Self-assembling nanosystems based on elastin-like recombinamers and antimicrobial peptides for biomedical applications

PROGRAMA DE DOCTORADO EN INVESTIGACIÓN BIOMÉDICA TESIS DOCTORAL: SELF-ASSEMBLING NANOSYSTEMS BASED ON ELASTIN-LIKE RECOMBINAMERS AND ANTIMICROBIAL PEPTIDES FOR BIOMEDICAL APPLICATIONS Presentada por Sergio Acosta Rodríguez para optar al grado de Doctor por la Universidad de Valladolid Dirigida por: Dr. José Carlos Rodríguez-Cabello A mi familia Our greatest weakness lies in giving up. The most certain way to succeed is always to try just one more time. Thomas Alva Edison Agradecimientos Esta Tesis es el resultado de varios años de trabajo, durante los cuales he vivido multitud de experiencias tanto profesionales como personales que me han permitido evolucionar y crecer como investigador y madurar como persona. Con estas líneas me gustaría agradecer a todo el mundo que ha compartido conmigo su tiempo durante estos años y que directa o indirectamente han contribuido al desarrollo de esta Tesis. En primer lugar, quería expresar mis agradecimientos a mis directores de tesis, Carlos, por la oportunidad de iniciar mi carrera investigadora en su grupo de investigación, por toda la formación recibida y enseñarme a trabajar en ciencia. Quiero continuar agradeciendo a todos los profesores e investigadores senior del grupo. Gracias Matilde, Luis, Javi, Merche y Alessandra por los buenos momentos, por vuestra amabilidad, vuestros consejos y vuestra ayuda. Durante todo este tiempo he tenido la oportunidad de compartir poyata con muchos y muy dispares compañeros, de los que he aprendido mucho de todos y con cada uno de ellos. Muchas gracias por regalarme tantos y tan buenos recuerdos, habéis sido una parte muy importante de este viaje. Gracias a Isra, Rocío, Menchu, Mohammed, Alicia, Arturo, Guille, Laura, Chus, Lucia, Soraya, Reinal, Eduardo, Doriana, Ito, Filippo, Leander, Miguel, Sofía, Juan, Marcos, Fernando, Irene y Sara. Y en especial a Tatjana que tanto me ha aguantado. Muchas gracias también a Irene y a Teresa por su amabilidad y comprensión, y a Beatriz por su simpatía y buena disposición. Quiero dar las gracias a Conrado por haberme acogido en su grupo de investigación en dos ocasiones, por haberme ayudado tanto desde el primer momento en mis andaduras minnesotas, y por ser además de un referente profesional, un buen amigo. I would like to thank all my colleagues from the MDRCBB, to Prof. Rudney and Prof. Gorr for offering me the opportunity to learn in their labs. I am also grateful to Ms. Ruoqiong Chen for her good humor and invaluable help. A mis amigos, por estar ahí, por escucharme y servir de vía de escape para desconectar siempre que lo he necesitado. Por último, mil gracias a mis padres y hermana, por creer en mí y por su incondicional apoyo en todo lo que hago. ¡Gracias por vuestra paciencia! A todos, ¡¡Gracias!! VI Enzimas endonucleasas (SapI, EarI, EcoRI, DpnI, XhoI, XbaI) y enzimas (T4 DNA ligasa, FastAP, SAP) ThermoFisher Scientific, EE.UU. Etanol Panreac Química S.L.U., España Extracto de levadura BD Biosciences, EE.UU. Fluoruro de fenilmetilsulfato (PMSF) Apollo Scientific, EE.UU. Gentamicina/Anfotericina ThermoFisher Scientific, EE.UU. Glicerol Sigma-Aldrich, EE.UU. Glicina Sigma-Aldrich, EE.UU. Glucosa Panreac Química S.L.U., España Glutaraldehído Sigma-Aldrich, EE.UU. Hemina Sigma-Aldrich, EE.UU. Hidróxido de sodio (NaOH) ThermoFisher Scientific, EE.UU. Kit de extracción de ADN en geles de agarosa “PureLink Quick Gel Extraction” Invitrogen, EE.UU. Kit de purificación de ADN MasterPure™ Epicenter, Illumina, Inc., EE.UU. Kit de purificación de plásmidos (NucleoSpin Plasmid) Macherey-Nagel, Germany Leupeptina Apollo Scientific, EE.UU. Marcador de ADN (1kb Plus Ladder) Invitrogen, EE.UU. Marcador de proteínas (Pierce Unstained) ThermoFisher Scientific, EE.UU. Resumen VII Medio de cultivo de auto-inducción (Terrific Broth,TB) Formedium, Reino Unido Medio de cultivo de lisogenia (Luria Bertani, LB) Conda, España Medio de cultivo Mueller Hinton tipo II (MHB) Sigma-Aldrich, EE.UU. Medio de cultivo Todd Hewitt Broth (THB) BD Biosciences, EE.UU. Medio de cultivo Tryptone Soy Broth (TSB) Sigma-Aldrich, EE.UU. Medios de cultivo (DMEM 1g/L glucosa, Medium 200) Gibco, EE.UU. Menadiona (vitamina K3) Sigma-Aldrich, EE.UU. Mucina gástrica porcina parcialmente purificada Sigma-Aldrich, EE.UU. Penicilina/Estreptomicina Gibco, EE.UU. Pepstatina A Apollo Scientific, EE.UU. Peptona (Triplicate peptone) Sigma-Aldrich, EE.UU. Proteosa-peptona Sigma-Aldrich, EE.UU. Suero fetal bovino (FBS) Gibco, EE.UU. Tampón fosfato salino (PBS) Gibco, EE.UU. Tetrametiletilendiamina (TEMED) Sigma-Aldrich, EE.UU. Tetróxido de osmio (OsO4) Sigma-Aldrich, EE.UU. Tinción de ADN (SimplySafe) Eurx, Polonia Tinción de viabilidad Filmtracer LIVE/DEAD ThermoFisher Scientific, EE.UU. Tripsina/EDTA 0.05% Gibco, EE.UU. VIII Tris(hidroximetil)aminometano (Tris) Sigma-Aldrich, EE.UU. Azul Tripan Invitrogen, EE.UU. Urea Sigma-Aldrich, EE.UU. β-Mercaptoetanol Sigma-Aldrich Adicionalmente, distintas cepas bacterianas o líneas celulares humanas fueron empleadas para llevar a cabo la construcción génica, test antimicrobianos y pruebas de citocompatibilidad (Tabla 2). Tabla 2. Cepas bacterianas y líneas celulares empleadas en el desarrollo de esta tesis. Ensayo Cepa bacteriana/Línea celular Proveedor Clonaje Escherichia coli XL-1 Blue competent cells Agilent, EE.UU. Subclonaje Escherichia coli XL-1 Blue subcloning Agilent, EE.UU. Producción recombinante Escherichia coli BLR (DE3) ThermoFisher Scientific, EE.UU. Actividad antimicrobiano Staphylococcus aureus 25923 ATCC, EE.UU. Actividad antimicrobiano Staphylococcus epidermidis 35984 ATCC, EE.UU. Actividad antimicrobiano Streptococcus gordonii DL-1 Prof. Joel Rudney (Universidad de Minnesota, EE.UU.) Actividad antimicrobiano Pseudomonas aeruginosa PAO1 Xen41 PerkinElmer Inc., Waltham, EE.UU. Resumen IX Actividad antimicrobiano Microcosmo oral 769NS Prof. Joel Rudney (Universidad de Minnesota, EE.UU.) Análisis de la citocompatibilidad Fibroblastos de prepucio humano (HFF-1) Life Technologies, España Análisis de la citocompatibilidad Fibroblastos primarios gingivales de origen humano ATCC, EE.UU. 2.2. Métodos 2.2.1. Síntesis recombinante de ELRs Todos los ELRs empleados durante el desarrollo de esta Tesis fueron producidos recombinantemente y de forma exclusiva para el desarrollo de la misma. En primer lugar, se llevó a cabo la construcción de los genes codificantes de los ELRs y AMP-ELRs mediante ingeniería genética siguiendo el método iterativo-recursivo previamente descrito.[2] Los genes codificantes de las secuencias monoméricas de los ELRs o de las secuencias específicas, como AMPs, fueron producidos por una empresa externa (NZYTech, Portugal). Estos genes se clonaron en el plásmido de clonación pDriveAll utilizando E. coli XL1 como cepa de clonación. Dicho plásmido proviene del vector de clonación comercial pDrive (Novagen), en el que se mutaron las dianas de reconocimiento de las endonucleasas tipo II SapI y EarI, para llevar a cabo la construcción de los genes de manera controlada y direccional.[2] Una vez alcanzada la construcción final, ésta se introdujo en el plásmido de expresión p7, desarrollado también en el grupo Bioforge a partir del vector comercial pET25(+) (Novagen, EE.UU.).[3] Este vector de expresión incorpora las secuencias de reconocimiento para su sobreexpresión en cepas E. coli con sistema de expresión basado en la ARN X polimerasa del fago T7. En nuestro caso, los plásmidos de expresión con los constructos deseados fueron transformados en la cepa BLR (DE3), cepa derivada de BL21 que carece de la recombinasa recA,[4] favoreciendo, de este modo, la expresión de secuencias poliméricas. Tras esto, se realizó un screening de las cepas hiperproductoras mediante electroforesis en condiciones desnaturalizantes (SDS-PAGE) y la cepa que mayor expresión mostró se empleó para inocular un fermentador de 15 L (Applikon Biotechnology). La fermentación se llevó a cabo en condiciones de pH, temperatura, agitación y concentración de O2 controladas durante 16-18 horas. Seguidamente, tras el lavado de las bacterias con tampón salino, se extrajo el biomaterial mediante disrupción de las mismas en un disruptor mecánico (homogeneizador GEA Lab PandaPLUS 2000, GEA Farm Technologies, Nueva Zelanda) y se purificó mediante ciclos de calentamiento-centrifugación (40 °C) y enfriamientocentrifugación (4 °C) consecutivos aprovechando la separación de fase reversible dependiente de la temperatura de los ELRs. El solvente para disolver los ELRs fue agua y se añadió una concentración de 1.5 M de NaCl para contribuir a la precipitación en caliente de los biomateriales. Tras 3-4 ciclos de purificación se verificó que los ELRs estén completamente puros mediante SDS-PAGE, se dializaron frente agua destilada (3 cambios) y agua ultrapura (1 cambio) para eliminar todo tipo de sales, se filtraron a través de filtros de 0.22 µm (Nalgene, EE.UU.) para conseguir un producto estéril que se liofilizaron (FreeZone 1, LABCONCO) y congelaron a -20 °C hasta su uso. 2.2.2. Modificación de ELRs Durante esta tesis se han realizado dos modificaciones químicas de ELRs expuestas a continuación. Resumen XI Liberación bloque protector Los AMP-ELRs híbridos fueron diseñados como pro-polipéptidos de tal modo que se incorporó un bloque protector con varias funciones: (i) aumentar expresión del constructo híbrido, (ii) proteger al microorganismo huésped de posibles efectos secundarios dañinos, (iii) incorpora una Met en el extremo C-terminal, de tal modo que es fácilmente escindible en un único paso mediante CNBr y (iv) su naturaleza catiónica facilita su separación del constructo de interés mediante precipitación ácida. La liberación del bloque protector se llevó a cabo disolviendo el ELR a una concentración de 20 mg mL-1 en ácido fórmico (70%) y añadiendo CNBr (200:1 relación molar CNBr:Met). Tras 20 h de incubación en oscuridad a T ambiente, el exceso de CNBr se eliminó en un rotavapor y se dializó frente a agua destilada (4 pasos). Modificación de ELRs con D-péptidos Se probó a incorporar en los ELRs péptidos antimicrobianos compuestos por D-amino ácidos, los cuales no son producidos de manera natural por el microorganismo productor E. coli. El péptido empleado (DGL13K) fue sintetizado químicamente por la empresa Pepscan (Holanda) con una pureza superior al 90% e incluyendo un grupo azida (N3) en su extremo C-terminal (N3-Gkiiklkaslkll-NH2). Los grupos azida reaccionan selectivamente con grupos ciclooctinos en condiciones fisiológicas sin precisar de solventes químicos que comprometan la citocompatibilidad. De este modo, se modificaron con grupos ciclooctino los grupos amino de las cadenas laterales de las Lys presentes en el ELR, siguiendo el protocolo previamente desarrollado en el grupo Bioforge.[5] XII 2.2.3. Caracterización físico-química de los ELRs Electroforesis en gel de poliacrilamida La electroforesis en geles de poliacrilamida se realizó en condiciones desnaturalizantes, añadiendo el detergente SDS, que desnaturaliza las proteínas, para una migración a través del gel dependiente de su peso molecular (MW). Para ello, empleamos el sistema de electroforesis vertical “MiniVE” de Hoefer (Amersham Pharmacia Biotec, Reino Unido) donde se corrieron las muestras a una intensidad fija de 25 mA por gel. Los geles se prepararon siguiendo el protocolo descrito por Laemmli.[6] Cuando el frente de proteínas salió del gel, éstos se tiñeron utilizando una disolución acuosa de CuCl2 (0.3 M). Las imágenes de los geles se tomaron con el sistema Gel Doc™ EZ Imager (Bio-Rad, EE.UU.). Espectrometría de masas y análisis de aminoácidos De manera complementaria, el peso molecular de los ELRs se evaluó mediante espectrometría de masas del tipo “Matrix-assisted laser desorption/ionization–Time-of-flight” (MALDI-TOF) utilizando el equipo Voyager STR (Applied Biosystems, EE.UU.). Además, la composición aminoacídica fue analizada mediante cromatografía tras la hidrolización de la muestra empleando un equipo de “High-Performance Liquid Chromatography” (HPLC) en gradiente Waters 600 acoplado a un detector de UV Waters 2487 (Waters, EE.UU.). Ambas técnicas se llevaron a cabo en el Laboratorio de Técnicas Instrumentales (LTI) de la Universidad de Valladolid. Caracterización del comportamiento térmico Para caracterizar el comportamiento térmico de los ELRs a altas concentraciones se empleó el equipo de calorimetría diferencial de barrido, Mettler Toledo 822e, con refrigeración mediante nitrógeno líquido. La Resumen XIII concentración utilizada fue de 50 mg mL-1 y se empleó un volumen de 20 µL. Los experimentos consistieron en una primera etapa isotérmica a 5 ºC durante 5 minutos, seguida de una etapa de calentamiento, desde 5 a 60 ºC, a una velocidad de 5 ºC min-1. La caracterización a bajas concentraciones (1 mg mL-1) se llevó a cabo mediante medidas de turbidimetría o de dispersión dinámica de luz (DLS). Se prepararon las muestras a 25 µM en agua ultrapura. Para medir la turbidimetría de las muestras respecto la temperatura se programaron rampas de calentamiento de 5 a 40 °C (con una velocidad de 0.25 °C min-1 y se registró el valor de la absorbancia de la muestra a 350 nm cada grado en el espectrofotómetro Cary 100UV-Vis (Agilent, EE.UU.). Para calcular la temperatura de transición mediante DLS, se registró la intensidad de luz dispersada cada 4 ° C desde 5 hasta 53 °C en un equipo ZetaSizer Nano (Malvern Instruments, Reino Unido). Las muestras se estabilizaron durante 2 minutos a cada temperatura y se midieron por triplicado con 11 corridas por medición. La temperatura a la cual la intensidad dispersada alcanzó el 50% de la intensidad se consideró como la temperatura de solución crítica baja. Dicroísmo circular Los espectros de dicroísmo circular (CD) se obtuvieron utilizando un espectropolarímetro Jasco J-810 (Jasco, EE.UU.) equipado con un controlador de temperatura en los servicios técnicos a la investigación (SSTTI) de la Universidad de Alicante (España). Las muestras de ELR se disolvieron a 5 mM y se midieron en células de cuarzo de 0.1 cm en el rango de 190-250 nm. La temperatura se estabilizó a 37 °C durante 10 minutos antes de la medición. Los porcentajes de estructura secundaria se determinaron usando el servidor web BeStSel (Beta Selection Structure, o XIV selección de estructuras beta) [7,8] en el rango de 200-250 nm, siempre y cuando el voltaje del dinodo se encontrara por debajo de 500 nm. Caracterización de nanoestructuras mediante técnicas de dispersión de luz dinámica La distribución del tamaño de las nanopartículas y el potencial zeta se midieron en agua ultrapura utilizando el equipo Zetasizer Nano (Malvern Instruments, Reino Unido), con un ángulo de dispersión de 173° y equipado con un láser HeNe (633 nm) con una potencia de salida de 10 mW. Las distribuciones de tamaño se analizaron en un rango de concentraciones por encima de la concentración micelar crítica. Cada muestra se midió por triplicado. Caracterización de nanoestructuras mediante microscopía electrónica de transimisión Las muestras para microscopía electrónica de transmisión (TEM) se prepararon sobre rejillas de cobre recubiertas con una película de carbono y con soporte de malla 300 (C300Cu). En primer lugar, las rejillas fueron hidrofilizadas mediante un tratamiento con plasma de 20 segundos de duración, usando para ello el equipo PDC-002 (Harrick Plasma, EE. UU.) a baja potencia (7,2 W aplicado a la bobina de RF). Tras esto, situamos 15 μL de las distintas soluciones de ELR pre-incubadas a la temperatura y concentración deseada, de agua ultrapura y de una solución de acetato de uranilo (1% p/v) sobre una tira de Parafilm® colocada sobre una superficie de vidrio atemperada a la temperatura deseada. Las rejillas