Abstract
Although various (bio)fabrication technologies have achieved revolutionary progress in the past decades, engineered constructs still fall short of expectations owing to their inability to attain precisely designable functions. Shrinkable and expandable (bio)materials feature unique characteristics leading to size-/shape-shifting and thus have exhibited a strong potential to equip current engineering technologies with promoted capacities towards applications in biomedicine. In this Progress Report, the advances of size-/shape-shifting (bio)materials enabled by various stimuli, are evaluated; furthermore, representative biomedical applications associated with size-/shape-shifting (bio)materials are also exemplified. Towards the future, the combination of size-/shape-shifting (bio)materials and 3D/4D fabrication technologies presents a wide range of possibilities for further development of intricate functional architectures.
Keywords: 3D fabrication, 4D fabrication, stimuli-responsive, size-shifting, shape-shifting, biomedical application
Graphical Abstract
This Progress Report systematically summarizes representative size-/shape-shifting mechanisms used towards engineering (bio)materials. With such unique maneuverability, numerous biomedical applications have been made possible in the fields of tissue engineering, imaging, drug delivery, 4D printing, medical device, and nanofabrication.

1. Introduction
Biomedical engineering is a multidisciplinary field that encompasses engineering, biology, and materials, among others, raising as a collection of novel strategies that apply engineering principles to healthcare and medicine.[1] Although tremendous progress has been achieved in this field, researchers have long-struggled to properly replicate biological tissues for two reasons, cellular and structural complexity.[2, 3] The former can be resolved by using stem cells, which can differentiate into somatic cells with desired functions.[4] As to the structural complexity, awareness is raised regarding the differences in cellular remodeling between two-dimensional (2D) and three-dimensional (3D) microenvironments.[5, 6] Studies have demonstrated that a 3D microenvironment could provide more physiologically relevant conditions to guide cell behaviors and regulate cell functions.[1] Therefore, numerous tissue-mimicking functional 3D architectures have been fabricated with a range of solutions solved by different techniques.[7, 8] As one example, significant efforts have been made to develop different (bio)fabrication techniques that can replicate intricate volumetric architectures with (bio)materials and specific cell types to establish the native-like microenvironments and biological functions.[9, 10]
Much work so far has focused on harnessing biomimetic technologies to rebuild microscale tissue mimics using other methods to improve their biofunctions.[11] However, uncertainty still remains regarding fabricating macroscale or nanoscale constructs, especially with multiscale or hierarchical structures. This could be a pivotal obstacle to overcome since such variations from the macroscale to the nanoscale may be the decisive factor for developing tissue functions in different dimensional regimes. For instance, cartilage has an oriented nanoscale fibril structure and nanopore spaces between the glycosaminoglycans networks; wherein hierarchical structure of the bone ranges from macro to nanoscale.[12, 13] The performances of tissue structures are different at the microscale while molecular constituents vary at the nanoscale.[7, 14–17] As such, controlling the particular architecture of an engineered scaffold across the multiple scales plays a vital role in forming biological functions of the target tissue.[18, 19] Different (bio)fabrication technologies have been attempted to construct biological tissues, such as electrospinning, gas-foaming, particulate-leaching, and 3D (bio)printing.[20] Among these technologies, 3D (bio)printing has shown exceptional advantages of precisely controlling the structural and architectural features of the constructed objects over other (bio)fabrication technologies.[21–23] However, biomimetic scaffolds inherently heterogeneous are challenging to fabricate directly with the above-mentioned methods due to limitations of the (bio)materials and the (bio)fabrication technologies. Recent progress of shrinkable and expandable materials, which can alter network spacing resulting in size or shape change, have further effectively extended the potential of 3D (bio)fabrication technologies by offering a complementary strategy over size or shape controls.[24] For instance, a typical microscale 3D-printed scaffold can shrink to sub-micron- or nanoscale-size through a programmed size-shifting behavior. This is the most critical feature distinguishing shrinkable or expandable (bio)materials from others utilized in tissue remodeling.
In this Progress Report, we first summarize in detail the most representative mechanisms of shrinkable or expandable (bio)materials, given from a materials science perspective. We then illuminate the state-of-the-art design strategies of size-/shape-shifting due to changes through pH, temperature, swelling, electrostatic interactions, nanoparticles, or post-crosslinking. Thereafter, we highlight representative biomedical applications enabled by (bio)material shrinking and expansion, with a focus on those fabricated by 3D (bio)fabrication strategies (Figure 1). Afterwards, a perspective towards the future outlook, which will expand this principle to other promising fields, is presented.
Figure 1. Brief summary of mechanisms in (bio)material shrinkage and expansion, as well as representative biomedical applications.
Major biomedical applications of shrinkable and expandable (bio)materials are drug delivery, diagnostic, tissue regeneration, disease modeling, and medical devices.
2. Mechanism of shrinkage/expansion
Shrinkable and expandable (bio)materials open new fields of applications in which a structure can be adapted or designed to be size-/shape-shifting. As such, this ability offers several advantages, such as significant volume increase or decrease.[25, 26] The increased volume allows a more porous structure, which benefits cell infiltration.[27] On the other hand, the opposite example is that instead of directly processing a construct with high resolution through (bio)fabrication, the lower resolution configuration can be formed first and then the structure shrunken to reach the final required high resolution.[28] Currently, considerable research efforts have been devoted to exploring different shrinking and expanding strategies.[29–31] The way to shift the shape or volume will impact the properties of the (bio)printed structure, such as resolution, mechanical strength, and biocompatibility. Therefore, it is important to understand the different mechanisms of shrinkage and expansion, which are divided into multiple categories and demonstrated below.
2.1. Temperature
As one of the external stimulus factors, temperature is perhaps the most widely used inspiration for material size-/shape-shifting due to its easy controllability. An important concept regarding the phase-change of materials is termed the critical solution temperature. When the material has a phase-transition from a soluble status to an insoluble status above this critical solution temperature, it is defined as a lower critical solution temperature (LCST), and by contrast, an upper critical solution temperature (UCST) is characterized if the phase-transition happens in the opposite way; i.e., the material is insoluble when the temperature rises above LCST while become soluble at low temperature (below LCST).[32] The expansion or shrinkage of temperature-responsive materials is the result of this reversible phase-transition. Therefore, the mechanism of shrinkage and expansion induced by temperature can be explained via the competition between intramolecular or intermolecular hydrogen bonding above or below the LCST or UCST.[33]
In general, thermoresponsive materials can be classified into different groups, such as poly(N-vinylalkylamides) and poly(N-alkyl substituted acrylamides).[34] Taking poly(N-isopropylacrylamide) (PNIPAAm) as an example, a well-characterized poly(N-alkyl substituted acrylamide) polymer, it has been widely utilized in many biomedical applications because of its physiologically relevant LCST of 32 °C.[33] As can be seen from Figure 2A, PNIPAAm exhibits a water-swollen, highly open hydrogel network when the temperature falls below LCST. At this moment, the primary intermolecular hydrogen bonding existing between PNIPAAm and water molecules results in the hydrophilicity of PNIPAAm. By contrast, the intramolecular hydrogen bonding between C=O and N-H groups contributes to a collapsed and compressed conformation of PNIPAAm network when the temperature goes above LCST. As a result, the PNIPAAm network releases the bound water leading to a shrunken state of the hydrogel.
Figure 2. The important elements evolved in inducing expansion and shrinking include.
A) pH, B) temperature, C) swelling, D) gas-foaming, E) nanotechnology, F) post-crosslinking.
The fundamental behaviors and related mechanisms of temperature-responsive materials have already been intensively investigated and understood, and meanwhile, various technological applications have been developed. For instance, Agarwal and co-workers reported a fibrous bilayer membrane consisted of a crosslinked PNIPAAm and thermoplastic polyurethane (TPU) layer, which could provide actuation movements through PNIPAAm deformation.[35] As a consequence of expansion or shrinkage of the PNIPAAm layer, this membrane exhibited opening or closing behaviors within a second of being placed in water above or below LCST of PNIPAAm. They demonstrated that this concept could downscale to even single fibers with a side-by-side morphology. With this design, a side-by-side fiber configuration displayed immediate reversible curling motion in buffers at different temperatures. More interestingly, the direction of this motion could also be controlled by the alignment of fibers at particular angles. When the aligned PNIPAAm film experienced shrinking at 40 °C, it twisted in the direction parallel to the fiber direction, leading to a tubular structure with PNIPAAm as the inside layer. Whereas at 4 °C, swelling behavior in the perpendicular direction to the fiber alignment resulted in the opening of helix and rolling diagonally in same direction forming a second tube with PNIPAAm as the outside layer. The complete inside-out movement of tube resulted in the motion with controlled direction.
2.2. pH
pH value-stimulated hydrogels are another major branch of the shrinkable and expandable families.[36, 37] The underlying mechanism is associated with the nature of pendant groups in the network of a hydrogel, which can be ionized and rebalance the density of charged ions within the hydrogel when pH is changed in the surrounding environment.[38] Followed by ion redistribution, alteration of electrostatic repulsion between adjacent hydrogel backbones with the same charge leads to hydrogel expansion or shrinkage. On the basis of this mechanism, the pH-sensitive hydrogels can be divided into cationic and anionic hydrogels.[39] The expansion and shrinkage behaviors of a cationic material, such as poly(ethylene imine) or chitosan, rely on the dissociation of cationic pendant groups within the hydrogel. When the surrounding pH goes below the acid-dissociation constant (pKa) of the hydrogel, amino/imine groups on the hydrogel chains are protonated, making the intrinsically neutralized hydrogel positively charged and thus responsible for expansion. With the migration of charged ions, a concentration gradient of ions is generated inducing hydrogel expansion due to difference in the osmotic pressure.[40] In regards to the anionic hydrogel, such as carboxymethyl chitosan, which shows an expanding behavior at higher pH levels (>pKa) due to the ionization of acidic groups, such as the carboxylic groups. Similar repulsion induced by ionized negatively charged groups within the hydrogel network leads to final expansion. In contrast, the shrinking behavior of pH-sensitive materials happens in the opposite way when the pH is reversely changed (Figure 2B).
A number of studies have demonstrated several types of pH-responsive hydrogels, such as poly(acrylic acid) (PAAc)-based,[39] poly(methacrylic acid) (PMAA)-based,[41–45] poly(vinyl alcohol) (PVA)-based,[46–48] and poly(acrylamide) (PAAm)-based materials.[49–52] Take PAAc as an example, the carboxylic groups on PAAc are dissociated via proton release at higher pH (>pKa), while they are protonated at lower pH (<pKa). Therefore, the material will expand due to repulsion generated by the electrostatic force between anionic networks at higher pH and shrink at lower pH values. Inheriting from this mechanism, PAAc shows significant shape-change at various pH levels, which endows extensive applications. For instance, Jo and co-workers conjugated PAAc to 2-hydroxyethyl methacrylate (HEMA) to form the copolymer p(AAc-co-HEMA) and used it as a pH-sensitive valve for flow control in a microfluidic device.[53] According to the ambient pH variant, this ‘smart’ valve regulated the flow rate inside the microsystem featuring a fast and powerful flow-switch. Along with this pH-sensitive valve design, it is expected that this novel microfluidic system could have potential applications in drug delivery and self-regulated sensors.
