In this thesis a novel model of human blood brain barrier developed on a microfluidic chip is presented. This model employed human induced Pluripotent Stem Cells (hiPSCs) differentiated in neurons and astrocytes, human pericytes and human Cerebral Microvascular Endothelial Cells (hCMEC/D3). This 3D microfluidic platform consisted in a PDMS device having four parallel microchannels. In one of the external channels, both conceived for cell culture medium, hCMEC/D3 were seeded along its walls forming a vessel-like structure. In the two central channels, conceived for being filled with hydrogels, astrocytes were embedded together with pericytes in a fibrin gel close to hCMEC/D3 channel, and in the other central channel neurons were embedded in a collagen gel. The goal of my project was dual: at first I studied the influence of pericytic and astrocytic cellular populations on the hCMEC/D3 monolayer in two cases: both when they were singularly co-cultured with HCMEC/D3, and also by maintaining cells with conditioned media by astrocytes/pericytes, which are culture media collected and stored after one day they fed astrocytes/pericytes in a standard flask. Thus these conditioned media, which according to many evidences in literature induced a more BBB phenotype on the hCMEC/D3 barrier, contained some cytokines released by cells. Secondly, I developed an electronic board in order to perform with 4 electrodes instead of 2 electrodes transendothelial electrical resistance (TEER, which is a parameter accounting for the tightness of the cell to cell junctions: the higher this resistance is, the tighter these junctions should be) measurements, since the measurements performed with four electrodes should be more precise and reliable. This electronic board was developed by manually soldering and microsoldering the integrated circuit components. An Arduino MINI chip was programmed to control the board operation, which is an extension circuit of an existing printed circuit board (EVAL-AD5933 EBZ by Analog Devices) available and purchased from the market. To study astrocyte/pericyte influence on hCMEC/D3 barrier two main assays were carried out: TEER measurement (recorded along all the 7 days cells were cultured in the microfluidic chip thanks to the cited board by inserting a couple of electrodes in the microfluidic device) and permeability assay (only at the final 7th day of culture, measured by evaluating the flux of FITC labeled dextran across the hCMEC/D3 barrier), which allowed to compute a permeability coefficient that was lower as the barrier was tighter. Thanks to immunocytochemistry confocal imaging was performed to evaluate the expression of BBB proteins and to evaluate cellular morphology. Moreover, neuron vitality was proved by performing calcium imaging, prior to calcium ion immunostaining and potassium chloride (KCl) stimulus. Briefly, when KCl was injected it forced to open calcium channels of neural cells and an increasing in fluorescence was recorded once calcium ion entered the cell. Results showed that conditioned media both by pericytes and by astrocytes did not provide any tightening effect on the hCMEC/D3 barrier of my model. In fact, as reported by TEER trends and by permeability coefficients measured, the values of the cells maintained with conditioned media were worse with respect to values obtained by cells maintained with standard culture media. Also in co-culture few co-culture conditions (only when hCMEC/D3 were co-cultured with pericytes at very low concentrations) showed better values of the aforementioned parameters. Interestingly, astrocytes as well were proven to form less complex networks in the fibrin gel when maintained with astrocyte conditioned medium: this analysis was performed through an appositely designed function of Imaris software called filament analysis. Calcium imaging proved that neurons were viable and functional. Finally, TEER measurement performed with the electronic board displayed more noisy values with respect to the 2 electrode configuration, since too high spurious oscillations were recorded in any measurement, probably due to the manually soldered connections that distorted the acquired signal.