hidrofilizadas se depositaron sobre la gota de ELR durante 90 s, seguidamente sobre la de agua ultrapura durante 60 s, y finalmente, sobre solución de tinción negativa durante otros 60 s, empleando papel de filtro (Whatman® Gel Blot GB003) para eliminar el exceso de solución después de cada paso. Las imágenes se Resumen XV tomaron usando un microscopio Tecnai Thermionic T20 operado a 200kV (SAI, Universidad de Zaragoza, España). Caracterización de nanoestructuras mediante criomicroscopía electrónica de transimisión Las muestras para el análisis mediante criomicroscopía electrónica de transimisión (Cryo-TEM) se disolvieron a una concentración de 25 µM en agua ultrapura a 5 °C durante 24 h y se calentaron a 37 °C durante 30 min. Se prepararon muestras Cryo-TEM antes y después del calentamiento para evaluar los cambios en el ensamblado de las muestras respecto a la temperatura. La preparación y visualización de muestras Cryo-TEM se realizó en la "Plataforma de Microscopía Electrónica" (CICbioGUNE, Universidad del País Vasco, España). Se colocaron cuatro microlitros de la muestra en una rejilla de malla 300 con recubrimiento de carbono lacey (LC300-Cu; Microscopy Sciences) y se incubaron dentro de la cámara de un Vitrobot Mark III (FEI Inc., Holanda) a la temperatura deseada (4 o 37 °C) y a una humedad relativa cercana a la saturación (95% HR) durante 30 s. La mayor parte del líquido en la rejilla se eliminó mediante transferencia (3 s en un desplazamiento de -3 mm) y se vitrificó sumergiéndolo en etano líquido, previamente enfriado con nitrógeno líquido a aproximadamente -180 ºC. Las imágenes se tomaron a temperatura de nitrógeno líquido en el microscopio electrónico de transmisión JEM-2200FS / CR (JEOL Europe, Croissy-sur-Seine, Francia) operado a 200 kV. Un filtro de energía en columna (filtro Omega) permitió obtener imágenes con contraste y relación señal/ruido mejorado. Las imágenes digitales se grabaron en una cámara Ultrascan4000 ™ de 4K × 4K (Gatan, Inc.) utilizando el software DigitalMicrograph ™ (Gatan, Inc.). XXII Para la evaluación de los biofilm formados en el reactor por goteo (drip flow biofilm reactor, DFBR), se recogió la biomasa adherida a los discos testados mediante sonicación en 330 o 660 µL de solución salina (NaCl al 0.9%) atemperada a 4 °C, dependiendo de si se trataban de biofilms de monoespecies S. gordonii o microcosmos orales, respectivamente. Posteriormente, se incubaron 100 µL del extracto bacteriano durante 2 horas a 37 °C para reactivar la actividad metabólica de las bacterias, luego se midió el ATP utilizando el kit BacTiter-Glo™. La solución de bacterias se mezcló con 100 µL de la solución del kit en una placa blanca de 96 pocillos y se incubó durante 5 minutos en la oscuridad. La luminiscencia se midió posteriormente en un luminómetro de microplacas (BioTek, EE.UU.). Adicionalmente, los biofilms de S. gordonii fueron evaluados por conteo de UFC. Para ello, se realizaron diluciones seriadas en NaCl al 0.9% a 4 °C, y 10 μL de cada dilución se sembraron en placas de agar THB. Las placas de agar se incubaron a 37 °C durante 16 h, luego se contaron las UFC. En el caso del modelo de biofilm 769-NS basado en un microcosmos oral, al contener al menos 103 taxones bacterianos diferentes, el control de las condiciones de incubación en placas de agar para un recuento de UFC completo y de confianza no es posible debido a los diferentes requisitos nutricionales y de tiempo de incubación de las diferentes cepas. De forma alternativa, después de la recolección de muestras de biofilm mediante sonicación, se usaron 500 µL de extracto bacteriano para la extracción y purificación de ADN mediante el kit de Purificación de ADN MasterPure™ (Epicenter, Illumina, Inc., EE. UU.). La concentración de ADN se cuantificó utilizando un espectrofotómetro Nanodrop 2000c (ThermoFisher Scientific, EE. UU.). Resumen XXIII Ensayo de citocompatibilidad La citocompatibilidad de los recubrimientos se probó utilizando células de fibroblastos de prepucio humano (HFF-1), que se adquirieron de Life Technologies S.A. (Madrid, España) y fibroblastos gingivales primarios humanos (HGF; PCS-201-018, ATCC, EE.UU.). Se sembraron 5000 células HFF-1 por cm2 en medio DMEM (suplementado con 15% de FBS y 100 U mL-1 a 100 μg mL-1 de penicilina-estreptomicina) a 37 °C y 10% de CO2. Mientras que para la expansión de los fibroblastos primarios se sembraron 20,000 células cm-2 en medio basal de fibroblastos (PCS-201030, ATCC, EE.UU.) suplementado acorde a las indicaciones del distribuidor (PCS-201-041, ATCC, EE. UU.) sin añadir rojo de fenol ni antibióticos, a 37 °C, 5% de CO2. En el primer caso, se emplearon células HFF-1 en pasajes entre 5 y 8 en todos los experimentos. Y los niveles de citocompatibilidad se determinaron utilizando el reactivo de viabilidad celular AlamarBlue (AB) (ThermoFisher Scientific, EE.UU.). Siguiendo las instrucciones del fabricante, después de los tiempos de incubación deseados (5 h, 48 h y 7 días) las superficies se incubaron durante 4 h a 37 °C con una solución de AB al 10% en medio de cultivo y se registró la fluorescencia usando un Lector de microplacas SpectraMax M2e (Molecular Devices, EE. UU.). En el segundo caso, las muestras fueron incubadas a 37 °C durante 3 h para permitir la unión celular, y se añadió un exceso de medio para el cultivo adicional. Después de 1 y 3 días, las muestras se tiñeron con el kit LIVE/DEAD® (ThermoFisher Scientific, EE.UU.) y se visualizaron en un microscopio de fluorescencia invertido (Nikon Eclipse Ti-E acoplado a una cámara digital Nikon DS-2MBWc, Nikon Corporation, Japón) para confirmar la viabilidad celular en los diferentes discos. Las imágenes fueron XXIV adquiridas y procesadas utilizando el software NIS-Elements Advanced Research (versión 4.5, Nikon Corporation, Japón). Análisis estadístico El análisis estadístico ha sido realizado mediante un análisis de la varianza (ANOVA) de una vía con la con la prueba post-hoc de comparación múltiple Holm-Sidak usando el software Statgraphics XVII. Las diferencias significativas se exponen del siguiente modo: (*) p < 0.05; (**) p < 0.001; (***) p < 0.0001. Los datos presentados en esta tesis se presentan como la media ± desviación estándar (n ≥ 3, según lo especificado en cada apartado concreto). 3. Resultados El Capítulo 2 se centra en el estudio del papel que desempeña la distribución de cargas en el autoensamblado y separación de fases de polímeros proteicos intrínsecamente desordenados (intrinsically disordered protein polymers, IDPPs). Para ello, una librería de ELRs basada en dibloques anfifílicos fue diseñada y producida alternando la longitud de cadena y la distribución de cargas mediante técnicas de ADN recombinante. Tras esto, se procedió a la caracterización físico-química de la capacidad de formación de estructuras supramoleculares a distintas escalas de tamaño (nano-, mesoy microescala) en el rango de concentraciones entre 25 µM y 2.5 mM. El estudio de la nanoestructuración de los ELRs mediante DLS, CD, TEM y medición ZP, determinaron que la influencia de la distribución de la carga compromete la capacidad de autoensamblado. Al aumentar la densidad de carga en la corona pudimos acceder a nuevos diseños estructurales con la capacidad de ensamblarse en estructuras micelares a bajas concentraciones y de evolucionar a estructuras jerárquicas de mayor orden que el ensamblado micelar. Resumen XXV A altas concentraciones (2.5 mM) se evaluó la capacidad de formar hidrogeles. Sólo aquellas construcciones que mostraron una mayor inestabilidad micelar, agregándose en entramados de micelas a bajas concentraciones presentaron una transición sol-gel cuando se incubaron a 37 °C. Los hidrogeles físicos fueron caracterizados mediante un análisis reológico y mediante microscopía electrónica, observando que la distribución de cargas aparte de la capacidad de gelación también influye en las propiedades de los hidrogeles resultantes. La incorporación de repulsión electrostática como una variable en la organización supramolecular permite alcanzar ensamblados alternativos e impulsar la transición de fase sol-gel. En el Capítulo 3 se describe el desarrollo de un polipéptido híbrido basado en AMPs y ELRs. En este capítulo se investigó el uso de AMPs como dominios autoensamblables y la relación entre AMPs y ELRs en el proceso de autoensamblado con el objetivo de crear nanoreservorios para AMPs. Los ELRs híbridos producidos se basaron en un diseño de dibloque anfifílico, donde los péptidos de diseño GL13K y 1018 se incorporaron en el bloque hidrofílico. Tras la síntesis recombinante, se caracterizó su comportamiento térmico y su capacidad de autoensamblado mediante turbidimetría, DLS, TEM y cryo-TEM. La combinación de ambos dominios funcionales resultó en un polipéptido polivalente con capacidad de auto-ensamblado dual, dominado por la naturaleza del AMP. Los AMPs dirigieron la formación de nanofibras por debajo de la temperatura de transición del ELR que, además, retuvieron el comportamiento termosensible de estos. De este modo, aumentando la temperatura se pudo dirigir selectivamente un segundo ensamblado en agregados de fibras. Finalmente, en los Capítulos 4 y 5 se presenta la creación de recubrimientos covalentes basados en ELRs y AMPs para la prevención de la formación de biofilms sobre dispositivos biomédicos. Más XXVI concretamente, en el Capítulo 4 se describe la asociación sinérgica entre las propiedades anti-adhesivas de un ELR y la actividad antibiofilm de un AMP, en un polímero proteico híbrido. Se produjo un polipéptido basado en tres bloques funcionales (el AMP GL13K, un ELR policatiónico y un motivo basado en cisteínas para el anclaje selectivo sobre superficies). Este diseño modular permitió la inmovilización covalente sobre superficies modelo de oro, formando monocapas autoensambladas (self-assembled monolayers, SAMs) como demostraron los análisis mediante XPS, ángulo de contacto y QCM-D. La actividad antibiofilm de las SAMs se testó frente a dos cepas de estafilococo, S. aureus y S. epidermidis, las cuales son responsables de la mayoría de infecciones nosocomiales relacionadas con dispositivos implantables. Se incubaron las superficies modificadas en condiciones estáticas y medio TSB suplementado con glucosa para favorecer la formación de biofilm y se observaron mediante microscopía de fluorescencia y electrónica de barrido. De forma complementaria, se cuantificó la reducción de biofilm valorando la cantidad de biomasa remanente mediante tinción con cristal violeta y mediante un test de actividad metabólica cuantificando el ATP. De este modo, se observó que la combinación del AMP y del ELR en un diseño híbrido permite fabricar recubrimientos covalentes con una actividad sinérgica incrementando el efecto antibiofilm del AMP. Además, los recubrimientos demostraron una excelente citocompatibilidad frente a células humanas. Siguiendo con esta línea de investigación, en el Capítulo 5 se explora la aplicación de recubrimientos basados en AMP-ELRs para implantes de titanio. Además del polipéptido híbrido recombinante empleado en el Capítulo 4, un segundo AMP-ELR fue desarrollado mediante derivación química, uniendo enantiómeros D del péptido GL13K al ELR. Discos de titanio comercial (grado II) fueron recubiertos covalentemente empleando Resumen XXVII organosilanos. La efectividad de la funcionalización y su resistencia se testó mediante XPS y midiendo el ángulo de contacto dinámico. Tras esto, se procedió a evaluar su actividad antibiofilm frente a dos modelos de biofilms orales (Streptococcus gordonii y un modelo de microcosmo oral) en condiciones dinámicas en un reactor de biofilm por goteo. Los ensayos antibiofilm y de citocompatibilidad mostraron la efectividad del estereoisómero D del AMP como agente antibiofilm y bactericida cuando se encuentra unido al ELR sin efectos secundarios tóxicos frente a células humanas. En resumen, esta Tesis aporta nuevas perspectivas al diseño y fabricación de nanosistemas autoensamblados basados en polímeros proteicos diseñados genéticamente. De manera específica, se describe la caracterización físico-química y biológica de múltiples ELRs e híbridos AMP-ELRs y confirma su potencial como plataformas multivalentes para el estudio de procesos de ensamblado de proteínas y para el desarrollo de biomateriales avanzados con propiedades antimicrobianas. 4. Conclusiones 4.1. Ingeniería genética, biosíntesis y caracterización de ELRs En esta Tesis, se ha demostrado con éxito el desarrollo de múltiples biomateriales basados en ELRs con diseño modular. Gracias a su naturaleza recombinante, se ha llevado a cabo la producción de ELRs introduciendo diferentes bloques funcionales. En el Capítulo 2, se diseñó y biosintetizó una biblioteca de ELRs con diseño dibloque (ELdcRs) anfifílicos con el fin de estudiar la influencia de la distribución de carga en la organización supramolecular de IDPPs. En los capítulos 3 y 4, se diseñó un sistema de producción recombinante para polipéptidos híbridos basados en AMPs y ELRs. La incorporación de un bloque protector en el extremo N-terminal XXVIII permitió la sobreexpresión de AMP-ELR híbridos evitando efectos tóxicos no deseados para el microorganismo huésped. Además, su diseño funcional permitió su posterior liberación sin introducir residuos aminoacídicos adicionales en el dominio AMP, que pudieran interferir en sus propiedades biológicas. La construcción de los distintos genes codificantes de los polipéptidos se evaluó mediante electroforesis en gel de agarosa y secuenciación de ADN. Los ELRs o AMP-ELRs se hiperexpresaron adicionalmente en Escherichia coli como proteínas heterólogas obteniendo rendimientos entre 220 y 600 mg por L de medio de cultivo. Además, su purificación mediante ciclos de transición inversa (ITC) permitió obtener productos altamente puros y monodispersos como se comprobó mediante técnicas de caracterización fisicoquímica como SDS-PAGE, MALDI-TOF y HPLC. Finalmente, en el Capítulo 5, demostramos que la versatilidad de los ELRs también puede explotarse para su conjugación química con péptidos bioactivos y así, explorar nuevos diseños que no pueden lograrse mediante síntesis recombinante. De esta manera, el enantiómero D de un AMP se bioconjugó a un ELR mediante química ‘clic’ (SPCC). 4.2. Distribución de carga como modulador molecular de la nanoestructuración en ELR Se ha demostrado que la repulsión electrostática juega un papel clave en el ensamblado supramolecular de polímeros proteicos intrínsicamente desordenados, utilizando ELdcRs como modelos moleculares La caracterización mediante TEM y DLS determinó que al aumentar la densidad de carga en dibloques IDPPs anfifílicos se contribuye a estabilizar el ensamblado micelar a bajas concentraciones, incluso en Resumen XXIX diseños altamente desproporcionados (ratio de longitud entre bloques hidrófílico:hidrofóbico <0.3). De este modo, la presencia de residuos cargados en el bloque hidrofílico permite romper los límites preestablecidos para el diseño de nanoestructuras basadas en ELdcRs. En contraposición, a altas concentraciones, la presencia de residuos cargados impulsó la desestabilización de la micela favoreciendo la formación de estructuras de orden superior a concentraciones superiores (a partir de 125 µM), como agregados de micelas. Además de contribuir a la transición de fase sol-gel y la consiguiente formación de geles físicos a 2.5 mM. En resumen, la distribución de carga parece ser un importante modulador del autoensamblado a diferentes escalas de longitud, afectando a la formación de estructuras jerárquicas y a la separación de fases permitiendo alcanzar arquitecturas moleculares alternativas con gran interés para la fabricación de biomateriales aplicables a la ingeniería de tejidos o administración de fármacos. Además, este trabajo no sólo contribuye a sentar las bases para el diseño racional de dispositivos jerárquicamente autoensamblados basados en ELdcRs, sino que también contribuye a arrojar luz sobre la compleja relación estructura-función de las IDPs y los parámetros que contribuyen a su transición de fase y, por lo tanto, a la formación de condensados biomoleculares u orgánulos sin membrana. 