Superchaotropicity, a New Mechanism for the Intracellular Delivery of Bioactive Molecules Using Boron Clusters

INTERNATIONAL DOCTORAL SCHOOL OF THE USC Giulia Salluce PhD Thesis Superchaotropicity, a New Mechanism for the Intracellular Delivery of Bioactive Molecules Using Boron Clusters Santiago de Compostela, 2023 Doctoral Programme in Chemical Science and Technology TESIS DE DOCTORADO SUPERCHAOTROPICITY, A NEW MECHANISM FOR THE INTRACELLULAR DELIVERY OF BIOACTIVE MOLECULES USING BORON CLUSTERS Giulia Salluce ESCUELA DE DOCTORADO INTERNACIONAL DE LA UNIVERSIDAD DE SANTIAGO DE COMPOSTELA PROGRAMA DE DOCTORADO EN CIENCIA Y TECNOLOGÍA QUÍMICA SANTIAGO DE COMPOSTELA 2023 DECLARACIÓN DEL AUTOR/A DE LA TESIS D./Dña. Giulia Salluce Título de la tesis: Superchaotropicity, a new mechanism for the intracellular delivery of bioactive molecules using boron clusters. Presento mi tesis, siguiendo el procedimiento adecuado al Reglamento y declaro que: 1) La tesis abarca los resultados de la elaboración de mi trabajo. 2) De ser el caso, en la tesis se hace referencia a las colaboraciones que tuvo este trabajo. 3) Confirmo que la tesis no incurre en ningún tipo de plagio de otros autores ni de trabajos presentados por mí para la obtención de otros títulos. 4) La tesis es la versión definitiva presentada para su defensa y coincide la versión impresa con la presentada en formato electrónico. Y me comprometo a presentar el Compromiso Documental de Supervisión en el caso que el original no esté depositado en la Escuela. En Santiago de Compostela, 17 de abril de 2023. Firma electrónica AUTORIZACIÓN DEL DIRECTOR / TUTOR DE LA TESIS Superchaotropicity, a new mechanism for the intracellular delivery of bioactive molecules using boron clusters. D. .......................Javier Montenegro García (director).......................................................................................... Dª. .....................Irene Lostalé Seijo (directora)....................................................................................................... D. .......................Juan R. Granja Guillán (tutor)........................................................................................................ INFORMAN: Que la presente tesis, se corresponde con el trabajo realizado por Dª. Giulia Salluce, bajo mi dirección/tutorización, y a utorizo su presentación , considerando que reúne l os r equisitos exigidos en el R eglamento de Estudios de Doctorado de la USC, y que como director de esta no incurre en las causas de abstención establecidas en la Ley 40/2015. De acuerdo con lo indicado en el Reglamento de Estudios de Doctorado, declara también que la presente tesis doctoral es idónea para ser defendida en base a la modalidad MONOGRÁFICA CON REPRODUCCIÓN DE PUBLICACIONES, en las que la participación de la doctoranda fue decisiva para su elaboración y las publicaciones se ajustan al Plan de Investigación. En Santiago de Compostela, 11 de abril de 2023 1 Table of Contents Table of Contents .............................................................................................. 1 Abbreviations and Acronyms ........................................................................... 7 Abstract ........................................................................................................... 11 Resumen .......................................................................................................... 17 Resumo ............................................................................................................ 26 INTRODUCTION ........................................................................................... 37 1. INTRACELLULAR TRANSPORT OF BIOLOGICALLY ACTIVE MOLECULES .............................................................................................. 38 1.1. The cellular membrane as a selective natural barrier .................... 38 1.2. Natural mechanisms of internalization ........................................... 40 1.3. Synthetic methods for intracellular delivery ................................... 42 1.3.1. Physical and mechanical methods .............................................. 44 1.3.2. Chemical methods ....................................................................... 45 1.3.2.1. Cell-penetrating peptides ..................................................... 45 1.3.2.2. Synthetic polymers ............................................................... 47 1.3.2.3. Lipids ..................................................................................... 49 1.3.2.4. Inorganic nanoparticles ........................................................ 50 1.3.2.5. Protein channels ................................................................... 50 1.4. The amphiphilicity paradigm ........................................................... 51 1.4.1. The counterion activation mechanism ....................................... 52 1.4.1.1. Anionic activators ................................................................. 54 1.4.1.2. Cationic amphiphiles ............................................................ 56 1.4.2. Limitations .................................................................................. 57 2. SUPERCHAOTROPIC BORON CLUSTERS ......................................... 59 2.1. Chaotropicity ..................................................................................... 59 2.2. Boron clusters .................................................................................... 61 2.2.1. Globular boron clusters properties ............................................. 65 2.2.2. Applications and therapeutic use of boron clusters ................... 67 2.2.3. Boron clusters as transmembrane carriers ................................ 68 GENERAL OBJECTIVE ................................................................................ 71 CHAPTER I ..................................................................................................... 73 1. OBJECTIVES .......................................................................................... 74 2 2. RESULTS & DISCUSSION .................................................................... 76 2.1. Carrier activity in model membranes ............................................... 76 2.2. Broadband carrier characteristics .................................................... 83 2.3. Membrane translocation in living cells ............................................ 88 2.4. New cargo types and biological activity enhancement .................... 92 3. CONCLUSION ...................................................................................... 101 4. MATERIALS & METHODS .................................................................. 103 4.1. Chemicals, peptides, and cell lines ................................................. 103 4.2. Vesicle preparation .......................................................................... 103 4.3. Transport experiments in HPTS/DPX vesicles .............................. 103 4.4. Activator efficiency .......................................................................... 104 4.5. Transport efficiency ......................................................................... 104 4.6. Leakage experiments in CF vesicles .............................................. 104 4.7. U-tube transport experiments ........................................................ 105 4.8. Isothermal Titration Calorimetry ................................................... 105 4.9. Dynamic Light Scattering ............................................................... 106 4.10. Cell culture and confocal imaging ................................................ 106 4.11. Cell viability assay ........................................................................ 106 4.12. Kanamycin A delivery in E. coli ................................................... 107 4.13. CRBN target engagement assay ................................................... 107 4.14. Cytosolic TAMRA-R8 concentration .............................................. 108 4.15. ICP-MS .......................................................................................... 109 4.16. Flow cytometry .............................................................................. 109 4.17. Synthesis and characterization of TAMRA-D-R8 ......................... 110 CHAPTER II ................................................................................................. 111 1. OBJECTIVES ........................................................................................ 112 2. RESULTS & DISCUSSION .................................................................. 115 2.1. Transmembrane transport in living cells ....................................... 115 2.1.1. Phalloidin–TRITC transport .................................................... 115 2.1.2. (KLAKLAK)2 peptide transport ................................................ 118 2.2. Cluster cytotoxicity ......................................................................... 119 3. CONCLUSION ...................................................................................... 121 4. MATERIALS & METHODS .................................................................. 123 4.1. Chemicals, peptides, and cell lines ................................................. 123 3 4.2. Synthesis of (KLAKLAK)2 ............................................................... 123 4.2.1. Solid-phase peptide synthesis ................................................... 123 4.2.2. Peptide cleavage and purification ............................................ 123 4.3. Confocal imaging ............................................................................. 124 4.4. Flow cytometry ................................................................................ 125 4.5. Cell viability assays ......................................................................... 125 CHAPTER III ................................................................................................ 127 1. OBJECTIVES ........................................................................................ 128 2. RESULTS & DISCUSSION .................................................................. 130 2.1. Internalization in living cells and intracellular localization ......... 130 2.2. Mechanism of internalization in living cells .................................. 134 2.2.1. Internalization in the presence of endocytosis inhibitors ....... 134 2.2.2. Endosomal disruption ............................................................... 137 2.2.3. Internalization with depolarized membrane ........................... 140 2.2.4. Cargo transport with depolarized membrane .......................... 143 3. CONCLUSION ...................................................................................... 144 4. MATERIALS & METHODS .................................................................. 146 4.1. Chemicals, peptides, and cell lines ................................................. 146 4.2. General protocol for confocal fluorescence imaging ....................... 146 4.2.1. Internalization of fluorescent clusters ..................................... 146 4.2.2. LysoTracker™ Green staining ................................................. 147 4.2.3. EGFP-hGal3 endosomal disruption assay in HeLa cells ......... 147 4.2.4. Phalloidin−TRITC transport with normal and depolarized membrane ............................................................................................ 147 4.3. Raman confocal imaging ................................................................. 148 4.4. General protocol for flow cytometry ............................................... 148 4.4.1. Internalization with endocytosis inhibitors ............................. 148 4.4.2. Endo-lysosomal disruption assay ............................................. 149 4.4.3. Internalization with normal to depolarized membrane .......... 149 4.4.4. Phalloidin−TRITC transport with normal and depolarized membrane ............................................................................................ 150 4.5. LysoTracker™ Green fluorescence intensity quantification ......... 150 GENERAL CONCLUSION .......................................................................... 151 REFERENCES. ............................................................................................. 153 4 Appendix I – Rights & Permissions. ............................................................ 175 Appendix II – List of publications derived from this Thesis. ...................... 185 5 12 In the first chapter of this thesis, we investigated halogenated dodecaborates as potential carriers for delivering charged and neutral compounds across hydrophobic barriers. The study involved experiments with artificial membranes and thermodynamic characterization, and the results showed that B12Br122– was the most effective non-covalent carrier across hydrophobic barriers. This cluster was able to transport a range of cargos, including positive, non-charged, and zwitterionic molecules, as well as larger polypeptides, except for negatively charged molecules. The boron clusters' carrier activity was found to be less restrictive than for amphiphilic activators, indicating that their activity is not limited to specific intermolecular interactions or salt bridges. Prompted by the success in vesicles, transport experiments in living cells were carried out. Confocal fluorescence microscopy was employed to study the potential of boron clusters to trigger membrane translocation of a model carboxytetramethylrhodamine-labelled R8 peptide (TAMRA-R8) in living cells. In line with expectation, peptide transport experiments in the absence of the clusters or in the presence of those clusters which displayed no or low activity in the vesicle models, namely B12H122– and B12Cl122–, resulted only in punctate fluorescence pointing to TAMRA-R8 trapped in endosomes. However, in the presence of B12Br122–, which was the most active in the liposomal experiments, diffuse peptide fluorescence was detected in the cytosol and the nucleus of the cells, which signalled the desired carrier activity in cells. The membrane-lytic iodinated cluster was also active but showed compromised cell morphology. Flow cytometry confirmed the enhancement of total TAMRA-R8 uptake in the presence of boron clusters in HeLa cells, with B12Br122– performing again as the most active carrier. MTT assay confirmed low cellular toxicity for the clusters in the transport experiments, even at 100 µM, except for the iodinated derivative. Since boron clusters showed a broad scope of accessible cargo types including neutral hydrophilic molecules, the possibility to deliver phalloidin, a rigid bicyclic heptapeptide that is long known in cell biology for its ability to bind to F-actin of the cytoskeleton, was investigated. Phalloidin-TRITC transport experiments with living HeLa cells showed that B12Br122– triggered its direct membrane passage to the cytosol and afforded an excellent staining of the F-actin target even at 500 nM cargo concentration. This protocol was transferable to other cell lines. In contrast, when attempting to use the prototypical octaarginine penetrating peptide AcR8 as an alternative non-covalent carrier for the same set of cell lines, only trace levels of cytosolic phalloidin were observed in all cell lines. 13 As a therapeutically relevant cargo, the transport of a proteolysis targeting chimera (PROTAC), and the antineoplastic monomethyl auristatin F (MMAF) were studied. We tested whether the use of boron clusters could contribute to enhance the activity of dBET1, a well-characterized PROTAC that is known for its undesirable low permeability and that should fall within the potential cargo scope of B12Br122–. For this bioactive cargo, B12Br122– was found to enhance by a factor of two the internalization of dBET1 in living cells. In the case of MMAF, the presence of the cluster improved the cytotoxic activity of this low-permeable auristatin. Antibiotics delivery is another area where novel carrier concepts are intensively being sought for, and vesicle studies had demonstrated transport of ampicillin and kanamycin A by the prototype chaotropic cluster carrier, B12Br122–. As a proof-of-principle, the potential to reduce the minimum inhibitory concentration of kanamycin A, an aminoglycoside antibiotic, with the brominated cluster was tested. In the absence of cluster, E. coli retained viability (60 %), but in the presence of the cluster carrier kanamycin A showed potent antibacterial activity (< 1 % viability). The combined transport experiments and the successful functional delivery of different bioactive cargos demonstrated that boron clusters, and prominently B12Br122–, are able to transport the intact agents through the cellular bilayer membrane, at physiologically relevant concentrations, and to induce the corresponding enhanced biological effects. Nevertheless, this avenue is still preliminary, especially for the delivery of (bio)macromolecules. In conclusion, in the first chapter, the concept of superchaotropic globular boron cluster anions as a chemically distinct class of membrane carriers has been introduced. The clusters obviate the traditional amphiphilic transport mechanism in that they operate by a direct chaotrope-mediated translocation. It has also been demonstrated that increasing the polarizability of the halogen substituents of the boron clusters in the series of halogens (Cl, Br, I), hence increasing the chaotropicity of the cluster, is key to control cargo-membrane interactions, and thus transport across the lipid bilayer. In the second chapter of this thesis, we wanted to investigate the effects of modifying the number of core boron atoms, and thus the number of substituents, using two groups of globular boron clusters: the decaand dodecaborate series. These sets present a different number of boron atoms in the core, respectively 10 and 12, hence each homologue cluster couple bears the same type of substituent but presents different size. 14 The first cargo used to investigate how the variation in the number of the core boron atoms affects their potential to trigger membrane translocation, was the neutral phalloidin-TRITC. By confocal microscopy and flow cytometry, it was observed that by increasing the size and polarizability of the substituents (H < Cl < Br < I), and thus, their chaotropicity, the transport capacity reached an optimal level. Above a certain degree of chaotropicity, cargo transportation diminished, and this was observed both for decaand dodecaborate cluster series. Then, a cationic model cargo was employed to validate the hypothesis on the correlation within size, chaotropicity, cytotoxicity, and transport activity: the proapoptotic peptide (KLAKLAK)2. Once more, B12Br122− showed the highest efficiency in peptide transport, visible by a considerable difference in cell viability between the samples incubated with and without the toxic peptide for cluster concentrations above 50 µM. In this case, B12Cl122− and B10I102− presented analogous results to the brominated cluster. MTT assay in HeLa cells treated with the clusters alone confirmed that a reduction in clusters’ size – from dodecato decaborate – induced an improvement in biocompatibility. In the case of the two iodinated clusters, a 2.5-fold increase in the IC50 was observed for B10I102−, compared to B12I122−. The inactive clusters were completely non-toxic, and the most active clusters – B12Br122− and B10I102− – presented an IC50 compatible with the concentrations required for intracellular delivery, making them suitable for transport across the cellular membrane. In conclusion, we demonstrated that a variation in the number of core boron atoms can be used to modulate the impact of the halogen substituents (Cl, Br, I), by modifying the shape and the extension of the polarizable surface of the cluster. Furthermore, the prevision that clusters with similar size and polarizability, and thus similar chaotropicity, present analogous transport capacity was confirmed. In fact, the transport activity of B10I102–, which presents a radius of 530 pm, was more akin to the one of B12Cl122– (r = 525 pm) or, surprisingly, B12Br122– (r = 560 pm) depending on the cargo and also on the cluster concentration. These results indicate the importance of both the size and the polarizability of superchaotropic boron clusters to modulate cargo-membrane interaction and, thus, intracellular delivery. The third and last chapter was focused on the biological aspects of the interaction between the boron clusters and the cellular environment. Internalization experiments in living cells showed that, despite the fact that negatively charged clusters should be repelled by anionic cell membranes, very small concentrations are required for cluster uptake (in the range of 1- 15 5 µM), and that, while hydrogenated clusters have a heterogeneous intracellular distribution, brominated clusters preferably accumulate in nuclei, with the exclusion of the nucleoli, where the abundance of negatively charged nucleic acids may repulse the superchaotropic anions. Regarding the internalization mechanism, vesicle experiments in the first part of this thesis suggested that superchaotropic boron clusters can cross lipidic membranes by direct translocation. In confocal microscopy experiments of living cells, the presence of fluorescent endosomes suggested the additional involvement of endocytic pathways. The results obtained with endocytic inhibitors pointed in the direction of a simultaneous combination of direct translocation and endocytic pathways. All together, these results indicated that anionic chaotropic clusters can be transported and diffuse across cell membranes due to their small size and membrane dynamic disrupting capabilities. The fact that the internalization of the most chaotropic cluster was less hindered by the presence of endocytic inhibitors, compared to the least chaotropic one, suggests that the more chaotropic the cluster is, the more it involves dynamic membrane disordering in its translocation process and can penetrate the cell regardless of the inhibition of endocytic pathways. In addition, endosomal membrane integrity was preserved after incubation with the cluster. This also makes possible that the cargo translocation and cytosolic release occurs, thanks to the superchaotropic carriers, without endosomal disruption. Furthermore, globular boron clusters are largely taken up by cells in the absence of membrane potential and can efficiently transport neutral hydrophilic cargos inside the cell in this condition. This behavior is opposite to the one described for up-to-date reported transmembrane carriers and might be important for the future design of superchaotropic carriers for hydrophobic drugs whose target cell presents low membrane potential. Overall, this research demonstrated that superchaotropic carriers could be the starting point of conceptually different cell biological, neurobiological, physiological, and pharmaceutical studies, thanks to their broad and distinct delivery spectrum, and to the completely new mechanism for cargo membrane translocation. 