. ABSTRACT Engineered vascularized human early metastatic niche for the study of breast cancer cells extravasation Introduction Cancer metastasis is the most lethal aspect of human cancer and represents a great challenge for worldwide research, being responsible of more than 90% of cancer-associated mortality [1]. Even if great advances in treatments related to primary tumor have been reached, due to new discoveries in cancer molecular and cell biology, the mechanism governing the metastatic process is far to be understood [2]. In spite of the high mortality rate, metastasis is a very inefficient process, due to the hostile environment that cancer cells have to cross. This process consists of a sequence of basic steps: local invasion, intravasation, survival in the circulation, extravasation and colonization [3]. Figure 1 | Cancer cells metastatic colonization. The local invasion of the primary tumor by CCs is followed by their intravasation into the tumor vasculature. Then CCs enter the circulatory system as single cell or clusters which are coated with platelets. The dissemination of CCs is influenced by circulatory patterns. On their arrest in capillaries in at distant sites, the CCs extravasate into the parenchyma of target organs to begin colonization [4]. In particular, the survival in the blood stream and the following extravasation step represent one of the most critical passages due to the exposition of cancer cells to the shear stress, turbulence and immune cells attack [4]. A key aspect during metastatic dissemination is that multiple players establish competitive mechanisms whereby cancer cells experience both pro- and anti-metastatic effects [4, 5]. It is now widely accepted that platelets contribute to metastatic dissemination by creating a physical shield which protects cancer cells from the harsh fluid dynamic conditions experienced into the blood stream and prevents the attack of natural killer cells [6, 7]. On the other side, other immune cells including neutrophils, play a more controversial role [5]. The activation of the coagulation cascade and the formation of platelet rich thrombi are the main responsible of preserving circulating tumor cells and allowing their extravasation. 5 Since targeting cancer cell extravasation could represent a promising strategy to limit metastatic dissemination, it is fundamental to analyze the complex heterotypic interactions occurring among each components of the early metastatic niche and that can drive this crucial process. Previous in vivo and in vitro studies have reported the use of drugs targeting specific components of the early metastatic niche, with a particular emphasis on platelets [8, 9]. In this context, platelet glycoprotein IIb/IIIa (αIIbβ3 integrin) inhibitor Eptifibatide is a clinically approved drug reducing the risk of acute cardiac ischemic events. Our hypothesis is that inhibition of αIIbβ3 integrin on platelets may affect multiple components of the early metastatic niche, impairing extravasation. In fact, it is known that integrin signaling plays a critical role in platelet activation which seems to be the switch leading the recruitment of neutrophils to extravasation site. Despite in vivo animal models allow to study the patho-physiology of a living organism, they do not fully recapitulate the complex events associated with cancer metastasis in a human body. Indeed, it is known that potentially effective therapeutics fail after promising preclinical trials because of the significant differences existing between mouse and human immune systems [10]. On the other hand, even if standard in vitro models allow to precisely analyze the biological interactions between multiple cells types, the lack of the physiological 3D structure and functionality characterizing in vivo tissues poses a significant limitation to understand complex biological systems. To overcome these limitations, we firstly chose to adopt a 3D microfluidic model, and then to use only human cells, in order to obtain more physiological results. These devices couple the analytical advantages related to the easiness of handling of traditional 2D in vitro models with the possibility to more reliably mimic the biochemically and biophysically complex phenomena of the in vivo human 3D microenvironment. Integrated microfluidic devices are nowadays largely used in biomedical research field thanks to the invention of soft lithography and microfluidic large scales integration [11]. In the last decades, many 3D microfluidic device models have been developed, and in this thesis, we based our chip on those developed previously by our group and by Professor Kamm [12]. Our device is characterized by self-organized human microvascular networks, through which a co-culture of cancer cells, platelets and neutrophils is perfused, allowing real time high resolution analysis of extravasation events. What makes this model so appealing is the possibility to analyze metastatic 6 progression within a model of the early metastatic niche mimicking the human environment and to reach robust and rapid scoring of intra/extravascular cells due to high-resolution imaging. Aim of the work The aim of this thesis is the development of a model, based on 3D microfluidic devices, which recreates the microenvironment of the early metastatic niche. Indeed, we wanted to set the most suitable conditions which will allow us to specifically reproduce the extravasation step of the metastatic process. To reach this goal, we will firstly test the best experimental conditions which lead to the development of a human vascular network within our device. Then, within this setting, we will include some key components of the blood circulation within endothelial vessels. In particular, to recreate the early metastatic niche, we will inject a suspension of CCs, platelets and neutrophils. We will be able to detect CCs extravasation events through real-time imaging analysis. Following the device development, we will exploit its features to deeply investigate the mechanisms involved in the extavasation process, focusing on the role of αIIbβ3 platelets integrin and its effect on the extravasation events and on the other components of the early metastatic niche. These innovative features will allow to closely represent the physiological CCs extravasation microenvironment. The association with the adoption of the αIIbβ3 inhibitor, which is already used in clinical treatment, will also lead to promising results, even for clinical applications. Materials and methods The microfluidic device has been obtained through soft-lithography and PDMS replica molding and the microfluidic channels are produced by the attachment of PDMS device to a glass slide through plasma bonding. Figure 2 | PDMS device production. A. SU-8 master was placed in a plate. B. PDMS obtained by mixing 10:1 polymer and cross-linker was filtered on SU-8 master. C. PDMS crosslinking occurs in oven at 65°C for at least 2 hours. D. Device’s inlet and outlet port were obtained with puncher of proper size. E. After device’s sterilization in autoclave, it was bounded on a glass slice through plasma surface treatment. F. the device ready to use. Our device is composed by three seeding channels. 