In questo lavoro di tesi viene presentato un nuovo modello di barriera ematoencefalica umana sviluppato su un chip microfluidico. Questo modello ha impiegato cellule staminali umane pluripotenti indotte (hiPSCs) differenziate in neuroni e in astrociti, periciti umani e Human Cerebral Microvascular Endothelial Cells (hCMEC/D3). Questa piattaforma microfluidica tridimensionale consiste in un dispositivo PDMS avente quattro microcanali paralleli. In uno dei canali esterni, entrambi concepiti per il mezzo di coltura cellulare, le hCMEC/D3 sono state seminate lungo le pareti formando una struttura simile ad un vaso sanguigno. Nei due canali centrali, concepiti per essere riempiti con idrogeli, gli astrociti sono stati incorporati insieme a periciti in un gel di fibrina vicino al canale delle hCMEC/D3 e nell'altro canale centrale i neuroni sono stati seminati in un gel di collagene. L'obiettivo di questo progetto è stato doppio: innanzitutto ho studiato l'influenza delle popolazioni cellulari pericitiche e astrocitiche sulla barriera formata da hCMEC/D3 in due casi: sia quando erano co-coltivate con HCMEC/D3, sia mantenendo le cellule con mezzi condizionati da astrociti o periciti, che sono i mezzi di coltura raccolti e conservati dopo un giorno che hanno nutrito astrociti o periciti in una flask standard: questi mezzi di coltura condizionati, che secondo molte evidenze presenti in letteratura inducono nelle hCMEC/D3 un fenotipo della barriera ematoencefalica più simile a quello in vivo, contenevano alcune citochine rilasciate dalle cellule. In secondo luogo, ho sviluppato una board elettronica per eseguire con 4 elettrodi invece che con 2 elettrodi misure di resistenza elettrica transendoteliale (TEER, che è un parametro che tiene conto di quanto le giunzioni tran le cellule siano sviluppate: maggiore è questa resistenza, più strette queste giunzioni dovrebbero essere), dato che le misure effettuate con quattro elettrodi dovrebbero essere più precise e affidabili. Questa board elettronica è stata sviluppata da saldando manualmente e i componenti del circuito integrato. Un chip Arduino MINI è stato programmato per piotare il funzionamento della board, che è quindi un circuito di estensione di una circuito stampato già esistente (EVAL-AD5933 EBZ da Analog Devices) acquistato sul mercato. Per studiare l'influenza degli astrocitici / periciti sulla barriera formata dalle hCMEC / D3 sono stati effettuati due test principali: la misura di TEER (registrata lungo tutti i 7 giorni di coltura cellulare, i cui dati sono stati ottenuti dal microfluidico grazie alla board sopracitata inserendo due elettrodi nel dispositivo microfluidico) e un saggio di permeabilità (solo al 7 ° giorno della coltura, misurato valutando il flusso di destrano coniugato al FITC), che permetteva di calcolare un coefficiente di permeabilità più basso se la barriera era più selettiva. Grazie all'immunoistochimica è stata eseguito l’imaging al confocale per valutare l'espressione delle proteine della barriera ematoencefalica e la morfologia cellulare. Inoltre, la vitalità neuronale è stata dimostrata eseguendo l'imaging del calcio: a seguito dell’immunoistochimica degli ioni calcio, uno stimolo stimolo di cloruro di potassio (KCl) apriva i canali cellulari di tale ione. Un aumento della fluorescenza veniva registrato una volta che lo ione calcio entrava nella cellula. I risultati hanno mostrato che i mezzi di coltura condizionati sia da periciti che da astrociti non hanno fornito alcun effetto benefico sulla barriera hCMEC/D3 di questo modello. Infatti, come riportato dagli andamenti di TEER e dai coefficienti di permeabilità misurati, i valori misurati sulle cellule mantenute con i mezzi condizionati erano peggiori rispetto ai valori ottenuti dalle cellule mantenute con mezzi di coltura standard. Anche in co-coltura poche condizioni (solo quando hCMEC/D3 sono state co-coltivate con periciti a concentrazioni molto basse) hanno mostrato valori migliori dei parametri menzionati. È interessante notare come gli astrociti formavano reti cellulari meno complesse nel gel di fibrina quando erano mantenuti con il mezzo condizionato da astrociti: questa analisi è stata eseguita attraverso una funzione del software Imaris appositamente progettata chiamata analisi dei filamenti. L'imaging del calcio ha dimostrato invece che i neuroni erano vitali e funzionali. Infine, la misura TEER eseguita con la board elettronica ha mostrato valori più rumorosi rispetto alla configurazione a 2 elettrodi, poiché in qualsiasi misurazione sono state registrate oscillazioni anomale troppo elevate, probabilmente a causa della saldatura manuale delle connessioni che ha distorto il segnale acquisito.
