Abstract Introduction Retina is a fundamental component of the eye in the transmission of the visual stimulus; it is a highly specialized tissue composed of different layers and cell types, including the Retinal Pigmented Epithelium (RPE), neurons, photoreceptors, and bipolar, amacrine and horizontal cells. The RPE is formed by a layer of epithelial cells held together by tight junctions, and is separated from the choroid by Bruch’s membrane, which is a highly dynamic acellular tissue, mainly responsible for providing support to the EPR and the choriocapillaris and for regulating the transport of substances in both directions; a dysfunction or degeneration of the RPE leads to the malfunction of the receptors. In fact, retina can be subject to various pathologies, first of all Age-related Macular Degeneration, which is one of the main causes of blindness worldwide and occurs with the onset of deposits of extracellular material that – starting from the RPE – reach the overlying Bruch's membrane. Most retinal diseases are particularly challenging and often the treatment can only be palliative, in fact for many of them nowadays there is no real cure that allows the recovery of vision lost; for this, an alternative and innovative solution has increasingly become necessary. One of the main options is in the field of tissue engineering and involves retinal cell transplant combined with a scaffold to replace damaged or dead cells with viable ones. Preliminary results have shown that the scaffold improves cell survival upon implantation compared to cell-only injections and promotes the formation of a functional epithelium; moreover, if the scaffold has adequate properties, it can act as a substitute for Bruch's membrane, which also gets damaged by Age-related Macular Degeneration. Among the different scaffold manufacturing methods, electrospinning is an effective option because it allows to obtain porous membranes with thicknesses equal to a few micrometres, made of nanometric fibres, thanks to an electric field. In this thesis research, to develop a model of Bruch’s membrane, the first goal was to create flat membranes with aligned fibres in order to assess their properties. With established electrospinning it is in fact possible to obtain two-dimensional structures, but in particular for retinal approaches, it would be more suitable to obtain a three-dimensional shape support, which could better adapt to the natural shape of the retina: in this regard, 3D electrospinning was used in the aim of manufacturing hemispherical membranes able to reproduce the patient specific curvature of the human eye. Materials and Methods First of all, in this thesis work a protocol was developed for the creation via electrospinning of flat membranes having a preferential organization of the fibres, in order to evaluate if this could significantly influence the properties of the scaffold, by comparison with a flat membrane instead having a random organization of the fibres and ultimately with the native Bruch’s membrane that we wanted to reproduce. For this purpose, starting from a solution of polycaprolactone (PCL) and silk fibroin, four potential electrospinning setups with different operational parameters have been developed. The alignment of the fibres in the resulting scaffolds was assessed by analysing SEM (Scanning Electron Microscope) images of the scaffolds using ImageJ software. To verify the support function of the scaffold, mechanical tensile tests were carried out on the scaffold with a random orientation of the fibres and on the one that had shown the best alignment; after four preconditioning cycles, the samples were brought to failure and from the data obtained it was possible to derive Young's elastic modulus, which is equal to the slope of the first linear region of the stress-strain plot, and the stress and strain at break, which define the ability of a material to resist tensile stress. Finally, its permeability was evaluated, that is the ability of the membrane to be crossed by a liquid. The second goal of this study concerned the fabrication of hemispherical membranes with a curvature similar to that of the human retina. For this purpose, a novel collector was developed and applied to the electrospinning setup; a mould in three modular pieces was designed, using SolidWorks® 2020 software, so that the base and the top could be reused and the hemisphere could be adapted to the specific patient, precisely replicating the geometry of his eye, which brings significant advantages. The mould was then manufactured thanks to the use of an FDM 3D printer (Fused Deposition Modelling), which allows you to precisely replicate the curvature of the retina and also makes it possible to create rounded edges; thanks to this feature, the silicone collector obtained from the mould had rounded edges, so as not to interfere with the electrospinning process, and a hole was also created at the base to allow electrical conduction. The assembly and the coating materials of the collector have been optimized by experimenting with different setups, keeping the operational parameters of the electrospinning constant. To analyse the variations in the distribution of the electric field, which depend on the different materials of the collector, simulations were carried out with COMSOL Multiphysics® software. The resulting scaffolds were selected basing on their shape, their stability and the uniformity of their fibres, and the best two were further analysed from a mechanical point of view and for their permeability, in a similar manner to what was previously described. Results As confirmed by SEM images, flat membranes with aligned