A Multifunctional Acoustic Tweezer for Heterogenous Assembloids Patterning

Acoustic tweezers, capable of complicated manipulation of bioparticles by acoustic radiation forces using a noninvasive and noncontact approach, are an innovative technology for patterning assemble organoids. Hence, acoustic tweezers exhibit considerable potential for forming programmable patterning of organoids with specific spatial structures. Furthermore, heterogeneous assembloids with complex arrangement patterns can be built through sequential assembling and culturing to explore polarized tissue development or disease metastasis in multiple organs. This study focuses on the structural assembly of organoids using an ultrasonic 2D matrix array to generate real‐time switching of different acoustic fields. In addition, a local renal injured assembloid is fabricated to study and verify its application in tissue engineering and disease modeling.


Introduction
Organoids are minitissues developed from primary or pluripotent cells using the self-assembly design to form complex 3D cellular structures in vitro, which facilitate the recapitulation of the key physiological features and structures of tissues. [1,2] Although organoids provide the advantage of self-assembling fertilized cell types with significantly improved temporal-spatial resolution over cellular or animal models, modeling complex interactions between the cells derived from different lineages is a challenge for organoids. [3] At present, self-assembled organoids are limited in providing a more advanced and accurate representation of human tissues or conducting more evolutionary scientific hypotheses, such as the crosslineage rearrangement of tissues. [4] Some recent studies suggested that integrated vascular, immune, or nerve cells allow for the creation of complex organoid structures via assembly from individual organoids. [5][6][7] However, the cocultured assembloids were aggregated from organoids without organized patterning, which experienced uncontrolled heterogeneity and inadequate repeatability. Especially programmable and defined spatial organizations could not be generated robustly. For instance, as a repeated functional unit, nephron damage might transpire sequentially in renal diseases. [8] The cellular and molecular mechanisms of injury-invoked repair processes that maintain homeostasis to regulate kidney function tightly are challenging to interrogate under any established model system. Programmable patterning assembloids can be leveraged to study the spatial arrangement using a histogenetic approach with a potential understanding of crosslineage cell-cell interactions, crosscellular niche-cell migration, and establishing a signaling center within the organoids under different cellular statuses. [9] Therefore, a reliable device that enables structural assembly of assembloids using a nonintervention and noncontact approach should be developed.
Acoustic tweezers, as an alternative to optical tweezers, [10,11] can generate greater radiation force at the same input energy while avoiding the limitations of the photothermal effects, phototoxicity, and complex optical corrections, which provide a noncontact and labeled-free manipulation. Some studies have reported that utilizing considerable depth and biocompatibility of the acoustic tweezering approach facilitate the application in cell sorting and tissue construction and manipulation in vivo. [12][13][14][15] Modulating the frequency and duration of action of the transducer's emission signal can produce surface acoustic waves (SAWs) with different waveforms, harmonic orders, and symmetry patterns at the gas-liquid interface. Utilization of SAW shall select all items in the acoustic wave without bias, as clusters of cells could be manipulated at the wave nodes to achieve large-scale patterning. [16,17] However, the individual selectivity was limited. In addition, SAW-type acoustic tweezers combined with microfluidic technology enable highthroughput cell sorting in a heterogeneous cell population, such as selecting cancer cells for diagnosis. [18][19][20] However, promising progress in cell lineage regulation has increased the demand for a more precise and programmable individualized manipulation method to construct in vitro models of heterogeneous organoids in regenerative medicine and drug screening. [21][22][23][24] SAW-type acoustic tweezer devices allow changing the direction of SAWs by exciting symmetrically placed transducers in a particular sequence. Similarly, through the signal phase control, the device can realize the translation of the standing wave nodes to achieve the basic assembly function. However, due to complex genetic architecture and synergy mechanisms between organs, there is still a considerable need to flexibly and accurately assemble heterogeneous organoids over programmable spatial distribution. [25][26][27] As the fundamental mechanism, the sound field distribution reaching the action target can be predesigned by modulating the sound pressure distribution of the bulk acoustic wave (BAW) in the target space. [28][29][30][31] Moreover, BAW-type acoustic tweezers are more appropriate for precise control of a single particle and do not require preset multiple pairs of transducers or reflectors. In addition, phase-array acoustic tweezers can dynamically construct holographic acoustic fields in the target area by real-time modulation of the amplitude and phase of each array element. Hence, this programmable BAW-type approach offers the great advantage of generating a single sound field using artificial acoustic structures. [32,33] Furthermore, the advantage of BAW in the precise selection and sequential assembly in real-time live bioparticle manipulation might offer outstanding biocompatibility, which is more appropriate for constructing complex heterogeneous assembloids. To test this hypothesis, acoustic tweezers should be tested in the application of patterning heterogeneous organoids into assembloids through a precisely programmed approach.
