Transition Metal-Catalyzed Annulations in Biological Media

INTERNATIONAL DOCTORAL SCHOOL OF THE USC Joan Miguel Ávila PhD Thesis Transition Metal-Catalyzed Annulations in Biological Media Santiago de Compostela, 2022 Doctoral Programme in Chemical Science and Technology TESE DE DOUTORAMENTO Transition Metal-Catalyzed Annulations in Biological Media Joan Miguel Ávila ESCOLA DE DOUTORAMENTO INTERNACIONAL PROGRAMA DE DOUTORAMENTO EN CIENCIA E TECNOLOXÍA QUÍMICA SANTIAGO DE COMPOSTELA 2022 D./Dña. Joan Miguel Ávila Título da tese: Transition Metal-Catalyzed Annulations in Biological Media 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, 22 de abril de 2022. Asdo. Joan Miguel Ávila AUTORIZACIÓN DEL DIRECTOR / TUTOR DE LA TESIS Transition Metal-Catalyzed Annulations in Biological Media D. José Luis Mascareñas Cid Dª. María Tomás Gamasa INFORMAN: Que la presente tesis, se corresponde con el trabajo realizado por D. Joan Miguel Ávila, 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 de Monográfica con reproducción de publicaciones, en los que la participación del doctorando fue decisiva para su elaboración y las publicaciones se ajustan al Plan de Investigación. En Santiago de Compostela, 22 de abril de 2022 Asdo. José Luis Mascareñas Cid Asdo. María Tomás Gamasa “A moderately good memory, one of the prerequisites of survival in chemistry” - Rolf Huisgen “María, ¿cómo se llamaba esa cosa que ya sabes taca y filtras…?” - Joan Miguel Dedicat al meu germà INTRODUCTION Chapter I 17 1. Chemical biology As the smallest unit of life, the cell has been widely studied.1 However, ultimately, the cell is not more than a soup of molecules that are perfectly organized, and therefore chemists, the experts in molecules, have been increasingly involved in interrogating and manipulating cellular and biological systems. This field of Chemical Biology encompasses a variety of topics that goes from the study of life from a chemical point of view, to the synthesis of biomolecules, the introduction of small molecules in biological systems or even the discovery of new chemical processes by taking inspiration from biology.2,3 For instance, the development of innocuous probes that are able to specifically label biomolecules or organelles,4 the synthesis of proteins by Native Chemical Ligation (NCL)5 or the use of optimized enzymes by direct evolution6,7 are some “recent” research topics within the field of chemical biology (figure 1). Figure 1. Some contributions of chemical biology to life science: a) Selected probes for organelle targeting and fluorescent staining: MitoTracker® as mitochondrial marker, Hoechst for nucleus staining and BODIPY® for lipids targeting. b) General scheme of a Native Chemical Ligation. c) Outline of an evolved enzyme to perform a specific reaction.8 1 R. Hook, Micrographia, Royal Society, 1665. 2 Nat. Chem. Biol. 2015, 11, 363. 3 K. L. Morrison, G. A. Weiss, Nat. Chem. Biol. 2006, 2, 3. 4 I. Johnson, M. T. Z. Spence, The Molecular Probes Handbook 11th Ed, Molecular Probes, 2010. 5 S. S. Kulkarni, J. Sayers, B. Premdjee, R. J. Payne, Nat. Rev. 2018, 2, 0122. 6 F. H. Arnold, Acc. Chem. Rev. 1998, 31, 125. 7 K. Chen, F. H. Arnold, Nat. Catal. 2020, 3, 203. 8 D. K. Romney, J. Murciano-Calles, J. E. Wehrmüller, F. H. Arnold, J. Am. Chem. Soc. 2017, 139, 10769. Chapter I 18 2. Bioorthogonal chemistry 2.1. What is bioorthogonal chemistry Despite the impressive progress of Chemical Biology over the last decades, there is still a lack of novel robust techniques and methodologies that contribute to the study of life in a harmless manner. In this sense, Bioorthogonal Chemistry was born as an approach to perform selected reactions in biological systems without perturbing the host metabolism (figure 2). Since its recent consolidation as an important branch of chemical biology, bioorthogonal chemistry has continued to grow in an exponential manner.9 In the meantime, it has also contributed to revolutionize related fields such as cell imaging, drug development or biotechnology, due to its ability to manipulate small exogenous compounds in biological and living settings. Bioorthogonal tools supplement other biological post-translation modification strategies for protein tagging like “genetically encoded fluorescent proteins”,10 antibody recognition11 or “avidine/streptavidinebiotin technology”,12,13 among others. Figure 2. Simplified representation of a bioorthogonal modification in a living system. Bioorthogonal reactions share some common features, such as: chemoselectivity, good performance under physiological conditions (that is, aqueous media, neutral pH, 37 ºC) and fast rates. The components of the transformations are generally exogenous to nature and consequently, should not be metabolized by the cell. They must be innocuous to the biological system and stable under physiological conditions. Fast kinetics is one of the most important attributes since it allows not only to reduce the exposition time of the abiotic components to the biological system, but also to decrease the concentration of reagents. Along the years there has been intensive research on the development of fast bioorthogonal reactions based on the activation of substrates with strain, or by electronic tuning.14 In contrast, the use of catalysts has been less explored. 9 T. Cañeque, S. Müller, R. Rodriguez, Nat. Rev. Chem. 2018, 2, 202. 10 G. U. Nienhaus, Angew. Chem. Int. Ed. 2008, 47, 8992. 11 C. Milstein, G. Winter, Nature 1991, 349, 293. 12 O. H. Laitinen, V. P. Hytönen, H. R. Nordlund, M. S. Kulomaa, Cell. Mol. Life. Sci. 2006, 63, 2992. 13 Q. Le, V. Nguyen, S. Park, Appl. Microbiol. Biotechnol. 2019, 103, 7355. 14 S. L. Scinto, D. A. Bilodeau, R. Hincapie, W. Lee, S. S. Nguyen, M. Xu, C. W. am Ende, M. G. Finn, K. Lang, Q. Lin, J. P. Pezacki, J. A. Prescher, M. S. Robillard, J. M. Fox, Nat. Rev. Methods Primers 2021, 1, 30. Chapter I 19 Together with high reaction kinetics, chemoselectivity is the most pursued feature, since the reaction must be compatible with an important number of functional groups and biomolecules present in living systems. Introducing other characteristics in the substrates, like soluble and lipophilic components to allow internalization, or targeting moieties to reach a specific location, are also important aspects to take into account when designing bioorthogonal tools.15 Sometimes, a better biological assimilation is preferred rather than a very high reactivity.16,17 The most common strategy used in the application of bioorthogonal chemistry involves two main steps (figure 3). First, a molecule of interest is modified to bear a bioorthogonal functional tag (I), and is assimilated by the biological system (II). Second, a paired reacting partner bearing a probe (III) selectively reacts with the tag within the biological habitat (IV).18,19 Although the focus of this research was initially set in bioconjugation approaches, bioorthogonal tools have been increasingly used for the controlled release of bioactive molecules,20 or even for the synthesis of designed products.21,22 Figure 3. Bioorthogonal bioconjugation standard strategy. I = Molecule of interest (blue hexagon) bearing the bioorthogonal handle (red square); II = Assimilation by the biological system; III = Inactivated probe partner (colorless star), bearing the biorthogonal counterpart tag; IV = Bioorthogonal transformation that leads to the bioconjugation of the molecule of interest. 15 Y. Tian, Q. Lin, ACS Chem. Biol. 2019, 14, 2489. 16 C. P. Ramil, M. Dong, P. An, T. M. Lewandowski, Z. Yu, L. J. Miller, Q. Lin, J. Am. Chem. Soc. 2017, 139, 13376. 17 J. Miguel-Ávila, M. Tomás-Gamasa, A. Olmos, P. J. Pérez, J. L. Mascareñas, Chem. Sci. 2018, 9, 1947. 18 E. M. Sletten, C. R. Bertozzi, Angew. Chem. Int. Ed. 2009, 48, 6974. 19 E. M. Sletten, C. R. Bertozzi, Acc. Chem. Res. 2011, 44, 666. 20 J. Wang, X. Wang, X. Fan, P. R. Chen, ACS Cent. Sci. 2021, 7, 929. 21 M. O. N. van de L’Isle, M. C. Ortega-Liebana, A. Unciti-Broceta, Curr. Opin. Chem. Biol. 2021, 61, 32. 22 D. P. Nguyen, H. T. H. Nguyen, L. H. Do, ACS Catal. 2021, 11, 5148. Chapter I 20 2.2. The birth of bioorthogonal chemistry: Hydrazone ligations The term Bioorthogonal Chemistry was coined for the first time in 2003 by Prof. Bertozzi.23 Despite its recent birth, it is difficult to ensure which is the first real bioorthogonal reaction due to its close relation to the selective modification of proteins.19 Nevertheless, a few reports from Rideout and coworkers between late 1980s and early 1990s can be considered among the first publications dealing with bioorthogonal principles.24,25,26 In these articles the authors demonstrated the viability of performing hydrazine-aldehyde condensations in the presence of bacteria or cells (figure 4a). The efforts were focused on the design of substrates than can be internalized by cells where the reaction could take place efficiently (figure 4b). The ligation was applied for the selective generation of cytotoxicity in specific cells. Figure 4. a) Reaction scheme by Rideout: hydrazine-aldehyde condensation. b) Representative description of a selective reaction inside cells by Rideout in 1990, before the definition of bioorthogonal chemistry. A) Incubation of normal and target cells, with both substrates X and Y. B) Selective internalization of X and Y in the target cell. C) Reaction inter and intracellular.26 This strategy was further improved by Bertozzi and coworkers in 1997 (figure 5). They succeeded in performing a similar type of condensation in Jurkat or HeLa cell surfaces, by expressing in excess, a mannose derivative (ManLev, 5 mM) bearing a ketone group in the cellular membrane. An exogenous hydrazine probe modified with biotin (1 mM) was used as coupling partner and allowed to react at room temperature for 2 h.27,28 Figure 5. Ketone-acylhydrazine condensation on mammalian cell surface, by Bertozzi and coworkers in 1997.27 23 H. C. Hang, C. Yu, D. K. Kato, C. R. Bertozzi, Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 14846. 24 D. Rideout, Science 1986, 233, 561. 25 D. Rideout, J. Jaworski, R. Dagnino, Biochem. Pharm. 1988, 37, 4505. 26 D. Rideout, T. Calogeropoulou, J. Jaworski, M. McCarthy, Biopolymers 1990, 29, 247. 27 L. K. Mahal, K. J. Yarema, C. R. Bertozzi, Science 1997, 276, 1125. 28 H. C. Hang, C. R. Bertozzi, J. Am. Chem. Soc. 2001, 123, 1242. Chapter I 21 Since these pioneering examples, it was not until 2003 that the first intracellular bioorthogonal version saw the light. The Schultz’s group described an intracellular condensation between Nacetyl-L-phenylalanine previously expressed in E. coli proteins and either a fluorescent or biotinylated hydrazine probe.29 However, the first mammalian intracellular version of this type of coupling was reported by the Kool’s group 10 years later. In this case, an exchange strategy that involves a hydrazone fluorescein derivative, coupled with a Föster Resonance Energy Transfer (FRET) fluorescence quencher (Q, figure 6), was designed. This inactive hydrazone (20 M) was exchanged along 1 h, with the target aldehyde (100 M) using 2,4-dimethoxyaniline as organic catalyst (10 mM, figure 6). The reaction promotes the release of the quencher as an aldehyde derivative, liberating a fluorescent hydrazone in the process.30 Figure 6. Hydrazone exchange in HeLa cells, developed by Kool and co-workers in 2016. PEG = poly(ethylene glycol) as flexible group for energy release.30 Despite the hydrazone condensation can be considered among the first bioorthogonal reactions, and a breakthrough in the field,31 it was quickly set apart mainly due to selectivity problems (competition with endogenous aldehydes, ketones) or side reactions like hydrolysis. Thus, the research moved towards other transformations like the Staudinger ligation or the azide-alkyne cycloaddition. 29 P. G. Schultz, Z. Zhang, B. A. C. Smith, L. Wang, A. Brock, C. Cho, Biochemistry 2003, 42, 6735. 30 E. T. Kool, L. H. Yuen, N. S. Saxena, H. S. Park, K. Weinberg, ACS Chem. Biol. 2016, 11, 2312. 31 E. T. Kool, D. K. Kölmel, Chem. Rev. 2017, 117, 10358. Chapter I 22 2.3. Staudinger ligation The Staudinger reduction was proposed by Meyer and Staudinger in 1919,32,33 and involves the conversion of organic azides into amines through the formation of iminophosphoranes (figure 7). Figure 7. Transformation proposed by Meyer and Staudinger in 1919.32,33 This reaction was brought to the field of chemical biology in 2000 by Bertozzi and coworkers,34 with the design of a ligation version to couple phosphines and azides (figure 8), which fits with the concept of a bioorthogonal transformation. These two functional groups, are totally abiotic and show an amazing orthogonality and biocompatibility.19 The ligation involves an orthophosphine terephthalic acid analogue. After the reaction with the corresponding azide, and nitrogen loss, the aza-ylide I is trapped by the ester group, undergoing a cyclization to give the oxazaphosphetane II. In the aqueous environment, the intermediate II is hydrolyzed, affording a p-carbamoylbenzoic acid analogue, with a phosphine oxide in its structure, as a final product. Therefore, in this case the ligation entails the formation of an amide bond.33 Figure 8. Mechanism of the Staudinger ligation designed by Bertozzi in 2000.34 These authors were able to metabolically incorporate azidosugars into the membrane of Jurkat or HeLa mammalian cells. The small size of the azide group is neglected by the biosynthetic machinery, allowing the assembly of glycoproteins bearing the azide function at specific points of the cell surface. The coupling partner (1 mM) containing the phosphine group was modified to exhibit a biotin for a posterior interaction with avidin, and allowed to react for one hour at room temperature in presence of incubated cells modified with azidosugar. 32 H. Staudinger, J. Meyer, Helv. Chim. Acta 1919, 2, 635. 33 C. Bednarek, I. Wehl, N. Jung, U. Schepers, S. Bräse, Chem. Rev. 2020, 120, 4301. 34 C. R. Bertozzi, E. Saxon, Science 2000, 287, 2007. Chapter I 23 The bioorthogonal coupling presented at least a two-fold improved efficiency, in comparation to the hydrazone ligation (figure 9).34 Furthermore, the group of Bertozzi expanded this methodology through the years, to carry out the reactions in parallel with the hydrazone condensation,35 or even in mice.36 Figure 9. Staudinger ligation performed on mammalian cell surface by Bertozzi in 2000.34 These studies demonstrated the high level of bioorthogonality of the azide functionality, which has been widely used in the field. Its inertness, small size and unique reactivity that arises from its 1,3 – dipole character makes the azide a very appealing functional group for abiotic ligations in biological settings.37 The azide moiety has been since then introduced into many other molecules that can be incorporated into biological polymers through metabolic pathways.38,39 Some of them are indicated in figure 10: a) acetylated N-azidoacetylmannosamine (Ac4ManNAz) which is metabolized into azido-sialic acid (siaNAz) and expressed on cell surface;34 b) azidohomoalanine (AHA) which is metabolized by cells as a methionine (Met) analogue;40 c) pazidophenyalanine (p-Az) that has been introduced by genetic code expansion;41 d) an azido 7deaza-2’-deoxyadenosine (Azido-dATP) analogue that has been used for the modification of DNA.42 2546 35 P. V. Chang, J. A. Prescher, M. J. Hangauer, C. R. Bertozzi, J. Am. Chem. Soc. 2007, 129, 8400. 36 J. A. Prescher, D. H. Dube, C. R. Bertozzi, Nature 2004, 430, 873. 37 M. F. Debets, C. W. J. Van der Doelen, F. P. J. T. Rutjes, F. L. Delft, ChemBioChem 2010, 11, 1168. 38 N. J. Pedowitz, M. R. Pratt, RSC Chem. Biol. 2021, 2, 306. 39 K. Lang, J. W. Chin, Chem. Rev. 2014, 114, 4764. 40 K. L. Kiick, E. Saxon, D. A. Tirrell, C. R. Bertozzi, Proc. Natl. Acad. Sci. USA 2002, 99, 19. 41 M. Tsao, F. Tian, P. G. Schultz, ChemBioChem 2005, 6, 2147. 42 S. H. Weisbrod, A. Marx, Chem. Commun. 2007, 1828 Chapter I 24 Figure 10. Some of the azido-biomolecules employed in the Staudinger ligation and further bioorthogonal transformations. As important as the internalized azide-tagged biomolecule, is the reaction partner. On this regard, the group of Bertozzi explored the use of smart probes, compounds that are fluorescent only after the reaction takes place.43 For this purpose, they developed a coumarin smart probe in which the fluorescence is quenched via an intramolecular charge transfer (ICT) pathway which involves the lone pair of electrons of a phosphine group. After the reaction with an AHA bearing protein (65 g/mL) and 200 M of coumarin, a 60-fold fluorescence increase was observed (figure 11).44 Since then, the use of smart probes for bioorthogonal labeling has become each day more popular until turning into a standard strategy in bioorthogonal chemistry. Figure 11. Staudinger ligation developed by Bertozzi using smart probes with intramolecular charge transfer quenching in 2003.44 Although the Staudinger reaction was the first bioorthogonal process that could be translated and widely used in cells and even in living systems, it presents important drawbacks. For instance, it is among the slowest bioorthogonal reactions.45,46 43 P. Shieh, C. R. Bertozzi, Org. Biomol. Chem. 2014, 12, 9307. 