Another common application of pH-sensitive materials lies in the drug delivery system. It is well-established that numerous microenvironments of the human body or disease sites are close to acidic. Therefore, shape-change of pH-sensitive materials in these acidic microenvironments offers new possibilities for drug delivery, including systemic delivery with nano/microparticles, localized delivery, and transdermal drug delivery with hydrogels.[54] For example, a pH-sensitive hydrogel formulated with hydroxyethyl cellulose/hyaluronic acid (HECHA) was crosslinked by divinyl sulfone for controlled release of isoliquiritigenin (ILTG) to cure Propionibacterium acnes.[55] At pH 7, as a result of hydrogel disassembly due to shape-shifting, an increase of ILTG release rate with better skin permeability was recorded, which achieved 70% inhabitation of acnes growth. Overall, due to easy controllable in responding to external pH variations, this type of pH-sensitive hydrogels has been applied as a promising platform in biosensor and drug delivery.
One exciting finding demonstrated a pH-sensitive hydrogel synthesized by PNIPAAm or PAAc that could form an ionic complex between the charged hydrogel and the oppositely charged bioactive species.[56, 57] To this end, the cationic hydrogel enabled to form an ionic complex with negatively charged bioactive agents, such as insulin. In contrast, the anionic hydrogel successfully created a complex with positively charged growth factors, such as transforming growth factor-β1. These hydrogels could be used as an injectable drug delivery platform as well, which would show the potential of enabling sustained drug release. Moreover, this strategy was further expanded by us to another field where the constructs were fabricated via the 3D bioprinting technology using charged bioinks and immersing them into corresponding polyion solutions of the opposite net charges.[28] The results indicated that rapid shrinking occurred rendering high-resolution constructs with reduced volumes.
2.3. Swelling
It is well-known that a hydrogel is commonly formed via chemical or physical crosslinking and possesses a relatively stable 3D network structure.[58] When placed into a thermodynamically compatible solvent, swelling is the fundamental behavior of the hydrogel.[59] At the beginning of contact with water molecules, these molecules interact with hydrogel chains and penetrate into the network (Figure 2C). As such, the unsolvated glassy phase is broken away from the sturdy hydrogel area with a moving barrier. This hydrogel network, in the rubbery phase, will expand and allow more water molecules to fill in the hydrogel network. This mechanism has been well-established, and a new method has been developed to visualize the dynamic deformation in real-time during the hydrogel swelling process.[60] On the basis of caged photoactivated fluorophores covalently attached to the gel network, the swelling of the hydrogel in a constrained geometry was monitored. With the aid of this technology, it was concluded that this swelling process was a continuous movement until reaching the boundary. When the osmotic pressure and elasticity force from the hydrogel network reached an equilibrium at a certain volume of the gel, the stretching of the hydrogel network stops, and the status is maintained.
A number of studies have reported hydrogel expansion through swelling.[29, 61, 62] For example, Tibbits and co-workers demonstrated a linearly expanding hydrogel/polymer composite stimulated by water.[63] This expanding construct was fabricated by 3D printing and contained alternating materials between rigid polymer discs and swelling hydrogels. After the construct was immersed in water, the hydrogel parts swelled and then expanded while the rigid discs stayed the same. The desirable length of this type of constructs could be regulated by adjusting the ratio of the swellable hydrogel segments to the sturdy polymer discs.
2.4. Gas-foaming
The conventional gas-foaming technology generates porous hydrogel structures through mixing with gas-containing or generating solutions. With the occupied space of polymer by gas, the whole volume of polymer is expanded. The process of gas-foaming within the polymer is divided into three steps: formulation of gas/polymer solution, nucleation of gas pores, and volume expansion of the gas bubble.[64] The key development of this technology is the formulation of gas/polymer solution, which is governed by the distribution of gas within the polymer matrix (Figure 2D). However, the gas-diffusion process usually is slow and typically takes multiple cycles to achieve a desirable distribution. Therefore, a variety of technologies have been focused on increasing the diffusion rate, such as increasing gas pressure or improving solution temperature.[65] Both methods have allowed the enhancement of gas diffusivity; thus, decreasing the diffusion time and cycles of gas. Another modified gas-foaming strategy, developed by Xie and co-workers, attempted to expand electrospun nanofiber mats to volumetric and porous structures through treatment with aqueous solution of sodium borohydride (NaBH4).[27, 66, 67] This modified gas-foaming technology allowed uncontrolled out-gassing and microbubble-formation from the aqueous NaBH4 solution. The resultant nanofibrous mats exhibited a highly porous layered morphology while maintaining the original nanotopography. It should also be pointed out that these structures with controllable gap widths and layer thicknesses could be regulated by different treatment times in the aqueous NaBH4 solution, as well as its concentrations.
Apart from chemical reactions in an aqueous system, other strategies such as ultrasonication are also explored to expand the polymer scaffolds.[68, 69] For instance, Kwon and colleagues developed the electrospun nanofibrous scaffolds and further expanded them by an ultrasonication process in the aqueous system.[69] These ultrasonicated nanofibrous scaffolds exhibited increased porosity and therefore, allowed excellent cell infiltration. Although the above methods allowed the expansion of nanofibrous scaffolds with high porosity and cellular infiltration, there was a risk of losing bioactive materials encapsulated in fibers during the gas-production process in an aqueous solution. Herein, Xie and co-workers conducted another expansion strategy via depressurization of subcritical CO2 fluid, whereby the CO2 liquid phase changed to the gas phase when the pressure was reduced rapidly.[70] As the liquid CO2 that penetrated the fibrous matrix transferred into gas bubbles, the electrospun nanofibrous scaffolds expanded dramatically. Compared with previous methods, this approach offered a complementary strategy in expanding bioactive materials-loaded nanofibrous scaffolds.
2.5. Other strategies
Nanoparticles have been utilized as drug delivery carriers for a long time due to their biocompatibility, targeting properties, and loading efficiencies.[71] For coordination polymers and porous coordination networks, nanoparticles also play a crucial role in regulating the network structures.[72] Kitagawa and co-workers proposed an expanding and shrinking pore-modulating mechanism by embedding pillared ligands ({[Cd(pzdc)(bpee)] 1.5 H2O}n) into the porous coordination polymers network.[73] As shown in Figure 2E, the expansion of the porous coordination polymers network resulted from electrostatic repulsion between oxygen, carboxylate, and nitrogen atoms, as well as azo groups. By contrast, hydrogen bonding after dehydration was responsible for the shrinkage. Another exciting strategy is using deformable nanoparticles to enhance hydrogel mechanical properties after expansion. Similar to the previous example, liposomes were embedded into a crosslinked hydrogel network as reinforcement factors.[74] After hydrogel swelling, the encapsulated molecules were released from liposomes and subsequently induced the secondary hydrogel network-formation between the original hydrogel network. Therefore, this rigid double-crosslinked network enabled enhanced mechanical properties compared with standard swelling hydrogels, which usually suffered from the swelling-induced weakening behavior. To this end, the existence of nanoparticles plays an essential role in forming expanding or shrinking behaviors by generating or removing spaces within hydrogel networks. Other shape-shifting capacities in complex structures based on this principle can be reasonably extrapolated to achieve additional applications.
The aforementioned strategies have shown their efficiencies in enabling shrinkable and expandable (bio)materials with the help of network reconstruction. Crosslinking ratio variation is another useful approach in designing the shrinkable and expandable (bio)materials. During photocrosslinking of a hydrogel, nonuniformly crosslinked networks are commonly found due to the nonuniform light density distribution. After photocrosslinking, the partially crosslinked hydrogel network is formed but extra crosslinkers or free monomer chains still exist within the hydrogel.[75] The former will result in the network becoming tighter after the post-crosslinking with remained crosslinkers, resulting in shrinkage of the post-crosslinked areas. The latter influences the free monomers within the hydrogel network to be removed after immersion into a solution, termed the desolvation process.[76] Following this desolvation process, the hydrogel volume shrinks as a result of the free monomer removal, which should be equivalent to the volume of the free monomers. The nonuniform swelling results in different expansion ratios in the network, while the nonuniform desolvation leads to varying shrinkage degrees (Figure 2F). The spatial distribution of crosslink densities can be regulated through managing the light patterns in the photopolymerization process. Based on this fact, studies have been demonstrated to transform 3D hydrogel films into complex 3D constructs upon shrinking or expansion due to different levels of photocrosslinking within single hydrogel pieces.[77]
3. 3D/4D fabrication and applications
3.1. Biomedical applications of shrinkable materials
3.1.1. Regenerative medicine and tissue/disease modeling
The 3D (bio)printing technology has been widely applied to in vivo tissue regeneration or in vitro tissue modeling.[78, 79] However, the producible tissue constructs sometimes lack practicality due to the resolution limitations. For instance, the resolution of extrusion bioprinting with hydrogels as bioinks is usually as large as sub-millimeter. To overcome this obstacle, different approaches have been devised to achieve more significant or convenient resolution-enhancement.[80] Very recently, we reported a new approach termed shrinking bioprinting via post-treatment of bioprinted structures.[28] In a typical example, an anionic hydrogel methacrylated hyaluronic acid (HAMA) was chosen as the bioink for 3D bioprinting and then post-neutralized by a cationic solution of chitosan. We first fabricated a cylindrical HAMA hydrogel as a proof-of-concept and immersed it into a 2.0% (w/v) chitosan solution for 24 h. This method enabled the hydrogel to shrink about 61% in height and diameter and resulted in 21% of final volume compared with the original hydrogel (Figure 3A). It was also validated that different anionic hydrogels, including gelatin methacryloyl (GelMA) and alginate, showed the same behavior. As a result of shrinking, the Young’s modulus of 2.0% (w/v) HAMA increased from 15.9 ± 1.8 kPa to 24.3 ± 3.0 kPa. It might be caused by the condensation of HAMA within the same volume after shrinking. To further explore this concept, microchannel-embedded HAMA hydrogel constructs were fabricated through the use of sacrificial printing; where the initial diameter was approximately 355 μm after selectively removing the pluronic fugitive inks. Interestingly, the diameter of the microchannel shrank to 174 μm after immersion in the 2.0 w/v% chitosan solution, by which the enhanced resolution was achieved. Similar to the previous strategy, melt electrospinning-printed polycaprolactone (PCL) mesh-enabled microchannel structures embedded in HAMA hydrogels were also investigated. Evidence suggested that the fabricated microchannels shrank from 39 μm to 10 μm under the same shrinking condition, approaching the single-capillary size (Figure 3B).
Figure 3. (Bio)fabrication of tissue constructs based on shrinkable materials.