7 Figure 3 Solidworks draw of three channels device. Two channels are filled with a fibrin gel embedding fibroblasts, which release essential factors for the growth of endothelial vascular networks developed in the central channel where endothelial cells, generating the microvasculature, are seeded. After the vascular network growth, a suspension of cancer cells, platelets and neutrophils is injected into the perfusable vessels. The real time confocal analyses of the process allow cancer cells trans-migration quantification and the observation of early metastatic niche behavior. The images obtained by confocal microscope were analyzed through image processing softwares such as ImageJ and Imaris. In order to obtain the physiological microenvironment of the early metastatic niche we set up a protocol for the isolation of platelets and neutrophils from human blood. Figure 4 | Human platelets and neutrophils isolation. A. Dilution of Buffy Coat with PBS 1:2. B. The blood diluted was dripped on Ficoll surface 2:1. C. The blood and Ficoll suspension was spun at 900 g for 30 minutes, then we obtained the density gradient stratification of blood cells. D. The down layer from stratification was diluted 1:1 with Dextran 3%, after 1 hour at room temperature we observed the stratification of neutrophils (the upper layer) and red blood cell (RBC) (the bottom layer). E. The upper fraction was taken and spun. F. The pellet obtained was suspended in NaCl solution at different concentration to induce RBC lysis. The interaction of αIIbβ3 inhibitor with each component of the early metastatic niche and the interplay among these components have been also investigated through analysis of cells morphology, proliferation, protein level, adhesion and invasion assay, which allow to statistically quantify the effects of these cellcell interactions. We estimated CCs morphology changes in terms of cells circularity and major/minor axis index calculated through the “Analyzed Particle” instrument of ImageJ software. This software was also used to determine the endothelial structure through calculation of the coverage area of the plate, and the CCs invasiveness counting the number of the cells placed in a certain area of the plate. 8 Results and discussion To generate the engineered early metastatic niche in which to study the extravasation process, we first determined the most suitable device for our scope. Thus, based on the previous experience of our group, we optimized the experimental conditions in terms of cell initial concentration, duration of the experiments, and timing and dosage of the αIIbβ3 inhibitor. We found that a concentration of 3,5 milion/ml of ECs, cultured for 4 days allowed to recreate a perfusable vascular network. Figure 5 | Experiment description. A. Seeding of ECs and Fibroblast in the devices, and at the same time 2D CCs culture were seeded. B. Device medium change. D. Device medium change and addition of 0,75% αIIbβ3inhibitor to the treated condition. At the same time, platelets and neutrophils were isolated from Buffy Coat and then incubate for 24 hours with CCs. (CCs, platelets and neutrophils were divided in treated and untreated conditions too). E. CCs, platelets and neutrophils were injected in the two central medium channels. The endothelial network developed within our device allowed the crossing of the injected CCs within endothelial vessels. Figure 6 | CC route within the endothelial vessels grown in device. The morphology of the grown vessels allowed the adhesion of CCs to the endothelial wall and their following extravasation. Figure 7 | Extravasation of a CC obtained with RealTime Imaging. The results obtained by the extravasation experiments demonstrate that αIIbβ3 inhibitor significantly reduced the number of extravasated cancer cells. 9 Figure 8 | Trans endothelial migration (TEM) quantification. Our results clearly indicate that the addition of platelets and neutrophils to cancer cells significantly increase their invasive potential, positively affecting extravasation. Furthermore, our experiments demonstrated that platelets always behave in a prometastatic way, while neutrophils role still remains controversial. In particular, our morphological studies show that CCs have a more rounded morphology when co-cultured with neutrophils, suggesting that neutrophils could drive cancer cells to a quiescent phase. Figure 9 | Quantification of CCs morphology through ‘circularity’ and major/minor axis ratio’ indices. Our data show for the first time that inhibition of αIIbβ3 integrin not only impairs platelets aggregation, but also affect each components of the early metastatic niche. More in detail, Eptifibatide affects cancer cells reducing their aggressiveness due direct and indirect effects of platelets. Moreover, αIIbβ3 inhibitor also acts on endothelial cells, strengthening the architecture of inter endothelial junctions impairing metastatic dissemination. Figure 10 | Effect of αIIbβ3 inhibition on ECs. A. Protein expression determined by Western blot of key player involved in junction regulation cells with or without αIIbβ3 inhibitor. B. Immunofluorescence staining of co-cultures of ECs, CCs, platelets and neutrophils of Ve-cadherine and its phosphorylation form on ECs with or without αIIbβ3 inhibitor. We want to highlight that one of the novel aspects of this thesis is constituted by the interdisciplinary approach combining engineering and biology, which allowed to get some insights in the cellular mechanisms underlying metastatic progression. This has been achieved through the analyses of cellcell interactions and drug effects in a highly controlled 3D in vitro model, closely mimicking the human early metastatic niche. To really impact the clinical side, engineered models could be employed such as drug 10 screening platforms. Another facet of this approach is the possibility to realize different engineered human 3D organs and to connect them, to test the effects of drugs or developed