44 G. A. Lemieux, C. L. De Graffenried, C. R. Bertozzi, J. Am. Chem. Soc. 2003, 125, 4708. 45 A. S. Cohen, E. A. Dubikovskaya, J. S. Rush, C. R. Bertozzi, J. Am. Chem. Soc. 2010, 132, 8563. 46 B. L. Oliveira, Z. Guo, G. J. L. Bernardes, Chem. Soc. Rev. 2017, 46, 4895. Chapter I 31 2.6. Strain-promoted azide-alkyne cycloadditions The use of metals for promoting bioorthogonal cycloadditions is appealing, because they don’t require especially reactive substrates. However, in recent years there have been many advances in a different strategy based on the use of strained reactants, which avoids the employ of metals. Inspired by Huisgen,80 but especially by the work of Wittig and Krebs with strained cycloalkynes,81 the group of Bertozzi investigated the strain-promoted azide-alkyne cycloaddition (SPAAC), in parallel with the development of the Staudinger ligation. They needed activated alkynes to perform the reaction, and found that the cycloaddition of cyclootyne and azides take place at room temperature without the need of a catalyst (figure 19). Figure 19. a) General conditions for SPAAC labeling. b) Common strained cyclooctynes and their kinetics.19 Other cycloalkynes with improved water solubility, or better kinetics due to a higher ring strain or the presence of electron withdrawing groups (EWG) such as fluorine atoms82,83 and fused aryl rings, have been described.84,85 The first application of the SPAAC transformation in a biological context appeared in 2004. An azido-sugar was expressed on cell surface of CHO and Jurkat mammalian cells and the labeling with the cyclooctyne (OCT) was demonstrated.86 Some years later, in 2007, Bertozzi and coworkers implemented a new difluorobenzocyclooctyne (DIFO) coupled with an Alexa Fluor 488 probe (100 M), for the SPAAC reaction. At 37 ºC, they could observe, cell-binding in a matter of minutes (figure 20).82 80 R. Huisgen, Proc. Chem. Soc. 1961, 357. 81 G. Wittig, A. Krebs, Chem. Ber. 1961, 94, 3260. 82 J. M. Baskin, J. A. Prescher, S. T. Laughlin, N. J. Agard, P. V. Chang, I. A. Miller, A. Lo, J. A. Codelli, C. R. Bertozzi, Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 16793. 83 N. J. Agard, J. M. Baskin, J. A. Prescher, A. Lo, C. R. Bertozzi, ACS Chem. Bio. 2006, 1, 644. 84 X. Ning, J. Guo, M. A. Wolfert, G. Boons, Angew. Chem. Int. Ed. 2008, 47, 2253. 85 J. C. Jewett, E. M. Sletten, C. R. Bertozzi, J. Am. Chem. Soc. 2010, 132, 3688. 86 N. J. Agard, J. A. Prescher, C. R. Bertozzi, J. Am. Chem. Soc. 2004, 126, 15046. Chapter I 32 This new version was applied to the study of glycan recycle metabolism, where the labelled modified-sugars on the cell surface are internalized, and further detected in endosomes and in the Golgi apparatus. Some years later the transformation was achieved in zebrafish,87 nematodes88 and even mouses.89 Figure 20. SPAAC coupling developed by Bertozzi and co-workers on cell surface in 2007.89 The first intracellular example came from a collaboration between Tirrel and Bertozzi groups, in 2010 (figure 21). Therein, the AHA aminoacid containing an azide group was metabolically incorporated after treatment with 1 mM of AHA. Afterwards, cells were supplied with 50 M of DIFO-coumarin, noticing fluorescence after 10 min at 37 ºC.90 The acquisition of threedimensional microscopic images confirmed the intracellular reaction. Unfortunately, almost half of the fluorescence came from membrane proteins, and additionally, some intracellular fluorescence background caused by non-specific side reactions, owing to the side reactivity of the cyclooctyne. Remarkably, the transformation is still nowadays pretty used. For instance, although preliminary, recently Xi and co-workers were able to carry out the cycloaddition inside mitochondria of mammalian cells.91 Figure 21. Intracellular version of SPAAC developed by Tirrel and Bertozzi in 2010.90 87 S. T. Laughlin, J. M. Baskin, S. L. Amacher, C. R. Bertozzi, Science, 2008, 320, 664. 88 S. T. Laughlin, C. R. Bertozzi, ACS Chem. Biol. 2009, 4, 1068. 89 P. V. Chang, J. A. Prescher, E. M. Sletten, J. M. Baskin, I. A. Miller, N. J. Agard, A. Lo, C. R. Bertozzi, Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 1821. 90 K. E. Beatty, J. D. Fisk, B. P. Smart, Y. Y. Lu, J. Szychowski, M. J. Hangauer, J. M. Baskin, C. R. Bertozzi, D. A. Tirrel, ChemBioChem 2010, 11, 2092. 91 L. Yi, X. Yang, X. Cai, D. Wang, Y. Gao, Z. Xi, RSC Adv. 2019, 9, 23. Chapter I 33 The strained azide-alkyne cycloaddition is not exempt of problems. The kinetics are slower than those of CuAAC (10-100 times),92 and usually suffers from non-specific reactions with nucleophiles of the biological media.90 Moreover, the cyclooctyne group is not easy to prepare and can be considered as a “bulky” moiety in comparation with the standard triazole obtained in a CuAAC. 2.7. Bioorthogonal ligations based on inverse electron demand diels-alder cycloadditions Given these problems, there has been an intense effort to develop new, more efficient alternatives to the strain-driven cycloadditions. In this context, the most popular bioorthogonal reaction nowadays is perhaps the inverse electron-demand diels-alder (IEDDA) or tetrazine ligation between a 1,2,4,5-tetrazine and an alkyne/alkene reaction. This [4+2] cycloaddition affords dihydropyridazine or pyridazine conjugates, showing the highest kinetics in the bioorthogonal reaction portfolio (figure 22).46 Figure 22. a) General scheme for the inverse electron demand diels-alder (IEDDA). b) Common cyclooctenecyclooctynes employed nowadays for the IEDDA and some of their kinetics in methanol-water mixtures.93,94,95,96 The first relevant report appeared in 2008 when Fox and co-workers demonstrated the biocompatibility and efficiency of the reaction in the presence of nucleophiles, biological media or even mammalian cell lysates.94 This was further confirmed by the group of Devaraj which performed the reaction on an antibody bearing a norbornane structure (NB) on pre-targeted cells.93 The authors introduced the use of a smart probe in this transformation, taking advantage of the resonant energy transfer between the selected BODIPY employed fluorophore and the tetrazine. Additionally, they corroborated, by the first time, an intracellular IEEDA cycloaddition by incubation of mammalian cells with a taxol analogue, able to bind the microtubular network and posterior reaction with transcyclooctene (TCO) as dienophile.97 92 S. I. Presolski, V. Hong, S. H. Cho, M. G. Finn, J. Am. Chem. Soc. 2010, 132, 14570. 93 R. Weissleder, S. A. Hilderbrand, N. K. Devaraj, Bioconjugate Chem. 2008, 19, 2297. 94 M. L. Blackman, M. Royzen, J. M. Fox, J. Am. Chem. Soc. 2008, 130, 13518. 95 J. Yang, Y. Liang, J. S. K. N. Houk, N. K. Devaraj, Chem. Eur. J. 2014, 20, 3365. 96 K. Lang, L. Davis, S. Wallace, M. Mahesh, D. J. Cox, M. L. Blackman, J. M. Fox, J. W. Chin, J. Am. Chem. Soc. 2012, 134, 10317. 97 S. Hilderbrand, R. Upadhyay, R. Mazitschek, R. Weissleder, N. K. Devaraj, Angew. Chem. Int. Ed. 2010, 49, 2869. Chapter I 34 The groups of Fox and Lemke demonstrated, almost simultaneously along 2012, that different dienophiles could be incorporated via genetic code expansion in bacteria98 or even in mammalian cells99,100 undergoing the reaction despite the size of the employed TCO, NB, or bicyclononyne (BCN). Prescher and co-workers designed a cyclopropene attached to a sugar analogue (9-AzNeuAc) improving efficiency of metabolism incorporation and cell surface labeling. The prior treatment of cells with 1 mM of the modified sugar, allowed them to bioconjugate a tetrazinebiotin (100 M) on mammalian Jurkat cell membrane after 1 h at 37 ºC (figure 23).101 Figure 23. Inverse electron demand diels-alder on cell surface using cyclopropane as bioorthogonal handle developed by Prescher and co-workers in 2012.101 The IEEDA cycloadditions have been used for different type of exciting applications, ranging from super-resolution microscopy,102,103 to the detection of radiotracers.104,105 Nonethelesss, this inverse Diels Alder cycloaddition also shown problems of side reactivity, because of the presence of the strained reactants, and the synthesis of some of the precursors is not trivial.46 98 J. L. Seitchik, J. C. Peeler, M. T. Taylor, M. L. Blackman, T. W. Rhoads, R. B. Cooley, C. Refakis, J. M. Fox, R. A. Mehl, J. Am. Chem. Soc. 2012, 134. 2898. 99 T. Plass, S. Milles, C. Koehler, J. Szymanski, R. Mueller, M. Wießler, C. Schultz, E. A. Lemke, Angew. Chem. Int. Ed. 2012, 51, 4166. 100 J. W. Chin, K. Lang, L. Davis, S. Wallace, M. Mahesh, D. J. Cox, M. L. Blackman, J. M. Fox, J. Am. Chem. Soc. 2012, 134, 10317. 101 D. M. Patterson, L. A. Nazarova, B. Xia, D. N. Kamber, J. A. Prescher, J. Am. Chem. Soc. 2012, 134, 18638. 102 G. Beliu, A. J. Kurz, A. C. Kuhlemann, L. Behringer-Pliess, M. Meub, N. Wolf, J. Seibel, Z. Shi, M. Schnermann, J. B. Grimm, L. D. Lavis, S. Doose, M. Sauer, Commun. Biol. 2019, 2, 261. 103 P. Werther, K. Yserentant, F. Braun, N. Klatwasser, C. Popp, M. Baalmann, D. Herten, R. Wombacher, Angew. Chem. Int. Ed. 2020, 59, 804. 104 T. Läppchen, R. Rossin, T. R. van Mourik, G. Gruntz, F. J. M. Hoeben, R. M. Versteegen, H. M. Janssen, J. Lub, M. S. Robillard, Nucl. Med. Biol. 2017, 55, 19. 105 J. L. Houghton, R. Membreno, D. Abdel-Atti, K. M. Cunanan, S. Carlin, W. W. Scholz, P. B. Zanzonico, J. S. Lewis, B. M. Zeglis, Mol. Cancer Ther. 2017, 16, 124. Chapter I 35 2.8. Photochemical ligations A few bioorthogonal, click-like reactions promoted by light have also been described. In 2008, Lin’s group demonstrated that photo-induced 1,3 – dipolar cycloadditions can be made compatible with biological media, such as bacterial lysates.106 This reaction, that was coined as “photoclick” later on,107 relies on the generation of a highly reactive nitrile-imine dipole after UV irradiation of a 2,5 – diaryl tetrazole, and the subsequent release of N2. This species, cyclizes spontaneously with a proper dipolarophile, either alkene or an electron deficient alkyne, to afford pyrazoline cycloadducts (figure 24).108 Figure 24. General scheme of the photoinduced 1,3-dipolar cycloaddition. This group successfully translated this transformation into E. coli bacteria by genetically encoding,109 or into mammalian cells by metabolic incorporation of homoallylglycine (HAG, 1 mM) as Met analogue (figure 25).110 Thus, the alkenyl-tagged HeLa cells were treated with 100 M of tetrazole, and set under 365 nm or two-photon IR laser irradiation along 2 min at 37 ºC. Nevertheless, despite the fluorogenic nature of the pyrazolines, the fluorescence obtained remained still low and controls showed background noise. Figure 25. Intracellular version of the “photo-click” reaction developed by Lin in 2010.110 106 J. S. Clovis, A. Eckell, R. Sustmann, R. Huisgen, Chem. Ber. 1967, 100, 60. 107 R. K. V. Lim, Q. Lin, Acc. Chem. Res. 2011, 44, 828. 108 W. Song, Y. Wang, J. Qu, M. M. Madden, Q. Lin, Angew. Chem. Int. Ed. 2008, 47, 2832. 109 W. Song, Y. Wang, J. Qu, Q. Lin, J. Am. Chem. Soc. 2008, 130, 9654. 110 W. Song, Y. Wang, Z. Yu, C. I. R. Vera, J. Qu, Q. Lin, ACS Chem. Biol. 2010, 5, 875. Chapter I 36 The “photo-click” also faces important limitations: (i) selectivity issues (the nitrile-imine intermediate has shown to react with any kind of nucleophile, even with solvents); (ii) slower kinetics than those of CuAAC; (iii) the requirement of UV light.111 Photochemistry has recently gained attention again, with the incorporation of hindered tetrazoles, the development of photo-triggered versions of other bioorthogonal transformations such as the SPAAC112 or IEDDAC113 as well, or more recently, the development of reactions in cellular settings.114,115 111 Z. Li, L. Qian, L. Li, J. C. Bernhammer, H. V. Huynh, J. Lee, S. Q. Yao, Angew. Chem. Int. Ed. 2016, 55, 2002. 112 A. A. Poloukhtine, N. E. Mbua, M. A. Wolfert, G. Boons, V. V. Popik, J. Am. Chem. Soc. 2009, 131, 15769. 113 S. V. Mayer, A. Murnauer, M. von Wrisberg, M. Jokisch, K. Lang, Angew. Chem. Int. Ed. 2019, 58, 15876. 114 Z. Yao, X. Xu, X. Zhang, Q. Xiong, S. Jiang, Z. Yu, Org. Biomol. Chem. 2019, 17, 6777. 115 J. Li, H. Kong, L. Huang, B. Chen, K. Qin, M. Zheng, Z. Yan, Y. Zhang, J. Am. Chem. Soc. 2018, 140, 14542. Chapter I 37 3. Transition metal-catalyzed transformations in biological settings 3.1. Organometallic catalysis. Reactions in aqueous media As shown in the introduction, among the different bioorthogonal reactions so far developed, the CuAAC, one of the most prominent transformations, is catalyzed by transition metals (TM, mainly copper). This process follows a typical organometallic mechanism based on elementary steps indicated in figure 26.116,117 This suggests that other organometallic catalyzed processes could be also compatible with biological conditions. However, despite organometallic catalysis is a powerful tool for a broad range of reactivity including redox transformations, cross-couplings, isomerizations or formal cycloadditions,118 it is yet underexplored in the biological field. This is in great part due to the established view that organometallic pathways are incompatible with water and oxygen, and also with biological components.119 Figure 26. Typical elementary organometallic steps, involving either inner or outer sphere mechanisms. 116 A. K. Å. Persson, PhD thesis, Palladium (II)-Catalyzed Oxidative Cyclization Strategies. 117 P. Destito, PhD thesis, Transition metal-promoted bioorthogonal transformations in biological media. 118 R. H. Crabtree, The Organometallic Chemistry of the Transition Metals, 6th Edition, Wiley, 2014. 119 A. Chanda, V. V. Fokin, Chem. Rev. 2009, 109, 725. Chapter I 38 However, in recent years a number of TM-catalyzed transformations have been demonstrated to tolerate water and the presence of air.120 This is the case of the Suzuki’s reaction, one of the most popular cross coupling transformations (figure 27).121,122 It has not only been performed in water but also in the presence of biological mixtures. It is worth to mention that the Sonogashira and Heck cross couplings have followed a similar evolution along the years.123,124,125 Figure 27. Standard conditions for Suzuki cross coupling.126 RuPhos: 2-Dicyclohexylphosphino-2’-6’- diisopropoxybiphenyl. The Suzuki cross-coupling starts with the oxidative addition of an organohalide to a palladium(0) (Pd0) complex. The resulting PdII complex reacts with an organometallic species by transmetalation, giving as a result a PdII intermediate bearing the two organic coupling partners. At this point, a reductive elimination step provides the desired organic compound containing the new C-C bond, recovering Pd0, able to re-enter into the catalytic cycle (figure 28).121,127 Figure 28. Suzuki-Miyaura mechanism.121,127 120 B. Cornils, W. A. Herrman, Aqueous-Phase Organometallic Catalysis, 2nd Edition, Wiley, 2004. 121 C. C. C. J. Seechurn, M. O. Kitching, T. J. Colacot, V. Snieckus, Angew. Chem. Int. Ed. 2012, 51, 5062. 122 D. G. Hall, Boronic acids Volume 2, 2nd Edition, Wiley, 2011. 123 P. G. Isengger, B. G. Davis, J. Am Chem. Soc. 2019, 141, 8005. 124 A. L. Satz, J. Cai, Y. Chen, R. Goodnow, F. Gruber, A. Kowalczyk, A. Petersen, G. Naderi-Oboodi, Lucja Orzechowski, Q. Strebel, Bioconjugate Chem. 2015, 26, 1623. 125 K. Kodama, S. Fukuzawa, H. Nakayama, T. Kigawa, K. Sakamoto, T. Yabuki, N. Matsuda, M. Shirouzu, K. Takio, K. Tachibana, S. Yokoyama, ChemBioChem 2006, 7, 134. 126 S. D. Dreher, S. E. Lim, D. L. Sandrock, G. A. Molander, J. Org. Chem. 2009, 74, 3626. 127 A. De Meijere, F. Diederich, Metal-Catalyzed Cross-Coupling Reactions Volume 1, 2nd Edition, Wiley, 2004. Chapter I 39 Using water soluble ligands like dpdbs, it is possible to carry out the couplings even in pure water (figure 29).128 Figure 29. Suzuki-Miyaura reaction developed in aqueous media by Buchwald and coworkers in 2005.128 Dpdbs: Sodium 2-Dicyclohexylphosphino-2’,6’-dimethoxybiphenyl-3’-sulfonate. The groups of Schultz and Davis explored the application of the Suzuki coupling for the modification of biomacromolecules in aqueous solvents.129 Therefore, Davis and co-workers demonstrated that it is possible to perform efficient bioconjugations to proteins in phosphate aqueous buffers after 30 min at 37 ºC, using 50 equivalents of palladium complex, and 500 equivalents of the corresponding phenylboronic acid (figure 30).130 Notably, the buffer solution employed (pH∼8) is basic enough to allow the key transmetalation step to occur in the biological media, without addition of any other base. Figure 30. Biocompatible Suzuki's version reported by Davis and co-workers in 2009.130 Bpin: Pinacol boronic ester. In addition to the Suzuki coupling other type of transition metal catalyzed cross couplings, like Sonogashira or metathesis transformations, have been also demonstrated to work in aqueous media, and in bioconjugation reactions. And more recently, the range of reactions has been expanded to include redox process, isomerizations, cyclizations and of course, cycloadditions, like the CuAAC or the RuAtAC.131,132 128 K. W. Anderson, S. L. Buchwald, Angew. Chem. Int. Ed. 2005, 44, 6173. 