fibres were successfully fabricated through the setup ("Setup B") which involved the use of a cylindrical collector with a speed rate of 775 rpm. In this setup the polymer solution was extruded from two mobile spinnerets (25 mm/s): the use of two needles allows to electrospin a certain volume of solution in half the time and furthermore, being mobile, scaffolds can be obtained of greater width and less thickness compared to electrospinning carried out by using a single fixed needle. To evaluate the influence of fibre directionality, the aligned scaffold was compared with a random scaffold manufactured using a protocol optimized in previous studies. The direction of the fibres (Figure A.1) has been confirmed by the histogram, in fact it is possible to notice how the random scaffold had an almost uniform distribution of fibres for each inclination, while the aligned scaffold shows a narrower and higher peak in the distribution, confirming that most of the fibres were oriented at ~ 97°. It was also found that the aligned scaffold had fibres with double diameter compared to the other (aligned: daverage = 5.066 ± 1.619 μm; random: daverage = 10.861 ± 5.395 μm) and also a greater thickness (aligned: s = 58.63 μm; random: s = 44 μm). Figure A.1 - Top: SEM images comparison (scalebar: 20 μm); bottom: quantitative evaluation of fibre alignment through ImageJ The stress-strain plots were obtained from the mechanical tests (Figure A.2): both curves show an initial linear trend, followed by a sudden change in slope. Their comparison showed that the mean Young's modulus was 43.364 ± 9.088 MPa in the aligned scaffold and 24.476 ± 5.906 MPa in the random one. The alignment of the fibres, on the other hand, meant that the failure occurred at lower stress and strain values, compared to the random ones; in the aligned one, the rupture occurred at a stress of 3.014 ± 0.165 MPa and at a deformation of 17.452 ± 2.697%, while in the random one at a stress of 6.967 ± 2.983 MPa and at a deformation of 79.477 ± 7.456%. Figure A.2 – Stress-strain plots of flat membranes Permeability was evaluated on the flat scaffolds and on the hemispherical ones in the same way, imposing pressure values within the physiological range. The resulting plot for the flat membranes can be found in Figure A.3: these curves show an exponential trend, probably due to a progressive failure of the scaffold fibres as a consequence of the increasing pressure; the difference between the two is not very pronounced, in fact the initial values are in the order of 10-8 and the final values in that of 10-7. Figure A.3 – Permeability curves for flat membranes For the production of hemispherical scaffolds, on the other hand, the combination of materials that gave the best results involved the use of a conductive flat collector coated in aluminium, while the new collector was directly exposed on the outside, but the inner surface of the hemisphere was instead covered in parchment paper, with an aluminium circle on the bottom. It had the correct shape, was stable and also had a more homogeneous fibre density than the other developed scaffolds. The properties of this scaffold (“Setup VI”) were compared with a second hemispherical scaffold (“Setup II”) which also had a correct stability and shape, but a less homogeneous fibre density; in this setup, the collector was internally covered in aluminium while the underlying plate collector was wrapped in polyethylene. For the computational analysis it was chosen to analyse the distribution of the y component of the electric field as it is perpendicular to the collector plane, at a height y = 17 mm so that Ey was observed right on top of the collector, making it possible to understand how different combinations of materials affected the electric field. The result (Figure A.4) for the selected scaffolds matched the experimental results. In fact, in setup II in correspondence with the insulated plate, the electric field is at a minimum, while it is uniformly equal to its maximum in the hemisphere and in the immediately surrounding area. Configuration VI instead had a conductive plate externally insulated and was conductive only at the bottom of the hemisphere, in fact the area corresponding to the silicone collector is blue, a sign that the electric field was at a minimum, while it increased both away from it , therefore proceeding towards the flat collector, and going towards the hole present at the base of the hemisphere. Figure A.4 –Ey distribution on top of the collector in setup II (left) and setup VI (right) From the morphological analysis it emerged that in the scaffold obtained from setup II the fibres were highly randomized and had an average diameter of 8.451 ± 2.455 μm. As regards setup VI, the resulting conductivity also influenced the distension of the fibres, in fact their average diameter is 2.151 ± 0.54 μm; furthermore, the deposition of the fibres is slightly more regular, almost grid-like (Figure A.5). Figure A.5 - A) Frontal view of the scaffold obtained from setup II; B) Frontal view of the scaffold obtained from setup VI; C) SEM image of the scaffold obtained from setup II (scalebar: 20 μm); D) SEM image of the scaffold obtained from setup VI (scalebar: 20 μm) Similarly to what was done for flat membranes, mechanical tensile tests were carried out and the corresponding stress-strain plots are found in Figure A.6. The hemispherical scaffold obtained from setup VI proved to be more resistant, in fact showing a higher average Young's modulus of 43.333 ± 27.538 MPa, a strain at failure of 13.220 ± 8.157% and an ultimate tensile strength of 2.110 ± 0.70 MPa. The other scaffold, on the contrary, had