16 17 Resumen El transporte de pequeñas biomoléculas, ácidos nucleicos (ADN y ARN), proteínas, nanomateriales sintéticos y fármacos al interior de células vivas es un aspecto crucial en las ciencias de la vida y ha permitido décadas de avances biomédicos. También suscita cada vez más interés en aplicaciones industriales y médicas. En los campos de las técnicas de imagen y la administración de fármacos, el transporte intracelular ha cobrado especial relevancia porque permite que las moléculas de interés alcancen sus dianas específicas dentro la célula. Sin embargo, la membrana celular actúa como una barrera, permitiendo el paso de algunas moléculas e impidiendo la entrada de otras. La permeabilidad selectiva de la membrana celular depende de una combinación de factores, como la composición de la propia membrana y la presencia de proteínas de transporte especializadas. La bicapa lipídica de la membrana crea una barrera permeable a la mayoría de las moléculas hidrofóbicas, como lípidos y esteroides, al tiempo que impide el paso de moléculas hidrófilas, como iones y pequeñas moléculas polares. El transporte de moléculas exógenas al interior de las células puede realizarse mediante diferentes mecanismos que, en general, pueden clasificarse en difusión pasiva, difusión facilitada y transporte activo, requiriendo este último el consumo de energía por parte de la célula. Un tipo de transporte activo especialmente importante es la endocitosis. Este fenómeno implica una modificación conformacional de la membrana plasmática para internalizar sustancias extracelulares mediante la formación de vesículas. Algunas moléculas de interés biológico o terapéutico no son capaces de atravesar por sí mismas la membrana celular o ser internalizadas sin quedar atrapadas en el interior del endosoma e incurrir en degradación lisosomal. Por este motivo, el desarrollo de métodos eficaces y seguros para el transporte al interior celular de moléculas bioactivas ha cobrado gran interés en los últimos tiempos. En las últimas décadas se han desarrollado una gran variedad de métodos físicos y químicos para el transporte intracelular de moléculas de interés biológico y terapéutico. Dentro de los métodos químicos, se encuentra toda una diversidad de moléculas transportadoras llamadas “carriers”. Estos pueden ser, entre otros, péptidos penetrantes de células, lípidos catiónicos, o nanopartículas. 18 La búsqueda de carriers sintéticos para transportar moléculas que no atravesarían de forma natural y espontánea la membrana celular ha recibido gran atención en las últimas décadas. En esta búsqueda ha prevalecido hasta ahora un enfoque conceptual: los investigadores se han centrado especialmente en el diseño de moléculas anfifílicas, con una parte hidrofóbica, que sirve de anclaje a la membrana, y una parte iónica, que atrae electrostáticamente el cargo para activar el transporte a través de la membrana de los complejos resultantes con carga neutra. El potencial del dominio de transducción del transactivador de la transcripción del virus de la inmunodeficiencia humana tipo 1 (VIH-1 Tat) para transportar cualquier conjugado a través de las membranas celulares ha atraído una considerable atención científica desde finales de los años 90. A partir de entonces se han desarrollado muchos péptidos penetrantes de células ricos en arginina basados en la secuencia viral mencionada. La capacidad de estos péptidos penetrantes de células ricos en guanidinio para atravesar las membranas celulares se ha explicado con el concepto de "bifilicidad dinámica". Este término se refiere a moléculas que pueden ser tanto hidrofílicas como lipofílicas, según las circunstancias, y que, a diferencia de las moléculas anfifílicas, pueden adaptarse a cualquier entorno. Se ha observado que cuando las poliargininas se unen a contraaniones hidrofóbicos, se comportan como moléculas lipofílicas y prefieren naturalmente el cloroformo al agua neutra. Sin embargo, la sustitución de los contraaniones hidrofóbicos por otros hidrofílicos convierte la hidrofobicidad de las poliargininas en hidrofilia. Las bicapas lipídicas de las membranas proporcionan una complejidad ambiental ideal para estos camaleones moleculares adaptables como las oligoy poliargininas. Estas propiedades pueden aprovecharse para el transporte a través de la membrana celular. De hecho, mediante la adición de una mezcla sinérgica de aniones hidrofílicos y anfifílicos, los complejos poliarginina-anión pueden mediar en la translocación de moléculas aniónicas a través de membranas artificiales y celulares. Para transportar moléculas polianiónicas más complejas, como los ácidos nucleicos, se han desarrollado cientos de vectores que evitan los problemas asociados al uso de virus recombinantes. La mayoría de los vectores artificiales propuestos comparten esta característica: la presencia de una porción catiónica que permite la unión a los ácidos nucleicos polianiónicos, y una parte hidrofóbica que da lugar al autoensamblaje de los anfifilos en estructuras micelares o en dobles capas alrededor del material genético para proteger el cargo de la degradación por nucleasas, y mejorar la permeación celular. 19 En general, los avances en el campo de la liberación intracelular han sido rápidos y amplios. Sin embargo, los transportadores anfifílicos desarrollados hasta ahora presentan algunas limitaciones intrínsecas debidas a su tendencia a la agregación, la unión inespecífica, el atrapamiento endocítico, la propensión a romper la membrana celular y la citotoxicidad. Para superar estos problemas, en esta tesis doctoral se presentan a los clústeres globulares de boro aniónicos como una nueva clase de transportadores directos a través de la membrana que abandonan la clásica topología anfifílica. Los clústeres de boro son especies inorgánicas, contrariamente a los transportadores desarrollados hasta ahora, y su afinidad por la membrana no se debe a la anfifilicidad, sino a que son supercaotrópicos. En consecuencia, y a diferencia de los transportadores de membrana convencionales, los clústeres de boro son muy solubles en agua y no se encapsulan ni forman agregados con su cargo. Además de estas diferencias químicas y físicas fundamentales, descubrimos que los clústeres supercaotrópicos también permiten el transporte directo a través de la membrana de una amplia gama de cargos hidrofílicos. Los clústeres globulares de boro de interés en este estudio son aniones estables con una doble carga negativa permanente, propiedades ligantes débiles y alta biocompatibilidad, ya que previamente se han empleado, entre otras aplicaciones, en la terapia contra el cáncer por captura de neutrones por el boro. Recientemente, hemos identificado que dodecaboratos del tipo B12X122– (X = H, Cl, Br e I) son supercaotrópicos, es decir, sus propiedades caotrópicas superan a las de los aniones más caotrópicos en la escala de Hofmeister. Por lo tanto, en una escala continua de solvatación acuosa, sus propiedades empiezan a parecerse a las de las especies hidrofóbicas iónicas y, a pesar de las grandes diferencias estructurales y fisicoquímicas, empiezan a mostrar una propensión genérica a asociarse dinámicamente a zonas hidrofóbicas, incluidas las bicapas lipídicas. Estas similitudes descubiertas recientemente nos llevaron a plantear la hipótesis de que los clústeres de boro, y potencialmente otros iones grandes con carácter supercaotrópico, podrían activar el transporte directo a través de la membrana de cargos moleculares hidrofílicos. En el primer capítulo de esta tesis, se estudiaron los dodecaboratos halogenados, lo que permitió identificar un buen candidato, mejor que los competidores anfifílicos, para el transporte no sólo de cargos catiónicos, sino también de compuestos neutros. En primer lugar, se llevaron a cabo experimentos en membranas artificiales y la caracterización termodinámica de la interacción de los clústeres de boro con diferentes cargos. El ensayo de liberación de las vesículas y los experimentos de transporte en tubo en U, a través de una capa de cloroformo, confirmaron inequívocamente que 20 mientras que el clúster yodado B12I122– (el más grande y caotrópico) era membranolítico, y los clústeres B12H122– (el más pequeño y menos caotrópico) y B12Cl122– eran poco o nada activos; el clúster bromado B12Br122–, resultó ser el mejor de esta serie y se comportó como un excelente transportador no covalente a través de barreras hidrofóbicas artificiales. Una vez que el clúster bromado emergió como el mejor exponente de una nueva clase de transportadores de membrana sintéticos altamente activos, analizamos el rango de los compuestos que pueden ser transportados, centrándonos en moléculas de interés biológico impermeables a la membrana. Entre los compuestos evaluados se incluyeron biomoléculas con cargas diferentes (como la acetilcolina y los aminoácidos), vitaminas, antibióticos, agentes bloqueantes neuromusculares y proteínas. Sorprendentemente, el clúster B12Br122– transportó muchos tipos de cargo, desde moléculas positivas a no cargadas y zwitteriónicas, o desde moléculas pequeñas como la acetilcolina a polipéptidos más grandes como la protamina, con la excepción de las moléculas cargadas negativamente como el glutamato y la albúmina, para las que no puede producirse atenuación de la repulsión cargo-transportador. La insensibilidad a la naturaleza química de los diversos grupos funcionales de los cargos transportados confirmó que la actividad transportadora de los clústeres de boro no se limita a residuos que establezcan puentes salinos o interacciones intermoleculares específicas, y es mucho menos restrictiva que en el caso de los activadores anfifílicos. Impulsados por el éxito de los experimentos en vesículas, decidimos llevar a cabo estudios de transporte en células vivas. Se empleó la microscopía confocal de fluorescencia para estudiar el potencial de los clústeres de boro para activar la translocación de membrana de un péptido modelo: una octaarginina marcada con carboxitetrametilrodamina (TAMRA-R8). En línea con lo esperado, los experimentos de transporte del péptido a baja concentración en ausencia de los clústeres o en presencia de aquellos clústeres que mostraron una actividad nula o baja en los modelos de vesículas, es decir B12H122– y B12Cl122–, sólo mostraron fluorescencia punteada en las micrografías confocales, lo que indicaba que el péptido marcado había quedado atrapado en los endosomas. Sin embargo, en presencia del clúster B12Br122–, que había resultado ser el más activo en los experimentos en vesículas liposomales, se detectó fluorescencia difusa del péptido en el citosol y en el núcleo de las células, lo que señalaba la presencia de una eficiente actividad transportadora en las células. El clúster B12I122– también resultó activo pero las células mostraban una morfología comprometida, lo que indicaba que el clúster yodado había sido tóxico para las células. Mediante citometría de flujo se confirmó el aumento de la internalización total del péptido marcado en presencia de clústeres de boro 21 en células HeLa, siendo B12Br122– el transportador más activo. El ensayo de viabilidad celular MTT confirmó la baja citotoxicidad de los clústeres en los experimentos de transporte, incluso a 100 µM, excepto para el derivado yodado. Dado que los clústeres de boro mostraban una amplia gama de tipos de cargos accesibles, incluyendo moléculas hidrofílicas neutras, investigamos la posibilidad de transportar faloidina, un heptapéptido bicíclico rígido conocido desde hace tiempo en biología celular por su capacidad de unirse a la actina F del citoesqueleto. Los experimentos de transporte de faloidinaTRITC, una faloidina marcada con el fluoróforo tetrametilrodamina, en células HeLa vivas demostraron que el clúster B12Br122– promovía el paso directo de la faloidina al citosol a través de la membrana y proporcionaba una excelente tinción de la actina F incluso a una concentración de cargo de 500 nM. Este protocolo era transferible a otras líneas celulares. Por el contrario, cuando se intentó utilizar un conocido péptido penetrante, i.e. la octaarginina, como transportador no covalente alternativo para el mismo conjunto de líneas celulares, sólo se observaron niveles extremadamente bajos de faloidina citosólica en todas las líneas celulares testadas. Las quimeras dirigidas a la proteólisis (PROTACs) son pequeñas moléculas con un brillante futuro como la próxima generación de fármacos para la eliminación de proteínas específicas no deseadas. Probamos si el uso de clústeres de boro podría contribuir a mejorar la actividad de dBET1, una PROTAC bien caracterizada que es conocida por su indeseable baja permeabilidad y que debería caer dentro del ámbito de potencial cargo transportado por el clúster bromado. Para este cargo bioactivo, se descubrió que el clúster B12Br122– reducía eficazmente el valor de IC50 de un ensayo competitivo en células en más de un factor de 2. El antineoplásico monometil auristatina F también vio su actividad citotóxica potenciada en presencia del clúster. La administración de antibióticos es otro campo en el que se buscan intensamente nuevos transportadores, y nuestros estudios en vesículas habían demostrado la eficacia del transporte de ampicilina y kanamicina A por el prototipo de transportador de clúster caotrópico, B12Br122–. Por ello, investigamos el potencial de este clúster para reducir la concentración inhibitoria mínima de la kanamicina A, un antibiótico aminoglucósido. En ausencia del clúster, E. coli conservó la mayor parte de la viabilidad (60 %), pero, en presencia del clúster transportador, la kanamicina A mostró una potente actividad antibacteriana, llevando la viabilidad de E. coli a menos del 1 %. 28 presentan algunhas limitacións intrínsecas debidas á súa tendencia á agregación, a unión inespecífica, o atrapamento endocítico, a propensión para romper a membrana celular e a citotoxicidade. Para superar estes problemas, nesta tese doutoral presentánse clústeres globulares de boro aniónicos como unha nova clase de transportadores directos de membrana que abandonan a clásica topoloxía anfifílica. Os clústeres de boro son especies inorgánicas, contrariamente aos transportadores desenvolvidos ata o de agora, e a súa afinidade pola membrana non se debe á anfifilicidade, senón a que son supercaotrópicos. En consecuencia, e a diferenza dos transportadores de membrana convencionais, os clústeres de boro son moi solubles en auga e non se encapsulan nin forman agregados co seu cargo. Ademais destas diferenzas químicas e físicas fundamentais, descubrimos que os clústeres supercaotrópicos tamén permiten o transporte directo a través da membrana dunha ampla gama de cargos hidrófilos. Os clústeres globulares de boro de interese neste estudo son anións estables cunha dobre carga negativa permanente, propiedades ligantes débiles e alta biocompatibilidade, xa que foron empregados anteriormente na terapia contra o cancro por captura de neutróns polo boro, entre outros. Recentemente, identificamos que dodecaboratos do tipo B12X122– (X = H, Cl, Br e I) son supercaotrópicos, é dicir, as súas propiedades caotrópicas superan ás dos anións máis caotrópicos na escala de Hofmeister. Por tanto, nunha escala continua de solvatación acuosa, as súas propiedades empezan a parecerse ás das especies hidrófobas iónicas e, a pesar das grandes diferenzas estruturais e fisicoquímicas, empezan a mostrar unha propensión xenérica a asociarse dinámicamente a zonas hidrófobas, incluídas as bicapas lipídicas. Estas similitudes descubertas recentemente leváronnos a expor a hipótese de que os clústeres de boro, e potencialmente outros ións grandes con carácter supercaotrópico, poderían activar o transporte directo a través da membrana de cargos moleculares hidrófilos. No primeiro capítulo desta tese, estudáronse os dodecaboratos haloxenados, o que permitiu identificar un bo candidato, mellor que os competidores anfifílicos, para o transporte non só de cargos catiónicos con carga oposta, senón tamén de compostos neutros. En primeiro lugar, leváronse a cabo experimentos en membranas artificiais e a caracterización termodinámica da interacción dos clústeres de boro con diferentes cargos. O ensaio de liberación das vesículas e os experimentos de transporte en tubo en U, a través dunha capa de cloroformo, confirmaron inequivocamente que mentres 29 que o clúster iodado B12I122– (o máis grande e caotrópico) era membranolítico, os clústeres B12H122– (o máis pequeno e menos caotrópico) e B12Cl122– eran pouco ou nada activos; o clúster bromado, B12Br122–, resultou ser o mellor desta serie e comportouse como un excelente transportador non covalente a través de barreiras hidrófobas artificiais. Unha vez que o clúster bromado emerxeu como o mellor expoñente dunha nova clase de transportadores de membrana sintéticos altamente activos, analizamos o rango dos compostos que poden ser transportados, centrándonos en moléculas de interese biolóxico impermeables á membrana. Entre os compostos avaliados incluíronse biomoléculas con cargas diferentes (como a acetilcolina e os aminoácidos), vitaminas, antibióticos, axentes bloqueantes neuromusculares e proteínas. Sorprendentemente, o clúster B12Br122– transportou moitos tipos de cargo, desde moléculas positivas a non cargadas e zwitteriónicas, ou desde moléculas pequenas como a acetilcolina a polipéptidos máis grandes como a protamina, coa excepción das moléculas cargadas negativamente como o glutamato e a albumina, para as que non se pode producir atenuación da repulsión cargo-transportador. A insensibilidade á natureza química dos diversos grupos funcionais dos cargos transportados confirmou que a actividade transportadora dos clústeres de boro non se limita a residuos que establezan pontes salinos ou interaccións intermoleculares específicas, e é moito menos restritiva que no caso dos activadores anfifílicos. Impulsados polo éxito dos experimentos en vesículas, decidimos levar a cabo estudos de transporte en células vivas. Empregouse a microscopía confocal de fluorescencia para estudar o potencial dos clústeres de boro para activar a translocación de membrana dun péptido modelo: unha octaarxinina marcada con carboxitetrametilrodamina (TAMRA-R8). En liña co esperado, os experimentos de transporte do péptido a baixa concentración en ausencia dos clústeres ou en presenza daqueles clústeres que mostraron unha actividade nula ou baixa nos modelos de vesículas, é dicir B12H122– e B12Cl122–, só mostraron fluorescencia punteada nas micrografías confocais, o que indicaba que o péptido marcado quedara atrapado nos endosomas. Con todo, en presenza do clúster B12Br122–, que resultara ser o máis activo nos experimentos en vesículas liposomais, detectouse fluorescencia difusa do péptido no citosol e no núcleo das células, o que sinalaba a presenza dunha eficiente actividade transportadora nas células. O clúster B12I122– tamén resultou activo pero as células mostraban unha morfoloxía comprometida, o que indicaba que o clúster iodado fora tóxico para as células. A citometría de fluxo confirmou o aumento da internalización total do péptido marcado en presenza de clústeres de boro en células HeLa, sendo 30 B12Br122– o transportador máis activo. O ensaio de viabilidade celular MTT confirmou a baixa citotoxicidade dos clústeres nos experimentos de transporte, mesmo a 100 µM, excepto para o derivado iodado. Dado que os clústeres de boro mostraban unha ampla gama de tipos de cargos accesibles, incluíndo moléculas hidrófilas neutras, investigamos a posibilidade de transportar faloidina, un heptapéptido bicíclico ríxido coñecido desde hai tempo en bioloxía celular pola súa capacidade de unirse á actina F do citoesqueleto. Os experimentos de transporte de faloidinaTRITC, unha faloidina marcada co fluoróforo tetrametilrodamina, en células HeLa vivas demostraron que o clúster B12Br122– promovía o paso directo da faloidina ao citosol a través da membrana e proporcionaba unha excelente tinción da actina F mesmo a unha concentración de cargo de 500 nM. Este protocolo era transferible a outras liñas celulares. Pola contra, cando se tentou utilizar un coñecido péptido penetrante, i.e. a octaarxinina, como transportador non covalente alternativo para o mesmo conxunto de liñas celulares, só se observaron niveis extremadamente baixos de faloidina citosólica en todas as liñas celulares testadas. As quimeras dirixidas á proteólise (PROTACs) son pequenas moléculas cun brillante futuro como a próxima xeración de fármacos para a eliminación de proteínas específicas non desexadas. Probamos se o uso de clústeres de boro podería contribuír a mellorar a actividade de dBET1, unha PROTAC ben caracterizada que é coñecida pola súa indesexable baixa permeabilidade e que debería caer dentro do ámbito de potencial cargo transportado polo clúster bromado. Para este cargo bioactivo, descubriuse que o clúster B12Br122– reducía eficazmente o valor de IC50 nun ensaio competitivo en células en máis dun factor de 2. O antineoplásico monometil auristatina F tamén viu a súa actividade citotóxica potenciada en presenza do clúster. A administración de antibióticos é outro campo no que se buscan intensamente novos transportadores, e os nosos estudos en vesículas demostraran a eficacia do transporte de ampicilina e kanamicina A por o prototipo de transportador de clúster caotrópico, B12Br122–. Por iso, investigamos o potencial deste clúster para reducir a concentración inhibitoria mínima da kanamicina A, un antibiótico aminoglucósido. En ausencia do clúster, E. coli conservou a maior parte da viabilidade (60 %), pero, en presenza do clúster transportador, a kanamicina A mostrou unha potente actividade antibacteriana, levando a viabilidade de E. coli a menos do 1 %. A combinación do éxito do transporte intracelular e o mantemento da funcionalidade de diferentes cargos bioactivos demostraron que os clústeres de boro, e principalmente o clúster bromado, son capaces de transportar as 31 moléculas de interese intactas e funcionais a través da bicapa lipídica da membrana celular. Con todo, esta tecnoloxía atópase nun estado aínda preliminar, especialmente para o transporte de macromoléculas de maior complexidade. En conclusión, no primeiro capítulo presentamos os clústeres de boro globulares supercaotrópicos como unha clase quimicamente distinta de transportadores de membrana. Os clústeres obvian o mecanismo tradicional de transporte baseado na anfifilicidade, xa que operan mediante unha translocación directa mediada pola caotropicidade propia destas moléculas. Tamén demostramos que aumentar a polarizabilidade dos substituíntes halóxenos dos clústeres de boro na serie de halóxenos (Cl, Br, I), aumentando desta maneira a caotropicidade do clúster, é fundamental para controlar as interaccións cargo-membrana e, por tanto, o transporte a través da bicapa lipídica. No segundo capítulo desta tese, a investigación centrouse nos efectos que podía producir unha modificación do número de átomos do núcleo do clúster, e por tanto o número de substituíntes, mediante o estudo de dous grupos de clústeres globulares de boro: as series de decaboratos e dodecaboratos. Estas series presentan diferente número de átomos de boro no núcleo, respectivamente 10 e 12, por iso é