1. Introduction
The effective intracellular delivery of biomolecular cargos is critical across several biomedical research areas, including gene therapies, cell reprogramming, vaccine development, drug delivery, and nanomedicine. Engineering solutions that facilitate transport of therapeutics and/or genetic engineering reagents into cells play a crucial role in advancing our understanding of cellular processes and developing therapeutic interventions
[1],
[2]. To deliver a desired cargo to a target cell successfully, it must cross the cell’s semi-permeable outer lipid membrane. To date, various approaches have been developed, including passive approaches based on viral
[3] or non-viral vectors
[4] that can facilitate the uptake process, as well as active approaches that are able to open micropores along cell membranes (e.g., sonoporation
[5], electroporation
[6], mechanical membrane disruption
[7],
[8],
[9], thermal membrane disruption
[10],
[11], and microinjection
[12]).
Among the vector-based methods
[13], various viral vectors are widely used in clinical and translational studies because of their efficiencies and specificities
[14]. However, clinical-grade viral vectors are limited by their high manufacturing costs, complex processing requirements, potential insertional genotoxicity and immunogenicity, and relatively low packaging capacities
[15]. Given these drawbacks, non-viral vectors such as polymer
[16],
[17],
[18], inorganic material
[19], lipid
[20], and exosome-based nanocarriers
[21],
[22] offer attractive alternatives. Different cargo types (e.g., therapeutic molecules, gene-editing tools, reprogramming factors, and proteins) can be loaded into nanocarriers and then delivered to target cells
[23]. Once the nanocarriers enter a cell, additional stimuli are often used to trigger the cargo release process
[24]. Even though there have been significant advances in the development of different nanocarriers to enhance the delivery efficiency
[25],
[26],
[27],
[28], multiple challenges remain to be addressed. These obstacles include nanocarrier aggregation
[29], slow uptake of cargo entering the cytoplasm, high ratios of residual nanocarriers (i.e., those left in the extracellular medium), and a lack of efficient and controllable mechanisms that facilitate cell-nanocarrier contact, membrane-nanocarrier bonding, and ultimately nanocarrier transport across the cellular membranes
[30].
Vector-free approaches, which can actively facilitate the formation of pores at cellular membranes for the diffusion of cargo into cells have attracted great interest
[23]. However, these approaches are often limited by low recovery rates, dependence on certain cell types, and/or costly equipment
[31]. These technologies are also susceptible to damaging critical cellular components, lowering cell viability, and/or reducing the effectiveness/functionality of the cargo delivered
[32]. For example, conventional microbubble-based sonoporation approaches usually require specific contrast agents contained within microbubbles
[5], which can cause irreparable damage to targeted cells, lowering their viability (e.g., to below 50% in some studies)
[33],
[34].
The convergence of acoustic waves and microfluidics (i.e., acoustofluidics) offers a powerful tool for active nanomaterial manipulation and delivery
[22],
[35],
[36],
[37],
[38],
[39],
[40],
[41],
[42],
[43],
[44],
[45],
[46],
[47],
[48],
[49],
[50]. Several groups have explored the development of various bubble-free sonoporation approaches
[51] to address issues associated with microbubble contrast agents, which leverage focused acoustic waves
[33], surface acoustic waves
[52],
[53], spiraling acoustical tweezers
[54], or hyperfrequency acoustic streaming
[55]. The throughputs of these approaches are still low because they are limited to the sonoporation of cells in a small area targeted by focused acoustic waves, in small microfluidic chambers, or in regions near the hyperfrequency acoustic source. To increase sonoporation throughput, we have recently developed a platform based on the application of standing bulk acoustic waves in a capillary that achieved high-throughput delivery of expression plasmid to human hematopoietic stem cells and other primary cell types
[56]. The motivation was to overcome the channel clogging by cell debris limiting prior methods of cell deformation in physical constrictions in microfluidic channels
[56],
[57]. Previous sonoporation approaches induce the opening of membrane pores
[23],
[58],
[59], but the cargo nevertheless relies on diffusion-based transport for entry into cells. This diffusion-based mechanism relies on high concentrations of cargo in the extracellular medium. It also leads to low cargo usage rates (i.e., ratios between cargos delivered to cells and cargos left in the extracellular medium), which is a critical factor that must be considered when delivering expensive cargos.