129 E. Brustad, M. L. Bushey, J. W. Lee, D. Groff, W. Liu, P. G. Schultz, Angew. Chem. Int. Ed. 2008, 47, 8220.130 J. M. Chalker, C. S. C. Wood, B. G. Davis, J. Am. Chem. Soc. 2009, 131, 16346. 130 J. M. Chalker, C. S. C. Wood, B. G. Davis, J. Am. Chem. Soc. 2009, 131, 16346. 131 P. Destito, C. Vidal, F. López, J. L. Mascareñas, Chem. Eur. J. 2021, 27, 4789. 132 P. H. Dixneuf, V. Cadierno, Metal-Catalyzed Reactions in Water, Wiley-VCH Verlag GmbH & Co. KgaA, 2013. Chapter I 40 3.2. Transition metal catalysis in biological habitats and living cells Translating organometallic reactions from water to biological mixtures is especially challenging owing to the presence of numerous components that can deactivate the catalyst or trap the intermediates of the reaction mechanism. However, as previously indicated for the CuAAC, this transition is possible. Initial examples demonstrating the viability of carrying out metal-promoted processes in biological environments appeared in 1985, when Cséplö and co-workers reported the hydrogenation of unsaturated fatty acids in presence of mesophyll protoplast plant cells using a water-soluble ruthenium catalyst, although the reaction generated toxicity.133 It took time until other organometallic processes could be performed in biological settings. I will now summarize some of the most relevant contributions, organized by reaction types. 3.2.1. Bond cleavage (uncaging) transformations In 2006, the Meggers group noticed the ability of the commercially available ruthenium(II) Cp*Ru(COD)Cl to promote the cleavage of allylcarbamates into their corresponding amines, in the presence of water, at open air, with the aid of external thiols. Moreover, a caged rhodamine 110 exhibiting no fluorescence, was used to check the viability of the methodology inside living mammalian cells. A 10-fold increase was observed in the cytoplasm of the cells, after cell treatment with 100 M of the allocated probe, 20 mol% of ruthenium catalyst and 5 equivalents of thiophenol for 15 min (figure 31). This value decreased to a 3.5-fold increase if no external thiol was added.134 This report, which was unnoticed for several years, is now considered, together with the CuAAC, a pioneering contribution demonstrating the viability of performing transition metal catalysis inside living mammalian cells. In 2014, the group of Mascareñas employed the uncaging reaction for the controlled release of DNA binders inside mammalian cells.135 Figure 31. Allylcarbamate deprotection of rhodamine 110, inside HeLa cells reported by Meggers and co-workers in 2006.134 133 L. Vigh, F. Joó, A. Cséplö, J. Biochem. 1985, 146, 241. 134 C. Streu, E. Meggers, Angew. Chem. Int. Ed. 2006, 45, 5645. 135 M. I. Sánchez, C. Pernas, E. Vázquz Sentís, J.L. Mascareñas, Chem. Sci. 2014, 5, 1901. Chapter I 47 3.2.5. Cycloisomerizations Gold complexes are well known π acids capable of activating alkynes. This activation can be used to promote cycloisomerization processes, even in cellular media. Figure 39. General scheme for the activation of gold sensitive probes. For instance, Tae and co-workers reported in 2009 the synthesis of an oxazole-TAMRA after cyclization with Au(III) of propargyl amide-TAMRA. The concurrent increase of fluorescence from the oxazole generation was used for Au detection in cells.167 Similar strategies were designed by Yoon,168,169 Ahn and co-workers.170 Another annulation strategy was published by Patra’s group. Therein, a fluorescein probe was derivatized with a 2-ethynylbenzoic acid moiety, with its consequent fluorescence quenching. The presence of Au(I) or Au(III) triggers an intramolecular cyclization, that would release an active fluorescein and the new created isochromene skeleton.171 In the same way as previous works, it was claimed to be performed inside mammalian cells, although only one preliminary experiment was shown. In these articles, the main objective was the creation of Au detection probes. In 2018, the group of Mascareñas reported a gold mediated cycloisomerization in mammalian cells. They demonstrated that gold chloride complexes can be activated in aqueous media without the need of a chloride scavenger, allowing phenylpropyonates to cycle and build fluorescent coumarins.172 Remarkably, the reaction can be also carried out concurrently and orthogonally with the previously mentioned ruthenium promoted allyl deprotection. This was the first time that two metals promoted bioorthogonal transformations inside living cells in a total independent manner, working in parallel, without observing cross reactivities. Figure 40. Cycloisomerization inside HeLa cells performed by Mascareñas group in 2018.172 Y. Yang, S. Lee, 167 J. Tae, Org. Lett. 2009, 11, 5610. 168 M. J. Jou, X. Chen, K. M. K. Swamy, H. N. Kim, H. Kim, S. Lee, J. Yoon, Chem. Commun. 2009, 7218. 169 J. H. Do, H. N. Kim, J. S. Kim, H. J. Kim, J. Yoon, Org. Lett. 2010, 12, 932. 170 O. A. Egorova, H. Seo, A. Chatterjee, K. H. Ahn, Org. Lett. 2010, 12, 401. 171 N. T. Patil, V. S. Shinde, M. S. Thakare, P. H. Kumar, P. R. Bangal, A. K. Barui, C. R. Patra, Chem. Commun. 2012, 48, 11229. 172 C. Vidal, M. Tomás-Gamasa, P. Destito, F. López, J. L. Mascareñas, Nat. Commun. 2018, 9, 1913. Chapter I 48 3.2.6. Carbene-mediated transformations Recently, our group demonstrated that carbene insertions can also be carried out inside mammalian cells, provided the substrates are appropriately engineered. Metal-carbene complexes had been previously used for the modification of biomolecules, especially with Rh,173 but had been never explored in a living system. In the current work, the group was able to assemble complex structures such as benzoquinoxalines, starting from alpha-keto diazocarbenes, orthoamino arylamines and benign Cu(II) salts.174 Interestingly, this process could discriminate between different cell lines thanks to the derivatization of a copper ligand with an integrintargeting motif, that could be further used for the cell selective generation of a bioactive benzoquinoxaline with mitochondria depolarizing properties (figure 41). Due to the wide reactivity of carbenoids, the authors foresee novel future applications of the carbene transfer chemistry into the field of bioorthogonal chemistry. Figure 41. Carbene mediated N-H insertion in selectively directed mammalian cells by Mascareñas group in 2021.174 4. Conclusions In summary, despite the well-established belief that transition metal catalysis was incompatible with biological systems, an increasing number of evidences indicate that this hypothesis is wrong, and that transition metal catalyzed reactions can be performed in complex biological media and even in living contexts. It is true, however, that most of the transformations are very poor in terms of catalytic efficiency, in great part owing to the deactivation of the catalyst or catalytic intermediates, but also because of transport, localization and toxicity problems. The field is in its infancy, and much remains to be progressed until we can truly add this abiotic chemistry to the armory of reactions that can be performed in cells and organisms. 173 K. A. Mix, M. R. Aronoff, R. T. Raines, ACS Chem. Biol. 2016, 11, 3233. 174 S. Gutiérrez, M. Tomás-Gamasa, J. L. Mascareñas, Angew. Chem. Int. Ed. 2021, 60, 22017. GENERAL OBJECTIVES General objectives 51 Considering all precedents and the current challenges in the field of bioorthogonal chemistry, this doctoral thesis aims the study and improvement of existing bioconjugation techniques, and the development of novel biocompatible transformations in native settings, relaying on the power of transition metal catalysis for the synthesis of complex structures. Specifically, we focus on: 1) Translation of CuAAC into mammalian cells: Previous works have been mainly focused on cell surface bioconjugation with some coy contributions inside cells with the use of nonconventional methodologies such as cell fixation or the use of azide-chelating copper complexes. The limitations have raised the question if a simpler process involving discrete substrates is plausible with the use of predefined Cu(I) complexes in absence of exogenous reductants. We have approached this goal, and the results are discussed in chapter II. 2) Novel cycloadditions in biological media: As described before, our group has gained some expertise in synthetic protocols in biological media and cellular settings based mainly in ruthenium catalysts. Thus, the transition metal has demonstrated to be a biocompatible and flexible for a diverse kind of reactivity. In chapter III, we expand the bioorthogonal reaction pool with a metal-promoted [2+2+2] cycloaddition, carried out for the first time inside mammalian cells. Therein, this transformation is applied for the one-step synthesis of the core skeleton of important biomolecules and fluorophores. 3) Synthetical photoredox catalysis approximation in native settings: Visible light photocatalysis has recently gained attention for biological purposes due to its biocompatibility and the spatiotemporal control it provides. Nevertheless, synthetic applications in cell environment are unknown until date. In chapter IV, we explore some photo-Meerwein arylations in aqueous and biological media investigating the bioorthogonality of these synthetic photoredox transformations. CHAPTER 2. COPPER AZIDE-ALKYNE CYCLOADDITION IN MAMMALIAN CELLS This chapter includes work published in Chemical Science as: J. Miguel-Ávila,a M. Tomás-Gamasa,a A. Olmos,b Pedro J. Pérez,b J. L. Mascareñas,a Chem. Sci. 2018, 9, 1947. aCentro Singular de Investigación en Química Biolóxica e Materiais Moleculares (CIQUS) and Departamento de Química Organica, Universidad de Santiago de Compostela, 15782, Santiago de Compostela, Spain. bLaboratorio de Catálisis Homogénea, Unidad Asociada al CSIC, CIQSO-Centro de Investigación en Química Sostenible, Departamento de Química, Universidad de Huelva, Campus de El Carmen s/n, 21007 Huelva, Spain. DOI: 10.1039/C7SC04643J, Open Acces article (RSC, Most impactful Chemical Biology articles of 2018) Chapter II 55 1. CuAAC inside live mammalian cells As mentioned in the introduction, the copper azide-alkyne aycloaddition, known as CuAAC, that could be considered as a Nobel Prize-like discovery, has been key to trigger the field of bioorthogonal chemistry. Even now, after years of progress in this area, the CuAAC, together with the emergence of the IEDDA, can be considered as the more relevant bioorthogonal transformation so far discovered.46 The simplicity of the substrates in terms of synthesis and handling, its intrinsic chemoselectivity and the compatibility with aqueous media, makes this reaction a priority choice for performing bioorthogonal ligations.14,37 However, translating this process to living cells and organisms is very challenging, mostly because of the instability of the copper complex, its potential sequestration by biological components and its presumable toxicity.77,78 As indicated in the introduction, most significant advances in this topic were carried out only in bacteria, that is much less complex than a mammalian cell. Initial attempts to carry out CuAAC reactions in mammalian cells were reported in 2006, using HeLa and HEK293T cell lines,68 and some years later, in 2008, employing mice derived cells.175 However, fixation of cells with paraformaldehyde before their treatment with substrates and the copper source was mandatory in both cases to observe reaction. Hence, this example cannot be considered a real in cellulo application of the click reaction. In 2012, the group of Ting relied on the smart concept of using “Cu chelating azides”, where the combination of copper salts with different 2-methylazidopyridines resulted in an improvement of several orders of magnitude on kinetics with respect to the use of other azides (figure 42c).176,177 The reason behind this acceleration of the CuAAC is the enhancement of electrophilicity that experiences the azido group once the neighbor pyridine is chelated with the metal. This is also an advantage in terms of orthogonality, owing to the preference of the copper to be chelated by a bidentate ligand. The group of Ting demonstrated that the reaction can be performed on HEK cells, incorporating an Hpg (200 M) aminoacid on the membrane. The reaction was carried with 200 M of pycolil azide, 40  of CuSO4 as copper source, 5 equivalents of BTTAA or THPTA ligand, 2.5 mM of NaAsc for 5 min at 37 ºC in PBS as cellular media. Albeit it was also reported to be performed inside A375 cells with metabolic incorporated Hpg and ethynyl-uridine (50 and 200 M, respectively), the expriments were carried out with fixed cells prior to substrate incubation, and conditions were forced with the use of 2 mM of CuSO4, 4 equivalents of THPTA and 5 equivalents of NaAsc for 1 h. 175 A. Salic, T. J. Mitchison, Proc. Natl. Acad. Sci. USA 2008, 105, 2415. 176 U. Sing, P. Slade, K. R. Gee, A. Y. Ting, Angew. Chem. Int. Ed. 2012, 51, 5852. 177 C. Uttamapinant, M. I. Sánchez, D. S. Liu, J. Z. Yao, K. A. White, S. Grecian, S. Clarke, K. R. Gee, A. Y. Ting, Nat. Protoc. 2013, 8, 1620. Chapter II 56 The first report of an intracellular mammalian cell CuAAC had to wait until 2014. Taran developed a different version of the same concept, modifying the standard tristriazole ligands with a chelating azide into its structure (figure 42d). These substrates underwent the reaction with alkynes with very good rates and efficiencies, using micromolar concentrations, and in presence of cell lysates.178 Figure 42. Comparison of CuAAC methodologies under diluted conditions. The chelating-azide substrate developed by Taran’s group (figure 42c) was implemented in HuH7 human cells, using a microtubule skeleton tracker modified with an alkyne moiety and the Cu chelating azide precursor conjugated to a TAMRA fluorophore. For this purpose, CuSO4 was incubated with one equiv. of the chelating azide and 50 equiv. of NaAsc. The mixture was added to a final concentration of 50 M into cells incubated with 62.5 nM of the alkyne probe at 37 ºC. After 2 h, the cells were fixed with paraformaldehyde, and a second incubation with Oregon green-tubulin tracker was performed to confirm the intracellular reaction and the colocalization with the generated red fluorophore (figure 43). Figure 43. Imaging experiments from Taran’s work. From left to right: blue channel for the DAPI-nuclei tracker probe, red channel for the product of the reaction with TAMRA, green channel for the Oregon green-tubulin tracker, and merging of pictures. Yellow color resulting from merging between red and green channel corroborates co-localization between the product and the tubulin tracker.178 A major disadvantage of these latter chelating azide strategies is that the copper catalyst has to be used in stoichiometric amounts, as it is trapped in the final adduct. 178 V. Bevilacqua, M. King, M. Chaumontet, M. Nothisen, S. Gabillet, D. Buisson, C. Puente, A. Wagner, F. Taran, Angew. Chem. Int. Ed. 2014, 53, 5872. Chapter II 63 Figure 52. Calibration of the coumarin probe 14. Additionally, we also prepared the azido-benzothiazole 18, that after undergoing CuAAC it exhibits green-red fluorescence (figure 53). This would open the door for future orthogonal reactions using both probes in parallel. This azide was obtained in 2 steps.184 The first step relays in the addition of an in situ prepared methyl carbanion of the benzothiazole 15, to the aldehyde 16, resulting in the alcohol 17 in a moderate 44% yield. Then, the hydroxyl group is eliminated under acidic conditions to yield the probe 18 (37%). Figure 53. Synthesis of the probe azido-benzothiazole 18.184 This azide reacts with propargylic alcohol using CuSO4 and NaAsc in a mixture 1:1 of waterethanol, albeit with low yield. We later found that while the cycloaddition product 19 is fluorescent, the reacting azide also presents some fluorescence background, which somewhat complicates the use of this chemistry for monitoring the transformation by fluorescence. Figure 54. Synthesis of the fluorescent triazole 19. 