polo que cada parella de clústeres homólogas leva o mesmo tipo de substituíntes pero presenta diferente tamaño. O primeiro cargo utilizado para investigar como a variación no número de átomos de boro do núcleo afecta o impacto dos substituíntes halóxenos no seu potencial para promover a translocación de membrana, foi de novo a faloidina marcada con tetrametilrodamina. Mediante microscopía confocal e citometría de fluxo, observouse que ao aumentar a polarizabilidade dos substituíntes (H < Cl < Br < I) e, en consecuencia, a súa caotropicidade, a capacidade de transporte alcanzaba un nivel óptimo, cunha eficiencia practicamente equivalente para o clúster B12Br122– á máxima concentración testada, e o clúster B10I102– á concentración máis baixa. Por encima dun certo grao de caotropicidade, o transporte do cargo diminuía, e isto observouse tanto para os clústeres decaboratos como para os dodecaboratos. Tamén se observou que a eficacia do transporte depende da concentración: por baixo do nivel óptimo de caotropicidade, os clústeres máis pequenos requiren concentracións máis altas para unha entrega eficaz do cargo; máis aló dese punto óptimo, as concentracións altas conducen a unha entrega menos eficaz. Outro cargo empregado como modelo para validar a hipótese sobre a correlación entre tamaño, caotropicidade, citotoxicidade e actividade de 32 transporte foi o péptido proapoptótico (KLAKLAK)2. Unha vez máis, B12Br122– demostrou ser o clúster máis eficiente no transporte deste péptido, o que se manifestou como unha considerable diferenza na viabilidade celular entre as mostras incubadas con e sen o péptido tóxico en presenza de concentracións de clúster superiores a 50 µM. Neste ensaio, os clústeres B12Cl122– e B10I102– presentaron tamén resultados excelentes, análogos aos de B12Br122–. O ensaio MTT en células HeLa, tratadas con concentracións crecentes de clústeres para determinar o IC50 de cada anión globular considerado neste capítulo, confirmou que a redución do tamaño dos clústeres – de dodecaboratos a decaboratos – inducía unha melloría na biocompatibilidade. No caso da parella de clústeres homólogos iodados, observouse unha diminución de 2.5 veces no IC50 para B10I102–, en comparación con B12I122–. Os clústeres inactivos eran completamente atóxicos, e os clústeres máis activos – B12Br122– e B10I102– – presentaban unha IC50 compatible coas concentracións requiridas para o transporte ao interior celular, o que os facía adecuados para o transporte a través da membrana en ensaios in vitro e, potencialmente, in vivo. En conclusión, demostramos que unha variación no número de átomos de boro no núcleo pode utilizarse para modular o impacto dos substituíntes halóxenos (Cl, Br, I), modificando a forma e a extensión da superficie polarizable do clúster. Ademais, demostrouse que os clústeres con tamaño e polarizabilidade semellantes, e por tanto unha caotropicidade similar, presentan unha capacidade de transporte similar, aínda que outros factores poden contribuír a este comportamento, pero non de forma tan previsible como se supoñía ao principio. De feito, a actividade de transporte de B10I102–, que presenta un radio de 530 pm, era máis parecida á de B12Cl122– (r = 525 pm) ou, sorprendentemente, á de B12Br122– (r = 560 pm, considerablemente máis grande que B10I102–) dependendo da molécula a transportar e tamén da concentración do clúster utilizada. Estes resultados indican a importancia do tamaño e da polarizabilidade dos clústeres de boro supercaotrópicos para modular a interacción cargomembrana e, por tanto, controlar a eficiencia da actividade de transporte intracelular. No terceiro e último capítulo, centrámonos nos aspectos biolóxicos da interacción entre os clústeres de boro e a membrana celular. Os experimentos de internalización en células vivas mostraron que, a pesar de que os clústeres, cargados negativamente, deberían ser repelidos polas membranas celulares aniónicas, bastaban concentracións moi pequenas 33 para a internalización dos mesmos nas células (no rango de 1-5 µM). Ademais, mentres que os clústeres hidroxenados tiñan unha distribución intracelular heteroxénea, os clústeres bromados acumulábanse preferentemente nos núcleos, con exclusión dos nucléolos, onde a abundante presenza de ácidos nucleicos cargados negativamente pode repeler aos anións supercaotrópicos. En canto ao mecanismo de internalización, os experimentos con vesículas realizados na primeira parte deste proxecto indicaron que os clústeres de boro supercaotrópicos poden atravesar membranas lipídicas por translocación directa. Por outra parte, a presenza de endosomas fluorescentes nos experimentos de microscopía confocal suxeriu a participación de vías endocíticas no proceso de translocación ao interior da célula. Os resultados obtidos en experimentos con inhibidores da endocitose apuntaron na dirección dunha combinación simultánea de translocación directa e endocitose. En conxunto, estes resultados indicaron que os clústeres caotrópicos aniónicos poden transportarse e difundir a través das membranas celulares grazas ao seu pequeno tamaño e á súa capacidade de alterar de forma dinámica a membrana. O feito de que a internalización do clúster máis activo se vise menos afectada pola presenza de inhibidores da endocitose, en comparación co inactivo, suxire que canto máis caotrópico é o clúster, máis implicada vese a alteración dinámica da membrana no proceso de translocación do clúster e leste pode penetrar na célula independentemente da inhibición das vías endocíticas. Por outra parte, a incubación das células con clústeres non afectou á integridade da membrana endosomal. Isto tamén fai posible o feito de que a internalización do cargo e a liberación no citosol se produzan sen rotura dos endosomas. Ademais, os clústeres globulares de boro, en ausencia de potencial de membrana, se internalizan nas células en maior medida, e poden transportar eficazmente cargos hidrófilos neutros ao interior da célula nestas condicións. Este comportamento é oposto ao descrito para transportadores de membrana descritos ata a data, como no caso dos péptidos penetrantes de células e outros transportadores anfifílicos. Isto podería ser importante para o futuro deseño de carriers supercaotrópicos para fármacos hidrófobos cuxa diana celular se caracterice por baixo potencial de membrana. En xeral, con esta investigación quixemos demostrar que os carriers supercaotrópicos poderían ser o punto de partida de estudos biolóxicos celulares, neurobiolóxicos, fisiolóxicos e farmacéuticos conceptualmente diferentes, grazas ao seu espectro amplo e distinto de cargos que poden 34 transportar e ao mecanismo completamente novo grazas ao cal permiten a translocación do cargo a través da membrana. 35 36 37 INTRODUCTION 44 Figure 3 | Schematic representation of the most common methods for intracellular cargo delivery. Membrane disruption-based methods, such as permeabilization and penetration, are physical approaches that induce the formation of transient pores into the plasma membrane for cargo internalization. Other methods use biochemical approaches for membrane permeabilization and cargo translocation (e.g., detergents, pore-forming proteins). As an alternative, cargoes can access the intracellular compartment through carrier-mediated delivery. The carrier/cargo complex can be internalized via endocytosis or by membrane fusion depending on its chemical or biological nature. Created with BioRender.com, adapted with permission from ref. 10 Copyright 2016 Macmillan Publishers Limited, part of Springer Nature. 1.3.1. Physical and mechanical methods One of the first methods used for the delivery inside the cell of exogenous compounds, especially macromolecules, was microinjection, which makes use of a glass capillary to transfer proteins, antibodies, or other macromolecules across the cell membrane or even the nuclear membrane.11 This technique is suitable when precise control of timing and amount of cargo to be delivered is crucial, but it is not adequate for the delivery of macromolecules in vivo or into a large number of cells or in living systems. Electroporation consists in applying voltage pulses to create transient pores in the membrane that, hence, can be crossed by exogenous molecules.12 The main advantage is that this technique can be applied to all cell types; though, sensitive cells cannot survive the treatment and the use of electrodes makes electroporation of limited use for in vivo applications.13 However, electroporation has been exploited for intradermal or intramuscular delivery of DNA vaccines against SARS-CoV-214 and for in vivo delivery of plasmid DNA to produce anti-viral monoclonal antibodies.15 11 Zhang, Y.; Yu, L.-C. BioEssays 2008, 30, 606–610. 12 Shi, J.; Ma, Y.; Zhu, J.; Chen, Y.; Sun, Y.; Yao, Y.; Yang, Z.; Xie, J. Molecules 2018, 23, 3044. 13 Torchilin, V. Drug Discov. Today Technol. 2008, 5, e95–e103. 14 Xu, Z.; Patel, A.; Tursi, N. J.; Zhu, X.; Muthumani, K.; Kulp, D. W.; Weiner, D. B. Front. Med. Technol. 2020, 2, 571030. 15 Pagant, S.; Liberatore, R. A. Pharmaceutics 2021, 13 (11), 1882. 45 Similarly, in optoporation, the plasma membrane is permeabilized through light pulses.16 Cell membrane can also be temporarily permeabilized by using a buffered solution containing 25 % ethanol,17 or a hyperosmolar NaCl solution combined with propanebetaine, a transduction compound. The latter method, named iTOP (induced transduction by osmocytosis and propanebetaine), provokes macropinocytotic uptake and intracellular release of the molecule of interest even in difficult-to-manipulate primary cells.18 Another method to internalize molecules inside the cells is microfluidics, in which cells are forced to pass through a hole which is smaller than the diameter of the cell itself, thus inducing fast deformation of the membrane and facilitating passive diffusion of substances dissolved in the eluting buffer. This high throughput technique has been proven to efficiently deliver different cargos to immune and stem cells (up to 20,000 cells/s).19 1.3.2. Chemical methods 1.3.2.1. Cell-penetrating peptides Cell-penetrating peptides (CPPs) are short sequences of natural and/or artificial amino acids that are able to cross the cell membrane. Since the discovery of the potential of the human immunodeficiency virus type 1 transactivator of transcription (HIV-1 Tat) transduction domain to transport any conjugate across cell membranes in the late ‘90s,20 CPPs have emerged as a new approach for the intracellular delivery of various therapeutic and diagnostic cargos such as drugs, imaging agents, nucleic acids, proteins, etc. They usually present low cytotoxicity and require low concentrations for the transport activity (micromolar range).21 The mechanism of uptake of CPPs has been investigated in many studies, and this topic has generated controversy over the years. Generally, these peptides are translocated inside the cell through different endocytic 16 Vogel, A.; Noack, J.; Hüttman, G.; Paltauf, G. Appl. Phys. B 2005, 81, 1015–1047. 17 O’Dea, S.; Annibaldi, V.; Gallagher, L.; Mulholland, J.; Molloy, E. L.; Breen, C. J.; Gilbert, J. L.; Martin, D. S.; Maguire, M.; Curry, F.-R. PLoS One 2017, 12, e0174779. 18 D’Astolfo, D. S.; Pagliero, R. J.; Pras, A.; Karthaus, W. R.; Clevers, H.; Prasad, V.; Lebbink, R. J.; Rehmann, H.; Geijsen, N. Cell 2015, 161, 674–690. 19 Sharei, A.; Zoldan, J.; Adamo, A.; Sim, W. Y.; Cho, N.; Jackson, E.; Mao, S.; Schneider, S.; Han, M. J.; Lytton-Jean, A. et al. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 2082–2087. 20 Vivès, E.; Brodin, P.; Lebleu, B. J. Biol. Chem. 1997, 272 (25), 16010–16017. 21 Heitz, F.; Morris, M. C.; Divita, G. Br. J. Pharmacol. 2009, 157, 195–206. 46 pathways, and the internalization rate is influenced by the concentration and type of CPP, as well as the characteristics of the cargo.22 Thus far, a wide variety of CPPs has been reported and patented, and, among other kinds of classification, CPPs can be divided according to their physical-chemical features into the following classes:23 - Cationic CPPs: characterized by the presence of mainly cationic residues such as arginine and lysine, which provides a net positive charge to the CPP. Some examples in this category are TAT,20 penetratin,24 polyarginines,25,26 or polylysines.27 - Hydrophobic CPPs: in most cases, derived from signal peptide sequences and containing apolar residues or a very low charge. To date, only a few hydrophobic CPPs, including SG3,28 Pep-7,29 and a fragment of fibroblast growth factor 12,30 have been reported.31 - Amphipathic CPPs: they may be chimeric peptides that contain two different regions: a hydrophobic domain and a hydrophilic domain,32 or acquire an amphiphilic character thanks to their secondary structure (secondary amphipathic peptides).33,34 These CPPs usually form stable noncovalent complexes with different types of cargo leading to intracellular delivery of the cargo. The amphipathic peptide Pep-1, for example, can form non-covalent interactions with other peptides and protein cargos to allow their membrane translocation.35 Another example is MPG, a peptide that 22 Rioboo, A.; Gallego, I.; Montenegro, J. An. Química 2019, 115, 9–21. 23 Milletti F. Drug Discov. Today 2012, 17 (15–16), 850–860. 24 Joliot, A.; Pernelle, C.; Deagostini-Bazin, H.; Prochiantz, A. Proc. Natl. Acad. Sci. U. S. A. 1991, 88, 1864–1868. 25 Wender, P. A.; Mitchell, D. J.; Pattabiraman, K.; Pelkey, E. T.; Steinman, L.; Rothbard, J. B. Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 13003–13008. 26 Futaki, S.; Suzuki, T.; Ohashi, W.; Yagami, T.; Tanaka, S.; Ueda, K.; Sugiura, Y. J. Biol. Chem. 2001, 276, 5836–5840. 27 Mai, J. C.; Shen, H.; Watkins S. C.; Cheng T.; Robbins P. D. J. Biol. Chem. 2002, 277, 30208–30218. 28 Gao, S.; Simon, M. J.; Hue, C. D.; Morrison, B. III; Banta, S. ACS Chem. Biol. 2011, 6, 484–491. 29 Gao, C.; Mao, S.; Ditzel, H. J.; Farnaes, L.; Wirsching, P.; Lerner, R. A.; Janda, K. D. Bioorg. Med. Chem. 2002, 10, 4057–4065. 30 Nakayama, F.; Yasuda, T.; Umeda, S.; Asada, M.; Imamura, T.; Meineke V.; Akashi, M. J. Biol. Chem. 2011, 286, 25823–25834. 31 Guo, Z.; Peng, H.; Kang, J.; Sun, D. Biomed. Rep. 2016, 4 (5), 528–534. 32 Raucher, D.; Ryu, J. S. Trends Mol. Med. 2015, 21, 560–570. 33 Crombez, L.; Aldrian-Herrada, G.; Konate, K.; Nguyen, Q. N.; McMaster, G. K.; Brasseur, R.; Heitz, F.; Divita, G. Mol. Ther. 2009, 17 (1), 95–103. 34 Pazo, M.; Salluce, G.; Juanes, M.; Lostalé-Seijo, I.; Gonzalez, F.; Garcia-Fandiño, R.; Montenegro, J. RSC Chem. Biol. 2021, 2, 503–512. 35 Morris, M. C.; Depollier, J.; Mery, J.; Heitz, F.; Divita, G. Nat. Biotechnol. 2001, 19, 1173–1176. 47 has been discovered able to form stable complexes with different nucleic acids.36 Moreover, proline-rich amphipathic peptides have been reported to show good membrane translocation abilities.37,38 Recently, a library of CPPs with variable cationic/hydrophobic residues balance and positions of the charged amino acids along a short oligoalanine scaffold has been developed to study the importance of these features on the cellular uptake of the CPPs and their cytotoxicity.39 Within this library, a peptide with a low number of arginine residues, which could be responsible for the formation of aggregates with the cargo, and hydrophobic residues, that could be selectively membrane-lytic, was identified; this minimal peptide (MP1) acquired a helical conformation when inserted in a hydrophobic environment, like cell membranes. The helical character was especially pronounced in the presence of anionic membranes, and this led to the opening of transient pores in endosomes (enriched in negatively charged lipids), allowing the release of the endosomal content. This work shows the importance of the amphiphilicity enhancement by helical control, since MP1 presents an amphipathic character, and thus can disrupt membranes, only in the endosomes when it is in a helical conformation, due to the resulting opposite orientation of the cationic and hydrophobic residues.34,40 1.3.2.2. Synthetic polymers Different polymers have been described to follow a membrane-fusion entry mechanism and, thus, function as carriers for the internalization of macromolecules inside the cell. Particularly, 2-pyridinecarboxaldehyde polymers have been reported to deliver proteins to which they are covalently attached, such as GFP and RNAse A;41 poly(oxanorborneneimide) (PONI) homopolymers allowed the internalization of engineered proteins (GFP, Cre recombinase, tdTomato) derivatized with terminal oligoglutamate “E-tags” to facilitate interaction with the polymer;42 finally, cell penetrating poly(disulfide)s (CPDs) and benzopolysulfanes (BPS) were proven to induce 36 Morris, M. C.; Vidal, P.; Chaloin, L.; Heitz, F.; Divita, G. Nucleic Acids Res. 1997, 25, 2730–2736. 37 Frankel, A. D.; Pabo, C. O. Cell 1988, 55, 1189-1193. 38 Fillon, Y. A.; Anderson, J. P. Chmielewski, J. J. Am. Chem. Soc. 2005, 127, 11798-11803. 39 Pazo, M.; Juanes, M.; Lostalé-Seijo, I.; Montenegro, J. Chem. Commun. 2018, 54 (50), 6919–6922. 40 Pazo, M. Membrane-targeted peptides for delivery of functional proteins. Ph.D. Dissertation, Universidade de Santiago de Compostela, Santiago de Compostela, 2021. 41 Sangsuwan, R.; Tachachartvanich, P.; Francis, M. B. J. Am. Chem. Soc. 2019, 141, 2376– 2383. 42 Lee, Y.-W.; Luther, D. C.; Goswami, R.; Jeon, T.; Clark, V.; Elia, J.; Gopalakrishnan, S.; Rotello, V. M. J. Am. Chem. Soc. 2020, 142, 4349–4355. 48 the cell uptake of protein cargos such as GFP, streptavidin or nanobodies.40,43,44,45 Regarding nucleic acids delivery, a novel class of transporters called chargealtering releasable transporters (CARTs) has recently been developed.46 Similar to conventional polycationic transporters, CARTs electrostatically complex their polyanionic cargo (typically mRNA), but the cargo release within cells is facilitated by a dynamic conversion to neutral byproducts (cationic amines are converted to neutral amides), thereby avoiding the toxicity issues frequently encountered with persistent cations.46,47 CARTs demonstrate great efficiency for the delivery of various cargo sizes, including an mRNA SARS-CoV-2 vaccine in murine models.48 Remarkably, the organ selectivity of these transporters in in vivo experiments can be tuned through alterations in the CART structure, without the use of monoclonal antibodies or targeting ligands.49 Delivery of mRNA via CARTs has shown promising results in preclinical studies, including the protection and/or cure of cancer in mice50 and the clearance of metastatic disease51 in mouse models. Polyhydrazide polymers have been recently employed for postpolymerization functionalization with cationic and hydrophobic aldehydes to generate amphiphilic polymeric vehicles for the intracellular delivery of siRNA52 or plasmid DNA.53 The in situ activation of the poly-(acryloyl hydrazide) scaffold with readily available aldehydes allowed the evaluation of transport activity in biological conditions without the need for isolation 43 Gasparini, G.; Bang, E.-K.; Molinard, G.; Tulumello, D. V.; Ward, S.; Kelley, S. O.; Roux, A.; Sakai, N.; Matile, S. J. Am. Chem. Soc. 2014, 136, 6069–6074. 44 Derivery, E.; Bartolami, E.; Matile, S.; Gonzalez-Gaitan, M. J. Am. Chem. Soc. 2017, 139, 10172–10175. 45 Cheng, Y.; Zong, L.; López-Andarias, J.; Bartolami, E.; Okamoto, Y.; Ward, T. R.; Sakai, N.; Matile, S. Angew. Chemie - Int. Ed. 2019, 58, 9522–9526. 46 McKinlay, C. J.; Vargas, J. R.; Blake, T. R.; Hardy, J. W.; Kanada, M.; Contag, C. H.; Wender, P. A.; Waymouth, R. M. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (4), E448–E456. 47 Benner, N. L.; McClellan, R. L.; Turlington, C. R.; Haabeth, O. A. W.; Waymouth, R. M.; Wender, P. A. J. Am. Chem. Soc. 2019, 141 (21), 8416–8421. 48 Haabeth, O. A. W.; Lohmeyer, J. J. K.; Sallets, A.; Blake, T. R.; Sagiv-Barfi, I.; Czerwinski, D. K.; McCarthy, B.; Powell, A. E.; Wender, P. A.; Waymouth, R. M.; et al. ACS Cent. Sci. 2021, 7 (7), 1191–1204. 49 Blake, T. R.; Haabeth, O. A. W.; Sallets, A.; McClellan, R. L.; Del Castillo, T. J.; VilchesMoure, J. G.; Ho, W. C.; Wender, P. A.; Levy, R.; Waymouth, R. M. Bioconjug. Chem. 2023. Doi: 10.1021/acs.bioconjchem.3c00019. 50 Haabeth, O. A. W.; Blake, T. R.; McKinlay, C. J.; Waymouth, R. M.; Wender, P. A.; Levy, R. Proc. Natl. Acad. Sci. U. S. A. 2018, 115 (39), E9153–E9161. 51 Haabeth, O. A. W.; Blake, T. R.; McKinlay, C. J.; Tveita, A. A.; Sallets, A.; Waymouth, R. M.; Wender, P. A.; Levy, R. Cancer Res. 2019, 79 (7), 1624–1634. 52 Priegue, J. M.; Crisan, D. N.; Martínez-Costas, J.; Granja, J. R.; Fernandez-Trillo, F.; Montenegro, J. Angew. Chemie - Int. Ed. 2016, 55 (26), 7492–7495. 53 Priegue, J. M.; Lostalé-Seijo, I.; Crisan, D.; Granja, J. R.; Fernández-Trillo, F.; Montenegro, J. Biomacromolecules 2018, 19 (7), 2638–2649. 49 and purification steps of the obtained supramolecular polymeric vector.52 It was also demonstrated that by tuning the degree of polymerization it was possible to deliver larger nucleic acid cargos. In fact, plasmid DNA required a longer polymer for the efficient DNA condensation, membrane translocation, and nuclear delivery.53 Additionally, with further modifications and adjustments, these amphiphilic polyhydrazones complexed and delivered mRNA with high efficiency and low cytotoxicity even at low polyhydrazone concentrations, confirming the adaptability of the technology to more sensitive and therapeutically relevant nucleic acids.54 1.3.2.3. Lipids Lipid-based drug delivery systems have been extensively explored in the delivery of nucleic acids, proteins, and small molecule drugs, and they have been developed in multiple modalities; to date the most common are micelles, liposomes, solid lipid-based nanoparticles, nanostructured lipid carriers, lipid-drug conjugates, and nanoemulsions.55 Lipid-based formulations have many advantages, such as biocompatibility, stability, versatility and tunability for diverse kind of payload and route of administration; as a consequence, they have been considered valuable candidates for the formulation of pharmaceuticals, as well as vaccines, diagnostics, and nutraceuticals.56 Different lipid carriers are already available for the use in human treatment at present time. Notable applications based on lipid nanoparticles (LNPs) include the first-ever U.S. Food and Drug Administration (FDA)-approved nano-drug, Doxil®, which employs LNPs to deliver the anticancer drug doxorubicin, and Epaxal®, for the delivery of hepatitis A protein antigens,57 or ONPATTRO® (patisiran), developed by Alnylam for the intravenous delivery of RNAi to treat hereditary transthyretin-mediated amyloidosis.58 In recent times, LNP-based delivery systems have been used to develop mRNA vaccines to face the COVID-19 pandemic, for example by 54 Juanes, M.; Creese, O.; Fernandez-Trillo, F.; Montenegro, J. MedChemComm 2019, 10, 1138–1144. 55 Zubair, A.-H. A.; Sheshe, S. M.; Bashir, M. R.; Sade, S. M. J. Appl. Life Sci. Int. 2021, 24 (3), 33–46. 56 Shrestha, H.; Bala, R.; Arora, S. J. Pharm. (Cairo) 2014, 2014, 801820. 57 Li, Y.; Ye, Z.; Yang, H.; Xu, Q. Acta Pharm. Sin. B 2022, 12 (6), 2624–2639. 58 Zhang, X.; Goel, V.; Robbie, G. J. J. Clin. Pharmacol. 2020, 60 (5), 573–585. 