an average Young's modulus of 25.544 ± 2.426 MPa, a strain at failure of 0.185 ± 0.068% and an ultimate tensile strength of 1.098 ± 1.369 MPa.. Figure A.6 – Comparison of stress-strain plots from the hemispherical scaffolds As far as the permeability of the hemispherical scaffolds is concerned, between the two curves (Figure A.7) there is a more relevant difference, in fact the curves start from 10-6, but then the one corresponding to setup VI ends at 10-5 while that of setup II reaches a value almost 10 times higher. Figure A.7 – Permeability curves for hemispherical scaffolds Discussion In this work, flat membranes with a preferential direction of the fibres were first developed. From the morphological comparison of the scaffold with aligned fibres and a scaffold with a random organization of the fibres, it emerged that the elastic modules of both were higher than that of Bruch’s membrane (E = 17 MPa), which gives both membranes the ability to resist manipulation. The best option turned out to be the random scaffold, having a greater thickness and fibres with a diameter greater than two orders of magnitude compared to those of the native membrane (s = 2-4 μm; d = 60 nm); its mechanical performance was in fact superior to that of Bruch’s membrane, and at the same time - compared to the other scaffold - it had more suitable dimensions. A possible solution to further bring the dimensions closer to the physiological values could be to optimize some operational parameters of the electrospinning, such as the voltage or the distance between the needle and the collector, but also to try and keep the temperature and relative humidity values constant during the process. For hemispherical scaffolds, on the other hand, it turned out that the best scaffold had a fibre diameter more similar to that of Bruch’s membrane and, also in this case, the scaffold proved to be more mechanically performing than the native membrane, and this is very positive, considering that the construct must be able to maintain its shape and withstand the necessary manipulations during the implantation phase. The analysed artificial scaffolds showed a significantly higher permeability than the physiological values, and this result guarantees the transport of substances as it happens physiologically. Conclusions Of the two goals initially set, it was possible to fabricate flat scaffolds with aligned fibres and also to develop hemispherical scaffolds that retain their shape. The combination of the practical work with the simulations carried out with COMSOL Multiphysics® also made it possible to fully understand the theoretical functioning of electrospinning and to verify each step carried out in the laboratory. Future developments of the work consist in the modification of some operational parameters of electrospinning, in order to further reduce the diameter of the scaffold fibres and their thickness, to make them more similar to Bruch’s membrane. Furthermore, it would be interesting to perform cytocompatibility and epithelial formation tests. Finally, considering that the ultimate purpose of this construct would be in vivo implantation, it might be useful to study the degradation times of the scaffold in a physiological environment.
Sommario Introduzione La retina è una componente dell’occhio fondamentale per la trasmissione dello stimolo visivo; è un tessuto altamente specializzato composto da diversi strati e tipologie cellulari, tra cui l’Epitelio Pigmentato Retinico (EPR), neuroni, fotorecettori, cellule bipolari, amacrine e orizzontali. L’EPR è formato da uno strato di cellule epiteliali tenuto insieme mediante giunzioni strette, ed è separato dalla coroide mediante la membrana di Bruch, che è invece un tessuto acellulare altamente dinamico, responsabile principalmente di fornire supporto all’EPR e ai coriocapillari e di regolare il trasporto di sostanze nelle due direzioni; una disfunzione o degenerazione dell’EPR porta al malfunzionamento dei recettori. La retina infatti può essere soggetta a diverse patologie, prima tra tutte la degenerazione maculare senile, che è tra le principali cause di cecità a livello globale e si presenta mediante l’insorgenza di depositi di materiale extracellulare che a partire dall’EPR raggiungono la sovrastante membrana di Bruch. La maggior parte delle patologie retiniche sono particolarmente problematiche e spesso l’unico trattamento può essere soltanto palliativo, infatti per molte di esse al giorno d’oggi non esiste una vera e propria cura che permetta il recupero della visione persa; per questo, si rende sempre più necessaria una soluzione alternativa e innovativa. Una delle principali opzioni proposte appartiene al campo dell’ingegneria dei tessuti e prevede un trapianto di cellule retiniche combinato con l’utilizzo di uno scaffold per sostituire le cellule danneggiate o morte con cellule vitali. Risultati preliminari hanno mostrato che lo scaffold migliora la sopravvivenza cellulare in seguito all’impianto rispetto a iniezioni di sole cellule e promuove la formazione di un epitelio funzionale; inoltre, se lo scaffold presenta adeguate proprietà, può fungere da sostituto della membrana di Bruch, anch’essa danneggiata dalla degenerazione maculare senile. Tra i diversi metodi di fabbricazione dello scaffold, l’elettrofilatura rappresenta un’efficace opzione perché permette di ottenere membrane porose con spessori pari a pochi micrometri, composte da fibre di dimensioni nanometriche grazie all’utilizzo di un campo elettrico. In questo lavoro di tesi, per sviluppare un modello della membrana di Bruch, il primo obiettivo è stato quello di creare delle membrane piane aventi fibre allineate, per