In this study, we present an acoustofluidics-based intracellular delivery approach that can be used to deliver functional cargos encapsulated in metal organic framework-based nanoparticles efficiently and biocompatibly to immortalized cancer cell lines (e.g., U937 and HeLa). Our approach leverages standing acoustic waves generated in a glass capillary coated with cargo-encapsulated nanoparticles. By tuning the acoustic wave frequency to align the generated acoustic pressure node line along the capillary wall, our approach can push cells in flow to the nanoparticle-coated capillary wall using acoustic radiation forces, achieving controllable cell-nanoparticle contact and facilitating membrane-nanoparticle attachment. This feature decreases the random nature of cell-nanoparticle interactions through Brownian movement that limits conventional nanocarrier-based delivery approaches
[55],
[60]. Our approach facilitates nanoparticle uptake by pressing the cells onto the nanoparticle coated capillary as well as increasing the membrane stress via acoustic radiation forces. Additionally, our approach does not need the bubbles or special acoustic contrast agents that are usually required in conventional sonoporation methods.
To investigate the mechanism of our approach and to characterize its performance, we performed finite element simulations and a series of experiments to deliver different types of cargo to U937 and HeLa cancer cell lines. The nanocarriers used in this research are based on zeolitic imidazolate framework-8 (ZIF-8) nanoparticles, that is, porous nanomaterials that can be loaded with different biomolecular cargos such as camptothecin, doxorubicin (DOX)
[61], DNA
[62], and clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9
[63]. Two types of cargo were used in this work, DOX and fluorescein isothiocyanate (FITC)-labeled bovine serum albumin (FBSA), to investigate the delivery of small-molecule and protein-based cargos. Both simulations and experimental results show that our acoustofluidic approach enables active cell-nanoparticle contact to facilitate cell-nanoparticle attachment as well as increases the membrane permeability to assist in nanoparticle uptake. Moreover, different cargos can be successfully delivered into the cells with enhanced efficiency. Altogether, this contrast agent-free, acoustofluidic method informs future intracellular delivery technologies that are biocompatible, efficient, and controllable for therapeutic applications and biophysical studies.
2. Experiments
2.1. Synthesis of cargo-loaded ZIF-8 nanoparticles
Preparation of DOX-loaded ZIF-8 (denoted as DOX@ZIF-8) nanoparticles was conducted following procedures reported previously
[50]. Different loading rates can be achieved by using different DOX concentrations during synthesis. We used 1 mL of 6 mg∙mL
−1 DOX in the experiments. The solutions were mixed with 200 μL of 0.8 mol∙L
−1 Zn(NO
3)
2 (pH adjusted to 8.0 using NaOH) and agitated vigorously for 10 min, followed by a dropwise addition of 2 mL of 3 mol∙L
−1 2-methylimidazole solution. After 15 min, the DOX encapsulated nanoparticles were collected following three cycles of centrifugation at 8000 r∙min
−1, washed with water, and dried at 60 °C overnight. For the preparation of FBSA-loaded ZIF-8 (denoted as FBSA@ZIF-8) nanoparticles, 100 μL of 2 mg∙mL
−1 FBSA was added into 2 mL of 3 mol∙L
−1 2-methylimidazole solution and stirred for 30 min. Then, 200 μL of 0.4 mol∙L
−1 Zn(NO
3)
2 was added to the mixture and magnetically stirred for 15 min. The yellow solution was centrifuged at 8000 r∙min
−1 and washed with deionized (DI) water three times, followed by drying at 65 °C overnight.
