At its completion in 2003, the Human Genome Project (1) was saluted as the tool to finally cure every cancer. Two decades later, this largely anticipated promise has yet to be delivered and the community of cancer researchers, which was once disproportionally focused on the central dogma of molecular biology, now embraces more holistic views. One of the most exciting frontiers of cancer research deals with understanding homeostatic processes during cancer development and how cancer cells respond to environmental stresses of a chemical and physical nature. In this context, the catabolic activity of autophagy is the key mechanism to maintain the balance between synthesis, degradation, and recycling of cytosolic components (2). These routine housekeeping functions provide a cellular mechanism to preserve homeostasis, enhance resilience to stresses and promote cell survival. Several environmental stresses activate autophagy, among those hypoxia, DNA damage, inflammation, and metabolic challenges such as starvation. Aside from challenges of chemical nature, cells are also exposed to stresses of mechanical nature, arising from environmental cues. Sensing of mechanical stress (mechanosensing) is mediated by force-induced conformational changes of mechanosensitive proteins directly or indirectly connected to the cytoskeleton and by mechanically activated ion channels (3, 4). Mechanosensing results in a modification of intracellular tension through reorganization of cytoskeletal and actomyosin contraction, which, in turn, integrate the mechanical signals into biochemical cascades (mechanotransduction), and, at longer time scale, lead to modification of gene expression (4). Hence, the physical properties of the microenvironment, such as extracellular matrix composition, stiffness, and architecture, have a profound impact on cellular genotype, phenotype, processes, tissue organization and overall biological function of the organism (4). This relation between mechanics and biological responses is also important during cancer transformation and progression, where the specific physical microenvironment of the tumor cells undergoes dramatic changes. These modifications of the tumor microniche are driven by enhanced cell contractility, increased pressure resulting from abnormal cell proliferation and growth of tumor mass, and alterations of composition, architecture and rheological properties of the surrounding extracellular matrix (5, 6). It has been reported that these mechanical changes correlate with activation of autophagy, which may be part of an integrated response to mechanical stresses employed by cancer cells to escape programmed cell death and to facilitate their adaptative response to the new mechanical environment (7). Furthermore, compelling evidence have suggested that autophagy impact several cancer hallmarks including cell motility and invasion, cancer stem cell viability and differentiation, epithelial to mesenchymal transition (EMT), resistance to apoptosis and anoikis, escape from immune surveillance and tumor cell dormancy (7, 8). However, the causative relation between cellular mechanics and autophagy and their interdependent role in cancer transformation are fragmentary and largely anecdotical. Here, we aim to review the autophagic process using a mechanical perspective and explore the crosstalk between mechanotransduction and cellular catabolism in order to access their possible contribution to cancer transformation and survival.

Vital functions of eukaryotic cells such as resistance to deformation, control of cellular shape, migration and transport of intracellular cargos depend on the activity of the cytoskeleton, an interconnected network of filamentous polymers, motor proteins and regulatory proteins (9). This network is composed by three interdependent structural components, namely microtubules, intermediate filaments, and microfilaments (actin) which are the engine of the cells as they convert chemical energy into mechanical energy via ATP-dependent polymerization and action of motor proteins. This mechanical energy is used to produce forces that displace cellular elements (e.g. formation of cellular protrusion, transport of cargos) and/or store elastic energy therein (e.g. cortical tension, cellular contractility). The whole process of autophagy being a sequence of membrane remodeling events is mechanically accomplished and coordinated by ATP-dependent cytoskeletal dynamics that lead to mechanical deformation and transport (10, 11). The cytoskeleton acts as an important framework for the modulation and control of correct positioning, tethering, docking, priming, fusion, and movement of organelles, such as autophagosomes and lysosomes. Actin cytoskeleton is composed by actin filaments and fibers whose assembly and disassembly generate web-like networks (Arp2/3-mediated branching) and bundles (formin-dependent crosslinking of filaments). These networks and bundles structurally support cellular membranes and determine their dynamics (12). Importantly, the action of molecular motors of the myosin family puts actin filaments under tension. Similar to a stretch coil, the release of this tension produces kinetic energy used for vesicle transport and membrane remodeling associated to autophagosome formation (13, 14). In addition, some myosins [i.e. myosin VI (15)] are directly involved in the transport of various cargos including autophagosomes (15). Furthermore, microtubules dynamics of polymerization and depolymerization and the action of associated motor proteins [i.e. kinesin and dyneins (16, 17)] orchestrate the movement of pre-autophagosomal structures and autophagosomes across the cytoplasm during the process of autophagosome maturation (18, 19) and autolysosome bidirectional transport (20). The cooperation and competition between actin and microtubules are responsible for a large part of cellular mechanics. Together, these ATP-dependent cytoskeletal processes provide the mechanisms to overcome the energy barriers imposed by membrane elasticity and resistance to deformation that affect each step of the autophagic process (21). Finally, intermediate filaments (i.e. keratins and vimentin), which do not have evident dynamics and lack motor proteins, are thought to provide mechanical stability to the cell and its organelles (22). Intermediate filaments play a key role in autophagosome and lysosome positioning by providing a resistance to their free, unregulated movement (23). For instance, networks of vimentin cables have been observed to form cages around cellular organelles including the nucleus, endoplasmic reticulum, and mitochondria (24). Consistently with this regulatory function, pharmacological disruption of the vimentin network results in defective flow of the autophagic process (autophagic flux), the perinuclear position of autophagic vesicles and a loss of their region-specific localization at different stages of the process (23).