valutarne le proprietà. Con l’elettrofilatura canonica è infatti possibile ottenere strutture bidimensionali, ma in particolare per approcci retinici, sarebbe più adeguato ottenere un supporto di forma tridimensionale, che si adatti meglio alla forma naturale della retina: a tal proposito in questo studio è stata utilizzata l’elettrofilatura 3D, una tecnologia molto recente e ancora poco diffusa e studiata, allo scopo di fabbricare membrane emisferiche in grado di riprodurre la curvatura paziente specifica dell’occhio umano. Materiali e Metodi Per prima cosa, in questo lavoro di tesi è stato messo a punto un protocollo per la creazione mediante elettrofilatura di membrane piane aventi un’organizzazione preferenziale delle fibre, in modo da valutare se ciò potesse influenzare in modo significativo le proprietà dello scaffold, a confronto con una membrana piana avente invece un’organizzazione casuale delle fibre e in ultima analisi con la membrana nativa di Bruch che si voleva riprodurre. A questo scopo, partendo da una soluzione di policaprolattone (PCL) e fibroina della seta, sono stati sviluppati quattro potenziali setup di elettrofilatura aventi diversi parametri operazionali. L’allineamento delle fibre negli scaffold risultanti è stato valutato analizzando le immagini SEM (Microscopio Elettronico a Scansione) degli scaffold mediante l’utilizzo del software ImageJ. Per verificare la funzione di supporto dello scaffold sono state effettuate delle prove meccaniche a trazione sullo scaffold avente orientamento casuale delle fibre e su quello che aveva mostrato il migliore allineamento; in seguito a quattro cicli di precondizionamento, i campioni sono stati portati a rottura e dai dati ottenuti è stato possibile ricavare il modulo elastico di Young, che è pari alla pendenza della prima regione lineare del grafico sforzo-deformazione, lo sforzo a rottura e la deformazione a rottura, che definiscono l’abilità di un materiale di resistere ad una trazione. Infine, ne è stata valutata la permeabilità, ossia la capacità della membrana di essere attraversata da un liquido. Il secondo obiettivo di questo studio riguardava la produzione di membrane emisferiche, aventi una curvatura analoga a quella della retina umana. A tale scopo è stato sviluppato un collettore innovativo da applicare al setup dell’elettrofilatura; si è progettato uno stampo in tre pezzi componibili, mediante il software SolidWorks® 2020, in modo che la base e il coperchio potessero essere riutilizzati e la semisfera potesse essere adattata allo specifico paziente, replicando in modo preciso la geometria del suo occhio, il che comporta notevoli vantaggi. Lo stampo è stato poi fabbricato grazie all’utilizzo di una stampante 3D FDM (Modellazione a Deposizione Fusa), che permette di replicare fedelmente la curvatura della retina e inoltre rende possibile creare degli spigoli arrotondati; grazie a questa accortezza, il collettore in silicone ottenuto a partire dallo stampo presentava degli spigoli smussati, in modo da non interferire con il processo di elettrofilatura, ed è stato anche previsto un foro alla base per consentire la conduzione elettrica. Il montaggio e i materiali di rivestimento del collettore sono stati ottimizzati sperimentando diversi setup, mantenendo costanti i parametri operazionali dell’elettrofilatura. Per analizzare le variazioni nella distribuzione del campo elettrico, che dipendono dai diversi materiali del collettore, sono state effettuate delle simulazioni con il software COMSOL Multiphysics®. Gli scaffold risultanti sono stati selezionati in base alla loro forma, alla loro stabilità e all’uniformità delle loro fibre, e i due migliori sono stati analizzati ulteriormente da un punto di vista meccanico e per la loro permeabilità, in maniera analoga a quanto precedentemente descritto. Risultati Come confermato dalle immagini al SEM, membrane piane con fibre allineate sono state fabbricate con successo tramite il setup (“Setup B”) che prevedeva l’utilizzo di un collettore cilindrico con velocità di rotazione pari a 775 rpm. In questo setup la soluzione polimerica viene estrusa da due spinneret mobili (25 mm/s): l’utilizzo di due aghi permette di elettrofilare una determinata quantità di volume di soluzione nella metà del tempo e inoltre, essendo mobili, si possono ottenere degli scaffold di larghezza maggiore e spessore minore rispetto all’elettrofilatura effettuata mediante l’utilizzo di un solo ago fisso. Per valutare l’influenza della direzionalità delle fibre, lo scaffold a fibre allineate è stato confrontato con uno scaffold a fibre random fabbricato tramite un protocollo ottimizzato in precedenti studi. La direzione delle fibre (Figura S.1) è stata confermata dal grafico, infatti è possibile notare come lo scaffold random presentava per ogni inclinazione una distribuzione pressoché uniforme di fibre, mentre lo scaffold allineato mostra un picco stretto ed alto nella distribuzione, confermando che la maggior parte delle fibre era orientata a ~97°. È inoltre risultato che lo scaffold allineato presentava delle fibre con diametro doppio rispetto all’altro (allineato: dmedio = 5.066 ± 1.619 μm; random: dmedio = 10.861 ± 5.395 μm) e anche uno spessore maggiore (allineato: s = 58.63 μm; random: s = 44 μm). Figura S.1 - In alto: confronto delle immagini al SEM (scala: 20 μm); in basso: valutazione quantitativa dell'allineamento mediante ImageJ Dalle prove meccaniche sono stati ricavati i diagrammi di sforzo-deformazione (Figura S.2): entrambe le curve presentano un iniziale andamento lineare, seguito da un repentino cambio di pendenza. Dal loro confronto è emerso che il modulo di Young medio era pari a 43,364 ± 9,088 