To provide a more comprehensive in vitro biological tissue patterning model to study kidney injuries or complicated disease modeling, we propose an online organoid assembly platform (OOAP) based on ultrasonic phase arrays for a programmable structural assembly. In contrast to the conventional vortex trapping methods, the proposed method allows dynamic adjustment and time division multiplexing control of different functional acoustic fields to achieve a precise selection, movement with lateral force justification, and rotation of a single organoid without surface treatment or modification. Therefore, assembloids with specific spatial structures can be constructed using a sequentially operated tweezing process, providing a promising platform to model physiological and pathological processes of organogenesis. In this article, we s uccessfully assembled cisplatin-induced local renal injured assembloids using the in-house developed device. Indeed, the construction of heterogeneous renal injured assembloids that presents corresponding histopathological characterizations via real-time spatial patterning shall facilitate delineating damage repair responses modeling interorgan crosstalks. Furthermore, the repeatability and biocompatibility of this system were verified for programmable organoid patterning in biological applications.

Real-Time Patterning Assembly Using a Live Online Organoid Assembly Platform
A live OOAP mounted on an inverted microscope was applied to achieve real-time assembly (Figure 1a), which also incorporated the cell culture system to maintain live cell condition. The isolation coupling module and ultrasonic acoustic tweezer system were implemented to discover positive coupling between the ultrasonic probe and existing standard cell culture consumables. During the assembly, organoids were maintained in conventional cell culture dishes. They were assembled structurally stepwise by image guidance, wherein the feedback closed loop controlled the temperature and carbon dioxide concentration, as illustrated in Figure S1, Supporting Information. The ultrasonic phase array affects the optical path of the inverted microscope during manipulation; hence, the ring structure diode was implemented as the cold light source for the system. Here, the angle between the light beam and ultrasonic phase array was separated by a horizontal angle of 30°under the optimized lighting and imaging conditions. In addition, the coupling module was implemented to form a trans-well-like structure to confirm the construction status of the acoustic fields and reduce the risk of contamination caused by placing the transducer. The trans-well-like structure was sealed using a Mylar film for acoustic transduction ( Figure S2, Supporting Information). The cultivation and external environments were isolated for sterility in addition to avoiding electrolyte contaminations of the solution. Sterile water was used as the ultrasonic coupling medium in the concave position.
Therefore, we laid a firm foundation for the reproducible and sustainable structural assembly of organoids and subsequent biological investigations using phase-array acoustic tweezers. In addition, the excitation system modulated the emission voltage and time delay of the 64 channel transducer, and holographic acoustic fields were constructed in the target space ( Figure 1c). Subsequently, we enabled a precise selection and rotation of individual organoids and their structural assembly using the acoustic radiation force and various functional fields. Finally, we employed OOAP to construct local renal injured assembloids ( Figure 1d) by real-time arrangement of live organoids in a specific pattern.

Generation and Application of Holographic Functional Acoustic Fields
In this study, we designed and constructed a series of holographic adjustable fields to utilize acoustic fields' flexible adjustment and construction through a phase-array transducer. As shown in Figure 2a, each transducer array element was connected to the corresponding channel of the multichannel electronic excitation system. Based on the preset focus information, we combined the position information of each array element to establish the propagation coefficient. Subsequently, the propagation coefficient matrix of each control focal combination and transducer array was obtained; in addition, the transmission phase and time delay of each element of the transducer were determined using the backpropagation principle and an iterative algorithm. [30] Thus, we succeeded in forming multiple focal points in the designated space.
The symmetric vortex can be viewed as a circular arrangement of eight focal points, known as control points, around the central location (Figure 2b), where the phase of the eight control points is π/8 along the circumference interval changes. The clockwise and counterclockwise changes in the phase generated positive and negative vortices, respectively. According to Equation (1), the vortex wave fronts are screw phase dislocations. [34] p in ¼ Pðr, θÞe ilφ e Àiωt (1) where in polar coordinates, ω denotes the angular frequency, and Pðr, θÞ is an axisymmetric field, typically a firstor higher-order Bessel function that satisfies the scalar wave equation. The phase factor expðilφÞ describes the direction of propagation of the vortex around the axis, φ refers to the azimuth angle, and l represents the topological charge. Its sign determines the chirality at which the wave front propagates. The phase changes in different handed directions produce tangential forces in the corresponding directions; this feature will be adopted in subsequent studies. When the weights of the eight control points were equal, a symmetric vortex was formed, in which the weak field at the center was surrounded by a uniform strong field. Next, by adjusting the spacing between the control points, the vortex aperture was tuned (Figure 2c and S3c,d,g, Supporting Information). However, the adjustment of the vortex aperture is not infinite, and when the control point spacing is greater than 5.2 λ or less than 0.4 λ, the shape of the vortex fails in the measurement and simulation experiments, as shown in Figure S3a,b, Supporting Information. Additionally, assigning different weights to the eight control points could generate an asymmetric vortex with a weak field surrounded by a strong inhomogeneous field (Figure 2d and S3e,f ). We characterized these acoustic fields using hydrophones (Figure 2e-g), and the pressure distribution was identical to the simulation results.