50 Pfizer/BioNTech (BNT162b2/Comirnaty®) and Moderna (mRNA-1273/ SpikeVax®).57,59 Lipid-based formulations have also been used for the intracellular delivery of protein cargos. For example, cholesterol was employed as a component of a protein transporter that bypasses the endocytic pathway.60 Other liposome-based carriers are inspired by soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins, as they induce the fusion of cellular vesicles with the plasma membrane through coiled-coil interactions given by helical peptides with which both liposomes and target cells are decorated.61 1.3.2.4. Inorganic nanoparticles Inorganic nanoparticles (NPs) coated with cell-derived membranes can interact with cell membranes and give rise to fusion mechanisms. For this reason, and thanks to their low toxicity and high cargo encapsulation efficiency, they have been lately employed as novel and versatile carriers, for in vitro assays as well as for systemic delivery or specific targeting in cancer treatment.62 In this field, gold nanoparticles (AuNPs) have gained special attention for the delivery of macromolecules using an electrostatic complexation strategy, where anionic proteins interact with cationic peptide tags present on the AuNPs.63 However, this tight interaction may lead to inefficient cargo release, so a formulation with hydrophobic components (such as linoleic acid) and a shorter binding tag was proposed to increase NPs’ interactions with the plasma membrane and give rise to a membrane-fusion delivery mechanism. This new type of NPs was called nanoparticle stabilized nanocapsules (NPSCs).64 1.3.2.5. Protein channels Transport of biomolecules across the cell membrane was also achieved through the assembly of protein units in multimeric channels. Many of these protein channels have a bacterial origin, like protein injection machines known as types III, IV and VI protein-secretion systems.65 59 Schoenmaker, L.; Witzigmann, D.; Kulkarni, J. A.; Verbeke, R.; Kersten, G.; Jiskoot, W.; Crommelin, D. J. A. Int. J. Pharm. 2021, 601, 120586. 60 Tai, W.; Zhao, P.; Gao, X. Sci. Adv. 2020, 6, eabb0310. 61 Marsden, H. R.; Tomatsu, I.; Kros, A. Chem. Soc. Rev. 2011, 40, 1572–1585. 62 Soprano, E.; Polo, E.; Pelaz, B.; del Pino, P. J. Nanobiotechnol. 2022, 20, 538. 63 Ray, M.; Lee, Y. W.; Scaletti, F.; Yu, R.; Rotello, V. M. Nanomedicine 2017, 12, 941–952. 64 Tang, R.; Jiang, Z.; Ray, M.; Hou, S.; Rotello, V. M. Nanoscale 2016, 8, 18038–18041. 65 Galán, J. E.; Waksman, G. Cell 2018, 172, 1306–1318. 51 Another type of protein channel discovered within bacterial structures is the large conductance mechanosensitive ion channel (MscL) from E. coli, which allows the internalization of peptides like phalloidin.66 Other pore-forming proteins are perforins, which belong to the membrane attack complexes/perforin-like family (MACPF).67 MACPFs aid the internalization of cationic cargos, like CytC or positively supercharged GFP, but not of neutral or negatively charged (negatively supercharged GFP) proteins.68 Pore-forming cholesterol-dependent cytolysins (CDCs) are another example of protein channels that can be used for reversible permeabilization of the plasma membrane to allow the intracellular transport of proteins or antibodies.40,69,70 1.4. The amphiphilicity paradigm As previously mentioned, the plasma membrane is a thin, flexible barrier that surrounds the cell and separates the inside from the extracellular environment. Its basic structure is determined by the bilayer of phospholipid molecules, which are amphiphilic in nature. Amphiphilic molecules have both hydrophilic (‘water-loving’) and hydrophobic (‘water-fearing’) regions. In the case of phospholipids, the hydrophilic head is composed of a phosphate group, which is negatively charged at physiological pH and attracts water molecules, and another polar group, such as choline, ethanolamine, serine or sugar, which also attract water molecules. The hydrophobic tail is composed of two long-chain fatty acid molecules, which are non-polar and do not interact well with water.4,71 Both parts are joined by a alcohol, such as glycerol or sphingosine. Because of their amphiphilic nature, when placed in an aqueous environment, phospholipids spontaneously self-assemble into energetically favourable structures, such as micelles, liposomes, or lamellar tubules. The hydrophobic tails face inward, away from the water, and the hydrophilic heads face towards the water environment (Figure 4). This arrangement 66 Doerner, J. F.; Febvay, S.; Clapham, D. E. Nat. Commun. 2012, 3, 990. 67 Dunstone, M. A.; Tweten, R. K. Curr. Opin. Struct. Biol. 2012, 22, 342–349. 68 Stewart, S. E.; Kondos, S. C.; Matthews, A. Y.; D’Angelo, M. E.; Dunstone, M. A.; Whisstock, J. C.; Trapani, J. A.; Bird, P. I. J. Biol. Chem. 2014, 289, 9172–9181. 69 Teng, K. W.; Ishitsuka, Y.; Ren, P.; Youn, Y.; Deng, X.; Ge, P.; Lee, S. H.; Belmont, A. S.; Selvin, P. R. eLife 2016, 5, e20378. 70 Walev, I.; Bhakdi, S. C.; Hofmann, F.; Djonder, N.; Valeva, A.; Aktories, K.; Bhakdi, S. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 3185–3190. 71 Biga, L. M.; Dawson, S.; Hardwell, A.; Hopkins, R.; Kaufmann, J.; LeMaster, M.; Matern, P.; Morrison-Graham, K.; Quick, D.; Runyeon, J. Anatomy & Physiology. OpenStax/Oregon State University, 2019. 52 creates a selectively permeable barrier, which allows some substances to pass through it while preventing others from crossing it.4 For this reason, the search for synthetic membrane carriers to transport molecules that would not naturally and spontaneously cross the cell membrane has always attracted great attention. In this quest, one conceptual approach has prevailed so far: researchers have focused their attention on the design of amphiphilic molecules, with a hydrophobic portion, that serves as an anchor to ensure membrane affinity, and an ionic part, that electrostatically attracts the cargo to trigger membrane transport of the resulting chargeneutralized complexes.72,73 Figure 4 | Phospholipid structure and bilayer assembly motif. Representation of the structure of an amphiphilic phospholipid molecule, formed by a polar hydrophilic phosphate “head” and a non-polar hydrophobic lipid “tail”. The phospholipid bilayer consists of two adjacent sheets of phospholipids, arranged tail to tail. The hydrophobic tails associate with one another, forming the interior of the membrane. The polar heads contact the fluid inside and outside of the cell. Adapted and reproduced with permission from ref. 71 licensed under a Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA). 1.4.1. The counterion activation mechanism Many arginine-rich CPPs have been developed based on the TAT peptide viral sequence, namely RKKRRQRRR.74,75,76 The only structural requirement of this kind of peptides for cell-permeating activity seemed to 72 Lostalé-Seijo, I.; Montenegro, J. Nat. Rev. Chem. 2018, 2, 258–277. 73 Lehn, J.-M. Science 1985, 227 (4689), 849–856. 74 Silhol, M.; Tyagi, M.; Giacca, M.; Lebleu, B.; Vivès, E. Eur. J. Biochem. 2002, 269 (2), 494–501. 75 Ziegler, A.; Nervi, P.; Dürrenberger, M.; Seelig, J. Biochemistry 2005, 44 (1), 138–148. 76 Rapoport, M.; Lorberboum-Galski, H. Expert Opin. Drug Deliv. 2009, 6 (5), 453–463. 53 be the presence of multiple guanidinium cations. These CPPs were able to transport a large variety of membrane-impermeable molecules inside the cell, including macromolecular cargos as large as GFP or nanoparticles, and this has attracted a strong scientific interest.77 The ability of guanidinium-rich CPPs to cross cell membranes has been explained with the concept of “dynamic biphilicity”. This term refers to molecules that can be both hydrophilic and lipophilic, depending on the circumstances, and, in contrast with amphiphiles, can adapt to any environment.75,78 The unusual behavior of oligoand polyarginines in biomembranes (sometimes called “arginine magic”) originates from counteranion scavenging.77,78,79,80 Guanidinium cations are known to scavenge and bind to various exchangeable anions (e.g., phosphates and carboxylates) both by Coulombic and hydrogen bonding interactions. It has been demonstrated that the biphilic character of these polycations depends on the nature of their counteranions.78 This property has not been observed for other polycations, such as polyhistidines or polylysines.81,80 Guanidinium cations prove less hydrophilic than ammonium cations because of the delocalization of the positive charges. The strong basic character of the guanidinium cation may give rise to the observed ‘‘occasional lipophilicity’’ of polyarginine. When placed in a neutral aqueous environment, the more acidic polylysine (intrinsic pKa = 10.5) can release protons to minimize the charge repulsion between proximal cations in the polymer. The insufficient acidity of the guanidinium cation (intrinsic pKa = 12.5) makes a similar reduction of individual pKa’s impossible for polyarginine. To reduce intramolecular charge repulsion, proximal guanidinium cations have to neutralize (or even invert) their charge not by proton release but by counteranion scavenging (Figure 5).81,82 77 Sakai, N.; Matile, S. J. Am. Chem. Soc. 2003, 10 (3), 14348–14356. 78 Sakai, N.; Futaki, S.; Matile, S. Soft Matter 2006, 2 (8), 636–641. 79 Rueping, M.; Mahajan, Y.; Sauer, M.; Seebach, D. ChemBioChem 2002, 3, 257–259. 80 Mitchell, D. J.; Kim, D. T.; Steinman, L. C.; Fathman, C. G.; Rothbard, J. B. J. Pept. Res. 2000, 55, 318–325. 81 Nishihara, M.; Florent, P.; Takeuchi, T.; Futaki, S.; Lazar, A. N.; Coleman, A. W.; Sakai, N.; Matile, S. Org. Biomol. Chem. 2005, 3, 1659–1669. 82 Takeuchi, T.; Sakai, N.; Matile, S. Faraday Discuss. 2009, 143, 187–203. 60 referred to water-structure-makers and chaotrope to water-structurebreakers.113 Later, chaotropic ions were also shown to exhibit direct binding affinity to hydrophobic species.114,115,116 However, while the classic and widely described hydrophobic effect was governed by entropic forces, the association of chaotropic ions to hydrophobic surfaces was enthalpy driven.117 In recent times, the chaotropic effect has risen in popularity in various areas of chemistry, ranging from molecular recognition to biologically relevant interactions with membranes, lipids, and proteins to material science.118,119,120,121,122 In 2018, Assaf and Nau clarified the relationship between kosmotropic, chaotropic, and hydrophobic solvation, and presented the chaotropic effect as an assembly motif, orthogonal to the hydrophobic effect. They illustrated how the chaotropic effect led chaotropes to bind to synthetic and biological hydrophobic structures, and pointed out that this phenomenon was especially pronounced for some large anions; since their chaotropic behavior exceeded the Hofmeister scale, these molecules were classified as superchaotropes, and included dodecaborates and derivatives, like COSAN and bigger boranes (Figure 7b).123,124 The effect of solutes with different characteristics on water structure can be figured out through the way that water molecules organize themselves around that particular solute. Water molecules have a tetrahedral coordination pattern, and they arrange and orientate themselves as shown in Figure 7a depending on the nature of the solute. In this perspective, the 113 Marcus, Y. Chem. Rev. 2009, 109 (3), 1346–1370. 114 Sachs, J. N.; Woolf, T. B. J. Am. Chem. Soc. 2003, 125, 8742–8743. 115 Gibb, C. L. D.; Gibb, B. C. J. Am. Chem. Soc. 2011, 133 (19), 7344–7347. 116 Assaf, K. I.; Ural, M. S.; Pan, F.; Georgiev, T.; Simova, S.; Rissanen, K.; Gabel, D.; Nau, W. M. Angew. Chemie - Int. Ed. 2015, 54, 6852–6856. 117 Biedermann, F.; Nau, W. M.; Schneider, H. J. Angew. Chemie - Int. Ed. 2014, 53 (42), 11158–11171. 118 Moussawi, M. A.; Leclerc-Laronze, N.; Floquet, S.; Abramov, P. A.; Sokolov, M. N.; Cordier, S.; Ponchel, A.; Monflier, E. et al. J. Am. Chem. Soc. 2017, 139 (36), 12793–12803. 119 Gale, P. A.; Howe, E. N. W.; Wu, X. Chem 2016, 1 (3), 351–422. 120 Li, W. Z.; Chen, H.; Shen, M. N.; Yang, Z.; Fan, Z.; Xiao, J.; Chen, J.; Zhang, H.; Wang, Z.; Wang, X. Q. Small 2022, 18 (15), 2108055. 121 Hu, X.; Guo, D. Angew. Chemie - Int. Ed. 2022, 61 (26), e202204979. 122 Naskar, B.; Diat, O.; Nardello-Rataj, V.; Bauduin, P. J. Phys. Chem. C 2015, 119 (36), 20985–20992. 123 Assaf, K. I.; Nau, W. M. Angew. Chemie - Int. Ed. 2018, 57 (43), 13968–13981. 124 Wang, W.; Wang, X.; Cao, J.; Liu, J.; Qi, B.; Zhou, X.; Zhang, S.; Gabel, D.; Nau, W. M.; Assaf, K. I.; Zhang, H. Chem. Commun. 2018, 54, 2098–2101. 61 chaotropic effect positions itself between the classic hydrogen bonding, where water molecules are strongly coordinated with apical orientation, and the hydrophobic effect, characterized by a preferred basal orientation of the water molecules that maximizes the proximity to bonding electrons, lone pairs, and the oxygen atom. Diversely, a chaotrope causes less directional ion–dipole interactions, which result in a lateral orientation of the water molecules under the alignment of their dipole moment (Figure 7b). The described modes of solvation are characterized by different thermochemical fingerprints that allow the discrimination and identification of the predominant driving force in a certain supramolecular assembly process in water.123 Figure 7 | Principal aqueous solvation patterns for anions. a, Presumed favourable orientations of water molecules around a cavity and at the surface of different solutes: kosmotropic and chaotropic ions, hydrophobic molecules, and void space. Electron lone pairs are visualized in yellow. b, Extended Hofmeister scale with specification of the superchaotropic, hydrophobic ionic, and superhydrophobic regions Reproduced with permission from ref. 123 Copyright 2018 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim. 2.2. Boron clusters Carbon-based chemistry has been at the forefront of chemical research for over two centuries. Despite the closeness in the periodic table, boron-based 62 chemistry has not experienced the same recognition and has mostly been used in service of carbon-based chemistry, rather than having a spotlight on itself.125,126 The history of boron clusters began in the early 1900s, with the synthesis of a series of neutral boranes (or boron hydrides) in the pioneering work of Alfred Stock and co-workers.127 However, the concept of boron cluster was introduced in 1954, when Lipscomb and co-workers deepened the study of boron catenation. This property – belonging to only a few elements – allows atoms of the same type to form bior three-dimensional structures through a series of covalent bonds and led to the theorization of the icosahedral borane B12H12.128 In fact, boron catenation generally results in threedimensional deltahedra (polyhedra with regular triangular faces), as opposed to carbon and silicon which form almost exclusively rings and chains. However, neutral boranes were unstable and waterand airsensitive, thus not suitable for further development and investigation.126 Posterior studies by Longuet-Higgins and Roberts suggested that the stable form of an icosahedral boron cluster would have an overall double negative charge.129 This theory was then validated by Hawthorne and Pitochelli, who reported in 1960 the first synthesis of double anionic dodecaboranes.126,130 In 1971, Robert E. Williams identified the structures of open boranes, illustrating that they were not (as had been previously assumed) simply fragments of an icosahedron. He arranged such structures into three families – closo, nido and arachno – and outlined that, considering dianionic borates [BnHn]2– as parent closo n-vertex polyhedra, the corresponding nido structure would have (n – 1) vertexes, and the arachno structure would have (n – 2) vertexes.131 In the same year, Kenneth Wade provided a rationalization of the shapes of boron clusters with delocalized skeletal bonding, in terms of the number of skeletal electron pairs (SEPs) these molecules possess. Wade clarified that species with n skeletal atoms and held together by (n + 1) SEPs adopt closo125 Axtell, J. C.; Saleh, L. M. A.; Qian, E. A.; Wixtrom, A. I.; Spokoyny, A. M. Inorg. Chem. 2018, 57 (5), 2333–2350. 126 Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457−2483. 127 Stock, A. Hydrides of Boron and Silicon. Cornell University Press: Ithaca, New York, 1933. 128 Eberhardt, W. H.; Crawford, B., Jr.; Lipscomb, W. N. J. Chem. Phys. 1954, 22, 989−1001. 129 Longuet-Higgins, H. C.; de V. Roberts, M. Proc. R. Soc. London, Ser. A 1955, 230, 110−119. 130 Pitochelli, A. R.; Hawthorne, M. F. J. Am. Chem. Soc. 1960, 82, 3228−3229. 131 Williams, R. E. Inorg. Chem. 1971, 10 (1), 210–214. 63 structures, if they are held together by (n + 2) pairs they adopt nidostructures, and when held together by (n + 3) SEPs they adopt arachnostructures (Figure 8).132,133,134 These principles are officially known as Polyhedral Skeletal Electron Pair Theory (PSEPT) but commonly referred to as the Wade’s (or Wade-Mingos) rules. Figure 8 | The deltahedral pattern, showing atom and SEP numbers. The closo, nido and arachno structures of boron clusters presenting from 6 to 15 skeletal electron pairs (SEPs), according to Wade’s rules. Reproduced with permission from ref. 133. Copyright 2013 Walter de Gruyter GmbH, Berlin/Boston. 132 Wade, K. J. Chem. Soc. D Chem. Commun. 1971, 15, 792–793. 133 Fox, M. A.; Wade, K. Pure Appl. Chem. 2003, 75 (9), 1315–1323. 134 Welch, A. J. Chem. Commun. 2013, 49 (35), 3615–3616. 64 Boron cluster skeletal structures may also contain carbon atoms: in this case, the substitution of a B–H with a C–H group generates a carborane, whose structure and existence were first predicted by Lipscomb and Hoffmann.135 The term “carborane” is a condensation of the IUPAC “carbaborane” and defines, de facto, a borane containing one or more carbon atoms in the polyhedral framework. Similarly, “metallaboranes” and “heteroboranes” refer to clusters containing metal or main-group heteroatoms.136 Due to their stability, facile synthesis, and versatile functionalization, twelve vertex monoand dicarboranes are the most studied. The latter exist in different isomeric forms (orto, meta, or para), depending on the relative position of C–H towards each other (Figure 9a). Regarding the shape, carboranes follow Wade’s rules. For example, considering the dicarborane C2B(n–2)Hn as the closo parent (with n between 5 and 14), they present the same structures as the ones illustrated for boranes in Figure 8. If a closo dicarborane undergoes a process known as deboronation and loses a B–H group, the resulting nido structure is known as dicarbollide. The atoms forming the flat face of dicarbollides with n = 12 (C2B9H112–) present the same hybridization as flat polygonal systems (e.g., benzene, cyclopentadienide), while the atoms forming the icosahedron retain the normal hybridization of cluster compounds. This characteristic hybridization at the flat face allows carbollides to complex transition metals and form metallacarboranes. Of special interest are bis(dicarbollide), compounds in which two carbollides clusters are complexed by a transition metal (e.g., Cr, Fe, Cu, Co). Among them, the most used and well-known are cobaltabis(dicarbollides), in which a central cobalt atom is sandwiched between two carboranyl clusters, and they are commonly known as COSANs (CObalt SANdwich). Bis(dicarbollide) compounds can form cisoid, gauche and transoid isomers (rotamers), depending on the relative position of the C–H bonds across the B10-B10’ axis (Figure 9b).137 It has been reported that, surprisingly, COSANs behave like amphiphiles without head-and-tail design,138,139,140 and they can self-assemble into 135 Hoffmann, R.; Lipscomb, W. N. J. Chem. Phys. 1962, 36 (12), 3489–3493. 136 Grimes, R. N. J. Chem. Educ. 2004, 81, 657−672. 137 Fernandez-Alvarez, R. Design of Boron Cluster-Containing Nanostructures in Solution. Ph.D. Dissertation, Charles University, Prague, 2020. 138 Fernandez-Alvarez, R.; Ďorďovič, V.; Uchman, M.; Matějíček, P. Langmuir 2018, 34 (12), 3541–3554. 139 Malaspina, D. C.; Viñas, C.; Teixidor, F.; Faraudo, J. Angew. Chemie - Int. Ed. 2020, 59 (8), 3088–3092. 65 membranes and vesicles.141 They have recently been used to generate synthetic vesicles,142 conjugated with fluorescent BODIPY for cell tracking,143 and they have been shown to interact with DNA144 and cavities like cyclodextrins.145 Figure 9 | Chemical structure of dicarborane and bis(dicarbollide) isomers. Pink spheres represent B–H bonds, black spheres C–H bonds, and purple spheres transition metals, e.g., cobalt. a, Dicarborane isomers, from left to right: orto (1,2–C2B10H12), meta (1,7–C2B10H12), and para (1,12–C2B10H12). b, Rotamer configurations of metallabis(dicarbollides). 2.2.1. Globular boron clusters properties In this dissertation, the term ‘globular boron cluster’ refers specifically to closo-boranes and halogen-substituted closo-boranes. 140 Hohenschutz, M.; Grillo, I.; Diat, O.; Bauduin, P. Angew. Chemie - Int. Ed. 2020, 59 (21), 8084–8088. 141 Bauduin, P.; Prevost, S.; Farràs, P.; Teixidor, F.; Diat, O.; Zemb, T. Angew. Chemie - Int. Ed. 2011, 50 (23), 5298–5300. 142 Tarrés, M.; Canetta, E.; Paul, E.; Forbes, J.; Azzouni, K.; Viñas, C.; Teixidor, F.; Harwood, A. J. Sci. Rep. 2015, 5, 7804. 143 Chaari, M.; Gaztelumendi, N.; Cabrera-González, J.; Peixoto-Moledo, P.; Viñas, C.; Xochitiotzi-Flores, E.; Farfán, N.; Salah, A. Ben; Nogués, C.; Núñez, R. Bioconjug. Chem. 2018, 29, 1763−1773. 144 Fuentes, I.; García-Mendiola, T.; Sato, S.; Pita, M.; Nakamura, H.; Lorenzo, E.; Teixidor, F.; Marques, F.; Viñas, C. Chem. - A Eur. J. 2018, 24 (65), 17239–17254. 145 Assaf, K. I.; Begaj, B.; Frank, A.; Nilam, M.; Mougharbel, A. S.; Kortz, U.; Nekvinda, J.; Grüner, B.; Gabel, D.; Nau, W. M. J. Org. Chem. 2019, 84 (18), 11790–11798. 66 It has been outlined here how globular boron clusters represent a distinctive class of covalent species: thanks to their unique molecular architecture, nonconventional bonding, and unusual chemistry, they exhibit specific properties not encountered in other types of compounds.146 As previously illustrated, clusters in their closo forms with n number of boron atoms contain n + 1 skeletal bonding electron pairs, resulting in anions of the type [BnHn]2− (with n = 5 to 12 and above).132 The presence of n + 1 SEPs emerges from the peculiar bonding situation of globular boron clusters, which cannot be explained by the classic two-center two-electron (2c-2e) bonding. Rather, they are characterized by the less common threecenter two-electron (3c-2e) bond type. This situation deviates from the standard Lewis model and is labeled as “nonclassical” by inorganic and organic chemists. 3c-2e bonds are often observed in the case of compounds that have boron, beryllium, or aluminum atoms at the center, and these molecules were typically – and equivocally – described as “electrondeficient” since a single pair of electrons is shared by three atoms. However, the unique electronic distribution of deltahedral cluster dianions makes them some of the most stable molecules in all of chemistry.136 The dicesium salt of the prototype borane B12H122−, for example, can survive temperatures above 810 ºC without decomposition. The source of the incredible stability and inertness of these dianionic icosahedra is the high delocalization of bonding electrons, which form a molecular orbital in the center of the cluster shared by all boron atoms and one set of atomic orbitals whose in-phase contributions result in a molecular orbital delocalized over the surface of the pseudospherical cage (Figure 10).147,148 Besides being astonishingly stable, globular boron clusters are water-soluble and present very low toxicity in humans. These characteristics make them biocompatible and especially suitable for biomedical applications.149 146 Plesek, J. Chem. Rev. 1992, 92 (2), 269–278. 