MPa nello scaffold allineato e 24,476 ± 5,906 MPa in quello casuale. L'allineamento delle fibre, invece, ha fatto sì che il cedimento avvenisse a valori di sforzo e deformazione inferiori, rispetto a quelli casuali; in quello allineato, la rottura è avvenuta ad uno sforzo di 3.014 ± 0.165 MPa e ad una deformazione di 17.452 ± 2.697 %, mentre in quello casuale ad uno sforzo di 6.967 ± 2.983 MPa e ad una deformazione di 79.477 ± 7.456 %. Figura S.2 - Curve di sforzo-deformazione delle due membrane piane La permeabilità è stata valutata sui due scaffold piani e su quelli emisferici allo stesso modo, imponendo dei valori di pressione compatibili con quelli fisiologici. Il grafico risultante per le membrane piane si può trovare in Figura S.3: queste curve presentano un andamento esponenziale, probabilmente dovuto ad un progressivo cedimento delle fibre dello scaffold a causa della pressione crescente; la differenza tra le due non è molto pronunciata, infatti i valori iniziali sono dell’ordine di 10-8 ed i valori finali in quello di 10-7. Figura S.3 - Curve di permeabilità per gli scaffold piani Per la produzione di scaffold emisferici invece, la combinazione di materiali che ha dato i risultati migliori prevedeva l’utilizzo di un collettore piano conduttivo rivestito in alluminio, mentre il collettore innovativo era direttamente esposto al suo esterno, ma la superficie interna della semisfera era invece rivestita di carta da forno, con un cerchio in alluminio sul fondo. Esso presentava la forma corretta, era stabile ed inoltre aveva una densità di fibre molto più omogenea rispetto agli altri scaffold sviluppati. Le prestazioni di questo scaffold (“Setup VI”) sono state confrontate con un secondo scaffold emisferico (“Setup II”) avente stabilità e forma corretta, ma una densità di fibre meno omogenea; in questo setup, il collettore era internamente ricoperto in alluminio mentre la piastra sottostante era avvolta in polietilene. Per l’analisi computazionale si è scelto di analizzare la distribuzione della componente y del campo elettrico in quanto perpendicolare al piano del collettore, ad un’altezza y = 17 mm in modo che Ey fosse osservato proprio all’apice del collettore, rendendo possibile capire come le diverse combinazioni di materiali influenzassero il campo elettrico. Il risultato (Figura S.4) per gli scaffold selezionati combaciava con i risultati sperimentali. Nel setup II infatti in corrispondenza della piastra isolata il campo elettrico è al minimo, mentre è uniformemente pari al massimo nella semisfera e nell’area immediatamente circostante. La configurazione VI prevedeva invece una piastra conduttiva, mentre il collettore era isolato esternamente e conduttivo soltanto sul fondo della semisfera, infatti la zona in corrispondenza del collettore in silicone è blu, segno che il campo elettrico era al minimo, mentre aumentava sia allontanandosi da esso, procedendo quindi verso il collettore piano, sia andando verso il foro presente alla base della semisfera. Figura S.4 – Distribuzione di Ey all’apice del collettore nel setup II (a sinistra) e VI (a destra) Dall’analisi morfologica è emerso che nello scaffold ottenuto dal setup II le fibre erano altamente randomizzate e avevano un diametro medio pari a 8,451 ± 2,455 μm. Per quanto riguarda il setup VI, la conducibilità risultante ha influenzato anche la distensione delle fibre, infatti il loro diametro medio è di 2,151 ± 0,54 μm; inoltre, la deposizione delle fibre è più regolare, quasi a griglia (Figura S.5). Figura S.5 - A) Vista frontale dello scaffold ottenuto con il setup II; B) Vista frontale dello scaffold ottenuto con il setup VI; C) Ingrandimento al SEM dello scaffold ottenuto dal setup II (scala: 20 μm); D) Ingrandimento al SEM dello scaffold ottenuto dal setup VI (scala 20 μm) Analogamente a quanto fatto per le membrane piane, sono state svolte delle prove meccaniche a trazione e i diagrammi di sforzo-deformazione corrispondenti si trovano in Figura S.6. Lo scaffold emisferico ottenuto dal setup VI si è dimostrato più resistente, mostrando infatti un modulo di Young medio più elevato e pari a 43,333 ± 27,538 MPa, una deformazione a rottura del 13,220 ± 8,157 % e una resistenza alla trazione di 2,110 ± 0,70 MPa. L’altro scaffold, invece, aveva un modulo di Young medio pari a 25,544 ± 2,426 MPa, una deformazione a rottura di 0,185 ± 0,068% e una resistenza alla trazione di 1,098 ± 1,369 MPa. Figura S.6 - Confronto delle curve di sforzo-deformazione nei due scaffold emisferici Per quanto riguarda la permeabilità degli scaffold emisferici, tra le due curve (Figura S.7) è presente una differenza più rilevante, infatti le curve partono da 10-6, ma poi quella corrispondente al setup VI termina a 10-5 mentre quella del setup II raggiunge un valore quasi 10 volte maggiore. Figura S.7 - Curve di permeabilità per gli scaffold piani (sinistra) ed emisferici (destra) Discussione In questo lavoro, in primo luogo sono state sviluppate delle membrane piane aventi una direzione preferenziale delle fibre. Dal confronto morfologico dello scaffold ottenuto con fibre allineate e uno scaffold invece con un’organizzazione casuale delle fibre, è emerso che i moduli elastici di entrambi erano superiori a quello della membrana di Bruch (E = 17 MPa), il che conferisce ad entrambe le membrane la capacità di resistere alle manipolazioni. L’opzione migliore è risultata essere lo scaffold random, avente uno spessore maggiore e fibre con un diametro maggiore di due ordini di grandezza rispetto a quelle della membrana nativa (s = 2-4 μm; d = 60 nm); le sue prestazioni meccaniche erano infatti superiori a quelle della membrana di Bruch, e contemporaneamente – rispetto