A) Photograph showing the size change of a HAMA hydrogel before (bottom) and after (top) shrinkage, as well as hydrogel diameter and volume change before and after shrinkage. B) Schematic of the sacrificial printing processed to fabricate microchannel patterns before and after shrinkage, and the diameter changes before and after shrinkage. C) Schematic of coaxial printing, the fabricated cannular structures, and their subsequent shrinkage. D) Live (green)/dead (red) staining and viability results of MCF-7 cells in GelMA/HAMA constructs before and after shrinkage. Reproduced with permission.[28]
One advantage of coaxial extrusion bioprinting is constructing desired cannular tissue structures through concentrically assembled nozzles.[81–83] This distinct process offers exciting opportunities to generate, for example, alternative vascular grafts. However, there is an urgent need, which yet is a significant challenge, to apply this technology to the production of small-diameter blood vessels.[84] Following this rationale, we coaxially bioprinted HAMA/alginate tubes with different concentrations of HAMA (from 0.5% to 2.5%). After soaking in a 2% chitosan solution, the greatest amount of shrinkage of the inner tube diameter was from 670 μm to 90 μm (Figure 3C). In addition, smaller tubular structures were fabricated with the aid of smaller nozzles, where smaller diameters ranging from approximately 30 μm to 150 μm were obtained after shrinking; thus, indicating the potential physiological relevancy in engineering small-diameter blood vessels or other cannular tissues of similar size ranges.
Considering that such a technology would be used in biomedical applications, the cell responses after shrinkage was of importance for this development. The viability of several cell types, including human umbilical vein endothelial cells (HUVECs), mouse skeletal muscle cells (C2C12), and human breast cancer cells (MCF-7), was evaluated after shrinking (Figure 3D). The results indicated that the volumetric shrinkage showed no significant influence on both MCF-7 and C2C12; however, HUVECs exhibited much higher sensitivity to the structural change. In addition, the cells response to different shrinking methods were also explored: single shrinking and successive shrinking. The viability results revealed that there was a dramatic decrease in cytotoxic with the successive shrinking approach. Overall, several proof-of-concept studies of different bioprinting-shrinking methods led to the conclusion that this unique shrinking bioprinting strategy is potentially broadly applicable and cell-friendly. In summary, this simple strategy also makes it possible to enhance the resolution of bioprinted hydrogel structures without necessarily having to improve bioprinter hardware or other parameters.
3.1.2. Diagnostic applications
The growing development of diagnostic devices, such as wearable sensors with multiplexed functions, allows for direct sensing functions and accurate operability.[85–87] However, an important question regarding the diagnostic portability and sensitivity remains to be solved. For instance, most diagnostic devices require samples or patients to be tested in a centralized laboratory; this in turn greatly diminished the diagnosis promptness, possibilities, or even accuracy, in resource-limited settings.[88, 89] Therefore, a smaller volume without losing required functions will present a wide range of possibilities for their further developments. To improve the flexibility of wearable devices, several approaches have been explored using size/shape-shifting materials.[90, 91] Very recently, Khine and colleagues reported a shrinkage-induced electrode by integrating a shape-memory polyolefin film with a thin metallic sheet.[92] After thermal shrinking, the thin electrode shrank more than 95% in volume, and more importantly, it provided a higher resolution as well as better conductivity; the former being increased by 20 times, while the latter proved an improvement of 600% (Figure 4A). It is worthwhile mentioning that these achievements were better than using the photolithography method by itself or other methods which employ shape-memory materials. The combination of shrinking materials and electrodes enable the further advancements of diagnostic medical devices, especially those used for flexible electronics. With this in mind, the same group developed a similar strategy to the production of multiple medical/biomedical devices, such as microfluidic chips,[93, 94] wearable devices,[95–97] and cell culture tools.[98–100] For instance, they reported the fabrication of a wearable strain sensor by embedding a wrinkled carbon nanotubes (CNTs) thin film with nano- and microstructures into Ecoflex and then shrinking into smaller size with the previously described method.[101] This wrinkled CNT thin film could be integrated with a soft stretchable elastomeric substitute without losing its original conductivity. As a result of the shrinking process, the sensor’s resolution for the measurement of strain was increased by more than 750% resulting in an ultrahigh sensitivity biosensor (Figure 4B). Therefore, it had enough sensitivity to perceive motions or even detect lesions when integrated to the joint sites of the knee, elbow, finger, or ankle.
Figure 4. Examples of shrinkable materials used in diagnostic devices.
A) Thermal-induced shrinkage of thin-film gold electrodes. Reproduced with permission.[92] B) i) Photograph of the CNT thin film at each step of process: as-deposited, shrunken, and the final wrinkled CNT thin film on a PS substrate. ii) Images of the strain sensor and the relative changes in resistance after stretching. iii) Monitoring human movement with the wrinkled CNTs-Ecoflex electrode. Reproduced with permission.[101] C) SEM images and brightfield reflection-mode optical micrographs of the woodpile photonic crystal with different shrinkage degrees under heating. D) i) Optical images of heated woodpile photonic crystals with different dimensions. ii) 3D-printed color constructs of the Eiffel Tower and 3D Chinese character “福” in red color. Reproduced with permission.[147]
In another study, Yang and coworkers developed photonic crystal (PC) structures, which reflect colors through the optical interference effects, via the 3D-printing method.[102] The wavelength is a function of crystal lattice spacing; therefore, as the matrix expands or shrinks in response to an outside input, the color shifts.[103] They first introduced a shrinking method to precisely control the reflected color by controlling the heating time. Results showed that the matrix could shrink to 55%, 71%, and 78% of the original volume after 12, 17, and 21 mins of heating at 450 °C (Figure 4C). At the same time, the reflectance of PCs transitioned from: no color, to yellow, to blue, and purple; each with their respective linear shrinkage. To extend the operational window of these devices to the UV-visible spectral range (from 100 nm to 700 nm), the lattice constants of the PCs should be decreased correspondingly; meanwhile, the constituent material must have as high of a refractive index as possible. Their findings established that the lattice constants were as small as 280 nm after shrinking, comparable to the finest periodicities in butterfly scales and two times smaller than the machine specifications. These findings established the possibility of reaching ultrahigh resolutions using 3D printing in concurrence with a standard method. Furthermore, the results also demonstrated evidence of enhancing the refractive index of crosslinked materials during the heating process, which exhibited a high efficiency in color generation. In addition, a further investigation focused on fabricating complex 3D structures at the microscale level using the coloring-by-shrinking method (Figure 4D). The printed Eiffel Tower could be generated as small as 39 mm in height, which enabled precise control of the wavelengths and polarization of the output light. This coloring-by-shrinking method was particularly attractive as an alternative way to overcome the resolution limitation when printing arbitrary colors and structures in all dimensional directions at the microscale. With further advancement of the technology, more attractive colorimetric constructs at the microscale would provide a powerful platform for integration with other devices.
In both approaches, fabrication and shrinkage of functional patterns towards a small scale were realized, allowing for diagnostic functions with greater accuracy and operability. Furthermore, enormous efforts have been made to construct sensors based on textiles and fabrics, which are ideal substrates for wearable applications.
3.1.3. Nanofabrication
Applications of nanotechnology have been investigated in drug and gene-delivery, bioimaging, and biomedical implants, among others.[104, 105] For example, Boyden and colleagues invented a strategy of 3D nanofabrication, termed implosion fabrication (ImpFab), by which 3D structures of nanoscale were achieved.[106] They chose polyacrylate/PAAm as the scaffolding material, which could be reduced into the nanoscale size by acid- or divalent cation-induced shrinkage. Due to the reaction of radicals generated from fluorescein photobleaching, the activated fluorescein molecules were crosslinked to the active acrylate groups within the hydrogel.[107] In ImpFab, fluorescent molecules carrying DNA, proteins, small molecules, or nanoparticles were deposited into the hydrogel matrix through two-photon lithography (Figure 5A). After the shrinking and dehydration, the functional 3D nanostructures were obtained, showing a 10-fold shrinkage in the linear dimension. Notably, multiple materials were able to be independently deposited in single constructs.
Figure 5. Nanofabrication relying on shrinkable materials.
A) Schematic of the fabrication process and the corresponding images, including patterning, deposition of small molecules (e.g., proteins, DNA, or nanoparticles), intensification of growing silver (blue) onto the gold nanoparticles (yellow), and shrinking. B) Nanofabricated metal structures. i, ii) SEM images of 2D structures in micrometer-scale resolution before and after sintering iii) Shrunken structures with 100-nm feature size of each line. iv-vi) Images obtained by the maximum-intensity projection of a 3D structure (iv) before shrinking, (v) after silver deposition, and (vi) after shrinking. Reproduced with permission.[106]
Considering that metal constructs with nanoscale features are required for applications such as nanophotonics, plasmonics, and metamaterials, the ImpFab method was applied to fabricating a conductive silver structure. The said construction was patterned with a 128 mW of laser power and sliver was deposited into the gold nanoparticle-conjugated hydrogel matrix (Figure 5B). The shrinking was conducted in 2-mM HCl with 0.05 v/v% Tween-20 for 6 h and again for another hour, followed by washing in 2-mM HCl without Tween-20 for 30 min. Nonetheless, initial results showed that the structure failed to show the conductive capacity since the sintering by plasma treatment, heating, and electrical discharge was not successful. Interestingly, we observed morphological changes similar to sintering through low-power irradiation (15 mW of laser power) of the samples. The shrunken silver wire showed a resistance of 0.38 Ω, which was 13.3% of the bulk silver conductivity. Using the same material, 3D metal structures were fabricated, varying the dimensions from several to hundreds of micrometers; at the same time, the shrinking and sintering did not alter the scaffold geometry. A set of metal structures was discontinuously patterned with different angles, and they retained the relative spatial relationships after shrinkage.
To extend ImpFab into the other applications, the fluorescent deposition could be replaced by other chemistries in the future, such as the DNA-addressed material deposition. Additionally, instead of using the two-photon lithography method, that limited the fabrication speed to 4 cm s−1, the ImpFab technique may also be applied to the production of larger-volume nanostructures with other faster patterning strategies. For example, a method of ultrafast two-photon polymerization was developed by Chen and co-workers.[108] Instead of using the sequential laser scanning process (speed at 0.1 Hz) of the typical two-photon lithography, the ultrafast random-access digital micromirror device scanner was introduced to control and generate multiple foci for parallel nanofabrication at the speed of 22.7 kHz. Moreover, to solve the issue of a limited working distance in the conventional two-photon polymerization and to extend the capability of large-scale fabrication, Chichkov and colleagues demonstrated a modified system with a broader working range.[109] The microscope objective, cover glass, and immersion oil could move together into the liquid material. They showed the fabricated 3D structure with a height of 7 mm, which was greater than the working distance of 170 μm that the microscope objective had.