Human blood brain barrier on a chip
CRESTANI, MICHELE
2016/2017
Abstract
In this thesis a novel model of human blood brain barrier developed on a microfluidic chip is presented. This model employed human induced Pluripotent Stem Cells (hiPSCs) differentiated in neurons and astrocytes, human pericytes and human Cerebral Microvascular Endothelial Cells (hCMEC/D3). This 3D microfluidic platform consisted in a PDMS device having four parallel microchannels. In one of the external channels, both conceived for cell culture medium, hCMEC/D3 were seeded along its walls forming a vessel-like structure. In the two central channels, conceived for being filled with hydrogels, astrocytes were embedded together with pericytes in a fibrin gel close to hCMEC/D3 channel, and in the other central channel neurons were embedded in a collagen gel. The goal of my project was dual: at first I studied the influence of pericytic and astrocytic cellular populations on the hCMEC/D3 monolayer in two cases: both when they were singularly co-cultured with HCMEC/D3, and also by maintaining cells with conditioned media by astrocytes/pericytes, which are culture media collected and stored after one day they fed astrocytes/pericytes in a standard flask. Thus these conditioned media, which according to many evidences in literature induced a more BBB phenotype on the hCMEC/D3 barrier, contained some cytokines released by cells. Secondly, I developed an electronic board in order to perform with 4 electrodes instead of 2 electrodes transendothelial electrical resistance (TEER, which is a parameter accounting for the tightness of the cell to cell junctions: the higher this resistance is, the tighter these junctions should be) measurements, since the measurements performed with four electrodes should be more precise and reliable. This electronic board was developed by manually soldering and microsoldering the integrated circuit components. An Arduino MINI chip was programmed to control the board operation, which is an extension circuit of an existing printed circuit board (EVAL-AD5933 EBZ by Analog Devices) available and purchased from the market. To study astrocyte/pericyte influence on hCMEC/D3 barrier two main assays were carried out: TEER measurement (recorded along all the 7 days cells were cultured in the microfluidic chip thanks to the cited board by inserting a couple of electrodes in the microfluidic device) and permeability assay (only at the final 7th day of culture, measured by evaluating the flux of FITC labeled dextran across the hCMEC/D3 barrier), which allowed to compute a permeability coefficient that was lower as the barrier was tighter. Thanks to immunocytochemistry confocal imaging was performed to evaluate the expression of BBB proteins and to evaluate cellular morphology. Moreover, neuron vitality was proved by performing calcium imaging, prior to calcium ion immunostaining and potassium chloride (KCl) stimulus. Briefly, when KCl was injected it forced to open calcium channels of neural cells and an increasing in fluorescence was recorded once calcium ion entered the cell. Results showed that conditioned media both by pericytes and by astrocytes did not provide any tightening effect on the hCMEC/D3 barrier of my model. In fact, as reported by TEER trends and by permeability coefficients measured, the values of the cells maintained with conditioned media were worse with respect to values obtained by cells maintained with standard culture media. Also in co-culture few co-culture conditions (only when hCMEC/D3 were co-cultured with pericytes at very low concentrations) showed better values of the aforementioned parameters. Interestingly, astrocytes as well were proven to form less complex networks in the fibrin gel when maintained with astrocyte conditioned medium: this analysis was performed through an appositely designed function of Imaris software called filament analysis. Calcium imaging proved that neurons were viable and functional. Finally, TEER measurement performed with the electronic board displayed more noisy values with respect to the 2 electrode configuration, since too high spurious oscillations were recorded in any measurement, probably due to the manually soldered connections that distorted the acquired signal.File | Dimensione | Formato | |
---|---|---|---|
Michele Crestani - Thesis final.pdf
non accessibile
Descrizione: Sommario dettagliato e testo della tesi
Dimensione
6.02 MB
Formato
Adobe PDF
|
6.02 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/10589/135399