4.3. Estudio del autoensamblado en polipéptidos híbridos diseñados genéticamente: interacción entre AMPs y ELRs Se demostró que los AMPs pueden ser empleados como dominios de autoensamblado para cadenas polipéptidicas de mayor tamaño mediante construcciones híbridas basadas en ELRs con diseño modular. XXX La evaluación del comportamiento térmico sugirió que la tendencia al ensamblado de los AMPs contribuye a la transición de fase de los ELRs, disminuyendo la Tt y aumentando la histéresis térmica. Los AMPs indujeron la formación de en nanofibras por debajo de la Tt, mientras que condujeron a la formación de agregados fibrilares por encima de la Tt. En función de la naturaleza del AMP, el mecanismo de nanoestructuración fue diferente, formando nanofibras de tamaños variables. Además, los nanosistemas híbridos conservaban termosensibilidad de los ELRs, demostrando de este modo, la fabricación de nanosistemas capaces de ensamblarse de manera dual. La interacción dinámica entre los AMPs y los ELRs en la formación de nanoestructuras con capacidad de estímulos abre un abanico de oportunidades que van desde la producción escalable de AMPs hasta el diseño de nanomateriales avanzados que recapitulen las propiedades de los AMPs y de los ELRs. 4.4. Desarrollo de monocapas autoensambladas con actividad antibiofilm basadas en AMPs y ELRs Hemos demostrado el anclaje covalente y de forma selectiva de ELRs y AMP-ELRs en superficies modelo de oro. La incorporación de un motivo basado en Cys en el extremo carboxilo terminal permitió la formación de monocapas autoensambladas. La inmovilización de los ELRs proporcionó un efecto de antiadherente que permitió reducir la formación de biofilms frente a dos cepas estafilocócicas de relevancia médica: S. aureus y S. epidermidis. Además, cuando el híbrido AMP-ELR fue inmovilizado, la actividad antiadherente del ELR y el efecto antimicrobiano del AMP convergieron sinérgicamente, mejorando el potencial de antibiofilm del recubrimiento. Resumen XXXI Los ensayos de viabilidad celular mostraron una excelente citocompatibilidad de los recubrimientos, destacando su potencial como recubrimientos seguros para evitar la formación de biofilms en dispositivos biomédicos. 4.5. Biofabricación de recubrimientos antibiofilm para prevenir infecciones asociadas a implante Se ha demostrado con éxito la biofabricación de recubrimientos basados en AMP-ELRs sobre superficies de titanio de grado médico. El motivo basado en Cys sirvió para inmovilizar los polipéptidos híbridos covalentemente de manera selectiva mediante organosilanos, permitiendo la flexibilidad de las moléculas en el recubrimiento. Los recubrimientos híbridos que contenían el enantiómero D del péptido GL13K fueron capaces de retener el potencial de antibiofilm del AMP frente dos modelos de biofilm oral en condiciones dinámicas. Los ensayos de citocompatibilidad demostraron que la toxicidad de los recubrimientos frente a bacterias no compromete la proliferación de células humanas primarias, confirmando, la fiabilidad de este método. La inmovilización de AMPs usando ELRs como plataformas multivalentes que pudieran incorporar biofuncionalidades adicionales y mejorar la respuesta biológica de los recubrimientos de antibiofilm. 6 Abstract 7 Abstract more thorough understanding of Nature and its building blocks feeds our ability to develop functional biomaterials and vice versa. Understanding the functions of proteins, the most abundant and diverse macromolecules of a cell, has been crucial for the development of advanced biomaterials for biomedical applications. In this sense, the advent of recombinant DNA technology provided a sustainable production method for of natural proteins, but also the engineering of new proteinbased materials, including recombinant protein polymers. Recombinant protein polymers, or recombinamers, are genetically engineered polypeptides based on conserved motifs found in structural proteins. Within this type of biomaterials, we can find elastin-like recombinamers (ELRs). ELRs are polypeptides based on the repetition of the pentapeptide Val-Pro-Gly-X-Gly, found in the hydrophobic domains of tropoelastin that allowed them to mimic its structural and biomechanical properties. Therefore, ELRs are characterized by intrinsic structural disorder, reversible phase transition behavior and excellent biological and mechanical properties. These properties in addition to low-complexity polymeric composition and outstanding tunability, make them an excellent candidate for the study of biological processes and the generation of tailored materials for biomedical applications. Thus, this Thesis is focused on the design, production and processing ELRs, targeting its biomedical application as self-assembled nanosystems and antimicrobial coatings. To this end, different modular designs will be studied using ELRs and antimicrobial peptides (AMPs) as building blocks. AMPs are short cationic peptides with the ability to kill microbes directly or indirectly modulating the host immune system and hence, they constitute one of the most A 8 promising alternative to conventional antibiotics in the treatment of drugresistant infections. A comprehensive review on biomedical applications of AMPs and ELRs is presented in Chapter 1. A great variety of options enables the processing of these biomaterials into gels, coatings or nanostructures for any of a number of biomedical applications, such as extracellular-matrix-mimicking 3D networks for tissue engineering or selfassembled nanocarriers for drug delivery. In particular, Chapter 2 is focused on the study of the role of charge distribution in self-assembly and phase separation of intrinsically disordered protein polymers (IDPPs). For this purpose, a library of ELRs based on amphiphilic diblocks was designed altering chain length and charge density. Physico-chemical characterization of the nanostructuration in a range of concentrations determined that charge distribution strongly affects selfassembly of IDPPs at different scale lengths. The incorporation of electrostatic repulsion as a variable in the supramolecular organization of ELRs enabled to reach alternative assemblies and to drive liquid-gel phase transition. Chapter 3 describes the development of a hybrid polypeptide based on AMPs and ELRs. This chapter seeks to explore the usage of AMPs as self-assembling domains (SADs) and to elucidate the interplay between AMPs and ELRs in the self-assembly with the further objective of creating nanoreservoirs for AMPs. The hybrid design consists in an amphiphilic diblock, where the designer peptides GL13K and 1018 were fused to the hydrophilic block. After recombinant biosynthesis, thermal behavior and self-assembly abilities were characterized. The combination of the two functional domains resulted in a multifaceted polypeptide with a dual selfassembly governed by the composition of the AMP-domain. The AMPs triggered the formation of nanofibers that retained the characteristic Abstract 9 thermo-responsiveness of the ELRs. Therefore, increasing the temperature, the hierarchical assembly into fibrillar aggregates could be driven. The interplay of the different SADs (ELR or AMP) offers opportunities for the development of new nanocarriers to deliver AMPs and to study the molecular mechanism that control the bactericidal properties of the AMP linked to biopolymers. Finally, Chapters 4 and 5 present the creation of covalent coatings based on AMPs and ELRs for the prevention of biofilm formation onto biomedical devices. Specifically, Chapter 4 describes the synergistic association of a low-fouling ELR and the antibiofilm designer peptide GL13K. For this purpose, a hybrid polypeptide based on three functional blocks (GL13K, polycationic ELR and a Cys-motif for the selective tethering onto surfaces) was bioproduced. The modular design enabled the covalent immobilization onto model gold surfaces, forming self-assembled monolayers (SAMs). Antibiofilm activity of the SAMs was tested against two staphylococcal strains of medical relevance in the development of nosocomial infections. The combination of the AMP and ELR in the hybrid design (AMP-ELR-coating) provided a synergistic activity increasing the antibiofilm effect of AMP-coatings under static conditions. Additionally, the nanocoatings demonstrated an excellent cytocompatibility against human cells. In the same vein, Chapter 5 explores the implementation of AMPELR for titanium implants. In addition to the recombinant hybrid polypeptide employed in Chapter 4, a second AMP-ELR was developed by chemical derivation attaching the D-enantiomer of the GL13K to the ELR backbone. Commercially pure (grade II) titanium discs were coated using organosilanes as covalent linkers. Then, antibiofilm activity was tested against two different oral biofilm models (monospecies Streptococcus gordonii 10 and an oral microcosm model) under dynamic conditions in a drip flow biofilm reactor (DFBR). Antibiofilm and cytocompatibility assays showed the strong antibiofilm and bactericidal effect of the D-enantiomer when is linked to an ELR with non-toxic side effects against human cells. In summary, this Thesis provides new insights on the design and fabrication of self-assembled nanosystems based on genetically engineered polymers. Specifically, it describes the physico-chemical and biological characterization of several novel ELRs and hybrid AMP-ELRs and confirms their potential as multivalent platforms for the study of protein self-assembly processes and for the development of advanced biomaterials with antimicrobial properties. Abstract 11 Chapter I: General introduction Chapter I General introduction General introduction 15 I 1.1. Recombinant protein polymers in biomaterials science: Elastin-like recombinamers n the past few decades, the combination of materials science, cellular biology, molecular biology and biochemistry has given rise to a new generation of biomaterials: genetically engineered biomaterials. Genetically engineered or recombinant biomaterials are protein-based biomaterials that aim to mimic the outstanding biological and mechanical properties of native proteins in vivo, and at the same time overcome some limitations of natural and synthetic biomaterials. In this sense, the application of recombinant DNA technology to biomaterials science provides a scalable, cost-effective and environmentally friendly source for protein materials with a precise control over the composition, molecular weight and monodispersity. Genetically engineered biomaterials, specifically recombinant protein polymers, are attractive for biotechnological and medical approaches because of their intrinsic biocompatibility and the extreme versatility. [15] Their recombinant nature enables to create modular building blocks with simplified repetitive sequences that recreate the properties of more complex proteins while simultaneously incorporating diverse structural and biological motifs. Thus, producing multivalent platforms, we can design tailored biomaterials with specific features for the desired application. Herein, we will focus on the polymeric derivatives of the elastin, the elastinlike recombinamers (ELRs) or polypeptides (ELPs). Elastin is a polymeric extracellular matrix (ECM) protein, found predominantly in tissues as varied as the skin, lungs, blood vessels, and I Chapter I 22 colleagues were the first to report the production of self-assembled nanoparticles based on elastin-like block co-recombinamers (ELbcRs).[54] They designed amphiphilic diblock co-recombinamers based on hydrophobic and hydrophilic individual blocks, with Tt below and above physiological temperature, respectively. Thus, at 37 °C, the hydrophobic block collapses triggering the formation of hydrophobic cores surrounded by a hydrophilic corona, made of hydrophilic blocks. If the temperature was increased above the Tt of the hydrophilic block, the corona would collapse forming macromolecular aggregates and undergoing phase separation (Figure 1.4). This means that the ELbcR can exist in at least three different phases: hydrated molecules in solution, nanoparticle assembly or micrometric aggregates. Due to the potential use of self-assembled structures as nanocarriers, self-assembly of diblock co-recombinamers has been exhaustively studied experimentally and theoretically.[55] First, evaluating self-assembling behavior of non-charged amphiphilic diblocks, Dreher et al. defined two physico-chemical requirements for the micellar assembly: significant difference must be between Tt of the hydrophilic and hydrophobic block, and a molecular weight ratio of 1:2 to 2:1 between both blocks is needed.[56] Then, after analyzing phase transition by spin probing continuous wave electron paramagnetic resonance (CW EPR) spectroscopy, Widder et al. specified that hydrophilic : hydrophobic weight ratio must be greater than or equal to 0.3 to guarantee micellar assembly.[57] Lastly, a minimum length of the hydrophobic block is also needed to produce stable nanoparticles (greater than 48 pentapeptides).[58] General introduction 23 I Figure 1.4. Schematic representation of the micelle formation and phase separation driven by an amphiphilic ELR with diblock design. Tt1 correspond with the Tt of the hydrophobic block, whereas Tt2 is the Tt of the hydrophilic one. On the other hand, we are able to control relevant parameters for the disposition of the nanostructures in solution.[59,60] Size and shape of the micelles can be modulated by tuning the size of the individual blocks in the diblock design.[58] It has been observed that size and shape can be modulated in diblock co-recombinamers introducing gradients of alternated hydrophilic and hydrophobic blocks.[61] As a function of the length of the individual blocks in the interfacial gradient, we can control the molecular weight, aggregation number, or the hydrodynamic diameter of the ELbcR-micelles, crucial parameters for the production of delivery nanocarriers. Furthermore, other supramolecular assemblies can be reached, including vesicles,[62,63] cylindrical micelles or nanofibers.[64–68] The exquisite versatility of ELRs enables the ability to introduce other protein selfassembly domains or bioactive molecules, recombinantly or chemically, by modification of their amino acid side-groups. Thus, exploiting biocompatibility[69] and biodegradability of ELRs,[70] tailored stimuliresponsive nanoparticles have been developed for different therapies. Biofabrication of ELR nanoparticles for cancer therapy has enabled to improve the pharmacokinetics and tumor accumulation of hydrophobic drugs,[56,71–73] tumor targeting,[74,75] decrease cancer drug resistance,[76,77] or Chapter I 24 increase cellular uptake,[78–80] in addition to develop emerging therapies such as local delivery of radioisotopes for brachytherapy[53] or immunotherapy.[68,81] Furthermore, fusion of bioactive peptides to the hydrophilic corona enables the production of self-assembled nanocarriers for other applications. It has been demonstrated that ELR nanocarriers can improve stability and half-life of small peptides and their effective delivery, including peptide drugs,[82] growth factors[83,84] or bacterial antigens.