3.1.4. 4D printing with shrinkable materials
Tibbits initially introduced the definition of 4D printing in 2014 as “3D printing + time”, which over the last few years has evolved as the shape, property, and functions of a 3D-printed construct that could change over time when it was stimulated by water, heat, light, or pH.[110–112] Smart materials are the fundamental elements of 4D printing, which can be categorized into various classes, such as shape-memory, self-sensing, self-adapting, self-repairing, and decision-making materials.[113, 114] On the other hand, 3D printing allows the building of target objects with several stimulus-responsive materials in a proper geometry, resulting in the shape-shifting behavior required for 4D changes.[115, 116] The shape-shifting behaviors in 4D printing can be generated from one-dimensional (1D) to 1D/2D/3D, 2D to 2D/3D, and 3D to 3D through the folding, bending, twisting, surface-curling, shrinking or contracting, and/or expanding processes.[117]
Qi and co-workers demonstrated the shrinking of constructs from 1D to 1D in a linear shape-shifting manner.[118] The shrunken structure was obtained by employing thermoresponsive shape-memory polymers and stimulating by the programmed heating cycles. Yet, the regular shape-memory cycles required the external force for maintaining the previous applied force in the shrinking. To overcome this problem, Hu and colleagues invented a combination of thermoresponsive materials with different expansion coefficients to create a flower-like architecture by transforming the planar sheet into a 3D shape.[119]
Needless to state, forming complex and double-curved shapes was challenged by the shrinkage of single materials and simple patterning since it required the growth gradients of the in-plane level and the thickness direction. With this in mind, Lewis and co-workers developed a multi-material 4D printing platform for creating curved geometries.[120] The ink was composed of polydimethylsiloxane (PDMS), short glass fibers, and fumed silica; wherein PDMS served as the base elastomer to bind the structure ribs and layers, The short glass fiber additive aimed to decrease the thermal expansivity, and the fumed silica was added to promote the rheological properties of the ink. By adjusting the thermal expansion coefficient of the ink with the tunable crosslink density, the shape-shifting behavior could be precisely controlled. The generated ink expansion coefficient (α) ranged from 32 × 10−6 to 229 × 10−6 °C−1, and the elastic modulus E varied from 1.5 to 1,245 kPa (Figure 6A). As shown in Figure 6B, the thermal response of the materials and the cross-sectional shape of the layers determined the curvature response of the printed bilayer structure. It should be pointed out that the curvature changing was reversible and repeatable when exposing the constructs to thermal cycles; however, the maximum in-plane growth achieved by the simple bilayer printing was limited to +6.4% when applying the temperature change of 250 °C. Therefore, an open-cell lattice was created as the bilayer structure, which presented the shape-shifting ability from 79% of shrinkage to 41% in expansion with 2 × 2-cell printed lattices. To further achieve spatial growth gradients, a heterogeneous lattice was applied for transforming the flat sheet into a spherical shape, including the cap-like and saddle-like structures. The system showed the multi-shape potential, i.e., when exposed to the decreased temperature it formed a cap-like shape, but immersing in a solvent (hexane) generated the saddle-like structure. To fully control the curvature, the bilayer lattices were printed using four different materials that enabled the generation of five levels of curvatures (Figure 6C). Finally, they utilized this multi-material printing method to produce the patch antenna with a liquid metal-containing eutectic gallium indium ink forming the human face-shape featuring complex geometry (Figure 6D–E), which would be challenging to obtain by the traditional shape-shifting behavior.
Figure 6. Shape-shifting in 4D printing controlled by multi-material shrinkage.
A) Changes of coefficient of thermal expansion and elastic modulus with varied ratios of crosslink to base. B) 2D lattices (i) printed with multi-material method, and (ii) their linear growth factors. C) Schematics of material combinations in the cross sections of multiplex bilayer lattices. D) The shape-shifting patch antenna (spherical cap shape) fabricated with the liquid metal ink. E) Design and transformation of a 3D-printed face shape. Reproduced with permission.[120]
3.2. Biomedical applications of expandable materials
3.2.1. Tissue regeneration
Tissue regeneration typically requires alternative scaffolds to replace the defect tissues and remodel the defect and surrounding tissues. An important point regarding remodeling of native tissue structures is their normally inherent heterogeneity and often complex physiological architectures. For example, many tissues are highly porous with interconnected pore networks that help to facilitate: cell growth, oxygen and nutrient exchange, as well as waste elimination. Hence, the substitute should be structurally similar, but functionally equivalent to the native tissue to be replaced. Given the functional mimetic and cell-friendliness requirements, a porous structure is commonly a fundamental strategy in designing replacement scaffolds.[18] To achieve this goal, various fabrication methods of producing replacement scaffolds have been attempted the last several decades, such as electrospinning, particle-leaching, and 3D printing.[121] Among these, the electrospinning technology offers significant advantages to fabricate highly porous scaffolds with interconnected networks. In addition, further modification of this technology, such as coaxial electrospinning, has been made to enhance the porosity of the scaffolds.[122] By far, unfortunately, most of the architectures fabricated by the electrospinning technology feature tightly packed nanofiber mats with extremely small pore sizes. Recently, a few examples have been reported to improve cell infiltration and tissue remodeling within these mats with simple yet effective modifications.[67, 70]
One such study investigated a gas-foaming approach to expand packed electrospun nanofibrous membranes into truly 3D structures along the z-direction so that the seeded cells had enough space to grow and migrate; hence, benefiting tissue regeneration.[70] Considering the mechanism that we previously discussed in Section 2.4, this gas-foamed electrospun scaffold showed a separated-layer morphology in the z-direction after 24 h of treatment with a 1-M NaBH4 aqueous solution. The thickness of the resultant was almost 36 times larger than that of the pre-expanded samples and 83.6% more porous than the original scaffolds (Figure 7A). As one of the key factors for scaffold development, cellular behaviors were evaluated by NIH/3T3 fibroblasts. The results revealed that the cells infiltrated and proliferated more in the expanded scaffolds when compared to the unexpanded ones (Figure 7B). Meanwhile, cellular morphologies were consistent on both expanded and unexpanded scaffolds, which indicated the retention of imparted anisotropic cues. Further work from the same group focused on the evaluation of host responses to expanded (bio)materials with the same strategy.[67] The results suggested more macrophage infiltration, and more M2-phenotype cells were found in the expanded scaffolds (Figure 7C). These findings demonstrated the positive influence on immunomodulation and enhanced new blood vessel-formation within the expanded scaffolds. Taking the above observations into account, it was concluded that expanded scaffolds were promising candidates for use as tissue-regenerative scaffolds with anti-inflammatory properties.
Figure 7. Examples of using expandable materials in tissue regeneration and modeling.
A) Expansion and characterizations of electrospun nanofibrous scaffolds before and after expansion. B) Hematoxylin-eosin staining of NIH/3T3 cells after 7 days of culturing on aligned and random PCL nanofibrous scaffolds before and after expansion. C) Quantification of immunobiological analysis of M2 macrophages within non-expanded and expanded PCL scaffolds in the subcutaneous implantation. Reproduced with permission.[27] D) i) Fabrication and expansion process of a sacrificially printed paper device. ii) Schematic and confocal images showing HUVEC and MCF-7 cell distributions within the printed paper devices non-expanded and expanded, as vascularized breast tumor models. Reproduced with permission.[123]
Another exciting application carried by our team demonstrated a microchannel-embedded paper device[123] with an expansion method. We first fabricated a perfusable microchannel structure surrounded by compacted bacterial cellulose nanofibrils using matrix-assisted 3D printing (Figure 7D).[124] Then, the precisely controlled gas-foaming strategy with different concentrations and immersion time of NaBH4 successfully maintained microchannel integrity with more surrounding porous structures, resulting in a better spatial cell infiltration and interactions. This study offered a complementary strategy in establishing an enabling vascularized tissue model and exhibited great potential for applications in point-of-care preclinical drug screening. Overall, further development of the gas-foaming technology is underdoing and continues to be a great impetus to the fields of materials science, regenerative engineering, and microenvironment modulation.
3.2.2. Drug delivery
Hydrogel materials have been used in drug delivery for many years owing to numerous advantages over other materials.[56] For instance, the intrinsic network structure of hydrogel allows targeted drug loading inside. Besides, abundant functional groups on hydrogel chains offer available conjugation opportunities for immobilizing loaded drugs. More importantly, hydrogels have been proven to be very biocompatible in many clinical studies compared with chemical or metal materials. Lastly, adjustable degradation of hydrogels plays a vital role in applying this type of material for in vivo applications. With the emergence of size-/shape-shifting hydrogels, such as shrinkable and expandable hydrogels, they exhibit such an excellent deformation that elevates their applications due to precisely controlled release with or without external stimulation. The general provocations of hydrogel shrinking and expansion for drug delivery purposes, are closely related to the surrounding environment, such as pH and temperature.[125] It should also be pointed out that the swelling nature of general hydrogels has the ability to release loaded agents from the hydrogel network. For example, the swelling property of hydrogels used for drug delivery has been reported by Magdassi and colleagues, where they fabricated different constructs loaded with sulforhodamine B by utilizing the digital light processing printing approach.[126] After immersing them in phosphate buffer for 24 h, the constructs with different shapes exhibited 3–15 times of expansion over their original sizes, and released all loaded dye molecules to the surrounding solution (Figure 8A). They also observed that the variation of expanding size was indispensable to the ratio of surface area to volume, which was different from 3D cube, cylinder, mat, sphere, or pyramid structures printed. Thus, those findings led them to conclude that the morphology of the constructs with complex geometries played an essential role in the swelling and drug release behaviors.
Figure 8. Expandable materials applied to drug delivery applications.
A) Photographs of 3D-printed constructs loaded with sulforhodamine B before and after 24 h of swelling in PBS (pH=7.4). Reproduced with permission.[126] B) Schematics of pH-responsive PMAA and chitosan copolymer for drug delivery applications. C) i) Confocal image of B16.F10 cells after 2 h of incubation with the hybrid gel. ii) Drug release profile under different pH conditions. iii) B16.F10 cell viability after co-culturing with the drug-loaded gel under different concentrations. Reproduced with permission.[127] D) Schematic of gold nanocage coated with the thermo-responsive copolymer PNIPAAm and their characterizations. E) Controlled drug release profiles of both i) anti-tumor drugs and ii) enzyme from gold nanocages coated by PNIPAAm. (iii) Cell viability of samples with and without laser irradiation. Reproduced with permission.[131]
Shrinking or expansion behavior of hydrogels in response to biophysical stimuli, such as pH, has witnessed considerable applications of drug delivery systems. Wu and co-workers developed a hybrid hydrogel (chitosan-poly(methacrylic acid), chitosan-PMAA) immobilized with quantum dots, which showed excellent expanding performances at different pH values (Figure 8B–C).[127] This expansion phenomenon was attributed to the redistribution of electrostatic repulsion resulting from the internal osmotic pressure. As a practical implementation, the phase-volume transition behavior of the hybrid nanogel significantly influenced the charged loading agents to release with variant pH into the surrounding environment. The release results implied that temozolomide as an excellent anti-tumor drug, was released faster in the acidic conditions at the tumor sites than in the natural environment. As a result, this delivery system exhibited remarkable anti-tumor efficacy, with a decrease of over 70% of B16.F10 melanoma viability after 24 h of incubation. In addition, the loaded quantum dots, which were anchored on the hydrogel chains, offered the opportunity to bioimage tumor cells, indicating the success of cellular diagnosis. Therefore, this pH-sensitive hydrogel presented satisfying anti-tumor effects towards melanoma cells, associated with excellent diagnostics-integrated functionalities.