2.2. Device fabrication
The acoustofluidics-based intracellular delivery device was composed of a lead zirconate titanate (PZT) plate-type transducer (Steminc-Piezo, USA) and a square capillary (VitroCom, USA). The transducer was fixed onto a glass slide (VWR, USA) using two pieces of double-sided adhesive tape attached to polyurethane spacers (thickness: 2 mm). Then, the capillary was bonded to the top of the transducer using epoxy (Permatex, USA). To generate standing acoustic waves, it is important to consider the acoustic frequency and the size of the glass capillary
[64]. A transducer with a resonant frequency of 710–720 kHz was used to generate standing bulk acoustsic waves in a 200 μm × 200 μm × 50 mm glass capillary. At this frequency, our device can generate a pressure node line along the capillary wall. Two tubes were connected to the end of the glass capillary. To coat the synthesized nanoparticles (e.g., ZIF-8, DOX@ZIF-8, and FBSA@ZIF-8) onto glass capillaries, nanoparticle solutions were injected into the glass capillaries, which were then placed horizontally in an oven. After drying each capillary at 65 °C overnight, a layer of nanoparticles remained on the bottom wall of the capillary.
2.3. Device operation
The fabricated device was mounted to an inverted microscope (Nikon, Japan). Cells were loaded into a 1 mL syringe (BD Biosciences, USA) and pumped into the glass capillary at a flow rate controlled by an automatic syringe pump (neMESYS, Cetoni, Germany). The excitation signal for the piezoelectric transducer was generated using a function generator (Agilent, USA) and then amplified with a power amplifier (Amplifier Research, USA). During the delivery experiments, a fan was used to cool the acoustofluidic device.
2.4. Release of FBSA and DOX from ZIF-8 nanoparticles
To evaluate the release of cargo from the cargo-encapsulated ZIF-8 nanoparticles, 2 mg of nanoparticles were dispersed in 10 mL of phosphate-buffered saline (PBS) at either pH 5.0 or 7.2. At different time points (1, 3.5, 4.5, 5.5, 15.5, 18.5, 36, 72, 96, 120, 140, 160, 180, and 240 h), 1 mL was taken from the solution and centrifuged at 6000 r∙min−1. For each measurement, 200 µL of the supernatant was added into a 96-well plate and measured with a microplate reader. After each measurement, the sample was returned to the bulk solution for future analysis.
2.5. Cell culture and viability assay
U937 and HeLa cells were purchased from ATCC (Virginia, USA). U937 and HeLa cells were cultured in Roswell Park Memorial Institute-1640 medium (Thermo Fisher Scientific, USA) and Dulbecco's Modified Eagle Medium (Thermo Fisher Scientific, USA), respectively, supplemented with 10% (v/v) fetal bovine serum. The medium was exchanged every two days. To evaluate cell viability, cells were stained with Calcein acetoxymethyl ester (C3100MP, Life Technologies, USA), and those exhibiting green fluorescence were counted as live cells. Cell viability was determined by the ratio of live cells to the total number of cells.
2.6. Delivery efficiency
The delivery efficiency was determined by counting the numbers of fluorescent cells in acquired microscopy images. The cell intensity was calculated using ImageJ (National institutes of Health, USA). Flow cytometry (BD FACSCanto II, USA) was used for the quantitative evaluation of the nanoparticle delivery efficiency. The collected cells were washed and redispersed in PBS before flow cytometry. For each characterization group, three samples were tested to calculate the standard deviation. The data were analyzed using FlowJo (BD Biosciences, USA).
