Quantitative comparison of DNA detection by GFP-lac repressor tagging, fluorescence in situ hybridization and immunostaining

Il-Han Kim, Jens Nagel, Otten Simone, Knerr Boris, Roland Eils, Karl Rohr, Steffen DietzelBMC 2007 · DOI

Abstract

Each of the tested methods revealed particular strengths and weaknesses, which should be considered when interpreting respective experimental results. Despite the required denaturation step, FISH signals in structurally preserved cells show a surprising similarity to signals generated before denaturation.

Introduction

In modern cell biology, understanding genomic organization relies heavily on fluorescence detection techniques to visualize specific DNA sequences or proteins. Three primary methods—GFP-tagging, immunostaining, and fluorescence in situ hybridization (FISH)—are commonly employed, yet each carries inherent limitations. GFP-tagging allows for live-cell imaging but may fail to detect non-fluorescent protein fractions. Immunostaining depends on antibody penetration, and FISH requires a harsh denaturation step that has long been suspected of distorting 3D chromatin architecture. This study provides a rigorous quantitative comparison to determine how consistently these techniques represent the same DNA structures within individual cells.

Experimental Methodology

The researchers utilized a specialized mouse cell line containing a 50-megabase transgene array. This large, repetitive DNA string was engineered for simultaneous detection via four distinct channels:

  1. Direct Fluorescence: GFP-labeled proteins bound to the DNA.
  2. Pre-FISH Immunostaining: Antibodies targeting the GFP before denaturation.
  3. FISH: Fluorescent DNA probes binding directly to the sequence after heat denaturation.
  4. Post-FISH Immunostaining: Re-staining the GFP following the FISH procedure to assess structural changes.

To ensure accuracy, the team captured 3D image stacks, applied deconvolution to remove optical blur, and used precise image registration to align color channels. The similarity between signals was quantified using Correlation Coefficients (CC).

Results and Discussion

The analysis revealed several critical insights into the performance and physical impact of these detection methods.

Structural Preservation and Volume Expansion

A significant finding was that FISH is remarkably "safe" regarding structural preservation. Despite the use of heat and formamide to melt DNA strands, the 3D footprint of the DNA locus remained highly similar to its pre-denaturation state. However, a "spreading" effect was observed; FISH signals were approximately 27% larger in volume than direct GFP signals. This expansion was even more pronounced in immunostaining measured after FISH, which showed a 73% increase in volume, suggesting that the FISH procedure physically opens up the chromatin structure.

Figure 4: Multi-method comparison of DNA detection signals. The grid displays direct GFP fluorescence, pre- and post-FISH immunostaining, and FISH signals for multiple nuclei, along with calculated correlation coefficients (CC) between channels.

The "Dark" GFP Fraction and RNA Interference

The study identified a "dark" fraction of GFP-fusion proteins. Antibodies frequently detected substructures that direct GFP fluorescence missed, implying that a portion of the GFP may be misfolded or quenched. Consequently, direct imaging might underestimate the true extent of a biological structure. Additionally, the researchers noted that omitting RNase during FISH leads to probes binding to nascent RNA transcripts, resulting in larger, fuzzier signals that could misrepresent DNA compaction levels.

Conclusion

While all three methods generally agree on the location and shape of DNA structures, researchers must account for their specific biases. FISH remains a reliable proxy for the in vivo state at light-microscopy scales, despite measurable volume expansion. Future studies using super-resolution microscopy may further refine these comparisons, particularly for smaller, endogenous DNA loci that might be more sensitive to experimental distortions than the mega-array used here.

Summary updated

Abstract

Each of the tested methods revealed particular strengths and weaknesses, which should be considered when interpreting respective experimental results. Despite the required denaturation step, FISH signals in structurally preserved cells show a surprising similarity to signals generated before denaturation.
BibTeX
@inproceedings{kim2007quantitative,
  author = {Il-Han Kim and Jens Nagel and Otten Simone and Knerr Boris and Roland Eils and Karl Rohr and Steffen Dietzel},
  title = {{Quantitative comparison of DNA detection by GFP-lac repressor tagging, fluorescence in situ hybridization and immunostaining}},
  booktitle = {BMC},
  year = {2007},
  doi = {10.1186/1472-6750-7-92},
}
Figure