On Nanomachines and Their Future Perspectives in Biomedicine

Nano/micromotors are a class of active matter that can self‐propel converting different types of input energy into kinetic energy. The huge efforts that are made in this field over the last years result in remarkable advances. Specifically, a high number of publications have dealt with biomedical applications that these motors may offer. From the first attempts in 2D cell cultures, the research has evolved to tissue and in vivo experimentation, where motors show promising results. In this Perspective, an overview over the evolution of motors with focus on bio‐relevant environments is provided. Then, a discussion on the advances and challenges is presented, and eventually some remarks and perspectives of the field are outlined.


Introduction
Motion is an essential function in nature across the different length scales. At the nanoscale, a fascinating example is the movement of kinesins, cytosolic proteins that "walk" along the cell microtubules transporting diverse cargo. Motion is crucial at cellular lever, for instance during cell division, fertilization, or cargo transport inside cells. Mimicking motion at the nanoscale was also envisioned in science-fiction, as exemplified in the well-known novels Fantastic voyage and Prey, or in comics and movies, e.g., in the Marvel universe. Scientists tried to reproduce motion at the nanoscale using synthetic devices, which are called nanomachines, nanoswimmers, or nanomotors.
Nanomotors are a concept of active matter that can selfnavigate environments outperforming Brownian motion by converting any type of energy into kinetic energy. The basic principle of the nanoscale motion requires gaining sufficient kinetic energy to overcome the viscous drag force of the fluid and the Brownian (random) motion of the particle suspended in a fluid. To do so, nanomotors' motion must be nonreciprocal Motors can use physical stimuli to gain fuel-free locomotion (Figure 1). In this regard, magnetic fields, light, and ultrasounds were preferred due to their easy tuneability, implementation, and often high speeds achieved by the motors. [18] Magnetic Fields: Motors displaying a magnetic functionality can respond to external magnetic fields that favor the locomotion (Figure 1a-i). [19] Usually, rotating or oscillating magnetic fields are employed, since they offer more possibilities than magnetic field gradients. However, the design of motors that respond to rotating/oscillating magnetic fields is rather complex. The main key parameter of magnetic motors is their magnetophoretic mobility (ξ), which broadly speaking depends on the magnetic material (magnetic susceptibility), the hydrodynamic radius of the motor, and the dynamic viscosity of the environment. In short, the "perfect" magnetic motor that aims to navigate in complex biological media such as serum, blood, or urine, should contain a magnetic material that actuates strongly and fast to a magnetic field (i.e., high saturating magnetization and an almost inexistent coercive field), and have a relatively small size.
Magnetic motors are a versatile option due to the variability in motor morphologies and the tuneability of the magnetic fields, reaching top velocities up to 2600 µm s −1 . [20] The earliest examples of magnetic motors that navigated in low viscous environments aimed for bacteria-like (or screw-like) shapes that responded to rotating magnetic fields for propulsion (Figure 1a-ii). [21][22][23][24] Nevertheless, spherical motors were also considered, where the tuning of the magnetic field together with geometry constrains of the motors allowed for intelligent mechanisms of locomotion. [25][26][27][28] Recent examples considered magnetic motors symmetrically or asymmetrically coated with DNA origami bundles. [29] Upon applying a rotating field, only Scheme 1. Evolution of nano/micromotors in biomedicine. Advances have been shown in the media where the motors navigate from simple, lowviscosity environments to more complex milieus including cells and tissue. In parallel, there has been a great development in both the ways to produce motion and in the diversity of envisioned applications. References to key pioneering examples are indicated by the superscripts a, [2] b, [3] c, [4] d, [5] e, [6] f, [7] g, [8] h, [9] i, [10] and j. [11] the asymmetrically coated motors showed straight trajectories, which responded to the type of the coating rather than the characteristics of the magnetic field. The authors suggested that the DNA bundles acted as flexible appendages that created thrust, similar to the whip-like motion of sperm flagella. More advanced designs were achieved upon incorporation of soft or porous materials to mimic the motion of microorganisms [30] or fish. [31] Additionally, the magnetic functionality has been often incorporated to other kinds of motors, allowing for magnetic separation of the motors from crowded media, [32,33] external guidance, [34][35][36][37] or heat delivery. [10] Light Sources: Light can trigger chemical reactions that result in enhanced locomotion (Figure 1b-i). Typically, light initiates chemical reactions on the surface of the motor that produce a gradient of uncharged molecules and/or ions, and thus allows for self-diffusiophoresis or self-electrophoresis. [38] The first examples reported the use of UV light. However, this highenergy light is likely harmful for living organisms and does not have enough tissue penetration, limiting their bio-related applications. In this context, near-infrared (NIR) and visible light are more suitable. Alternatively, photothermal effects have also been reported to induce the motor locomotion. In this case, the light does not trigger a chemical reaction, but leads to the release of heat in the close proximity of the motor, which creates a gradient of temperature and hence particle motion. In all cases when using light as a stimulus, the electron band gap of the material interacting with the light source is a core aspect to consider. This key parameter affects the photochemical reactions on the surface of the motor, and materials with a lowerenergy band gap are typically preferred.
Together with magnetically propelled motors, light is a ubiquitous source that can penetrate tissue (e.g., NIR light) and thus, pose applicability in nanomedicine. Light-triggered motors were first explored in aqueous solutions, and usually relied on NIR or UV light sources. These types of motors have achieved top velocities up to 25 µm s −1 (Figure 1b-ii). [39] For example, Xuan et al. reported nanobottle-shaped carbonaceous motors that could propel under the actuation of NIR light by a thermophoretic mechanism. [40] These motors showed a programmable behavior when subjected to ON-and-OFF cycles where the NIR light was applied or removed, displaying always similar velocity profiles. A more advanced design was proposed by Xing and co-workers, who made a multi-responsive motor that responded to both NIR light and (bio)chemical fuels. [41] The motor consisted of a Janus (asymmetric) particle made of dendritic mesoporous silica-carbon, where one of its hemispheres was decorated with platinum nanoparticles. The motors moved with top velocities of ≈22 µm s −1 due to a thermal gradient in its proximity induced by the presence of NIR responsive carbon.