Temperature-responsive hydrogels, as another stimuli-responsive drug delivery platform, play a pivotal role in controllable drug delivery. For example, PNIPAAm-based materials featuring size-/shape-shrinkable and expandable properties result in drug release.[128] Various biomedical applications have been carried out in drug delivery since their volume-transition temperature (LCST of 32 °C) is close to human body temperature (37 °C), which could spontaneously activate the release of loaded drugs when injected into the body.[56, 129, 130] It should also be highlighted that PNIPAAm has remarkably easy accessibility without any other external stimuli to trigger drug release. Accordingly, Xia and co-workers presented the effective release of doxorubicin from PNIPAAm-coated gold nanocages under heating.[131] They loaded the drug into the cores of gold nanocages, which are widely used as photoinitiators in photothermal therapy to generate proper heating to local tissues (Figure 8D–E).[132] When the temperature went up under laser irradiation, the polymer coating layer (PNIPAAm) shrunk leading to drug release because of the collapse of the hydrogel network and exposure of the pores on the gold nanocages. This release process was so fast that it showed a burst release profile within 1 min after heating and it took just 10 mins to release all the loading drugs.
Although various strategies of applying PNIPAAm for drug delivery applications have been developed and studied, uncertainties remain as to their insufficient biocompatibility and imprecise controllable release. Since most of the studies of these materials are used with living tissues and cells, low cytotoxicity is of importance in drug carrier selection. Many attempts have been made to address this issue, such as conjugation PNIPAAm with superior biocompatible natural materials, such as silk or chitosan.[133] With the contribution from these natural materials, PNIPAAm copolymers presented enhanced biocompatibility to fibroblasts and tumor cells without affecting the temperature-sensitivity. Therefore, all these valuable hallmarks make shrinkable and expandable (bio)materials as powerful platforms for loading and delivering drugs and are broadly applicable to a variety of systems and conditions.[134] Despite the fact that these materials have been proven promising and useful, very little optimization work for using them in clinical studies has been carried out. Similar to other novel functional materials, most investigations so far reported, are only limited to the laboratory level. More in-depth characterizations should be encouraged for the use of temperature-sensitive materials in translational drug delivery applications.
3.2.3. Imaging tool
Contrary to the ImpFab process, we first applied the swellable polymer to expand cell and tissue structures physically, facilitating cell and tissue imaging results.[135] It was termed as expansion microscopy (ExM), a process combining labeling, gelation, digestion, expansion, and microscopic imaging (Figure 9A). After expansion within water, nearly 4.5-fold linear expansion was finally recorded as can be seen from Figure 9B. Particular emphasis was placed on the transparent post-ExM constructs that were simultaneously achieved, eliminating scattering under microscopy. A trifunctional fluorescent label was developed that was integrated into the polymer networks to make the expanded samples visualizable (Figure 9C). The custom label contained the methacrylate group that could participated in the polymerization. Meanwhile, by adding the oligonucleotide, the label could attach to the secondary antibody of biological samples through complementary sequence. Then, the targeted samples were infused with the mixture of monomer sodium acrylate, comonomer acrylamide, and the crosslinker (N, N’-methylenebisacrylamide). Ammonium persulfate initiated the polymerization of the above monomers with the aid of tetramethylethylenediamine. Once the samples were fully integrated with polymer networks, the protease-digested tissue-polymer composites achieved desirable mechanical homogenization.
Figure 9. Expansion microscope.
A) i) Schematic of the collapsed polyelectrolyte network and (ii) its structure after expansion in water. B) Photographs of a mouse brain slice before expansion (i) and after expansion through ExM (ii). C) Schematic of the custom fluorescence label and its anchoring to cell microtubules. D) Images of a mammalian brain slice showing the presynaptic (anti-Bassoon, blue) and postsynaptic (anti-Homer1, red) markers (i and iii) before expansion and (ii and iv) after expansion. E) 3D super-resolution image of a mouse brain hippocampus after expansion. Neurons (green) and synapses markers (Bassoon in blue and Homer in red). Reproduced with permission.[135]
In the fixed human embryonic kidney (HEK) 293 cells, the post-ExM samples were imaged for microtubules. Compared to the detail-missing results of pre-ExM cells visualized by a super-resolution structured illumination microscope, the distinguished individual microtubules could be conveniently observed after ExM under the spinning disk confocal microscope. The effective resolution was measured around 60 nm, which was equal to the theoretical lateral resolution of the confocal system used (250 nm) divided by the expansion fold (4.5). In addition, the ExM was implemented to image the neurons and synaptic markers of the sliced brain tissue (Figure 9D). To observe the cytosolic yellow fluorescent protein in neurons and presynaptic Basoon and postsynaptic markers Homer1; they further developed the multicolor ExM through synthesizing the labels with different colors. ExM also enabled to image scalable objects, for instance, a sliced mouse brain (500 μm × 180 μm × 100 μm of length, width, and thickness, respectively) was scanned with the fast diffraction-limited microscope (Figure 9E). Neurons in the CA1 stratum lacunosum moleculare showed spine morphology, on which they observed the postsynaptic marker Homer1. More importantly, the invagination of spiny excrescences into the mossy fiber bouton could be recognized, whereas it could only be observed with the facilitation of electron microscopy as previously reported. Therefore, the ExM could potentially be suggested to visualize the nanoscale structures within multiscale samples, especially valuable for investigating the neural circuits and other tissues with intricate hierarchical structures.
With the emergence of the ExM strategy, other related expansion approaches have also been proposed. We demonstrated iterative expansion microscopy (iExM) to expand the pre-ExM specimen further. The first swellable polyelectrolyte gel network was generated based on the original ExM protocol using the chemically cleavable crosslinker, which allowed the first gel could be dissolved and open space for the second round of swellable polymer embedment and expansion. The iExM enabled ~4.5 × 4.5 or ~20× of final expansion and achieved 25-nm resolution of cell and tissue imaging using the conventional microscope.[136] Protein-retention ExM (proExM) was carried out by anchoring proteins to the swellable polymer, such as fluorescent antibodies and proteins.[137] These findings raised the prospects of imaging native proteins instead of using custom-made labels in the original ExM. We further explored the application of expansion microscopy on RNA imaging, which was defined as ExFISH.[138] By covalent conjugating RNA to the swellable polymer, fluorescent in situ hybridization (FISH) was used to image the post-expansion specimen. They achieved super-resolution RNA imaging indicating both RNA structure and location of the intact brain tissue or other thick tissues. Although the modified methods have been investigated to image proteins and RNA, other types of molecules, including lipids, glycans remain challenging to image using ExM. To expand the potential of ExM on the versatile biomolecules, we described the click-ExM, combining the click labeling of various biomolecules with the ExM, which enabled to image the proteins, nucleic acids, lipids, glycans, and small molecules of cells and tissues.[139] Furthermore, the expansion pathology (ExPath) was innovated to optimize clinical specimens in order to be accurately observed by the ExM.[140] ExPath could be utilized to visualize human tissue samples even those samples has been treated through formalin fixation, paraffin embedding, fresh-frozen tissues, or even staining with hematoxylin and eosin. For example, the post-ExPath sample confirmed the tertiary podocyte foot changes in kidney disease, promoting diagnostic imaging with high resolution. As a crucial tool for neuroscience and development studies, zebrafish, was also evaluated by ExM, tracing the fine structure of radial glia, visualizing the synaptic connections and subsynaptic proteins in nanoscale.[141] In addition, a strategy was established to image another important biological model, Caenorhabditis elegans (C. elegans), which was challenged to image due to the impermeability to antibodies.[142] Expansion of C. elegans (ExCel) enabled to visualize fluorescent proteins after 20× linear expansion, which offered a new toolbox for mapping synaptic proteins; thus, identifying previously unreported proteins at cell junctions. Recently, genes were also expanded using the untargeted expansion sequencing (ExSeq), mapping the RNA in the mouse brain with the nanoscale resolution.[143]
3.2.4. 4D printing with expandable materials
Apart from shrinkage applied to size/shape-shifting behaviors, other 4D printing mechanisms, such as stretching or folding, were using the expansion strategy.[63, 144] Tibbits and co-workers fabricated ring and disk structures with the hydrophilic UV curable polymer to form the linear-stretching, ring-stretching, or folding structures. The UV-curable polymer was deposited along with the rigid plastic part using an inkjet printer to build the 3D construct with low crosslinking density, followed by exposing to water and creating the hydrogel with the expansion up to 200% of the original volume.[63] The controllable shape-shifting behavior was achieved by adjusting the ratio of expandable hydrogel and rigid plastic.
Based on the property of hydrogel swelling and inspired by the botanical system, Lewis and colleagues explored the ability to print the hydrogel-based inks containing cellulose fibrils that were employed to mimic the cell wall of plants.[145] They used viscoelastic inks consisted of N, N-dimethylacrylamide, nanofibrillated cellulose, nanoclay, glucose oxidase, glucose, and photoinitiator. The swelling of the fabricated constructs could be controlled by the orientation of the cellulose fibrils, which was increased when embedding aligned fibrils (Figure 10A). Using a bilayer-structured system, curvature was achieved by depositing inks with different swelling properties on the top and bottom layers (Figure 10B). The elastic modulus, structure thickness, and the ratio of two layers were all of importance to impact shape-shifting during this process. They successfully created the Gaussian curvature and functional folding flower shapes with the combination of circular and orthogonal bilayer lattices (Figure 10C). This shape-shifting behavior occurred in minutes cause the water could be uptaken rapidly into filament structures. However, this shape-expansion was not reversible, which could be resolved by replacing poly(N,N-dimethylacrylamide) matrix with PNIPAAm, exhibiting reversible changing under the different temperatures of the water.
Figure 10. Applications of expandable materials in 4D printing.