Because of the dense structure of the organoids, we found that their acoustic impedance was higher than that of the culture solution. We simulated the force of the organoid in these acoustic fields using the COMSOL Multiphysics software. Here, we simplified the organoid into a sphere, and the acoustic radiation force could be calculated using the Gor'kov gradient potential ðUÞ as follows. [35] F ¼ À∇U (2) where U can be expressed by the complex acoustic pressure p and its spatial derivative as follows. www.advancedsciencenews.com www.small-structures.com Figure 2. Control strategies of the single-organoid manipulation and heterogeneous assembloids construction. a) Forming the focused vortex field in the designated space through eight control points. b-d) Symmetrical and asymmetrical vortex fields with different handedness and shapes were formed by modulating the handedness, spacing, and weight of eight control points. e-g) Acoustic field measurements of symmetrical and asymmetrical vortex fields. h-k) Shape of the multifunction acoustic fields and the force in the horizontal direction were analyzed using simulations by the COMSOL Multiphysics software. l) Discrete angle control strategy for the single rod-like organoid. m) Continuous angle control strategy for the single elliptical organoid. n) Compact assembly by changing the acoustic vortex aperture size. o) Sequential assembly using the asymmetric acoustic vortex in different directions. p) Revocation of error assembly object using the multifocal field, scale bar: 500 μm. 2200288 (4 of 12) where V denotes the volume of the particle, ω is the angular frequency of the emission, ρ is the density, and c is the speed of sound; the subscripts 0 and p are the culture solution and particle, respectively. Here, the density of the kidney organoids and speed of sound are 1044.0 kg m À3 and 1564.0 m s À1 , respectively. [36] Figure 2h-k shows the direction of the radiation force on the organoid and 3D shape of the acoustic fields in the numerical simulation. The direction of the acoustic radiation force acting on the organoids is from the region of high acoustic potential energy to the region of low acoustic potential energy. We intend to combine these tunable fields to develop structural manipulation strategies.

Control Strategies of Heterogeneous Assembloid Construction
First, the phase matrix transducer was utilized to flexibly construct various shapes and multifunction holographic fields; subsequently, we attempted to switch and multiplex those fields according to their characteristics rationally. Therefore, a series of construction strategies for heterogeneous assembloids were developed, yielding a precise selection, movement, and rotation of individual organoids and structurally assembling them in different situations. Moreover, we defined the control sequence as comprising a series of launch events, which are excitation signals to form different acoustic fields with a fixed duty cycle. Heterogeneous assembloids were fabricated by repeatedly emitting control sequences with different functions.
The precise selection of individual organoids was achieved by alternating the positive and negative vortex launch events in the control sequence. In contrast to single-directional vortex capturing, the alternate emissions of forward and reverse vortex field events contributed to counteract mutually due to tangential forces on the class, leading to mutually reducing the effects of tangential forces on the organoids rather than rotating organoids in the vortex.
Subsequently, steering manipulation was adopted to adjust fused sites between organoids prior to structural assembly. We proposed two control strategies for organoids with different morphologies. As shown in Figure 2i, the first steering strategy was appropriate for organoids with irregular shapes, such as rodand hetero-shaped organoids. In the control sequence, a twin with a different angle was inserted into the positive and negative vortex fields alternating the launch event. The neutralization of the tangential force between the positive and negative vortices can guarantee the consistency of the capture state and position during the movement, whereas twin fields with different angles were similar to invisible tweezers to clamp the sample and adjust the angle discretely (Movie S1, Supporting Information). In contrast, for spherical-shaped organoids (cell spheroids) at the early stage of structural differentiation, we utilized the tangential force caused by the positive and negative vortices (Movie S2, Supporting Information). As shown in Figure 2m, inspired by the pulse-width modulation control, we adjusted the proportion of the launch events of the positive and negative vortices in the control sequence, which can achieve the angle adjustment of different directions and speeds (Movie S3, Supporting Information).
After adjusting the position and angle of the organoid using previous methods, the assembly operation was realized using two methods. The first method adjusted the apertures of the positive and negative vortices. As shown in Figure 2n, by trapping the assembling targets in the weak field area surrounded by strong fields, we gradually adjusted the aperture and voltage to alternate the range of weak and strong fields to achieve small organoids to make a contact (Movie S4, Supporting Information). The second method adjusted the weights of the eight control points through the asymmetric positive and negative vortex fields (Figure 2o) propelled by the strong field, and the weak field area faced the assembling target, as available in Movie S5, Supporting Information. This method was applied to obtain the structural patterning. Compared with the symmetric vortex field assembly, we designed to modulate the field strength to obtain the vortex distribution nonuniformly. Moreover, the strong field ensured the directivity and reliability of the assembly target; in contrast, the weak field reduced the structural influence on the unfinished assembloid. In addition, the operation can be abolished simultaneously using setting multiple focus points to detach the unwanted assembled objects (Figure 2p). Therefore, this strategy is appropriate for constructing various organoids in size or assembloids, which allowed us to apply acoustic tweezers for organoid manipulation under optimized corresponding acoustic control parameters.