Ultrasound Waves: Acoustic energy was also employed to propel particles (Figure 1c-i). This approach relies on the production of gas bubbles that aid for the motor mobility. [42,43] These bubbles are the result of the decomposition of volatile molecules (e.g., perfluorocarbons) upon exposure to ultrasounds. This approach is able to offer good directionality and strong propulsion, especially if high frequencies are used. The force, which with the particle is propelled, directly depends on the amplitude of the ultrasound wave and the volume of the particle, offering a convenient option to propel large particles. However, the shape of the motors that undergo ultrasound propulsion is extremely important. Tubular or bottle-like structures Adv. Biology 2023, 7, 2200308 Figure 1. External-actuated motors. Motor locomotion is triggered by external, physical stimuli, such as a) magnetic fields, either employing gradients or oscillating AC fields; b) light sources that can involve photocatalysis or heat generation; and c) ultrasounds. i) For each type, the schematic for the mechanism of motion is shown. Arrows indicate locomotion. Examples include a,ii) the screw-like magnetic motors that move under the influence of a magnetic field. Reproduced with permission. [20] Copyright 2018, Elsevier; b,ii) Janus Ag-poly(methyl methacrylate) motors that respond to light. Reproduced with permission. [39] Copyright 2019, American Chemical Society; and c,ii) Au-Ru rods that perform multiple trajectories upon ultrasound actuation. Reproduced with permission. [44] Copyright 2012, American Chemical Society.
are often employed, to allow the gas bubbles to be released from the cavity and to improve the locomotion. Alternatively, flexible materials (e.g., polypyrrole) that respond to small-amplitude oscillations can be employed, since the motion of the flexible part creates the motion. Overall, speeds up to 200 µm s −1 have been reported for these motors (Figure 1c-ii). [44]

Fuel-Powered Motors
Other approaches considered the energy harvest from (bio) chemical reactions to boost the motor mobility (Figure 2). The first examples employed inorganic catalysts, but more biofriendly approaches emerged, and enzymes were identified as the more relevant engines to power motors in biological environments. Lately, the use of polymers on the surface of the motors has also shown promising results.
Inorganic Nanoparticles: Metal nanoparticles or combinations thereof pose high catalytic activity independently of the environmental conditions (Figure 1a-i). With the aim to produce motors, the metals usually (partially) cover the surface of the motor and leverage catalytic reactions to result in locomotion. [45] The mechanism of locomotion considers the uneven distribution of product molecules from the substrates that will result in a gradient of matter in the immediate surrounding of the motor, thus leading to self-diffusiophoresis and the propulsion of the particle. The first motors employed either platinum (Pt) coatings or combinations of rare earths and transition metals (e.g., InGaAs/GaAs/Cr/Pt or InGaAs/Cr/Pt). These metals can react with hydrogen peroxide (H 2 O 2 ) to accelerate its decomposition into water (H 2 O) and oxygen (O 2 ) bubbles, which propels the motor. However, they are extremely toxic, specially to be used in biomedical applications, and the required high concentrations of fuel (i.e., H 2 O 2 ) to propel the objects are not compatible with biological settings. Therefore, there was a need to transition toward more "inert" metals, such as nickel (Ni), silver (Ag), or gold (Au), which were expected to be less harmful. In general, these motors paved the way for the use of chemical reactions to produce locomotion with top velocities up to 970 µm s −1 (Figure 2a-ii). [46] More recently, inorganic nanoparticles that experience core degradation have been utilized as an alternative to metal catalysis. Examples include magnesium (Mg) or zinc (Zn) corebased motors that under specific conditions (generally acid environments) undergo redox reactions where the particle core is oxidized into ions, producing gas bubbles than aid the locomotion, and leads eventually the disintegration of the motor. Although benefiting from the fact of being controllable degradable and avoiding toxic fuels, the effect of the ion release in in vivo applications remains to be understood in detail. Nevertheless, this kind of motors has shown great motion performance, achieving top velocities up to 200 µm s −1 .
Enzymes: Enzymes appeared as an alternative to metal catalysts owing to the fact that their protein nature is more acceptable for biological systems, and they exhibit unique catalytic ability (Figure 1b-i). [5,34,[47][48][49][50] They often show fast turnover rates with high substrate specificity, using substrates that are available in a living organism and producing typically harmless compounds that the body can further metabolize. In the beginning, the major challenges were however to preserve the enzyme integrity and activity upon conjugation to the carrier particle. Enzymes are usually fragile and enormously dependent on the environmental conditions. Drastic changes of pH or salt concentration as well as effects from immobilization methods to anchor them to the particle surface might lead to the (partial) loss of the enzymes' 3D structure and thereby their function.
Adv. Biology 2023, 7, 2200308 Figure 2. Fuel-driven motors. Motors move due to (bio)chemical reactions that include the use of a) inorganic nanoparticles, through metal catalysis or core disintegration, b) single or multiple enzymatic reactions and c) surface polymerization or polymer disassembly. i) For each type, the schematic for the mechanism of motion is shown. Arrows indicate locomotion. Examples include a,ii) the use of TiO 2 -coated SiO 2 motors that move via bubble propulsion upon addition of H 2 O 2 . Reproduced with permission. [46] Copyright 2018, American Chemical Society; b,ii) a six-enzyme cascade reaction to propel stomatocytes by self-diffusiophoresis. Reproduced with permission. [57] Copyright 2016, American Chemical Society; and c,ii) Janus motors showed enhanced diffusion upon polymerization of polymer brushes on their surface. Reproduced with permission. [64] Copyright 2021, Royal Society of Chemistry.
Not surprisingly, the first enzyme powered motors had very low speeds. However, great improvements were made in the field, such as the direct positioning of the enzymes instead of their stochastic distribution, the use of combinations of enzymes that convert substrates to products via cascade reactions, or enzyme purification. [51,52] Catalase, urease, and glucose-oxidase are the most common enzymes used for this purpose, probably due to their easy availability in sufficiently large amounts. Later, other enzymes were employed, such as β-galactosidase [53] or collagenase. [10,54] Although those enzymes displayed good performance as independent engine units, it was rapidly understood that concerted enzymatic reactions would lead toward more powerful and advanced motors. [49] Concerted reactions involve those cascade reactions where the product of one enzyme becomes the substrate of the second, and so on. Thus, the tandem glucose oxidase/catalase is among the most exploited enzyme combinations since the H 2 O 2 released after glucose conversion could be employed to generate O 2 bubbles and enhance the mobility of the motor while removing the toxic side product of the first reaction. [6,55,56] Tandem reactions were improved by Wilson's group, where they designed a stomatocyte-like motor that contained an ensemble of six enzymes working cooperatively (Figure 2b-ii). [57] The cascade reaction was initiated by the addition of glucose, which was transformed along the enzymatic network, speeding the motor up to ≈8 µm s −1 . Alternately, we have reported a dual-enzyme motor that displayed glucose-oxidase and trypsin and could therefore use two independent substrates (glucose and peptides) to gain long-lasting motion. [58] To date, several reports illustrate high-speed enzyme-powered motors (e.g., 60 µm s −1 when employing cascade reactions [57] ), making them a viable alternative to inorganic nanoparticles.