147 McKee, M. L. Inorg. Chem. 2002, 41 (5), 1299–1305. 148 King, R. B. Chem. Rev. 2001, 101 (5), 1119–1152. 149 Gabel, D. Pure Appl. Chem. 2015, 87 (2), 173–179. 67 Figure 10 | Delocalization of bonding electrons in icosahedral dodecaboranes. Calculated structures (based on X-ray crystal structures) and representative Kohn−Sham molecular orbitals of A1g symmetry, demonstrating delocalization of the electron density in closo-[B12H12]2−, as described by King.148 Adapted and reproduced with permission from ref. 126. Copyright 2018, American Chemical Society. 2.2.2. Applications and therapeutic use of boron clusters Even though boron cluster species are quite expensive to produce, due to the limited quantities of boric oxide derivatives to create B−B and B−H bonds, they have been used since the 1960s for different applications, ranging from material science to medicine. Their first application was in advanced jet and rocket propulsion systems, but that era lasted only a few years. The current applications for boron cluster compounds are based either on specific properties of the basic element itself or on the unique properties of the molecules.146 In the last decades, different procedures for the synthesis of pure and functionalized boron clusters have been unceasingly developed.150,151,152 As a consequence of the plethora of spatial and electronic structures provided by these compounds, they found applications as new magnetic,153 semiconductive,154 superconductive,155 and hydrogen storage materials.156,157 Furthermore, many globular boron cluster derivatives have been exploited for a variety of applications in medicinal chemistry thanks to their remarkable stability and biocompatibility.136 The high affinity of boranes and carboranes towards biomacromolecules like proteins and nucleic acids 150 Piazza, Z. A.; Hu, H.-S.; Li, W.-L.; Zhao, Y.-F.; Li, J.; Wang, L.-S. Nat. Commun. 2014, 5, 3113. 151 Hou, C.; Tai, G.; Wu, Z.; Hao, J. ChemPlusChem 2020, 85, 2186−2196. 152 Đorđevic, S.; Solà, M.; Radenkovic, S. Inorg. Chem. 2022, 61, 10116−10125. 153 Scheifers, J. P.; Zhang, Y.; Fokwa, B. P. T. Acc. Chem. Res. 2017, 50, 2317–2325. 154 Duret, G.; Quinlan, R.; Bisseret, P.; Blanchard, N. Chem. Sci. 2015, 6, 5366–5382. 155 Nagamatsu, J.; Nakagawa, N.; Muranaka, T.; Zenitani, Y.; Akimitsu, J. Nature 2001, 410, 63–64. 156 Wang, Y.-J.; Xu, L.; Qiao, L.-H.; Ren, J.; Hou, X.-R.; Miao, C.-Q. Int. J. Hydrog. Energy 2020, 45, 12932–12939. 157 Wang, Y.-J.; Wang, G.-L.; Guo, M.-M.; Miao, C.-Q.; Chen, H.-P.; Zhai, H.-J. Int. J. Hydrog. Energy 2021, 46, 24225–24232. 68 could be exploited to create pharmacophore conjugates with therapeutic molecules that benefit from the properties of native therapeutic molecule and noncovalent interactions of boron clusters with biological material, thus resulting in a therapeutic activity enhancement.158 Among the most recent applications in chemical biology and life sciences149 to be mentioned is the use of boron clusters for fluorescent cell tracking,141 for the release of liposomal content,159 and more recently also for the cellular delivery of therapeutics nucleic acids.160 The most prominent and well-known application of globular boron clusters is Boron Neutron Capture Therapy (BNCT). BNCT uses the combination of 10B isotopes and low-energy neutrons to treat cancer. 10B is a stable, nonradioactive, isotope of boron, that represents 20 % of the total boron isotopes in Nature. In BNCT, 10B isotopes are selectively taken up by cancer cells carried by a tumor-localizing drug, and, when the tissue is bombarded with low-energy neutrons, 10B atoms capture the neutrons and release alpha particles, which damage and kill the cancer cells. The therapy is highly targeted, as the boron isotopes only accumulate in the cancer cells and not in the surrounding healthy tissues, thus minimizing side effects.161,162 To date, BNCT represents a promising treatment modality, with data suggesting the safety and efficacy of treatment in patients with advanced tumors. This treatment has been studied clinically in a variety of disease sites, including lung cancers, breast cancers, extramammary Paget’s disease, hepatocellular carcinoma, glioblastoma multiforme, meningioma, head and neck cancers, sarcomas, cutaneous malignancies, recurrent cancers, paediatric cancers, and metastatic disease.163 2.2.3. Boron clusters as transmembrane carriers As already mentioned, different research groups observed that boron clusters displayed some characteristics that made them resemble and behave like amphiphiles without head-and-tail design.138,139,140 Particularly, 158 Cebula, J.; Fink, K.; Boratyn, J. Goszczyn, T. M. Coord. Chem. Rev. 2023, 477, 214940. 159 Awad, D.; Bartok, M.; Mostaghimi, F.; Schrader, I.; Sudumbrekar, N.; Schaffran, T.; Jenne, C.; Eriksson, J.; Winterhalter, M.; Fritz, J.; et al. ChemPlusChem 2015, 80 (4), 656– 664. 160 Kaniowski, D.; Suwara, J.; Ebenryter-Olbińska, K.; Jakóbik-Kolon, A.; Nawrot, B. Int. J. Mol. Sci. 2022, 23, 14793. 161 Hawthorne, M. F. Angew. Chem. - Int. Ed. 1993, 32, 950−984. 162 Nedunchezhian, K.; Aswath, N.; Thiruppathy, M.; Thirugnanamurthy, S. J. Clin. Diagnostic Res. 2016, 10 (12), ZE01-ZE04. 163 Malouff, T. D.; Seneviratne, D. S.; Ebner, D. K.; Stross, W. C.; Waddle, M. R.; Trifiletti, D. M.; Krishnan, S. Front. Oncol. 2021, 11, 601820. 69 they showed affinity for artificial and natural membranes,114,145 and, when functionalized with iodine, presented liposomolytic activity.159 The behavior of superchaotropic anions in relation to hydrophobic structures and lipid membranes suggested that boron clusters, and potentially other large ions with superchaotropic character, could be employed to activate the transport of hydrophilic molecules across the lipid hydrophobic bilayer.116, 121,164 164 Assaf, K. I.; Suckova, O.; Al Danaf, N.; Von Glasenapp, V.; Gabel, D.; Nau, W. M. Org. Lett. 2016, 18 (5), 932–935. 76 2. RESULTS & DISCUSSION 2.1. Carrier activity in model membranes Prototypal cationic peptides such as oligoarginines do not translocate through zwitterionic phosphocholine lipid vesicles on their own.43,77,83 Therefore, the capability of boron clusters to activate the transport of a heptaarginine peptide (WR7) was investigated first in large unilamellar vesicles. The well-established HPTS/DPX assay,165 that uses 8hydroxypyrene-1,3,6-trisulfonate (HPTS) and p-xylene-bis-pyridinium (DPX), was implemented to monitor peptide-transport activation (Figure 12a). In a typical time-resolved fluorescence experiment, HPTS emission is monitored during the sequential addition of the globular cluster carrier (t = 50 s) and the peptide cargo (t = 100 s, Figure 12b). A surfactant (Triton X100) is added at the end of the experiment (t = 600 s) to release all vesicle content and normalize the fluorescence intensity data. For the iodinated cluster B12I122–, the addition of cluster alone led to an increase in fluorescence (Figure 13a, b). This observation indicated that this cluster, the largest and most chaotropic one, is disrupting the lipid membrane, which we independently confirmed by dynamic light scattering (DLS, Figure 13c) and an alternative vesicle leakage assay166 (carboxyfluorescein assay, Figure 13d, e). All other clusters studied herein preserved membrane integrity in the same sets of vesicular leakage experiments (Figure 13d-g), which ruled out the formation of pores or transitory bilayer disruption and encouraged us to investigate them as potentially viable membrane carriers. The parent B12H122–, which is the smallest and least chaotropic in this series, did not show significant transport of WR7 in vesicles (Figure 12b, pink trace). However, the chlorinated and brominated clusters, which are intermediate in size and chaotropicity, caused the desired time-resolved fluorescence responses that signalled successful transport of the heptaarginine peptide WR7 across the lipid membrane (Figure 12b, green and red traces). 165 Takeuchi, T.; Bagnacani, V.; Sansone, F.; Matile, S. ChemBioChem 2009, 10, 2793−2799. 166 Sorochkina, A. I.; Kovalchuk, S. I.; Omarova, E. O.; Sobko, A. A.; Kotova, E. A.; Antonenko, Y. N. Biochim. Biophys. Acta - Biomembr. 2013, 1828 (11), 2428–2435. 77 Figure 12 | Carrier activity obtained from the HPTS/DPX assay. a, Schematic representation of the transport of otherwise impermeable analytes facilitated by superchaotropic cluster anions with encapsulated HPTS/DPX probe/quencher pair used for signalling. In the presence of suitable carriers, cargo is carried into the vesicles and the cationic quencher DPX is transported from the endoliposomal phase to the bulk, resulting in increased fluorescence of HPTS as the optical signal of successful cargo transport. b, Changes in HPTS emission (I) (λex= 413 nm, λem = 511 nm) in EYPC⊃HPTS/DPX vesicles as a function of time during the addition of the membrane carriers B12Br122− and B12Cl122−, the established activator pyrenebutyrate, as well as the inactive parent B12H122−, as negative control; clusters (150 µM) added at t = 50 s, WR7 at t = 100 s and Triton X-100 at t = 600 s, for calibration. c, Dependence of fractional activity Y for WR7 transport on the concentration of different clusters, and the corresponding Hill curve fits in comparison to reference compound pyrenebutyrate. d, Transport efficiency of B12Br122− towards selected impermeable analytes of biological/medicinal relevance; values obtained by non-linear regression analysis (Hill equation), see Table 1. e, Comparable transport of WR7 (circles) and WK7 (triangles or squares) activated by B12Br122− (red, solid lines) versus exclusive transport of WR7 activated by pyrenebutyrate (black, dashed lines). f, Successful transport of membrane-impermeable analytes (100 µM) affected by B12Br122− (red, solid line) in the HPTS/DPX assay and transport failure for the established amphiphilic activator pyrenebutyrate (black, dashed lines). g, ‘Conventional’ (dashed line) versus reverse addition (solid line), demonstrating that the sequence of addition does not markedly affect transport efficiency. h, Transport efficiency in anionic DMPE/DPPG/CHOL⊃HPTS/DPX vesicles as a function of B12Br122− (red, solid line) versus pyrenebutyrate (black, dashed line) concentration at constant cargo concentration (WR7). Concentrations of carrier and cargo, when fixed, were 40 and 20 µM, respectively, unless explicitly stated. U-tube transport experiments, across a bulk chloroform layer, unambiguously confirmed that the brominated cluster served as a noncovalent carrier across a hydrophobic barrier (Figure 14a, b). To quantitatively characterize the transport efficiency of the new cluster carriers, the normalized fluorescence responses in the vesicle experiments (Figure 12b) were plotted against cluster concentration to afford doseresponse curves (Figure 12c and Figure 15a). We extracted, by Hill analysis, 78 the salient parameters: the maximal activity (Ymax), the concentration needed to achieve 50 % of Ymax (EC50), and the activator efficiency (Ea).81 The data (Figure 12b, Table 1) showed that transport is activated for the chlorinated boron cluster but becomes most efficient for the brominated one, with Ymax = 95 % and Ea = 6.1. This activity rivals the pyrenebutyrate gold standard in the field (Ea ≈ 5),81 which we used throughout as a control (see structure in Chart 1). From a molecular design point of view, it transpired that the transport activity critically depends on cluster size and type. B12H122–, the least chaotropic cluster, is inactive in the vesicle experiments, although it shows an onset of transport in the U-tube experiments (Figure 14a, b). The largest and most chaotropic one, B12I122–, appears to show too high activity (membrane affinity) as it causes membrane disruption even in the absence of cargo. The chlorinated cluster already shows sizable transport activity, but the “sweet spot” in this homologous cluster series is reached for B12Br122–, the prototypical superchaotropic carrier. Table 1 | Characteristic membrane transport parameters of boron clusters as activators of WR7 transport in vesicles.a,b Compound Ymax (%) c EC50 (µM) d Ea e Pyrenebutyrate 100 103 4.9 B12H122– f n.a. h n.a. --- B12Cl122– g 10 42 0.69 B12Br122– f 95 48 6.1 B12I122– f d.l i d.l. --- a All experiments were conducted with 20 µM WR7 as cargo. b Determined by Hill analysis. c Maximal activity; 5 % error (s.d.) unless stated differently. d Effective concentration to reach 50 % activity of Ymax; 10 % error (s.d.). e Activator efficiency; 10 % error (s.d., calculated by considering error propagation with respect to both, Ymax and EC50). f Used as sodium salt. g Used as cesium salt. h n.a. = no detectable activity. i d.l. = dye leakage, no determination possible. 79 Chart 1 | Chemical structures of the low molecular weight analytes studied. Predominant charges at neutral pH are shown. 80 Figure 13 | Assays testing for stability versus leakage of the vesicles. a, Response of the HPTS/DPX assay towards a membrane disrupting agent and chemical structures of the fluorescent dye HPTS and the quencher DPX. In this assay, large unilamellar vesicles are loaded with the HPTS/DPX probe/quencher pair, such that the fluorescence of HPTS is quenched. While the addition of a membrane-compatible carrier causes only a response in the presence of suitable cargo (see Figure 12) the addition of a membrane-disrupting agent (B12I122–) induces probe and quencher efflux, resulting in dilution with a concomitant sudden increase in fluorescence even in the absence of cargo. b, Changes in fractional HPTS emission intensity (λex = 413 nm, λem = 511 nm) of EYPC⊃HPTS/DPX vesicles (13 µM phospholipids in 81 10 mM Tris, 107 mM NaCl, pH 7.4) during the addition of 40 µM of clusters. The cluster B12I122– (red trace) was identified as lytic agent because the mere addition of the cluster led to a fluorescence increase. None of the other clusters (B12H122–, B12Cl122–, and B12Br122–) produced any perturbation in the vesicle membrane as monitored by a constant fluorescence intensity (baseline, black traces). c, Size distribution (by DLS) of EYPC⊃HPTS/DPX vesicles before and after the addition of 200 µM cluster and 20 µM WR7 to differentiate the clusters that retain membrane integrity (B12H122–, B12Cl122–, and B12Br122–) from the one that lyses the vesicles (B12I122–). d, Working principle of the carboxyfluorescein (CF) leakage assay and chemical structure of the fluorescent probe. In this assay, large unilamellar vesicles are loaded with CF at high concentration, such that the fluorescence is quenched. The addition of a membranedisrupting agent induces dye efflux, resulting in a CF dilution with the concomitant increase in fluorescence. e, Changes in fractional CF emission intensity (λex = 492 nm, λem = 517 nm) of EYPC⊃CF vesicles (13 µM phospholipids in 10 mM HEPES 107 mM NaCl, pH 7.5) during the addition of 40 µM of clusters and 20 µM WR7. The cluster B12I122– (red trace) was identified as a lytic agent because its mere addition led to a fluorescence increase. None of the other clusters (B12H122–, B12Cl122–, and B12Br122–) produced any perturbation in the vesicle membrane (even in the presence of WR7) as monitored by a constant fluorescence intensity (baseline). f, Changes in the fractional CF emission intensity (λex = 492 nm, λem = 517 nm) of DMPE/DPPG/CHOL⊃CF vesicles (13 µM phospholipids in 10 mM Tris, 140 mM NaCl, pH 7.4) upon addition of 300 µM of WR7 (red trace) or B12Br122– (black trace), demonstrating lysis caused by the former and retained membrane integrity in the presence of the latter. g, Compatibility of the carrier B12Br122– (blue, up to ca. 1 mM) with anionic liposomes (DMPE/DPPG/CHOL 1/2/1) in comparison to the strongly membrane-disrupting WR7 peptide (black) and pyrenebutyrate (red). EC50 for WR7 = 190 ± 50 µM and B12Br122– = 960 ± 100 µM. In b, e and f, Triton-X 100 was added at t = 600 s to induce vesicle lysis and allow data normalization. 82 Figure 14 | Visualization of transport through a hydrophobic phase in U-tube experiments. a, Photographs of the tri-phasic U-tube system at the start of an experiment (t = 0, top) and after 24 h (bottom). In the U-tube experiments, the trans compartment contained initially only the fluorescent HPTS dye and the cis compartment was administered with combinations of carrier (boron cluster), peptide cargo (WR7), and DPX, which serves as a quencher of HPTS. While the trans phase is initially strongly fluorescent (top), transport of DPX from the cis phase (10 mM Tris, 107 mM NaCl, pH 7.4) across the chloroform phase leads to a fluorescence quenching. b, HPTS fluorescence measured in the trans buffer (2 mM HPTS, 10 mM Tris, 107 mM NaCl, pH 7.4) upon addition of the following agents to the cis buffer: 40 M B12Br122– and 20 M WR7 (▲); 20 mM DPX (×); 20 M WR7 and 20 mM DPX (); 40 M B12H122–, 20 M WR7, and 20 mM DPX (); 40 M B12Br122– and 20 mM DPX (•); 40 M B12Br122–, 20 M WR7, and 20 mM DPX (■);. The experiments demonstrate that there is (i) no significant fluorescence quenching in the absence of added DPX quencher (▲), (ii) no significant transport of the DPX quencher in the absence of carrier and peptide cargo (×), (iii) no significant transport of the DPX quencher in the presence of peptide cargo but in the absence of a carrier (), (iv) slow and inefficient transport of DPX in the presence of peptide cargo and the least efficient carrier, B12H122– (), (v) faster and more efficient transport of DPX in the absence of peptide cargo and the very efficient carrier, B12Br122– (•), and (vi) the fastest and most efficient transport of DPX in the presence of peptide cargo and the very efficient carrier, B12Br122– (■). The experiments show that B12Br122– is itself an effective carrier of DPX, but that the carrier/peptide cargo system leads to the most effective transport. c,d, Calorimetric evidence for direct interaction of B12Br122− with peptide cargos as obtained from raw ITC data (top) and apparent reaction heats obtained from the integration of the calorimetric traces (bottom) for the sequential injections of 2 mM of B12Br122− into c, 250 µM WR7 and d, 450 µM WK7. 15 % error for Ka and ± 0.5 kcal mol−1 for ΔH and TΔS (duplicate measurements). The calorimetric titrations show enthalpically driven non-stoichiometric intermolecular interactions between the cluster and the peptides. 83 2.2. Broadband carrier characteristics Once the brominated cluster, B12Br122–, had emerged as exponent of a new class of highly active synthetic membrane carriers, we focused on the scope of compounds that can be transported.167 For amphiphilic activators with anionic headgroups such as pyrenebutyrate, the synergy of favourably oriented hydrogen bonds and permanent electrostatic interactions that the guanidinium groups can form with the carboxylates are considered prerequisites for efficient cargo translocation.77,168 As a consequence, while the transport of arginine-rich peptides is straightforward, that of lysine-rich peptides – which can be reversibly deprotonated to minimize charge repulsion – is already substantially more difficult169 and requires a meticulous design of suitable receptors in the carriers.170 As a challenge, we studied first the corresponding heptalysine (WK7) and then others as alternative cargos for B12Br122– (Figure 12d). The cargo screening experiments were conducted at a constant B12Br122– concentration of 40 µM, such that the extracted EC50 values and Ymax values can be used to define a relative scale of transport efficiency, Et, where we set the value for the oligoarginine reference to 10. Surprisingly, without further carrier design, B12Br122– showed very similar transport efficiency for WK7 as for WR7 (Figure 12d, e), which clearly breaks the trend for amphiphilic activation (e.g., by pyrenebutyrate) that fails to trigger any signal of oligolysine membrane transport (Figure 12e). This non-canonical membrane translocation of the WK7 peptide by the boron cluster was further characterized by isothermal titration calorimetry (ITC), which confirmed an enthalpy-driven interaction of the B12Br122– cluster with both cationic peptides in homogeneous solution (Figure 14c, d). Intriguingly, the binding affinity of the boron cluster was even slightly stronger with the lysine peptide (WK7) than with the arginine one (WR7), which demonstrates that other interactions but the conventional ones (Coulombic, salt bridges, or hydrogen bonding) are important contributors to the cluster-peptide affinity and translocation. The dynamic, enthalpically driven binding of globular boron clusters to their cargo is a reflection of their generic affinity to hydrophobic matter 167 Matile, S.; Vargas Jentzsch, A.; Montenegro, J.; Fin, A. Chem. Soc. Rev. 2011, 40, 2453– 2474. 168 Rothbard, J. B.; Jessop, T. C.; Lewis, R. S.; Murray, B. A.; Wender, P. A. J. Am. Chem. Soc. 2004, 126 (31), 9506–9507. 169 Robison, A. D.; Sun, S.; Poyton, M. F.; Johnson, G. A.; Pellois, J. P.; Jungwirth, P.; Vazdar, M.; Cremer, P. S. J. Phys. Chem. B 2016, 120 (35), 9287–9296. 170 Pan, Y. C.; Barba-Bon, A.; Tian, H. W.; Ding, F.; Hennig, A.; Nau, W. M.; Guo, D. S. Angew. Chemie - Int. Ed. 2021, 60 (4), 1875–1882. 84 (chaotropic effect).123 Accordingly, chaotropic carriers should be potentially capable of transporting not only cationic but also neutral targets. This “broadband carrier” hypothesis was validated with membrane-impermeable molecules of biological interest. Our targets included differently charged biomolecules (such as acetylcholine and amino acids), vitamins, antibiotics, neuromuscular blocking agents, and proteins. Remarkably, B12Br122– transported many types of cargo, ranging from positive to non-charged and zwitterionic molecules, or from small ones such as acetylcholine (molecular mass of 146 Da) to larger polypeptides such as protamine (molecular mass of 4,500 Da), with the exception of the negatively charged molecules glutamate and albumin, for which no carrier-cargo charge attenuation can occur (Figure 12d, Figure 15b and Table 2). Most striking was the very fast transport kinetics, within seconds for most cargo types (Figure 12f), which becomes competitive with uptake through membrane pores or channels.171 The insensitivity towards the chemical nature of the diverse functional groups in the transported cargos (Chart 1) confirmed that the carrier activity of the boron clusters is not limited to residues that entertain salt bridges or specific intermolecular interactions, and is much less restrictive than for amphiphilic activators. Indeed, the prototype amphiphile pyrenebutyrate showed no activity for any of the newly introduced targets (Figure 12f). 