Optimization of Acoustic Tweezer Parameters for Structural Assembling
To demonstrate that the phase acoustic tweezer could be used as a noninductive device for structural assembling, we performed a series of characterizations on the performance of acoustic tweezers. We first refined the velocity of organoids in the symmetric vortex under different acoustic parameters. Normal kidney organoids with a size in the range of 400-500 μm were placed at the weak field edge of the 1.6 λ vortex field, with the measured diameter of the weak field area of 800 μm. Subsequently, we recorded that the average speed of the organoids moved from the edge to center of the vortex under ultrasound (Figure 3a). The field was turned off for 15 min to recover the organoid after ultrasonication for 15 min. After repeating the procedure three times, we collected the organoids from the acoustic fields, and characterized them by immunofluorescence (IF) staining, followed by analyzing the proportion of proliferating and apoptotic cells. We measured the moving speed under different voltages at a transmit duty cycle of 2% (Figure 3b), and at different duty cycles at a voltage of 24 V (Figure 3c). We have theoretically analyzed the physical effects of electrical control parameters (duty cycle and excitation voltage) in acoustic manipulation.
In the process of time multiplexing, the formation of each holographic acoustic field is called one launch event, and the pulse period of the launch event is T where the excitation duty www.advancedsciencenews.com www.small-structures.com Figure 3. Optimization of acoustic tweezer parameters for structural assembling. a) Schematic of evaluating the moving speed by varying voltage and duty cycle, subsequent characterization of cell proliferation, and apoptotic rate after acoustic tweezers manipulated with different peak acoustic pressure values. b) Analysis of the speed of organoids moving from the edge to the center of the vortex acoustic field at different voltages. **P < 0.01, ***P < 0.001, and ****P < 0.0001 compared with 24 V voltage group (n = 10). c) Analysis of the speed of organoids moving from the edge to center of the vortex field at varying duty cycles. ****P < 0.0001 and "ns," no significant differences compared to the duty cycle group of 2% (n = 10). d) Cell apoptotic and proliferation rate of organoids after acoustic tweezers manipulation with different peak acoustic pressure values. ****P < 0.0001 and "ns," no significant differences compared to the apoptotic rate of the control group. ####P < 0.0001, #P < 0.05, and "ns," no significant differences com- www.advancedsciencenews.com www.small-structures.com cycle is D T . The following equation can express the increment of particle velocity Δv in a cycle of emission events.
where m p is the particle mass, r is the particle radius,¯n is the directional vector in the vertical plane, v is the particle velocity, and ∇U is the average potential energy of the acoustic radiation. Through the experimental test data, the particle motion velocity is less than 200 μm s À1 . Therefore, by equation where ρ is the density of the organoid, satisfying the Stokes formula to characterize the drag force on the particle. During the initial acceleration period of the acoustic field effect, the horizontal acoustic radiation force can be expressed as And the hysteresis effect has almost no significant effect on the acceleration. Thus, D T increases the acceleration time of the particle and shortens the deceleration time, and after a period of acceleration, the particle moves at approximately steady speed (Δv % 0) in the later part of a launch event cycle owing to the increasing hysteresis, so the steady speed can be approximated as v ¼ 1 6πηr Thus, for a constant voltage, the increase of D T not only shortens the time of particle acceleration to the maximum velocity, but also increases the final velocity. In addition, as the incident acoustic energy increases, the greater the radiation force applied to the particle, so the acoustic energy density ε of the incident acoustic wave can be described by the following equation.
where p is the sound pressure. To consider the universality of the theory, we assume that the incident wave is a plane wave and the acoustic energy density ε can be rewritten as Consider that the acoustic energy density during a cycle of launch events can be expressed as From this, it can be noted that the energy carried by the acoustic wave is related to the square of the amplitude of the incident acoustic wave. Therefore, enhancing the excitation voltage can improve the energy of the incident acoustic wave. At the same time, this can also be achieved by adding the duty cycle, [14] which is equivalent to increasing the number of cycle repetitions.
Both of the aforementioned approaches allowed significantly modify the speed of manipulation. However, in subsequent experiments, we found that bio samples such as organoids are sensitive to peak acoustic pressure. In order to make the acoustically assembled process more efficient, simply increasing the voltage could lead to excessive peak acoustic pressure and impact the expression of organoid activity and functionality. As experimental results shown in Figure 3b,c, adjusting the duty cycle of emission events at 24 V excitation voltage can also effectively enhance the particle movement speed, and the movement effect is consistent with the regulation of increasing the excitation voltage individually. Additionally, we combined the hydrophone measurement of the peak negative pressure (PNP) at different voltages to evaluate the bioeffect of sound pressure on organoids (Figure 3d,e). According to the IF assay, the proliferation and apoptosis rates of organoids at a PNP of 0.334 MPa (12 V) and 0.667 MPa (24 V) did not significantly differ from those of the control group. However, PNP greater than 0.993 MPa (36 V) initiated impairments on cell proliferation, followed by enhanced apoptosis rate of organoids. In particular, when PNP reached 1.680 MPa (60 V), approximately half of the organoid cells appeared to undergo apoptosis. To sustain the organization structure and functionality of organoids, we strictly controlled PNP of the multifunctional acoustic fields below 0.667 MPa (24 V) in the subsequent experiments and confirmed the structural assembly according to the expected arrangement pattern. Table S1, Supporting Information, shows the characterization of other acoustic fields used for structural construction.