It is worth highlighting that artificial enzyme mimics (i.e., molecules or particles that can perform similar kind of reactions than (oxidative) enzymes) have become a fascinating alternative to enzymes. An interesting example is the use of nitric oxide (NO) producing enzyme mimics as part of the catalytic reaction to produce locomotion in motors. [33,[59][60][61] Polymers: Polymers were often used as materials for the core particle of the motor. [62] Recently, polymers have been employed as a power unit to propel motors (Figure 1c-i), inspired by the propulsion systems in microorganisms, such as Listeria monocytogenes, a food-born bacterium that propels by polymerizing actin filaments after infecting mammalian cells. [63] The concept of polymer-driven motors is based on the idea of asymmetry caused when polymer brushes either polymerize or detach from the particle surface. Both effects are dynamic processes that occur stochastically and bring the system to an out-ofequilibrium state, which allows for the enhanced diffusion of the motor. To the best of our knowledge, the first example that reported the use of polymers to induce locomotion employed silica (SiO 2 )-based core particles coated with multilayers of poly(vinyl pyrrolidone) and poly(methacrylic acid). [11] These polymeric layers were susceptible to pH changes and disassembled at basic pH, resulting in enhanced diffusion. A follow-up example considered the polymerization from the surface of a SiO 2 particle to produce motion. [64] Poly(hydroxyethyl methacrylate) brushes were synthesized in situ from the nanomotors via an atom-transfer radical polymerization, which was sufficient to propel the motors (Figure 2c-ii). These motors showed indications of swarming (collective motion) when high concentrations of motors were employed, mimicking the behavior of microorganisms such as bacteria or sperm cells. Additionally, polymers have been used as macromolecular scaffolds whereto immobilized enzyme mimics (i.e., metalloporphyrins), which eventually resulted in locomotion. [61] Nonetheless, there are a variety of challenges to be overcome before this type of motors can become competitive to the other concepts discussed thus far. For instance, polymerization procedures can have slow reaction rates required or organic solvents, inert atmosphere, or metal catalysts. On the other hand, the loss of the polymer brushes leads to an increasing viscosity in the close proximity of the motor, which may impair the locomotion. The top speeds reached thus far are only ≈3 µm s −1 . However, these are early days of polymer-based motors, and the wider and detailed exploration of this concept is yet to come.

Motors in Bio-Relevant Environments
The successful creation of self-propelling particles that could navigate in low-viscous, simple aqueous environments (vide supra), encouraged researchers to go further and test these motors in bio-relevant scenarios. The first stage was to evaluate their interaction with 2D cell cultures and assess the opportunities they could offer. The advances in the field gave eventual rise to the incorporation of motors into 3D cell cultures and finally tissue/organs in living organisms toward potential applications in diagnosis or therapy. [65][66][67][68][69] It should be noted that in the following paragraphs only selected reports of motors will be presented that are potentially suitable for applications in biomedicine. Therefore, neither inorganic catalysts-powered motors nor visible light-triggered motors or ultrasound-fostered motors will be further discussed here, although there are reports in literature that introduce the interaction of these types of motors with living organisms. [70][71][72] For the first case, locomotion involves either the use or production of toxic molecules, although advances are being made in the field to operate with ultra-low H 2 O 2 concentrations. [73][74][75] For the second type, although visible light offers a convenient source that is clean and long-lasting, it presents poor tissue penetration, it has been difficult to be applied in a biomedical context. In the last case, remaining challenges to control the stimuli (especially in vivo) currently limit their use in biomedicine.

High Viscous, Complex Environments
Biological environments are often more viscous than water and crowded with biomolecules (e.g., cell media, extracellular matrix, cytosol). Therefore, motors capable of moving in these regimes are required to be employed in biomedicine. In addition to viscosity as the major constrain to motors' motion, other aspects such as boundaries (e.g., walls, steps), obstacles (e.g., fibers), or surfaces (e.g., rough versus smooth) will impair their locomotion.
A step forward illustrated the use of magnetic motors in more viscous environments. Owing to their capacity to strongly respond to a magnetic field, magnetic motors typically stand as the best candidates for navigation in complex media. However, the velocity can be compromised if the size of the motor is too large, or the viscosity is too high. For example, the dynamics of helical magnetic motors has been investigated in viscoelastic [76] and non-Newtonian fluids, [77] sperm, [78] blood, [79] and in different biological media, including cell extract and the zebrafish yolk sac, [80] showing that the viscosity impacted the velocity of the motors. It is worth highlighting that magnetic motors can also work as microscale viscometers, as the dynamic viscosity of the fluid is related to the particle velocity. Studies have been carried out where magnetic motors allowed for the quantification of rheological characteristics of the fluids. [81][82][83] Other examples where magnetic motors have been used in complex media include the navigation in the vitreous body of the eye, [84] and the mobility on the surface of moisture paper chips. [85] In the latter case, follow-up investigation considered the transportation of keratinocytes (Figure 3a-i). [86] Polystyrene beads (4 and 20 µm in diameter) coated with magnetite nanoparticles were used as the motors and incubated together with the cells. The cells interacted with and attached to the motors within 72 h and were transported by the motors upon applying a magnetic field. An alternative approach has been employed for transportation and delivery of microdroplets. [87] Carbon-coated iron motors were modified to display either a hydrophobic or hydrophilic surface. When hydrophobic surfaces were considered, the motors were attached to the microdroplets by van der Waals interactions and moved the droplets under the actuation of a magnetic field gradient. Conversely, when hydrophilic surfaces were employed, the motors did not attach but their locomotion toward the magnetic field enthralled the microdroplets together with the surrounding liquid and moved them. These are only few examples of what motors can do regarding transport of cargo, as recently reviewed by Xu et al. [88] Light-propelled motors have also been studied in more complex media. [89][90][91][92] For instance, NIR-actuated motors showed reduced amyloid aggregation in an amyloid-β fibril crowded environment (Figure 3a-ii). [93] The motor presented a Janus geometry, decorated with Au nanoparticles that allowed for the locomotion via a thermophoretic process and amyloid-β inhibitor molecules to impair the amyloid aggregation. There was a 66% decrease of the amyloid aggregation when the motors were present. Finally, they assessed the penetration in a blood-brain barrier model, where the results suggested a slight improvement when using their motors.