3. Results and discussion
3.1. Mechanism and design of the acoustofluidics-based intracellular delivery device
Fig. 1(a) shows a schematic of our acoustofluidics-based intracellular delivery device, which is composed of a square glass capillary coated with cargo-encapsulated nanoparticles and a piezoelectric transducer to generate incident acoustic waves. Details of the device fabrication can be found in the Section 2. Notably, the capillary’s inner dimension is smaller than a quarter wavelength of the acoustic waves generated from the transducer. This geometric constraint enables the generation of a pressure node line along the capillary’s inner wall (see
Fig. 1(b) for a cross-sectional illustration). As cells are flowed through the capillary coated with cargo-encapsulated nanoparticles (
Fig. 1(c)), the acoustic radiation force moves the cells to the pressure node line and pushes them against the capillary wall, where the cells can directly contact the coated nanoparticles. This controllable cell-nanoparticle interaction can facilitate nanoparticle transfer and attachment on the cell’s membrane. Moreover, the acoustic radiation force applied on the membrane and the reaction force applied through the cell membrane-capillary wall interaction can increase membrane stress and slightly deform the cells. These forces facilitate nanoparticle entry across the cell membrane. Based on these capabilities, our acoustofluidics-based intracellular delivery device enables controlled delivery of cargo-encapsulated nanoparticles to cells.
In this study, ZIF-8 nanoparticles were used as the nanocarriers because of their ability to encapsulate various molecular cargos and deliver them into cells through endocytosis
[61],
[65].
Fig. 1(d) provides a flow diagram detailing the steps for delivering cargo-encapsulated ZIF-8 nanoparticles and releasing the cargo. Cells that are flowed through our device with acoustic waves on are pushed to the glass capillary wall, enabling cell-nanoparticle contact. This process facilitates the cell-nanoparticle bonding and the intracellular entry of the nanoparticles via endocytosis. When the cargo-encapsulated ZIF-8 nanoparticles are delivered to the cytoplasm within endosomes, cargo is released by endosomal escape as the ZIF-8 nanocarriers gradually break down in this acidic environment
[63],
[66]. Note that no acoustic waves are applied to cells during the cargo release process. Whether acoustic waves can facilitate the cargo release is an interesting question that will be explored in future studies.
3.2. Capillaries coated with cargo-encapsulated ZIF-8 nanoparticles
We performed a series of experiments to characterize the properties of the cargo-encapsulated ZIF-8 nanoparticles used in this study and better understand the cargo encapsulation and release processes. Two types of cargos (i.e., DOX and FBSA) were used to investigate our device’s ability to deliver small molecule and protein-based cargos
[61],
[67]. By following the procedures summarized in the Section 2, we synthesized DOX@ZIF-8 and FBSA@ZIF-8 nanoparticles and then characterized the synthesized nanoparticles using multiple techniques, including transmission electron microscopy (TEM), dynamic light scattering (DLS), powder X-ray diffraction (PXRD), and ultraviolet-visible (UV-Vis) spectroscopy (
Figs. 2).
From the TEM images (
Figs. 2(a) and
(b)), we found that the diameters of DOX@ZIF-8 and FBSA@ZIF-8 nanoparticles were approximately 150 nm. The size distributions measured by number-weighted DLS measurements (
Fig. 2(c)) indicated that 60%, 83%, and 82% of the particles were in the 100–250 nm range for ZIF-8, DOX@ZIF-8, and FBSA@ZIF-8, respectively. The cargo-loaded ZIF-8 nanoparticles exhibited a more polydisperse size distribution compared to ZIF-8. This could be due to aggregation of the amorphous precursors induced by FBSA and DOX encapsulation during the ZIF-8 nucleation stage
[67]. The PXRD distributions (
Fig. 2(d)) for ZIF-8, DOX@ZIF-8, and FBSA@ZIF-8 nanoparticles show similar trends, indicating that these nanoparticles have similar crystalline structures. The packaging of cargo into the ZIF-8 nanocarriers was characterized using UV-Vis spectroscopy. The FBSA@ZIF-8 nanoparticles (
Fig. 2(e)) exhibit an absorbance peak at 500 nm, indicating the successful encapsulation of FBSA proteins
[67]. The DOX@ZIF-8 nanoparticles (
Fig. 2(f)) have an absorbance peak at 490 nm, indicating successful loading of DOX molecules
[68]. The pH-dependent cargo release was investigated by dissolving cargo-encapsulated nanoparticles in solutions with different pH values and measuring the absorbance spectrum change through UV-Vis spectroscopy. The result (Fig. S1 in Appendix A) shows that cargo can remain in the ZIF-8 nanoparticles in a non-acidic (pH 7.4) environment for at least 10 days. In contrast, the ZIF-8 nanoparticles gradually decompose in an acidic (pH 5.0) environment.