184 A. Herner, I. Nikic, M. Kállay, E. A. Lemke, P. Kele, Org. Biomol. Chem. 2013, 11, 3297. Chapter II 64 3.1.3. CuAAC reactions with azides 13 The initial CuAAC experiments were carried out in a plate reader (96 wells), what allows us to perform reactions by triplicate. The procedure consistes of first preparing a 2.5 mM solution of the Cu(II)-tristriazole complexes just before each use, by mixing a fresh water solution of CuSO4 and an aqueous solution of the corresponding ligand L (5% DMSO, 2 equiv. respect Cu) for 10 minutes in a 0.5 mL Eppendorf, at room temperature under air. Concurrently, a mixture containing 115 M of azide and 230 M propargyl alcohol in H2O or PBS is prepared in a 2 mL Eppendorf tube starting from fresh azide solution in DMSO and aqueous fresh solution of propargyl alcohol. 87 L of the solution with the substrates is added into the corresponding well, followed by 3 L of the Cu(II)-L solution. Finally, 10 L of a 25 mM NaAsc is added into the wells, thus promoting the formation of the Cu(I) species and eventually triggering the cycloaddition under air at room temperature. In summary, the final conditions are 100 M of azide, 200 M of alkyne, 75 M of the Cu(II)-L complex (75 mol%) and 2.5 mM of NaAsc (if present) in H2O (0.5% of final DMSO content). Figure 55. Schematic representation of the reaction screening methodology employed in this work. As expected from previous reports, initial cycloaddition experiments with the probe 13 gave good results using H2O as solvent. All ligands were able to promote the reaction in 20 min (table 1), but practically no fluorescence increase was found when no ligand was used. The yields were calculated using the calibration curve previously shown (figure 55), and are indicated in table 1. Chapter II 65 Entry Ligand Time (min) Yield (%) 1 L1 20 56 2 L2 20 47 3 L3 20 49 4 L4 20 61 5 - 20 - 6 L1 30 56 7 L2 30 48 8 L3 30 50 9 L4 30 61 10 - 30 - Table 1. Catalytic activity of the different Cu(I)-L complexes (considering the complete reduction) with the hydroxicoumarin 13. Conditions: 100  M of azido-probe 13, 200  M of propargyl alcohol, 75  M of Cu(I)-L (from CuSO4, and L preincubation), 2.5 mM NaAsc, in water at 25 ºC for 20 or 30 min. As shown in the table 1, the new designed Cu(I)-L4 complex affords the product in 61% yield, slightly better than when L1 is employed, the more commonly used in bioorthogonal experiments. No further increase in yields was detected after 30 min of reaction. These results, that confirms the water tolerance of the reaction, led us to carry out a preliminary test of the viability of performing these reactions in mammalian cells. The complex of Cu(I)-L4 was freshly prepared in a similar way to that used in the experiments in the plate reader. More in detail, both the copper salt (CuSO4) and the ligand were mixed in a 1:2 stoichiometry in H2O for 1 h to ensure the complete complexation of the copper, to a final concentration of 2.5 mM (Cu(II)- L). Then 6.25 equiv. of NaAsc were added and the mixture was kept for 30 min, to ensure the reduction of the copper (Cu(I)-L). A solution of this mixture (12 L) in 388 L of DMEM was added to a culture of cells (HeLa, 400.000), for a final concentration of 75 M. After 30 min incubation, cells were thoroughly washed with DMEM twice to ensure that the extracellular copper was removed. Afterwards, cells were treated with a solution of 100 M of probe 13 and 200 M of propargyl alcohol, in 400 L of DMEM. After 1 h, the cells were washed with DMEM (x2) and observed under the microscope. This protocol, that is quite general for the cellular experiments, minimizes extracellular reactions, because of the washing steps. Chapter II 66 Figure 56. Schematic representation of the methodology employed in cellular experiments in this work. Unfortunately, we didn’t observe the appearance of blue fluorescence at all, characteristic of the product (figure 57). We hypothesized that this failure, might be related to the formation of a very low amount of the product, which is not particularly highly fluorescent, or perhaps with internalization problems of the azide probe. Figure 57. Fluorescence micrographies of HeLa cells with probe 13 and the in situ made Cu(I)-L4 complex. Protocol: cells were incubated with 75  M of Cu(I)-L4 complex (from CuSO4, L and NaAsc preincubation), washed and then incubated with 100  M probe 13, and 200  M propargyl alcohol at 37 ºC for 1 h. A) incubation of cells with probe 13 and propargyl alcohol. B) Incubation of cells with substrates, and freshly prepared Cu(I)-L4 complex. C) Bright-field from B. Fluorescence obtained in the blue channel (λexc = 385 nm, λem = 410-480 nm). Chapter II 67 3.1.4. Selection of probe 20 In view of previous results and having in mind the higher stability of alkyl azides when compared to aryl azides,185 we considered to test the behavior of 9-(azidomethyl)-anthracene 20, that was easily prepared by reaction of the corresponding bromide with NaN3. In addition, its partial hydrophobicity could favor the cellular internalization. We were glad to confirm that this probe undergoes the CuAAC annulation with propargyl alcohol to give the expected triazole 21, when treated with CuSO4 and NaAsc in a H2O-tBuOH mixture (1:1, not totally solubilized, 1.18 M). The product was isolated in a 41% yield (figure 58a). The triazole 21 is 155-fold more fluorescent than the azide precursor 20, which facilitates monitoring the process by fluorescence (figure 58b and c). Figure 58. a) Synthesis of triazole 21, b) Fluorescence calibration curve. c) Absorbance and fluorescence spectra. 185 J. H. Boyer, F. C. Canter, Chem. Rev. 1954, 54, 1. Chapter II 68 3.1.5. CuAAC reactions with anthracenyl azide 20 With this new probe 20 in hand, we performed a detailed study of the CuAAC at a scale of 100 M in water and PBS, using 2 equiv. of propargyl alcohol, 75 M of Cu(I)-L complexes (75 mol%) and 33 equiv. of NaAsc to trigger the reaction, following the stablished methodology developed with the coumarin 13 (plate reader). In the absence of ligands, i.e., when only CuSO4 and NaAsc were employed, the reaction gaves poor yields even after 24 h. Similarly, if no NaAsc was used to trigger the reaction, no fluorescence enhancement was observed, independently of the ligand employed. Gratifyingly, using the preformed and reduced complex of the Cu(II) salt with L4 we observed a 55% yield in water and 54% in PBS after 10 min of reaction, while with Cu(I)-L1 and Cu(II)-L3 led to lower yields (around 30%). After 20 minutes, Cu(I)-L4 provided almost 70% yield in both water and PBS. These results were further corroborated by HPLC-UV, where similar outcomes were found after 1 h in H2O. Besides, full conversion of the azide was corroborated (see experimental section, Chapter II S4). Figure 59. Catalytic activity of the different Cu(I)-L complexes (considering the complete reduction, L) with the antracenyl azide probe 20. Conditions: 100  M of azido-probe 20, 200  M of propargyl alcohol, 75  M of Cu(I)-L (from CuSO4 and L preincubation), 2.5 mM NaAsc, in water or PBS at 25 ºC for 10 or 20 min. Chapter II 69 We next performed studies along time for comparing the activities of the different Cu(I)-L complexes (figure 60). When the reaction was carried out with Cu(I)-L4, the plateau was reached in 20 min, while with Cu(I)-L1 full conversion required around 10 minutes extra, and Cu(I)-L3 other additional 20 minutes. Therefore, the best results were obtained using L4 as ligand. We then checked the influence of the catalyst loading, finding out that we can go down to 25 mol% (25 M) and the yield is not specially affected. Meanwhile, in the case of L1 or L3, decreasing the catalyst loading below 50 mol% (50 M) entailed a drastic decrease in the reactivity. Figure 60. Experiments along time with Cu(I)-L1, Cu(I)-L3 and Cu(I)-L4. Conditions: 100  M of azido probe 20, 200  M of propargyl alcohol, the corresponding amount of Cu(I)-L (from CuSO4 and L preincubation), 2.5 mM NaAsc, in water or PBS at 25 ºC along the time. Chapter II 70 3.2. Cellular experiments with in situ prepared tristriazole complexes Cellular experiments were carried out with the anthracenyl azide 20, using similar conditions than described before: 100 M of azide 20, 2 equiv. of propargyl alcohol and 75 M of the corresponding Cu(I)-L complex, previously prepared from the incubation of CuSO4, the selected ligand and NaAsc. In contrast with previous experiments in HeLa cells using probe 13, we did observe the rise of fluorescence (figure 61) across the cytoplasm and in vesicles, with cells showing an unaltered morphology, which was more intense when the experiment was carried out with ligand L4 (figure 61, panel F), although with other ligands we also observed reactivity. Extracellular reactivity could be discarded due to the extensive washing protocols. In addition, this was further confirmed by the observation of a total lack of reactivity in control experiments using extracellular media. Figure 61. Fluorescence micrographies of HeLa cells with different Cu(I)-L complexes. Protocol: cells were incubated with 75  M of Cu(I)-L (from CuSO4, Ln and NaAsc preincubation), washed and then incubated with 100  M of azide 20, and 200  M of propargyl alcohol at 37 ºC for 1 h. A) Control without Cu. B) Ligandless Cu(I). C) Cu(I)-L2. D) Cu(I)-L3. E) Cu(I)-L1. F) Cu(I)-L4. Fluorescence obtained in the blue channel (λexc = 385 nm, λem = 410-480 nm). Chapter II 71 It is well known that the cell presents a reductive environment.186,187 Thus, we wondered whether the Cu(II)-L complex could be directly reduced inside cells to the active Cu(I) complexes. In fact, as it can be observed in figure 62, panel A, in experiments carried out with Cu(II)-L3 without preforming the reduction of Cu(II) complex with NaAsc, there is some residual fluorescence, that can be slightly increased with the subsequent addition of NaAsc to the cell incubation media (figure 62, panel B). The addition of well-known Cu chelators such EDTA in the cellular media didn’t affect the fluorescence outcome (figure 62, panel C). Nonetheless, the intensity of the signals is clearly lower than that obtained when the Cu(II)-L complexes were pretreated with NaAsc. Figure 62. Fluorescence micrographies of additional experiments in HeLa cells. Protocol: cells were incubated with 75  M of the chosen Cu catalyst, washed and then incubated with 100  M of azide 20, and 200  M of propargyl alcohol at 37 ºC for 1 h. A) Cu(II)-L3 (from CuSO4 and L3 preincubation without NaAsc). B) Cu(II)-L3 (from CuSO4 and L3 preincubation) plus posterior incubation with 6.25 equiv. NaAsc, with respect to Cu content. C) Cu(I)-L3 (from CuSO4, L3 and NaAsc preincubation) and 2 equiv. EDTA with respect to Cu content. Fluorescence obtained in the blue channel (λexc = 385 nm, λem = 410-480 nm) It is important to indicate that in the above cellular experiments we didn't not observe signs of toxicity, and cell shapes observed in brightfield microscope were the natural cell morphology (see MTT assays later). 186 H. J. Forman, J. Fukuto, M. Torres, Signal Transduction by Reactive Oxygen and Nitrogen Species: Pathways and Chemical Principles 2003, Kluwer Academic Publishers. 187 B. A. Wagner, V. G. J. Rodgers, G. R. Buettner, Cell Biochem. Biophys. 2013, 67, 477. Chapter II 72 3.3. CuAAC reactions with predefined copper (I) complexes In the previous experiments, the copper(I) complexes were made by in situ reduction of the corresponding Cu(II) precursors, which requires the use of a reductant, and doesn't ensure how much of the Cu(I) is produced. It would be highly desirable to use well-defined Cu(I) complexes instead. For this goal, we selected some several copper(I) complexes that have been previsouly described: [Cu(Tpa*)]PF6188 (C1), the commercially available phosphinite CuI{P(OEt)3189 (C2), a carbene containing catalyst [Cu(IAd)2]PF6190 (C3) and a phosphite CuBr{PPh2(OPh-2-OMe)}191 (C4) (figure 63). Figure 63. Representation of the different complexes employed in this section. Complex C1 [Cu(Tpa*)]PF6, was prepared in a dry box by mixing [Cu(NCMe)4]PF6 with 1 equivalent of trispyrazol ligand (Tpa*) (2 h). Figure 64. Synthesis of C1. C3 complex was synthetized through the formation of the carbene ligand IAd from the IAdBF4 salt and tBuONa, in presence of [Cu(NCMe)4]PF6. 188 E. Haldón, M. Delgado-Rebollo, A. Prieto, E. Álvarez, C. Maya, M. C. Nicasio, P. J. Pérez, Inorg. Chem. 2014, 53, 4192. 189 Before testing the phosphinite C2, Bromotris(triphenylphosphonie)copper(I) Cu(PPh3)3Br, and Cu(phen)PPh3Br were tried. However, in conjunction with the azide, it gave an absurd high amount of background fluorescent noise. Likely the Staudinger reduction took place, activating the probe. Then, an oxidized phosphine was tested in order to avoid the Staudinger process. 190 S. Díez-González, S. P. Nolan, Angew. Chem. Int. Ed. 2008, 47, 8881. 191 S. Lal, J. McNally, A. J. P. White, S. Díez-González, Organometallic 2011, 30, 6225. Chapter II 79 Figure 74. Fluorescence calibration of triazole 21 Then, we compared the performance of the two complexes bearing the BTTE ligand (L3), the predefined C5, and the in situ prepared Cu(I)-L3, at a concentration of 50 M (table 2). In fact, Cu(I)-L3 furnished 0.41 M of product, what corresponds to 2.7% yield, an impressive result in comparation to Cai’s protocol, taking into account that we made no efforts to decrease the GSH presence inside the cell. More thought-provoking was the formation of 2.71 M of product when C5 was employed, which is traduced into an astonishing 18% yield. Moreover, this complex still performs well at 25 M, reaching a 10.4% of yield with no apparent cytotoxicity. Catalyst Cat. concentration (M) Product formation (M) Product formation/cell (M/cell) Yield (%) Cu(I)-L3 50 0.41 2.39·10-06 2.7 C5 50 2.71 1.58·10-05 18 C5 25 1.57 9.16·10-06 10.4 Table 2. Yield quantification of CuAAC reaction inside HeLa cells. Protocol: cells were incubated with the chosen amount of Cu(I)-L3 (from CuSO4, L3 and NaAsc preincubation) or C5 (in absence of NaAsc), washed and then incubated with 100  M of azide 20 and 200  M of propargyl alcohol at 37 ºC for 1 h. Extraction of cell content with MeOH and fluorescence experiments led to the quantification of intracellular concentration of 21. It cannot be discarded that the performance of the catalysts (in situ and predefined) could be associated to different cellular uptake. Using ICP-MS experiments we could measure the amount of intracellular copper, which is indicated in figure 75a. These data confirm that the predefined Cu(I) complex C5 was internalized 3 times better than its in situ generated equivalent Cu(I)-L3. The TPP analogue Cu(I)-L4 was also efficiently internalized. Interestingly, in all the cases we observed that the predefined complexes (C1-C4) are better internalized than the in situ made tristriazole complexes using NaAsc (Cu(I)-L). These results, which confirm the good internalization of C4, are in consonance with its good performance in cellular experiments, despite the yields we had observed previously were low. In addition, the ICP analysis also corroborated the intracellular location of Cu as it can be observed in figure 75b. Chapter II 80 Figure 75. a) ICP-MS experiments for the quantification of the cellular uptake of copper catalyst. b) ICP-MS results of the copper accumulated in the membrane of HeLa cells. Protocol: HeLa cells were incubated with 75  M of Cu(I)- L complexes (from CuSO4, L and NaAsc preincubation) and 50  M of C complexes (absence of NaAsc). Of course, key to complete the cellular studies is the study of cytotoxicity of the copper complexes. Remarkably, under our reaction conditions, we observed no substantial cytotoxicity with most of them. We performed the colorimetric MTT (3-(4,5-dimethylthiazol-2-yl)-2,5diphenyl-2H-tetrazolium bromide) assays (figure 76) to evaluate the cytotoxicity of the complexes employed in this study in a range between 25-150 M. We found that the predefined complexes C1-C4 presented toxicity at concentration above 50 M. In contrast, the tristriazoles prepared in situ (Cu(I)-L) were less toxic, even after 12 h. Importantly, cells incubated with 50 M of the most active predefined catalysts, C1 and C5, showed a viability of ca. 85% and 70% after 2h respectively. Satisfactorily, a concentration of 25 M of C5 led to a survival over 80%. Figure 76. a) Cytotoxicity study of the most significant Cu(I) complexes in the range of 50 and 100  M, after 2h. b) Cytotoxicity study at 150  M after 12 h. Chapter II 81 4. Conclusion In conclusion, we have demonstrated that water soluble copper(I) complexes featuring designed ligands can readily enter mammalian cells and promote intracellular CuAAC annulations of small, abiotic and freely diffusible molecules. Our results indicate that using