An interesting approach to emphasize the uniqueness of enzyme-driven motion in complex environments is to employ gels made of biopolymers that can act as substrates of the enzymes. Besides being larger and more complex molecules, biopolymers can mimic the extracellular environment that motors need to face when navigating in living organisms (e.g., mucus layer, collagen matrix). For example, Zhang et al. showed that hyaluronidase-decorated motors were able to navigate in hyaluronic acid gels, with top speeds of 8 µm s −1 by digesting the gel in the nearest proximity. [94] A similar approach was reported by Choi and co-workers, who designed ureasepowered motors able to move in mucin gels, mimicking the Helicobacter pylori mechanism of infection. [95] The gel became less viscous in the surroundings of the motors owing to the enzymatic activity and allowed for locomotion. Collagenase-propelled motors could navigate in collagen fiber networks with different densities mimicking the extracellular matrix, achieving top velocities of ≈30 µm s −1 . [10,54] An advanced example considered the combination of bottlebrush polymers as scaffolds for the deposition of catalase, which could navigate a tumor microenvironment model (Figure 3a-iii). [96] The motor consisted of a tadpole-like structure combining hydrophilic and hydrophobic polymers that in combination with catalase displayed ballistic motion in a viscous 3D collagen gel model with top velocities

In Vitro and In Vivo: From Cells to Animal Models
The next step forward was made when motors were used in cell culture. First examples employed 2D cell cultures with the aim to investigate the interactions and advantages motors could bring to biomedicine. More advanced motors were then chosen for experiments in 3D cell models, which offer better resemblance to tissues. Lastly, the most advanced motors were explored in animal models, and their toxicity, locomotion, and potential in, e.g., drug delivery or signaling were assessed.
Magnetic motors have been investigated in their interactions with cells and tissue. Magnetic motors have been usually employed for targeted drug delivery, [37] mechanical destruction of cells, [97] or thermal therapies. [10,98] A recent approach used a gold-nickel (Au-Ni) rod-like motor with a dual performance, namely, biosensing and drug delivery. [99] On the one hand, the motor was able to identify and selectively bind to miRNA fragments, which were detected by fluorescence spectroscopy. On the other hand, the drug delivery performance was assessed in MCF-7 cells using doxorubicin-loaded motors. The cell viability decreased down to 80% after 24 h treatment using relatively low concentrations (200 × 10 −6 m doxorubicin/0.05 mg mL −1 motor), while higher concentrations were needed for passive particles (e.g., 70% viability using 1 × 10 −3 m doxorubicin/0.5 mg mL −1 SiO 2 particles). A different example exploited the use of magnetic motors for targeting the precise locations in a rat leg. [100] The motors consisted of Chlorella pyrenoidosa coated with magnetite nanoparticles that serve to magnetic guidance and eventual heat dissipation upon NIR illumination. Chlorella pyrenoidosa was chosen as it is widely commercialized as a nutritional supplement. The motors were injected into a rat leg and guided to a specific location for interaction with the muscle fibers. Upon irradiation with light, the muscle fibers could contract in a controlled fashion with no apparent damage to the surrounding tissue.
An advantage of magnetic fields is that they provide sufficient energy to propel larger (>1 µm) motors. For instance, magnetic poly(vinyl alcohol) hydrogel-based motors in the millimeter size range were employed to activate the migration of T cells by recruiting the chemokines that the motor released. [101] Jurkat T cells and the motors were placed in opposite ends of a microchannel and let to interact. The cells showed positive chemotaxis toward the motors, as both the number of cells increased over time (from 110 after 4 h exposure to the motors up to >200 cells after 12 h), and the distance between the motors and the cells became shorter.
Light-driven motors were used to cross 3D cell spheroids that resembled a tumor microenvironment. [102] The motors consisted of biodegradable Janus polymerosomes made of poly(ethylene glycol)-b-poly(D,L-lactide) block copolymers with a hemisphere decorated with Au nanoparticles. These motors responded to NIR light and moved by thermal diffusiophoresis due to the Au coating. They showed a twofold increase in spheroid penetration compared to regular polymerosomes when incubated together with 3D HeLa cells spheroids. Light-driven motors were also employed for 2D and 3D cell penetration and cytosolic placement was carried out by Zhou et al. [91] Janus calcium carbonate/ Au motors were used, which responded to NIR radiation and moved by thermophoresis besides releasing gas bubbles. These motors showed indications of lysosomal escape, likely due to the gas release that burst the vesicles and impaired the cell viability in 2D cell cultures after delivering the payload doxorubicin. 3D Adv. Biology 2023, 7, 2200308  [86] Copyright 2022, The Authors, published by Wiley-VCH. ii) Janus nanomotors were able to navigate in a β-amyloid crowded solution while assisting to protein disaggregation. Scale bars are 1 µm. Reproduced with permission. [93] Copyright 2020, American Chemical Society. iii) Polymer-based motors carrying catalase enzyme were employed in 3D collagen gels. The motors displayed no motion in the absence of H 2 O 2 but gained locomotion upon addition of the fuel. Reproduced with permission. [96] Copyright 2019, American Chemical Society. b) Motors in cell models and tissue. i) Light-triggered motors were used as MRI probes to detect tumors in mice. The T 1 signal due to Gd complexes was higher than the controls. Reproduced with permission. [103] Copyright 2021, Wiley-VCH. ii) Urease-powered motors were able to penetrate 3D cell spheroids and promote their disruption from the inside. Reproduced with permission. [111] Copyright 2019, American Chemical Society. iii) Self-disintegrating motors were injected in mice for the treatment of Helicobacter pylori. After treatment, both histology photographs and confocal laser scanning microscopy images showed a depletion in the bacteria colonies compared to the controls. Reproduced with permission. [118] Copyright 2021, Wiley-VCH.
cell spheroids were further created to investigate the penetration in tissue-like environments. The motors appeared localized inside the spheroids after 30 min and showed significant differences compared to the controls where no NIR light was utilized. An advanced example used a mesoporous SiO 2 core, where gold nanoparticles were deposited on one hemisphere to aid the locomotion, and gadolinium (Gd) complexes decorated the second hemisphere to serve as a contrast agent (Figure 3b-i). [103] The motor actuation under NIR illumination enhanced the penetration of the motors into tumor models, causing also cellular damage to the tumoral cells due to the heat release during the thermal-triggered diffusion of the motors. These motors were lastly used for magnetic resonance imaging (MRI), owing to the role of Gd as a T 1 -contrast agent. The findings showed a twofold increase of the MRI response both in vitro (compared to the commercial contrast agent) and after injection in mice models (compared to controls without NIR irradiation), suggesting that the motors might perform diagnosis analysis.