Next, the prepared DOX@ZIF-8 and FBSA@ZIF-8 nanoparticles were coated on the inner walls of the glass capillaries (Section 2). The uncoated capillaries presented a smooth wall (
Fig. 2(g)) and no fluorescence within the channel (
Fig. 2(j)). Bright-field and fluorescence microscopy was used to confirm successful coating of DOX@ZIF-8 (
Figs. 2(h) and
(k)) and FBSA@ZIF-8 (
Figs. 2(i) and
(l)) nanoparticles onto the capillary wall, respectively. The capillaries coated with ZIF-8 nanoparticles were subsequently used for intracellular delivery experiments.
3.3. Acoustofluidics-enabled cell movement, deformation, and permeability change
To investigate our device’s capabilities, such as acoustic wave-induced cell movement, deformation, and permeability change, we performed both finite element simulations and cell tracking experiments. The finite element simulations were performed in COMSOL Multiphysics 5.4 (Stockholm, Sweden) at an excitation frequency of 710 kHz and an amplitude of 20–25 Vpp (matching the experimental parameters, Vpp: voltage-peak to peak). The simulation results are given in
Figs. 3(a)–(d). The displacement field (
Fig. 3(a)) shows a resonant mode at the excitation frequency with dominated
z-direction displacements. The standing acoustic pressure field in the fluid domain (
Fig. 3(b)) shows a low-pressure region near the top boundary. The simulated acoustic radiation force (
Fig. 3(c)) shows a strong +
z component. Therefore, cells randomly distributed in the capillary can be moved to the top wall, as illustrated by the cell movement trajectories induced by the acoustic waves (
Fig. 3(d) and Movie S1 in Appendix A). To validate the cell movement predicted by finite element simulations, experiments were performed by loading cells into a capillary and monitoring cell movement induced by acoustic waves through a microscope. The captured images (
Figs. 3(e)–(g)) show that randomly distributed cells in a capillary can be moved rapidly to the capillary’s top wall when acoustic waves are generated.
Fig. 3(f) reveals the cell trajectories by stacking together multiple frames of a recorded video. This experimentally acquired cell trajectory is consistent with the simulation result in
Fig. 3(d). In addition to the case without flowing cell culture medium through the system, we performed similar experiments where cell solutions were passed through capillaries at a flow rate of 5 μL∙min
−1. Movies S2 in Appendix A show our device can push flowing cells to the capillary wall using acoustic waves.
To investigate our device’s capability to increase membrane permeability, we loaded a solution containing both U937 cells and DOX@ZIF-8 nanoparticles into two uncoated capillaries (one with acoustic waves off and one with the acoustic waves on). The tested cells were collected, washed, and imaged using fluorescence microscopy. In the control group, where acoustic waves were absent, the observed fluorescence intensities of the cells (
Fig. 3(h)) were notably weak. Conversely, in the group exposed to active acoustic waves, the cells exhibited higher fluorescence intensities from DOX (
Fig. 3(i)), suggesting markedly enhanced uptake of DOX@ZIF-8 nanoparticles. These results indicate that our acoustofluidic device increases nanoparticle uptake. The observed enhancement in nanoparticle internalization could be related to the increased lipid membrane stress and slight cell deformation observed during acoustofluidic treatment (Fig. S2 in Appendix A)
[58],
[69], which are induced by the acoustic radiation force directly applied to the cells, the reaction force from cell–cell interactions, and the reaction force from the cell-capillary wall interaction.