side-metabolites in a body-like platform. Furthermore, given the suitability of the developed microfluidic model to answer biological questions in a reliable human environment, employing only few cells, it raises the possibility to develop an effective personalized medicine approach. In this context, highly tailored treatments according with the specific response of patient derived cells embedded in microfluidic devices could be developed improving the efficacy of the existing treatments. Conclusions In this thesis, through the development of a 3D microfluidic device we have characterized the behavior of four fundamental components of the early metastatic niche, endothelium, CCs, platelets and neutrophils. The characterization includes analysis of cell morphology, proliferation, protein levels and adhesion/TEM/invasion assays. Furthermore, we quantified the effect of the treatment with αIIbβ3 inhibitor both on each single component of the early metastatic niche, and on the whole extravasation process. Our findings allow to suggest that competitive mechanisms occurring in the early metastatic niche regulate the process of extravasation. In conclusion, this thesis highlights the importance of the combination of engineering and biology bridging the gap between traditional in vitro assays and in vivo models to reach more ambitious goal contributing to the development of new clinical therapies. Bibliography [1] B. B. d. a. A. J. R. N. Reymond, « "Crossing the endothelial barrier during metastasis,» NATURE REVIEWS | CANCER VOLUME 13 | DECEMBER 2013. . [2] L. Hutchinson, «Understanding metastasis,» NATURE REVIEWS | CLINICAL ONCOLOGY VOLUME 12 | MAY 2015. [3] M. L. a. R. O. Hynes, «The initial hours of metastasis: the importance of cooperative host-tumor cell interactions during hematogenous dissemination,» Cancer Discov. 2012 December ; 2(12): 1091–1099. . [4] J. M. &. A. C. Obenauf, «Metastatic colonization by circulating tumour cells,» N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6. [5] E. H. E. A. C. T. A. K. L. N. a. R. B. Z. Granot, «Tumor Entrained Neutrophils 11 Inhibit Seeding in the Premetastatic Lung,» Cancer Cell 20, 300–314, September 13, 2011. . [6] M. L. a. R. O. Hynes, «The initial hours of metastasis: the importance of cooperative host-tumor cell interactions during hematogenous dissemination,» Cancer Discov. 2012 December ; 2(12): 1091–1099. . [7] F. W. W. H. B. S. H. D. K. M. e. a. Im JH, «Coagulation facilitates tumor cell spreading in the pulmonary vasculature during early metastatic colony formation.,» Cancer Res. 2004;64:8613-9.. [8] K. I. S. K. L. M. S. S. Z. D. e. a. Massberg S, «Platelets secrete stromal cellderived factor 1alpha and recruit bone marrow-derived progenitor cells to arterial thrombi in vivo.,» J Exp Med. 2006;203:1221-33. [9] S. B. S. K. W. N. O. S. Schumacher D, «Platelet-derived nucleotides promote tumorcell transendothelial migration and metastasis via P2Y2 receptor.,» Cancer Cell. 2013;24:130-7. [10] S. J. K. S. K. V. C. G. L. L. e. a. Fridlender ZG, «Polarization of tumorassociated neutrophil phenotype by TGF-beta: "N1" versus "N2" TAN.,» Cancer Cell. 2009;16:183-94. [11] A. M. S. a. Y. Huang, «Chip in a lab: Microfluidics for next generation life science research,» BIOMICROFLUIDICS 7, 011302 (2013). [12] J. A. W. J. F. C. Y. Y. S. &. R. D. K. M. B. Chen, «On-chip human microvasculature assay for visualization and quantification of tumor cell extravasation dynamics,» NATURE PROTOCOLS | VOL.12 NO.5 | 2017. .
Ingegnerizzazione della nicchia pre-metastatica vascolarizzata in dispositivo microfluidico 3D per lo studio dell’extravasazione di cellule derivate da tumore alla mammella Introduzione La formazione di metastasi rappresenta l'aspetto più letale delle patologie tumorali ponendo una grande sfida per i ricercatori di tutto il mondo. Sono infatti responsabili del 90% della mortalità associata al cancro [1]. Nonostante siano stati fatti grandi progressi nei trattamenti relativi ai tumori primari, grazie a nuove scoperte nel campo della biologia cellulare e molecolare, rimane ancora da chiarire il meccanismo alla base del processo metastatico [2]. Benché il tasso di mortalità associato alla presenza di metastasi sia elevato, il processo metastatico risulta essere molto inefficiente a causa dell'ambiente ostile che le cellule tumorali devono attraversare. Questo processo consiste in una sequenza di steps fondamentali: invasione locale, entrata nel circolo ematico, sopravvivenza all’interno della circolazione sanguigna, extravasazione e colonizzazione [3]. Figura 8 | Metastatizzazione di cellule tumorali. L’invasione locale del tumore primario è seguita dall’intravasazione delle cellule tumorali all’interno della circolazione tumorale. In seguito le cellule tumorali entrano nella circolazione sistemica come cellule isolate o come cluster avvolte dalle piastrine. La disseminazione delle cellule tumorali è influenzata da caratteristiche della circolazione. Al loro arresto all’interno dei capillari lontani dal sito di origine, le cellule tumorali extravasano nel parenchima dell’organo bersaglio per la colonizzazione [4]. In particolare la sopravvivenza nel flusso sanguigno e la seguente fase di extravasazione rappresentano i passaggi più critici caratterizzati dall'esposizione delle cellule tumorali a turbolenze e a forze di taglio dovute al flusso sanguigno, e dall’attacco da parte delle cellule del sistema immunitario [4]. Un aspetto di fondamentale importanza durante la diffusione metastatica è rappresentato dall’instaurarsi di meccanismi di competizione tra i vari componenti della nicchia pre-metastatica 1 in base ai quali possono produrre effetti pro- o anti-metastatici sulle cellule tumorali [4, 5]. Ad oggi è universalmente riconosciuto l’effetto pro-metastatico delle piastrine, che contribuiscono alla disseminazione metastatica formando uno scudo fisico che protegge le cellule tumorali dalle ostili condizioni fluidodinamiche sperimentate nel flusso sanguigno, e sottraggono le cellule tumorali dall’attacco di cellule killer del sistema immunitario [6, 7]. Un diverso discorso va fatto per altre cellule immunitarie, tra le quali i neutrofili, che svolgono un ruolo molto più controverso [5]. L'attivazione della cascata di coagulazione e la formazione di trombi piastrinici sono i principali responsabili della sopravvivenza delle cellule tumorali circolanti e della loro extravasazione. Dal momento che una strategia promettente per contrastare la diffusione metastatica potrebbe essere quella di individuare come target terapeutico l’extravasazione delle cellule