Enzyme-powered motors were typically assessed in 2D cell model cultures to determine the interactions of motors and cells. These motors were usually based on mesoporous (hetero) silica particles [104,105] or liposomes, [106,107] mainly employing urease or catalase as the engine units toward drug delivery applications, [65,108] antibacterial properties, [109] or triglyceride degradation. [110] Although 2D models are useful as initial screening platforms to understand the principles of uptake and cytotoxicity, more realistic examples consider 3D cell spheroids, which are more similar to the real environment in organisms. For instance, Hortelão et al. have reported urease-powered motors that could penetrate 3D bladder cancer spheroids and disrupted them upon internalization (Figure 3b-ii). [111] Similarly, collagenase-powered motors decorated with magnetic nanoparticles showed to increase the uptake of 3D bone cell spheroids and impair their viability after a hyperthermia treatment. [10] However, the use of enzyme-based motors in vivo is challenging since the enzymes need to be protected or hidden, their activity must be preserved, and sufficient substrates have to be available, among others. Joseph et al. developed an asymmetric polymerosome that responded to glucose gradients (chemotaxis) and indications of crossing a blood-brain barrier model in rats were claimed. [112] Similarly, Hortelão et al. demonstrated the capacity of urease-powered motors in an animal model. [113] The motors consisted of a mesoporous SiO 2 core decorated with urease for locomotion and Au nanoparticles for particle radiolabeling and tracking (by positron emission tomography imaging) inside the mice models. The motors showed accumulation mainly in the liver 5 min after intravenous injection, and negligible accumulation in lungs, kidneys, or bladder. When intravesical injection in the bladder was considered, the motors showed indications of retention in the organ and displayed a better distribution within compared to the controls (passive particles).
Wang's group pioneered the development of self-disintegrating motors that presumably pose the advantage of zerowaste and self-destruction while performing their locomotion. These motors gained high speed due to the disintegration of the core particle via redox reactions and a bubble propulsion mechanism that allowed them to be employed in vivo at an early stage. [114] Multiple examples have been reported on the use of these motors, usually related to gastrointestinal applications.
An initial example included the design of a magnesium-based motor with an enteric coating that prevented core degradation in a mouse stomach after oral administration. [115] The locomotion was initiated when the motors arrived in the intestine. The pH increased to 7 and above dissolving the coating and producing H 2 bubbles that resulted in motors moving with top speeds up to 60 µm s −1 . Similar systems were employed for the in vivo treatment of Helicobacter pylori in mice stomach, [116] or the treatment of iron deficiency (anemia). [117] Inspired by this idea, other groups developed core-disintegrating motors for in vivo applications. For instance, Wu et al. reported a calcium oxide nanobottle loaded with Pt nanoparticles and presenting an external silicon coating (Figure 3b-iii). [118] These motors were administered to mice and showed evidence of neutralizing the abnormally lowered pH in infected mice stomach within 20 min. After 24 h, they were able to return the acid pH in the stomach to normal values. Mice infected with Helicobacter pylori were recovered from infection after 5 days of treatment. A different approach was used by Xu et al., when they employed magnesium-based motors for rheumatoid arthritis therapy. [119] During locomotion, this motor released hydrogen gas that, besides propelling the motor, scavenged the reactive oxygen species produced during the rheumatoid process and alleviated the inflammation.
In addition to disintegrating inorganic core particles, polymeric particles were also explored toward self-disintegrating motors. For instance, Wan and co-workers reported a hybrid polyamide/L-arginine motor that could move in cells by conversion of arginine into citrulline and NO in the presence of H 2 O 2 . [59] These motors showed top speeds of up to ≈15 µm s −1 with increasing arginine-to-polyamide mass ratio (i.e., from 5 to 20) or increasing H 2 O 2 concentration (up to ≈4 µm s −1 for a mass ratio of 10).
Polymer-powered motors were a more recent addition to the field. Although these motors do not have high speed thus far, the vast toolbox of synthetic/polymer chemistry offers tremendous opportunities to modify different aspects of the polymer and by doing so, tunes the resulting locomotion performance. To the best of our knowledge, only two examples have reported the use of polymer-propelling motors in cell culture. In the first case, Gisbert-Garzarán et al. designed a motor containing a polymer that responded to redox changes in the close environment via a redox-responsive linker. [120] The motors showed indications of lysosomal escape, likely due to the proton sponge effect. Similarly, we have recently reported a polymer conjugated to a SiO 2 particle through a pH-responsive linker. The linker was cleaved and the polymer brushes were released when exposed to acid pH or cell media, increasing the motor locomotion. [121] The motors displayed velocities up to ≈3 µm s −1 in cell media when Janus (asymmetric) motors were used with the uptake of the motors by RAW 264.7 macrophages increased up to 5 × compared to motors without the pH-responsive linker, probably due to the enhanced diffusion of the motors.

Challenges and Advances in the Field
Motors have undeniably enriched the opportunities of (semi)synthetic materials to interact with cells and tissue.
Nonetheless, the current state of the field still falls short of the expectations that are in particularly fueled by TV shows and movies. With that in mind, this section offers a discussion on future aspects to be considered including motor swarming (collective behavior), theoretical models and simulations, and energy requirements for crossing biological membranes.

Collective Behavior: Swarming
Crowding and collective motion is a typical feature in living organisms. Nature offers multiple examples of organized, mobile models that increase the chances of success toward a specific goal compared to the individuals. For example, flocks of birds that orient together during migrations, schools of fish that protect themselves against predators, or colonies of ants that build bridges to cross gaps. This aspect is not exclusive for animals or insects. Bacteria colonies also show swarming (collective) behavior when the concentration of bacteria increases. [122][123][124] Sperm cells have been demonstrated to move 1.5× faster when swarming than individual spermatozoa, [125] increasing the chances of fertilization.
