The field of Single Cell biology 1 has revolutionized the way we understand biological processes and a new era is started. In a few words it can be explained as the detection of the same traits in every single cell inasample,notjustasthemeanvalueofthebulk.Thisrevolutionhasaddeda new level of resolution to what we can “see” and makes us better understand the complexity of a sample made of different entities: cells. Characterization of single cells has increased our understanding of cell phenotypes, the dynamics and trajectory of their development, and their network. All of this has been possible thanks to the fast development of a large group of technologies: “single-cell multi omic technologies”. Among all of these novelties high-plex spatial proteomic represents a small but crucial niche for many reasons: it has a single cell resolution associated with spatial localization, it evaluates the presence of protein and not RNA signal (bypassing post transcriptional modification), it analyses whole cells without losing any types of them as often happens in tissue disaggregation or other single cell technologies, it allows retrospective study on Formalin Fixed Paraffin Embedded (FFPE) sample collection, easily integrable with other -omic technologies. In 2017 we developed a method called Multiple Iterative Labeling by Antibody Neodeposition (MILAN technology) “Which implies multiple stainings of a tissue section with multiple antibodies..” 2 Within this scenario my PhD thesis develops. Briefly it concerns the analysis, with MILAN technology, of human or mouse, healthy or pathologic tissues, at single cell level in order to classify cell populations, quantify and localize them in situ and recognize their specific interaction and characteristics. It’s a selection of results of these 3 last years of research after the publication of the MILAN technology method. The first part is related to the application of MILAN methods to different types of tissue. This has allowed us to characterize immune infiltrate presence in a rare dataset of Uterine leiomyosarcomas and to identify cell population in a mouse model of Idiopathic Pulmonary Fibrosis (IPF) by measuring percentage of population and their expression as a drug transporter. These papers represent the application of methods at different levels and its use to answer specific scientific questions. Subsequently an overview of the available technologies for multiplex immunohistochemistry is presented with “their advantages and challenges, the comparison with MILAN, and provide the basic principles on how to interpret high-dimensional data in a spatial context.” Then, because of the availability of a large lymphoid tissue image data set, for which there were no present specific algorithms, we created the” Bayesian Reduction for Amplified Quantization in Umap Embedding” (BRAQUE) algorithm, a tailored pipeline of analysis for data coming from highly multiplexed imaging data sets. The last part of the project is dedicated to the classification at single cell level of normal human lymph nodes with the previously created bioinformatic tool BRAQUE. This has been possible thanks to a huge effort of sample collection and analysis of almost 100 samples stained with more than 80 antibodies. Even if the project is still ongoing, we have already obtained some interesting results presented in chapter 5-6 (manuscripts in preparations). The database of lymph nodes cell populations will be subsequently integrated with scRNAseq data. 1. Schier, A. F. Single-cell biology: beyond the sum of its parts. Nat. Methods 17, 17–20 (2020). 2. Bolognesi, M. M. et al. Multiplex Staining by Sequential Immunostaining and Antibody Removal on Routine Tissue Sections. J. Histochem. Cytochem. 65, 431–444 (2017).
Con la “Single Cell Biology”1 una nuova era è iniziata e con essa il nostro modo di di capire i processi biologici è stato rivoluzionato. In breve si tratta di misurare la stessa caratteristica in ogni singola cellula di un campione, non semplicemente estrarre il dato come una media fra tutte. Questa rivoluzione aggiunge un nuovo livello di risoluzione a quello che possiamo “vedere” e ci permette di comprendere meglio la complessità di un campione composto da diverse entità: le cellule. Caratterizzare singolarmente le cellule ci permette di approfondire la nostra conoscenza rispetto a fenotipo, sviluppo e network. Tutto questo è possibile grazie allo sviluppo di nuove tecnologie: le “single cell-omic technologies". Tra tutte queste novità la proteomica spaziale rappresenta un piccolo ma cruciale settore per diverse ragioni: la risoluzione a singola cellula è associata alla localizzazione spaziale, si misura la presenza di proteine e non RNA (quindi segnali dopo le modifiche trascrizionali), vengono analizzate tutte le cellule senza perderne determinati tipi come spesso succede nella disgregazione dei tessuti o in altre tecnologie a singola cellula, infine consente studi retrospettivi utilizzando campioni Fissati in Formalina e Inclusi in Paraffina (FFPE) ed è facilmente integrabile con altre tecnologie “-omiche”. Nel 2017 abbiamo sviluppato e pubblicato un lavoro metodologico chiamato Multiple Iterative Labeling by Antibody Neodeposition (MILAN technology) “che consiste nella colorazione multipla dello stesso campione di tessuto con diversi anticorpi..” 