tumorali, è di fondamentale importanza analizzare le complesse ed eterotipiche interazioni che avvengono tra i componenti della nicchia pre-metastatica. Precedenti studi in vivo e in vitro riportano l'uso di farmaci che mirano ad agire su componenti specifici della nicchia pre-metastatica, con particolare attenzione al ruolo delle piastrine [8, 9]. In questo contesto, l’Eptifibatide, un inibitore della glicoproteina piastrinica IIb / IIIa (aIIbβ3 integrina), è un farmaco clinicamente testato che riduce il rischio di eventi ischemici cardiaci acuti. La nostra ipotesi è che l'inibizione dell'integrina αIIbβ3 sulle piastrine possa influenzare i molteplici componenti della nicchia pre-metastatica, compromettendo l'extravasazione. È infatti noto che i segnali prodotti dall’integrina svolgono un ruolo fondamentale nell'attivazione delle piastrine, le quali sembrano rappresentare l'interruttore per il reclutamento di neutrofili al sito di extravasazione. Nonostante i modelli animali in vivo permettano di studiare la fisiopatologia di un organismo vivente, non sono in grado di riprodurre completamente la complessa sequenza di eventi che porta alla formazione di metastasi nel corpo umano. Per questa ragione molte terapie ritenute potenzialmente efficaci dopo una promettente fase di sperimentazione preclinica falliscono successivamente, a causa delle significative differenze esistenti tra il sistema immunitario umano e quello murino [10]. D'altra parte, 2 nonostante i modelli in vitro standard permettano di analizzare con precisione le interazioni biologiche tra diversi tipi di cellule, la mancanza di una struttura fisiologica tridimensionale e delle funzionalità caratterizzanti dei tessuti in vivo pongono significative limitazioni nella comprensione di sistemi biologici complessi. In questo scenario, per superare i limiti degli approcci sopra elencati, abbiamo scelto di utilizzare un modello microfluidico 3D unicamente con cellule di origine umana per ottenere risultati più fisiologici. Questi dispositivi combinano i vantaggi di analisi dei modelli tradizionali 2D in vitro con la possibilità di riprodurre in maniera più affidabile i complessi fenomeni biochimici e biofisici del microambiente 3D in vivo. I dispositivi micro-fluidici integrati vengono oggi ampiamente utilizzati nel campo della ricerca grazie all'invenzione della soft-lithography e dell'integrazione su larga scala della microfluidica [11]. Negli ultimi decenni sono stati sviluppati molti modelli di dispositivi microfluidici 3D. In particolare, in questa tesi siamo stati ispirati dal modello precedentemente sviluppato dal nostro gruppo di ricerca e dal Professor Kamm [12]. Il nostro dispositivo è caratterizzato da un network micro-vascolare umano auto-organizzato, all’interno del quale viene perfusa una co-cultura di cellule tumorali, piastrine e neutrofili, consentendo l'analisi degli eventi di extravasazione. Ciò che rende interessante questo modello è la possibilità di ricostruire un network micro-vascolare 3D che consenta la riproduzione di ogni fase della nicchia metastatica precoce, in particolare della migrazione trans-endoteliale (TEM) delle cellule tumorali per mezzo di tecniche di imaging ad alta risoluzione. Obiettivo Lo scopo di questa tesi è lo sviluppo di un dispositivo microfluidico 3D in grado di ricreare il microambiente strutturale della nicchia pre-metastatica. Il nostro obiettivo è infatti quello di riuscire ad impostare le condizioni più idonee possibile che ci permettano di riprodurre specificamente la fase di extravasazione del processo metastatico. Vogliamo quindi implementare le migliori condizioni sperimentali che conducano allo sviluppo di una rete di vasi di origine umana all'interno del nostro dispositivo. Questa impostazione è in grado di consentire la simulazione 3 della circolazione sanguigna nei vasi endoteliali e, attraverso l'iniezione di una sospensione di cellule tumorali, piastrine e neutrofili, la rilevazione dell’extravasazione delle cellule tumorali grazie al Real-time imaging. Sfruttando le caratteristiche sviluppate nel device micro-fluidico, vogliamo approfondire i meccanismi che regolano il processo di extravasazione, focalizzandoci sul ruolo dell’integrina αIIbβ3, sia sugli eventi di extravasazione, che sugli altri componenti della nicchia pre-metastatica. Questi aspetti innovativi permettono di rappresentare in maniera affidabile il microambiente fisiologico di extravasazione delle cellule tumorali. L’associazione con l’utilizzo di un farmaco inibitore dell’integrina αIIbβ3, già utilizzato nel trattamento clinico, può condurre a promettenti risultati nell’applicazione clinica. Materiali e metodi Il dispositivo micro-fluidico è stato ottenuto attraverso soft-lithography e replica molding con PDMS, ed i canali micro-fluidici sono stati prodotti dall'unione tra il dispositivo PDMS e un vetrino tramite trattamento al plasma. Figura 9 | Produzione del device in PDMS. A. Il master in SU-8 viene posto sul fondo di una petri. B. Si ottiene il PDMS mischiando 10:1 le fasi polimerica e reticolante, che vengono poi collate sul SU-8 master. C. La reticolazione del PDMS avviene in forno a 65°C per almeno due ore. D. I punti di accesso del device sono ottenuti con puncher della corretta misura. E. Dopo aver sterilizzato I device in autoclave, questi vengono incollati ad un vetrino tramite trattamento al plasma. F. Device pronto all’uso. Il nostro dispositivo è compost da tre canali di semina. Figura 10 | Device a tre canali Due dei quali sono riempiti con una sospensione di gel di fibrina e fibroblasti, i quali rilasciano fattori essenziali per la crescita della rete vascolare endoteliale sviluppata nell'ultimo canale di semina. Il canale di semina centrale viene infatti 4 seminato con cellule endoteliali che generano la micro-vascolarizzazione. Dopo la crescita della rete vascolare, una sospensione di cellule tumorali, piastrine e neutrofili viene iniettata nel network endoteliale perfondibile. Analisi del processo effettuate con microscopio confocale real-time consentono la quantificazione della trans migrazione di cellule tumorali e l'osservazione dei comportamenti nella nicchia premetastatica. Le immagini ottenute dal microscopio confocale sono state analizzate tramite software di elaborazione di immagini come ImageJ e Imaris. Per ottenere il microambiente fisiologico della nicchia pre-metastatica abbiamo perfezionato un protocollo per l'isolamento di piastrine e di neutrofili dal sangue umano. Figura 11 | Isolamento di piastrine e neutrofili da sangue umano. A. Il Buffy Coat viene diluito 1:2 in PBS. B. Il sangue diluito