HCR initiator across two probes fundamentally solves the problem of non-specific probe binding causing amplified background. This dramatically simplifies experimental design and increases reliability.
  • Unoptimized Probe Sets are Now Viable: Because of automatic background suppression, researchers can now use large sets of probes for multiple targets without the tedious optimization previously required. This democratizes complex multiplexed imaging.
  • HCR v3.0 Enables New Quantitative Possibilities: The enhanced signal-to-background and robustness of HCR v3.0 make analog (qHCR) and digital (dHCR) mRNA quantitation more accessible and accurate, both in fixed tissues and in live cells via flow cytometry.
  • Versatility Across Scales: The technique is shown to work from single-molecule resolution (dHCR) to whole-tissue imaging, and across different biological systems (vertebrates, bacteria).
  • A Unified Platform: The paper presents HCR v3.0 as a versatile platform that can be adapted for various quantitative imaging needs, from high-throughput screening to detailed anatomical studies.

Key Results

  • In vitro validation: Gel electrophoresis showed minimal leakage of HCR hairpins without initiator, and split-initiator probes dramatically reduced polymer formation when only one probe of a pair was present, or when the target was absent, compared to standard probes.
  • In situ validation in chicken embryos:
    • Increasing probe set size with standard probes (v2.0) led to a dramatic increase in background and a decrease in signal-to-background ratio.
  • Using split-initiator probes (v3.0), increasing probe set size resulted in no measurable change in background and a monotonic increase in signal-to-background ratio.
  • Confocal images showed significantly lower background with split-initiator probes, even with unoptimized probe sets.
  • Pixel intensity histograms demonstrated non-overlapping distributions for signal and background with v3.0, compared to overlapping distributions with v2.0.
  • Multiplexed imaging: Achieved 4-channel multiplexed imaging in whole-mount chicken embryos using large, unoptimized split-initiator probe sets, with signal-to-background ratios ranging from 27 to 59. This performance surpassed previously optimized standard probe sets.
  • qHCR imaging: Demonstrated high accuracy and precision for analog mRNA quantitation in anatomical context using 2-channel redundant detection with split-initiator probes.
  • dHCR imaging: Achieved high colocalization fractions (≈0.84) for digital mRNA absolute quantitation using large, unoptimized split-initiator probe sets, significantly improving upon previous results with standard probes.
  • qHCR flow cytometry: Showed high signal-to-background (101.8 ± 0.5 for HEK cells, 26 ± 5 for E. coli) and excellent accuracy/precision for analog mRNA quantitation in human and bacterial cells.

Limitations and Future Work

  • Limitations:
    • While split-initiator probes significantly reduce background, the paper acknowledges that non-specific binding of the bridge strand in Scheme C (a precursor to Scheme E) could still lead to background in later stages. Scheme E (HCR v3.0) avoids this by integrating amplification into a single stage.
  • The paper notes that for very low expression levels, qHCR signal might not be observable above autofluorescence, highlighting the complementary nature of qHCR and dHCR.
  • The paper implicitly suggests that while probe optimization is no longer necessary, it might still offer marginal improvements in specific, highly challenging cases, though the benefit of automatic suppression is so large that it's often not worth the effort.
  • Future Work:
    • The authors present HCR v3.0 as a robust and versatile platform, implying its broad applicability across various biological questions and organisms.
  • The development of quantitative read-out and read-in analyses using HCR v3.0 is highlighted as opening a "new era for in situ hybridization."
    • The paper suggests that the enhanced ease-of-use and performance of HCR v3.0 will facilitate broader adoption and application in diverse research areas.

Summary updated

Abstract

ABSTRACT In situ hybridization based on the mechanism of hybridization chain reaction (HCR) has addressed multi-decade challenges to imaging mRNA expression in diverse organisms, offering a unique combination of multiplexing, quantitation, sensitivity, resolution, and versatility. Here, with third-generation in situ HCR, we augment these capabilities using probes and amplifiers that combine to provide automatic background suppression throughout the protocol, ensuring that even if reagents bind non-specifically within the sample they will not generate amplified background. Automatic background suppression dramatically enhances performance and robustness, combining the benefits of higher signal-to-background with the convenience of using unoptimized probe sets for new targets and organisms. In situ HCR v3.0 enables multiplexed quantitative mRNA imaging with subcellular resolution in the anatomical context of whole-mount vertebrate embryos, multiplexed quantitative mRNA flow cytometry for high-throughput single-cell expression profiling, and multiplexed quantitative single-molecule mRNA imaging in thick autofluorescent samples. SUMMARY In situ hybridization chain reaction (HCR) v3.0 exploits automatic background suppression to enable multiplexed quantitative mRNA imaging and flow cytometry with dramatically enhanced ease-of-use and performance.
BibTeX
@misc{choi2018third,
  author = {Harry M. T. Choi and Maayan Schwarzkopf and Mark E. Fornace and Aneesh Acharya and Georgios Artavanis and Johannes Stegmaier and Alexandre Cunha and Niles A. Pierce},
  title = {{Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust}},
  year = {2018},
  doi = {10.1101/285213},
  howpublished = {bioRxiv},
}

Loading related content...

Figure