all’altro scaffold – presentava delle dimensioni più adatte. Una possibile soluzione per avvicinare ulteriormente le dimensioni ai valori fisiologici potrebbe essere di ottimizzare alcuni parametri operazionali dell’elettrofilatura, come ad esempio il voltaggio o la distanza tra ago e collettore, ma anche di cercare di mantenere costanti i valori di temperatura e umidità relativa durante il processo. Per gli scaffold emisferici invece, è risultato che lo scaffold migliore presentava un diametro delle fibre più simile a quello della membrana di Bruch e, anche in questo caso, lo scaffold si è rivelato più performante meccanicamente rispetto alla membrana nativa, e ciò è molto positivo, considerato che il costrutto deve essere in grado di mantenere la propria forma e resistere alle necessarie manipolazioni in fase di impianto. Tutti gli scaffold artificiali analizzati presentavano una permeabilità sensibilmente maggiore rispetto ai valori fisiologici, e questo risultato garantisce il trasporto di sostanze come avviene fisiologicamente. Conclusioni Dei due obiettivi inizialmente posti, è stato possibile realizzare degli scaffold piani con fibre allineate e anche sviluppare degli scaffold emisferici che mantenessero la propria forma. L’accostamento al lavoro pratico delle simulazioni svolte con COMSOL Multiphysics® inoltre ha permesso di comprendere a pieno il funzionamento teorico dell’elettrofilatura e di verificare ogni passo svolto in laboratorio. Sviluppi futuri del lavoro consistono nell’ottimizzazione di alcuni parametri operazionali dell’elettrofilatura, allo scopo di ridurre ulteriormente il diametro delle fibre degli scaffold e il loro spessore, per renderli più simili alla membrana di Bruch. Inoltre, sarebbe interessante effettuare dei test di citocompatibilità e di formazione di epitelio. Infine, considerato che lo scopo ultimo di questo costrutto sarebbe l’impianto in vivo, potrebbe essere utile studiare i tempi di degradazione dello scaffold in ambiente fisiologico.
Fabrication of hemispherical electrospun scaffolds for retinal tissue engineering featuring a novel setup
Restelli, Anna
2019/2020
Abstract
Abstract Introduction Retina is a fundamental component of the eye in the transmission of the visual stimulus; it is a highly specialized tissue composed of different layers and cell types, including the Retinal Pigmented Epithelium (RPE), neurons, photoreceptors, and bipolar, amacrine and horizontal cells. The RPE is formed by a layer of epithelial cells held together by tight junctions, and is separated from the choroid by Bruch’s membrane, which is a highly dynamic acellular tissue, mainly responsible for providing support to the EPR and the choriocapillaris and for regulating the transport of substances in both directions; a dysfunction or degeneration of the RPE leads to the malfunction of the receptors. In fact, retina can be subject to various pathologies, first of all Age-related Macular Degeneration, which is one of the main causes of blindness worldwide and occurs with the onset of deposits of extracellular material that – starting from the RPE – reach the overlying Bruch's membrane. Most retinal diseases are particularly challenging and often the treatment can only be palliative, in fact for many of them nowadays there is no real cure that allows the recovery of vision lost; for this, an alternative and innovative solution has increasingly become necessary. One of the main options is in the field of tissue engineering and involves retinal cell transplant combined with a scaffold to replace damaged or dead cells with viable ones. Preliminary results have shown that the scaffold improves cell survival upon implantation compared to cell-only injections and promotes the formation of a functional epithelium; moreover, if the scaffold has adequate properties, it can act as a substitute for Bruch's membrane, which also gets damaged by Age-related Macular Degeneration. Among the different scaffold manufacturing methods, electrospinning is an effective option because it allows to obtain porous membranes with thicknesses equal to a few micrometres, made of nanometric fibres, thanks to an electric field. In this thesis research, to develop a model of Bruch’s membrane, the first goal was to create flat membranes with aligned fibres in order to assess their properties. With established electrospinning it is in fact possible to obtain two-dimensional structures, but in particular for retinal approaches, it would be more suitable to obtain a three-dimensional shape support, which could better adapt to the natural shape of the retina: in this regard, 3D electrospinning was used in the aim of manufacturing hemispherical membranes able to reproduce the patient specific curvature of the human eye. Materials and Methods First of all, in this thesis work a protocol was developed for the creation via electrospinning of flat membranes having a preferential organization of the fibres, in order to evaluate if this could significantly influence the properties of the scaffold, by comparison with a flat membrane instead having a random organization of the fibres and ultimately with the native Bruch’s membrane that we wanted to reproduce. For this purpose, starting from a solution of polycaprolactone (PCL) and silk fibroin, four potential electrospinning setups with different operational parameters have been developed. The alignment of the fibres in the resulting scaffolds was assessed by analysing SEM (Scanning Electron Microscope) images of the scaffolds using ImageJ software. To verify the support function of the scaffold, mechanical tensile tests were carried out on the scaffold with a random