Based on the established optimal acoustic tweezer manipulation parameters, a green fluorescent protein (GFP)-labeled organoid with wild-type organoid was assembled. After 2 h of acoustic tweezers application (15 min of effect and 15 min of mutual recovery circulation) and 6 h of online culturing, we removed the isolation module and then transferred the assembloid to an incubator for another 10 h of cultivation. It can be observed that the two organoids fused with a defined orientation. Moreover, the physiological structure of kidney organoids was maintained, and no apoptotic signals were observed ( Figure S6a-c, Supporting Information). These preliminary data indicated that acoustic tweezers can manipulate and assemble organoids without impairing their cellular state or cell-cell arrangement.
Cell environment and cell-cell modulation are the fundamental mechanisms for organ genesis or disease modeling. Compacted cell-cell interactions occur in drug-resistant cold tumors with high cell density and elasticity [37] and in early-developmental stages. [38] In contrast, in the central nervous system, a part of the bloodbrain barrier, the remainder of the cell-cell modulation is less intense. [39] Hence, it would be interesting to mimic different contact elasticities of organoids during their assembly. Therefore, to demonstrate the ability of the construct compacted with a loose assembly between organoids in vitro programmed in the same pattern, a certain structure was assembled (Figure 3f ). The manipulation procedure employed a 1.6 λ symmetric vortex field. In this sequential assembly, the first three organoids were maneuvered to preset positions around the GFP-labeled organoids. Subsequently, the four organoids were assembled by gradually modulating the aperture of the vortex fields. Next, the organoids were "clapped" www.advancedsciencenews.com www.small-structures.com to develop a strong interaction between organoids using a more potent lateral acoustic radiation force. Finally, the fifth organoid was pushed with a GFP-labeled organoid at the center using a 1.6 λ asymmetric vortex acoustic field. Contact was obtained and maintained to construct weak connections between the organoids (Movie S6, Supporting Information). After 18 h of online assembling and culturing, a polarized pentagonal sphere assembly was formed (Figure 3g-m), without losing the intrinsic nephron structure (Figure 3n-p,r-t). Notably, without causing additional cell apoptotics (Figure 3q,u), it suggested that acoustic tweezers served as a promising tool for the personalized construction of complex assemblies that satisfy compact or loose interactions within the assembly in real time.

Formation and Characterization of Specified Regional Renal Injured Assembloids
The nephron is a critical functional unit of the kidney. In humans, the kidney comprises %750 000-2 000 000 nephrons. [40] Therefore, damage to a certain percentage of nephrons causes loss of kidney functions, such as a reduced glomerular filtration rate below 90%, which can increase urinary albumin excretion. [41] Thus, chronic kidney diseases progress from a filtration rate in the range of 30%-60% to end-stage renal diseases at a filtration rate in the range of 15%-30% (Figure 4a). However, owing to the limitations of the existing technology, organoid or cell models are often globally damaged by the in vitro model, which cannot mimic the progression of kidney failure (Figure 4b).
Hence, to construct more complex heterogeneous assembloids to imitate the physiological and pathophysiological conditions in vivo, we applied acoustic tweezers to structurally assemble heterogeneous kidney organoids and construct specific regional renal injured assembloids. First, a GFP-labeled organoid was exposed to a cisplatin of 50 μM to promote an acute kidney injury (AKI), which was then assembled with a wild-type organoid. After 2 h (15 min of effect and 15 min of recovery mutual circulation) of sound wave-guided symmetric vortex fusion and 16 h of culturing, heterogeneous kidney organoids could form a strong interaction ( Figure S7a-c, Supporting Information). Cell apoptosis primarily occurred in GFP-labeled injured organoids, whereas the untreated wild-type organoid-maintained podocytes and distal tubule structures ( Figure S8a-c, Supporting Information). Therefore, the feasibility of using acoustic tweezers to construct compound-treated regionally injured heterogeneous assembloids was confirmed.
Subsequently, to explore the interaction between the kidney organoids and their regionally damaged tissue (Figure 4a), we constructed central (Figure 4b and S9a-h, Supporting Information) and peripheral (Figure 4i and S10a-h, Supporting Information) injured AKI assembloids. After acoustic tweezer assembly and 18 h online culturing (Figure 4c,d,f,g,m,n,p,q), both the central and peripheral injured AKI assembloids were completely fused. In addition, the AKI organoids in both assembloids exhibited prominent apoptotic responses to cisplatin, and the hyperproliferative regions were seen mainly in the peripheral regions of the normal organoids (Figure 4e,h,i,o,r,s).