viene colato sulla superficie del Ficoll in proporzione 2:1. C. La sospensione di sangue e Ficoll viene centrifugate a 900 g per 30 minuti, dopo di che si ottiene la stratificazione per gradiente di densità delle cellule del sangue. D. Lo strato inferiore viene diluito 1:1 Dextran 3%, e dopo l’incubazione di 1 ora a temperature ambiente si osserva la stratificazione dei neutrofili (strato superiore) e dei globuli rossi (strato inferiore). E. Lo strato superiore viene raccolto e centrifugato. F. Il pellet ottenuto viene risospeso in soluzioni a diversa concentrazione di NaCl per indurre la lisi dei globuli rossi. Sono state inoltre studiate le interazioni tra l'inibitore dell’integrina αIIbβ3 e ciascun componente della nicchia premetastatica, e le interazioni tra i componenti stessi attraverso analisi morfologica, proliferazione, valori proteici, adesione e potere invasivo delle cellule tumorali. Abbiamo infatti stimato i cambiamenti morfologici delle cellule tumorali in termini di circolarità e di un indice di rapporto tra l’asse maggiore e minore calcolati attraverso lo strumento " Analyzed Particle " del software ImageJ. Questo software è stato utilizzato anche per determinare l’integrità della struttura endoteliale attraverso il calcolo della superficie occupata dalle cellule sulla piastra da coltura, e per valutare l'invasività delle cellule tumorali contando il numero di cellule collocate in una determinata sezione della piastra da cultura. 5 Risultati e discussione Prima di tutto, abbiamo testato il dispositivo più adatto per il nostro scopo e abbiamo ottimizzato le condizioni sperimentali in termini di concentrazione iniziale delle cellule, durata degli esperimenti, tempi e dosaggi della somministrazione dell’inibitore dell’integrina αIIbβ3. Abbiamo trovato che una concentrazione di 3,5 milioni/ml di cellule endoteliali, coltivate per 4 giorni, permette un network vascolare perfondibile. Figura 12 | Descrizione dell’esperimento. A. Semina di cellule endoteliali e fibroblasti nel device. Contemporaneamente semina 2D di cellule tumorali. B. Cambio del mezzo nei device. D. Cambio del mezzo nei device e aggiunta del 0,75% di inibitore dell’integrina αIIbβ3 alle condizioni trattate con il farmaco. Isolamento di piastrine e neutrofili da Buffy Coat e loro incubazione per 24 hours con cellule tumorali, anch’esse divise tra controllo e trattamento con il farmaco. E. iniezione di cellule tumorali, piastrine e neutrofili nei canali centrali del mezzo. Il network endoteliale sviluppato all’interno del nostro dispositivo permette il passaggio nel lume dei vasi endoteliali delle cellule tumorali iniettate. Figura 13 | Percorso di una cellula tumorale nei casi endoteliali. La morfologia dei vasi sviluppati consente l’adesione alla parente endoteliale da parte delle cellule tumorali e la loro successiva extravasazione. Figura 14 | Extravasazione di una cellula tumorale osservata con Real-Time Imaging. 6 I risultati ottenuti dall'esperimento di extravasazione dimostrano la validità della nostra teoria. Infatti l'inibitore dell’integrina αIIbβ3 ostacola significativamente il potere invasivo delle cellule tumorali riducendo il numero di cellule tumorali extravasate. Figura 8 | Quantificazione della migrazione trans endoteliale (TEM). I nostri risultati indicano chiaramente che l'aggiunta di piastrine e neutrofili alle cellule tumorali aumenta significativamente il loro potenziale invasivo, influenzando positivamente l'extravasazione. I nostri esperimenti hanno inoltre dimostrato che le piastrine si comportano sempre in modo pro-metastatico, mentre il ruolo dei neutrofili rimane ancora controverso. In particolare, in accordo con i nostri studi morfologici che mostrano come le cellule tumorali cocoltivate con neutrofili assumano una forma arrotondata, abbiamo supposto che i neutrofili potrebbero indurre le cellule tumorali in una fase di quiescenza, ma questa ipotesi ha bisogno di ulteriori indagini per essere validata. Figura 9 | Quantificazione della morfologia di cellule tumorali attraverso i parametri circolarità e rapporto tra asse maggiore e minore. I nostri dati mostrano per la prima volta che l'inibizione dell'integrina αIIbβ3 non solo compromette l'aggregazione delle piastrine, ma influenza anche ogni componente della nicchia metastatica precoce. Più in dettaglio, l’Eptifibatide colpisce le cellule tumorali riducendo la loro aggressività a causa dell'azione diretta e indiretta delle piastrine. Inoltre, l'inibitore αIIbβ3 agisce anche sulle cellule endoteliali, rafforzandone l'architettura delle giunzioni interne che ostacolano la disseminazione metastatica. 7 Figura 10 | Effetto dell’inibizione dell’integrina αIIbβ3 sull’endotelio. A. Analisi tramite Western blot dell’espressione proteica di fattori chiave della regolazione delle giunzioni cellulari trattate e no trattate con il farmaco. B. Immunofluorescenza dell’espressione da parte delle cellule endoteliali di Ve-caderine e della loro forma fosforilata nelle tetra-culture di cellule endoteliali, tumorali, piastrine e neutrofili, trattate e non trattate con l’inibitore dell’integrina αIIbβ3. Vogliamo sottolineare l’innovazione apportata da questa tesi nel proporre un approccio interdisciplinare che unisce l'ingegneria e la biologia con lo scopo di fornire una mappa più dettagliata dei meccanismi cellulari su cui si basa la progressione metastatica. Questi risultati sono stati raggiunti attraverso l'analisi delle interazioni cellula-cellula e gli effetti del farmaco in un modello in vitro 3D altamente controllato, che mima strettamente il microambiente della nicchia pre-metastatica umana. Inoltre, data la validità del modello microfluidico sviluppato nel rispondere alle problematiche biologiche in un ambiente umano affidabile, impiegando solo poche cellule, aumenta la possibilità di sviluppare un efficace approccio alla medicina personalizzata. In questo contesto, i trattamenti altamente personalizzati sulla base della risposta specifica di cellule derivate da paziente, incluse in dispositivi microfluidici, potrebbero essere sviluppati migliorando l'efficacia dei trattamenti esistenti. Conclusioni In questa tesi, attraverso lo sviluppo di un dispositivo micro-fluidico 3D abbiamo caratterizzato il comportamento di quattro componenti fondamentali della nicchia pre-metastatica: endotelio, cellule tumorali, piastrine e neutrofili. La caratterizzazione comprende l'analisi della morfologia cellulare, della proliferazione, del livello proteico e dell'adesione / TEM / invasione. Abbiamo inoltre quantificato l'effetto del trattamento con un inibitore dell’integrina