orientation of the fibres and on the one that had shown the best alignment; after four preconditioning cycles, the samples were brought to failure and from the data obtained it was possible to derive Young's elastic modulus, which is equal to the slope of the first linear region of the stress-strain plot, and the stress and strain at break, which define the ability of a material to resist tensile stress. Finally, its permeability was evaluated, that is the ability of the membrane to be crossed by a liquid. The second goal of this study concerned the fabrication of hemispherical membranes with a curvature similar to that of the human retina. For this purpose, a novel collector was developed and applied to the electrospinning setup; a mould in three modular pieces was designed, using SolidWorks® 2020 software, so that the base and the top could be reused and the hemisphere could be adapted to the specific patient, precisely replicating the geometry of his eye, which brings significant advantages. The mould was then manufactured thanks to the use of an FDM 3D printer (Fused Deposition Modelling), which allows you to precisely replicate the curvature of the retina and also makes it possible to create rounded edges; thanks to this feature, the silicone collector obtained from the mould had rounded edges, so as not to interfere with the electrospinning process, and a hole was also created at the base to allow electrical conduction. The assembly and the coating materials of the collector have been optimized by experimenting with different setups, keeping the operational parameters of the electrospinning constant. To analyse the variations in the distribution of the electric field, which depend on the different materials of the collector, simulations were carried out with COMSOL Multiphysics® software. The resulting scaffolds were selected basing on their shape, their stability and the uniformity of their fibres, and the best two were further analysed from a mechanical point of view and for their permeability, in a similar manner to what was previously described. Results As confirmed by SEM images, flat membranes with aligned fibres were successfully fabricated through the setup ("Setup B") which involved the use of a cylindrical collector with a speed rate of 775 rpm. In this setup the polymer solution was extruded from two mobile spinnerets (25 mm/s): the use of two needles allows to electrospin a certain volume of solution in half the time and furthermore, being mobile, scaffolds can be obtained of greater width and less thickness compared to electrospinning carried out by using a single fixed needle. To evaluate the influence of fibre directionality, the aligned scaffold was compared with a random scaffold manufactured using a protocol optimized in previous studies. The direction of the fibres (Figure A.1) has been confirmed by the histogram, in fact it is possible to notice how the random scaffold had an almost uniform distribution of fibres for each inclination, while the aligned scaffold shows a narrower and higher peak in the distribution, confirming that most of the fibres were oriented at ~ 97°. It was also found that the aligned scaffold had fibres with double diameter compared to the other (aligned: daverage = 5.066 ± 1.619 μm; random: daverage = 10.861 ± 5.395 μm) and also a greater thickness (aligned: s = 58.63 μm; random: s = 44 μm). Figure A.1 - Top: SEM images comparison (scalebar: 20 μm); bottom: quantitative evaluation of fibre alignment through ImageJ The stress-strain plots were obtained from the mechanical tests (Figure A.2): both curves show an initial linear trend, followed by a sudden change in slope. Their comparison showed that the mean Young's modulus was 43.364 ± 9.088 MPa in the aligned scaffold and 24.476 ± 5.906 MPa in the random one. The alignment of the fibres, on the other hand, meant that the failure occurred at lower stress and strain values, compared to the random ones; in the aligned one, the rupture occurred at a stress of 3.014 ± 0.165 MPa and at a deformation of 17.452 ± 2.697%, while in the random one at a stress of 6.967 ± 2.983 MPa and at a deformation of 79.477 ± 7.456%. Figure A.2 – Stress-strain plots of flat membranes Permeability was evaluated on the flat scaffolds and on the hemispherical ones in the same way, imposing pressure values within the physiological range. The resulting plot for the flat membranes can be found in Figure A.3: these curves show an exponential trend, probably due to a progressive failure of the scaffold fibres as a consequence of the increasing pressure; the difference between the two is not very pronounced, in fact the initial values are in the order of 10-8 and the final values in that of 10-7. Figure A.3 – Permeability curves for flat membranes For the production of hemispherical scaffolds, on the other hand, the combination of materials that gave the best results involved the use of a conductive flat collector coated in aluminium, while the new collector was directly exposed on the outside, but the inner surface of the hemisphere was instead covered in parchment paper, with an aluminium circle on the bottom. It had the correct shape, was stable and also had a more homogeneous fibre density than the other developed scaffolds. The properties of this scaffold (“Setup VI”) were compared with a second hemispherical scaffold (“Setup II”) which also had a correct stability and shape, but a less homogeneous fibre density; in this setup, the collector was internally covered in aluminium while the underlying plate collector was wrapped in polyethylene. For the computational analysis it was chosen to analyse the distribution of the y