A-C) Plant-inspired expandable materials. A) Images of cellulose fibrils aligned in isotopic. Scale bar represents 200 μm. B) Schematic showing the direct ink writing, cellulose fibril alignment induced by shear stress, and swelling of the printed structure. C) Photographs showing the 4D-printed flower structure with time-lapse sequences during the swelling process. Scale bar represents 5 mm, insert = 2.5 mm. Reproduced with permission.[145] D-F) Self-folding droplets driven by the osmotic pressure. D) Schematic of two droplets joining and two strips bending induced by the osmotic pressure. E) Photographs showing the folding of two droplet-filled strips into a circle through outer-layer expansion and inner-layer shrinkage. F) A flower-like shape folded into a hollow sphere presented in (i) photographs and (ii) simulation. Reproduced with permission.[148]
The other type of 4D printing based on expansion was investigated by Hayley and co-workers. They programmed the osmolarity gradient of lipid droplets leading to the self-folding of structures. Results indicated that droplets with lower or higher osmolarity were joint with the interface bilayer and shrank or swelled, respectively, until the osmolarities came to the same (Figure 10D). Two strips of droplets with different osmolarities were printed as a connecting pattern, and it spontaneously folded over 3 h (Figure 10E). A hollow sphere that is hard to obtain with the common printing method was generated by folding the flower-like shape, where the outside layer with the highest osmolarity resulted in expansion (Figure 10F).
4. Future perspectives
This Progress Report summarizes the recent progress of shrinkable and expandable (bio)materials for biomedical applications, such as tissue regeneration, in vitro tissue or disease modeling, drug delivery, (bio)imaging, and 4D printing, among others. Different factors that determine the process of shrinking or expansion were discussed, which include temperature, pH, gas-foaming, crosslinking ratio, and nanotechnology. Despite the current advances in the research and applications of shrinkable and expandable (bio)materials, more possibilities in this field remain to be explored. For example, more detailed understandings and investigations are required to apply these (bio)materials and processing methods towards clinical translations. Therefore, we aim to provide some future perspectives regarding the development of shrinkable and expandable (bio)materials below, which may be prioritized as a basis for an improvement plan (Figure 11).
Figure 11. The future perspective of (bio)fabrication of shrinkable and expandable (bio)materials, for which the convergence of implants, diagnostic devices, as well as other possible applications is inevitable.
First of all, the true potential of the size-/shape-shifting approaches lie in their abilities to achieve sophisticated structures or motions with programmed steps, which are concocted from relatively simple (bio)fabrication procedures. However, the size-/shape-shifting behavior of most shrinkable and expandable (bio)materials are induced by external stimuli, such as pH, temperature, and swelling. Nevertheless, a number of high-profile cases which implemented these strategies have failed to meet the required biocompatibility for the construct due to detrimental strains caused on the cells. Another adverse result from the aforementioned cases is the alternative of the mechanical properties of size-/shape-shifting (bio)materials after it has gone through shrinkage or expansion. This change could have an unpredictable influence on cell behaviors and their functions within the hydrogels. To counteract this, numerous efforts have focused upon the improvement of (bio)materials that prove to have low cytotoxicity and increased mechanical stability. Even so, processing a (bio)material that fits the previously mentioned needs: retention of shifting capability, reduced cytotoxicity, increased mechanical stability, all while maintaining its original properties, is still challenging. Therefore, more research must be to continue to find new (bio)materials that meet all requirements in order to exploit novel strategies for gentle size-/shape-shifting processes.
In addition, the strategies of shrinkable and expandable (bio)materials used for tissue repair have been intensively investigated in vitro and in vivo. Surprisingly, most of them are implanted after deformations have been formed and little attention has been devoted to using them in situ for various diseases, to a good extent wasting the advantages of controllable size-/shape-shifting. For instance, patients with severe spinal compression fractures, who suffer a significant loss in the vertebral height of the body, need vertebroplasty surgery. The current method for patient recovery is utilizing bone cement in aid of a balloon to increase and rebuild the height of the collapsed bone. However, due to the probability of unprecise control of the balloon size, this technique may lead to insufficient control of volume accuracy. Therefore, programmable expandable materials with desired mechanical properties would have more hopeful prospects in precisely controlling the replaced volume. This in situ size-/shape-shifting approach has the potential to be applied in designing a new class of clinical tools not only limited to compression fractures of the spine. With additional efforts, we believe that the future of shrinkable and expandable (bio)materials with more precise control will look bright for clinical applications.
Finally, an improvement in diagnostic technologies with shrinkable and expandable (bio)materials is needed. As we mentioned in Section 3.2.3, we demonstrated a series of interesting applications of expansion strategies to microscopy. Key to the development of diagnostic technology is the ability to accurately visualize objectives in multiple dimensions. As such, ExM offers significant advantages of amplifying detecting signals and generates more precise diagnostic results. For the healthcare area, with the rapid rise of point-of-care diagnostics, the concept of expansion mini-microscope (ExMM) was also proposed by us, by combining the ExM technique with portable and low-cost image devices.[146] Therefore, further efforts are needed to explore the combined technologies for diagnostic applications.
5. Conclusion
Overall, shrinkable and expandable (bio)materials have been applied as a (bio)fabrication platform to integrate the advantages of (bio)fabrication technology and the intrinsic properties of the (bio)materials. As such, different size-/shape-shifting strategies have been developed with more controllable and precise deformation methods. With such desirable properties, numerous biomedical applications have been carried out in the field of tissue regeneration, disease modeling, diagnostic, 4D (bio)printing, and drug delivery. This Progress Report summarizes representative size-/shape-shifting mechanisms, such as those based on pH, temperature, swelling, electrostatic interactions, nanotechnology, and crosslinking ratio, as well as related biomedical applications in different areas. Towards the future, shrinkable and expandable (bio)materials processed with different (bio)fabrication technologies will likely open the door to new possibilities in developing in situ tissue implants for advanced cell or drug delivery, and portable ultra-sensitive diagnostic device. More inspiring strategies will be further driven to offer new avenues in these exciting areas for enhancing this unique type of (bio)fabrication scenario.
Acknowledgments
The authors gratefully acknowledge funding from the National Institutes of Health (R00CA201603, R21EB025270, R21EB026175, R01EB028143, R03EB027984, R01HL153857, UG3TR00327, R21EB030257), the National Science Foundation (NSF-CBET-1936105), and the Brigham Research Institute.
Biography
Mian Wang received his B.S. degree in Biotechnology from Shandong University, and his Ph.D. degree in Chemical Engineering from Northeastern University. His previous work focused on nanotechnology-based drug delivery for immunotherapy and anti-inflammation. Currently, he is a postdoctoral research fellow in Prof. Y. Shrike Zhang’s group at Brigham and Women’s Hospital, Harvard Medical School. His research centers on novel bioink development and digital light processing-based 3D bioprinting.

Wanlu Li obtained her B.S. degree from Jilin University in 2015. She is currently a Ph.D. candidate in the School of Biomedical Engineering at Shanghai Jiao Tong University. Since 2018, she worked with Prof. Y. Shrike Zhang at Brigham and Women’s Hospital, Harvard Medical School as a joint Ph.D. candidate. Her research focuses on the digital light processing-based 3D bioprinting, the exploitation of the 3D brain model for the neuronal diseases, and the development of printable bioinks.

Dr. Zhang received a B.Eng. in Biomedical Engineering from Southeast University (2008), a M.S. in Biomedical Engineering from Washington University in St. Louis (2011), and a Ph.D. in Biomedical Engineering at Georgia Institute of Technology/Emory University (2013). Dr. Zhang is currently an Assistant Professor at Harvard Medical School and Associate Bioengineer at Brigham and Women’s Hospital. Dr. Zhang’s research is focused on innovating medical engineering technologies, including bioprinting, organs-on-chips, microfluidics, and bioanalysis, to recreate functional tissues and their biomimetic models towards applications in precision medicine.

Footnotes
Conflict of Interest
Y. S. Zhang is on the Scientific Advisory Board of Allevi, Inc., which however, did not support or participate in this work.
References