[85] Hydrogels Supramolecular assemblies based on ELRs can be extended to the micro and macroscale for the development of three dimensional networks that simulate the native ECM for their use in tissue engineering. ELRs with modular designs enabled to obtain tailored multifaceted scaffolds with tunable architecture, biofunctionalities, mechanical properties and biodegradability, key parameters that strongly affect cell behavior. [86–88] Thus, matrix-cell communication can be modulated through mechanotransduction, [89,90] exposition of cell-binding domains [91–93] or growth factors, [94–96] or by the matrix remodeling, [97–100] which results in the control of cell behavior, including adhesion, migration, proliferation and stem cell fate. Furthermore, other biomaterials can be integrated with ELRs to produce hybrid or composite scaffolds with improved features. Thus, ELRs have been combined with protein materials (e.g. collagen,[101–104] fibrin,[105] peptide amphiphiles[106]) polysaccharides (e.g. hyaluronic acid,[107–109] chitosan[110]) or synthetic polymers (e.g. polyethylene glycol[111,112]). On the other hand, ELRs are also compatible with multiple manufacturing technologies in addition to the great variety of crosslinking General introduction 25 I strategies, based on physical[113–116] or chemical crosslinks.[5,99,117–120] Scaffold fabrication techniques such as electrospinning or salt leaching can be used for the production of ELR scaffolds with precise architectures.[121–125] Coatings Biofunctionalization of surfaces is an effective approach to improve biological response of indwelling devices. In this sense, properties of the ELRs have been exploited for the improvement of clinical materials. Intrinsic properties of the ELRs, such as thermo-responsiveness[126] or biocompatibility,[127] can be transferred to the coated biomaterial by the functionalization of these recombinant polypeptides. ELR coatings were also applied to vascular grafts, demonstrating their intrinsic ability to enhance the hemocompatibility and inhibit thrombosis related complications in vivo.[128,129] In contrast to tropoelastin, cells do not recognize the polyVPGXG and consequently, ELRs show an intrinsic low-fouling behavior that prevents the unspecific protein attachment and cell binding.[130,131] However, if cellbinding motifs, such as the integrin-mediated adhesion tripeptide RGD, is incorporated into the ELR backbone, cell adhesion and proliferation are greatly increased.[132,133] The control of cell-binding motifs exposure by the thermoresponsive behavior of the ELR coating enables to harvest cell sheets with potential application in tissue regeneration.[134] Furthermore, as different cell types recognize preferentially different cell-binding motifs, the regulation of these bioactive cues density and spatial location can be used to produce cell patterns.[135,136] Besides binding motifs, other functional domains can be incorporated onto the ELR coatings such as different bioactive peptides[137] and crosslinking sites,[138] providing Chapter I 26 innovative surface engineering approaches to develop tailored bioactive substrates for multiple biomedical applications. In conclusion, ELRs are attractive biomaterials to create nanostructured devices for tissue engineering and drug delivery. Their inherent reversible phase transition behavior, biocompatibility and feasibility to introduce functionalities, chemically or by co-expression, make them excellent candidates for multiple biomedical applications. General introduction 27 I 1.2. Antimicrobial biomaterials: Antimicrobial peptides In recent years, antimicrobial resistance (AMR) to conventional antibiotics has become an uncontrollable health problem. Last report from the World Health Organization to the Secretary-General of the United Nations calls for the global cooperation in the development of alternatives to treat drug-resistant infections, a threat that is estimated to provoke 10 million annual deaths and USD 2.9 trillion expected cumulative loses in the Organization for Economic Cooperation and Development (OECD) countries, in 2050.[139] Overuse and overdependence of antimicrobials (antibiotics, antifungals and antivirals), heavy metals and biocides has driven to AMR in the environment.[140] Against this, health organizations must establish strict policies to control the usage of this compounds in health services, industry, agriculture and animal husbandry, but also it is a crucial need to develop new drugs to treat drug-resistant infections. In this sense, antimicrobial peptides (AMP) are a promising alternative to conventional antibiotics. Antimicrobial peptides (AMPs) are short (10-50 amino acids), amphiphilic, generally cationic peptides produced by a wide variety of life forms, including invertebrate, plant and animal species.[141] Reflecting their wide diversity of sources, AMPs have a wide variety of conformational structures[142] and they are commonly classified according to the structure adopted in solution. Thus, we can distinguish 3 extensive groups: α-helical peptides, β-sheet peptides and extended or random-coil peptides.[143] (i) αhelical family consists of AMPs with helical structures (e.g. magainins and LL-37). (ii) β-sheet family is composed by AMPs with β-sheets usually stabilized by disulfide bonds, forming rigid structures (e.g. human α and βdefensins). (iii) Extended/random-coil family contains AMPs with Chapter I 28 undefined secondary structures or random coils which are often rich in specific amino acids such as tryptophan and arginine (e.g. indolicidin) or histidines (e.g. histatin). Their amphipathicity allows them to acquire different conformations and interact with multiple molecular targets.[144] The modes of action are wide and varied, specific to each AMP. Traditionally, they were considered as membrane pore formers because of the high affinity with bacterial membranes and consequently their disruption,[144] but also AMPs are able to target intracellular components, inhibiting metabolic processes, such as nucleic acid synthesis, cell wall biosynthesis or protein folding, among others[145][146] and impeding biofilm formation.[147,148] In this sense, AMPs play an important role as part of the innate immune response, showing broad-spectrum activity against most bacterial pathogens, antiparasitic,[149] antiviral,[150] antifungal activities[151] and even against some kinds of cancer.[152] Additionally, recent studies have demonstrated their ability to interact and modulate the host response, defining them as host-defense peptides (HDPs) due to their multifaceted roles[153] (Figure 1.5). There exist thousands of different AMPs, data repository of antimicrobial peptides (DRAMP) database contains 19,899 entries of natural and designer AMPs (October 2019), 5,084 of which are general AMPs (containing natural and synthetic AMPs) and 14,739 patented. [154] However, currently there are only 76 AMPs in preclinical or clinical stages. [154] Despite the promising properties of natural AMPs, they also show potential limitations that hinder their clinical use and commercialization. [142] First, AMPs are highly sensitive to environmental conditions. The presence of salts in the medium or pH changes can compromise their antimicrobial potential. [155–157] Second, AMPs can be rapidly inactivated by proteolysis or by adsorption on plasma proteins, showing short plasma half-life. Third, General introduction 29 I some of them present cytotoxic or hemolytic associated activities. Fourth, despite the development of resistance occurs in a much lesser degree than conventional antibiotics, pathogens are also able to develop mechanisms of resistance against AMPs. Last, the cost of chemical synthesis hampers the scale-up of their production. Figure 1.5. Schematic representation of the multiple bioactivities developed by AMPs and potential applications. Adapted with permission from reference [147]. Fortunately, several strategies have been developed to overcome these limitations. The increasing number of structure-activity relationship studies along with the application of computational modelling enable to shed light into the biological function of the distinct residues in order to produce simplified AMPs with improved features, including selectivity, high potency or the ability to penetrate mammal cells to combat intracellular infections.[147,158] Moreover, the use of D-enantiomeric forms,[159] β-amino acids[160] and other non-canonical amino acids allow not only to improve bactericidal properties but also to decrease proteolytic degradation and evade resistance mechanisms.[161,162] Chapter I 30 Finally, one of the main drawbacks that limit the widespread clinical use is their chemical synthesis. In this sense, multiple recombinant approaches have been developed as scalable and cost-effective production using distinct host organisms. E. coli is the most commonly used due to their fast and low-cost growth and also because is a well-established expression system.[163,164] However, other alternatives have been used such as eukaryotic cells (e.g. fungi, plants or mammal cells).[165–167] 1.2.1. Therapeutic applications of antimicrobial peptides AMPs constitute a promising alternative to conventional antibiotics for the treatment of drug-resistant infections due to their antimicrobial and immunomodulatory properties. Despite the aforementioned limitations, their potent bactericidal activity in addition to their low-frequency of resistance emergency [168] can be efficiently used for the treatment of certain infections. One avenue that is being explored is to use them as topical agents. In humans, AMPs are produced by different epithelial cells (e.g. keratinocytes, conjunctival epithelial cells) and they act as multifunctional modulators of wound healing.[169] Several studies revealed that they are upregulated in multiple stages.[170] Apart from the inherent antimicrobial activity, they play important roles in chemotaxis and angiogenesis, particularly important in the regeneration of skin.[171] In fact, their ability to accelerate healing of chronic wounds has been demonstrated, acting as a damper of inflammation[172] and enhancing regeneration and wound closure.[173] In this sense, AMPs have been used for the treatment of infected wounds,[174][170] specifically for the treatment of drug-resistant pathogens,[175] such as resistant forms of Staphylococcus aureus, the most frequently isolated pathogen from human skin infections.[176] General introduction 31 I Furthermore, AMPs have demonstrated that they can show synergistic effects with other antibiotic molecules, including other AMPs,[177,178] conventional antibiotics[179–181] or metallic nanoparticles.[182,183] The wide variety of AMPs and their multiple potential molecular targets opens up possibilities for the design of combined therapies for the treatment of drugresistant infections, increasing efficacy and preventing side-effects and resistance. Thus, synergetic activity can be reached by co-delivery or through chemical conjugation. Chemical conjugation Chemical conjugation of AMPs consists in the covalent bonding of AMPs with other molecules in order to improve their antimicrobial properties or delivery (see for instance [184–186]). AMPs can be functionalized to generate more stable peptides (e.g. amidation of the terminal carboxyl group), but also to incorporate additional properties. For example, several studies have demonstrated that coupling AMPs with conventional antibiotics increases the bactericidal activity against resistant strains, such as vancomycin resistant Enterococci.[187,188] To enhance membrane permeability, AMPs have been also combined with lipid chains and cell-penetrating peptides, which resulted in an increase of the selectivity and toxicity.[189–191] AMPs can be conjugated with other biomaterials in order to improve their delivery. Morris et al. demonstrated that PEGylation minimizes cytotoxicity of a synthetic AMP.[192] They created hybrid AMPs by Nterminal PEGylation that were tested successfully for the treatment of pulmonary infections. Furthermore, conjugation of AMPs with nanoparticles has been explored. Coating AMPs onto the surface of nanoparticles aims to increase the local concentration and bioavailability in Chapter II: Charge distribution as molecular modulator of nanostructuration in ELRs Chapter II Charge distribution as a molecular modulator of the nanostructuration in ELRs Charge distribution as molecular modulator of the nanostructuration of ELRs 41 II 2.1. Introduction rotein-based polymers, and their recombinant versions (recombinamers), are a group of materials typically inspired by the repetition and combination of conserved peptide motifs present in structural proteins such as resilin, collagen or elastin, amongst others. [231] Consequently, their unique physical, chemical and biological properties have enabled various applications in molecular biology, drug delivery or regenerative medicine.[15,16] In this context, elastin-like recombinamers (ELRs) or polypeptides (ELPs) have to be highlighted. Recombinant polymers that are typically based on the repetition of a tropoelastin-derived pentapeptide Val-Pro-Gly-Xaa-Gly,[232] which provides extraordinary elastic properties and a thermo-responsive phase-transition behavior mainly characterized by a lower critical solution temperature (LCST) phase behavior driven by the reversible formation of molecular aggregates above a certain temperature.[233] Additionally, due to their low complexity, their intrinsic disordered nature and their recombinant production, ELRs can be described to be intrinsically disordered protein polymers (IDPPs).[234] Thus, ELRs constitute a powerful model for understanding the structural properties that drive the phase separation of complex intrinsically disordered protein regions (IDPRs).[27,235] IDPRs are protein sequences that, despite the lack of a defined 3D structure carry out vital biological functions, such as cell communication, gene regulation or cell cycle control.[25,236,237] Moreover, IDPRs can fold into different dynamic conformations and undergo phase separation depending on the presence of binding partners or the environmental conditions, which may lead to the formation of so-called biomolecular condensates.[28,238] Biomolecular condensates, are defined as dynamic, protein-rich, supramolecular assemblies that provide microenvironments which are P Chapter II 42 chemically different from the surrounding medium.[238] Phase transition, and thus biomolecular condensate formation, is driven by low complexity proteins that can assemble into liquid, gel or solid-like condensates stabilized by weak and multivalent interactions.[239–241] Appropriate assembly between these IDPRs predetermines their correct functionality. Thus, an understanding of the complex physicochemical forces that govern the formation and organization of these biomolecular condensates is critical for understanding their function and to derive implications for cellular processes and IDPR-associated diseases.[238,242] In this sense, IDPPs, especially ELRs are a powerful tool for studying the parameters that modulate formation of these protein-rich condensates. Thus, biomolecular condensates based on engineered protein and ribonucleoprotein systems have been produced in order to provide interesting insights into the dynamic intermolecular interactions of ELRs during coacervate formation.[243,244] Charge interactions are one of the parameters that mediate folding of IDPRs.[245–247] However, this parameter is still poorly understood. As such, herein, we aim to evaluate the influence of charge distribution and density on the phase separation of model IDPPs (ELRs) with an amphiphilic diblock design, from the solute state to hydrophobically crosslinked hydrogels, via nanostructured assemblies. The self-assembly and structural parameters of elastin-like diblock co-recombinamers (ELdcRs) that contribute to the nanostructuration have been extensively studied due to their potential application of these systems as nanocarriers for drug delivery or nanovaccine production.