Moreover, the functional maturation affects the ability of the hPSC-derived kidney organoids to predict drug screening and Corresponding IF staining of e) PCNAþ proliferative cell and h) CASP3þ apoptotic cell in GFP-labeled central injured AKI assembloid, consistent with that in the i) non-GFP labeled central injured AKI assembloid. j) Schematic of the defined regional distribution of the central injured AKI assembloid. Area-1 (A1): the AKI organoid region, Area-2 (A2): the intersection of the AKI and normal organoids, Area-3 (A3): the intersection of normal organoids, and Area-4 (A4): the single normal organoid with direct contact with the culture medium. k) Cell apoptotic and proliferation rate within different regions of the central injured AKI assembloid. l) Schematic of the patterned peripheral injured AKI assembloid. Bright-field and fluorescence images of the peripheral injured AKI assembloid after the structural patterning via m,p) acoustic tweezers and n,q) after 18 h online culturing . Corresponding IF staining of o) PCNAþ proliferative cells and r) CASP3þ apoptotic cells in GFP-labeled peripheral injured AKI assembloid, resembling that in the s) non-GFP-labeled peripheral injured AKI assembloid (s). t) Defined regional distribution of the peripheral injured AKI assembloid. u) Cell apoptotic and proliferation rate in different regions of the peripheral injured AKI assembloid. ****P < 0.0001, ***P < 0.001 versus adjacent region. ####P < 0.0001 and "ns," no significant differences in the cell proliferation rate versus adjacent region (n = 20). Scale bars: c-i,l,o,r,s) 500 μm.
www.advancedsciencenews.com www.small-structures.com disease modeling accurately; we thus concentrated on primary kidney proximal tubules due to their role as the primary target for drug-induced nephrotoxicities. As the cubilin-mediated specific endocytosis into proximal tubules was confirmed by dextran uptake, [42,43] we assessed the tubule function within individual kidney organoids and assembloids via incubating them with 70-KDa Rhodamine-labeled dextran, leading to the selective reabsorption of the dextran into proximal tubules ( Figure S11, Supporting Information) despite the relatively immature feature of the proximal tubule as demonstrated previously. [44] Notably, programmable assembloids showed no impairment in the function of kidney organoids ( Figure S11a-h, Supporting Information). However, the colocalization of dextran and LRP2 (proximal tubules marker) was not observed in the injured region of the central injured AKI assembloid ( Figure S11, Supporting Information), suggesting that the uptake function of proximal tubule was damaged induced by cisplatin. All these above findings further confirmed that the acoustically patterned AKI assembloid had the potential to explore the regionally injured heterogeneity.

Conclusion and Discussion
In this study, we demonstrated an ultrasonic phase array-based online structural assembling platform that integrates the precise selection, movement, rotation, and accurate combination of organoids under different conditions for structural construction and culturing heterogeneous assembloids of cisplatininduced local renal injury. In particular, our acoustic tweezer device can construct predesigned holographic acoustic fields in the target space through a precise modulation of the amplitude and phase of the arrays using the electronic system. We rationally switched and reused these fields according to their own characteristics to generate a series of multifunctional holographic fields with controllable parameters and relied on the acoustic radiation force to achieve predetermined control behavior. This unique control strategy can help fine tune and personalize the manipulation and assembly of microsized biological samples, such as organoids.
The application of acoustic tweezers in the form of BAW to fragile biological samples, such as organoids, which are harsh for culturing environmental conditions, is a challenging task. Our acoustic structural assembling technique was successfully demonstrated to stably and precisely trap angular-adjusted organoids and structurally construct the heterogeneous assembloids using a contactless and noninvasive approach. Moreover, the introduced technique does not require the preconstruction of specific-shaped fluidic channels or arrangement of paired transducers. Therefore, precise adjustment of the organoid positions and sequential assembly in a normal Petri dish can be achieved. These features not only facilitate the development of assembloids with differentiated polarized structures, such as the brain tissue, tumor infiltration with metastasis, or multiple organs, but also provide a new technology for exploring the pathophysiological process of diseases and drug target screening.
Our acoustic tweezer assembly system further enriches and refines the applicability of acoustics for individualized control in biological sample processing. The acoustic tweezers technology represented by SAW realizes on-chip high-throughput sorting and detection functions for cells and other organisms by combining it with microfluidic technology, which yield impressive results. Simultaneously, controllable pairing and separation of biological particles can be realized using the harmonic and cross-standing wave effects. Since the SAWs are generated globally on the gas-liquid partition surface, all particles in the acoustic field region are affected. Moreover, the SAWs' morphology variation is limited by the transducer position and the chamber morphology. Therefore, automating the multisamples with multistep sequential assembly is challenging, as collective manipulation is more appropriate for processing homogeneous samples. In the application of the construction of heterogeneous models, such as complex assembloids, the arrangement of individuals in special or sequential order was strictly required and hence in a programmable operation manner. Nevertheless, the existing BAW-type acoustic tweezers, which use a single-array element or acoustic lens, are limited by the singleness of the acoustic field, and the manipulation behavior of particles is monotonous; thus, the integration and application of the characteristics of different forms of fields to achieve various functions are difficult. Compared to these methods, the advantage of our holographic acoustic tweezer assembly platform lies in the dynamic adjustment of the handedness, size, and morphology of the acoustic field used in the electronic systems. Indeed, BAW-type acoustic tweezers can realize various functional manipulations of bioparticles by time division multiplexing of acoustic fields with different characteristics.