αIIbβ3 sia su ciascun singolo componente della nicchia pre-metastatica, sia su tutto il processo di extravasazione. I nostri risultati permettono di suggerire che i meccanismi competitivi che si instaurano nella nicchia pre-metastatica regolano il processo di extravasazione. In conclusione, questa tesi sottolinea l'importanza della combinazione di 8 approcci ingegneristici e biologici, che consentono di colmare il divario tra i tradizionali test in vitro e modelli in vivo per raggiungere obiettivi più ambiziosi che contribuiscano allo sviluppo di nuove terapie cliniche. Bibliografia [1] B. B. d. a. A. J. R. N. Reymond, « "Crossing the endothelial barrier during metastasis,» NATURE REVIEWS | CANCER VOLUME 13 | DECEMBER 2013. . [2] L. Hutchinson, «Understanding metastasis,» NATURE REVIEWS | CLINICAL ONCOLOGY VOLUME 12 | MAY 2015. [3] M. L. a. R. O. Hynes, «The initial hours of metastasis: the importance of cooperative host-tumor cell interactions during hematogenous dissemination,» Cancer Discov. 2012 December ; 2(12): 1091–1099. . [4] J. M. &. A. C. Obenauf, «Metastatic colonization by circulating tumour cells,» N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6. [5] E. H. E. A. C. T. A. K. L. N. a. R. B. Z. Granot, «Tumor Entrained Neutrophils Inhibit Seeding in the Premetastatic Lung,» Cancer Cell 20, 300–314, September 13, 2011. . [6] M. L. a. R. O. Hynes, «The initial hours of metastasis: the importance of cooperative host-tumor cell interactions during hematogenous dissemination,» Cancer Discov. 2012 December ; 2(12): 1091–1099. . [7] F. W. W. H. B. S. H. D. K. M. e. a. Im JH, «Coagulation facilitates tumor cell spreading in the pulmonary vasculature during early metastatic colony formation.,» Cancer Res. 2004;64:8613-9.. [8] K. I. S. K. L. M. S. S. Z. D. e. a. Massberg S, «Platelets secrete stromal cell-derived factor 1alpha and recruit bone marrow-derived progenitor cells to arterial thrombi in vivo.,» J Exp Med. 2006;203:1221-33. [9] S. B. S. K. W. N. O. S. Schumacher D, «Platelet-derived nucleotides promote tumor-cell transendothelial migration and metastasis via P2Y2 receptor.,» Cancer Cell. 2013;24:130-7. [10] S. J. K. S. K. V. C. G. L. L. e. a. Fridlender ZG, «Polarization of tumorassociated neutrophil phenotype by TGF- 9 beta: "N1" versus "N2" TAN.,» Cancer Cell. 2009;16:183-94. [11] A. M. S. a. Y. Huang, «Chip in a lab: Microfluidics for next generation life science research,» BIOMICROFLUIDICS 7, 011302 (2013). [12] J. A. W. J. F. C. Y. Y. S. &. R. D. K. M. B. Chen, «On-chip human microvasculature assay for visualization and quantification of tumor cell extravasation dynamics,» NATURE PROTOCOLS | VOL.12 NO.5 | 2017. .
Engineered vascularized human early metastatic niche for the study of breast cancer cells extravasation
CRIPPA, MARTINA
2016/2017
Abstract
. ABSTRACT Engineered vascularized human early metastatic niche for the study of breast cancer cells extravasation Introduction Cancer metastasis is the most lethal aspect of human cancer and represents a great challenge for worldwide research, being responsible of more than 90% of cancer-associated mortality [1]. Even if great advances in treatments related to primary tumor have been reached, due to new discoveries in cancer molecular and cell biology, the mechanism governing the metastatic process is far to be understood [2]. In spite of the high mortality rate, metastasis is a very inefficient process, due to the hostile environment that cancer cells have to cross. This process consists of a sequence of basic steps: local invasion, intravasation, survival in the circulation, extravasation and colonization [3]. Figure 1 | Cancer cells metastatic colonization. The local invasion of the primary tumor by CCs is followed by their intravasation into the tumor vasculature. Then CCs enter the circulatory system as single cell or clusters which are coated with platelets. The dissemination of CCs is influenced by circulatory patterns. On their arrest in capillaries in at distant sites, the CCs extravasate into the parenchyma of target organs to begin colonization [4]. In particular, the survival in the blood stream and the following extravasation step represent one of the most critical passages due to the exposition of cancer cells to the shear stress, turbulence and immune cells attack [4]. A key aspect during metastatic dissemination is that multiple players establish competitive mechanisms whereby cancer cells experience both pro- and anti-metastatic effects [4, 5]. It is now widely accepted that platelets contribute to metastatic dissemination by creating a physical shield which protects cancer cells from the harsh fluid dynamic conditions experienced into the blood stream and prevents the attack of natural killer cells [6, 7]. On the other side, other immune cells including neutrophils, play a more controversial role [5]. The activation of the coagulation cascade and the formation of platelet rich thrombi are the main responsible of preserving circulating tumor cells and allowing their extravasation. 5 Since targeting cancer cell extravasation could represent a promising strategy to limit metastatic dissemination, it is fundamental to analyze the complex heterotypic interactions occurring among each components of the early metastatic niche and that can drive this crucial process. Previous in vivo and in vitro studies have reported the use of drugs targeting specific components of the early metastatic niche, with a particular emphasis on platelets [8, 9]. In this context, platelet glycoprotein IIb/IIIa (αIIbβ3 integrin) inhibitor Eptifibatide is a clinically approved drug reducing the risk of acute cardiac ischemic events. Our hypothesis is that inhibition of αIIbβ3 integrin on platelets may affect multiple components of the early metastatic niche, impairing extravasation. In fact, it is known that integrin signaling plays a critical role in platelet activation which seems to be the switch leading the recruitment of neutrophils to extravasation site. Despite in vivo animal models allow to study the patho-physiology of a living organism, they do not fully recapitulate the complex events associated with cancer metastasis in a human body. Indeed, it is known that potentially effective therapeutics fail after promising preclinical trials because of the significant differences existing between mouse and human immune systems [10]. On the other hand, even if standard in vitro models allow to precisely analyze the biological interactions between multiple cells types, the lack of the physiological 3D structure and functionality characterizing in vivo