appropriate ligands, it is possible to tune the cell uptake and reactivity of Cu(I) complexes, and importantly, confirm the viability of using discrete copper species to promote efficient CuAAC annulations in the challenging interior of mammalian cells. Indeed, an independently isolated Cu(I)– tris(triazolylmethyl)amine complex, C5, that can be stored without degradation when kept under nitrogen, is capable of promoting the intracellular transformation even in the absence of NaAsc. This new complex displays better cellular uptake and better intracellular reactivity than that observed for the in situ made Cu(I)-L3 complex. This complex is therefore working as an “offthe-shelve” catalyst to promote challenging intermolecular annulations inside mammalian cells. The copper complex C5 circumvents some of the actual limitations of the “in vivo” CuAAC chemistry, since it avoids the use of excess of ligands or the use of reductants such as ascorbate. CHAPTER 3. Ru [2+2+2] CYCLOADDITION IN BIOLOGICAL MEDIA AND MAMMALIAN CELLS This chapter includes work published in Chemical Science as: J. Miguel-Ávila,a M. Tomás-Gamasa,a J. L. Mascareñas,a Angew. Chem. Int. 2020, 59, 17628. aCentro Singular de Investigación en Química Biolóxica e Materiais Moleculares (CIQUS) and Departamento de Química Organica, Universidad de Santiago de Compostela, 15782, Santiago de Compostela, Spain. DOI: 10.1002/anie.202006689, Open Acces articles (Very Important Paper) Chapter III 85 1. Transition metal-promoted cycloadditions As shown in the introduction and in the previous chapter, the queen mothers of bioorthogonal chemistry are cycloaddition reactions, including the strain-derived cycloadditions or the CuACC.192 These reactions have been manly used for bioconjugations, and in some cases for biological interrogation,14,20 but usually not for synthetic purposes, namely, the synthesis of complex structures with potential biological properties. This is surprising, as cycloaddition reactions are extremely appealing form a synthetic point of view, owing to the intrinsic complexity increase they achieve, fact that is already exemplified with the parent Diels-Alder reaction. Figure 77. General scheme for a Diels-Alder reaction. Despite its intrinsic synthetic interest, the Diels-Alder reaction present important limitations, such as the need of specific substrates with tuned electronic properties.193,194 In this context, the development of metal-catalyzed cycloadditions is especially attractive, as it allows a much wider range of reactivity. Coordination of the -electrons of unsaturated systems to transition metals change the electronic properties of the former, and therefore its reactivity, that can engage in different type of reactions, including cycloadditions. Pd, Rh, Ru, Ni, Co and Ir are common metals used in different type of catalytic processes, including formal cycloadditions.195,196,197,198 Importantly, in many cases, these reactions can be tuned by playing with the metal ligands. 192 S. S. Nguyen, J. A. Prescher, Nat. Rev. Chem. 2020, 4, 476. 193 F. A. Carey, R. J. Sundberg, Advanced Organic Chemistry 5th Ed, Springer, 2008. 194 J. Clayden, N. Greeves, S. Warren, Organic Chemistry 2nd Ed, Oxford, 2012. 195 A. Roglans, A. Pla-Quintana, M. Solà, Chem. Rev. 2021, 121, 3, 1894. (for 2+2+2) 196 B. M. Trost, Z. Zuo, J. E. Schultz, Chem. Eur. J. 2020, 26, 15354. (7 membered rings) 197 M. Gulías, A. Collado, B. Trillo, F. López, E. Oñate, M. A. Esteruelas, J. L. Mascareñas, J. Am. Chem. Soc. 2011, 133, 7660. (for 2+2) 198 M. Gulías, F. López, J. L. Mascareñas, Pure Appl. Chem. 2011, 83, 495. (for 3+2) Chapter III 86 Ruthenium, is one of the metals that has demonstrated a better performance to catalyze formal cycloadditions. This is also of particular interest for potential bioorthogonal applications, as it has shown high tolerance in biological environments. Commercially available ruthenium complexes like Cp*Ru(COD)Cl or [CpRu(NCMe)3]PF6, have demonstrated to be very effective to promote different type of synthetically relevant cycloadditions.199,200 Some mentionable examples are collected in figure 78: a) a [2+2] reactionreported by the group of Mitsudo in 1994;201 b) a [4+2] process described by Trost and co-workers in 2000;202 c) a [2+2+2] cycloaddition published by Itoh’s group in 2003.203 Figure 78. Ruthenium catalyzed cycloadditions. a) [2+2] Cycloaddition reported by Mitsudo in 1994.201 b) [4+2] Cycloaddition reported by Trost in 2000.202 c) [2+2+2] Cycloaddition reported by Itoh in 2003.203 This last type of reaction, the [2+2+2] cycloaddition, is highly attractive from the constructive and complexity increase point of view, as from very simple substrates it is possible to build aryl rings, generating very interesting structures. 199 L. Severa, J. Vávra, A. Kohoutová, M. Cízková, T. Sálová, J. Hývl, D. Saman, R. Pohl, L. Adriaenssens, F. Teplý, Tetrahedron Lett. 2009, 50, 4526. 200 J. Palaniraja, S. M. Roopan, Chem. Sci. Rev. Lett. 2014, 3, 93. 201 T. Mitsudo, H. Naruse, T. Kondo, Y. Ozaki, Y. Watanabe, Angew. Chem. Int. Ed. Engl. 1994, 33, 580. 202 B. M. Trost, R. E. Brown, F. D. Toste, J. Am. Chem. Soc. 2000, 122, 5877. 203 Y. Yamamoto, T. Arakawa, K. Itoh, J. Am. Chem. Soc. 2003, 125, 12143. Chapter III 87 The reaction has been proposed to take place through the following general mechanism: the catalytic cycle starts with the coordination of the two alkynes to the Ru(II) complex. Subsequently, and oxidative cycloametallation, which is the rate-determining step, gives a ruthenacyclopentatriene and its tautomeric ruthenacyclopentadiene form (II). After the coordination of the external alkyne (III), subsequent [2+2] cycloaddition yields a ruthenabicycloheptratriene (IV), that quickly suffers a rearrangement, triggering the ring opening and forming a ruthenacycloheptatetraene (V). Finally, a reductive elimination step leads to the formation of the corresponding arene (VI) that is displaced, completing the cycle.195,204 Figure 79. Mechanism for the ruthenium promoted [2+2+2] rerported by Itoh in 2003.203 R = CO2Me. Colors represent electron movement in the different compounds. 204 K. Tanaka, Transition-Metal-Mediated Aromatic Ring Construction, Wiley, 2013. Chapter III 88 2. TM-promoted [2+2+2] cycloadditions in aqueous and biological media: antecedents Alkynes are well-established bioorthogonal reactants, and key partners for instance for the CuAAC. Therefore, they might be used for other type of reactions in biological environments. Considering the synthetic relevance of the above commented metal-promoted [2+2+2] cycloadditions, one could wonder about the viability of performing this transformation in biological media. A few examples that demonstrate their water compatibility have been described, not only with ruthenium,205 but also with rhodium or cobalt catalysts. 206,207,208 However, solubility issues tend to be a problem. The groups of Oshima and Wan demonstrated that using a tris(msulfonatophenyl)-phosphine (tppts) as ligand,209 the reaction is more efficient and can be carried out in mixtures with high water proportion.210,211 Figure 80. TM [2+2+2] reaction performed in water using soluble ligands.210,211 In terms of bioorthogonality, and considering the possibility of carrying out this type of cycloadditions in complex biological media, there is a nice contribution by Teplý and collaborators in 2009. Herein, they reported an intramolecular [2+2+2] cycloaddition of designed triynes in bacteria cell lysates, at room temperature (figure 81).212 This result suggest that the multicomponent cycloaddition can be carried in complex aqueous media, at least in an intramolecular form. 205 V. Cadierno, S. E. García-Garrido, J. Gimeno, J. Am. Chem. Soc. 2006, 128, 15094. 206 L. Yong, H. Butenschön, Chem. Commun. 2002, 23, 2852. 207 B. Heller, B. Sundermann, H. Buschmann, H. Drexler, J. You, U. Holzgrabe, E. Heller, G. Oehme, J. Org. Chem. 2002, 67, 4414. 208 Y. Wang, S. Huang, T. Lin, F. Tsai, Tetrahedron 2010, 66, 7136. 209 Complex preformed in situ at 100 ºC under deoxygenated conditions 210 F. Xu, C. Wang, X. Li, B. Wan, ChemSusChem, 2012, 5, 854. 211 H. Kinoshita, H. Shinokubo, K. Oshima, J. Am. Chem. Soc. 2003, 125, 7784. 212 L. Adriaenssens, L. Severa, J. Vávra, T. Sálová, J. Hývl, M. Cízkova, R. Pohl, D. Saman, F. Teplý, Collect. Czech. Chem. Commun. 2009, 74, 1023. Chapter III 95 Entry Alkyne Ru1 (mol%) Milieu Yield (%)4 11 27 5 DCE 755 2 27 25 H2O 60 3 27 50 H2O 43 4 27 25 PBS 94 5 27 50 PBS 78 6 27 25 DMEM 95 7 27 50 DMEM 95 8 27 25 Lysate2 62 9 27 25 Lysate3 89 10 28 25 H2O traces 11 28 25 PBS 63 Table 3. Milieu scope for the [2+2+2] reaction to obtain phtalanes. Conditions: 1 mM propargyl ether 26, 4 equiv. alkyne 27 or 28, the proper amount of Ru1 in the chosen solvent at r. t. for 2 h. 1Standard conditions used by Itoh. 2Cells lysates 2 mg/mL prepared from water dilution. 3Cells lysates 2 mg/L prepared from PBS dilution. 4Yields were determined by NMR using TMB as internal standard 5Isolated yield. We next explored the performance of other type of ruthenium complexes, which have previously shown to compatible with biological settings. In this sense, we prepared the quinoline containing complexes Ru3 and Ru4,136,139 the bis-allyl-Ru dimer (Ru5), and its acetate-Ru monomer (Ru6).172 We also considered the commercially available [Cp*Ru(NCMe)3]PF6 (Ru7)76. The quinoline-ruthenium complexes were prepared following reported procedures.136 Starting from kynurenic acid, the alcohol is brominated with P2O5 and tetrabutylammonium bromide (TBAB) to give 31 in a 10% of yield. Allylation of the carboxylic acid and coupling with Nmethyl-3-aminopropanol led to ligand 32 in 65% yield. Finally, the complex Ru3 is formed in 70% yield by mixing the ligand 32 with [CpRu(NCMe)3]PF6 Figure 87. Synthesis of Ru3. Chapter III 96 Ru2 can be synthetized from the ligand 32 by coupling it with 4-tryphenylphosphinobutanoic acid (30% of yield), using EDC as activating agent, and its posterior treatment with [CpRu(NCMe)3]PF6, leading to the complex Ru2 in 65% of yield. Figure 88. Synthesis of Ru2. The complex Ru4 was also prepared according to literature protocols.216 3-Amino-4hydroxybenzoic acid is condensed in acid media with acrolein to construct the quinoline core (33) in quantitative yields. Its posterior allylation to give 34, and hydrolysis affords the quinoline acid 35 in good yields. Ru4 is formed by mixing the ligand 35 with [CpRu(NCMe)3]PF6 in 74% yield. Figure 89. Synthesis of Ru4. The bisallyl Ru(IV) complex Ru5 was obtained following a procedure previously developed in the group by mixing RuCl3, isoprene and 2-methoxyethanol (38% yield). Ru6 was synthetized from Ru5 and silver acetate in a 60% of yield.172 The formation of the complexes was corroborated by HPLC-MS. Figure 90. Synthesis of ruthenium (IV) complexes Ru5 and Ru6. 216 T. Völker, E. Meggers, ChemBioChem 2017, 18, 1083. Chapter III 97 We then tested the different complexes under the standard conditions: 1 mM of 26, 4 equivalents of 27 and 25 mol% of the chosen ruthenium catalyst in PBS for 2 h. However, when Ru2 was tested, we didn’t observe the formation of the desired cycloadduct (table 4, entry 1). It is known that this kind of Ru(IV) quinoline-catalysts are activated by nucleophilic addition of exogenous,136 or endogenous nucleophiles,139 such as PhSH or GSH to the allyl ligand, leading to an active Ru(II) species. Therefore, we tested the reaction with Ru2 in presence of different nucleophilic species, such as PhSH, aniline or 2-phenylethanol (PhEtOH), observing the formation of the product in 6, 10 and 14% of yield respectively (entry 3-5). When PhSH and Ru3 were combined, only a 5% of product was obtained (entry 6). We also tested the performance of a ruthenium(II) complex, such as Ru4. In this case, an increase of the yield up to 25% was observed (entry 6). Furthermore, we explored Ru5 and its monomer Ru6, as they were recently applied by the group for redox isomerization processes in HeLa cells. Pleasingly, both furnished over 70% yield with a 50 mol% of catalyst loading (entry 8-10). Entry Ru Cat. loading (mol%) Additive Yield (%)5 11 Ru2 25 - 2 Ru2 25 NaAsc3 - 3 Ru2 25 PhSH4 6 4 Ru2 25 Aniline4 10 5 Ru2 25 PhEtOH4 14 6 Ru3 50 PhEtOH4 5 71 Ru4 25 25 8 Ru5 252 73 9 Ru5 122 59 10 Ru6 50 80 11 Ru7 25 38 Table 4. Catalysts with bioorthogonal potential under study. Conditions: 1 mM of propargyl ether 26, 4 equiv. of 27, the chosen Ru catalyst in PBS for 2 hours. 1Allowed to react for 24 h. 2The amount of ruthenium is the double due the dimeric nature of the complex. 3Concentration of 25  M. 41equiv. of the additive was employed. 5Yields were determined by NMR using TMB as internal standard. All the above results confirm the viability of the intermolecular cycloaddition. However, the product of these reactions is not fluorescent, which made difficult their translation and monitoring in more complex settings, especially under biological and living conditions. Chapter III 98 4.3.1. [2+2+2] Intermolecular cycloadditions that give fluorescent products Anthraquinones (AQ) are fascinating secondary metabolites that some organisms such plants, fungi insects or bacteria are able to synthesize. Similarly to phtalanes, these structures are synthesized from acetylCoA and malonylCoA by the multidomain polyketide synthase, following an elaborated multistep pathway, where the main transformation consist in a consecutive decarboxylative Claisen thioester condensation (figure 91).217 An enormous number of different post-translational modifications can take place, leading to a large family of biomolecules,218 which involves a great variety of bioactive compounds, and importantly, some of them with fluorescent properties.219 Figure 91. Plausible biological pathway for the synthesis of anthraquinones. Remarkably, mammalian cells lack the metabolic machinery to make these polyketide products. Thus, we considered that a fruitful translation of the above [2+2+2] to intracellular environment would provide mammalian cells a way to have access to these metabolites. In 2003, Itoh and co-workers reported a smooth synthesis of anthraquinones (AQ) starting from bispropargylic ketones, 220 based on a a ruthenium-catalyzed [2+2+2] cycloaddition. The reaction takes places in DCE, at 60 mM, using 4 equiv. of the partner alkyne, and 5 mol% of Ru1 (65% yield). Figure 92. Synthesis of anthraquinones reported by Itoh in 2003.220 217 C. Hertweck, Angew. Chem. Int. Ed. 2009, 48, 4688. 218 C. Olano, C. Méndez, J. A. Salas, Nat. Prod. Rep. 2010, 27, 571. 219 G. M. Ziarani, R. Moradi, N. Lashgari, H. G. Kruger, Metal-Free Synthetic Organic Dyes 1st Ed, Elsevier, 2018. 220 Y. Yamamoto, K. Hata, T. Arakawa, K. Itoh, Chem. Commun. 2003, 1290. Chapter III 99 On the basis of this reaction, and considering the cycloaddition studies described in the previous section, we decided to investigate the synthesis of AQs in the presence of biological media. The required diyne precursors were prepared by addition of selected Grignard reagents (ethynylmagnesium bromide or equivalent) to the corresponding dialdehydes, followed by oxidation with Dess-Martin periodinane (DMP) to afford the desired ketones (figure 93). Figure 93. Synthesis of diynes employed in this work. While o-phthalaldehyde is commercially available, the other aldehydes were easily synthesized as indicated in the following scheme: Figure 94. synthesis of aldehyde 33. With the diynes in hand, we tested the cycloaddition chemistry. Employing Itoh's conditions in DCE, with diyne 31 and 1-ethynyl-4-methoxybenzene (27), we observed the formation of the anthraquinone 41, which was isolated in 55% (table 5, entry 1). Importantly, we confirmed that this product exhibits a very good blue fluorescence (λexc = 430 nm, λem = 536 nm). We then moved to aqueous and biologically relevant conditions, using 1 mM concentrations of the diyne 31, 4 equivalents of 27, and TMB as NMR internal standard to determine the yields. As shown in table 5, with 25 mol% of the catalyst Ru1, 41 was obtained in 59% yield in water. This mixture is slightly heterogeneous, mainly because of the low solubility of 31, suggesting that it is an "on water" reaction. With 50 mol% of the catalyst the yield increased to 79%, recovering the unreacted diyne (see experimental section, Chapter III S4). Curiously, reactions performed under stoichiometric conditions suffered an important decrease in yield. Chapter III 100 Entry Ru1 (mol%) Milieu Yield (%)3 11 5 DCE 554 2 5 H2O 6 3 10 H2O 25 4 25 H2O 59 5 50 H2O 79 6 100 H2O 68 7 25 H2O2 27 8 25 H2O/ACN 20 Table 5. Milieu scope for the ruthenium promoted [2+2+2] cycloaddition to obtain anthraquinones 41. Conditions: 1 mM of 31, 4 equiv. of alkyne 27, Ru1, at room temperature for 2 h. 1Itoh standard conditions. 2Deoxygenated solvent. 