[54–57,248,249] Although charge density has not yet been considered, charged ELdcR designs have been employed for the development of self-assembled nanocarriers and as fusion proteins. [54,85,250,251] Most recently, Schiller and coworkers have demonstrated the Charge distribution as molecular modulator of the nanostructuration of ELRs 43 II ability of these systems to form subcellular compartmentalization in vivo and to produce dynamic protein membranes, thus directly linking them with biomolecular condensates formation. [252–254] In this study, the length and charge density of amphiphilic diblocks designs were varied to produce charged amphiphiles with unbalanced length ratios between the hydrophobic and hydrophilic blocks (charged unbalanced ELdcRs, cuELdcRs). Circular dichroism (CD) spectroscopy, dynamic light scattering (DLS), transmission and scanning electron microscopy (TEM and SEM) and rheological characterization were performed in order to obtain a comprehensive overview of hierarchical selfassembly of the cuELdcRs on the nanoand microscale. On the basis of the results, charge repulsion is seen to be a critical parameter for nanostructure complexity and consequently, contributes to hierarchical assembly into higher-order structures, such as physically crosslinked gels. 2.2. Results and Discussion In order to study the influence of charge on self-assembly, an ELdcR library was designed based on the sequence of a diblock ELR developed previously in our group.[85] The original diblock design, referred to as E-I, was based on an anionic hydrophilic block E [(VPGXG)50; X=V/E in a 4:1 ratio] and a hydrophobic block I [(XPGVG)60, X=I], with an LCST above and below physiological temperature, respectively. This design enables the formation of self-assembled micelles or physical hydrogels depending on the concentration under physiological conditions.[250] In order to study the influence of charge density and size on the corona during the self-assembly, the length and composition of the hydrophilic block was varied; the hydrophobic block was the same for all the constructs. Thus, four new ELdcRs were recombinantly produced as heterologous proteins in E. coli: Chapter II 44 (i) an uncharged diblock with the same amino acid length for the E-I but based on uncharged L-serine as polar amino acid (S-I), (ii) a diblock in which the length of the charged hydrophilic block E was decreased by half (E1/2-I), (iii) a diblock in which only the glutamic acid pentapeptides of the original hydrophilic E-block were mantained (oE-I) and (iv) a related system in which these pentapeptides were decreased by half (oE1/2-I). A schematic representation of the ELdcR library designs, hydropathy plots and complete amino acid sequences and size ratios of the ELdcRs can be found in Figure 2.1a and Table 2.1, respectively. Condensation of the charged pentapeptides in the oE-blocks may have an impact on the hydropathy of the block, as shown the Kyte-Doolitle plots (Figure 2.1b). The Kyte Doolittle algorithm predicts the hydrophilic and hydrophobic tendencies of a polypeptide chain by the progressive evaluation (from the N-terminus to the C-terminus) of the average hydropathy following the Kyte-Doolittle scale, where the larger the number is, the more hydrophobic the amino acid. The most hydrophilic amino acids are arginine (-4.5) and lysine (-3.9), whereas the most hydrophobic ones are isoleucine (4.5) and valine (4.2). Purification by inverse transition cycling (ITC) enabled us to obtain highly pure and monodisperse products with extreme control of the sequence,[58] as confirmed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) (Figure 2.1c), by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) (Figure 2.1-2.5 and Table 2.2) and by high performance liquid chromatography (HPLC) (Table A2.1). Charge distribution as molecular modulator of the nanostructuration of ELRs 45 II Figure 2.1. (a) Schematic representation of the ELdcR sequences (b) and their hydropathy plot calculated using the Kyte-Doolittle scale. (c) CuCl2-stained SDS-PAGE gel of the purified ELdcRs. The first lane from left to right is the Pierce™ Unstained Protein MW Marker (ThermoFisher). Lanes 2 to 6 are the different ELdcRs. Self-assembly was studied at different scales in aqueous solution. First, ELdcR behavior was evaluated on the nanoscale to assess the influence of charge density and the unbalanced hydrophilic/hydrophobic block length ratio on nanostructure formation. CD analyses confirmed the disordered secondary structure of the diblocks (Figure 2.2a and 2.2b). Consistent with previous studies, differences in the composition of the hydrophilic blocks barely affected the secondary structure of the ELdcRs. [58] Indeed the Amino acid Position 100 200 300 Score -1 0 1 2 66.2 45.0 35.0 25.0 kDa (a) (c) VPGEG VPGSG IPGVG [(VPGVG)2VPGEG (VPGVG)2]5 (IPGVG)60 (VPGEG)10 [(VPGVG)2VPGEG (VPGVG)2]10 (IPGVG)60 (IPGVG)60 (IPGVG)60 (VPGEG)5 E-I E1/2-I oE-I oE1/2-I (VPGSG)50 (IPGVG)60 VPGVG S-I Amino acid Position 100 200 300 400 500 Score -1 0 1 2 Amino acid Position 100 200 300 400 500 Score -1 0 1 2 Amino acid Position 100 200 300 400 Score -1 0 1 2 Amino acid Position 100 200 300 Score -1 0 1 2 E-I E1/2-I oE-I S-I oE1/2-I (b) Chapter II 46 characteristic negative peak for IDPs at around 197 nm appeared in all the CD spectra as did the characteristic maximum at 210 nm indicating the presence of β-turns in ELRs.[255,256] However, it must be noted that slight differences in secondary structure were found among the ELdcRs after deconvolution of the CD data using the BeStSel algorithm (Figure 2.2b).[7,8] A decrease in ordered structures (higher amount of undefined structures) was found for cuELdcRs. Although this did not affect their self-assembly, it appears that a high charge density in the reduced and condensed hydrophilic region may increase the disordered nature of the polypeptides, mainly by reducing the anti-parallel proportion. Table 2.1. Hydrophilic-hydrophobic block length ratio, comparison of the theoretical and experimental molecular weights (MW) calculated by MALDI-TOF and complete sequence of the ELdcRs. Ratio Theo. MW (Da) Exp. MW (Da) Sequence E-I 1:1.2 46972.8 46916.3±7.1 MESLLP-[(VPGVG)2-VPGEG- (VPGVG)2]10-VG-(IPGVG)59-IPGV S-I 1:1.2 46070.2 45996.3±16.9 MESLLP-(VPGSG)50-VG-(IPGVG)59IPGV E1/2-I 1:2.4 36585.8 36609.8±8.9 MESLLP-[(VPGVG)2-VPGEG- (VPGVG)2]5-VG-(IPGVG)59-IPGV oE-I 1:6 30593.4 30580.5±3.6 MESLLP-(VPGEG)10VG-(IPGVG)59IPGV oE1/2-I 1:12 28396.0 28380.3±4.4 MESLLP-(VPGEG)5VG-(IPGVG)59IPGV Nanostructure formation was followed by DLS, measuring the scattered light as a function of temperature. All ELdcRs exhibited a similar transition temperature (Tt) phase behavior in their DSC thermograms (below 25 °C, Figure 2.2c and A2.3), in accordance with the hydrophobic Charge distribution as molecular modulator of the nanostructuration of ELRs 47 II block (I) present in all ELdcRs. Differences in the Tt measured by DLS and DSC corresponded mainly to differences in the ELdcR concentrations used in each technique. Table 2.2. LCST (degrees celsius) of the ELdcRs measured by DLS and DSC in ultrapure water at physiological pH. DLS (°C) DSC (°C) E-I 20.1 19.6 S-I 24.4 18.2 E1/2-I 18.7 17.8 oE-I 23.8 17.2 oE1/2-I 22.5 19.1 DLS temperature trends and the low scattering intensity, below the Tt, suggested complete hydration and solubility of the ELdcR molecules (Table 2.2). When the temperature was increased above the Tt of the hydrophobic block (I), the collapsed I-block formed hydrophobic cores surrounded by the hydrophilic blocks. The increase in scattering intensity suggested the formation of spherical micelles triggered by the hydrophobic block collapse, as described previously.[54,56,58] Micellar assembly was studied in a concentration range (25-250 µM) in order to compare the nanostructuration of the diverse ELdcRs. In comparison with E-I polypeptide, DLS measurements and TEM images revealed that all the ELdcR designs were able to form similar nanostructures in solution at low concentrations (25 µM) (Figure 2.3). The unbalanced diblock designs also allowed self-assembly into spherical micelles despite of the fact that the length ratio between both blocks was outside the limits (Table 2.1) previously predetermined in the literature for micellar formation of uncharged amphiphilic ELdcRs (1:2 ≤ length ratio ≤ 2:1 [56] and the most recent, hydrophilic : hydrophobic ratio ≥ 0.3).[57] Chapter II 54 SEM micrographs show that the physical gel formed by E-I had a different appearance to those formed by E1/2-I and oE-I. However, the pore size was similar in E-I and E1/2-I gels (around 2 µm), but larger (up to ≈ 3 µm) for oE-I. Moreover, in the later, the pore wall thickness was so small that some pores were joined together to form “macropores”. In light of these results, we characterized the rheological properties of the hydrogels in order to quantify their viscoelastic properties. First, the linear viscoelastic region was determined by carrying out a swept of the strain amplitude. As can be seen from Figure A2.5, no significant changes were observed in the complex modulus magnitude up to a strain amplitude of around 1%. A 0.3% strain was selected for all the oscillatory measurements. Figure 2.6. Rheological characterization of the ELdcRs that undergo a liquid-gel phase transition at 37 °C: (a) Evolution of storage (G’) and (b) loss (G’’) moduli as a function of frequency. (c) Evolution of phase angle (δ) as a function of frequency and (d) ELdcR hydrogel viscosity as a function of shear rate at 37 °C under continuous flow. Frequency (Hz) 0,1 1 10 100 d (degrees) 0,01 0,1 1 10 100 Frequency (Hz) 0,1 1 10 100 G'' (Pa) 0 200 400 600 800 1000 Frequency (Hz) 0,1 1 10 100 G' (Pa) 0 1000 2000 3000 4000 (b)(a) Shear rate (s-1) 0,1 1 10 100 Viscosity (Pa·s) 0,1 1 10 100 1000 E-I E1/2-I oE-I (c) (d) Charge distribution as molecular modulator of the nanostructuration of ELRs 55 II Figure 2.6 shows the frequency response of both the storage (G´, Figure 2.6a) and loss (G´´, Figure 2.6b) moduli. While no noticeable dependence of G´ on frequency was observed for E-I, both E1/2-I and oEI exhibit a clear evolution with frequency. At 1 Hz, the higher storage modulus corresponds to E-I (see Table 2.3), whereas E1/2-I and oE-I have a value of around 2.5 kPa. As for G´´, a clear difference was observed at frequencies lower than 1 Hz. Thus, whereas no significant frequency dependence was found for EI, a strong evolution was observed for oE-I and, especially, for E1/2-I. As can be seen in Table 2.3, the loss modulus for E1/2-I at 1 Hz is considerably higher than for oE-I and E-I (in a ratio of 2.7 and 15, respectively). At the highest frequencies, similar values of about 150 Pa were found for all hydrogels. The phase angle can be calculated from the ratio of the loss and storage modulus the phase angle can be calculated (Figure 2.6c). A very high elastic behavior was suggested for E-I since phase angles lower than 1° were obtained. This elastic behavior was partially lost in the other two hydrogels at low frequencies, especially for E1/2-I, where values higher than 10° were reached. The phase angle evaluated at 1 Hz confirmed this trend (Table 2.3). Table 2.3. Mean G’, G’’ and δ for the different ELdcRs at 37 °C. Values were chosen from the linear viscoelastic region (LVR) (0.3% strain) at a frequency of 1 Hz. E-I E1/2-I oE-I G' (Pa) 3404.7±428.2 2824.3±261.3 2420.7±45.6 G'' (Pa) 32.03±2.2 480.2±58.6 180.9±5.2 δ (degrees) 0.55±0.05 9.7±1.57 4.27±0.05 Finally, the viscosity of the three hydrogels was measured as a function of shear rate using flow measurements. All three hydrogels exhibited a shear Chapter II 56 thinning behavior, with the hydrogel viscosity decreasing linearly by up to four orders of magnitude for the three orders of magnitude swept for shear rate (Figure 2.6d). Amphiphilic block co-polypeptides can form self-assembled micelles in solution with block length and distribution being key parameters for supramolecular assembly and stabilization.[56,61] The hydrophilic corona is responsible for avoiding aggregation and maintaining the hydrophobic core in solution. Thus, small coronas cannot sustain the micellar conformation and a minimum length ratio between the two blocks in the amphiphilic diblock design is required.[56,57] However, if the corona is charged, we face a different scenario in which electrostatic repulsion between charges enables the established limits to be overcome and new molecular designs to be obtained. Thus, an increase in charge density in the corona seems to contribute to micelle stabilization of highly unbalanced diblocks at low concentrations. Therefore, micellar assembly is favored in the oE1/2-I design (length ratio 1:12), just as it is also favored in the non-charged ELdcR (S-I). Electrostatic repulsion then contributes to micellar destabilization, promoting hierarchical assembly into higher-order structures. In the other three charged ELdcRs (E-I, E1/2I and oE-I), micellar assemblies evolve into larger assemblies on the mesoscale upon increasing the concentration. As such, electrostatic repulsion between charges seems to be the driving force for micellar destabilization, and the hydrophilic block length appears to be important for stabilization of the higher-order assemblies. Finally, charge distribution also contributes to liquid-gel transition and viscoelastic behavior of the hydrogel. Hydrophilic block length and composition modulate the rheological properties. Thus, whereas E-I (1:1.2 length ratio) forms elastic hydrogels with a viscoelastic behavior Charge distribution as molecular modulator of the nanostructuration of ELRs 57 II independent of the frequency, E1/2-I, which has a higher length ratio (to 1:2.4) but the same composition shows a strong frequency dependence for the loss modulus (G’’). In addition, an increase in charge density (oE-I, length ratio 1:6) compensates the frequency dependence observed in the E1/2-I hydrogel despite the condensed oE-block. In summary, the incorporation of charged residues into the backbone of IDPPs leads to the emergence of molecular interactions that trigger new molecular dynamics and behaviors, such as shear thinning of the gel form, which may find applications in tissue engineering or drug delivery. Our results also shed light on the folding of IDPPs that may be useful for understanding complex protein aggregates based on IDPRs such as biomolecular condensates. 2.3. Conclusion We have studied charge distribution as a molecular trigger in the supramolecular assembly of IDPPs. An IDPP library based on ELdcRs with an amphiphilic diblock design have been bioproduced, varying the hydrophilic block while maintaining the hydrophobic block composition in order to evaluate their influence on nanostructuration and phase transition. In addition, we have demonstrated that incorporating charges into the molecular design strongly contribute to assembly on the nano-, mesoand microscale. Electrostatic repulsion enables the micellar structuration of unbalanced diblock co-recombinamers at low concentrations and can, in turn, further contribute liquid-gel transition at higher concentrations. Finally, charge distribution seems to modulate the protein polymer entanglements within hydrogels and, hence, viscoelastic behavior. As such, this study represents a successful step toward the design of self-assembled Chapter II 58 protein devices for biomedical applications but also toward gaining further insight into the molecular mechanisms that govern the formation and maturation of biomolecular condensates. 