As indicated by the numerical and experimental results, our acoustofluidic device is capable of overcoming Brownian movement between nanoparticles and target cells to facilitate cell-nanoparticle attachment and changing membrane permeability. The combination of these capabilities enables effective intracellular transport of the nanoparticles. Next, we explored whether our acoustofluidic device could be used for delivering ZIF-8 nanoparticles loaded with different cargos (i.e., FBSA and DOX) into various types of cells (i.e., U937 and HeLa cells).
3.4. Delivery of cargo-encapsulated ZIF-8 nanoparticles to different types of cells
We performed additional experiments to demonstrate the delivery of cargo-encapsulated ZIF-8 nanoparticles using our acoustofluidics-based approach. To investigate the delivery of ZIF-8 nanoparticles loaded with different cargos (e.g., FBSA and DOX), experiments were performed by running two groups of U937 cells (concentration: 2 × 10
6 cells mL
−1, flow rate: 5 μL·min
−1) through acoustofluidic devices, one with FBSA@ZIF-8 and one with DOX@ZIF-8 coatings. The collected cells were washed three times with PBS, seeded in a Petri dish, cultured for 30 min, and then imaged with a fluorescence microscope (TE2000-U, Nikon, Japan). The acquired microscopy images are shown in
Figs. 4(a)–(d). The control groups without acoustic waves show low fluorescence intensities (
Figs. 4(a) and
(c)). In contrast, higher fluorescence intensities are observed for cell populations exposed to acoustic waves (green in
Fig. 4(b) and red in
Fig. 4(d)), indicating higher nanoparticle delivery yields. To characterize the intensity differences between groups with acoustic waves on and off, the mean fluorescence intensity for cells in each group was calculated using ImageJ (Fig. S3 in Appendix A). By comparing these results for the delivery of DOX@ZIF-8 nanoparticles, the mean fluorescence intensity for the acoustics-on group was at least twice that of the acoustics-off group. The delivery of FBSA@ZIF-8 nanoparticles by the acoustic waves shows greater intensity enhancement.
To investigate the effects of nanocarrier (ZIF-8 only) decomposition and acoustic waves on cell viability, calcium AM-based cell viability assays (detailed procedures in Section 2) were performed on U937 cells treated under three different conditions: ① acoustics off without ZIF-8 coating, ② acoustics on without ZIF-8 coating, and ③ acoustics on with ZIF-8 coating. The results are presented in
Fig. 4(e). Comparing the results of conditions ① and ②, we found a slight decline in cell viability, from 93% to 85%, when acoustic waves were applied in devices without ZIF-8 nanocarriers. Comparing the results of conditions ② and ③, we observed that cell viability at 24 h decreased slightly from 85% to 78%. After the 24 h period, the viability for the ZIF-8 coating group is about 80%. These results indicate that the ZIF-8 nanocarrier decomposition and the used acoustic waves have minimal effects on cell viability.
To determine whether our acoustofluidic device can deliver cargo-encapsulated ZIF-8 nanoparticles into cells and whether the cargo released into the cells remained functional, we performed experiments using U937 cells under different delivery conditions (i.e., acoustics off/on with DOX@ZIF-8 coating). The cargo used in these experiments is DOX, a common chemotherapeutic agent that can cause cell apoptosis
[70]. We examined the DOX-related cytotoxicity resulting from our nanoparticle delivery by characterizing cell viability at 0.5, 2, 8, and 24 h after flowing U937 cells through our acoustofluidic device. The collected cells were washed with PBS and cultured prior to performing calcium AM-based viability assays. The results in
Fig. 4(f) show that cell viability for the group without acoustic waves is about 90% in 0.5 to 24 h. However, for the group treated with acoustic waves, the cell viability at 24 h declines to 53%, demonstrating the effectiveness of acoustofluidics-facilitated delivery. The comparison between groups with and without acoustic waves indicates that the DOX@ZIF-8 nanoparticles have been delivered into cells, releasing the DOX cargo. Moreover, the DOX cargo released into cells remains functional
[61].