A great variety of biophysical stimuli elicit cellular responses and determine cellular functions ( Figure 3A ) (111). Much of these stimuli stem from short-scale ( Figure 3A , blue boxes) interaction between the cells and their physicochemical microenvironment. Cells have been reported to sense and respond to the a) physical status of the extracellular matrix (e.g. composition, stiffness, topography and density) by exerting traction forces on the substrate (112–114), b) geometrical cues (e.g. size, confinement, curvature) affecting cortical and membrane tension (115, 116), c) presence of surrounding cells (e.g. cell crowding) and their physical activity (e.g. pulling and pushing causing cell-cell shear and normal forces), and the chemical composition of the interstitial and luminal fluids (e.g. osmotic pressure inducing cell swelling or shrinkage and consequent variation in membrane tension) (117). In addition, cells are subject and respond to large-scale mechanical forces ( Figure 3A , red boxes) such as shear stresses and fluid pressure due to flow of liquids or solid material in the lumen of tubular structures (e.g. gut, blood vessels and urinary tract), and lateral stretch and compression of tissues required for physiological function of lungs, muscles, and digestive system, among others (118–121). Short and long-scale force (summarized in Figure 3A ) elicit reactive and adaptive cellular responses that primely involve active processes mediated by the cell cytoskeleton (4). This can be activated by the direct action of external mechanical cues on the cytoskeleton via sensing mechanisms involving various mechanosensors at the cell surface, such as mechanically activated ion channels (e.g. TRP and piezo), proteins sensing tension and curvature of the plasma membrane (e.g. BAR proteins) and of the cytoskeleton (e.g. filamin), and adhesion protein complexes (e.g. focal adhesion, adherens junctions) (122). These mechanosensors translate mechanical inputs into biochemical signals via mechanotransducers (AKA mechanotransduction process) that control cytoskeletal organization, membrane trafficking, gene expression profile and ultimately cellular function as a whole (4, 123, 124) ( Figure 3B ). Mechanosensing is generally achieved by a force-dependent conformational change of the sensing protein that may lead to the opening of a channel (typically calcium channels), which subsequently activates a cellular response via an electrochemical signal, or through the dissociation of proteins (mechanotransducer) from the sensing complex. In its freely diffusive form, the mechanotransducer participates in enzymatic reactions (e.g. phosphorylation), either as the enzyme or the substrate. As examples of both cases, the rise of calcium and/or the activation of protein phosphorylation cascades will lead to short- and long-term adaptation to mechanical stimuli. Acting as an essential part of the innate adaptive mechanisms of the cell, the autophagic response aids in the management of mechanical challenges and allows the cell to adapt to the everchanging physical environment ( Figure 3B ). In general terms, mechanical cues may affect the autophagic process in two ways: firstly, via specific crosstalk between mechanotransduction and autophagy regulatory proteins (e.g. mTORC, AMPK) (125, 126) responsible for the initiation and/or inhibition of the autophagic process and/or secondly, via the unspecific cooperation/competition mechanisms between mechanical processes and autophagy to recruit cytoskeletal elements and phospholipid membranes (127). A growing body of evidence demonstrates that indeed mechanical cues feed into the signaling required for the activation of autophagy (7, 128–130). Conversely, despite being highly plausible, the competition for cellular components between the two processes and the consequences of such, have still to be addressed by the literature. A final point of convergence is the regulatory role of autophagic catabolism and recycling of biological components in managing the turnover of cellular components necessary for proper execution of mechanical processes. In the following sections, we will discuss the crosstalk and interactions between cell mechanics and autophagy. Next, we will discuss in detail some the most relevant and better known connection between the mechanotransduction machinery and the autophagic process.