In the context of motors and biomedicine, swarms of particles could be of utmost importance. It is envisioned that synergistic interactions between the motors when in a swarm would boost their velocity to accomplish motion in turbulent fluids (like blood) or provide a more robust motors' configuration to move in complex biological media. In this regard, magnetic fields have been reported to produce swarms of motors. [22,97,[126][127][128] A recent example included the use of magnetic motors (based on magnetite, Fe 3 O 4 ), that could swarm into a millimeter-size flock with top speeds of ≈8 mm s −1 . [127] The motors moved under a combination of a gradient magnetic field and a rotating magnetic field, which provided spatial control and high-speed, respectively. The motors were employed as a proof-of-concept for tomographic imaging using magnetic particle imaging that was based on the nonlinear magnetization behavior of the swarm constituents. Another interesting example combined Fe 3 O 4 particles with an enzyme that has thrombolytic properties. [128] These motors could swarm in the presence of a rotating magnetic field gradient, which resulted in locomotion and additional spinning of the swarm (Figure 4a-i). This effect enhanced the fluid shear stress acting on the thrombi and accelerated their disintegration. In vitro studies demonstrated a tenfold increase in the efficiency of this approach compared to the enzyme alone (Figure 4a-ii), and a total disintegration of the clot after ≈20 min in animal studies (Figure 4a-iii).
Light can also contribute to swarm formation. [92,[129][130][131][132] An elegant approach has been reported by Sun et al. [133] where they used NIR light-driven motors based on a polydopamine coating to control the dynamics of the swarm. The motion was found to be dependent on particle concentration. For small groups of motors, they move by self-thermophoresis, while for larger number of motors, the locomotion was dominated by convective flows. The motors' concentration also controlled their directionality, showing negative phototaxis at low concentrations and positive phototaxis at high numbers (Figure 4b-i-iv). This was attributed to the balance between hydrodynamic drag and thermophoretic forces. They incorporated the motors into an ink that could be printed and reproduced images and text based on the light response (Figure 4b-v). Another examples where light was used to control swarming relied on switchable photoresponsive colloids. [132] This approach was based on goldtitania (Au/TiO 2 ) motors that showed negative or positive phototaxis toward blue or green light, respectively, in the presence of H 2 O 2 . According to this, the motors would swarm under green light and redispersed under blue light illumination. Interestingly, when active and passive particles were combined in the same medium and illuminated, the phoretic flow field could be visualized. The passive particles organized in a hexagonal pattern around the active particle when exposed to green light, whereas the passive particles were pushed away far from the active particle during blue light illumination. An interesting example that used visible light described a biomimetic motor that resembled phototactic microorganisms. [134] This motor presented a Janus shape made of carbon nitride/polypyrrole that showed negative chemotaxis toward low intensity light and positive chemotaxis toward high intensity light, being nonmobile at intermediate intensities or when both (high and low) intensities were applied simultaneously. Interestingly, the mechanism of diffusion depended on the light intensity: as for low intensities, the motor would move via self-diffusiophoresis while as for high intensities it would rely on self-thermophoresis. Further, the motor exhibited spinning schooling behavior (such as a school of fish) when divergent light was applied.
Ultrasounds have also been used to control swarming. Tang et al. demonstrated the modulation of a swarm of light-driven motors using acoustic waves. [135] The approach was based on Au/TiO 2 microbowls that responded to both UV light and ultrasounds. When the outer layer of the microbowl was Au, the motors flocked under ultrasound actuation and dispersed upon irradiation with UV light, flocking again when the light was off (Figure 4c-i). The opposite behavior was observed when the outer layer of the microbowls was TiO 2 . In this case, the particles disaggregated under ultrasound actuation and collapsed upon UV illumination (Figure 4c-ii).
Physico-chemical interactions (or chemical reactions) are a means to control particle interactions. [64,113,[136][137][138] For instance, a predator-prey system has been reproduced by Mou and coworkers. [137] They used zinc oxide (ZnO) nanospindles as the predator (active) particles and TiO 2 nanospheres as the prey (passive) particles. The TiO 2 nanospheres moved and dispersed when the ZnO nanospindles navigated toward them (Figure 4d-i). This behavior was explained by the local electric field created upon hydrolysis of both kinds of particles in aqueous solution. The release of ions generated a diffusioelectric gradient that made the particles attract or repel each other (Figure 4d-ii). A similar approach was reported by Liang et al., presenting microswarms with leader-follower behavior. [138] They employed TiO 2 particles of two different sizes that responded to an electric field. Electrohydrodynamic and diffusiophoretic interactions were responsible for the flocking of the particles. The small particles (followers) were immobile until a larger particle (leader) passed by, collecting the followers due to the increase in the electrohydrodynamic force, which would favor the interactions. If multiple leaders were present, the flock was unstable.

Theoretical Models and Simulations
Theoretical predictions stand as a convenient means to predict and/or elucidate the systems' behavior offering complementary, instructive, or supportive information to experiments. For instance, the characteristics of cell membranes have been predicted [139] or dynamics of microorganisms in confined spaces have been identified using molecular simulations.
Adv. Biology 2023, 7, 2200308 Figure 4. Swarms of motors. a), i) Swarms of magnetic motors were formed under a rotating magnetic field gradient. These swarms were able to unclog thrombi ii) in vitro and iii) in vivo with high efficiency. Reproduced with permission. [128] Copyright 2022, Wiley-VCH. b) A light-responsive swarm was evaluated for light-triggered imprints. The velocity of i,ii) one single motor was substantially lower than iii,iv) the swarm. Controlled swarming of v1-3) the motors under light illumination and v4-6,v10-12) schematic illustration of the v7-9,v13-15) corresponding optical microscopy images after printing. Reproduced with permission. [133] Copyright 2019, American Chemical Society. c) Swarming of Au/TiO 2 microbowls could be controlled under ultrasound actuation together with UV illumination. When the outer part of the microbowls was Au, i) the swarm redispersed upon UV illumination, while ii) it compacted if TiO 2 was on the outer shell. Reproduced with permission. [135] Copyright 2019, Wiley-VCH. d) Physico-chemical interactions were used to mimic a predator-prey system. i) The predator particle (trajectories in red and green) would move toward the swarm of prey particles that escaped from it. ii) The local electrophoretic field generated by the particles drove the interactions between the two groups, explaining the locomotion. Reproduced with permission. [137] Copyright 2020, American Chemical Society.