The purpose of selectively assembling healthy and injured organoids in a programmed manner is to simulate the disease progression in vivo as heterogeneous damaging of functional units in AKI. Remarkably, current organoid models were globally induced; an assembly of healthy organoids that selectively induces the injury cannot replicate the direct interaction between heterogeneous organoids. Although the future development of microinjection technology may become a solution to complete selective injury, the spillover of injected drugs and the uncertainty of the spreading in the injured area are inevitable. Therefore, assembling heterogeneous organoids can improve the spatial resolution of model construction and most closely mimic the in vivo scenario of AKI.
However, this study has some limitations. Similar to other organoid-related studies, [45] owing to the lack of resident vascular networks in the assembloids, it limits their lifespan cultured in vitro by impeding continuous transport of water and nutrients into the organoid. This limitation may be ameliorated by the subsequent combination of the artificial vessel structure into the assembloids. Furthermore, the presented device partially blocks the optical path of the inverted microscope owing to the limitations caused by the current form of the transducer. In the future study, therefore, we will design and process a hollow torus array to reduce the influence on the optical path. Furthermore, because of a large difference between the horizontal and vertical directions of the vortex acoustic radiation force, the manipulation on the vertical direction was not satisfactorily advanced. In addition, in our future study, we will investigate the compounding of the spatial acoustic field; thus, the goal of our next phase of study will be achieving a controllable levitation based on the current horizontal manipulation.

Experimental Section
Construction of Acoustic Tweezers System: An acoustic tweezer system comprised a 2D matrix array and an excitation system. The 2D matrix array was fabricated and characterized using the optimized protocols from our previous study. [13] It was designed to operate at 3 MHz with 64 elements. Used piezoelectric 1-3 composite was prepared from a piece of PZT-5H bulk ceramic using the dice-and-fill technique. The Piezo CAD simulation confirmed that the required grinding thickness was 41 μm, which was dicing bladed to a kerf of 50 μm and pitch of 187.5 μm. Later, the PZT-5 H 1-3 composite was sputtered with a chrome/gold (Cr/Au) electrode at a thickness of 200 nm/500 nm on both polished surfaces. In front of the PZT-5H 1-3 composite, an acoustic matching layer, compound alumina powder of 23 μm, and Epo-Tek 301 epoxy (Epoxy Technologies, MA, USA) were cast. The matching layer was polished to 250 μm after curing for 24 h at 25°C and sliced into 6 mm Â 6 mm squares containing 64 array elements (8 Â 8). Thereafter, on the back of the PZT-5 H 1-3 composite, each array was cut to 0.75 mm Â 0.75 mm and an electrode incision kerf was determined with a pitch and depth of 187.5 and 200 μm, respectively. Subsequently, Epo-Tek 301 epoxy pasting was used to connect 64 array elements to a single customized flexible circuit that included 64 electrode traces; it was fabricated using polyimide. Next, 64 coaxial cables were welded to 64 channels in a flexible circuit. Finally, a curable resin shell, fabricated by 3D printing, was used to fix the flexible circuit and ultrasonic arrays with a fully cast insulating epoxy. At this stage, the 2D array transducer was completed. Furthermore, we selected the Verasonics Vantage 256 System (Verasonics Inc., WA, USA) to drive the 64 channels of a selffabricated phase array and achieve a better construction and modulation of acoustic fields.
Fabrication of Online Organoids Assembly Platform: OOAP involves an acoustic tweezer system, a cell cultivation system, and a coupling module. The entire platform was mounted on an inverted microscope (IX73, Olympus Corporation, Tokyo, Japan) to assist real-time image-guided assembly and culture observations. For subsequent position adjustment, the phase-array acoustic tweezer was fastened on a precision 3D displacement platform with a standard connector and self-made jig (WeNext Technology Co., Shenzhen, China). In addition, the support skeleton of the culture system was 3D printed using nylon and a reserved humidifying waterway to design and manufacture cell cultivation systems. Around the reserved positions of the skeleton, visible cutting glasses were tightly installed to guarantee a certain sealing condition. In the middle of the cultivation system, a cold-led light source and customized silicone flexible heating plate were placed on the reserved holes. At the top of the cultivation system, we could install complete covers or holes at the center, which allowed the transducer access. After the temperature and gas sensors were installed inside, control wires were connected to the cell culture environment controller (TOWN-INT TECH Co., Shanghai, China) used to adjust the concentration of the perfusion gas and environment temperature. Finally, isolation caps for 60 mm Petri dishes, fabricated using photosensitive resin (WeNext Technology Co., Shenzhen, China), were combined with the Mylar film to form the coupling module.
Numerical Simulation: The COMSOL Multiphysics software was used to simulate the functional field using the finite-element method. Differences between the vortex and simulation parameters are shown in Figure S4a-d, Supporting Information.