171 Ghale, G.; Lanctôt, A. G.; Kreissl, H. T.; Jacob, M. H.; Weingart, H.; Winterhalter, M.; Nau, W. M. Angew. Chemie - Int. Ed. 2014, 53 (10), 2762–2765. 85 Figure 15 | Transport activity. a, Transport activity, Y, in EYPC⊃HPTS/DPX vesicles as a function of cluster concentration at constant cargo concentration (WR7, 20 µM). b, Transport activity, Y, for B12Br122– (40 µM) in EYPC⊃HPTS/DPX vesicles as a function of the concentration of different cargo types; red lines correspond to the resulting fit with the Hill equation. 92 performed. In addition to the control experiments using a hydrolysisresistant peptide (the enantiomeric peptide TAMRA-D-R8) and the HPLC analysis of cytosolic extracts (Figure 18d, e), we performed transport experiments with the non-conjugated TAMRA probe (Figure 20). These experiments revealed that – in contrast to the labelled TAMRA-R8 peptide – the isolated fluorescent dye is not carried by B12Br122–. Even at fifty times higher dye concentrations than the peptide, no relevant TAMRA fluorescence signal was observed in the cells. Figure 20 | TAMRA uptake control experiments in HeLa cells. Cells were incubated with 1 µM TAMRA-R8 (as a positive control, left) and 1, 10, or 50 µM non-conjugated TAMRA (second to fourth columns) in the absence (top row) or in the presence of either 10 µM B12Br122– or 10 µM B12H122– (middle and bottom) diluted in HKR buffer for 1 h, washed for 5 min with 0.1 mg/mL heparin and HKR buffer, and imaged by confocal fluorescence microscopy. Where detectable, images show TAMRA-R8 or TAMRA fluorescence (red) and the brightfield in the insets; scale bars are 50 µm. 2.4. New cargo types and biological activity enhancement Beyond enhancing membrane translocation of cationic peptides in vesicles and living cells, boron clusters also show a broad scope of accessible cargo types including neutral hydrophilic molecules (Figure 12d, f; Table 2). One of the successfully transported non-charged hydrophilic cargos in vesicles is phalloidin, a rigid bicyclic heptapeptide that is long known in cell biology for 93 its ability to bind to F-actin of the cytoskeleton.174 At the same time, phalloidin is notorious for resisting internalization, and cell fixation and membrane permeabilization are traditionally employed for cytoskeleton labelling purposes.175 Thus, phalloidin delivery in living cells has been explored by covalent modifications of the cargo itself with polycationic dendrimers,175 or by membrane-disrupting strategies such as optoporation,176 the addition of pore-forming toxins,69 or redox-sensitive polymer-based strategies.177 However, a routine strategy for phalloidin delivery, such as one based on the addition of a low-molecular mass noncovalent carrier, has been elusive. Phalloidin-TRITC transport experiments with living HeLa cells showed that B12Br122– triggered its direct membrane passage to the cytosol and afforded an excellent staining of the F-actin target even at 500 nM cargo concentration (Figure 21). This protocol was transferable to GT1-7 mouse hypothalamic GnRH neuronal cells, human retinal pigmentary epithelial cells ARPE-19, and adenocarcinoma human alveolar basal epithelial cells A549, as shown in Figure 22a-d. In contrast, when attempting to use the prototypical octaarginine penetrating peptide AcR8 as an alternative noncovalent carrier for the same set of cell lines, only trace levels of cytosolic phalloidin were observed in all cell lines (Figure 22a-d). Under the transport experimental conditions, the superchaotropic cluster B12Br122– also showed a lower toxicity than its penetrating peptide competitor AcR8 (Figure 22e). 174 Melak, M.; Plessner, M.; Grosse, R. J. Cell Sci. 2017, 130, 525−530. 175 Li, K.; Pu, K.-Y.; Cai, L.; Liu, B. Chem. Mater. 2011, 23, 2113−2119. 176 Dhakal, K.; Black, B.; Mohanty, S. Sci. Rep. 2014, 4, 6553. 177 de Vries, W. C.; Grill, D.; Tesch, M.; Ricker, A.; Nüsse, H.; Klingauf, J.; Studer, A.; Gerke, V.; Ravoo, B. J. Angew. Chemie - Int. Ed. 2017, 56 (32), 9603–9607. 94 Figure 21 | B12Br122−-assisted phalloidin-TRITC transport into living HeLa cells. Cells were incubated with 0.5, 2.5 and 5 µM phalloidin-TRITC (red, from left to right) in the absence (top row) and presence of different concentrations of B12Br122− cluster (10, 25 and 50 µM; second, third and fourth rows, respectively) in HKR buffer for 3 h, subsequently stained with Hoechst (blue), washed for 5 min with 0.1 mg/mL heparin in HKR buffer and imaged by confocal fluorescence microscopy; bright-field images in insets. Representative images of two biological replicates. Staining results were insensitive to sequence of addition (see Figure 23). Scale bars are 50 µm. 95 Figure 22 | B12Br122–-assisted phalloidin-TRITC transport into living cells. a, HeLa, b, GT1-7, c, ARPE-19, and d, A549. Cells were incubated with 7 µM phalloidin-TRITC (red) in the absence (left) and presence of 50 µM AcR8 (middle) or 50 µM of B12Br122– cluster (right) in HKR buffer for 3 h, subsequently stained with Hoechst (blue), washed with DMEM without phenol red, and imaged by confocal fluorescence microscopy; bright-field images in insets. Micrographs from each cell line were processed independently, experiments were performed twice. Scale bars are 25 µm. e, MTT viability assay of the corresponding cell lines (HeLa, GT1-7, ARPE-19, and A549, from left to right) at different concentrations (0, 25, 75, 250, and 500 µM) of B12Br122– (black) or AcR8 (red); data are mean ± s.d. of five measurements. According to the dynamic nature of chaotropic interactions, transport should be insensitive to the sequence of addition of cluster and cargo. To demonstrate this, cellular uptake of phalloidin-TRITC after varying 96 sequences of addition and pre-incubation was investigated (Figure 23). Regardless of whether cluster was added first and phalloidin later, or the reverse, or whether both were added at the same time, with or without a pre-incubation time, efficient phalloidin transport and actin staining was observed in all cases. This robustness can be an advantage in comparison to methods that require a previous adsorption or fixation of the carrier to the cargo. In alternative approaches that require adsorption or encapsulation, e.g., lipofection, the order of mixing and the incubation time between cargo and carrier affects the transport efficiency and, thus, it needs to be carefully optimized for a successful delivery. The actual insensitivity of the boron clusters to the incubation process, in vesicles and in cells, rules out the possibility of any pre-adsorption requirement. Figure 23 | B12Br122–-assisted phalloidin-TRITC transport into living cells with different sequence of addition and incubation modes. HeLa cells were treated with 2.5 µM phalloidin-TRITC (red) in the absence and presence of different concentrations of B12Br122– cluster (10, 25, and 50 µM; left to right column) in HKR buffer for 3 h with varying sequence of addition: a, by adding first phalloidin-TRITC (cargo) to the cells, subsequently the cluster; b, by adding first the cluster, subsequently the cargo; c, by premixing cargo and cluster and adding the mix to the cells immediately or d, same as in c, but after 20 min pre-incubation of the cargo-cluster mixture. In the next step, cells were stained with Hoechst (blue), washed for 5 min with 0.1 mg/mL Heparin and HKR buffer, and imaged by confocal fluorescence microscopy; brightfield images in insets, inset scale bars are 50 µm. 97 Proteolysis targeting chimeras (PROTACs) are small molecules with a bright future as the next generation of drugs for the removal of specific unwanted proteins. We tested whether the use of boron clusters could contribute to enhance the activity of dBET1, a well-characterized PROTAC that is known for its undesirable low permeability178 and that should fall within the potential cargo scope of B12Br122– (neutral, molecular mass of 785 Da). The internalization of the PROTAC in the absence and presence of the cluster was assessed by its ability to bind to the Cereblon (CRBN) E3 ligase by using the NanoBRET™ TE intracellular E3 ligase assay (Figure 24a, b). In this assay, a cluster-enhanced uptake of dBET1 was indeed observed (factor 2-3 decrease in IC50 value), which illustrates the versatility of the new carriers. We also demonstrated the cluster-mediated intracellular transport of monomethyl auristatin F (MMAF, zwitterionic, molecular mass of 732 Da), an antineoplastic drug with considerably lower permeability in comparison to other auristatins.179 For this bioactive cargo, the B12Br122– cluster was found to effectively reduce its IC50 value by more than a factor of 2, as assessed through the viability of HeLa cells (Figure 24c, d). Antibiotics delivery is another area where novel carrier concepts are intensively being sought for, and vesicle studies described here had demonstrated transport of ampicillin and kanamycin A by the prototype chaotropic cluster carrier, B12Br122– (Figure 12d, f). As a proof-of-principle, we investigated its potential to reduce the minimum inhibitory concentration of kanamycin A, an aminoglycoside antibiotic. Aminoglycosides function by binding to the bacterial 30S ribosomal subunit; consequently, effective passage through the cell wall and plasma membrane is essential for aminoglycosides to reach their intracellular targets.180 The antibiotic resistance of the Gram-negative Escherichia coli Top10 strain to the action of kanamycin A (3.5 µg/mL) was investigated in the absence and presence of B12Br122– (Figure 24e). In the absence of cluster, E. coli retained viability (60 %), but, in the presence of the cluster carrier, kanamycin A showed potent antibacterial activity (< 1 % viability). The fact that incubation with the cluster alone did not affect bacterial survival up to 1 mM demonstrates that the combination of both, antibiotic and carrier, is essential to prompt the biological response. 178 Winter, G. E.; Mayer, A.; Buckley, D. L.; Erb, M. A.; Roderick, J. E.; Vittori, S.; Reyes, J. M.; di Iulio, J.; Souza, A.; Ott, C. J.; et al. Mol. Cell 2017, 67 (1), 5–18.e19. 179 Doronina, S. O.; Mendelsohn, B. A.; Bovee, T. D.; Cerveny, C. G.; Alley, S. C.; Meyer, D. L.; Oflazoglu, E.; Toki, B. E.; Sanderson, R. J.; Zabinski, R. F.; et al. Bioconjug. Chem. 2006, 17 (1), 114–124. 180 John, T.; Thomas, T.; Abel, B.; Wood, B. R.; Chalmers, D. K.; Martin, L. L. Biochim. Biophys. Acta - Biomembr. 2017, 1859 (11), 2242–2252. 98 The combined transport experiments and the successful functional delivery of different bioactive cargos demonstrate that boron clusters, and prominently B12Br122–, are able to transport the intact agents through the cellular bilayer membrane, at physiologically relevant concentrations, and to induce the corresponding enhanced biological effects. Nevertheless, this avenue is still preliminary, especially for the delivery of (bio)macromolecules. 99 Figure 24 | Boron cluster-induced enhancement of biological activity. a, Doseresponse experiment for dBET1 binding to the CRBN E3 Ubiquitin ligase, quantified with the CRBN NanoBRET target engagement assay, in the presence of 0, 25, and 50 µM B12Br122–. In this assay, HEK293 cells are transfected with a plasmid expressing a CRBNNanoLuc fusion protein for 24 h and subsequently incubated with a labelled tracer molecule (0.5 µM) and the test compounds. dBET1 acts as a competitive inhibitor and decreases the NanoBRET signal between CRBN-NanoLuc and the tracer. Solid points indicate means of three technical replicates; error bars indicate standard deviation. Values were normalized to the controls without dBET1, as indicated in the Methods section. b, Corresponding IC50 values calculated with the CRBN Target Engagement assay. Crossbars and error bars indicate mean and standard deviation, respectively, of three independent experiments (two with technical triplicates, one with technical duplicates); values obtained 100 in each experiment are represented by different shapes. c, Viability of HeLa cells after incubation with different doses of MMAF in the presence of 0, 5, and 10 µM B12Br122–. Solid points indicate the mean of three technical replicates; error bars indicate standard deviation. d, Corresponding IC50 values of MMAF. Crossbar and error bars indicate mean and standard deviation, respectively, of four independent experiments (each one with technical triplicates); each experiment is represented by a different shape. e, E. coli Top10 viability in the presence of different concentrations of kanamycin A monosulfate (03.5 µg/mL) and B12Br122– (0-1000 µM) in LB broth at 37 ºC; data are mean ± s.d. of triplicate measurements. 101 3. CONCLUSION The rational design of effector molecules with biological activity is constrained by physicochemical concepts traditionally derived from the observation of molecules or processes found in nature. The paradigm of amphiphilicity has, in particular, governed the design of membrane carriers for the last fifty years.73 The hydration-thermochemical properties of superchaotropic anions differ from those of hydrophobic or amphiphilic solutes and, on a continuous scale of solvation in water, they fall in between hydrophobic ions and conventional chaotropes.123 The chaotropic effect, that is, the interaction of superchaotropes with hydrophobic phases, surfaces, and concavities, is enthalpically driven, by a combination of desolvation effects as well as strong dispersion interactions, and it differs from the thermochemical signature of the classical hydrophobic effect, which is entropically driven.123 Our results introduce superchaotropic globular boron cluster anions as a chemically distinct class of membrane carriers. The clusters obviate the traditional amphiphilic transport mechanism in that they operate by a direct chaotrope-mediated translocation (Figure 11b), as jointly experimentally confirmed by vesicle, ITC, U-tube, and cellular assays for different cargos (Figure 12d). The here reported enthalpy-driven complexation and non-canonical transport of cationic peptides with either guanidinium or ammonium moieties (Figure 12e), the retained transport regardless of the sequence of cargo/carrier addition (Figure 12g), the independence of cargo uptake on membrane charge (Figure 12h), and the efficient translocation of selected neutral hydrophilic cargos (Figure 12d, f) in vesicles and cells distinguish these new globular anionic carriers. In this complementary membrane transport concept, the low dehydration penalty and the strong dispersion interactions of superchaotropic clusters minimize the repulsion between the hydrophilic molecules and the membrane barrier and, thus, allow the direct passage of a broad scope of cargos across lipid membranes. The use of superchaotropic clusters expands the molecular toolbox to affect transport of membrane-impermeable hydrophilic molecules, which will have implications for cell-biological, neurobiological, and physiological investigations. Owing to the biocompatibility181 and broad cargo scope of the boron clusters, as well as the enhancements in bioactivity of different 181 Hey-Hawkins, E.; Viñas, C. (Eds.) Boron‐based Compounds: Potential and Emerging Applications in Medicine. John Wiley & Sons, 2018. 108 integration time of 1 s). Background correction was carried out by subtracting the signal of a sample without tracer. Values of each B12Br122– concentration series were normalized to the BRET readout of the controls without dBET1. Data were analyzed with R (v. 4.0.3).185 4.14. Cytosolic TAMRA-R8 concentration Cytosolic extracts were obtained according to a previously described protocol186 by incubation with digitonin, a steroidal saponin that preferentially permeabilizes cholesterol rich membranes, such as the plasma membrane, with minor effects on intracellular membranes. Briefly, HeLa cells were seeded at 260,000 cells/well in 6-well plates, washed the next day twice with HKR, incubated with 1 µM TAMRA-R8 (the L enantiomer) in the presence or absence of 10 µM B12Br122– for 1 h, washed twice with HKR, three times with 2 mg/mL heparin in HKR, and once with ice-cold PBS containing calcium and magnesium. Cells were incubated on ice with 600 µL of 35 µg/mL digitonin in PBS Ca/Mg for 10 min, the supernatant with the cytosolic fraction collected, and cells washed with 200 µL of PBS Ca/Mg, combining this supernatant with the previous extract. The non-cytosolic fraction was collected by incubation of the cells with 800 µL of 1 % Triton X-100 in PBS. TAMRA fluorescence of the extracts was determined in a plate reader (Tecan Infinite 200Pro, λex = 555 nm, λem = 585 nm) and concentrations were calculated by using a calibration curve with serial dilutions of TAMRA-R8. For the complementary HPLC analysis, phosphate buffer was replaced by TBS (20 mM Tris-HCl, pH 7.2, 150 mM NaCl, 0.5 mM CaCl2, 0.5 mM MgCl2) and digitonin extraction was performed as indicated above. An aliquot of these extracts was used for β-hexosaminidase activity determination. Cytosolic extracts were lyophilized and resuspended in 1:10 volumes of H2O:CH3CN 1:1 with 1 % TFA and analyzed by HPLC [RP-HPLC Agilent Luna 5U C18 100 Å, H2O (0.1 % TFA)/CH3CN (0.1 % TFA) 100:0 (0→5 min); 100:0→5:95 (5→20 min)] by monitoring the 555-nm absorbance of the TAMRA chromophore. The quality of fractionation was assessed by lysosomal βhexosaminidase activity, using 4-nitrophenyl 2-acetamido-2-deoxy-β-Dglucopyranoside as substrate. Briefly, 20 µL of extract were incubated with 80 µL of 7.5 mM substrate in 100 mM citrate buffer, pH 4.7, for 40 min at 37 °C, and the reaction was stopped by addition of 200 µL of 0.2 M Tris solution. Absorbance at 405 nm was measured in a plate reader. As blank, wells containing only the substrate were used. The enzymatic activities 186 Evans, B. C.; Hocking, K. M.; Kilchrist, K. V.; Wise, E. S.; Brophy, C. M.; Duvall, C. L. ACS Nano 2015, 9 (6), 5893–5907. 109 were found to be 3.2 ± 2.0 % in the presence of the peptide and 5.4 ± 1.0 % in the presence of peptide and cluster, confirming a high purity of the cytosolic fractions. 4.15. ICP-MS HeLa cells, seeded at 260,000 cells/well in 6-well plates the day before, were washed with HKR and incubated for 3 h with 2.5 mL/well of 50 µM of each boron cluster diluted in HKR. Cells were washed with HKR containing 0.1 mg/mL heparin, twice with HKR, and subsequently lysed with concentrated nitric acid (69 % HNO3). Cells from nine wells were pooled for each sample. Lysates were diluted prior to analysis by ICP-MS in an Agilent 7700x equipped with a MicroMist glass low-flow nebulizer, a double-pass spray chamber with a Peltier system (2 ºC), and a quartz torch. A calibration curve for the element boron (B) between 10 and 1000 µg/L was prepared with the element germanium (Ge) as internal standard. The ICPMS instrument parameters were as follows: RF Power: 1550 W, sample depth: 8, carrier gas flow: 1.1 L/min, nebulizer pump speed: 0.1 rps, S/C temperature 2 ºC. Other parameters were set as follows: extract 1: 0, extract 2: –175, omega bias: –100, omega lens: 12.6, cell entrance: –40, cell exit: –60, deflect: 0.4, plate bias: –60, QP bias: –15, OctP RF: 180, OctP bias –18, He gas: 3.6, discriminator: 4.5 mV, analog HV: 1730 V, pulse HV: 954 V. 4.16. Flow cytometry HeLa cells were seeded at 10,000 cells/well in 96-well plates. The next day, they were incubated for 1 h with the indicated compounds diluted in HKR. Cells were subsequently washed for 5 min with HKR containing 0.1 mg/mL heparin, washed again with HKR, and trypsinized. Trypsin was neutralized with PBS containing 2 % FBS and 5 mM EDTA. TAMRA fluorescence was excited with a green laser (532 nm) and measured on a Guava easyCyte BG HT collecting the emission at 620/52 nm (Orange-G channel) and using InCyte v. 3.2. (GuavaSoft, Millipore). Data were analyzed with R (version 4.0.3)185 and the packages CytoExploreR (v. 1.0.8)187 and ggcyto (v. 1.18.0).188 Cells with typical FSC and SSC parameters were selected, and the median fluorescence intensity calculated for each sample (MFI). Each condition was measured in triplicate. 187 Hammill, D. CytoExploreR: Interactive Analysis of Cytometry Data. R package version 1.0.8 https://dillonhammill.github.io/CytoExploreR/, 2020. 188 Van, P.; Jiang, W.; Gottardo, R.; Finak, G. Bioinformatics 2018, 34 (22), 3951–3953. 110 4.17. Synthesis and characterization of TAMRA-D-R8 TAMRA-D-R8 was synthesized via manual Fmoc solid-phase peptide synthesis, using Fmoc-Rink amide resin (loading: 0.19 mmol/g), as previously described.39 TAMRA-D-R8 was obtained after RP-HPLC purification with an overall yield of 17 % (15 mg) in 99 % purity. It was characterized on an RP-HPLC Agilent SB-C18 column, H2O (0.1 % TFA)/ CH3CN (0.1 % TFA) 95:5→5:95 (0→12 min)]. Rt: 5.96 min. MS (ESI): 1,124.7 (9, [M+2H+4TFA]2+), 1,067.9 (17, [M+2H+3TFA]2+), 1,011.0 (14, [M+2H+2TFA]2+), 712.4 (37, [M+3H+3TFA]3+), 674.2 (100, [M+3H+2TFA]3+), 636.3 (95, [M+3H+TFA]3+), 598.2 (36, [M+3H]3+), 534.5 (24, [M+4H+3TFA]4+), 506.0 (36, [M+4H+2TFA]4+), 477.5 (48, [M+4H+TFA]4+), 449.1 (62, [M+4H]4+). 111 CHAPTER II DECAAND DODECABORATE CLUSTERS: CHANGING THE SIZE OF BORON CORE AS A TOOL FOR MODULATING THE IMPACT OF HALOGEN SUBSTITUENTS. 112 1. OBJECTIVES In the previous chapter, the superchaotropicity of boron clusters has been described as a new disruptive concept for the transport of certain hydrophilic cargos, as opposed to the amphiphilic dogma. It has been shown that increasing the polarizability of the halogen substituents of the dodecaborate boron clusters, with the formula B12X122−, in the series of halogens (X = Cl, Br, I), hence increasing the chaotropicity of the cluster,116 is key to control cargo-membrane interactions, and thus transport across the lipid bilayer. The previous chapter was focused on how the substituent type (B12X122−, X = H, Cl, Br, or I) modulated the cluster chaotropicity, and therefore the ability to interact with and deliver cargos across model and cell membranes. In this study, only the type of substituent was modified and the dimensions of the boron clusters were limited to the dodecaborate series. However, the size of the boron clusters can also be controlled by changing the number of boron atoms in the core, which, in the current study, is equal to the number of atoms in the outer layer. The modification of the number of core atoms, and thus the number of halogen substituents, will in turn impact the size and polarizability of the cluster entity. In this section we have investigated two families of boron clusters, namely B12X122− and B10X102− (Figure 25), and studied the impact of the size of the cluster and the halogen substitution in their transport capacity of model cargo molecules to the cell cytosol. Figure 25 | Globular boron cluster structures. Space-filling molecular models (top) and chemical structures (bottom, pink spheres represent B−H, green spheres are B−Cl, orange spheres are B−Br, and purple spheres are B−I) of decaborate (B10X102−) and dodecaborate (B12X122−) clusters. As previously pointed out, Assaf and Nau introduced a chaotropicity scale123 based on ion hydration parameters proposed by Marcus.113,189,190 This scale 189 Marcus, Y. J. Solution Chem. 1994, 23, 831–848. 190 Marcus, Y. Ion Properties. Marcel Dekker: New York, 1997. 