2.4. Experimental section 2.4.1. Materials Glassware was dried at 120 °C overnight prior to use. Chemicals were purchased from Sigma-Aldrich and used as received unless otherwise mentioned. Ultrapure water (15 MΩ·cm) was used from Milli-Q A-10 Synthesis and Elix 10 Millipore. Restriction enzymes, shrimp alkaline phosphatase (SAP) and T4 ligase were purchased from ThermoFisher Scientific Inc. (USA). For the construction and expression of the encoding genes of the ELdcRs, we employed the cloning vector pDriveAll and the expression vector p7. These plasmids were constructed in our group from the commercial pDrive (Qiagen, Germany) and pET-25b (+) vectors (Novagen), respectively. [3] The DNA miniprep purification kit (Nucleospin® plasmid) and the Quick Gel Extraction Kit (PureLink™) were purchased from Macherey-Nagel (Germany) and Invitrogen (USA), respectively. 2.4.2. Recombinant synthesis of the ELdcRs To generate the battery of genes that encodes the ELdcR library the iterative-recursive method was employed. [2] Gene construction was performed in the pDriveAll vector using Escherichia coli strain XL-1 blue (Agilent, USA). Encoding genes were then cloned into a the p7 vector (Figure A2.1 and A2.2) and transformed for expression in E. coli Charge distribution as molecular modulator of the nanostructuration of ELRs 59 II BLR(DE3) strain (Novagen). Transformation in E. coli BLR(DE3) was carried out following the simple method developed by Chung et al. [257] Hyperexpression of the heterologous polypeptides was screened and the best clones were selected for the fermentation. Bacterial fermentation was performed in Terrific Broth (Formedium, UK) supplemented with glycerol (8 mL L-1) in a 15‐L bioreactor (Applikon biotechnology, USA), at 37 °C with agitation (500 rpm). Then, bacteria were collected by centrifugation and resuspended in pre-chilled ‘wash’ buffer (20 mM Tris Base, 140 mM NaCl, pH=7). This step was repeated until supernatant was clear. Then, bacteria were resuspended in ‘disruption’ buffer (20 mM Tris Base, 1 mM EDTA, 1mM PMSF, pH=8) and the polypeptides were liberated from the inclusion bodies by disruption in a GEA Lab PandaPLUS 2000 homogeneizer (GEA Farm Technologies, New Zealand). ELdcRs were purified by inverse transition cycling (ITC)[2] adding 1.5 M NaCl for precipitation above the Tt. Pure products were dialyzed against deionized and ultrapure water for three days (12,000 MwCO, Medicell Membranes Ltd, UK), neutralized with NaOH (pH≈7.0), filtered (0.22 µm Nalgene™, ThermoFisher Scientific, USA), lyophilized and stored at -20 °C until further use. A final yield of 425-640 mg L-1 of cells was obtained, depending the diblock construct. The monodispersity and purity of the ELdcRs were evaluated by sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS-PAGE), matrix assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry and by high performance liquid chromatography (HPLC). MALDI-TOF and HPLC were carried out in the Laboratorio de Técnicas Instrumentales of the University of Valladolid. Chapter II 60 The theoretical hydrophibicity of the ELdcRs was calculated using the ProtScale algorithm and the Kyte-Doolittle scale.[258,259] Agarose gel electrophoresis Agarose gels were prepared with and run in Tris-acetate EDTA (TAE) buffer (40 mM Tris-acetate, 1mM EDTA, pH 8). Analytical and preparative gels were run at 90 and 60 V, respectively. SimplySafe™ (EURx®, Poland) was used as stain for detecting nucleic acids in the agarose gel. SDS-PAGE Proteins were run under denaturing conditions on polyacrylamide gels following the protocol described by Laemmli.[6] Gels were stained using CuCl2 solution (0.3 M in distilled water). 2.4.3. Thermal behavior characterization The thermal behavior of the diblock co-recombinamers were characterized by dynamic light scattering and differential scanning calorimetry. The transition temperature at low concentrations was calculated measuring by DLS at 25 µM in ultrapure water. Thus, the scattered light intensity was recorded every 4 °C from 13 to 53 °C. Samples were stabilized for 2 min at each temperature and measured in triplicate with 11 runs per measurement. The temperature at which the scattered intensity reached 50% of the intensity was considered as the low critical solution temperature. DSC experiments were performed using a Mettler Toledo 822e equipment with liquid nitrogen cooler. Thus, 20 µL of at 50 mg mL-1 solution of the ELdcR in ultrapure water was placed in a standard 40-µL Charge distribution as molecular modulator of the nanostructuration of ELRs 61 II aluminium pan and sealed. The samples were stabilized at 5 °C for 5 min, then the temperature was increased by 5 °C min-1 from 5 to 60 °C. 2.4.4. Circular dichroism CD spectra were recorded using a Jasco J-810 spectropolarimeter (Jasco, USA) equipped with a temperature controller (Research Technical Services, University of Alicante, Spain). Samples were dissolved at 0.2 mg mL-1 and measured in 0.1 cm quartz cells in the range 190-250 nm. The temperature was stabilized at 37 °C for 10 min prior to measurement. Secondary structure percentages were determined using the BeStSel (Beta Structure Selection)[7,8] web server in the 200-250 nm range (when the dynode voltage was below 500 nm). Data were smoothed using a 15pt Savitzky-Golay filter. 2.4.5. Nanoparticle size and zeta potential analysis The nanoparticle size distribution and zeta potential were measured in ultrapure water using a Zetasizer Nano (Malvern Instruments, UK), with a 173° scattering angle and equipped with a HeNe laser (633 nm) with an output power of 10 mW. Size distribution was analyzed in a range of concentrations (25–250 µM) above the critical micelle concentration of the E-I polypeptide, [85] to evaluate nanoparticle stability and aggregation. Zeta potential measurements were carried out in the same concentration range, at 37 °C. Each sample was measured in triplicate. 2.4.6. Transmission electron microscopy TEM samples were prepared on 300-mesh carbon coated copper grids (C300Cu) with negative staining. To that end, grids were rendered hydrophilic by plasma treatment using a PDC-002 plasma cleaner (Harrick Chapter II 62 Plasma, USA). Low power setting (7.2 W applied to the RF coil) for 20 s. Then, 15 μL of the pre-incubated ELdcR (1 h at 37 °C at a concentration of 25 µM), ultrapure water and uranyl acetate (1% w/v) solutions were dropped on Parafilm® strip over a pre-heated (37 °C) glass surface. Plasma treated grids were placed onto the ELR drop for 90 s, on ultrapure water for 60 s, and finally, on the negative staining solution for another 60 s. Blotting filter paper (Whatman® Gel Blot GB003) was used to remove excess solution after every step by touching the edge of the grid. Images were taken using a Tecnai Thermionic T20 microscope operated at 200kV (SAI, University of Zaragoza, Spain). 2.4.7. Physical hydrogel formation and characterization Samples were dissolved in ultrapure water at a concentration of 2.5 mM in ultrapure water at 4 °C, then incubated at 37 °C for 30 min to qualitatively test the ability to form physical hydrogels. Rheological analysis The mechanical properties of the hydrogels were tested by performing oscillatory shear and flow measurements in an AR2000 rheometer (TA Instruments) using a parallel plate with a diameter of 40 mm. Measurements were performed with a sample volume of 1300 µL (gap = 1100 µm) 37 °C, controlling the temperature with a Peltier plate. After sample deposition at 4 °C, gel formation was accomplished in situ. Data were recorded using TRIOS software (v4.1.1.33073). For the oscillatory shear measurements, a strain sweep was performed from 0.01% to 15% at an angular frequency of 1 Hz to test the LVR. Frequency sweeps were carried out sequentially from 0.1 to 50 Hz, with a constant strain of 0.3% (corresponding to the LVR). The storage (G´) and loss modulus (G´´) were obtained from the Charge distribution as molecular modulator of the nanostructuration of ELRs 63 II rheological measurements. The loss factor (tan δ ≡ G´´/G´, where δ is the phase angle between the output response to the input stimulus) and the complex modulus magnitude ((|G*|2 = (G´)2 + (G´´)2 were then calculated using the values obtained. Flow measurements were employed to measure the viscosity. Samples were conditioned with a pre-shear of 500 s-1 for 1 min, then the viscosity was measured in a flow ramp from 500 to 0.1 s-1 using a continuous ramp in a logarithmic descending series of discrete steps. Overall measurement took 10 min, acquiring 10 points for each order of magnitude. All measurements were performed at least in triplicate. Scanning electron microscopy In order to visualize the morphology of those samples that formed hydrogels, samples were incubated for 1 day at a concentration of 2.5 mM at 37 °C. They were then cryo-fractured in liquid N2 and lyophilized. SEM micrographs were obtained using a FEI Quanta 200 FEG (FEI Company, USA) microscope in low vacuum mode (SAI, University of Zaragoza, Spain). A 20 nm layer of Pd was applied with a Leica EM ACE200 vacuum coater to avoid charging effects. SEM images were analyzed using the ImageJ software. Chapter III: Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs Chapter III Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 73 III 3.1. Introduction ntimicrobial peptides (AMPs) or host defense peptides (HDPs) are one of the most promising alternatives to conventional antibiotics in the treatment of the increasingly frequent drugresistant infections.[260,261] Their broad-spectrum antimicrobial activity and their immunomodulatory properties rely on their amphipathicity that confers the ability to interact with several molecular targets,[148,262] or even undergo self-assembly. Recent investigations have demonstrated their ability to form supramolecular assemblies including fibers,[208] nanoparticles,[209] twisted nanoribons[210] or hydrogels.[211–214] Moreover, peptide assembling can predetermine their biological effects, enhancing antimicrobial activity, [217] but it may induce cytotoxic effects[218] or other unforeseen side effects.[263] As mentioned in Chapter 1, AMPs present some drawbacks that limit their clinical use. They are highly sensitive to environmental conditions, they show short plasma-half-life because their sensitiveness to proteolytic degradation and the high cost of their chemical synthesis hinder the scale-up of their production.[141] Therefore, in many cases, their therapeutic use has been restricted to topic applications. To overcome these limitations, AMPs often require molecular carriers to control their delivery. In this context, AMPs have been integrated into synthetic or natural polymers in order to decrease their cytotoxicity, protect them from the protease degradation and control their delivery.[264] Additionally, the use of self-assembled polymers enables to obtain interesting architectures for their biomedical application, including vesicles, fibers or hydrogels.[265] In this sense, recombinant smart polymers constitute highly interesting candidates. On the one hand, their extreme versatility and sequence control given by their inherent recombinant nature provides a sustainable alternative for the co-production of AMPs as fusion partners. A Chapter III 74 On the other hand, their stimulus-responsiveness can be used in a synergistic combination with the self-assembly behavior of the AMPs to design advanced functional nanomaterials for biomedical applications. Therefore, we aim to investigate in this chapter the interplay of selfassembling AMPs and smart protein polymers. As smart polymers, we employed an ELR with modular design based on an amphiphilic diblock, where the AMP was fused into the hydrophilic block through a spacer. In this way, both potential self-assembling domains (SADs) (AMP and hydrophobic block of the ELR) were located in the two ends of the molecule. We designed two different hybrid polypeptides containing the well-characterized designer peptides: GL13K and 1018. These AMPs have similar physicochemical properties and both have proven to self-assemble in solution.[210,218] Moreover, GL13K and 1018 show potent antibiofilm properties combined with broad spectrum bactericidal activity and immunomodulatory properties, respectively.[172,261] Herein, we propose an alternative approach for the design of hierarchical self-assembled nanomaterials exploiting the self-assembly behavior of AMPs and the thermo-responsiveness of the ELRs. We hypothesized that the combination of both SADs in both ends of a hybrid polypeptide could lead the hierarchical assembly through a dual process. First, AMP-domain would drive the self-assembly hiding the antimicrobial sequences and consequently protecting them from the environment. Then, the thermo-responsive ELR-domain would trigger a secondary assembly creating molecular aggregates increasing the local concentration of the peptide. Following this rationale, nanoreservoirs for AMPs could be developed. This approach seeks to shed light into the mechanisms that govern into the supramolecular assembly of hybrid protein-polymers based Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 75 III on AMPs in order to set the basis for the de novo design of functional nanocarriers to safely treat drug-resistant infections. 3.2. Results and Discussion 3.2.1. Molecular design strategy and production Hybrid AMP-ELRs were engineered and biosynthesized by recombinant DNA technology using a modular design (Figure 3.1). Two different domains can be differentiated. On the one hand, the ELR domain was based on the amphiphilic, non-charged, diblock design, SI, developed in Chapter 2. SI polypeptide contains hydrophilic [(VPGSG)50] and hydrophobic block [(IPGVG)60] whose LCST are below and above physiological temperature, respectively. At 37 °C, the hydrophobic block (I) collapses into hydrophobic cores surrounded by hydrophilic corona (S) and therefore, driving the formation of micellar nanostructures. On the other hand, designer AMP sequences (GL13K or 1018) were located on the N-terminus, connected to the hydrophilic block (S) through a flexible polyGly spacer. In this way, both potential SADs were located in two ends of the molecule (Figure 3.1a). Figure 3.6. (a) Schematic representation of the modular design of the hybrid polypeptide and sequence of the polypeptides. (b) Copper stained SDS-PAGE of the pure ELRs. AMP Spacer ELR ELR Composition SI MESLLP-(VPGSG)50-VG(IPGVG)59 IPGV GL13K - SI GKIIKLKASLKLL-VLG10L-(VPGSG)50-VG(IPGVG)59 IPGV 1018-SI VRLIVAVRIWRR-VLG10L-(VPGSG)50-VG(IPGVG)59 IPGV (a) 66.2 45.0 35.0 25.0 kDa (b) Chapter III 76 AMP-ELR were produced as pro-polypeptides. We introduced a sacrificial ELR-block in the N-terminus of the AMP, named HE, in order to: (i) protect to the producing strain to toxic side effects of the AMP during the fermentation; (ii) increase the expression levels; (iii) enable site-specific cleavage, a Met was incorporated in its C-terminus that allowed us to release the AMP-ELRs with no extra amino acid that may affect the AMP bioactivities; and (iv) facilitate the AMP-ELR purification with an His-tag intended for the selective removal of the HE-block and the uncleaved constructs from the AMP-ELRs. After recombinant production, ITC purification and chemical cleavage of the sacrificial block, AMP-ELRs were purified with a nickel-charged agarose resin (Figure A3.1 and A3.2). Monodisperse and highly pure products were obtained as revealed SDS-PAGE, MALDI-TOF and HPLC analysis (Figure 3.1b, A3.3, A3.4 and Table A3.3). 