Additional experiments of the DOX@ZIF-8 delivery and DOX release processes were performed. HeLa cells suspended in PBS were treated using acoustofluidic devices coated with DOX@ZIF-8 nanoparticles. To determine how cells interact with the coated nanoparticles on the capillary walls, we recorded a video of HeLa cells flowing through a capillary (cell concentration: 2 × 10
6 cells mL
−1, flow rate: 5 μL⸱ min
−1). Our results (
Fig. 5(a) and Movie S3 in Appendix A) show that the acoustic waves generated by a PZT transducer installed at the bottom side of the capillary can force cells to move through the capillary along the top wall. We observed that the original nanoparticle coating (∼5 μm thick) becomes thinner (
Fig. 5(b)), after running 1 mL of the cell solution through our device. This observation results from nanoparticles coated on the capillary wall being transferred to the membranes of the targeted cells, as shown in a bright-field microscopy image acquired at 5 min after the experiment (
Fig. 5(c)). These experimental observations confirm that our acoustofluidic device can facilitate cell-nanoparticle contact. To investigate the release of cargo from DOX@ZIF-8 nanoparticles, the collected HeLa cells were washed with PBS, cultured for 6 h, stained, and then imaged via fluorescence microscopy. The acquired fluorescence images show that DOX (red) was well distributed in the cytoplasm (
Fig. 5(d)). The observed DOX cargo was released from the successfully delivered DOX@ZIF-8 nanoparticles via an endosomal escape process whereby the ZIF-8 framework begins to decompose when exposed to the acidic environment of the maturing endosome. To test if the DOX cargo released into HeLa cells is functional, microscopy images (Fig. S4 in Appendix A) were recorded at 8 h post-delivery. Here, the test group exposed to acoustic waves (Fig. S4(b)) has more apoptotic cells and fewer cells attached to the Petri dish, compared to the control group that were not exposed to acoustic waves (Fig. S4(a)). The viability of HeLa cells significantly decreases over time (e.g., decreases to 73% after 2 h, 55% after 8 h, and 42% after 24 h), further confirming that the release of the cytotoxic chemothereutic agent reduces viability of the treated populations (Fig. S4(c)). To assess the efficiency of our acoustofluidic device quantitatively, we studied the delivery of DOX@ZIF-8 nanoparticles to HeLa cells and further characterized the delivery efficiencies through flow cytometry (Section 2 for procedures). From the results in
Figs. 5(e) and
(f), we find that the delivery efficiencies for cases without and with acoustic waves are ∼18% and ∼91%, respectively. The results also show that the measured mean fluorescence intensity increases from 2 × 10
2 (acoustics off) to 5 × 10
2 (acoustics on), indicating more DOX@ZIF-8 nanoparticles were delivered to the cells.
Through the above studies presented in this subsection, we have investigated and characterized different aspects of the intracellular delivery performance of our acoustofluidic approach. Our results demonstrate that ZIF-8 nanoparticles loaded with small molecules and proteins can be successfully delivered into acoustofluidically manipulated cells. The feasibility of our approach for suspension and adherent cells (such as U937 and HeLa cells) has been validated, and cell viability measurements (
Fig. 4(e)) confirm that the acoustic waves and ZIF-8 decomposition minimally impact cell viability. Data from our flow cytometry studies (
Figs. 5(e) and
(f)) indicate that the delivery efficiency can be enhanced by introducing acoustic waves. We also investigated the mechanism of delivering cargo-encapsulated ZIF-8 nanoparticles and the release of functional cargos. The results (
Figs. 5(a)–(d)) indicate that our acoustofluidic device enables controllable contact between flowed cells and nanoparticles coated on the wall of the device’s microcapillary, facilitating cell-nanoparticle attachment.