3Yields were determined by NMR using TMB as internal standard. 4Isolated yield. The transformation tolerates more complex biological media such as PBS, DMEM and cell lysates (65, 37, 28% respectively with 25 mol% of Ru1). These results could be improved by increasing the catalyst loading to 50 mol%, leading to acceptable yields of 59 and 43% when DMEM and HeLa lysates were used, respectively. Entry Ru1 (mol%) Milieu Yield (%)1 1 5 PBS 13 2 10 PBS 23 3 25 PBS 65 4 50 PBS 67 5 25 DMEM 37 6 50 DMEM 59 7 25 Lys 28 8 50 Lys 43 Table 6. Milieu scope for the ruthenium promoted [2+2+2] cycloaddition to obtain anthraquinone 41. Conditions: 1 mM of 31, 4 equiv. of alkyne 27, Ru1 in different milieu at room temperature for 2h (4 hours for lysates). 1Yields were determined by NMR using TMB as internal standard. Chapter III 101 The reaction was also studied along time. As can be notice in figure 95, plateau was reached between 30 min and 1 h, when water, PBS or even DMEM were employed, and independently of the amount of catalyst employed. However, in lysates the process was slower. Figure 95. Experiments along time of the ruthenium promoted [2+2+2] cycloaddition of 31 and 27. Conditions: 1 mM of 31, 4 equiv. of alkyne 27, Ru1 in different milieu at room temperature. Yields were determined by NMR using TMB as internal standard. Lowering the concentration of the diyne 31 led to a decrease in the yield (from 59 to 36%, table 7, entry 1). By using higher catalyst loadings (50 mol%), the yield was raised to 53% (entry 2), although still below the value of 79% obtained at 1 mM. On the other hand, varying the equivalents of the alkyne partner was also translated into a significant yield decay, although with 2 equiv. the yield was still acceptable (entries 5 and 6). The catalyst was not active after 2 h of reaction, as proved by adding additional substrate and observing no further reaction (entry 7). Chapter III 102 Entry Ru1 (mol%) Deviation from standard conditions Milieu Yield (%)1 1 25 31 (0.5 mM) H2O 36 2 50 31 (0.5 mM) H2O 53 3 25 1 equiv. 27 H2O 30 4 50 1 equiv. 27 H2O 32 5 25 2 equiv. 27 H2O 43 6 50 2 equiv. 27 H2O 49 7 50 +1 equiv. 31 after 2 h PBS 63 8 50 +4 equiv. 31 after 2 h PBS 68 9 50 + 50 mol% Ru1 after 2 h PBS 70 10 25 37 ºC PBS 40 11 50 37 ºC PBS 64 Table 7. Effect on the deviation of the standard conditions to the Ru promoted [2+2+2] transformation. Standard conditions: 1 mM of 31, 4 equiv. of 27, 25 or 50 mol% of Ru1, in water or PBS at room temperature for 2 h. 1Yields were determined by NMR using TMB as internal standard. Finally, other ruthenium complexes employed in the studies of section 5.1, were also tested. In general, they were less efficient in PBS, than the parent complex Ru1 (table 8). In particular, Ru2, Ru3 and Ru4 failed to promote the transformation. Entry Ru1 (mol%) Catalyst Yield (%)1 1 50 Ru6 5 2 25 Ru5 12 3 50 Ru5 10 4 25 Ru7 30 5 50 Ru7 33 Table 8. Scope of other Ruthenium catalysts. Conditions: 1 mM of 31, 4 eq of 27, 25 or 50 mol% of the selected Ru catalyst, PBS, at room temperature for 2 h. 1Yields were determined by NMR using TMB as internal standard. Chapter III 103 All-around, these experiments suggest that cellular experiments will require 50 mol% or even up to 1 equiv. of Ru catalysts, because the concentration in that crowded media will be much lower, and the reactions will be performed at physiological temperature (37 ºC). As previously mentioned, the anthraquinone 41 obtained through the ruthenium promoted [2+2+2] cycloaddition displays excellent fluorescent properties, exhibiting a bright emission between 450-750 nm, when excited at 385 nm. However, both the reactants and the ruthenium catalysts don’t present fluorescence at all (figure 96). Hence, this system can be considered as adequate for our cellular experiments, and allowing us to monitor the progress of the reactions in real time using fluorescence microscopy. Figure 96. Absorbance vs fluorescence spectra of the diyene 31 and anthraquinone 41. Absorbance experiments were carried out in water at 100  M and fluorescence at 20  M with excitation at λem = 385 nm. In addition, we decided to extend the scope, and explore some other substrates, either modifying the substituents of the mono or the diyne. When dimethoxy-bearing diynes 37 or the internal diynes 38 were submitted to the reaction conditions, products 42 or 43 were obtained in modest yields. These lower values could be attributed to the deactivation of the dieynes by EDGs, but also to the lower solubility (at least of 38). Indeed, this diyne demonstrated to perform better when acetonitrile was added as co-solvent (table 9, entry 5 and 6). The reaction of trifluoromethylphenylacetylene 28 with 31 afforded the product 44 in yields over 80% (entry 8). Chapter III 104 Entry Diyne Alkyne Ru1 (mol%) Milieu Yield (%)2 1 38 27 25 H2O 24 2 38 27 50 H2O 45 3 38 27 25 PBS 40 4 38 27 50 PBS 48 5 38 27 25 H2O/MeCN1 43 6 38 27 50 H2O/MeCN1 59 7 37 27 50 H2O 40 8 31 28 25 PBS 81 Table 9. Substrate scope for the ruthenium promoted [2+2+2] cycloaddition to obtain anthraquinones. Conditions: 1 mM of the diyne, 4 equiv. of the alkyne, 25-50 mol% of Ru1, in water, PBS or 1a mixture H2O/MeCN (8:2) at r t. for 2 h, without any precaution to avoid oxygen. 2Yields were determined by NMR using TMB as internal standard. Nevertheless, the fluorescent properties of these anthraquinone are not as good as the one shown by product 41, and therefore we didn’t further check their reactivity in cells. Figure 97. Absorbance and fluorescence comparison among diynes 31, 37, 38 and anthraquinones 41, 42, 43. Absorbance experiments were performed at 100  M in water, and fluorescence at 40  M using λ = 385 nm as excitation wavelength. Chapter III 111 The results are summarized in table 10, When the bromo derivative 52 was employed, we were able to observe the product in reasonable yields (around 40%) either with H2O or with an organic co-solvent. However, with the phosphonium derivative, the reaction mixture presented a slightly turbid aspect when ran in pure H2O, and the process was less efficient, recovering most of the starting diyne. Entry Alkyne Milieu Yield (%)1 1 52 H2O 40 2 52 H2O/MeCN (8:2) 42 3 52 H2O/MeOH (8:2) 39 4 52 H2O/MeOH (5:5) 40 5 53 H2O trace 6 53 H2O/MeCN (9:1) 17 7 53 H2O/MeCN (8:2) 47 8 53 H2O/MeCN (5:5) 50 9 53 H2O/MeOH (5:5) 54 Table 10. Co-solvent screening for Ru promoted [2+2+2] cycloaddition to obtain TPP-Anthraquinones. Conditions: 1 mM of the diyne 31, 4 equiv. of the proper alkyne, 25-50 mol% of Ru1, at r.t. for 2 h. 1Yields were determined by NMR using TMB as internal standard. Although the reaction with the TPP-alkyne 53 gave modest yields, likely because of physical solubility problems, we considered that it would be sufficient for its detection inside cells. Consequently, we moved to cellular experiments which were carried using standard protocols with alkyne 53 (150 M diyne 31 (50 M) and Ru2 (50 M)), and allowing the reaction to take place for 4 hours to maximize the formation of product. Afterwards, the cellular content soluble in methanol was extracted and treated for its posterior HPLC-MS detection. Chapter III 112 Gratifyingly, we were able to detect the compound AQ-TPP (55) in these extracts confirming the intracellular formation of the product (figure 106a), although the intensity of the extracted ion chromatogram was quite low, and difficult to quantify. Remarkably, some product in the reaction media could be also appreciated in lower intensity (figure 106b). Figure 106. Extracted ion chromatogram of the formation of 55 in cellular experiments. Protocol: cells were incubated with 50  M Ru1 for 30 min, after the washes steps, cells were incubated with 50  M of 31 and 150  M of 53, at 37 ºC for 4 h. a) Chromatogram of the methanolic extract. b) Chromatogram of the DMEM reaction media. c) Chromatogram of the wash media prior cellular extraction. In view of these results, we decided to check whether non-fluorescent products, like the phthalans, could also be formed and detected in cells by MS following the same approach, the diyne 26 and the alkyne 53 bearing the TPP moiety for the detection, were subjected to the reaction conditions. Accordingly, a representative amount of product 56 was detected in the methanol extract (figure 107a). In this case, we also detected product in the washings steps performed prior to methanol extraction, what could mean that some product can diffuse outside the cell. Chapter III 113 Figure 107. Extracted ion chromatogram of the formation of 56 in cellular experiments. Protocol: cells were incubated with 50  M Ru2 for 30 min, after the washes steps, cells were incubated with 50  M of 26 and 150  M of 53, at 37 ºC for 4 h. a) Chromatogram of the methanolic extract. b) Chromatogram of the DMEM reaction media. c) Chromatogram of the wash media prior cellular extraction. Curiously, in this case when we ran the same experiments in cells using Ru1 instead of Ru2, a higher amount of product 56 was formed inside the cell (figure 108, a). Moreover, the levels of final product detected in the reaction media and wash media, were always very low (figure 108, b and c). Figure 108. Extracted ion chromatogram of the formation of 56 in cellular experiments. Protocol: cells were incubated with 50  M Ru1 for 30 min, after the washes steps, cells were incubated with 50  M of 26 and 150  M of 53, at 37 ºC for 4 h. a) Chromatogram of the methanolic extract. b) Chromatogram of the DMEM reaction media. c) Chromatogram of the wash media prior cellular extraction. Chapter III 114 4.6. Localized ruthenium-promoted intermolecular [2+2+2] cycloadditions in specific organelles In previous work of the group, a Ru2 derivative (Ru8, figure 109) was demonstrated to accumulate in mitochondria, performing its catalytic activity (uncaging reactions) in this organelle.139 This new complex provided a colocalization with the organelle of around 88%.229 Thus, we found appealing to test if this complex is also active in the intracellular [2+2+2] cycloadditions and if it can express its activity in the mitochondria surroundings . We selected the optimized cycloaddition that provides the product AIE-AQ (50). Experiments were performed incubating the cells with of Ru8 (50 M) for 1 h. After two washing steps, cells were treated with 50 M of 31 and 150 M of the triphenylethylene-alkyne 49 for 3 h at 37 ºC. As negative control, some cells were treated with Ru1, that lack the localization tag. In order to confirm the mitochondrial localization of the product, a green fluorescent mitochondrial marker (mitotrackerMT) was also used. As can be observed in figure 105, when Ru1 was employed, the product fluorescence spread through the cell (figure 109, panel A). The absence of overlay with the fluorescent signal of the mitotracker confirms the lack of colocalization. In contrast, transformations that were promoted by Ru8 exhibited a localized fluorescence (figure 109, D), predominantly matching with the mitochondrial dye (figure 109, panel F). 229 0.88 of Mander’s coefficient M1 (385-500 nm) Chapter III 115 Figure 109. Fluorescence micrographies of co-localization experiments in HeLa cells. Protocol: cells were incubated with 50  M Ru1 or Ru8 for 3 h, then were washed and incubated with 50  M of 31 and 150  M of 49, at 37 ºC for 1 h. A and D: Red channel (λexc = 550 nm, λem = 570-690 nm) corresponding to intracellular transformation with cells incubated with 31, 49 and Ru8. B and E: Green channel (λexc = 470 nm, λem = 490-580 nm) corresponding to mitochondrial marker fluorescence. C and F: merging of A and B, and D and E. A, B and C were performed with Ru1 and D, E and F with Ru8. Chapter III 116 In previous works of the group, the cytotoxicity of different Ru catalysts were already analyzed, with Ru2 and Ru8 showing no remarkable cytotoxicity.139 So, we performed an MTT assay to evaluate the cytotoxicity of Ru1, diyne 31, the alkyne 49 and the resulting AIEgen 50 in a range of 10-100 M for 12 h as shown in figure 110. Importantly, no remarkable cytotoxicity was observed with any component below 100 M. At this concentration, a moderate cytotoxicity was noticeable for Ru1, and only a slight decrease in cell viability was displayed for 31, which still kept over 80%. Figure 110. Cell viability of cells treated with 31, 49, 50 and Ru1 compounds between 10-100  M for 12 h. 5. Conclusions As conclusion, our results demonstrate that transition metals can promote challenging [2+2+2] multicomponent cycloadditions in live mammalian cells. The use of abiotic reactants with bidentate metal coordination abilities (diyne) seems instrumental for the success of the process. In addition to provide a cutting-edge addition to the palette of cell-compatible metal-mediated transformations, our work demonstrates the viability of synthesizing complex, biorelevant polycycles inside cells from simple precursors, and therefore goes beyond the more common uncaging or ligation reactions. Our intracellular synthetic strategy allows to generate products that otherwise cannot be delivered to the cell, as well as controlling their spatial distribution by using suitable ruthenium reagents. Indeed, the possibility of generating the desired products in specific subcellular locations, just by changing the targeting characteristics of the reagents, is exciting. While further work to increase the efficiency of the reactions is needed, our discoveries should further foster this young field of intracellular metal catalysis, and trigger important applications in chemical and synthetic biology, and in biomedicine. Chapter 4: PHOTOREDOX CATALYSIS IN LIVE MAMAMALIAN CELLS Chapter IV 119 1. Introduction into visible-light photochemistry Photochemical transformations rely on the use of light to trigger reactions that otherwise cannot take place under normal thermal conditions (figure 111a). In some cases, light promoted reactions can be carried out with the assistance of catalysts, that are called photocatalysts, (PC) that absorb light and transfer the energy to a substrate, that is not able to absorb the light “per se” (or requires a higher energetic light source). Thus, they act as intermediates between light and substrates, in a process called “sensitization”, that finishes with the excited substrate, in a higher energy position, and ready to trigger the corresponding reaction. So, a good photocatalyst must be not only a good light absorber but also be able to state in the excited state without suffering important changes in its structure. In figure 111b, it can be observed the internal electronic reorganization that suffers the photocatalyst before its quenching, which can be radiative, non-radiative or undergo a reaction.230,231 Figure 111. a) Simplified representation of potential energy surface along the course of the reaction. In red the photochemical pathway, in blue the thermal pathway. B) Simplified representation of the potential energy surface for a photocatalyst electronic rearrangement. Since the late 2000s there has been a drastic increase in the number of contributions in the field of photocatalysis and nowadays a huge amount of photocatalysts can be found, from cheap and green organic photocatalysts to organometallic complexes, mainly based on Ru and Ir metals.232 230 B. König, Chemical Photocatalysis 2nd Ed, De Gruyter, 2020. 231 J. K. McCusker, Chem. Soc. Rev. 2016, 45, 5803. 232 M. H. Shaw, J. Twilton, D. W. C. MacMillan, J. Org. Chem. 2016, 81, 6898. Chapter IV 120 Figure 112. Compendium of common photocatalysts. Upper par; some of the most common organic photocatalysts.233 Below; some of the most common organometallic photocatalysts.234 Within the plausible bimolecular quenching of the PC by a chemical reactant, we can distinguish two different mechanisms. One of them is known as photoredox catalysis, and relies on the concept that the excited triplet state of the catalyst can perform single electron transfer processes (SET),235,236 leading to high energy radical intermediates (figure 113). The other mechanism, far less explored, is named energy transfer (Et), which is based in the direct transference of the energy gathered in the excited triplet state (3PC1). In this process, PC will act as energy donor and the corresponding substrate as energy acceptor, leading to the excited substrate after molecular orbitals overlapping, and electron exchange (Dexter mechanism).237 Figure 113. Oxidative and reductive quenching cycle for a PC. D: electron donor. A: electron acceptor. 233 N. A. Romero, D. A. Nicewicz, Chem. Rev. 2016, 116, 10075. 234 K. Teegardin, J. I. Day, J. Chan, J. Weaver, Org. Process Res. Dev. 2016, 20, 1156. 235 J. W. Tucker, C. R. J. Stephenson, J. Org. Chem. 2012, 77, 1617. 236 C. K. Prier, D. A. Rankic, D. W. C. MacMillan, Chem. Rev. 2013, 113, 5322. 