2 La tesi di dottorato si sviluppa in questo scenario. In pratica concerne l’analisi, con la tecnologia MILAN, di tessuti umani e murini, sani e patologici, a risoluzione cellulare, per classificare, quantificare e localizzare in situ le diverse popolazioni cellulari e per identificarne le specifiche interazioni e caratteristiche. Raccoglie gli articoli significativi di questi 3 anni di ricerca che hanno seguito la pubblicazione del metodo MILAN. La prima parte contiene l'applicazione di MILAN su due differenti tipi di tessuti. In una casistica di Leiomiosarcomi Uterini è stato caratterizzato l’infiltrato infiammatorio; in un modello murino di Fibrosi Polmonare Idiopatica (IPF) sono state identificate le diverse popolazioni cellulari nel polmone, misurata la loro presenza in percentuale e la loro espressione di un trasportatore transmembrana. Questi due articoli rappresentano l'applicazione del metodo a due livelli differenti e il suo utilizzo per rispondere a specifiche domande scientifiche. Segue un articolo che contiene una panoramica sulle tecnologie disponibili per immunoistochimica/immunofluorescenza con approccio multiplex, vantaggi e svantaggi, il confronto con la tecnologia MILAN. Fornisce inoltre i principi base per interpretare dati high-dimensional nel loro contesto spaziale. Poi, per analizzare un grosso data set di tessuti linfoidi, mancando uno specifico algoritmo di analisi, ne abbiamo creato uno nuovo: BRAQUE ovvero Bayesian Reduction for Amplified Quantization in Umap Embedding. Questo algoritmo è specificamente pensato per analizzare dati prodotti da immagini in immunofluorescenza multiplex. Infine l’ultima parte del progetto è dedicata alla classificazione a singola cellula del tessuto linfoide umano, su una casistica di circa 100 linfonodi, colorati e analizzati con più di 80 marcatori. Il progetto è ancora in corso, tuttavia negli ultimi due capitoli sono presentati risultati preliminari. I dati delle popolazioni identificate saranno integrati con dati provenienti da single-cell RNA sequencing. 1. Schier, A. F. Single-cell biology: beyond the sum of its parts. Nat. Methods 17, 17–20 (2020). 2. Bolognesi, M. M. et al. Multiplex Staining by Sequential Immunostaining and Antibody Removal on Routine Tissue Sections. J. Histochem. Cytochem. 65, 431–444 (2017).
(2023). Tissue-based high-dimensional landscaping in inflammation, repair and transformation. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2023).
Tissue-based high-dimensional landscaping in inflammation, repair and transformation
BOLOGNESI, MADDALENA MARIA
2023
Abstract
The field of Single Cell biology 1 has revolutionized the way we understand biological processes and a new era is started. In a few words it can be explained as the detection of the same traits in every single cell inasample,notjustasthemeanvalueofthebulk.Thisrevolutionhasaddeda new level of resolution to what we can “see” and makes us better understand the complexity of a sample made of different entities: cells. Characterization of single cells has increased our understanding of cell phenotypes, the dynamics and trajectory of their development, and their network. All of this has been possible thanks to the fast development of a large group of technologies: “single-cell multi omic technologies”. Among all of these novelties high-plex spatial proteomic represents a small but crucial niche for many reasons: it has a single cell resolution associated with spatial localization, it evaluates the presence of protein and not RNA signal (bypassing post transcriptional modification), it analyses whole cells without losing any types of them as often happens in tissue disaggregation or other single cell technologies, it allows retrospective study on Formalin Fixed Paraffin Embedded (FFPE) sample collection, easily integrable with other -omic technologies. In 2017 we developed a method called Multiple Iterative Labeling by Antibody Neodeposition (MILAN technology) “Which implies multiple stainings of a tissue section with multiple antibodies..” 2 Within this scenario my PhD thesis develops. Briefly it concerns the analysis, with MILAN technology, of human or mouse, healthy or pathologic tissues, at single cell level in order to classify cell populations, quantify and localize them in situ and recognize their specific interaction and characteristics. It’s a selection of results of these 3 last years of research after the publication of the MILAN technology method. The first part is related to the application of MILAN methods to different types of tissue. This has allowed us to characterize immune infiltrate presence in a rare dataset of Uterine leiomyosarcomas and to identify cell population in a mouse model of Idiopathic Pulmonary Fibrosis (IPF) by measuring percentage of population and their expression as a drug transporter. These papers represent the application of methods at different levels and its use to answer specific scientific questions. Subsequently an overview of the available technologies for multiplex immunohistochemistry is presented with “their advantages and challenges, the comparison with MILAN, and provide the basic principles on how to interpret high-dimensional data in a spatial context.” Then, because of the availability of a large lymphoid tissue image data set, for which there were no present specific algorithms, we created the” Bayesian Reduction for Amplified Quantization in Umap Embedding” (BRAQUE) algorithm, a tailored pipeline of analysis for data coming from highly multiplexed imaging data sets. The last part of the project is dedicated to the classification at single cell level of normal human lymph nodes with the previously created bioinformatic tool BRAQUE. This has been possible thanks to a huge effort of sample collection and analysis of almost 100 samples stained with more than 80 antibodies. Even if the project is still ongoing, we have already obtained some interesting results presented in chapter 5-6 (manuscripts in preparations). The database of lymph nodes cell populations will be subsequently integrated with scRNAseq data. 1. Schier, A. F. Single-cell biology: beyond the sum of its parts. Nat. Methods 17, 17–20 (2020). 2. Bolognesi, M. M. et al. Multiplex Staining by Sequential Immunostaining and Antibody Removal on Routine Tissue Sections. J. Histochem. Cytochem. 65, 431–444 (2017).File | Dimensione | Formato | |
---|---|---|---|
phd_unimib_069860.pdf
Accesso Aperto
Descrizione: Tissue-based high-dimensional landscaping in inflammation, repair and transformation
Tipologia di allegato:
Doctoral thesis
Dimensione
4.48 MB
Formato
Adobe PDF
|
4.48 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.