Field Synthesis and Analysis: To structure more functional fields serving the structural assembly, the iterative backpropagation algorithm [30] was used to generate eight focal points arranged in a circle around the central position by calculating the transmission phase and amplitude of each element; the phases of the eight focal points varied at π/8 intervals along the circle. We entitled these focal points as the control points. Subsequently, different chiral symmetric vortices could be generated by changing the phase of the control point clockwise and counterclockwise. In addition, the aperture of the vortex could be controlled by changing the spacing between the control points. The weight distribution of the eight control points of the spiral dislocation was adjusted to obtain an asymmetric vortex. Furthermore, we dynamically adjusted and controlled time-division multiplexing for different functional vortices and twins to realize the heterogeneous assembly construction. The duty cycle of the launch events and emission voltage was adjusted using a Verasonics Vantage 256 system (Verasonics Inc., WA, USA) to research ultrasound parameters appropriate for manipulation. Subsequently, we measured the acoustic field pressure distribution using a commercial polyvinylidene fluoride needle hydrophone (NH0500, Precision Acoustics Ltd., Dorchester, UK) and this was compared with the simulation results that were evaluated using FOCUS and COMSOL Multiphysics.
Cell Culture: Human ESC lines H1 (Wi Cell) and H1-expressing GFP under the control of the Oct4 promoter were provided by Dr. Xiao Zhang (GIBH, Guangzhou, China). [46] Both were maintained on Matrigel (Corning, USA)-coated plates in mTeSR Plus medium (Stem Cell Technologies, Vancouver, Canada). All cell types were incubated at 37°C in 5% CO 2 and at relative humidity environment of 95%.
Kidney Organoid Differentiation: Kidney organoids were prepared with slight modifications to that presented in a previous study. [44] In summary, H1 hESCs and H1-expressing GFP were differentiated after reaching a confluence of %70%-80%. For embryoid body (EB) formation, cells were dissociated using dispase (1 mg mL À1 ; Gibco) for 6 min at 37°C, washed three times with PBS (HyClone, Logan, UT, United States) to completely aspirate dispase, and detached with a cell lifter. Subsequently, prior to resuspension in BSA poly (vinyl alcohol) essential lipid medium (BPEL medium, supplemented with 8 μM CHIR99021, 3.3 mM Y27632, and 1 mM β-mercaptoethanol), cells were gently pipetted into evenly sized fragments and evenly distributed into a six-well ultralow attachment plate (Corning, NY, United States). On day 2, half of the medium was replaced with BPEL and 8 mM CHIR99021. For the next stage, initiated on day 3, the self-organized EBs were precipitated in 50 mL falcon tube followed by rinsing twice in DMEM high glucose. Subsequently, EBs were replaced to the 6-well ultralow attachment plate and incubated in the Stage II culture medium without additional growth factors at 37°C and 5% CO 2 . Half of Stage II was changed every two days to promote kidney organoid differentiation until day 14. BPEL and Stage II medium were prepared as described in a previous study. [44,47] Cisplatin-Induced Acute Kidney Injured Organoids: According to an existing protocol, [48,49] normal and GFP-labeled organoids cultured until D9-12 were plated in low-adherence 96-well plates (4515, Corning) and exposed to 50 μM cisplatin (PHR1624, Sigma Aldrich) diluted in Stage II medium. After 24 h of treatment, samples were collected as shown in Figure S5a,b, Supporting Information and used for the construction of local renal injured assembloids. The remaining AKI organoids were used for IF characterization.
Formation of Local Renal Injured Assembloids: Local renal injured assembloids, assembled from six heterogeneous organoids, were constructed to explore the interaction between the normal and AKI organoids. The assembly scheme was as follows. First, we constructed the pattern of the central injured AKI assembloid by surrounding one GFP-labeled AKI organoid with five non-GFP-labeled normal organoids. Second, we assembled the pattern of peripheral injured AKI assembloid, whose vertex was a GFP-labeled AKI organoid, and the other part was formed by five non-GFP-labeled normal organoids. Local renal injured assembloids were cultured to complete fusion to explore the interaction of multiple heterogeneous organoids.
Assembloids Embedding and Frozen Section Preparation: As described in an existing study, [50] organoids and assembloids were fixed with 4% paraformaldehyde/PBS for 30 min at 25°C (room temperature) and washed three times with PBS. Next, the samples were dehydrated in 30% sucrose solution at 4°C overnight to avoid excessive ice crystals in the frozen sections.
Subsequently, the blocks were transferred into a metal container with a bottom size of 1 mm Â 1 mm and gently embedded in Optimal Cutting Temperature cryoblocks (OCT, 4583, Tissue-Tek). Thereafter, the container was placed in À80°C freezer for quick freezing. Finally, the cryopreserved samples were serially sectioned (Leica CM 1520, Leica Microsystems Co., Germany) at a thickness of 10 μm for the next procedure.
Statistical Analysis: Based on the influence of the acoustic field on organoids under the action of different ultrasonic parameters, we used Xviewer (Olympus Corporation, Tokyo, Japan) to calibrate the images or videos captured by the microscope. Later, the speed of organoid movement was tracked and measured using the ImageJ 1.6 software (National Institutes of Health, USA).
All statistical data were presented as mean AE standard error of the mean. One-way analysis of variance was used to analyze differences between three or more groups, followed by the Greenhouse-Geisser correction. Statistical analyses were performed using GraphPad Prism 9.0.0 (GraphPad Software, San Diego, CA, USA), and statistical significance was set at P < 0.05.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.