Publikationsdatenbank

Improvement of the similarity spectral unmixing approach for multiplexed two-photon imaging by linear dimension reduction of the mixing matrix (2021)

Art

Zeitschriftenartikel / wissenschaftlicher Beitrag

Autoren

Rakhymzhan, Asylkhan

Acs, Andreas

Hauser, Anja E.

Winkler, Thomas H.

Niesner, Raluca A. (WE 2)

Quelle

International journal of molecular sciences

Bandzählung: 22

Heftzählung: 11

Seiten: Artikel 6046

ISSN: 1422-0067

Sprache

Englisch

Verweise

URL (Volltext): https://www.mdpi.com/1422-0067/22/11/6046

DOI: 10.3390/ijms22116046

Pubmed: 34205072

Kontakt

Institut für Veterinär-Physiologie

Oertzenweg 19 b
14163 Berlin
+49 30 838 62600
physiologie@vetmed.fu-berlin.de

Abstract / Zusammenfassung

Two-photon microscopy enables monitoring cellular dynamics and communication in complex systems, within a genuine environment, such as living tissues and, even, living organisms. Particularly, its application to understand cellular interactions in the immune system has brought unique insights into pathophysiologic processes in vivo. Simultaneous multiplexed imaging is required to understand the dynamic orchestration of the multiple cellular and non-cellular tissue compartments defining immune responses. Here, we present an improvement of our previously developed method, which allowed us to achieve multiplexed dynamic intravital two-photon imaging, by using a synergistic strategy. This strategy combines a spectrally broad range of fluorophore emissions, a wave-mixing concept for simultaneous excitation of all targeted fluorophores, and an unmixing algorithm based on the calculation of spectral similarities with previously measured fluorophore fingerprints. The improvement of the similarity spectral unmixing algorithm here described is based on dimensionality reduction of the mixing matrix. We demonstrate its superior performance in the correct pixel-based assignment of probes to tissue compartments labeled by single fluorophores with similar spectral fingerprints, as compared to the full-dimensional similarity spectral unmixing approach.