113 allows the direct comparison of structurally quite different ions by comparing the water-structural entropy for ionic hydration of each one. The higher values of water-structural entropy during hydration imply that their desolvation will lead to water structure recovery, with an increase of the number of hydrogen bonds. Thus, binding of superchaotropic anions (those with a TΔSstruct > 9 kcal · mol−1 or, in other words, those that disrupt more than two hydrogen bonds in their solvation shell)123 to other molecules will have an unfavourable entropic component. The globular clusters taken into account in this chapter are displayed in Table 3, together with their radius, polarizability, and water-structural entropy, the latter estimated according to Marcus theory.113,189,190 Table 3 | Chaotropicity and polarizability scales. Ordering of superchaotropic anions according to their chaotropicity, according to their size (radius, r; calculated from the diameter as the distance between the outer atoms including the van der Waals radii obtained from their geometry-optimized structures) and water-structural entropy for ionic hydration (TΔSstruct, values estimated from cluster size).123 Polarizability (α) values were taken from ref. 191, except B12I122−, which was taken from ref. 116, and B10I102−, which was estimated by reducing in a 15 % the polarizability of the previous one, as it was observed for the other deca and dodeca-borates pairs. Anion r (pm) TΔSstruct (kcal mol−1) α (Å3) Anion r (pm) TΔSstruct (kcal mol−1) α (Å3) B10H102− 393 11.5 18.9 B12H122− 400 14.8 22.0 B10Cl102− 480 14.0 36.9 B12Cl122− 525 15.3 44.0 B10Br102− 500 14.5 48.7 B12Br122− 560 16.3 57.1 B10I102− 530 15.4 71.5 B12I122− 590 17.1 84.1 It has also to be taken into account that, in 1965, Kaczmarczyk and Kolski investigated the relationship between the polarizability (α) of deca- /dodecaborates and not only their volume but also their geometry. They observed that, while the icosahedral dodecaborates can be inscribed in a sphere with a defined radius r (see Table 3), the boron atoms of decaborate clusters form a bicapped square antiprism which can be inscribed in a prolate spheroid. So, while the polarizability of dodecaborates is isotropic and clearly proportional to the volume,116 for decaborates it is not so obvious. However, they calculated that, over a wide range of eccentricities, the polarizability of a spheroidal anion is still proportional to the volume. They also predicted that the ratio between the polarizabilities of the cores of two homologous clusters, decaand dodecaborate, should remain approximately constant and close to 0.83.191 Values of polarizability, that 191 Kaczmarczyk, A.; Kolski, G. B. Inorg. Chem. 1965, 4 (5), 665–671. 114 together with charge density also impact the hydration pattern of these anions,123 are given in Table 3. Therefore, in this chapter, we have selected the decaand dodecaborate clusters for a systematic investigation of the correlation between cluster size and halogen substituent and their biological activity. The objectives for this chapter are: • To compare the ability of these clusters to transport into the cells a neutral cargo, such as phalloidin. • To investigate their potential to carry a cationic cargo into cells, the proapoptotic (KLAKLAK)2 peptide. • To determine the toxicity of these clusters in cells. • To identify the principles that relate cluster activity and structure. 115 2. RESULTS & DISCUSSION 2.1. Transmembrane transport in living cells 2.1.1. Phalloidin–TRITC transport The first cargo used to investigate the impact of the number of the core boron atoms, halogen substituents and chaotropicity of the clusters on their potential to trigger membrane translocation, was a tetramethylrhodaminelabelled phalloidin (phalloidin–TRITC). As previously mentioned, phalloidin is a rigid bicyclic heptapeptide that is commonly employed in cell biology to label F-actin of the cytoskeleton.174 This molecule constitutes a non-charged and hydrophilic model cargo that is ideal to monitor potential membrane translocation and intracellular functional activity by cytoskeleton labelling (Figure 26). Confocal microscopy of living HeLa cells provided a visual and qualitative outcome of the internalization of the fluorescent cargo with the eight clusters tested. As already demonstrated in the previous chapter, the parental, and thus the smallest and less polarizable cluster of the B12X122− group, namely B12H122−, appeared inactive as a transmembrane carrier, since no F-actin staining was observed (Figure 26a). The B12Cl122− derivative, slightly larger and more chaotropic than B12H122−, showed negligible transport activity, which became moderate at the highest concentration of the chlorinated cluster tested (75 µM, Figure 26b, bottom micrograph). As previously observed, the brominated cluster (B12Br122−) showed an excellent transport capacity and cellular morphology, which confirmed the optimal level of chaotropicity of the brominated cluster for the transport of functional phalloidin–TRITC without damaging the cell membrane and thus providing an excellent F-actin staining (Figure 26c). On the other hand, the higher chaotropicity of the B12I122− would lead to a much stronger interaction of this cluster with both the cargo and the cell membrane. This excessive affinity of the iodinated derivative was reflected in the formation of fluorescent cargo aggregates and in the compromised morphology of the cells that would indicate potential toxicity (Figure 26d). We next investigated the potential delivery of phalloidin using the smaller decaborate clusters. As it would be expected for the decrease in size, polarizability and chaotropicity, the B10X102− series showed a significant decrease in transport efficiency of the phalloidin cargo (Figure 26e-h). The two smallest decaborates – B10Cl102− and B10H102− – appeared inactive, as no F-actin staining was observed (Figure 26e, f). In contrast to the optimal 116 dodecaborate brominated cluster (B12Br122−), the smaller B10Br102− showed just an onset of phalloidin transport and weak actin labelling (Figure 26g). However, the iodinated B10I102− presented the best transport activity for phalloidin–TRITC within the decaborate group, that was already remarkable at the lowest concentration tested (25 µM, Figure 26h, top micrograph), and comparable to the one observed with the brominated dodecaborate (B12Br122−). In addition, it was also observed that the cells presented a much healthier morphology in the presence of the smaller version of iodinated cluster B10I102−, which suggested that the reduction in the number of atoms reduced cluster’s toxicity in comparison to B12I122− (Figure 26h, bright-field micrographs). Figure 26 | Cluster-assisted phalloidin–TRITC transport into living HeLa cells. Confocal microscopy images of HeLa cells incubated with 2.5 µM phalloidin–TRITC (red) in the presence of different concentrations of globular decaand dodecaborate clusters (25, 50 and 75 µM; first, second, and third row, respectively) in HKR buffer for 3 h, subsequently stained with Hoechst (blue), washed with HKR buffer and imaged by confocal fluorescence microscopy; bright-field images in insets. Representative images of three biological replicates. Scale bars are 50 µm. To quantify and confirm the confocal microscopy results, the median fluorescence intensity of HeLa cells incubated with the eight clusters and the fluorescently labelled phalloidin was measured by flow cytometry. In all the conditions, the outcome was consistent with what was previously detected by microscopy (Figure 27). Particularly, it was observed that by increasing the polarizability of the substituents (H < Cl < Br < I), and thus the size of the decaborate cluster series, the transport capacity reached an optimal level. Above a certain degree of chaotropicity and concentration, toxicity increased, and cargo transportation diminished, and this was observed both for decaand dodecaborate cluster series. It was also observed that transport efficiency is concentration dependent: below the optimal level of chaotropicity, smaller clusters require higher concentrations for effective cargo delivery; beyond that point, high concentrations lead to toxicity and 117 less efficient delivery. The smallest and least chaotropic boron clusters – B12H122−, B10H102−, and B10Cl102− – were confirmed as inactive carriers for intracellular transport (Figure 27, bars). B12Cl122− and B10Br102− also showed comparable activity, as already observed via confocal microscopy. Interestingly, B12Br122− and B10I102− showed quantitatively almost identical phalloidin–TRITC internalization at the lowest concentration tested (25 µM); at higher concentrations, the latter resulted less active and likely more toxic, as could be inferred by the number of cells counted by the cytometer. Although a certain amount of phalloidin cytosolic delivery could lead to toxicity,192,193 this is unlikely to manifest at this short incubation time, and the live cell concentration in the samples analysed by flow cytometry provided a first experimental insight into the potential cytotoxicity of each cluster in each condition. These data were coherent with the previsions of chaotropicity ascribed to the different boron clusters, i.e., the smallest ones and least chaotropic were less toxic, while the clusters with bigger and more polarizable substituents showed higher toxicity (Figure 27, diamonds). Figure 27 | Quantification of cluster-assisted phalloidin–TRITC transport into living HeLa cells. Flow cytometry data of HeLa cells incubated with 2.5 µM phalloidin– TRITC in the absence or presence of different concentrations of globular decaand dodecaborate clusters (25, 50 and 75 µM) in HKR buffer for 3 h, subsequently washed with PBS, trypsinized, and analysed by flow cytometry. Bars represent the median fluorescence intensity (MFI) of phalloidin–TRITC in each condition. Black diamonds (scale on the right axis) represent cell concentration normalized on untreated cells. Error bars represent s.d. of three biological replicates, each one with three technical replicates. 192 Wehland, J.; Osborn, M.; Weber, K. Proc. Natl. Acad. Sci. U. S. A. 1977, 74 (12), 5613– 5617. 193 An, M.; Wijesinghe, D.; Andreev, O. A.; Reshetnyak, Y. K.; Engelman, D. M. Proc. Natl. Acad. Sci. U. S. A. 2010, 107 (47), 20246–20250. 124 95:5→5:95 (0→12 min)]. MS (ESI, H2O): 1525.1 (11, [M+H]+), 762.2 (100, [M+2H]2+), 508.6 (94, [M+3H]3+), 381.6 (14, [M+4H]4+). See Figure 29. Figure 29| (KLAKLAK)2 characterization. RP-HPLC [Agilent SB-C18 column, H2O (0.1 % TFA) / CH3CN (0.1 % TFA) 95:5→5:95 (0→12 min)] (Rt 6.0 min) and ESI-MS of peptide (KLAKLAK)2. 4.3. Confocal imaging For confocal microscopy studies, HeLa cells were seeded the day before on a 96-well glass bottom plate (Cellvis) at a density of 10,000 cells/well. Cells were washed with HKR buffer (5 mM HEPES, 137 mM NaCl, 2.68 mM KCl, 2.05 mM MgCl2, 1.8 mM CaCl2, pH 7.4). Phalloidin–TRITC and the boron clusters were diluted in HKR buffer (5 µM phalloidin–TRITC; 50, 100 or 150 µM boron clusters), then added sequentially to the cells (first the phalloidin, then the clusters; final concentrations: 2.5 µM phalloidin– TRITC; 25, 50 or 75 µM boron clusters), and incubated for 3 h at 37 ºC, 5 % CO2. Subsequently, nuclei were stained with 1 µM Hoechst 33342 for 20 min, washed with HKR, and immediately imaged in DMEM w/o phenol red using Fusion software (Andor) with a Dragonfly spinning disk confocal microscope mounted on a Nikon Eclipse Ti-E and equipped with an Andor Sona 4BV6U sCMOS digital camera. Images were processed with FIJI v. 2.1/1.53e.184 125 4.4. Flow cytometry HeLa cells were seeded the day before on a 96-well plate (Costar) at a density of 10,000 cells/well. Phalloidin–TRITC and the boron clusters were diluted in HKR buffer (5 µM phalloidin–TRITC; 50, 100 or 150 µM boron clusters), then added sequentially to the cells previously washed with HKR buffer (first the phalloidin, then the clusters; final concentrations: 2.5 µM phalloidin–TRITC; 25, 50 or 75 µM boron clusters), and incubated for 3 h at 37 ºC, 5 % CO2. Cells were subsequently washed with PBS, trypsinized, and trypsin was neutralized with PBS containing 2 % FBS and 5 mM EDTA. TRITC fluorescence was excited with a green laser (532 nm) and measured on a Guava easyCyte BG HT collecting the emission at 620/52 nm (OrangeG channel) and data were analyzed using InCyte v. 3.2. (GuavaSoft, Millipore). Cells with typical FSC and SSC parameters were selected, and the median fluorescence intensity calculated for each sample (MFI). Each condition was measured in triplicate. 4.5. Cell viability assays Cell viability was determined via MTT assay, a colorimetric assay for assessing cell metabolic activity through the reduction of tetrazolium dye MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] to its insoluble formazan, which has a purple color, by the NAD(P)H-dependent cellular oxidoreductase enzymes.199,200 For MTT assays in the presence of the boron clusters, HeLa cells were seeded the day before in 96-well plates at 10,000 cells/well. Cells were incubated with the clusters diluted in DMEM (range 0–2000 µM) for 3 or 24 h. For the viability assays in the presence of the boron clusters and the toxic peptide (KLAKLAK)2, HeLa cells were seeded the day before in 96-well plates at 10,000 cells per well. The toxic peptide and the boron clusters were diluted in HKR buffer (100 µM peptide; 50, 100, 150 or 200 µM boron clusters), then added sequentially to the cells (first the peptide, then the clusters; final concentrations: 50 µM peptide; 25, 50, 75 or 100 µM boron clusters), and incubated for 3 h at 37 ºC, 5 % CO2. Thereafter, the solutions were carefully removed, and cells were incubated for 24 h with complete medium. 199 Mosmann, T. J. Immunol. Methods. 1983, 65, 55–63. 200 Stockert, J. C.; Horobin, R. W.; Colombo, L. L.; Blázquez-Castro, A. Acta Histochem. 2018, 120, 159–167. 126 For all types of assays, at the end of the incubation, cells were incubated with complete DMEM containing 0.5 mg/mL MTT for 2 h, then the medium was carefully removed, and formazan crystals dissolved by addition of DMSO. The absorbance at 570 nm was measured with a plate reader (Tecan Infinite F200Pro). To eliminate the background, control wells corresponding to 0 % viability were treated with 1 % Triton X-100 just before MTT addition, and the mean absorbance value of these wells was subtracted from each sample absorbance value. Then, data were normalized to the mean value of untreated cells (100 % viability). Each condition was measured in 5 replicates. 127 CHAPTER III MECHANISM OF INTERNALIZATION AND INTRACELLULAR LOCALIZATION OF DODECABORATE ANIONS: A BIOLOGICAL INSIGHT. 128 1. OBJECTIVES After having examined the chaotropic properties of globular boron clusters to modulate cargo-membrane interaction and, thus, intracellular delivery of hydrophilic molecules, the last chapter of this Thesis will be focused on investigations of the internalization mechanism and the biological interactions of boron clusters with cell membranes. Particularly, the objective of this chapter is to investigate the following features of chaotropic boron clusters: • The uptake threshold and the extent of internalization in living cells. • The potential intracellular localization. • The mechanism of internalization in cells, and specifically: - The internalization in the presence of endocytosis inhibitors. - Whether chaotropic boron clusters cause endosomal disruption. - The importance of membrane potential on the intracellular transport of chaotropic boron clusters. For this study, we have selected the superchaotropic cluster with the best carrier activity and the inactive cluster from the dodecaborate series, which were employed in their native form, B12Br122− and B12H122−, and in a fluorescent version to allow visualization and quantification of the clusters inside living cells by different microscopy techniques. The hydrogenated cluster of low chaotropicity, B12H122−, was labelled with a 7−nitrobenzofurazan (NBD), which has an excitation peak at 467 nm, and an emission peak 539 nm (green).201 The brominated boron cluster, B12Br122−, with an optimal chaotropicity for membrane transport, was labelled with a 5(6)−carboxyfluorescein (CF), which has an excitation peak at 493 nm, and an emission peak 517 nm (green).202 The chemical structures of the fluorescently labelled clusters are represented in Figure 30. 201 https://www.aatbio.com/fluorescence-excitation-emission-spectrum-graphviewer/nbd_nitrobenzofurazan. 202 https://www.aatbio.com/fluorescence-excitation-emission-spectrum-graphviewer/fam_carboxyfluorescein. 129 Figure 30 | Fluorescently labelled boron cluster structures. Chemical structures of B12H11S−NBD− (left) and B12Br11O−CF3− (right); grey spheres represent B−H, orange spheres are B−Br, and pink spheres are B. 130 2. RESULTS & DISCUSSION 2.1. Internalization in living cells and intracellular localization To study the internalization of dodecaborate clusters in living cells, HeLa cells were first incubated with the fluorescent clusters, B12H11S−NBD− and B12Br11O−CF3−, at concentrations ranging from 1 to 100 µM for one hour; as a control, cells were also incubated with the corresponding unbound fluorophores, NBD and CF (Figure 31). Cells were then washed with HKR buffer to remove the membrane weakly associated clusters, and directly observed by microscopy. The corresponding confocal images showed that B12H11S−NBD− was detectable inside cells above a threshold concentration of 5 µM, and heterogeneously distributed throughout the cytosol, especially in membranous locations such as the endoplasmic reticulum (Figure 31a). No fluorescent signal was detected in the nuclei of the cells. We should consider that the solvatochromism of the NBD chromophore could increase the quantum yield in hydrophobic environments such as lipid membranes.203 Therefore, in these experimental conditions, cluster presence in the cell nucleus could not be completely ruled out as the absence of nuclear fluorescence could be related to the absence of membranes in the nucleoplasm. The control probe (NBD) was not detectable at any tested concentration, indicating that the unbound probe could not enter living cells (Figure 31b). In contrast to the hydrogenated cluster, the brominated labelled cluster, B12Br11O−CF3−, was already detectable inside cells at 1 µM, and clearly visible above 10 µM. At the lowest concentration (1 µM), the cluster seemed to accumulate in endosomes, but also in the nuclei of the cells (Figure 31c). At higher concentrations (e.g., 10 µM), the brominated cluster, besides its evident presence in various parts of the cell, such as the endoplasmic reticulum and endosomal compartments, was strongly localized in the nuclei, with the exclusion of nucleoli, which are rich in negatively charged nucleic acids204 that may repulse the superchaotropic anions (Figure 31c). The control fluorophore (CF) was not detected inside cells, and only at 50 µM showed a weak signal with heterogenous distribution. At the very high concentration of 100 µM, we could observe certain accumulation of this dye in the mitochondria (Figure 31d). Given these observations, the preferential nuclear localization of B12Br11O−CF3− could be imputed to the 203 Mazères, S.; Schram, V.; Tocanne, J. F.; Lopez, A. Biophys. J. 1996, 71 (1), 327–335. 204 Martin, R. M.; Ter-Avetisyan, G.; Herce, H. D.; Ludwig, A. K.; Lättig-Tünnemann, G.; Cardoso, M. C. Nucleus 2015, 6 (4), 314–325. 131 boron cluster itself, and not to the fluorescent probe to which it is associated. Figure 31 | Fluorescently labelled boron clusters and control fluorophores internalization in living cells. HeLa cells were incubated for 1 h with B12H11S−NBD− (a), NBD (b), B12Br11O−CF3− (c), and CF (d) diluted from 1 to 100 µM in HKR buffer, subsequently washed and imaged by confocal fluorescence microscopy; bright-field images in insets. Scale bars are 25 µm. To get further insights into the intracellular distribution of the nonfluorescently labelled clusters in living cells, confocal Raman microscopy was employed. First, the Raman spectra of B12H122− and B12Br122−, in the form of solid sodium salt (Na2B12H12 and Na2B12Br12), were determined individually (Figure 32). The Raman spectra of the clusters showed the intense bands of the symmetric in-phase breathing involving the stretching vibrations along the boron-boron (  B-B) and the boron-substituent (  B-X) bonds at around 751 cm−1 and 201 cm−1, respectively (see red circles in Figure 32),205,206,207 that were selected for further studies. 205 Muetterties, E. L.; Merrifield, R. E.; Miller, H. C.; Knoth, W. H.; Downing, J. R. J. Am. Chem. Soc. 1962, 84 (13), 2506–2508. 206 Leites, L. A. Chem. Rev. 1992, 92 (2), 279–323. 1 10 0 100 132 Figure 32 | Raman spectra of hydrogenated and brominated dodecaborate clusters. Raman spectra of the two clusters used in internalization studies in the form of solid sodium salt: Na2B12H12 (left) and Na2B12Br12 (right). Spectra were obtained with a Raman confocal microscope with the laser operating at 532 nm. Red circles indicate the peaks selected for the identification of the dodecaborate anions when incubated with cells, namely 750.63 cm−1 for B12H122−, and 200.95 cm−1 for B12Br122− (see Figure 33). The Raman spectra of HeLa cells alone is very complex due to the sum of the spectra of the different cellular components (see Figure 33b and Figure 34b). However, the assignment of the predominant bands of the main cellular components for they discrimination has been previously described.208,209 For instance, the region around 780 cm−1, is associated to the pyrimidine ring of the nucleic acids, the region around 1,000 cm−1 corresponds to the phenylalanine of proteins, the region around 2,850 cm−1 is mostly associated to the C−H bonds of the lipid’s alkyl chains, and the broad peak at 2,950 cm−1 corresponds to organic matter.208,209 HeLa cells were incubated with the boron clusters diluted in HKR and studied by Raman confocal microscopy. It should be noted that the concentration of the clusters had to be increased to the mM range to discriminate the signals from the background of the cells.210 In these conditions, incubation of HeLa cells with B12H122− indicated that the most characteristic Raman peak of this cluster at 754 cm−1 (  B-B +  B-H) could be perfectly resolved from the background of the Raman signal of the cells (Figure 33c, black arrow). Therefore, to visualize the intracellular distribution of this cluster, a 2 cm−1 wide filter, located at 754 cm−1, was applied to the obtained confocal Raman image.207 The filtered confocal Raman micrographs indicated a random distribution of the hydrogenated clusters inside HeLa cells (Figure 33a, third micrograph). To investigate the potential colocalization with the 207 Leites, L. A.; Bukalov, S. S.; Kurbakova, A. P.; Kaganski, M. M.; Gaft, Y. L.; Kuznetsov, N. T.; Zakharova, I. A. Spectrochim. Acta Part A Mol. Spectrosc. 1982, 38 (10), 1047–1056. 208 Toporski, J.; Dieing, T.; Hollricher, O. (Eds.) Confocal Raman Microscopy. 2nd edition; Springer: New York, 2018. 