Investigating pathological extracellular matrix architecture

Abstract

[eng] The composition and architecture of the extracellular matrix (ECM), and their dynamic alterations, play an essential regulatory role in numerous cellular processes. Furthermore, structural, and biochemical properties of the ECM are central in regulating cell response via mechanical, chemical, and topological cues detected by receptors on the cell membrane, which induce cytoskeleton and/or cell nucleus rearrangements affecting gene expression. Indeed, distinct ECM architectures in the native stroma depend on tissue type, function, and composition. For instance, ECM anisotropy and stiffness are associated with altered ECM degradation and remodeling in cancer. In turn, abnormal ECM architecture favors tumor progression and invasion. Moreover, numerous diseases are associated with mutations in genes encoding ECM components, leading to deficient mechanical properties and altered ECM structure. Thus, there is an increasing interest to exploit and consolidate this knowledge to improve patients’ diagnosis, treatment, and care.

In this work, we examined the use of cutting-edge open-source software to characterize the properties of ECM fibers and fibril bundles in patients’ samples and in vitro model tissues. We focused on Collagen I-III in lung cancer and on Collagen VI-related muscular dystrophies (COL6-RMD) because these pathologies present a clear link between altered ECM architecture and patients’ outcomes, which requires further investigation and may be exploited at the clinical level. For this purpose, we initially generated phantom images of fibrils networks (created artificially in silico or obtained from bioengineered models in vitro mimicking isotropic/anisotropic ECM) to test and validate the analysis of ECM fibrils with different bioengineering tools. Then, we focused on ECM images obtained from patient cell cultures or tissue biopsies and evaluated ECM descriptors as novel potential biomarkers.

We collaborated with the Institut de Recerca Sant Joan De Déu (IRSJD) to analyze ECM models obtained from fibroblasts of patients affected by COL6-RMD. This is a rare subset of neuromuscular diseases related to deficiency in collagen type VI (COL6) expression, which affects young patients causing a broad range of severe disabilities, which worsen with time and shorten their lifespan. Currently, analysis of in vitro cultures of patients’ fibroblasts is employed as the first screening to select cases of COL6-RMD, and image-based classification has been suggested to improve this assay. Employing similar protocols, fibroblasts of different sources and malignant origin have been used to obtain decellularized cell-derived matrices (CDM), which retain topography and biochemical features, preserving their native microenvironment. Therefore, we produced and analyzed patients’ CDM to provide new in vitro personalized ECM models and improve early recognition of COL6-RMD subtypes with novel fiber-based ECM descriptors.

In parallel, in collaboration with the CIBERES (Centro de Investigación Biomédica en Red de Enfermedades Respiratorias), we characterized the fibrillar properties of Non-Small Cell Lung Cancer (NSCLC) patients’ tissues. NSCLC is among the tumors with the highest death rate and recurrence; therefore, novel biomarkers may provide helpful complementary data with clinical impact. Fiber descriptors extracted by CT-FIRE have been demonstrated to capture relevant prognostic changes in the collagen structure of several tumors. In NSCLC, collagen architecture influences immune system activation, and increased fibrosis is already reported to correlate with patients’ prognosis. Therefore, we studied alterations of collagen fiber networks in the two most common NSCLC subtypes, Adenocarcinoma and Squamous Cell Carcinoma, to provide novel collagen structure-based biomarkers.

Ultimately this work, along with the tools presented, paves the way for addressing the need for novel ECM-based descriptors, to stratify patients and evaluate their response to experimental treatments.

[spa] La composición y arquitectura de la matriz extracelular (MEC), y sus alteraciones dinámicas, juegan un papel regulador importante en numerosos procesos patológicos. Por ejemplo, la anisotropía y la rigidez de la MEC están asociadas con la remodelación alterada de la MEC en el cáncer. A su vez, esta arquitectura favorece la progresión e invasión tumoral. Además, numerosas enfermedades están asociadas con mutaciones en los genes que codifican los componentes de la MEC, lo que conduce a propiedades mecánicas deficientes y una estructura alterada de la MEC. Por lo tanto, existe un interés creciente por explotar y consolidar este conocimiento para mejorar la atención de los pacientes. En este trabajo, nos centramos en las distrofias musculares relacionadas con el colágeno VI (COL6-RMD) y el cáncer de pulmón de células no pequeñas (NSCLC), para investigar biomarcadores relacionados con la MEC in vitro y en los tejidos de los pacientes que se implementarán en el entorno clínico. Empleamos la segmentación automática de imágenes para cuantificar las características fibrilares e investigamos la asociación con la información clinicopatológica de los pacientes. Por lo tanto, producimos y analizamos matrices derivadas de células de pacientes para proporcionar nuevos modelos de ECM personalizados in vitro y nuevos descriptores de ECM basados en fibra de los subtipos COL6-RMD.

Se ha demostrado que los descriptores de fibra basados en imágenes capturan cambios pronósticos relevantes en la estructura del colágeno de varios tumores. En el NSCLC, la arquitectura del colágeno influye en la activación del sistema inmunitario y ya se reconoce que el aumento de la fibrosis se correlaciona con el pronóstico de los pacientes. Por lo tanto, estudiamos las alteraciones de las redes de fibras de colágeno en los dos subtipos de NSCLC más comunes, el adenocarcinoma y el carcinoma de células escamosas, y demostramos que los descriptores de la estructura del colágeno son posibles biomarcadores basados en imágenes con aplicaciones de diagnóstico y pronóstico. En última instancia, este análisis, junto con las herramientas presentadas, es prometedor para abordar la necesidad de nuevos descriptores, para estratificar a los pacientes y evaluar su respuesta a los tratamientos experimentales.