Malignant transformation is accompanied by a progressive loss of tissue homeostasis and perturbations of tissue architecture. It is widely recognized that a critical component of this transformation involves alterations in the mechanical phenotype of the cell and of the surrounding microenvironment, creating a peculiar mechanical milieu predominantly composed of cancer cells surrounded by a dense extracellular matrix (6, 195, 196). In addition, a set of accessory cells may be found in the tumor microenvironment, including blood and lymphatic vascular cells, lymphocytes, inflammatory cells and cancer associated fibroblasts (6, 195, 197). Depending on the context and stage of cancer development, autophagy has been recognized as a “double-edged sword” ( Figure 4 , left panel), as it can act as a mechanism for either tumor-suppression or tumor-promotion depending on the cellular context in which it acts (198, 199). Consistent with its role in promoting cell survival and rejuvenating cellular components, autophagy serves as a quality-control mechanism, detecting changes in organelle architecture and protein folding and thus preventing tumor initiation. On the other hand, these same mechanisms promote cancer cell survival and escape from apoptosis. This occurs by aiding the responses against environmental stress and generating the energy needed for unregulated growth and metastasis through the recycling and degradation of cellular organelles ( Figure 4 ) (198, 200–202).

In the context of solid tumors, several mechanical aspects of the tumor microenvironment contribute to the tumor-promoting function of autophagy ( Figure 4 , right panel). When confined by the extracellular matrix, cancer spheroids experience forces exerted by the expanding tumor mass as a result of unchecked proliferation and the resistance to deformation of the surrounding stromal tissue (203, 204). This causes increased interstitial pressure (203) and generates shear stress within the tumor microenvironment (7, 205, 206). Eventually, this mechanical stress affect cell growth directly, by compressing cancer cells, and/or indirectly, by compressing the surrounding blood and lymphatic vessels (207). Due to the sustained compression of the vasculature within the tumor, poor tissue perfusion causes hypoxia and eventually necrosis within the tumor (208). Hypoxia promotes epithelial to mesenchymal transition (EMT), a reorganization of the cytoskeleton and dissolution of the epithelial cell-cell junctions. This enables dynamic cell elongation, directional motility (209), and consequently an increase in the metastatic potential of the tumor cells. Furthermore, during EMT the composition of intermediate filaments changes, switching from keratin to vimentin (195, 210). Furthermore, during this process the actin cytoskeleton becomes hypercontractile through the TGF-β-dependent activation of pathways such as Rho GTPases, p38MAPK and ERK1/2 (211). This pathway activation triggers actin reorganization and formation of cellular protrusions, including lamellipodia and filopodia (212, 213). Furthermore, TGF-β and hypoxia also promote the formation of cancer associated fibroblasts (214, 215), which interact with each cellular component of the tumor microenvironment.