tissues poses a significant limitation to understand complex biological systems. To overcome these limitations, we firstly chose to adopt a 3D microfluidic model, and then to use only human cells, in order to obtain more physiological results. These devices couple the analytical advantages related to the easiness of handling of traditional 2D in vitro models with the possibility to more reliably mimic the biochemically and biophysically complex phenomena of the in vivo human 3D microenvironment. Integrated microfluidic devices are nowadays largely used in biomedical research field thanks to the invention of soft lithography and microfluidic large scales integration [11]. In the last decades, many 3D microfluidic device models have been developed, and in this thesis, we based our chip on those developed previously by our group and by Professor Kamm [12]. Our device is characterized by self-organized human microvascular networks, through which a co-culture of cancer cells, platelets and neutrophils is perfused, allowing real time high resolution analysis of extravasation events. What makes this model so appealing is the possibility to analyze metastatic 6 progression within a model of the early metastatic niche mimicking the human environment and to reach robust and rapid scoring of intra/extravascular cells due to high-resolution imaging. Aim of the work The aim of this thesis is the development of a model, based on 3D microfluidic devices, which recreates the microenvironment of the early metastatic niche. Indeed, we wanted to set the most suitable conditions which will allow us to specifically reproduce the extravasation step of the metastatic process. To reach this goal, we will firstly test the best experimental conditions which lead to the development of a human vascular network within our device. Then, within this setting, we will include some key components of the blood circulation within endothelial vessels. In particular, to recreate the early metastatic niche, we will inject a suspension of CCs, platelets and neutrophils. We will be able to detect CCs extravasation events through real-time imaging analysis. Following the device development, we will exploit its features to deeply investigate the mechanisms involved in the extavasation process, focusing on the role of αIIbβ3 platelets integrin and its effect on the extravasation events and on the other components of the early metastatic niche. These innovative features will allow to closely represent the physiological CCs extravasation microenvironment. The association with the adoption of the αIIbβ3 inhibitor, which is already used in clinical treatment, will also lead to promising results, even for clinical applications. Materials and methods The microfluidic device has been obtained through soft-lithography and PDMS replica molding and the microfluidic channels are produced by the attachment of PDMS device to a glass slide through plasma bonding. Figure 2 | PDMS device production. A. SU-8 master was placed in a plate. B. PDMS obtained by mixing 10:1 polymer and cross-linker was filtered on SU-8 master. C. PDMS crosslinking occurs in oven at 65°C for at least 2 hours. D. Device’s inlet and outlet port were obtained with puncher of proper size. E. After device’s sterilization in autoclave, it was bounded on a glass slice through plasma surface treatment. F. the device ready to use. Our device is composed by three seeding channels. 7 Figure 3 Solidworks draw of three channels device. Two channels are filled with a fibrin gel embedding fibroblasts, which release essential factors for the growth of endothelial vascular networks developed in the central channel where endothelial cells, generating the microvasculature, are seeded. After the vascular network growth, a suspension of cancer cells, platelets and neutrophils is injected into the perfusable vessels. The real time confocal analyses of the process allow cancer cells trans-migration quantification and the observation of early metastatic niche behavior. The images obtained by confocal microscope were analyzed through image processing softwares such as ImageJ and Imaris. In order to obtain the physiological microenvironment of the early metastatic niche we set up a protocol for the isolation of platelets and neutrophils from human blood. Figure 4 | Human platelets and neutrophils isolation. A. Dilution of Buffy Coat with PBS 1:2. B. The blood diluted was dripped on Ficoll surface 2:1. C. The blood and Ficoll suspension was spun at 900 g for 30 minutes, then we obtained the density gradient stratification of blood cells. D. The down layer from stratification was diluted 1:1 with Dextran 3%, after 1 hour at room temperature we observed the stratification of neutrophils (the upper layer) and red blood cell (RBC) (the bottom layer). E. The upper fraction was taken and spun. F. The pellet obtained was suspended in NaCl solution at different concentration to induce RBC lysis. The interaction of αIIbβ3 inhibitor with each component of the early metastatic niche and the interplay among these components have been also investigated through analysis of cells morphology, proliferation, protein level, adhesion and invasion assay, which allow to statistically quantify the effects of these cellcell interactions. We estimated CCs morphology changes in terms of cells circularity and major/minor axis index calculated through the “Analyzed Particle” instrument of ImageJ software. This software was also used to determine the endothelial structure through calculation of the coverage area of the plate, and the CCs invasiveness counting the number of the cells placed in a certain area of the plate. 8 Results and discussion To generate the engineered early metastatic niche in which to study the extravasation process, we first determined the most suitable device for our scope. Thus, based on the previous experience of our group, we optimized the experimental conditions in terms of cell initial concentration, duration of the experiments, and timing and dosage of the αIIbβ3 inhibitor. We found that a concentration of 3,5 milion/ml of ECs, cultured for 4 days allowed to recreate a perfusable vascular network. Figure 5 | Experiment description. A. Seeding of ECs and Fibroblast in the devices, and at the same time 2D CCs culture were seeded. B. Device medium change. D. Device medium change and addition of 0,75% αIIbβ3inhibitor to the treated condition. At the same time, platelets and neutrophils were isolated from Buffy Coat and then incubate for 24 hours with CCs. (CCs, platelets and neutrophils were divided in treated and untreated conditions too). E. CCs, platelets and neutrophils were injected in the two central medium channels. The endothelial network developed within our device allowed the crossing of the injected CCs within endothelial vessels. Figure 6 | CC route within the endothelial vessels grown in device. The morphology of the grown vessels allowed the adhesion of CCs to the endothelial wall and their following extravasation. Figure 7 | Extravasation of a CC obtained with RealTime Imaging. The results obtained by the extravasation experiments demonstrate that αIIbβ3 inhibitor significantly reduced the number of extravasated cancer cells. 