- [1].Chen FM, Liu X, Prog Polym Sci 2016. 53 86–168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Abdulghani S, Mitchell GR, Biomolecules 2019. 9(11). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Shaaban M, Chowdhury S, Nolen BJ, Nat Struct Mol Biol 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Yin X, Mead BE, Safaee H, Langer R, Karp JM, Levy O, Cell Stem Cell 2016. 18(1) 25–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Eltom A, Zhong G, Muhammad A, Adv Mater Sci Eng 2019. 2019. [Google Scholar]
- [6].Charbe NB, Zacconi FC, Amnerkar N, Pardhi D, Shukla P, Mukattash TL, McCarron PA, Tambuwala MM, Curr Cardiol Rev 2019. 15(3) 188–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Mandrycky C, Phong K, Zheng Y, MRS Commun 2017. 7(3) 332–347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Tetsuka H, Shin SR, J Mater Chem B 2020. 8(15) 2930–2950. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Nikolova MP, Chavali MS, Bioact Mater 2019. 4 271–292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Ong CS, Yesantharao P, Huang CY, Mattson G, Boktor J, Fukunishi T, Zhang H, Hibino N, Pediatr Res 2018. 83(1) 223–231. [DOI] [PubMed] [Google Scholar]
- [11].Zhu L, Luo D, Liu Y, Int J Oral Sci 2020. 12(1) 6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Qasim M, Chae DS, Lee NY, Int J Nanomedicine 2019. 14 4333–4351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Alarçin E, Guan X, Kashaf SS, Elbaradie K, Yang H, Jang HL, Khademhosseini A, Reg Med 2016. 11(8) 849–858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Tan J, Saltzman WM, Biomaterials 2004. 25(17) 3593–601. [DOI] [PubMed] [Google Scholar]
- [15].Geckil H, Xu F, Zhang X, Moon S, Demirci U, Nanomedicine 2010. 5(3) 469–484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Prince E, Kumacheva E, Nat Rev Mater 2019. 4(2) 99–115. [Google Scholar]
- [17].Gaharwar AK, Singh I, Khademhosseini A, Nat Rev Mater 2020. 5(9) 686–705. [Google Scholar]
- [18].O’brien FJ, Mater Today 2011. 14(3) 88–95. [Google Scholar]
- [19].Stratton S, Shelke NB, Hoshino K, Rudraiah S, Kumbar SG, Bioact Mater 2016. 1(2) 93–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Li J, Chen M, Fan X, Zhou H, J Transl Med 2016. 14 271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Datta P, Dey M, Ataie Z, Unutmaz D, Ozbolat IT, NPJ Precis Oncol 2020. 4 18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Kim SH, Yeon YK, Lee JM, Chao JR, Lee YJ, Seo YB, Sultan MT, Lee OJ, Lee JS, Yoon SI, Hong IS, Khang G, Lee SJ, Yoo JJ, Park CH, Nat Commun 2018. 9(1) 1620. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Yan Q, Dong H, Su J, Han J, Song B, Wei Q, Shi Y, Engineering 2018. 4(5) 729–742. [Google Scholar]
- [24].Ligon SC, Liska R, Stampfl J, Gurr M, Mulhaupt R, Chem Rev 2017. 117(15) 10212–10290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Apsite I, Biswas A, Li Y, Ionov L, Adv Funct Mater 2020. 1908028. [Google Scholar]
- [26].van Manen T, Janbaz S, Zadpoor AA, Mater Today 2018. 21(2) 144–163. [Google Scholar]
- [27].Jiang J, Carlson MA, Teusink MJ, Wang H, MacEwan MR, Xie J, ACS Biomater Sci Eng 2015. 1(10) 991–1001. [DOI] [PubMed] [Google Scholar]
- [28].Gong J, Schuurmans CCL, Genderen AMV, Cao X, Li W, Cheng F, He JJ, Lopez A, Huerta V, Manriquez J, Li R, Li H, Delavaux C, Sebastian S, Capendale PE, Wang H, Xie J, Yu M, Masereeuw R, Vermonden T, Zhang YS, Nat Commun 2020. 11(1) 1267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Momeni F, Liu X, Ni J, Mater Des 2017. 122 42–79. [Google Scholar]
- [30].Peralta Ramos ML, Gonzalez JA, Fabian L, Perez CJ, Villanueva ME, Copello GJ, Mater Sci Eng C Mater Biol Appl 2017. 78 619–626. [DOI] [PubMed] [Google Scholar]
- [31].Xiong J, Chen J, Lee PS, Adv Mater 2020. e2002640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Zhang Q, Weber C, Schubert US, Hoogenboom R, Mater Horiz 2017. 4(2) 109–116. [Google Scholar]
- [33].Nagy PI, Int J Mol Sci 2014. 15(11) 19562–633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Shi Q, Liu H, Tang D, Li Y, Li X, Xu F, NPG Asia Mater 2019. 11(1) 1–21. [Google Scholar]
- [35].Liu L, Jiang S, Sun Y, Agarwal S, Adv Funct Mater 2016. 26(7) 1021–1027. [Google Scholar]
- [36].Gupta P, Vermani K, Garg S, Drug Discov Today 2002. 7(10) 569–79. [DOI] [PubMed] [Google Scholar]
- [37].Vargason AM, Anselmo AC, Mitragotri S, Nat Biomed Eng 2021. [DOI] [PubMed] [Google Scholar]
- [38].Koetting MC, Peters JT, Steichen SD, Peppas NA, Mater Sci Eng R Rep 2015. 93 1–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Deen GR, Loh XJ, Gels 2018. 4(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [40].Gupta NV, Shivakumar HG, Iran J Pharm Res 2012. 11(2) 481–93. [PMC free article] [PubMed] [Google Scholar]
- [41].Zhang J, Peppas NA, Macromolecules 2000. 33(1) 102–107. [Google Scholar]
- [42].Kim B, La Flamme K, Peppas NA, J Appl Polym Sci 2003. 89(6) 1606–1613. [Google Scholar]
- [43].Albright V, Zhuk I, Wang Y, Selin V, van de Belt-Gritter B, Busscher HJ, van der Mei HC, Sukhishvili SA, Acta Biomater 2017. 61 66–74. [DOI] [PubMed] [Google Scholar]
- [44].Lavrador P, Esteves MR, Gaspar VM, Mano JF, Adv Funct Mater 2020. 2005941. [Google Scholar]
- [45].Rizzo F, Kehr NS, Adv Healthc Mater 2021. 10(1) e2001341. [DOI] [PubMed] [Google Scholar]
- [46].Shin MK, Spinks GM, Shin SR, Kim SI, Kim SJ, Adv Mater 2009. 21(17) 1712–1715. [Google Scholar]
- [47].Zhao L, Niu L, Liang H, Tan H, Liu C, Zhu F, ACS Appl Mater Interfaces 2017. 9(43) 37563–37574. [DOI] [PubMed] [Google Scholar]
- [48].Jang J-H, Park S-Y, Sensors and Actuators B: Chemical 2017. 241 636–643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [49].Bassil M, Davenas J, Tahchi ME, Sens Actuators B: Chem 2008. 134(2) 496–501. [Google Scholar]
- [50].Techawanitchai P, Ebara M, Idota N, Asoh T-A, Kikuchi A, Aoyagi T, Soft Matter 2012. 8(10) 2844–2851. [Google Scholar]
- [51].Guo W, Lu CH, Orbach R, Wang F, Qi XJ, Cecconello A, Seliktar D, Willner I, Adv Mater 2015. 27(1) 73–8. [DOI] [PubMed] [Google Scholar]
- [52].Lee BP, Konst S, Adv Mater 2014. 26(21) 3415–9. [DOI] [PubMed] [Google Scholar]
- [53].Beebe DJ, Moore JS, Bauer JM, Yu Q, Liu RH, Devadoss C, Jo BH, Nature 2000. 404(6778) 588–90. [DOI] [PubMed] [Google Scholar]
- [54].Wells CM, Harris M, Choi L, Murali VP, Guerra FD, Jennings JA, J Funct Biomater 2019. 10(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [55].Kwon SS, Kong BJ, Park SN, Eur J Pharm Biopharm 2015. 92 146–154. [DOI] [PubMed] [Google Scholar]
- [56].Li J, Mooney DJ, Nat Rev Mater 2016. 1(12). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [57].Wang Z, Wang Z, Lu WW, Zhen W, Yang D, Peng S, NPG Asia Mater 2017. 9(10) e435–e435. [Google Scholar]
- [58].Zhang YS, Khademhosseini A, Science 2017. 356(6337). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [59].Ahmed EM, J Adv Res 2015. 6(2) 105–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [60].Achilleos EC, Prud’Homme RK, Christodoulou KN, Gee KR, Kevrekidis IG, Chem Eng Sci 2000. 55(17) 3335–3340. [Google Scholar]
- [61].Lui YS, Sow WT, Tan LP, Wu Y, Lai Y, Li H, Acta Biomater 2019. 92 19–36. [DOI] [PubMed] [Google Scholar]
- [62].Mitchell A, Lafont U, Hołyńska M, Semprimoschnig C, Addit Manuf 2018. 24 606–626. [Google Scholar]
- [63].Raviv D, Zhao W, McKnelly C, Papadopoulou A, Kadambi A, Shi B, Hirsch S, Dikovsky D, Zyracki M, Olguin C, Raskar R, Tibbits S, Sci Rep 2014. 4 7422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [64].Jin FL, Zhao M, Park M, Park SJ, Polymers (Basel) 2019. 11(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [65].Sonnenburg J, Gao J, Weiner J, Macromolecules 1990. 23(21) 4653–4657. [Google Scholar]
- [66].Jiang J, Chen S, Wang H, Carlson MA, Gombart AF, Xie J, Acta biomater 2018. 68 237–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [67].Jiang J, Li Z, Wang H, Wang Y, Carlson MA, Teusink MJ, MacEwan MR, Gu L, Xie J, Adv Healthc Mater 2016. 5(23) 2993–3003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [68].Jeon J, Park JH, Wi S, Kim K-H, Kim S, Ind Eng Chem 2019. 79 437–442. [Google Scholar]
- [69].Lee JB, Jeong SI, Bae MS, Yang DH, Heo DN, Kim CH, Alsberg E, Kwon IK, Tissue Eng Part A 2011. 17(21–22) 2695–2702. [DOI] [PubMed] [Google Scholar]
- [70].Jiang J, Chen S, Wang H, Carlson MA, Gombart AF, Xie J, Acta Biomater 2018. 68 237–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [71].Li C, Wang J, Wang Y, Gao H, Wei G, Huang Y, Yu H, Gan Y, Wang Y, Mei L, Acta Pharm Sin B 2019. 9(6) 1145–1162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [72].Maji TK, Kitagawa S, Pure Appl Chem 2007. 79(12) 2155–2177. [Google Scholar]
- [73].Bureekaew S, Shimomura S, Kitagawa S, Sci Technol Adv Mater 2008. 9(1) 014108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [74].Wu F, Pang Y, Liu J, Nat Commun 2020. 11(1) 4502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [75].Zhao Z, Qi HJ, Fang D, Soft Matter 2019. 15(5) 1005–1016. [DOI] [PubMed] [Google Scholar]
- [76].Zhao Z, Wu J, Mu X, Chen H, Qi HJ, Fang D, Macromol Rapid Commun 2017. 38(13). [DOI] [PubMed] [Google Scholar]
- [77].Wan X, Luo L, Liu Y, Leng J, Adv Sci 2020. 7(16) 2001000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [78].Zhang B, Gao L, Ma L, Luo Y, Yang H, Cui Z, Engineering 2019. 5(4) 777–794. [Google Scholar]
- [79].Zhang YS, Khademhosseini A, Engineering in vitro human tissue models through bio-design and manufacturing, Springer, 2020. [Google Scholar]