component of the electric field as it is perpendicular to the collector plane, at a height y = 17 mm so that Ey was observed right on top of the collector, making it possible to understand how different combinations of materials affected the electric field. The result (Figure A.4) for the selected scaffolds matched the experimental results. In fact, in setup II in correspondence with the insulated plate, the electric field is at a minimum, while it is uniformly equal to its maximum in the hemisphere and in the immediately surrounding area. Configuration VI instead had a conductive plate externally insulated and was conductive only at the bottom of the hemisphere, in fact the area corresponding to the silicone collector is blue, a sign that the electric field was at a minimum, while it increased both away from it , therefore proceeding towards the flat collector, and going towards the hole present at the base of the hemisphere. Figure A.4 –Ey distribution on top of the collector in setup II (left) and setup VI (right) From the morphological analysis it emerged that in the scaffold obtained from setup II the fibres were highly randomized and had an average diameter of 8.451 ± 2.455 μm. As regards setup VI, the resulting conductivity also influenced the distension of the fibres, in fact their average diameter is 2.151 ± 0.54 μm; furthermore, the deposition of the fibres is slightly more regular, almost grid-like (Figure A.5). Figure A.5 - A) Frontal view of the scaffold obtained from setup II; B) Frontal view of the scaffold obtained from setup VI; C) SEM image of the scaffold obtained from setup II (scalebar: 20 μm); D) SEM image of the scaffold obtained from setup VI (scalebar: 20 μm) Similarly to what was done for flat membranes, mechanical tensile tests were carried out and the corresponding stress-strain plots are found in Figure A.6. The hemispherical scaffold obtained from setup VI proved to be more resistant, in fact showing a higher average Young's modulus of 43.333 ± 27.538 MPa, a strain at failure of 13.220 ± 8.157% and an ultimate tensile strength of 2.110 ± 0.70 MPa. The other scaffold, on the contrary, had an average Young's modulus of 25.544 ± 2.426 MPa, a strain at failure of 0.185 ± 0.068% and an ultimate tensile strength of 1.098 ± 1.369 MPa.. Figure A.6 – Comparison of stress-strain plots from the hemispherical scaffolds As far as the permeability of the hemispherical scaffolds is concerned, between the two curves (Figure A.7) there is a more relevant difference, in fact the curves start from 10-6, but then the one corresponding to setup VI ends at 10-5 while that of setup II reaches a value almost 10 times higher. Figure A.7 – Permeability curves for hemispherical scaffolds Discussion In this work, flat membranes with a preferential direction of the fibres were first developed. From the morphological comparison of the scaffold with aligned fibres and a scaffold with a random organization of the fibres, it emerged that the elastic modules of both were higher than that of Bruch’s membrane (E = 17 MPa), which gives both membranes the ability to resist manipulation. The best option turned out to be the random scaffold, having a greater thickness and fibres with a diameter greater than two orders of magnitude compared to those of the native membrane (s = 2-4 μm; d = 60 nm); its mechanical performance was in fact superior to that of Bruch’s membrane, and at the same time - compared to the other scaffold - it had more suitable dimensions. A possible solution to further bring the dimensions closer to the physiological values could be to optimize some operational parameters of the electrospinning, such as the voltage or the distance between the needle and the collector, but also to try and keep the temperature and relative humidity values constant during the process. For hemispherical scaffolds, on the other hand, it turned out that the best scaffold had a fibre diameter more similar to that of Bruch’s membrane and, also in this case, the scaffold proved to be more mechanically performing than the native membrane, and this is very positive, considering that the construct must be able to maintain its shape and withstand the necessary manipulations during the implantation phase. The analysed artificial scaffolds showed a significantly higher permeability than the physiological values, and this result guarantees the transport of substances as it happens physiologically. Conclusions Of the two goals initially set, it was possible to fabricate flat scaffolds with aligned fibres and also to develop hemispherical scaffolds that retain their shape. The combination of the practical work with the simulations carried out with COMSOL Multiphysics® also made it possible to fully understand the theoretical functioning of electrospinning and to verify each step carried out in the laboratory. Future developments of the work consist in the modification of some operational parameters of electrospinning, in order to further reduce the diameter of the scaffold fibres and their thickness, to make them more similar to Bruch’s membrane. Furthermore, it would be interesting to perform cytocompatibility and epithelial formation tests. Finally, considering that the ultimate purpose of this construct would be in vivo implantation, it might be useful to study the degradation times of the scaffold in a physiological environment.| File | Dimensione | Formato | |
|---|---|---|---|
|
Restelli Anna - Master's Thesis.pdf
solo utenti autorizzati dal 23/05/2022
Descrizione: Tesi di Laurea Magistrale - Anna Restelli
Dimensione
6.24 MB
Formato
Adobe PDF
|
6.24 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/10589/175114