237 F. Strieth-Kalthoff, M. J. James, M. Teders, L. Pitzer, F. Glorius, Chem. Soc. Rev. 2018, 47, 7190. Chapter IV 127 In 2018, Ward and Wenger reported a concurrent photoredox and enzymatic catalytic process for the obtention of enantioenriched amines using a water-soluble Ir photocatalyst coupled to the bacteria machinery that recycles the non-desired enantiomer, leading to a high enantiomeric excess after several cycles.271 Very recently, a number of photocatalytic processes to generate reactive intermediates for protein mapping have been reported. Most of them rely on the generation of carbenes or nitrenes.272,273 271 X. Guo, Y. Okamoto, M. R. Schreier, T. R. Ward, O. S. Wenger, Chem. Sci. 2018, 9, 5052. 272 J. B. Geri, J. V. Oakley, T. Reyes-Robles, T. Wang, S. J. McCarver, C. H. White, F. P. RodriguezRivera, Dann L. Parker, Jr, E. C. Hett, O. O. Fadeyi, R. C. Oslund, D. W. C. MacMillan, Science 2020, 367, 1091. 273 H. Wang, Y. Zhang, K. Zeng, J. Qiang, Y. Cao, Y. Li, Y. Fang, Y. Zhang, Y. Chen, JACS Au 2021, 1, 1066. Chapter IV 128 5. Objectives There is clearly an increasing interest in the use of photocatalytic processes for bioorthogonal chemistry and for biological purposes. However, the number of applications, especially in mammalian cells, is very small, and in particular, the use of photocatalytic processes for synthetic purposes in biological and cellular media is essentially unknown. Considering the previous results of the photo-Meerwein transformation carried out in water by the Xiao’s group, we questioned whether a similar photoredox process could be performed in the presence of complex biological mixtures. We hypothesized that the diazonium reagent could be more efficiently activated than other biological components, ensuring selectivity, in native settings. The resulting carbon radical might react with an appropriately designed acceptor, like a quinone, more efficiently that with other components, which would ensure bioorthogonality. Chapter IV 129 7. Results and discussion 7.1. Photocatalyzed arylations of unsaturated substrates in aqueous media In order to investigate if this aryl radical chemistry can be performed in biological complex environments, we considered the use of quinones as coupling partners owing to their well-known radical trapping properties.274 Although König had previously reported the arylation of benzoquinone with diazonium salts in DMSO, (90% yield, 2 h),250 we decided to study the arylation of naphthoquinones because of their interesting properties as secondary metabolites and their relation with anthraquinones.275 We explored the ability of 1,4-naphtoquinone 57 to act as a radical trapping agent of anisole diazonium salt 58 (table 11) producing the naphtoquinone 59a. Reactions were carried out weighing the solids in Schlenk flasks. After the purge with Ar, the corresponding deoxygenated solvent is added, gently stirred and set under irradiation for the chosen reaction time with fan refrigeration. After extraction, yields were quantified by NMR using trimethoxybenzene as internal standard. The formation of the product of the double addition (59b) was also detected. Delightfully, initial experiments at 200 mM of diazonium salt 58 and 5 equivalent of 57 in DMSO as solvent, reflected the excellent behavior of compound 57 as radical scavenger, observing full conversion of the diazonium salt 58, after 2 h of reaction under blue light source and using 1 mol% Ru(bpy)3(PF6)2, from now on Ru(bpy)3, as PC (68% of the monoaddition product 59a, entry 1). Eosin Y was also tested as PC in a 7.5 mol% and white-light as well, yielding 80% yield (entry 2). Gratefully, the reaction also tolerates water as reaction media, obtaining 50% of 59a in a mixture DMSO/H2O (8:2). Nevertheless, higher amount of water led to precipitation of substrate 57. Water solubility studies demonstrated that a combination of THF/H2O 1:1 provided a homogeneous crude, and an excellent 85% yield. Moreover, under more complex biological media such as PBS the reactivity was maintained (74%, entries 6 and 7). 274 O. Fónagy, E. Szabó-Bárdos, O. Horváth, J. Photochem. Photobiol. A. 2021, 407, 113057. 275 P. Babula, V. Adam, L. Havel, R. Kizek, Curr. Pharm. Anal. 2009, 5, 47. Chapter IV 130 Entry Catalyst Cat. loading (mol%) Milieu Yield (%)1 1 Ru(bpy)3 1 DMSO 68/322 2 Eosin Y 7.5 DMSO 80/202 3 Ru(bpy)3 1 DMSO/H2O 1:1 50/8 4 Ru(bpy)3 1 DMSO/H2O 8:2 40/traces 5 Ru(bpy)3 1 DMSO/PBS 1:1 21/traces 6 Ru(bpy)3 1 THF/H2O 1:1 85/15 7 Ru(bpy)3 1 THF/PBS 1:1 74/24 Table 11. Preliminary experiments for the naphthoquinone arylation. Conditions: 200 mM of 58, 5 equiv. of 57, Ru(bpy)3 at the chosen concentration under 32 W blue-light, in different solvents at room temperature for 2 h. whitelight was employed. 1Yields were determined by NMR using nitromethane as internal standard. 2Isolated yield. In contrast, under more diluted conditions such as 10 mM (table 12), the use of THF was not beneficial, leading to dirty reaction crude and low yields. However, at this concentration the reaction took place satisfactorily in a mixture 1:1 DMSO/H2O (73% yield of 59a, entry 3) and DMSO/PBS (70% yield of 59a, entry 6). Unfortunately, in the absence of a cosolvent yields were poor in both, water and PBS solvents (below 20%), mainly due to solubility issues of quinone 57 (entries 7 and 8). Entry Cat. loading (mol%) Milieu Yield (%)3 1 5 THF/H2O 1:1 32 2 5 THF/PBS 1:1 35 3 5 DMSO/H2O 1:1 73 41 5 DMSO/H2O 1:1 34 5 5 DMSO/H2O 8:2 38 6 5 DMSO/PBS 1:1 70 7 5 H2O 20 8 5 PBS 16 92 10 H2O 19 Table 12. Preliminary experiments for the naphthoquinone arylation. Conditions: 10 mM of 58, 5 equiv. of 57, Ru(bpy)3 at the chosen concentration under 32 W blue-light irradiation, in different solvents at room temperature for 2 h. 11 equiv. of 57 was employed. 2Reaction at 5 mM of 58. 3Yields were determined by NMR using nitromethane as internal standard. Chapter IV 131 In view of these promising results and considering the excellent fluorescent properties of AIE derived anthraquinones developed in Chapter III, we explored the possibility of synthetizing a fluorescent product based on this skeleton. We anticipated that the addition of an aryl radical into triphenylethylene would be feasible, leading to the formation of a well-known AIEgen (figure 122). Figure 122. Our proposal for the formation of tetraphenylethylene. The addition of an aryl radical into diphenylethylene was already reported by König using ethyl2-(4-bromophenyl)acetate as source of aryl radical in DMSO, using his “conPET” strategy. This strategy allows the catalyst to accumulate two photons of different wavelength through a first excitation step, then reduction and a second excitation, emulating the photosynthesis pathway.276 They obtained a mixture of saturated and unsaturated product (2.8:1) in good yields (64%, figure 123).277 Figure 123. König's reported version of aryl addition into diphenylethylene.277 Initial experiments were performed at 200 mM of 58 with 5 equiv. of triphenylethylene, using 1 mol% of ruthenium complex in DMSO. The tetraphenyl product 60 was isolated in 54% yield after 2 h. When a mixture of DMSO/PBS 9:1 was used, a 27% of final compound was obtained. However, only traces were detected in a 1:1 mixture of DMSO/PBS. Despite these are preliminary results, they already underscore the potential of this approach, confirming a significative reactivity in aqueous media or buffered solutions, what stresses the viability of performing bioorthogonal photochemistry. Figure 124. Diazonium salt addition into triphenylethylene. 276 I. Ghosh, T. Ghosh, J. I. Bardagi, B. König, Science 2014, 346, 725. 277 I. Ghosh, B. König, Angew. Chem. Int. Ed. 2016, 55, 7676. Chapter IV 132 7.2. Photocatalytic formal-cycloadditions in aqueous and biological settings Considering our results in the previous sections on metal-catalyzed intermolecular cycloadditions we wondered whether it would be possible using photoredox catalysis, to make aqueous and biocompatible formal cycloadditions. Towards this objective, we took as reference a formal [4+2] benzannulation of biaryldiazonium salts with alkynes, reported by the group of Zhou in organic solvents.278 This reaction likely proceeds by reduction of diazonium salt by the excited photocatalyst, and posterior addition into the alkyne. They performed the reaction at 330 mM of 2-biphenyldiazonium salt 61, adding 5 equiv. of methyl propiolate (or 3 equiv. if diazonium salt is added fractionated), under visible-light irradiation in acetonitrile for 12 h, provides the final product 62 in 74% yield (figure 125). The process excels displaying a high quantum yield of 0.35, which lead us to presuppose some selectivity in complex environments.279 Figure 125. Standard conditions for [4+2] photo-cycloaddition developed by Zhou and co-workers in 2012.278 We considered that if this reaction works in water, it could be adapted to bioorthogonal environments as the radical chain that leads to the cycloadduct should occur in preference to other secondary processes. The mechanism can be seen as an expansion of the one proposed by König for the intermolecular couplings. In this case, the key radical intermediate II generated by the radical addition into the alkyne, would undergo a ring closing evolving to the new radical intermediate III. The stabilized radical would be oxidized, regenerating the PC and leading to the intermediate IV. A final re-aromatization step would provide the corresponding phenanthrene product. Figure 126. Proposed mechanism by the group of Zhou.278 278 T. Xiao, X. Dong, Y. Tang, L. Zhou, Adv. Synth. Catal. 2012, 354, 3195. 279 M. Majek, F. Filace, A. Jacobi von Wangelin, Beilstein J. Org. Chem. 2014, 10, 981. Chapter IV 133 In our hands, the replication of Zhou’s conditions for the reaction of 2-biphenyldiazonium salt (61) with methyl propiolate in organic solvents to yield methylcarboxylatephenanthrene (62), provided almost the same results when 5 equiv. of methyl propiolate were employed (70%, table 13, entry 1). The solvent must be deoxygenated to avoid the quenching of the photocatalyst (entry 2). However, this behavior might have no consequences inside cells, due to their reductant and controlled environment. Since the methyl propiolate is a liquid, the previous established protocol had to be adapted as follows. In the Schlenk flask, the solids were weighted, purged with Ar, and dissolved with the chosen deoxygenated solvent to a final concentration of 330 mM of 61 and 1 mol% of Eosin Y. At this point the mixture is gently stirred and 5 equivalents of the oil-liquid alkyne methyl propiolate is added and set under irradiation and fan refrigeration for the selected reaction time, and finally quantified by NMR using nitromethane as internal standard. The reaction was successfully performed in PBS, using 20% of DMSO as cosolvent to obtain a homogeneous mixture, obtaining the product in 66% of yield after 12 h (entry 4), although after only 30 min a big amount of precipitate corresponding to the product was observed. Gratifyingly, the reaction run in pure PBS led to a 62% yield, thus confirming that a cosolvent is not needed. When non deoxygenated PBS is employed only a 10% of decrease was observed in contrast to MeCN (entry 2, 6), in consonance with a lower oxygen concentration in H2O than MeCN.280 Using a mixture of PBS/DMSO 8:2 as a solvent, a precipitate was quickly formed in 30 min. Furthermore, in the case of pure PBS, a biphasic mixture was observed in a few seconds, suggesting that the activation of the diazonium salt is drastically faster in water than MeCN, under these conditions. On this wise, we could decrease the reaction time from 12 to 2 h, reaching the same results in a robust way (entries 7 and 8). In further experiments we will study the kinetics of the reaction with more detail. 280 M. Quaranta, M. Murkovic, I. Klimant, Analyst, 2013, 138, 6243 Chapter IV 134 Entry Milieu Variation Time (h) Yield (%)1 1 MeCN 12 702 2 MeCN Non-deoxygenated 12 15 3 MeCN 24 62 4 PBS 20% DMSO 12 66 5 PBS 12 62 6 PBS Non-deoxygenated 12 51 7 PBS 20% DMSO 2 66 8 PBS 2 62 9 Neat 12 34 10 Neat 2 37 Table 13. Solvent and reaction time screening. Conditions:330 mM of diazonium salt 61, 5 equiv. of methyl propiolate, EosinYNa2 1 mol%, under 32 W white light irradiation, in MeCN, PBS or in neat at room temperature for 2, 12 or 24 h. 1Yields were determined by NMR using nitromethane as internal standard. 2Isolated yield. In view of the results reported, we investigated other biocompatible photocatalysts. Specially, Ru(bpy)3 was selected, due the considerable higher excited state lifetime, in comparison with Eosin Y.233,234 Besides, as mentioned along the introduction, this type of bipyridine or phenantroline ruthenium catalysts have been employed in uncaging experiments reporting good results even in cells. Initial results in MeCN as solvent showed similar catalytic activity of Ru(bpy)3, with only a slight decrease in the yield (table 14, entry 1). However, carrying out the reaction in water or PBS resulted in excellent yields up to 84%. Indeed, this catalyst showed better performance than Eosin Y under these conditions after 2 h (entries 4 and 5). Additionally, Riboflavin, an intracellular biomolecule with photoredox properties, was tested to check its ability to catalyze the reaction. This catalyst has been already studied in both organic photocatalysis and bioorthogonal applications.281,282 But in this case, only a small amount of product was detected under blue irradiation (entry 6). 281 S. A. Castro, E. Ruggiero, A. Ruiz-de-Angulo, E. Rezabal, J. C. Mareque-Rivas, X. Lopez, F. LópezGallego, L. Salassa, Chem. Sci. 2017, 8, 4619. 282 J. Gurruchaga-Pereda, V. Martínez-Martínez, E. Rezabal, X. Lopez, C. Garino, F. Mancin, A. L. Cortajarena, L. Salassa, ACS Catal. 2020, 10, 187. Chapter IV 135 Some iridium complexes have been also employed for protein modification.272,273 In this regard, we tested the commercially available tris[2-phenylpyridinate-C2,N]iridium (III) (Ir(ppy)3, but only 19% of yield was obtained (entry 7). Rhodamine and TMRE has been also reported by Chen and co-workers for alcohol unmasking inside mammalian cells.265 Nevertheless, TMRE wasn’t able to promote the cycloaddition and Rhodamine B led to a 20% of 61 (entries 8-9). Entry Catalyst Milieu Light Variations Yield (%)1 1 Ru(bpy)3 MeCN Blue 66 2 Ru(bpy)3 MeCN Blue PC 0.5 mol% 56 3 Ru(bpy)3 PBS Blue 24 h 75 4 Ru(bpy)3 PBS Blue 84 5 Ru(bpy)3 H2O Blue 84 6 Riboflavin MeCN Blue 11 7 Ir(ppy)3 MeCN Blue 19 8 Rhodamine B PBS White 20 9 TMRE PBS White traces Table 14. Catalyst screening. Conditions: 330 mM of diazonium salt 61, 5 equiv. of alkyne, 1 mol% of the chosen PC, under 32 W light irradiation, in MeCN or PBS at room temperature for 2h. 1Yields were determined by NMR using nitromethane as internal standard. With these results in hand, we made a quick analysis of the scope using some selected alkynes and Ru(bpy)3 as PC in MeCN or PBS as solvent. The use of the internal alkyne diphenylacetylene led to a modest isolated yield of 27% of 63 in MeCN, but no reaction in PBS was observed, consequence of the poor solubility. Similar, reaction with TMS protected acetylene gave 50% of yield in MeCN of the deprotected phenanthrene 64, but failed in PBS. In contrast, with the use of the water-soluble propiolic acid the reaction achieves the formation of product 65 in an excellent 82% yield in PBS, in absence of an organic cosolvent. The use of other alkynes with strong EWGs such as 1-ethynyl-4-fluorobenzene and ethynyltrifluorotoluene (28) provided compounds 66 and 67 in 60% and 82% yield, respectively, which is in concordance with a better radical stabilization after the aryl radical addition. Chapter IV 136 Figure 127. Alkyne scope. Conditions: 330 mM of diazonium salt 61, 5 equiv. of alkyne, 1 mol% Ru(bpy)3, under 32 W blue light irradiation, in MeCN (a) or PBS (b) at room temperature for 2h. Isolated yields. Along the study of the reaction, several control experiments were concurrently performed. As expected, with no light irradiation no reaction took place, even when no special precautions in order to avoid environmental light were applied (table 15, entries 1-4). When reactions were performed under light irradiation without a catalyst, moderate yields could be obtained. Depending on the reaction conditions, the yield could drastically vary. For instance, the use of ACN as solvent (entries 5-7), or green as light source (entry 9), only provided traces of product. However, a 50% yield was obtained in reaction performed in H2O (entry 13) or without any control of the temperature (entry 11). Thankfully, under standard conditions in PBS and blue light this problem was more subtle, obtaining around 24% of yield at room temperature (entry 17) or 36% when the reaction was performed at a physiological temperature of 37 ºC (entry 21). Entry Catalyst Milieu Light Variations Yield (%)2 1 PBS Dark1 - 2 PBS - 3 Ru(bpy)3 PBS Dark1 - 4 Ru(bpy)3 PBS - 5 MeCN Blue trace 6 MeCN Green trace 7 MeCN White trace 8 PBS Blue 24 9 PBS Green trace 10 PBS White 36 11 PBS Blue No fan 52 12 PBS Blue 37 ºC 36 13 H2O Blue 48 Table 15. Solvent and reaction time screening. Standard conditions: 330 mM of diazonium salt 61, 5 equiv. of alkyne, 1 mol% Ru(bpy)3 if used, under 32 W light irradiation in ACN, PBS or water at room temperature for 2h. 1Covered with black polyethylene plastic. 