By mediating extracellular matrix stiffness, the cancer associated fibroblasts can regulate the cancer cell cytoskeleton (216–218). These changes to the cellular cytoskeleton during transformation or EMT drastically alters their mechanical phenotype and in particular the degree of tension exerted on neighboring cells and the extracellular matrix, leading to increased migration, invasion and dissemination potential (201, 219). To successfully metastasize, tumor cells migrate locally and invade into surrounding tissue to gain vasculature access, and subsequently intravasate through the basal membrane and detach from the extracellular matrix to become circulating tumor cells (220, 221). Ultimately, circulating tumor cells that survive in circulation can extravasate from the bloodstream and engraft in secondary tissue sites and thus forming metastatic foci (222). All cells that travel in the bloodstream experience fluctuating levels of shear stress. Hemodynamic shear stress, caused by the movement of blood along the cell surface, is influenced by both the fluid viscosity and fluid flow velocity (210). Shear stress can also be caused by frictional interaction with endothelial cells (6, 223). Equally, tumor cells within the bloodstream must survive harsh conditions, including extracellular matrix detachment-induced apoptosis (i.e. anoikis), immune system assault, along with the variations in shear stress (222). The physiological shear stress (0.5–3 Pa) caused by blood flow may suppress cancer cell proliferation but may also promote migration and adhesion (224–227). Substantial evidence suggests that mechanical stress, such as compressive and shear forces, in the tumor milieu boosts malignant progression by inducing autophagy (129, 130, 172, 228). Consistent with this, cervical cancer cells exposed to pulses of laminar shear stress of 2 Pa (over 3 and 6 minutes) undergo autophagy, by a lipid raft-mediated p38MAPK dependent process, and delay apoptotic cell death (130). However, shear stress is not necessarily beneficial to the cancer cell. Conversely, elevated levels of shear stress (6 Pa), as occurs during intense exercise, has been shown to promote tumor cell death (229). Furthermore, fluid shear stress in the range of 0.05 to 1.2 Pa is shown to trigger cancer cell death through apoptosis and autophagy in several cancer cell lines, including hepatocarcinoma, osteosarcoma, oral squamous carcinoma, and carcinomic alveolar basal epithelia. Interestingly, this fluid shear stress induced death did not occur in non-cancerous cells (230). Thus, it appears that depending on both the intensity and the duration of the shear stress, autophagy may act as either a pro- or anti-survival mechanism. Furthermore, it has been shown that autophagy induced by shear or compressive stress plays a role in cytoskeletal remodeling and in the recycling of proteins essential for cancer progression (149, 172). Indeed, increased tissue stiffness is implicated in the control of several tumor features, such as growth, invasion, and metastasis (203, 231, 232). Accordingly, it has been observed that the stiffness of cancerous tissue of breast, hepatic and liver origin is higher than that of the corresponding respective physiological context (233–236). Extracellular matrix stiffening in tumors is produced as consequence of stroma reorganization, through the excessive activity of extracellular matrix proteins and enzymes that covalently cross-link collagen fibers and other extracellular matrix components (237, 238). Collagen crosslinking enhances integrin activation, focal adhesion maturation, intracellular contraction and thus causes a subsequent increase in the stiffness of the actin cytoskeleton, which may favor cancer cell migration and invasion (214, 239–241). The higher extracellular matrix stiffness also plays a role in the onset of the malignant phenotype: cytoskeletal tension leads to increased cell- extracellular matrix adhesions and disruption of cell-cell junctions (242). Enhanced collagen deposition in the extracellular matrix leads to activation of the Hippo signaling pathway (159, 162, 243) with a consequent loss of contact inhibition. Autophagy is also reported to have a pivotal role at the center of these processes. Autophagy is reported to be compromised in contact-inhibited cells in both 2D or 3D-soft extracellular matrix cultures. In such cells, YAP/TAZ (previously mentioned to be regulated by mechanical forces) fail to co-transcriptionally regulate the expression of myosin-II genes, resulting in the loss of F-actin stress fibers, which leads to impairment in autophagosome formation. This loss of F-actin stress fibers is also associated with a reduction in the number