3.2.2. Phase behavior characterization Given the thermo-sensitiveness of the ELRs, we studied the thermal behavior of the three recombinamers in aqueous solution. We monitored the evolution of the optical density at 350 nm (OD350) during consecutive heating and cooling cycles in the range 5-40 °C (Figure 3.2). All the three polypeptides (ELRs or AMP-ELRs) showed a reversible LCST behavior with a transition temperature (Tt) below physiological temperature. However, a slight increase of the OD350 after cooling down the samples was observed. This may indicate that the process was not completely reversible. Below the Tt (Table 3.2), SI polypeptides were soluble. Raising temperature above the Tt, triggered the collapse of the I-block and the formation of hydrophobic cores which stabilized in solution by a surrounding hydrophilic coronas (S-blocks) and thus, SI self-assembled into micellar Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 77 III nanostructures.[58,85] When the samples were cooling down, micelles disassembled and the OD350 decreased. However, a minimal fraction of the hydrophobic interactions between Ile-side chains seemed to remain and hence, OD350 was greater than the pre-heated state, as it occurs in the poly(VPGLG).[266] Furthermore, as well as poly(VPGLG), thermal hysteresis appeared during cooling cycles, possibly due to these hydrophobic interactions. Figure 3.2. Thermal behavior of the ELRs monitored by turbidimetry. Optical density at 350 nm (OD350) evolution as a function of the temperature demonstrated that all the ELRs behave a reversible liquid-liquid phase separation. The presence of the AMPs contributed to the phase transition and thermal hysteresis. Similarly, hybrid AMP-ELRs also behave a reversible phase transition but with substantial differences. First, OD350 values reached above the Tt were significantly greater than SI. SI self-assembled into stable nanostructures in solution. Consequently, OD350 was below 0.1. In contrast, Chapter III 78 OD350 was much higher for the hybrid AMP-ELRs when the temperature was above the Tt. As previously observed for AMP in solution, AMPdomains may self-interact and therefore, they contributed to assembly into larger aggregates that increased the OD350.[210] Comparing to SI, OD350 values were one or two orders of magnitude greater for 1018-SI and GL13K-SI, respectively. Table 3.1. Transition temperatures (Tt) and hysteresis (ΔTt) of the ELRs. Then, the presence of AMP affected the phase transition (Table 2) in comparison with the SI polypeptide. Tt decreased and thermal hysteresis increased in spite of AMP-domains introduced charged residues. This suggests that the interactions between AMPs contributed to the cooperative phase transition of the ELR-domains, reducing Tt. Coherently, the AMPdomains also enhanced thermal hysteresis. AMP aggregation may drive the formation of ordered structures[210,267] which could increase the intermolecular order in the coacervate state and hence, the hysteresis. This behavior is consistent with previous studies where order-promoting domains (polyAla) were introduced in the ELR backbone.[116] 3.2.3. Self-assembly dynamics of the hybrid AMP-ELRs Phase transition characterization suggested that AMP-domains may play an important role in the supramolecular assembly of the AMP-ELRs. They may contribute to the formation of higher-order aggregates and to the stabilization of the assembled structure, thus increasing thermal hysteresis. SI GL13K-SI 1018-SI Tt heating (°C) 24.1 22.7 19.3 Tt cooling (°C) 21.8 17.7 15.3 Hysteresis (ΔTt) (°C) 2.3 5.0 4.0 Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 79 III Therefore, we proceeded to characterize the self-assembly dynamics in particular detail below and above the Tt of the I-block. At 5 °C, below the LCST, ELR molecules were completely soluble. SI polypeptide did not form any nanostructures (Figure 3 and 4a). DLS distribution around 10 nm corresponded with the soluble macromolecules.[268] In contrast, the presence of the AMPs within the recombinant polypeptides triggered the self-assembly into fibrillar nanostructures. The aggregation of the peptides 1018 and GL13K had been previously observed in solution and it was found to be dependent on the pH or the presence of salts.[210,218] Electrostatic repulsion between the positively charged side-groups have to be neutralized to favor the interaction between the peptides and to form higher-order assemblies. Interestingly, self-assembly of the hybrid polypeptides occurred in salt-free solution and at slightly acid pH (≈6). This suggests that the combination of a larger polypeptide chain with the AMP may have induced a cooperative effect that facilitated the aggregation of the AMP-domains. Additionally, fibrillar nanostructures formed by the AMP-ELRs underwent a dynamic behavior. In both cases, nanofibers evolved over time but different aggregation patterns were observed for each hybrid polypeptide. 1018-SI formed longer nanofibers than GL13K-SI as TEM characterization revealed (Figure 3.3, second and third column). The growth of GL13K-SI nanofibers seemed to be spatially constrained, thus limiting the fiber elongation in favor of nanofibers with repeated patterns. DLS analysis verified the size differences between both fibrillar populations after short periods of time (Figure 3.4a). However, it must be noted that the large nanofibers found in 1018-SI samples after 1 day of incubation could not be monitored by DLS, possibly because large fibrillar aggregates Chapter III 86 mention that the presence of the spacer between the AMP-domain and the ELR diblock, did not compromise the hierarchical assembly. Thus, the introduction of functional spacers, including sensitive sequences to biological or physical stimuli (e.g. protease degradation or pH) opens up a range of possibilities in the development of release-controlled nanocarriers for AMPs. Figure 3.6. Cryo-TEM micrographs of the AMP-ELRs: (a) and (c) correspond to GL13KSI samples and (b) and (d) to 1018-SI samples. (a, b) After the incubation at 5 °C for 24 h, the AMP triggered the fibrillar assembly, whereas (c, d) when the fibers were incubated at 37 °C, the coacervation of the ELR drove the formation of aggregates. (e) Schematic representation of the hierarchical self-assembly of the hybrid polypeptides. 3.3. Conclusions In conclusion, the use of AMPs combined with stimuli-responsive protein polymers is a promising strategy for the design of self-assembled nanomaterials for biomedical applications. We have shown that AMPs can be used as SADs to trigger the assembly of larger protein polymers, and depending on the AMP, different nanostructures can be achieved. Moreover, their combination with thermo-responsive polypeptides in modular designs enables to fabricate hierarchical nanostructures formed by a dual assembling process. Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 87 III Therefore, our nanosystem represents a sound strategy in the fabrication of smart biomaterials incorporating AMPs. Their recombinant nature facilitates the edition of the modular design and the incorporation of other bioactive motifs with extreme control, in addition to a scalable method for their sustainable production and potential widespread use. This investigation provides a new insight in the development of delivery strategies for combating drug-resistant infections, but also engenders enormous opportunities for the production of self-assembled devices for tissue engineering and regenerative medicine applications taking advantage of the immunomodulatory properties of AMPs. 3.4. Experimental section 3.4.1. Gene construction ELR encoded genes construction was performed using previously described procedures (Chapter 1 and 2). Encoding genes for the AMPs were purchased from NZYTech, Lda. (Portugal) and cloned into a modified pDrive plasmid flanked by EarI restriction sites,[47] using E. coli XL-1 blue (Agilent, USA) as cloning strain. Then, by iterative recursive method, the final genetic constructions (HE-AMP-SI) were completed. 3.4.2. Bioproduction and purification All the polypeptides used in this work were recombinantly bioproduced. Briefly, encoding genes were cloned into pET-25b (+) expression vectors and transformed into E. coli BLR (DE3) for heterologous expression. After overnight fermentation in a 15 L bioreactor (Applikon Biotechnology, The Netherlands), the polypeptides were purified by inverse transition cycling (ITC) adding 1.5 M NaCl for their warm precipitation.[47] After 3 cycles, we assessed the purity and Chapter III 88 moniodispersity by SDS-PAGE and we dialyzed them against ultrapure water, lyophilized and stored them at -20 °C. The yields observed ranged from 380 to 600 mg L-1 of purified ELR per liter of bacterial culture. 3.4.3. Protective block cleavage and purification AMP-ELRs were designed and bioproduced in E. coli BLR(DE3) as pro-polypeptides (HE-AMP-SI, Table A3.1). After the recombinant expression, we verified by SDS-PAGE that the pro-AMP-ELRs were completely pure (Figure A3.1). We removed the sacrificial block, HE. For this, pro-polypeptides were incubated with CNBr solution (70% formic acid, FA) in a molar ratio 1:200, Met:CNBr. The reaction was released for 20 h at room temperature in the darkness and under anaerobic conditions. Then, CNBr was eliminated in a rotary evaporator. The ELRs were resuspended in ultrapure water and dialyzed. After four dialyzing steps against cold ultrapure water and lyophilization, we purified the cleaved AMP-ELR using HisPur™ Ni-NTA resin (ThermoFisher Scientific, USA) by batch methodology. In short, lyophilized products were dissolved in denaturing buffer (4M urea, 20 mM sodium phosphate, 500 mM NaCl) in order to prevent physical interactions between the AMP. 30 mL of the dialyzed solution was mixed with 15 mL of the resin in 50 mL-tubes and incubated at 200 rpm for 3 h at 4 °C. Then, resin was centrifuged. Due to the presence of the His-tag within the sacrificial block, HE and HE-AMPSI polypeptides bonded the resin, whereas the AMP-ELR was in the supernatant. After two purification steps, AMP-ELR polypeptides were completely purified (Figure A3.2). Finally, the polypeptide solutions were dialyzed, filtered (0.22 µm Nalgene™, ThermoFisher Scientific, USA) lyophilized and stored at -20 °C until further use. Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 89 III Monodispersity and purity of the hybrid AMP-ELRs were assessed by SDS-PAGE (Figure 3.1, A3.2), MALDI-TOF (Figure A3.3 and Table A3.2) and HPLC (Table A3.3). MALDI-TOF and HPLC analysis were performed in the ‘Laboratorio of técnicas instrumentales’ (LTI) at the University of Valladolid (Spain). 3.4.4. Phase transition characterization Thermal behavior was evaluated by turbidimetry. Absorbance at 350 nm was measured in the range of 5-40 °C with a scan rate of 1 °C with a Cary 100 UV-Vis spectrophotometer (Agilent). Heating and cooling ramps were performed at 0.25 °C min-1. All the samples were prepared in ultrapure water at a concentration of 25 µM. 3.4.5. Physical characterization of the nanostructuration Self-assembly dynamics of the hybrid polypeptides were evaluated below and above the transition temperature of the hydrophobic block (I) and compared with the ELR control, SI. For this, 25 μM solutions in ultrapure water were prepared under sterile conditions and incubated at 5 or 37 °C for 10 min, 1 h, 4 h, 1 d, 3 d, and 7 d. After that, nanostructuration of the polypeptides were analyzed by dynamic light scattering and transmission electron microscopy. Dynamic light scattering Nanoparticle size distribution was evaluated using a Zetasizer Nano (Malvern Instruments, UK), with a 173° scattering angle and equipped with a HeNe laser (633 nm) with an output power of 10 mW. Each sample was measured in triplicate. Chapter III 90 Transmission electron microscopy TEM samples were prepared on 300-mesh carbon coated copper grids with negative staining. For this, first, grids were rendered hydrophilic by plasma treatment in a PDC-002 plasma cleaner (Harrick Plasma, USA) at low power setting (7.2 W applied to the RF coil) for 20 s. Then, 15 μL of the pre-incubated ELRs, ultrapure water and uranyl acetate (1% w/v) solutions were dropped on Parafilm® strip over a pre-chilled (5 °C) or preheated (37 °C) glass surfaces. Plasma treated grids were placed onto the ELR drop for 90 s, on ultrapure water for 60 s, and finally, on the negative staining solution for another 60 s. Blotting filter paper was used to remove excess solution after every step. Images were taken using a Tecnai Thermionic T20 microscope operated at 200kV (SAI, University of Zaragoza, Spain). Circular dichroism spectroscopy ELRs or AMP-ELRs solutions were prepared at 5 µM in pre-chilled ultrapure water and incubated at 5 or 37 °C for 10 min, 1 h, 4 h, 1 d, 3 d, and 7 d. The CD signal was measured from a 200 µL solution in a quartz cuvette (1 mm path-length) using a CD spectrometer (Jasco J-815, Easton, MD, USA). It was scanned over a range of 260-190 nm with a data pitch of 1 nm, a scanning rate of 50 nm min-1 and a response time of 2 s. All measurements were subtracted from the background signal from ultrapure water in the quartz cuvette and repeated in triplicates. 3.4.6. Cryogenic TEM To test the thermo-responsiveness of the fibrillar structures formed by the AMP-ELRs, we preformed nanofibers and then we evaluated by cryo- Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 91 III TEM the behavior of the nanostructures. For this, 25 µM solutions in ultrapure water of the three ELRs (SI, GL13K-SI and 1018-SI) were incubated at 5 °C for 24 h, in order to drive the AMP fibrillar assembly. Then, samples were heated at 37 °C for 30 min. Cryo-TEM samples were prepared before and after heating to evaluate changes in the nanostructuration. Cryo-TEM samples preparation and visualization were carried out at the ‘Electron Microscopy Platform’ (CICbioGUNE, University of the Basque Country, Spain). For this, four microliters of the sample were placed onto a glow-discharged 300-mesh lacey-carbon coated grids (Lacey Carbon film on 300 mesh copper; LC300-Cu; Electron Microscopy Sciences) and incubated inside the chamber of a Vitrobot Mark III (FEI Inc., The Netherlands) at 4 ˚C and at a relative humidity close to saturation (95% rH) for 30 s. Most of the liquid in the grid was removed by blotting (3 s at an offset of -3 mm) and vitrified by plunging into liquid ethane, previously cooled with liquid nitrogen at approximately -180 ºC. Images were collected at liquid nitrogen temperature on a JEM2200FS/CR (JEOL Europe, Croissy-sur-Seine, France) field emission gun transmission electron microscope operated at 200 kV. An in-column energy filter (Omega filter) produced images with improved contrast and signalto-noise ratio by zero-loss filtering. The energy slit width was set up at 15 eV. Digital images were recorded on a 4K × 4K Ultrascan4000™ chargecoupled device (CCD) camera (Gatan, Inc.) using DigitalMicrograph™ (Gatan, Inc.) software. 3.4.7. Minimal inhibitory concentration The antimicrobial activity of the hybrid polypeptides was evaluated by MIC assays against Gram-positive Streptococcus gordonii DL-1 and Gram- Chapter III 92 negative Pseudomonas aeruginosa PAO, Xen 41 (PerkinElmer Inc., Waltham, MA, USA) following the protocol described elsewhere.[12] Briefly, 3-5 colonies of S. gordonii or P. aeruginosa were picked from fresh plates and inoculated in a sterile 15-mL falcon tube with 2 mL of brain heart infusion (BHI) broth or Mueller-Hinton Broth (MHB), respectively. After overnight incubation at 37 °C under aerobic conditions (P. aeruginosa.) or with 5% CO2, inocula were diluted until 5 × 105 CFU mL-1 in BHI or MHB and 90 μL of this solution was incubated with a 10 μL serial peptide dilution from a 1280 µM stock solution in 96-well polypropylene plates at 37 °C for 20 h. The bacterial growth was evaluated visually and by optical density (OD) at 570 nm in a Synergy HT plate reader (BioTek, Winooski, VT). Additionally, designer peptides GL13K-NH2 and 1018-NH2 (purchased by Pepscan (The Netherlands) with a purity of more than 97%) were tested as positive controls. Each experiment was repeated at least three times. Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 93 III 3.5. Appendix Table A3.2. Complete sequence of the engineered block co-recombinamers. Sequence HE MESLLPVG5H6G5[(VPGVG)2(VPGEG)(VPGVG)2]10VM HE-GL13K-SI MESLLPVG5H6G5[(VPGVG)2(VPGEG)(VPGVG)2]10 VM GKIIKLKASLKLLV LG10L (VPGSG)50VG(IPGVG)59IPGV HE-1018-SI MESLLPVG5H6G5[(VPGVG)2(VPGEG)(VPGVG)2]10VM VRLIVAVRIWRR VLG10L (VPGSG)50VG(IPGVG)59IPGV GL13K-SI GKIIKLKASLKLLVLG10L(VPGSG)50VG(IPGVG)59IPGV 1018-SI VRLIVAVRIWRRVLG10L(VPGSG)50VG(IPGVG)59IPGV Figure A3.7. Copper stained SDS-PAGE (12.5%) electrophoresis of (a) HE-1018-SI and (b) HE-GL13K-SI (lane 1) recombinamers and band pattern after chemical treatment with CNBr (lane2). Lane M is Pierce™ Unstained Protein MW Marker. Chapter III 94 Figure A3.8. Analysis of the purification of 1018-SI (a) and GL13K-SI (b) after CNBr cleavage by 12.5% SDS-PAGE visualized with copper staining. Lane M is Pierce™ Unstained Protein MW Marker (ThermoFisher). We obtained highly pure and monodisperse products after two purification steps by batch method, using the HisPur™ Ni-NTA resin. Lanes 1: After CNBr treatment. Lanes 2: After incubation 1 with the resin. Lanes 3: Wash 1. Lanes 4: Eluted from the resin. Lanes 5: After incubation 2 with the resin. Lanes 3: Wash 2. Lanes 4: Eluted (2) from the resin. Figure A3.4 . MALDI-TOF spectra of the purified ELR SI. Self-assembly of genetically engineered polymers: the interplay between AMPs and ELRs 95 III Figure A3.4 . MALDI-TOF spectra of the purified AMP-ELRs GL13K-SI and 1018-SI. Table A3.3. Comparison of the theoretical and experimental MW calculated by MALDI-TOF. Theo. MW (Da) Exp. MW (Da) SI 46070 46006±52 GL13K-SI 47702 47845±34 1018-SI 47814 47528±43 Development of self-assembled monolayers with antibiofilm activity based on AMPs and ELRs 103 IV 4.1. Introduction he aging of the population combined with technological developments in the field of biomaterials have resulted in the increasing use of biomedical devices. However, the implantation of temporary or permanent medical devices implicitly increases the risk of bacterial infections.[271] Indeed, biofilm-related infections and the increasing appearance of multidrug resistant bacteria make medical device-associated infections a significant economic and medical concern.[272,273] Although a wide range of bacteria are implicated in such infections, staphylococci, especially Staphylococcus aureus and Staphylococcus epidermidis, are of particular importance. Indeed, S. aureus and S. epidermidis are the leading cause of reported healthcare-associated infections and tare particularly relevant in biomaterials-associated infections (BAIs) due to their ability to form biofilms, their prevalence and the occurrence of multi-drug resistant phenotypes.[274,275] Significant efforts have been made to design bactericidal and antifouling coatings to prevent bacterial adhesion and inhibit biofilm formation on biomaterials,[276,277] including antimicrobial agent releasingbased, anti-adherent and contact-killing coatings. Traditional antibiotics,[278] metallic nanoparticles, such as Ag[279] or Zn,[280] and polymers[281] are the most widely used approaches and have proved effective against staphylococcal strains, although with significant limitations, such as antibiotic resistance,[282] toxicity at high concentrations,[283,284] and lack of biocompatibility.[285] As such, in recent years, the use of AMPs has increased markedly to develop antibiofilm coatings,[219] thereby demonstrating their potential applicability for the prevention of indwelling device-associated infections. T Chapter IV 104 As mentioned in Chapter 1, AMPs have been directly immobilized onto biomedical materials[223,286–289] or bioconjugated with a polymeric scaffold that may improve their antimicrobial activity,[229,290,291] and they can also incorporate additional functionalities such as low-fouling behavior,[292,293] multiple and synergistic antimicrobial peptides,[294] or an ability to promote tissue integration.[295,296] Furthermore, the immobilization of AMPs may improve some of their limitations, minimize their toxic side effects and improve their susceptibility to proteases.[297] The main limitation for the production and improvement of AMPbased coatings for clinical materials is the high cost of chemical manufacture, which impedes scale-up.[298] As such, the recombinant production of AMPs has been studied and several examples have been produced using this methodology.[299,300] Despite this, the recombinant production of AMPs usually involves the use of expensive techniques for purification of the final product that increase final costs, thereby hampering large-scale production.[301,302] In this regard, and because of their thermosensitive behavior, ELRs enable the efficient recombinant production and simple purification of fusion proteins in a cost-effective scalable process. [303,304] In addition, chimeric AMP coproduction with ELRs delivers highly monodisperse and pure products and, most importantly, enables the synthesis of sophisticated antimicrobial designs with improved properties and applications by taking advantage of the elastin-like smart behavior and their potential complex molecular architecture. Examples of the latter are films for wound healing[202] and self-assembling antimicrobial nanoparticles.[203] On the other hand, ELR-based matrices and coatings have been shown to exhibit protein antifouling activity, preventing the unspecific attachment of proteins.[130,131,305] Thus, the combination of AMPs Development of self-assembled monolayers with antibiofilm activity based on AMPs and ELRs 105 IV and ELRs is an attractive alternative for the development of advanced antimicrobial coatings for biomedical materials. In this chapter, we develop self-assembled monolayers (SAMs) based on a multifunctional design, in which the antibiofilm properties of an AMP are enhanced upon combination with a low-fouling ELR. Antibiofilm properties of the coatings were tested against two staphylococcal singlestrain biofilm models (S. aureus and S. epidermidis). Strong antibiofilm activity and cytocompatibility of these coatings was demonstrated, thus confirming the potential of recombinant approaches for the production of antimicrobial coatings with powerful features for biomedical devices. 4.2. Results and Discussion 4.2.1. Rationale of the hybrid polypeptide design and biosynthesis The ELRs used in this study are based on a multimodular design. The first ELR, referred to as VC and used as control, comprises two modules, namely a polycationic backbone (VPGXG)40 (where X is Val and Lys in a 5:1 ratio) and a cysteine-based C-terminal grafting motif (Cys-cys-motif) for covalent immobilization onto surfaces (Table 4.1). Cysteine side-chains have demonstrated that are excellent candidates for the selective covalent and functional immobilization of peptides and proteins onto multiple surfaces and biomedical materials for the biofabrication of surfaces with antimicrobial properties,[219,306] cell adhesion selectivity,[307,308] or enzymatically active.[309] In contrast, the hybrid polypeptide (AMP-ELR), referred to as GVC, also incorporates the designer peptide GL13K at its N-terminus via a flexible spacer comprising ten glycines.[310] Polycationic ELRs provide a positive environment, and specific C-terminal attachment facilitates Chapter IV 106 molecular flexibility that may increase the antimicrobial potential[311] while minimizing cytotoxic side reactions.[312] Table 4.1. MWs and sequences of the AMP and ELRs used in this chapter. MW (Da) Sequence GL13K 1528.02 GKIIKLKASLKLLC-NH2 VC 21038.0 MESLLPVG (VPGVG VPGKG (VPGVG)4)8 VCC GVC 22670.0 GKIIKLKASLKLLV LG10L VG (VPGVG VPGKG (VPGVG)4)8 VCC EGVC 45241.3 MESLLP [(VGPVG)2VPGEG(VGPVG)2]10V LG10LVM GKIIKLKASLKLLVLG10L VG (VPGVG VPGKG (VGPVG)4)8 VCC The GVC recombinamer was designed produced as an EGVC, with a sacrificial block (E) being included immediately before the GL13K sequence as part of the modular design. This E block plays the same role as HE-block described in Chapter 2. Moreover, their polyanionic composition (Table 4.1) enables, after cleavage, the easy separation from the final product (GVC) by ITC in a single step in addition to the uncleaved original product (EGVC) using the same method. After recombinant production and purification, purity was assessed by diverse characterization techniques, molecular weights and monodispersity were verified by SDS-PAGE (Figure A4.2) and MALDI-TOF (Figure A4.3 and Table A4.1), and amino acid composition by HPLC analysis (Table A4.2). Development of self-assembled monolayers with antibiofilm activity based on AMPs and ELRs 107 IV Figure 4.1. Schematic representation of the modular composition of the AMP-ELR and production of the AMP/ELR/AM-ELR self-assembled monolayers (SAMs) on gold surfaces. 4.2.2. SAMs biofabrication and characterization For the study of the bioactivities as covalent coatings, GVC polypeptides and controls (GL13K peptide and VC polypeptide) were covalently to model gold surfaces. Cys-Cys-motif allowed to generate SAMs from the functionalization of thiols on gold substrate (Schematically represented in Figure 4.1). Then, the surfaces were characterized by water contact angle (WCA), X-ray photoemission spectroscopy (XPS), and quartz crystal microbalance with dissipation (QCM-D). Physical and chemical characterization was performed, measuring the coating wettability by WCA and quantifying the elemental composition by XPS, respectively. Finally, to assess the effectiveness of AMP/ELR/AMP-ELR deposition, we used QCM-D to obtain the thickness of the coatings. The wettability of the coatings was assessed by measuring the static contact angle of ultrapure water drops on the surfaces with a stabilization time of 15 s. The wettability of pristine gold surfaces (WCA = 76.4 ± 2.5°) decreased after being coated with immobilized GL13K peptides (WCA = Chapter IV 108 85.6 ± 3.8°) (Figure 4.2). The hydrophobicity of the surface increased as the hydrophobic residues of these amphipathic peptides became exposed at the solid/air interface. [287] The VC recombinamer, in contrast, is a hydrophilic cationic molecule, therefore the VC coatings showed higher wettability (WCA = 58.6° ± 3.4°) than pristine gold and GL13K-coated surface. However, due to the presence of the GL13K peptide, GVC coatings were slightly more hydrophobic than VC coatings (WCA = 67.5° ± 3.1°), thus indicating that GL13K folds and exposes its hydrophobic residues at the solid/air interface. [210] Figure 4.2. Static water contact angle of the gold surfaces when the AMP/ELRs are covalently attached. Depending on the physicochemical properties of the molecules, the wettability of the surfaces changes significantly (**p<0.001). At least 10 different measurements for each surface are represented in the box diagram. Error bars represent standard deviation values. The elemental composition was quantified by XPS. The atomic ratios of the coatings are shown in Table 4.2. After immobilization of the peptide/polypeptides C 1s, O 1s and N 1s peaks increased (Figure A4.4), Au GL13K VC GVC Water Contact Angle (º) 40 50 60 70 80 90 100 ** Development of self-assembled monolayers with antibiofilm activity based on AMPs and ELRs 109 IV which resulted in an increase of the atomic ratios, demonstrated the deposition of the protein materials. Quantitatively, N/Au ratio (N 1s, characteristic peak for proteins) increased from an initial 0.04 in the pristine gold surfaces up to 0.74, 1.91 and 2.52 for GL13K, VC and GVC-coatings, respectively. Furthermore, when gold surfaces were modified with thiols, the energy for the Au 4f peak shifted, [313] thus indicating covalent attachment to the surface (Figure A4.5). Table 4.2. Quantitative XPS analysis of the covalently anchored AMP/ELR SAMs on gold surfaces. Relative atomic ratios of the most representative elements (C, O and N) of the coated molecules are shown respect to substrate element, Au. Finally, the thickness and the area density of the SAMs were estimated quantitatively using the QCM-D technique. First, the viscous penetration depth (δ) corresponding to ultrapure water, the solvent used, was estimated [314]. The decay rate of the oscillating wave with the distance from the sensor surface is indicated by (Eq. 4.1). 𝛿=√2𝜂 𝜌𝜔 Where η and ρ are the viscosity and density of the solution employed for the measurement, respectively, and ω is the oscillation frequency. For ultrapure water (η = 0.93 mPa·s and ρ = 0.998 g/cm3), δ is approximately 140 nm at 15 MHz (third overtone, n = 3). C / Au O / Au N / Au Au 0.56 0.08 0.04 GL13K 3.52 0.78 0.74 VC 9.24 2.23 1.91 GVC 10.71 2.87 2.52 (Eq. 4.1) Chapter IV 110 QCM-D measurements were performed at 23°C. The simultaneously measured shifts in frequency (normalized to the corresponding overtone, n), Δfn/n (Figure 4.3a), and energy dissipation, ΔDn (Figure 4.3b), obtained at n = 5 (25 MHz) are plotted as a function of time. Although measurements were carried out up to the 13th overtone (65 MHz), only the fifth harmonic is shown in Figure 3 for clarity. Frequency and dissipation changes corresponding to the fifth, seventh and ninth overtones are reported in the Appendix (Figure A4.6). Three events can be identified in the transient evolution: (i) flow of ultrapure water to establish the baseline, (ii) flow of the AMP/ELR solution, (iii) rinsing with ultrapure water. when frequency changes are considered (Figure 4.3a), at the beginning of the deposition stage, the slope of the frequency change was slightly higher for GVC than for VC deposition. In addition, the frequency stabilized for GL13K and VC during the deposition stage, whereas a roughly linear decrease of frequency with time was observed for GVC. Specifically, at the end of the deposition stage, the frequency changes observed were −3.6, −12, and −18 Hz, for GL13K, VC, and GVC, respectively. During the final rinsing stage, a slight increase in frequency of between +1 and +2 Hz was observed for all the biomaterials, thus indicating a minor desorption of molecules. As far as dissipation is concerned (Figure 4.3b), a plateau was reached for GL13K and VC solutions during the deposition stage, whereas a slight slope in the dissipation was observed for GVC, similar to the frequency evolution. The rinsing stage resulted in a decrease in dissipation, and at the end of the rinsing stage a dissipation of close to zero was observed for the GL13K peptide, whereas values of around 1 × 10-6 and 3 × 10–6 were found for the ELRs GVC and VC, respectively. Development of self-assembled monolayers with antibiofilm activity based on AMPs and ELRs 111 IV Figure 4.3. (a) Normalized frequency and (b) energy dissipation shifts measured at 23 °C at the fifth (n = 5) overtone for GL13K, VC, and GVC. Three events are distinguished: first, an ultrapure water stabilization flow for 2 min; second, the AMP/ELR solutions were exposed for 25 min; and finally, a stage of rinsing with ultrapure water for 20 min. As a whole, the time evolution of the frequency change is similar for both ELRs (VC and GVC) and clearly differs from that for the peptide. The difference in molecular weights between these molecules may explain this behavior. Because overtones are split in terms of both frequency and dissipation changes (Figure A4.6), the simple Sauerbrey model is not valid, therefore a viscoelastic model that enables thin film areas and masses to be calculated from multiple harmonics is required. In this case, a Voigt viscoelastic model based on a single layer was used.[315,316] In this model, the adsorbed film is