2Yields were determined by NMR using nitromethane as internal standard. Chapter IV 143 8. Conclusions As conclusion, our results exemplified the first example of a photocatalyzed Meerwein transformation in live mammalian cells. We show that a rationalized design of the reaction components allows to generate aryl radicals, that can be efficiently trapped in a chemoselective manner to generate aromatic cycloadducts in a smooth manner, in presence of biomolecules, biological media, or cells without disturb the reaction. LC-MS experiments verify the intracellular and photochemical generation of the phenanthrene 62, while no product was detected in absence of light or PC. We are aware that additional cellular experiments must be done to fully characterize the transformation, yet, we are excited with the results obtained, and the group is investigating the potential of photocatalysis in biological settings. OVERALL CONCLUSIONS Overall conclusions 147 This PhD work has aimed at the development of formal cycloaddition reactions promoted by transition metals or by photochemical means that can be carried out in aqueous and biological media, and even in the interior of live mammalian cells. General conclusions: 1) We have demonstrated that well-defined water-soluble copper (I) complexes featuring designed ligands, can be efficiently internalized into mammalian cells and promote intracellular CuAAC of designed abiotic substrates. Remarkably, we have proven that predefined Cu(I) complexes are also able to promote the intracellular transformation, even at lower catalyst loadings, while avoiding the use of exogenous reductants such as ascorbate, and hence, improving the biocompatibility of the methodology. 2) We have described the first multicomponent [2+2+2] formal cycloaddition in live mammalian cells. The combination of ruthenium as transition metal catalyst and designed diynes or triynes with adequate coordination abilities, ensure the formation of key ruthenacycles that can be trapped by external alkynes even inside mammalian cells. Importantly, we have used the technology to build polyketide-derived secondary metabolites, from simple precursors, inside cells that lack the corresponding biological pathway to make them. This strategy goes beyond the most common uncaging or ligations reactions, providing for one of the more complex abiotic reactions so far carried out in live settings. 3) We have also used the methodology to build products showing aggregation-induced emission (AIE) properties, which overcome the concentration quenching of fluorescence characteristic of classical fluorophores. By definition, this type of compounds is highly insoluble, and hence experience a poor cellular uptake. We have proved that Ru promoted [2+2+2] cycloaddition can intracellularly generate the AIE product, that otherwise cannot be delivered into the cell. In addition, we have demonstrated that the use of tailored ruthenium reagents provides a spatial control over the product generation. 4) We have presented the first example of an intracellular photoredox formal cycloaddition, specifically a visible light-driven Meerwein reaction. We show that a rationalized design of the reaction components allows to generate aryl radicals, that can be efficiently trapped in a chemoselective manner to generate aromatic cycloadducts in a smooth manner. Further research to broaden the scope of the transformation to include novel functional groups and fluorescent products or the use of irradiation of higher wavelength, is highly desirable for future applications in native cellular environments. RESUMO DA TESE DOCTORAL Resumo 151 A química bioortogonal naceu recentemente como unha ferramenta para a investigación celular, revolucionando campos próximos como a microscopia, o desenvolvemento de fármacos ou a biotecnoloxía debido á súa capacidade para manipular compostos esóxenos dentro dos sistemas biolóxicos sen perturbar o metabolismo do hóspede. A pesar do seu éxito, o descubrimento de novas reaccións bioortogonais segue sendo un reto hoxe en día, debido a natureza das mesmas: os compoñentes deben ser inocuos e solubles en medios biolóxicos, facilmente interiorizables, reactivos en condicións fisiolóxicas e capaces de interactuar rápida e especificamente nun ambiente tan complexo como os medios biolóxicos. O capítulo I serve de introdución ao campo da química bioortogonal e a súa evolución ao longo dos anos, poñendo en contexto os logros acadados paralelamente ao desenvolvemento desta tese de doutoramento. En primeiro lugar, preséntase un breve resumo das contribucións da química biolóxica, tales como a síntese de sondas, ligazón nativa de péptidos ou a evolución dirixida de proteínas, para a continuación facer énfase no uso de reaccións bioortogonais como ferramenta de bioconxugación. Finalmente, introdúcense as transformacións máis influentes neste campo, diferenciando claramente as reaccións realizadas na superficie das células ou as que se desenvolven no interior das células, o que supón un gran reto científico. A data de comezo da química bioortogonal establécese entre os anos 1980 e 1980, co traballo do grupo de Rideout baseado na condensación de hidracinas, previa acumulación dos substratos no interior da célula. Non obstante, non se falou oficialmente da bioortogonalidade ata 2003, ano no que o grupo de Bertozzi, que co paso do tempo se convertería nun referente no campo da química bioortogonal na superficie das células, acuñou o termo por primeira vez. Curiosamente, a condensación de hidracinas viu o seu verdadeiro potencial no campo das glicoproteínas presentes na membrana celular, polo que os azucres modificados con cetonas poderían bioconxugarse facilmente en ausencia de aldehidos ou outras cetonas no exterior da célula. Actualmente, unha das reaccións que mellor cumpre os requisitos da química bioortogonal é a redución de Staudinger, unha transformación que foi redescuberta por Bertozzi no ano 2000, quen a utilizou para levar a cabo procesos de bioconxugación. Ademais, esta reacción supuxo un punto de partida para o uso de sondas intelixentes. Estas sondas son compostos cuxa fluorescencia está enmascarada pola presenza dun grupo funcional cun par de electróns libres que poden participar nunha transferencia de electróns intramoleculares promovida por luz (PET). Esta transformación utilizouse con gran éxito na superficies celulares, aínda que a súa aplicación no interior das células permaneceu esquiva. Resumo 152 No seguinte apartado preséntase o concepto de química "click" e a súa relación coa química bioortogonal. Coméntanse as súas características tales como: la estabilidade dos substratos, modularidade e velocidade da reacción, condicións de reacción suaves, e incluso en condicións fisiolóxicas. Neste campo, as cicloadicións 1,3 – dipolar entre azidas – alquino para dar triazoles destacan sobre todas as demais, especialmente a versión catalizada polo cobre (I). Esta versión desenvolta por Sharpless, Fokin e Medal en 2002, e trasladada posteriormente ao campo da química biolóxica en 2003 polo grupo de Tirrell, converteuse na reacción bioortogonal por excelencia, coa mellor cinética, substratos estables, esóxenos, pero sobre todo simples e pequenos, facilmente sintetizables e incorporables á maquinaria biolóxica. Por todo isto, no capítulo II ponse o foco na versión promovida polos catalizadores de cobre, tamén coñecida como CuAAC (cicloadición catalizada por cobre de azidas e alquinos). A sección de bioconxugación remata presentando os exemplos máis recentes da química bioortogonal: as reaccións de cicloadición Diels - Alder con demanda inversa de electróns (IEDDAC) e os procesos fotoclick. Ambas transformacións presentan unha cinética comparable á cicloadición entre azida-alquino catalizada por cobre e, polo tanto, teñen o potencial necesario para levar a cabo as reaccións a concentracións moi baixas. Desafortunadamente, nestes procesos tamén se producen reaccións laterais, o que limita as súas posibles aplicacións. A continuación, entra en escena a química organometálica, destacando o seu achegamento á química en auga xunto coa súa recente contribución á química biolóxica, cos traballo dos grupos de Schultz e Davis, ou a conxugación de proteínas mediante acoplamentos tipo Suzuki. Por último, discútese o gran impacto da química organometálica na expansión da química bioortogonal. Neste apartado faise énfase nas técnicas de liberación de moléculas activas, destacando o uso de catalizadores de metais de transición como rutenio, paladio ou mesmo ouro, para a desprotección de grupos alilo, alilcarbamato ou propargilo. A contribución do grupo de Meggers en 2006, que conseguiu levar a cabo estas reaccións en células vivas de mamíferos, foi pioneira e promotora deste tipo de procesos. Non obstante, o foco principal desta sección sobre a química organometálica reside no inexplorado campo da química bioortogonal sintética que, a día de hoxe, ten un número moi reducido de aplicacións en ambientes celulares. A conclusión inmediata que se extrae deste capítulo é que a química bioortogonal é unha rama moi nova da química biolóxica cun gran potencial aínda por explorar. Por exemplo, os procesos bioortogonais baseados no uso de metais de transición apenas foron estudados, probablemente debido á crenza de que son incompatibles coa auga ou tóxicos para o ambiente celular ou, en definitiva, para a vida. É certo que estas reaccións presentan eficiencias menores que as das transformacións utilizadas para a bioconxugación, pero a existencia da CuAAC, xunto cun amplío abanico de reactividade da química organometálica, pon de manifesto o enorme potencial do uso Experimental section: Chapter IV 255 Figure S2. Crude NMR with 61 and methyl propiolate in presence of cells (1·106 cells/mL) S4 DIAZONIUM SALT STABILITY ASSAYS Stability of diazonium salt 61 was tested in different solvents (DMSO, H2O and PBS) at room temperature, following its decomposition by 1H NMR. For studies in DMSO or H2O, NMR tubes were prepared by dissolving 10 mg of the diazonium salt 61 (0.037 mmol) and 2.1 mg of trimethoxybenzene (DMSO, 0.012 mmol ,0.33 equiv.) or 2.4 L of dimethyl sulfone (H2O and PBS, 0.037 mmol, 1 equiv.) as internal standard, in 1mL of the corresponding deuterated solvent. For PBS experiments, a sample of 1 mL of non-deuterated PBS was lyophilized and diluted with 1 mL of D2O to obtain deuterated PBS. In figure S3 are summarized the stability results after 30 min or 60 min. Diazonium salt did not suffer degradation in DMSO or H2O under 30 min. Nevertheless, a 15% or 20% of degradation was observed in phosphate buffer, after 30 or 60 min, respectively. This behavior could be attributed to the exchange of the counterion BF4. Figure S3. Evolution of diazonium salt 61 in different solvents. 100 97 92 87 100 97 86 100 86 81 72 30 60 0 20 40 60 80 100 Remaining 61 (%) Time (min) DMSO H2O PBS Experimental section: Chapter IV 256 4.1 Representative NMR spectra of the evolution of diazonium salt 61 in solution Figure S4. NMR spectra of 61 in DMSO at time “0” using TMB as internal standard (1H:1H). Figure S5. NMR spectra of 61 in DMSO after 1h using TMB as internal standard (1H:1H). Experimental section: Chapter IV 257 S5.- VIABILITY ASSAYS The toxicity was analyzed by MTT assays in HeLa cell line. A comparison study of the different compounds was carried out. MTT assay: 15000 cells per well were seeded in 96-well plates two days before treatment with different concentrations of Ru(bpy)3 (10-100 µM), substrates or product (10-100 µM). The effect of the intracellular generation of product was also evaluated. Moreover, the viability of the cells after the irradiation was also tested. After 6, 12 or 24 h of incubation, HEPES containing 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) was added to a final concentration of 0.5 mg/mL. Cells were then incubated for 4 h to allow the formation of formazan precipitates by metabolically active cells. A detergent solution of 10% SDS and 0.01 M HCl was then added and the plate was incubated overnight at 37 °C to allow the solubilization of the precipitates. The quantity of formazan in each well (directly proportional to the number of viable cells) was measured by recording changes in absorbance at 570 nm in a microtiter plate reading spectrophotometer (Tecan Infinite 200 PRO). 6.- PHOTO-MEERWEIN ARYLATION INSIDE LIVING CELLS For the identification of 62 generated in the photo-promoted reaction inside HeLa, A549 or Vero cells, a total of 12 x 106 cells growing in 8 plates of 100 mm were used. The experiments were performed in plates of 100 mm as follows: 2000000 cells per well were seeded in 100 mm plated two days before treatment. For each measurement, eight plates were used. Cells were incubated with catalyst Ru(bpy)3 (50 µM) and diazonium salt 61 (100 M) and propargyl alcohol (100 µM) for 30 min. Prior to irradiation, the cells were washed twice with fresh DMEM and then treated with HEPES. Well plates containing cells were disposed as shown in illustration 2. Irradiated at 15 cm of distance under fan refrigeration for 30 min. Illustration 2. Montage for cell irradiation. Experimental section: Chapter IV 258 Afterwards, the reaction media was collected for analysis in a 50 mL Falcon. Prior to extraction, cells were washed with 3 mL of DMEM, followed by 3 mL of PBS and the washings were also collected separately in two 50 mL Falcons. Then the cell monolayer was treated with 1 mL of MeOH. After 5 min and pipetting up this solution was transferred to a 15 mL Falcon. Finally, we obtained 8 mL of methanolic extracts from the eight plates employed. All the samples were lyophilized for 3 days and dissolved in CH3CN until reaching a theoretical concentration of 250 µM. For the quantification of the product, the obtained samples (250 µM in CH3CN) were centrifuged at 13500 rpm for 15 minutes and the supernatant was collected. In the case of the methanolic extract, it was diluted 1:4 using CH3CN / H2O (6:4). However, in the case of the samples of the reaction media and washing steps, no dilution was required. Each sample was injected in a Bruker Elute coupled with time TOF using a column Zorba eclipse BXDC18 2.1 x 10 mm 1.8 µm and a flow rate of 0.5 mL/min at room temperature. For the solvent system, initial conditions H2O / CH3CN (40:60) were used for 1 min and followed by a gradual change over 7 min to CH3CN. Then, initial conditions were recovered in a gradual change over 0.2 min, and maintained for 1.8 min. S7. REFERENCES [1] B. Xing, C. Ni, J. Hu, Angew. Chem. Int. Ed. 2018, 57, 9896. [2] M. T. Molina, C. Navarro, A. Moreno, A. G. Csáky, Org. Lett. 2009, 11, 4938. [3] H. Jiang, Q. Xu, Z. Wang, J. Of Supercritical Fluids, 2009, 49, 377. [4] T. Xiao, X. Dong, Y. Tang, L. Zhou, Adv. Synth. Catal. 2012, 354, 3195. [5] C. Wang, S. Rakshit, F. Glorius, J. Am. Chem. Soc. 2010, 132, 14006. [6] M. Tobisu, K. Yamakawa, T. Shimasaki, N. Chatani, Chem. Commun. 2011, 47, 2946. [7] K. Nemoto, H. Yoshida, N. Egusa, N. Morohashi, T. Hattori, J. Org. Chem. 2010, 75, 7855. [8] D. García-Cuadrado, P. de Mendoza, A. A. C. Braga, F Maseras, A. M. Echavarren, J. Am. Chem. Soc. 2007, 129, 6880. [9] I. J. S. Fairlamb, J. D. Firth, Org. Lett. 2020, 22, 7057. LIST OF PUBLICATIONS List of publications 262 Title Discrete Cu(I) complexes for azide-alkyne annulations of small molecules inside mammalian cells Authors Joan Miguel Ávila, María Tomás Gamasa, Andrea Olmos Verge, Pedro J. Pérez, José Luis Mascareñas Cid. Year 2018 Journal Chemical Science Volume, pages 9, 1947-1952 Impact Factor 9.825 PhD student contribution Synthesis of probes, the majority of catalysts and its characterization. Experiments in plate reader and HPLC, UV-Fluorescence, EPR. Cells data treatment (no cell handling or HPLC cell analysis) Title Intracellular Ruthenium-Promoted (2+2+2) Cycloadditions Authors Joan Miguel Ávila, María Tomás Gamasa, José Luis Mascareñas Cid. Year 2020 Journal Angewandte Chemie International Edition Volume, pages 59, 17628-17633 Impact Factor 15.34 PhD student contribution Synthesis of probes, the majority of catalysts and its characterization. Reaction studies, UV-Fluorescence. Cells data treatment (no cell handling or HPLC cell analysis).