Huge efforts are devoted to investigating theoretical models and simulations of active matter. Modeling has helped experimentalists to analyze and interpret data. For example, the meansquared displacement model is extensively used in the field and sets the basis for particle tracking analysis and the discrimination of the types of motion (i.e., Brownian (random) motion vs ballistic (directed) nanopropulsion) (Figure 5a). [140] This simple model is valid for most of the scenarios where motors navigate in homogenous environments and only differentiation between Brownian and ballistic motion is required. However, when complex or less common locomotion properties are observed, more elaborate tools are needed to assess the locomotion, for instance when motors underwent deceleration. [121] Theoretical predictions have also facilitated the understanding of the self-diffusiophoretic principle of Janus motors. For example, Lauga and Spagnolie [1,141] and Gaspard and Kapral [142] described and advanced several aspects of the selfdiffusiophoretic mechanisms for Janus motors. Specifically, they extended the molecular derivations of the Langevin equations to describe the translational and rotational dynamics of rigid active colloids in nonequilibrium environments. [143] Thus, they correlated the relations between the fluid and the colloids considering boundary conditions, which are often neglected. The model could i) predict the force and torque, which are not present for ordinary Brownian motion equations, ii) be used to extrapolate to thermophoretic active colloids in the presence of an external temperature gradient, and iii) incorporate nonrigid particles to evaluate their motion. Further, simulations of Janus motors with catalytic domains showed to break the axial symmetry and result in rotational motion, [144] the dynamics of a Janus motor was determined where propulsion was not possible due to reversible kinetics, [145] or the relation between the fluid velocity and the active particle originated from thermal and molecular fluctuations was evaluated. [146,147] Mathematical models were also useful to explain the motion of motors in the presence of boundaries. [148,149] Microorganisms and self-propelling particles present complex trajectories when navigating close to obstacles (e.g., walls, surfaces, other objects) due to nanoscale, physical interactions. [150][151][152] For example, Spagnolie et al. investigated the capture and escape of motors and spheres (Figure 5b). [141,152] These findings suggested the existence of a critical trapping radius in which the motor was attracted to the sphere by passive hydrodynamic interactions, resulting in an orbiting motion pattern. Fluctuations in the system or geometrical defects in the motor could lead to the escape, and robust trapping was only expected to occur upon direct interaction of the motor and the obstacle.

Crossing Biological Barriers
A major challenge that motors face when it comes to biomedical applications (e.g., in drug delivery) is their capacity to attain cytosolic placement. It is widely accepted that nanoparticles in general can be endocytosed by cells and thereby, confined in endosomes that eventually lead to the dissolution or exocytosis of the particles. Consequently, endo/lysosomal escape is essential yet a demanding task that needs to be accomplished. Strategies to address this issue include electroporation [153] or the use of magnetic nanoparticles in the presence of a magnet during incubation with cells [80] to avoid endocytosis. Alternately, nanoparticles are equipped with polymer coatings [154] or selfpenetrating peptides [155] that burst the endosomes upon acidification following endocytosis. However, these approaches often result in cell stress and impaired cell viability, leading to irreversible cell injury and necrosis.
Not surprisingly, the question has been posed if motors could have sufficient thrust to cross lipid membranes. As a first step, the forces that rule out particle-mediated cell membrane penetration need to be understood, [13,102,120,[156][157][158][159][160][161][162] to design motors that are powerful enough to penetrate membranes. Considering the motor dynamics, its motion can be dissociated into rotational and translational motion, each contributing to their interaction with the membrane. A molecular modeling study suggested that rotation of the particle is key to compromise membrane integrity. [156] These results showed that rotating particles enhanced the uptake, since the membrane bending energy was reduced and the membrane dynamics (and even the fluidity) changed. However, excessive rotation could lead to membrane disruption. The rupture was observed using particles of 15 and 20 nm in diameter with rotation speeds of 0.6 deg ns −1 (≈10 6 s −1 ). This is in agreement with models that predict that motors have low rotational speeds, [144] as translational motion is more relevant. Taken together, this indicates that motors move with slower rotational speeds than passive colloids, which suggests that rotational motion only is not strong enough to produce membrane rupture.
Further, investigations into the interaction of membranes and self-propel particles have been carried out (without rotational motion), using theoretical and numerical simulations. [160] The results indicated that the interaction between a motor and a membrane can occur in three different manners: i) trapping of the motor within the membrane, ii) penetration through the membrane and membrane healing (no rupture), and iii) penetration without healing (rupture) (Figure 5c). Each scenario depends on the interplay of two parameters, referred to as the reduced activity and the particle-to-membrane thickness ratio. The reduced activity accounts for the active force of the particle relative to the repulsive steric force upon contact with the membrane. Therefore, trapping is more likely to happen for low reduced activity (i.e., motors with very low speeds). Penetration would occur for higher numbers of reduced activity. The membrane was predicted to not be compromised if the particle size is similar to the membrane thickness (up to 3-4× larger). However, there are no experimental results to support this prediction.
The experimental approaches so far considered have often employed millimetric size motors actuated by external stimuli. [163,164] To the best of our knowledge, the only example that showed results on nanosized motors crossing a biological tissue was reported by Wang et al. [165] In this article, a motor consisting of a mesoporous silica particle half-coated with a Pt layer and self-assembled monolayer (SAM) extracted from Staphylococcus aureus was fabricated with the aim to cross an in vitro and in vivo intestinal epithelium model. The Pt responded to an NIR laser to induce propulsion up to 2 µm s −1 . The SAM camouflaged the motor when interacting with the cell barrier. The in vitro cell model consisted of a co-culture of Caco-2 and HT29-MTX cells in transwell plates. The motors penetrated the cell barrier 4.5× micrometers more than passive particles. Both electron microscopy and confocal laser scanning microscopy images revealed that the motors were found in the cilia and inside the cells of the epithelium model compared to controls where no uptake by cells was observed. Further, the motors were investigated in mice models, showing successful accumulation in the intestine, without any toxicity during the 3 weeks of treatment.

Conclusion and Outlook
The field of nano/micromotors has shown great advances since the first motors were designed. Simple modes of locomotion using chemical catalysis and external stimuli such as magnetic fields, light, and ultrasounds have evolved toward more complex mechanisms of mobility, which include the use of enzymatic reactions, self-disintegrating motors, and more recently polymer coatings. All these advances were possible thanks to the collaborative efforts of different disciplines, e.g., chemistry, material science, physics, mathematics, and molecular/cell biology. This led to a better understanding of nanoscale motion, to creative motor designs, and to advanced motors that could self-propel in many different milieus. Consequently, the scope of the field broadened beyond nanomedicine.
From the first examples that navigated in aqueous environments and gels, motors were evaluated in their interactions with 2D and 3D cell cultures, tissue, and more recently in living organisms. In addition to assessing the fundamental locomotion properties, different potential applications are explored including biosensing, diagnosis, drug delivery, and basic nanosurgery. Opportunities for collective motion of motors start to be realized beyond using magnetic motors. In parallel, more accurate mathematical models are developed to unravel mechanisms that do not follow the classical modes of locomotion.