- [80].Miri AK, Mirzaee I, Hassan S, Mesbah Oskui S, Nieto D, Khademhosseini A, Zhang YS, Lab Chip 2019. 19(11) 2019–2037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [81].Pi Q, Maharjan S, Yan X, Liu X, Singh B, van Genderen AM, Robledo-Padilla F, Parra-Saldivar R, Hu N, Jia W, Xu C, Kang J, Hassan S, Cheng H, Hou X, Khademhosseini A, Zhang YS, Adv Mater 2018. 30(43) e1706913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [82].Cao X, Ashfaq R, Cheng F, Maharjan S, Li J, Ying G, Hassan S, Xiao H, Yue K, Zhang YS, Adv Funct Mater 2019. 29(31). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [83].Liu W, Zhong Z, Hu N, Zhou Y, Maggio L, Miri AK, Fragasso A, Jin X, Khademhosseini A, Zhang YS, Biofabrication 2018. 10(2) 024102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [84].Zhou X, Nowicki M, Sun H, Hann SY, Cui H, Esworthy T, Lee JD, Plesniak M, Zhang LG, ACS Appl Mater Interfaces 2020. 12(41) 45904–45915. [DOI] [PubMed] [Google Scholar]
- [85].Rogers JA, Someya T, Huang Y, Science 2010. 327(5973) 1603–7. [DOI] [PubMed] [Google Scholar]
- [86].Kim D-H, Lu N, Ma R, Kim Y-S, Kim R-H, Wang S, Wu J, Won SM, Tao H, Islam A, Science 2011. 333(6044) 838–843. [DOI] [PubMed] [Google Scholar]
- [87].Cao Q, Kim HS, Pimparkar N, Kulkarni JP, Wang C, Shim M, Roy K, Alam MA, Rogers JA, Nature 2008. 454(7203) 495–500. [DOI] [PubMed] [Google Scholar]
- [88].Carrell C, Kava A, Nguyen M, Menger R, Munshi Z, Call Z, Nussbaum M, Henry C, Microelectron Eng 2019. 206 45–54. [Google Scholar]
- [89].Zhang W, Guo S, Carvalho WSP, Jiang Y, Serpe MJ, Anal Methods 2016. 8(44) 7847–7867. [Google Scholar]
- [90].Copaci D-S, Blanco D, Martin-Clemente A, Moreno L, Int J Adv Robot Syst 2020. 17(1) 1729881419886747. [Google Scholar]
- [91].Villoslada A, Flores A, Copaci D, Blanco D, Moreno L, Rob Auton Syst 2015. 73 91–101. [Google Scholar]
- [92].Pegan JD, Ho AY, Bachman M, Khine M, Lab Chip 2013. 13(21) 4205–9. [DOI] [PubMed] [Google Scholar]
- [93].Chen CS, Breslauer DN, Luna JI, Grimes A, Chin WC, Lee LP, Khine M, Lab Chip 2008. 8(4) 622–4. [DOI] [PubMed] [Google Scholar]
- [94].Nguyen D, McLane J, Lew V, Pegan J, Khine M, Biomicrofluidics 2011. 5(2) 22209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [95].Pegan JD, Zhang J, Chu M, Nguyen T, Park SJ, Paul A, Kim J, Bachman M, Khine M, Nanoscale 2016. 8(39) 17295–17303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [96].Chu M, Nguyen T, Pandey V, Zhou Y, Pham HN, Bar-Yoseph R, Radom-Aizik S, Jain R, Cooper DM, Khine M, NPJ Digit Med 2019. 2 8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [97].Kim J, Chou EF, Le J, Wong S, Chu M, Khine M, Adv Healthc Mater 2019. 8(13) e1900109. [DOI] [PubMed] [Google Scholar]
- [98].Piltti KM, Cummings BJ, Carta K, Manughian-Peter A, Worne CL, Singh K, Ong D, Maksymyuk Y, Khine M, Anderson AJ, Methods 2018. 133 81–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [99].Taylor D, Dyer D, Lew V, Khine M, Lab Chip 2010. 10(18) 2472–5. [DOI] [PubMed] [Google Scholar]
- [100].Nguyen D, Sa S, Pegan JD, Rich B, Xiang G, McCloskey KE, Manilay JO, Khine M, Lab Chip 2009. 9(23) 3338–44. [DOI] [PubMed] [Google Scholar]
- [101].Park SJ, Kim J, Chu M, Khine M, Adv Mater Technol 2016. 1(5) 1600053. [Google Scholar]
- [102].Liu Y, Wang H, Ho J, Ng RC, Ng RJH, Hall-Chen VH, Koay EHH, Dong Z, Liu H, Qiu CW, Greer JR, Yang JKW, Nat Commun 2019. 10(1) 4340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [103].Lin S.-y., Fleming J, Hetherington D, Smith B, Biswas R, Ho K, Sigalas M, Zubrzycki W, Kurtz S, Bur J, Nature 1998. 394(6690) 251–253. [Google Scholar]
- [104].Bobo D, Robinson KJ, Islam J, Thurecht KJ, Corrie SR, Pharm Res 2016. 33(10) 2373–87. [DOI] [PubMed] [Google Scholar]
- [105].De M, Ghosh PS, Rotello VM, Adv Mater 2008. 20(22) 4225–4241. [Google Scholar]
- [106].Oran D, Rodriques SG, Gao R, Asano S, Skylar-Scott MA, Chen F, Tillberg PW, Marblestone AH, Boyden ES, Science 2018. 362(6420) 1281–1285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [107].Scott MA, Wissner-Gross ZD, Yanik MF, Lab Chip 2012. 12(12) 2265–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [108].Geng Q, Wang D, Chen P, Chen SC, Nat Commun 2019. 10(1) 2179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [109].Obata K, El-Tamer A, Koch L, Hinze U, Chichkov BN, Light: Sci Appl 2013. 2(12) e116–e116. [Google Scholar]
- [110].Tibbits S, Archit Des 2014. 84(1) 116–121. [Google Scholar]
- [111].Tibbits S, The emergence of “4D printing”, TED conference, 2013. [Google Scholar]
- [112].Pei E, Assem Autom 2014. [Google Scholar]
- [113].Layani M, Wang X, Magdassi S, Adv Mater 2018. 30(41) e1706344. [DOI] [PubMed] [Google Scholar]
- [114].Ding Z, Yuan C, Peng X, Wang T, Qi HJ, Dunn ML, Sci Adv 2017. 3(4) e1602890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [115].Ashammakhi N, Ahadian S, Zengjie F, Suthiwanich K, Lorestani F, Orive G, Ostrovidov S, Khademhosseini A, Biotechnol J 2018. 13(12) e1800148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [116].Kuang X, Chen K, Dunn CK, Wu J, Li VCF, Qi HJ, ACS Appl Mater Interfaces 2018. 10(8) 7381–7388. [DOI] [PubMed] [Google Scholar]
- [117].Khoo ZX, Teoh JEM, Liu Y, Chua CK, Yang S, An J, Leong KF, Yeong WY, Virtual Phys Prototy 2015. 10(3) 103–122. [Google Scholar]
- [118].Wu J, Yuan C, Ding Z, Isakov M, Mao Y, Wang T, Dunn ML, Qi HJ, Sci Rep 2016. 6 24224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [119].Zhang Q, Zhang K, Hu G, Sci Rep 2016. 6 22431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [120].Boley JW, van Rees WM, Lissandrello C, Horenstein MN, Truby RL, Kotikian A, Lewis JA, Mahadevan L, Proc Natl Acad Sci U S A 2019. 116(42) 20856–20862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [121].Murphy WL, Dennis RG, Kileny JL, Mooney DJ, Tissue Eng 2002. 8(1) 43–52. [DOI] [PubMed] [Google Scholar]
- [122].Udomluck N, Koh WG, Lim DJ, Park H, Int J Mol Sci 2019. 21(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [123].Li H, Cheng F, Li W, Cao X, Wang Z, Wang M, Robledo-Lara JA, Liao J, Chavez-Madero C, Hassan S, Xie J, Trujillo-de Santiago G, Alvarez MM, He J, Zhang YS, Biofabrication 2020. 12(4) 045027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [124].Cheng F, Cao X, Li H, Liu T, Xie X, Huang D, Maharjan S, Bei HP, Gómez A, Li J, Nano lett 2019. 19(6) 3603–3611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [125].Kamaly N, Yameen B, Wu J, Farokhzad OC, Chem Rev 2016. 116(4) 2602–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [126].Larush L, Kaner I, Fluksman A, Tamsut A, Pawar AA, Lesnovski P, Benny O, Magdassi S, Journal of 3D printing in medicine 2017. 1(4) 219–229. [Google Scholar]
- [127].Wu W, Shen J, Banerjee P, Zhou S, Biomaterials 2010. 31(32) 8371–81. [DOI] [PubMed] [Google Scholar]
- [128].Heskins M, Guillet JE, J Macromol Sci-Chem 1968. 2(8) 1441–1455. [Google Scholar]
- [129].Dai Y, Ma P, Cheng Z, Kang X, Zhang X, Hou Z, Li C, Yang D, Zhai X, Lin J, ACS Nano 2012. 6(4) 3327–38. [DOI] [PubMed] [Google Scholar]
- [130].Zhang XZ, Jo Lewis P, Chu CC, Biomaterials 2005. 26(16) 3299–309. [DOI] [PubMed] [Google Scholar]
- [131].Yavuz MS, Cheng Y, Chen J, Cobley CM, Zhang Q, Rycenga M, Xie J, Kim C, Song KH, Schwartz AG, Wang LV, Xia Y, Nat Mater 2009. 8(12) 935–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [132].Huang S, Duan S, Wang J, Bao S, Qiu X, Li C, Liu Y, Yan L, Zhang Z, Hu Y, Adv Funct Mater 2016. 26(15) 2532–2544. [Google Scholar]
- [133].Luckanagul JA, Pitakchatwong C, Ratnatilaka Na Bhuket P, Muangnoi C, Rojsitthisak P, Chirachanchai S, Wang Q, Rojsitthisak P, Carbohydr Polym 2018. 181 1119–1127. [DOI] [PubMed] [Google Scholar]
- [134].Zagorska-Dziok M, Sobczak M, Pharmaceutics 2020. 12(5). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [135].Chen F, Tillberg PW, Boyden ES, Science 2015. 347(6221) 543–548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [136].Chang JB, Chen F, Yoon YG, Jung EE, Babcock H, Kang JS, Asano S, Suk HJ, Pak N, Tillberg PW, Wassie AT, Cai D, Boyden ES, Nat Methods 2017. 14(6) 593–599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [137].Tillberg PW, Chen F, Piatkevich KD, Zhao Y, Yu CC, English BP, Gao L, Martorell A, Suk HJ, Yoshida F, DeGennaro EM, Roossien DH, Gong G, Seneviratne U, Tannenbaum SR, Desimone R, Cai D, Boyden ES, Nat Biotechnol 2016. 34(9) 987–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [138].Chen F, Wassie AT, Cote AJ, Sinha A, Alon S, Asano S, Daugharthy ER, Chang JB, Marblestone A, Church GM, Raj A, Boyden ES, Nat Methods 2016. 13(8) 679–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [139].Sun D.-e., Fan X, Shi Y, Zhang H, Huang Z, Cheng B, Tang Q, Li W, Zhu Y, Bai J, Nat Methods 2020. 1–7. [DOI] [PubMed] [Google Scholar]
- [140].Zhao Y, Bucur O, Irshad H, Chen F, Weins A, Stancu AL, Oh EY, DiStasio M, Torous V, Glass B, Stillman IE, Schnitt SJ, Beck AH, Boyden ES, Nat Biotechnol 2017. 35(8) 757–764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [141].Freifeld L, Odstrcil I, Forster D, Ramirez A, Gagnon JA, Randlett O, Costa EK, Asano S, Celiker OT, Gao R, Martin-Alarcon DA, Reginato P, Dick C, Chen L, Schoppik D, Engert F, Baier H, Boyden ES, Proc Natl Acad Sci U S A 2017. 114(50) E10799–E10808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [142].Yu CJ, Barry NC, Wassie AT, Sinha A, Bhattacharya A, Asano S, Zhang C, Chen F, Hobert O, Goodman MB, Haspel G, Boyden ES, Elife 2020. 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [143].Alon S, Goodwin DR, Sinha A, Wassie AT, Chen F, Daugharthy ER, Bando Y, Kajita A, Xue AG, Marrett K, bioRxiv 2020. [Google Scholar]
- [144].Ding H, Zhang X, Liu Y, Ramakrishna S, Int J Adv Manuf Tech 2019. 105(11) 4633–4649. [Google Scholar]
- [145].Gladman AS, Matsumoto EA, Nuzzo RG, Mahadevan L, Lewis JA, Nat Mater 2016. 15(4) 413–8. [DOI] [PubMed] [Google Scholar]
- [146].Zhang YS, Santiago GT, Alvarez MM, Schiff SJ, Boyden ES, Khademhosseini A, Curr Opin Biomed Eng 2017. 1 45–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [147].Liu Y, Wang H, Ho J, Ng RC, Ng RJ, Hall-Chen VH, Koay EH, Dong Z, Liu H, Qiu C-W, Nat commun 2019. 10(1) 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [148].Villar G, Graham AD, Bayley H, Science 2013. 340(6128) 48–52. [DOI] [PMC free article] [PubMed] [Google Scholar]