of ATG16L1 puncta per cell and with decreased co-localization of ATG9A-LC3, suggesting an alteration in the trafficking of key autophagy proteins and thus a defective autophagic response (244). Furthermore, compressive stress-induced autophagy can promote secretion of matrix metalloproteinase-2 and the turnover of the focal adhesion paxillin, boosting the invasiveness of the HeLa cervical cancer cell line (129). In line with these results, it has been suggested that paxillin binds directly to LC3 to stimulate focal adhesion disassembly in MDA-MB-231 human breast cancer and in B16.F10 mouse melanoma cell lines, and furthermore promote metastasis in vivo in the 4T1 mouse mammary tumor model (149). Another mechanism of force sensing in cancer involves filamin A. This actin and actin-integrins crosslinker is down-regulated in human bladder cancer, reducing autophagy in these cancer cells, as indicated by the decrease in the levels of LC3-II and decrease in LC3-I (245). It has been further reported that upon overexpression, filamin A attenuates autophagy and suppresses the invasive ability in cancer cells. The mechanism of action may involve the inhibition of matrix metalloproteinases expression, regulation of integrin function and enhances apoptosis (245–247). Interestingly, YAP/TAZ signaling has been shown to stimulate filamin A transcription to maintain actin anchoring and crosslinking under mechanical tension (248). This could be a potential mechanism for cancer cells, to control autophagy through a crosstalk between YAP/TAZ and cytoskeletal elements. Low mechanical stress has been shown to activate Caveolin-1, triggering the FAK/Src and ROCK/p-MLC pathways, which are involved in the reorganization of the cytoskeleton, cell motility, focal adhesion dynamics and breast cancer cell adhesion (227). PI3K/AKT activation and β-Catenin-TCF/LEF-dependent activity downstream from Caveolin-1 also correlates to increased VEGF expression and thus greater angiogenic potential of tumor (249). Shear stress-induced Caveolin-1 activation can induce PI3K/AKT/mTOR signaling and metalloprotease activity, which have been shown to promote cell motility and metastasis of breast carcinoma cells (250). Conversely, it was determined that phosphorylated Caveolin-1 functions to activate autophagy through binding to the Beclin-1/VPS34 complex under oxidative stress and to protect against ischemic damage (251). These data suggest that Caveolin-1 function might be cell-context dependent (252), resulting in different autophagic outcomes. Interestingly, similarly to autophagy, Caveolin-1 has also been implicated both in tumor suppression and progression (253, 254). Although potentially protective in bourgeoning tumors, higher levels of either Caveolin-1 mRNA or protein have been reported in varying cancers strongly correlating with poor survival in advanced cancer patients (227). This further implies that Caveolin-1 has a role in the metastatic process, as evidenced by increased migration, invasion and anchorage-independent growth (255). Furthermore, recent studies have unveiled the existence of an interplay between the primary cilium and autophagy in the regulation of cancer development and progression (256–258). In addition to being considered as a survival mechanism in tumorigenesis, excessive accumulation of autophagosomes may induce autophagic cell death or apoptosis (259–262), which, in the context of cancer, limits tumor growth and spread. Recently, Wang and collaborators showed that acute shear stress (10 Pa for 60 min) promotes autophagosome accumulation, which is accompanied by increased fusion of autophagic vesicles with multivesicular bodies, and reduction of autophagosome-lysosome fusion, in HeLa and MDA-MB-231 cell lines (263). Furthermore, the inhibition of autophagosome degradation, induced by mechanical stress, is associated with increased release of autophagic components in extracellular nanovesicles, possibly through a Ca²⁺-dependent pathway involving autophagy, multivesicular bodies and exosomes (263). Thus, exosome secretion might provide a supplementary pathway to maintain cellular homeostasis when the autophagy pathway is damaged or insufficient to degrade large amounts of damaged proteins and prevent cell death (263). In conditions of mechanical stress, these results suggest a possible crosstalk between degradative and secretory autophagy to maintain cellular homeostasis and tumor cell survival (264). Furthermore, following pathological stress, harmful nucleic acids, molecular chaperones, cytosolic proteins, and misfolded proteins are released into the extracellular space through exosomes and may contribute to tumor progression and metastasis (265, 266).