4. Conclusions and Prospects
This study presents an effective acoustofluidics-based intracellular nanoparticle delivery platform that eliminates the need for bubbles or other specialized acoustic contrast agents typically required by conventional sonoporation methods
[71]. Specifically, our approach leverages standing acoustic waves generated in a glass capillary to direct flowing cells to the capillary wall, which is coated with cargo–encapsulated ZIF-8 nanoparticles, thereby promoting membrane-nanoparticle attachment. Our experimental results revealed that cells were pushed against the capillary wall, slightly deformed, and transported along the wall, indicating that cells were subjected to acoustic radiation forces and shear forces. Because of these forces, the cell membrane stress increases, potentially altering the membrane permeability. Additionally, our results show that the ZIF-8 nanoparticles are capable of encapsulating small molecules and protein-based cargos while enabling the controlled release of these cargos upon intracellular uptake. The combination of the above mechanisms enables our approach to achieve effective intracellular delivery.
Applying our acoustofluidic approach, we have performed a series of experiments to deliver ZIF-8 nanoparticles loaded with different cargos (DOX or FBSA) to two immortalized cancer cell lines (e.g., U937 cells and HeLa cells). The intracellular delivery of our nanoparticles (
Figs. 5(e) and
(f)) suggests that acoustic waves can significantly enhance the delivery efficiency of nanomaterials by inducing cell membrane distortion and constant membrane-nanoparticle contact. Furthermore, our results (
Figs. 4(a) to
(d)) confirm that two types of cargos (DOX and FBSA) can be successfully delivered to the cells, demonstrating the potential applications of our approach for delivering small molecules and protein-based cargos to cells beyond Brownian motion within microfluidic systems. Through investigation, characterization, and validation, this study has established an acoustofluidics-based intracellular delivery method that does not rely on microbubbles and advances state-of-the-art sonoporation technologies
[72]. In the future, we will investigate the feasibility of our approach for delivering other types of cargo, such as nucleic acids and CRISPR/Cas9-based gene-editing constructs, and other carriers. We will test our approach using difficult-to-transfect primary human cell types (e.g., hematopoietic stem and progenitor cells and T cells) and investigate the feasibility of using the treated cells for
in vivo studies. We anticipate that our acoustofluidics-based intracellular delivery platform can be applied for rapid manufacturing of engineered cell products for gene/cellular therapies as well as fundamental biophysical studies of cell mechanics in a controllable, scalable, and biocompatible manner.
Declaration of competing interest
Jason N. Belling, Liv K. Heidenreich, Steven J. Jonas, and Paul S. Weiss have patents related to this work. Zhishang Li, Zhenhua Tian, Joseph T. Rich, Haodong Zhu, Zhehan Ma, Hunter Bachman, Liang Shen, Yaosi Liang, Xiaolin Qi, Yao Gong, Shujie Yang, Wenfen Zhang, Peiran Zhang, Yingchun Fu, Yibin Ying, Yanbin Li, and Tony J. Huang declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
We acknowledge the support from the National Institutes of Health (R01GM141055), the National Science Foundation (CMMI-2104295), the China Scholarship Council, and the NIH/NCATS UCLA CTSI (UL1TR001881) through the UC Center for Accelerated Innovation. This material is based upon work supported by the National Science Foundation Graduate Research Fellowship (1644868). Paul S. Weiss and Steven J. Jonas acknowledge support from the UCLA Innovation Fund MedTech Innovator Award and the Challenge Initiative at UCLA, and seed funding provided through a UCLA David Geffen School of Medicine Regenerative Medicine Theme Award. Steven J. Jonas acknowledges the support provided by the NIH Common Fund through an NIH Director’s Early Independence Award co-funded by the National Institute of Dental and Craniofacial Research and Office of the Director, NIH under award number (DP5OD028181). Jason N. Belling thanks the NIH for a predoctoral fellowship supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under Award Number (F31HL149356).
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2024.11.030.
PDF
(1761KB)