209 Kann, B.; Offerhaus, H. L.; Windbergs, M.; Otto, C. Adv. Drug Deliv. Rev. 2015, 89, 71– 90. 210 Tarrés, M.; Canetta, E.; Paul, E.; Forbes, J.; Azzouni, K.; Viñas, C.; Teixidor, F.; Harwood, A. J. Sci. Rep. 2015, 5, 7804. 133 endocytic compartments, a second filter was placed at the particular wavenumber of the Raman band typically assigned to cellular phospholipids (2,856 cm−1, Figure 33a, fourth micrograph).208,209 As no particular colocalization emerged from the merged pictures, the heterogeneous intracellular distribution of B12H11S−NBD− was confirmed (Figure 33a, fifth micrograph). These results were consistent with the fluorescent data obtained for the NBD labelled hydrogenated cluster of low chaotropicity that is not strongly internalized into cells and appears disseminated throughout the whole cell. Figure 33 | Raman confocal study of B12H122− internalization in living cells. HeLa cells were incubated for 1 h with 5 mM B12H122− in HKR buffer, then washed and imaged by confocal Raman microscopy (laser operating at 532 nm). Pictures in panel a show the bright-field image (first micrograph), the Raman chemical images at 2,935 cm−1 (width 250 cm−1), 754 cm−1 (width 2 cm−1), and 2,856 cm−1 (width 5 cm−1), respectively second, third and fourth micrographs from left to right, representing all cell components (proteins + nucleic acids + lipids), B12H122−, and phospholipids, and the fifth micrograph represents a merged picture of the Raman images of B12H122− (green) and phospholipids (magenta). Scale bars are 9 µm. Panels b and c represent Raman spectra of HeLa cells alone (b), and HeLa cells incubated with B12H122− (c), the black arrow indicates the peak corresponding to the presence of the cluster, i.e., 754 cm−1. Note that Figure 33b and Figure 34b show the same spectrum to ease the comparison with the spectra representing cells incubated with the boron clusters. The same filtering was applied to the B12Br122− 201 cm−1 predominant band (peak indicated by the black arrow in the spectrum in Figure 34b; filtered Raman image in Figure 34a, third micrograph),205 and for phospholipids (Figure 34a, fourth micrograph). In contrast to the hydrogenated cluster, and in agreement with the previously observed nuclear localization of fluorescent B12Br11O−CF3−, the merged picture of the filtered images corresponding to cluster and lipids showed a preferably nuclear distribution of B12Br122− (Figure 34a, fifth micrograph). 140 The incubation of HeLa cells expressing EGFP-hGal3 with the endosomolytic control, i.e. AcR8, produced the expected punctate motif, pointed out by white arrows in the right panel of Figure 38, while cells incubated with B12Br122− showed diffused cytosolic green fluorescence just like the untreated control, indicating that no endosomal disruption had occurred (Figure 38, left panel), and confirming that the results obtained in the LysoTracker experiment were most probably due to the quenching of the probe. Figure 38 | Endosomal disruption in p-EGFP-hGal3 transfected HeLa cells incubated with B12Br122−. HeLa cells expressing EGFP-hGal3 (green) were incubated with different concentrations of B12Br122− (left panel) or AcR8 as an endosomolytic control (right panel) as indicated in Materials & Methods section and subsequently imaged by confocal fluorescence microscopy. White arrows indicate the puncta corresponding to broken endosomes. Scale bars are 50 µm. 2.2.3. Internalization with depolarized membrane The different concentration of ions and the different permeability of the membrane for these ionic species generates a charge difference between both sides of the membrane, or membrane potential.228,229 The most abundant ions that control the membrane potential are Na+, K+ and Cl−, with a higher concentration outside the cell for the Na+ and Cl− ions, and a 228 Bonzanni, M.; Payne, S. L.; Adelfio, M.; Kaplan, D. L.; Levin, M.; Oudin, M. J. Biol. Open 2020, 9 (1), bio048553. 229 Alberts, B.; Johnson, A.; Lewis, J. et al. Molecular Biology of the Cell. 4th edition. Garland Science: New York, 2002. Ion Channels and the Electrical Properties of Membranes. 10 2 10 12 122 2 12 122 0 12 122 12 122 100 12 122 141 higher concentration inside the for K+. This gradient for the Na+ and K+ is maintained through the active pumping of 3 Na+ ions to the outside and 2 K+ ions to the inside by the Na+/K+-ATPase in the plasma membrane, leading to an intracellular concentration of K+ of ~140 mM and an extracellular concentration of ~5 mM. The potassium channels in the plasma membrane will allow the diffusion of the K+ ions outside the cell, down their concentration gradient, until the loss of positive charges, summed to the presence inside the cell of impermeable anionic molecules, creates an electrical field strong enough to counteract the efflux of K+. This electrical field is the equilibrium potential of this ion. The resting membrane potential of a cell will be a combination of the equilibrium potential of the ions involved weighted by their corresponding permeabilities. Typically, the cellular transmembrane potential is negative, and in resting Hela cell corresponds to −51 ± 2.4 mV,230 even though it may vary along the cell cycle.228 As in resting cells the permeability of the plasma membrane to K+ ions is much higher than for the other two, the membrane potential will be mostly driven by the equilibrium potential of K+.228,229,231 It has been previously observed and described, in this chapter, that membrane depolarization enhances the internalization of the anionic boron clusters in living cells (Figure 36). To determine whether the enhancement of cluster uptake required a minimum voltage or varied continuously with the membrane potential, the uptake experiment of the fluorescently labelled boron clusters was performed using a series of isotonic buffers based on PBS whose potassium ion concentration varied between 0 mM (maintaining the negative resting membrane potential) and 140 mM (K+PBS), concentration equivalent to that found intracellularly, which should bring membrane potential close to zero.168 Cluster internalization was quantified by flow cytometry and the results confirmed that, as opposed to what occurs with cationic cell penetrating peptides,168 boron cluster uptake increased with the increase in the external concentration of K+, both for B12H11S−NBD− and B12Br11O−CF3−, supporting what has been previously observed and demonstrating that cluster uptake (in terms of amount of internalized cluster per cell) increases constantly and 230 Stein, M. A.; Mathers, D. A.; Yan, H.; Baimbridge, K. G.; Finlay, B. B. Infect. Immun. 1996, 64 (11), 4820–4825. 231 Aidley, D.; Stanfield, P. Ion Channels. Cambridge University Press, 1996. 142 in direct proportion to the external concentration of potassium ions (Figure 39a).232 Surprisingly, it was also observed that the brominated labelled cluster entered the totality (100 %) of the cells regardless of the status of membrane potential, while the hydrogenated cluster percentage of internalization increased as this voltage approached to zero (Figure 39b). This suggested that the elevated chaotropicity of the brominated cluster endorsed the relocalization and homogeneous distribution of the dodecaborate anion among the whole cell population. The most relevant and surprising aspect of this result is that it suggests that chaotropicity helps surpass the classic dogma of the uptake dependence on the carrier-to-lipid ratio (Figure 39b). Figure 39 | Quantification of cluster internalization through normal to depolarized plasma membrane in living cells. HeLa cells were incubated with 5 µM B12H11S−NBD− or 2 µM B12Br11O−CF3−,233 with isotonic buffers containing decreasing concentrations of Na+ and increasing concentrations of K+ to alter membrane potential as indicated in Materials & Methods section, and subsequently analyzed by flow cytometry. Bars in a represent the mean of median fluorescence intensity (MFI) of B12H11S−NBD− or B12Br11O−CF3− normalized on the mean MFI value of cells incubated with B12H11S−NBD− or B12Br11O−CF3− in PBS without potassium (0 mM K+); each value was normalized on the corresponding control, indicated as 1. Bars in b represent the percentage of green fluorescence-positive cells. In both graphs, error bars represent s.d. of three technical replicates. 232 The uptake increase was not as remarkable as the one observed in Figure 36 as, in this case, cluster incubation time was considerably reduced because of protocol requirements168 (5 min in Figure 39 versus 60 min in Figure 36). Furthermore, the concentrations used were different (5 µM B12H11S−NBD− and 2 µM B12Br11O−CF3− in Figure 39 versus 10 µM both clusters in Figure 36). 233 Concentrations for the two clusters are not identical because the higher chaotropicity of B12Br11O−CF3−, compared to B12H11S−NBD−, made it difficult to appreciate the differences of uptake in the tested experimental conditions when it was used at 5 µM (data not shown), so the concentration of the brominated cluster had to be decreased to 2 µM. 0.0 0. 1.0 1. 2.0 2. 0 3 0 10 140 0 2 0 100 12 0 3 0 10 140 12 11 12 11 3 143 2.2.4. Cargo transport with depolarized membrane To provide further insight into, and potentially exploit, the phenomenon of cluster uptake enhancement by cells with depolarized membrane, we explored cargo intracellular delivery in this particular condition. Phalloidin−TRITC internalization was investigated by incubating cells with the active cluster B12Br122− and the cargo for 2 hours in PBS (resting negative membrane potential) or K+PBS (depolarized membrane). Flow cytometry and fluorescence confocal microscopy experiments proved that neutral hydrophilic cargo transport with the anionic boron cluster benefited from membrane depolarization, given that a phalloidin−TRITC internalization increase was observed in the presence of high external K+ concentration, both using 10 µM or 50 µM B12Br122− (Figure 40a, b). These experiments confirm the potential of the counterintuitive concept of using membrane depolarization to enhance the intracelular delivery with synthetic carriers. Figure 40 | B12Br122−-assisted phalloidin−TRITC transport in living cells with normal and depolarized membrane. HeLa cells were incubated with 2.5 µM phalloidin−TRITC (red, in micrographs) in the presence of different concentrations of B12Br122− in in PBS (light blue bars in a and box in b) or K+PBS (blue bars in a and box in b) as indicated in Materials & Methods section, subsequently stained with Hoechst (blue, only for confocal microscopy), and analyzed by flow cytometry (a) or imaged by confocal fluorescence microscopy (b); bright-field images in insets. Bars in a represents the median fluorescence intensity of phalloidin−TRITC taken up by cells; error bars represent s.d. of three technical replicates. Scale bars in b are 50 µm. 144 3. CONCLUSION In this chapter we have investigated new biological aspects of the interaction between the boron clusters and the cellular environment. Internalization experiments in living cells showed that, even though negatively charged clusters should be repelled by anionic cell membranes, very small concentrations are required for cluster uptake (in the range of 15 µM), and that, while hydrogenated clusters have a heterogeneous intracellular distribution that seemed to exclude the nucleus, brominated clusters preferably accumulate in the cytosol and the nuclei of cells, with the exclusion of the nucleoli, where the abundance of negatively charged nucleic acids may repulse the superchaotropic anions. Regarding the mechanism of internalization, vesicle experiments suggested that superchaotropic boron clusters can cross lipid membranes by direct translocation (see Chapter I). On the other hand, the results obtained in cells, that showed the presence of fluorescent endosomes by confocal microscopy and some reduction of the uptake in the presence of inhibitors of endocytosis, also suggested the involvement of endocytic pathways, such as macropinocytosis and clathrin-mediated endocytosis. Importantly, the fluorescently labelled brominated cluster (B12Br11O−CF3−), was much less affected by endocytic inhibitors than the hydrogenated one. The fact that the internalization of the most chaotropic cluster was less hindered by the presence of endocytic inhibitors, compared to the least chaotropic one, suggests that the more chaotropic the cluster is, the more it involves dynamic membrane disordering in its translocation process and can penetrate the cell regardless of the inhibition of endocytic pathways. All together, these results indicate that the chaotropicity assists boron clusters, and their attached fluorescent probes, with their transport and diffusion across cell membranes thanks to their membrane dynamic associating capabilities. Although a certain amount of the clusters was detected in endosomes by confocal microscopy, the clusters reach the cytosol and the nuclei of cells without altering the endosomal integrity. These results suggest that, when active as membrane transporters, boron clusters escape the endosomes without disrupting them. This raises the possibility that the cargo translocation and cytosolic release mediated by the superchaotropic carriers may also occur without endosomal disruption, but this will need to be confirmed in future experiments. 145 Furthermore, globular boron clusters are largely taken up by cells in the absence of membrane potential and can efficiently transport neutral hydrophilic cargos inside the cell in this condition. This behavior contrasts with the classical reported cationic transmembrane carriers,168 and opens new directions for membrane potential controlled delivery and the future design of superchaotropic selective carriers towards cellular type targets of lower membrane potential as, for instance, highly proliferative cancer cells.234 234 Yang, M.; Brackenbury, W. J. Front. Physiol. 2013, 4, 185. 146 4. MATERIALS & METHODS 4.1. Chemicals, peptides, and cell lines Boron clusters (as sodium salts) were from Katchem (Czech Republic). Fluorescently labelled boron clusters were obtained from W. M. Nau’s group (the synthesis of B12Br11O−CF3− is described in Zhang et al., 2022).235 AcR8 and Pep-1 were synthesized as described in ref. 34 (for Pep-1 see also ref. 35). The plasmid pEGFP-hGal3 was a gift from Tamotsu Yoshimori (Addgene plasmid #73080; http://n2t.net/addgene:73080; RRID: Addgene_73080).224 LysoTracker™ Green DND-26, and Lipofectamine 2000 were from ThermoFisher Scientific. 5(6)-Carboxyfluorescein (CF), 5-(Nethyl-N-isopropyl)-amiloride (EIPA), and sodium azide (NaN3) were from Sigma-Aldrich. Wortmannin (Wmn) was from Fluorochem. Dynasore (Dyn) was from Merck Millipore. Chlorpromazine (Cpz) hydrochloride was from Cymit Química. Phalloidin−TRITC was from Hello Bio. PBS without KCl was prepared using 140 mM NaCl, 30 mM Na2HPO4, 1.76 mM KH2PO4, 1 mM CaCl2, 0.5 mM MgCl2. K+PBS was prepared using 140 mM KCl, 30 mM Na2HPO4, 1.76 mM KH2PO4, 1 mM CaCl2, 0.5 mM MgCl2. HeLa cells were obtained from ATCC and maintained in DMEM supplemented with 10 % FBS and 1 % Penicillin-Streptomycin-Glutamine mix, at 37 ºC, 5 % CO2, and 95 % humidity. 4.2. General protocol for confocal fluorescence imaging For confocal fluorescence microscopy studies, cells were treated as described below, and imaged in DMEM w/o phenol red using Fusion software (Andor) with a Dragonfly spinning disk confocal microscope mounted on a Nikon Eclipse Ti-E equipped with an Andor Sona 4BV6U digital camera. Images were processed with FIJI v. 2.1/1.53e.184 4.2.1. Internalization of fluorescent clusters HeLa cells were seeded the day before on a 96-well glass bottom plate (Cellvis) at a density of 10,000 cells/well. Cells were washed with HKR buffer (5 mM HEPES, 137 mM NaCl, 2.68 mM KCl, 2.05 mM MgCl2, 1.8 mM CaCl2, pH 7.4). Fluorescent clusters (B12H11S−NBD− or B12Br11O−CF3−) or fluorophores alone as a control (NBD and CF) were diluted in HKR buffer (1, 5, 10, 50, and 100 µM), then added to the cells, 235 Zhang, J.; Gabel, D.; Assaf, K. I.; Nau, W. M. Org. Lett. 2022, 24, 9184−9188. 147 and incubated for 1 h at 37 ºC, 5 % CO2. Subsequently, cells were washed with 0.1 mg/mL heparin in HKR for 5 min and then with HKR buffer, and finally imaged in DMEM w/o phenol red. 4.2.2. LysoTracker™ Green staining HeLa cells were seeded the day before on a 96-well glass bottom plate (Cellvis) at a density of 10,000 cells/well. Cells were washed with HKR buffer, incubated for 1 h at 37 ºC, 5 % CO2 with B12Br122− diluted in HKR buffer (0, 10, 25, 50 µM), and then stained with LysoTracker™ Green DND26 according to manufacturer’s protocol (probe diluted 1:20,000 in DMEM to a final concentration of 50 nM, incubated onto cells for 30 min at 37 ºC, 5 % CO2). Subsequently, cells were washed with HKR buffer, and finally imaged in DMEM w/o phenol red. 4.2.3. EGFP-hGal3 endosomal disruption assay in HeLa cells HeLa cells were seeded on an 8-well glass bottom plate (Ibidi) at a density of 30,000 cells/well. The next day, cells were transfected with a plasmid carrying the gene for EGFP-hGal3 using Lipofectamine 2000 according to manufacturer's protocol and replacing the transfection mixture with fresh medium after 4 h. The day after transfection, cells were washed with HKR buffer, incubated with B12Br122− (0, 10, 25, 50, 75, 100 µM) or AcR8 (10, 25 µM) diluted in HKR buffer, for 3 h at 37 ºC, 5 % CO2, then washed again with HKR buffer, and imaged in DMEM w/o phenol red. 4.2.4. Phalloidin−TRITC transport with normal and depolarized membrane HeLa cells were seeded the day before on a 96-well glass bottom plate (Cellvis) at a density of 10,000 cells/well. Before cargo and carrier incubation, cells were washed three times with PBS without KCl or K+PBS. Phalloidin−TRITC and the boron clusters were diluted in the two buffers (PBS without KCl or K+PBS; 5 µM phalloidin−TRITC; 20 or 100 µM B12Br122−), then added sequentially to the cells (first the phalloidin, then the cluster; final concentrations: 2.5 µM phalloidin−TRITC; 10 or 50 µM boron clusters), and incubated for 2 h at 37 ºC, 5 % CO2. Subsequently, nuclei were stained for 20 min with 1 µM Hoechst 33342 diluted in the corresponding buffer, and immediately imaged in DMEM w/o phenol red. 148 4.3. Raman confocal imaging For Raman confocal microscopy, cells were seeded on round coverslips (12 mm diameter) in 24-well plate at a density of 50,000 cells/well the day prior the experiment. Then, cells were incubated with 5 mM B12H122− or B12Br122− diluted in HKR buffer for 1 h at 37 ºC, 5 % CO2, washed with HKR buffer, and imaged by confocal Raman microscopy. The cell sample coverslips were put on a glass microslide and immersed in PBS during the measurements, in order to keep the cells in good condition for longer than the measurement time of an image scan. To record the spectra of the bare clusters, they were placed in the form of solid sodium salt (Na2B12H12 or Na2B12Br12) on a glass dish. Raman spectra were recorded by rasterscanning the sample under a confocal Raman microscope Alpha300 R (WITec GmbH, Ulm, Germany). An area of 30 × 30 µm was divided in 100 lines, and 100 spectra per line were recorded using a 532 nm laser, with an integration time of 0.1 s at a laser power of 19 mW, and a Zeiss EC Epiplan Neofluar 50×/0.8NA water immersion microscope objective (10,000 total spectra with a 300 nm lateral resolution). The microscope was equipped with a 200 × 200 × 20 µm piezoelectric bracket and a 127 × 1024 CCD backilluminated detector, refrigerated through Peltier to −60 ºC, and using a 600 g/mm diffraction. Reflected and elastically scattered photons were rejected by an edge filter. Cosmic ray spikes were removed and images were processed with Project FIVE v. 5.1 (WITec GmbH, Ulm, Germany). 4.4. General protocol for flow cytometry A Guava EasyCyteTM cytometer (EMD Millipore) was used for all flow cytometry experiments. Data were analysed with InCyte software included in GuavaSoft 3.2 (Millipore). 4.4.1. Internalization with endocytosis inhibitors HeLa cells were seeded on a 96-well plate (Costar) at a density of 10,000 cells/well. The next day, cells were kept on ice (with DMEM) for 30 min, or treated with K+PBS, or with wortmannin (100 µM), dynasore (80 µM), EIPA (50 µM), chlorpromazine (30 µM), or sodium azide (1 %) diluted in DMEM without serum and incubated for 30 min at 37 ºC. After the treatment, these solutions were replaced by DMEM without serum (or K+PBS) containing 10 µM B12H11S−NBD− or B12Br11O−CF3−, and the same amount of the corresponding inhibitor. After incubation for 1 h, cells were washed with PBS, trypsinized, and trypsin was neutralized with PBS containing 2 % FBS and 5 mM EDTA. NBD and CF fluorescence was excited 149 with a blue laser (488 nm) and the emission collected at 512/18 nm (Green-B channel). Cells with typical FSC and SSC parameters were selected, and the median fluorescence intensity (MFI) calculated for each sample. Each condition was measured in triplicate. Data were normalized to untreated controls, incubated only with B12H11S−NBD− or B12Br11O−CF3−. 4.4.2. Endo-lysosomal disruption assay HeLa cells were seeded the day before on a 96-well plate (Costar) at a density of 10,000 cells/well. Cells were washed with HKR buffer. Cells were washed with HKR buffer, incubated for 1 h at 37 ºC, 5 % CO2 with B12Br122− (0, 10, 25, 50, 100 µM), AcR8 (50 µM), or Pep-1 (30 µM) diluted in HKR buffer and then stained with LysoTracker™ Green DND-26 according to manufacturer’s protocol (probe diluted 1:20,000 in DMEM, incubated onto cells for 30 min at 37 ºC, 5 % CO2). Subsequently, cells were washed with PBS, trypsinized, and trypsin was neutralized with PBS containing 2 % FBS and 5 mM EDTA. LysoTracker™ Green fluorescence was excited with a blue laser (488 nm) and the emission collected at 512/18 nm (Green-B channel). Cells with typical FSC and SSC parameters were selected, and the median fluorescence intensity (MFI) calculated for each sample. Each condition was measured in triplicate. Data were normalized to the untreated cells stained with LysoTrackerTM. 4.4.3. Internalization with normal to depolarized membrane Five different isotonic versions of PBS at pH 7.4 were prepared using 140, 105, 70, 35, or 0 mM NaCl and, respectively, 0, 35, 70, 105, or 140 mM KCl (the latter called K+PBS). All the buffers contained 30 mM Na2HPO4, 1.76 mM KH2PO4, 1 mM CaCl2, and 0.5 mM MgCl2. HeLa cells were seeded on a 96-well plate (Costar) at a density of 10,000 cells/well. The next day, cells were washed three times with the different isotonic buffers, and incubated with 5 µM B12H11S−NBD− or 2 µM B12Br11O−CF3− diluted in the same buffers for 5 min at 37 ºC, then trypsinized for 10 min with 10× trypsin-EDTA diluted to 1× in the same buffers, and trypsin was neutralized with the same buffers containing 2 % FBS and 5 mM EDTA, in order to maintain the same K+ concentration, and thus membrane polarization status, along the whole protocol. NBD and CF fluorescence were excited with a blue laser (488 nm) and the emission collected at 512/18 nm (Green-B channel). Cells with typical FSC and SSC parameters were selected, and the median fluorescence intensity (MFI) calculated for each sample. Each condition was measured in triplicate. Data were