As we have highlighted in the previous sections, the relation between cell mechanics and autophagy goes two ways. In the context of cancer, autophagy regulates multiple metastasis-related signaling pathways associated with cell mechanics depending on cell type and tumor microenvironment. Autophagic protein LC3-II mediates the targeted degradation of focal adhesion proteins such as Src and paxillin (149, 267) to promote focal adhesion disassembly and turnover and lead to cell migration. Furthermore, integrins can be differently recycled and degraded, depending on their conformation, activation by ECM proteins, and binding of effector proteins, such as TLNs and FERMTs/kindlins. Integrins trafficking, recycling and degradation affect their availability at the plasma membrane, focal adhesion dynamics and Rho GTPase-mediated cytoskeleton remodeling to facilitate cell motility (127, 268). While nutrient starvation (269) increases integrin internalization and ECM degradation (270), hypoxia (271) promotes recycling of specific integrin. Since nutrient starvation and hypoxia are both hallmarks of tumor microenvironment, further investigation into how autophagy regulates integrin trafficking may provide insight into the overall role of autophagy in cancer metastasis and lead to the understanding of how microenvironmental stress act on cell mechanics to induce cancer cell exit from the primary tumor.

In the last decade a plethora of new treatments has been introduced that significantly improved the survival of cancer patients. Despite this, highly aggressive cancers often develop primary or acquired resistance that finally cannot be treated. To this aim, new therapeutic approaches are required to overcome drug resistance and improve treatment response. Based on the reviewed literature, considering the “mechanobiology of autophagy” might represent a novel and promising approach. Indeed, even if, to the best of our knowledge, to date there are no drugs approved by the FDA or currently being investigated in clinical trials that consider the mechanobiology of autophagy in their approach there are examples of proteins/pathways that are modulated by mechanical forces thus affecting autophagy.

As previously mentioned, a pathway that is activated in cancer cells, which is regulated by mechanical forces, is the Hippo–YAP/TAZ pathway, whose inhibition has been shown promising results in reducing therapy resistance [for recent reviews see (272, 273)]. Interestingly, blockade of this pathway also reduces autophagy (244), which is targeted by several drugs currently under investigation in clinical trials, suggesting that dual inhibition of YAP-TAZ pathway and autophagy could improve treatment response. Mechanics also regulate epidermal growth factor receptor (EGFR), a protein that is regularly amplified or mutated in glioblastomas, and where autophagy is enhanced promoting cell survival (274). Inhibition of autophagy, in addition to radiotherapy, already showed positive results which might be further improved by considering the mechanics of cancer. Consistently, inhibition of Janus-associated kinase (JAK) by Ruxolitinib, a drug currently used in myeloproliferative neoplasms which inhibits cell contractility, preventing signaling downstream of focal adhesions, has recently been shown to induce autophagy (275), thus a combination of ruxolitinib with pharmacological inhibitors of autophagy needs to be followed for cancer treatment. Another drug that is currently being studied in clinical trials is losartan, an angiotensin II receptor blocker, which reduces intratumoral interstitial fluid pressure in solid tumors (276). Interestingly, this drug has also been shown to inhibit autophagy promoting autophagic cell death in cancer cells (277), again confirming the importance of targeting autophagy and mechanobiology in cancers.

In recent years, autophagy has emerged as one of the key regulators of cellular, tissue, and organism homeostasis. Vibrant research in this field has brought to light the intricacies of autophagy’s molecular machinery, together with its biochemical regulation and biomedical consequences associated with its impairment. The complex mechanobiology regulating cellular mechanical and biochemical processes is also a bourgeoning field. In this review, we hoped to bring to light the role of physical forces in autophagy regulation and their potential implications in both physiological as well as pathological conditions. More importantly, we hoped to raise questions to help investigate the mechanical requirements of autophagy and appreciate the extent to which mechanical signals affect this process. For instance, a diet rich in saturated fatty acids can negatively impact autophagic flux in neurons (278, 279). Interestingly, as the steric conformation of these phospholipids is known to mechanically decrease membrane bending, it could consequently impair autophagy by preventing vesicle fusion. However, the mechanical role of phospholipids is largely overlooked in the literature and it could represent an important area for future investigation. Similarly, to provide new frontiers for exploration, areas worthy of investigation are the action of cytoskeletal dynamics, the mechanical interplay between cellular processes, and the role of environmental cues. To achieve this a paradigm shift is required, one that adopts modern interdisciplinary approaches combining cell biology, physics, and engineering (280). To this end, cutting-edge techniques such as superresolution microscopy and the control of the mechanochemical environment (281) (e.g. by incorporating biomimetic substrates and microfluidics) will open exciting opportunities and perspectives. Combined, these technological and conceptual new directions will lead to a better understanding of autophagy and mechanisms onsetting related diseases, which in turn would pave the way to the identification of new pharmacological targets.

MH-C, LM, PL, AR, and CB prepared the figures. JP and FP investigated the intersection of autophagy and cell mechanics in the literature by data mining (Cytoscape). MH-C, LM, JP, FP, PL, PA, GO, EM, AR, and CB reviewed the literature. MH-C, LM, GO, EM, AC, AR, and CB wrote the manuscript. GO, EM, AR, and CB proofed the manuscript. AR and CB conceptualized and supervised the work. All authors contributed to the article and approved the submitted version.

AR and CB are supported by ANID PIA192015; EM and AC by ANID PIA 172066; GO by CONICYT FONDAP-15130011, IMII P09/016-F, and FONDECYT 1180241; EM by FONDECYT 1200499; AC by FONDECYT 1171075 and FONDAP no 15130011. AR acknowledges PUENTE 004/2019 and 012/2020 from the Pontificia Universidad Católica de Chile. LM, PL, JP, and FP acknowledge IPREint20 from the Pontificia Universidad Católica de Chile.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.