9 Figure 8 | Trans endothelial migration (TEM) quantification. Our results clearly indicate that the addition of platelets and neutrophils to cancer cells significantly increase their invasive potential, positively affecting extravasation. Furthermore, our experiments demonstrated that platelets always behave in a prometastatic way, while neutrophils role still remains controversial. In particular, our morphological studies show that CCs have a more rounded morphology when co-cultured with neutrophils, suggesting that neutrophils could drive cancer cells to a quiescent phase. Figure 9 | Quantification of CCs morphology through ‘circularity’ and major/minor axis ratio’ indices. Our data show for the first time that inhibition of αIIbβ3 integrin not only impairs platelets aggregation, but also affect each components of the early metastatic niche. More in detail, Eptifibatide affects cancer cells reducing their aggressiveness due direct and indirect effects of platelets. Moreover, αIIbβ3 inhibitor also acts on endothelial cells, strengthening the architecture of inter endothelial junctions impairing metastatic dissemination. Figure 10 | Effect of αIIbβ3 inhibition on ECs. A. Protein expression determined by Western blot of key player involved in junction regulation cells with or without αIIbβ3 inhibitor. B. Immunofluorescence staining of co-cultures of ECs, CCs, platelets and neutrophils of Ve-cadherine and its phosphorylation form on ECs with or without αIIbβ3 inhibitor. We want to highlight that one of the novel aspects of this thesis is constituted by the interdisciplinary approach combining engineering and biology, which allowed to get some insights in the cellular mechanisms underlying metastatic progression. This has been achieved through the analyses of cellcell interactions and drug effects in a highly controlled 3D in vitro model, closely mimicking the human early metastatic niche. To really impact the clinical side, engineered models could be employed such as drug 10 screening platforms. Another facet of this approach is the possibility to realize different engineered human 3D organs and to connect them, to test the effects of drugs or developed side-metabolites in a body-like platform. Furthermore, given the suitability of the developed microfluidic model to answer biological questions in a reliable human environment, employing only few cells, it raises the possibility to develop an effective personalized medicine approach. In this context, highly tailored treatments according with the specific response of patient derived cells embedded in microfluidic devices could be developed improving the efficacy of the existing treatments. Conclusions In this thesis, through the development of a 3D microfluidic device we have characterized the behavior of four fundamental components of the early metastatic niche, endothelium, CCs, platelets and neutrophils. The characterization includes analysis of cell morphology, proliferation, protein levels and adhesion/TEM/invasion assays. Furthermore, we quantified the effect of the treatment with αIIbβ3 inhibitor both on each single component of the early metastatic niche, and on the whole extravasation process. Our findings allow to suggest that competitive mechanisms occurring in the early metastatic niche regulate the process of extravasation. In conclusion, this thesis highlights the importance of the combination of engineering and biology bridging the gap between traditional in vitro assays and in vivo models to reach more ambitious goal contributing to the development of new clinical therapies. Bibliography [1] B. B. d. a. A. J. R. N. Reymond, « "Crossing the endothelial barrier during metastasis,» NATURE REVIEWS | CANCER VOLUME 13 | DECEMBER 2013. . [2] L. Hutchinson, «Understanding metastasis,» NATURE REVIEWS | CLINICAL ONCOLOGY VOLUME 12 | MAY 2015. [3] M. L. a. R. O. Hynes, «The initial hours of metastasis: the importance of cooperative host-tumor cell interactions during hematogenous dissemination,» Cancer Discov. 2012 December ; 2(12): 1091–1099. . [4] J. M. &. A. C. Obenauf, «Metastatic colonization by circulating tumour cells,» N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6. [5] E. H. E. A. C. T. A. K. L. N. a. R. B. Z. Granot, «Tumor Entrained Neutrophils 11 Inhibit Seeding in the Premetastatic Lung,» Cancer Cell 20, 300–314, September 13, 2011. . [6] M. L. a. R. O. Hynes, «The initial hours of metastasis: the importance of cooperative host-tumor cell interactions during hematogenous dissemination,» Cancer Discov. 2012 December ; 2(12): 1091–1099. . [7] F. W. W. H. B. S. H. D. K. M. e. a. Im JH, «Coagulation facilitates tumor cell spreading in the pulmonary vasculature during early metastatic colony formation.,» Cancer Res. 2004;64:8613-9.. [8] K. I. S. K. L. M. S. S. Z. D. e. a. Massberg S, «Platelets secrete stromal cellderived factor 1alpha and recruit bone marrow-derived progenitor cells to arterial thrombi in vivo.,» J Exp Med. 2006;203:1221-33. [9] S. B. S. K. W. N. O. S. Schumacher D, «Platelet-derived nucleotides promote tumorcell transendothelial migration and metastasis via P2Y2 receptor.,» Cancer Cell. 2013;24:130-7. [10] S. J. K. S. K. V. C. G. L. L. e. a. Fridlender ZG, «Polarization of tumorassociated neutrophil phenotype by TGF-beta: "N1" versus "N2" TAN.,» Cancer Cell. 2009;16:183-94. [11] A. M. S. a. Y. Huang, «Chip in a lab: Microfluidics for next generation life science research,» BIOMICROFLUIDICS 7, 011302 (2013). [12] J. A. W. J. F. C. Y. Y. S. &. R. D. K. M. B. Chen, «On-chip human microvasculature assay for visualization and quantification of tumor cell extravasation dynamics,» NATURE PROTOCOLS | VOL.12 NO.5 | 2017. .File | Dimensione | Formato | |
---|---|---|---|
Crippa Martina_Tesi Magistrale_Ingegneria Biomedica.pdf
solo utenti autorizzati dal 12/07/2020
Descrizione: Tesi Magistrale_Martina Crippa
Dimensione
5.14 MB
Formato
Adobe PDF
|
5.14 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/135400