Despite all the advances, the current motors are not fulfilling the high expectations with many fundamental questions and challenges unsolved. As pointed out by Wang and Gao more Adv. Biology 2023, 7, 2200308 ii) The mean-squared displacement (MSD) plots show linear or parabolic trends depending on whether the particle motion is Brownian or directed. Reproduced with permission. [140] Copyright 2012, American Chemical Society. b) Models have predicted the particle i) trapping and orbiting around obstacles that will ii) eventually escape the radius of capture. Reproduced with permission. [152] Copyright 2015, Royal Society of Chemistry. c) Theoretical models have been employed to determine the efficiency of particle crossing a lipid membrane. Three scenarios are provided (from top to bottom): particle trapping, penetration with membrane healing, and penetration without healing. Reproduced with permission. [160] Copyright 2019, American Institute of Physics. than 10 years ago, nanomachines offer many opportunities to share, also in the field of nanomedicine. [166,167] Some of the questions they discussed have been addressed in the past years, i.e., cooperative motion, [64,84,126,133,135,137,138] transport of heavy cargo, [88] navigation in body fluids, [78,84,93,96] and applications in vitro and in vivo. [48,66,94,114] In my opinion, the next step forward requires attention to the following five key aspects (Figure 6): 1) Efforts have considered either simple, aqueous media or complex, high-viscous media (e.g., gels, protein crowded solutions). However, no experiments have been done in viscosity-changing media, i.e., inhomogenous environments. In the context of nanomedicine, motors have to cross from an extracellular media to the inside of the cell. [168,169] Assuming cytosolic placement (by either escaping the endo/lysosomes or direct incorporation into the cytosol), the viscosity of the media where the motors are placed will change drastically. Usually, motors are incorporated in cells in diluted solutions with low viscosity, which allows for relatively high speeds of the motor. Upon cytosolic placement, the motors will face an ultra-high crowded solution where the major components are amino acids, proteins, and a high amount of ions, thus impairing their mobility. [170] Therefore, although the most advanced motors achieved high speeds in simulated conditions, experiments should be done in more realistic scenarios, which include a real transition from a low-viscous to a high-viscous environment in order to identify the impact on the motor mobility. Models to operate under these conditions could include gradients within the solution, which are a gentle transition from two distinct media, or physical barriers that separate two media, offering a steep hurdle for the motor to cross. 2) In line with the previous, conditions for mobility only consider static fluids where no flow is applied. However, in many of the envisioned applications, laminar or turbulent flows are present, especially when body fluids are considered (e.g., nanosurgery to clear arteries from clogs, treatment against bacterial infections in the intestine, or fertility treatments).
Simulations of magnetic motors in turbulent flow conditions revealed that vortices formed in the motor boundaries Adv. Biology 2023, 7, 2200308 Figure 6. Future perspectives of nano/micromotors. Viscosity-changing environments would impair the motor velocity, as it inversely depends on the dynamic viscosity of the media. Complementary, laminar, or turbulent flows would alter the motor trajectories and affect their locomotion. Further, when considering controlled swarming and locomotion, the type of stimulus and the interactions between the motors should be considered. A major challenge remains the design of motors that can use their inherent thrust/power to propel across biological membranes and not to be trapped in the interface. Additionally, new applications for motors could include the use in nanobiopsy with minimal damage for the cells. Finally, the biosafety and toxicity of motors may pose a hurdle for biomedical applications and minimizing the risks must be thoroughly considered.
affected its velocity. [171] Thereby, analyzing the mobility of motors in flow regimes is relevant to understand the realistic performance of the motors in this conditions. 3) Swarming is thus far an underappreciated aspect of motors.
Microorganisms, e.g., bacteria or motile cells, display swarming behavior when reaching high concentrations of individuals. This behavior also seems to be affected by boundary conditions, i.e., physical barriers or obstacles that the individuals encounter during locomotion. It remains to be fully elucidated how these organisms "choose" collective behavior, as it appears that they only interact with the nearest neighbors in short-range interactions. This suggests that an individual does not perceive what happens outside a very close volume. However, the coordinated swarm behaves as a single individual. The same was observed in pluricellular organisms, such as birds, fish, reindeers, and even humans. The question now is if this behavior is unique for living organisms, or similar concepts could be applied to nonliving matter. In other words, does matter need to be alive to be able to swarm? So far, there is little knowledge and experimental data on this type of collective motion. Most of the examples reported to date use light-triggered and magnetically actuated motors, which employ thermophoretic or diffusioelectric gradients in the first case and dipolar interactions in the second case, to achieve particle agglomeration. However, the theoretical considerations when it comes to swarming of chemical-triggered motors become more challenging. Is it swarming particle-density dependent as for instance observed in bacteria? Can it only be achieved by electro/thermophoretic gradients or what other weak forces can trigger collective behavior (van der Waals, hydrophobic)? Further, it is known that the speed and thrust of swarming microorganisms increase compared to the individuals. Initial investigations suggested that this also happens for motors. [64,133,134] However, are these speeds sufficiently high to move in complex environments?
This last question leads to the next point to address: crossing biological membranes. Until now, most motors still rely on common endocytosis to be phagocytized by cells and then perform a task after internalization. This means that the motors' power has not been actually used to cross and penetrate membranes or to escape from endolysosomes using the kinetic energy a motor may pose. For this to happen, the power of the motor must be such that it can cross through a biological membrane. So far, there is only a single study where atomic force microscopy experiments have determined that the minimum force to break a bacterial cell membrane was ≈20 nN. [172] However, the corresponding force for mammalian cells remains unknown; and so is the force (thrust) that a motor can achieve during locomotion. 4) Finally, an improvement in the use of motors may consider their applicability in exploiting functions while moving.
There are examples in literature where the mechanical power of the motors was used to carry out tasks such as clearing clogs in arterial models or doing in situ rheology at the nanoscale. An interesting application where nanomachines could have an impact is their use in nanobiopsy. It is known that many diseases begin with the release of specific biochemical signals (chemokines) at usually very low concentrations, which makes them very challenging to detect. In this regard, motors, having smaller sizes than conventional biopsy tools and being able to operate at low concentrations and with high affinity to specific molecules, could be a means to detect biological signals at subcellular level. Designs may consider cytosolic placement of the motors (as discussed above) and recovery of the motor upon cell intervention with the minimal impact for the cell. Further, while the sensing capacity of the individual motors might not be enough, controlled swarming may increase the chances of molecule detection at ultra-low concentrations. 5) Biosafety and toxicity aspects are important and require more attention in the future before nanomachines can become clinical reality. Although initial reports point toward some level of biocompatibility, [173,174] the interaction of motors and cells might affect the viability of the cells or even cause extensive damage to the surrounding tissue in particular for swarming.