Letter Cite This: Nano Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/NanoLett
Intracellular Delivery of Nanomaterials via an Inertial Microfluidic Cell Hydroporator Yanxiang Deng,† Megan Kizer,‡ Miran Rada,⊥ Jessica Sage,⊥ Xing Wang,‡ Dong-Joo Cheon,⊥ and Aram J. Chung*,†,§,∥ †
Department of Mechanical, Aerospace, and Nuclear Engineering and ‡Department of Chemistry and Chemical Biology, Center for Biotechnology and Interdisciplinary Studies (CBIS), Rensselaer Polytechnic Institute (RPI), Troy, New York 12180, United States § School of Biomedical Engineering and ∥Department of Bio-Convergence Engineering, Korea University, Seoul 02841, Republic of Korea ⊥ Department of Regenerative and Cancer Cell Biology, Albany Medical College (AMC), Albany, New York 12208, United States S Supporting Information *
ABSTRACT: The introduction of nanomaterials into cells is an indispensable process for studies ranging from basic biology to clinical applications. To deliver foreign nanomaterials into living cells, traditionally endocytosis, viral and lipid nanocarriers or electroporation are mainly employed; however, they critically suffer from toxicity, inconsistent delivery, and low throughput and are time-consuming and labor-intensive processes. Here, we present a novel inertial microfluidic cell hydroporator capable of delivering a wide range of nanomaterials to various cell types in a single-step without the aid of carriers or external apparatus. The platform inertially focuses cells into the channel center and guides cells to collide at a T-junction. Controlled compression and shear forces generate transient membrane discontinuities that facilitate passive diffusion of external nanomaterials into the cell cytoplasm while maintaining high cell viability. This hydroporation method shows superior delivery efficiency, is high-throughput, and has high controllability; moreover, its extremely simple and low-cost operation provides a powerful and practical strategy in the applications of cellular imaging, biomanufacturing, cell-based therapies, regenerative medicine, and disease diagnosis. KEYWORDS: Intracellular delivery of nanomaterials, macromolecule delivery, cell hydroporator, inertial microfluidics
R
Alternatively, membrane-disruption-based methods introduce transient nanopores in the cell membrane utilizing physical means (for example, mechanical force,14 electric fields,15 thermal deviation,16 and electromagnetic radiation17) to accommodate the influx of exogenous nanomaterials into the cytoplasm. Because delivery is achieved by passive cargo transport through the created membrane openings, the process is inherently independent of carrier types and properties. Though various nanomaterials dispersed in solution have been delivered to various cell types, it should be noted that current physical cell membrane disruption approaches cause excessive damage to both cells and delivery nanomaterials and suffer from limited throughput and inefficient protocols.1,18,19 In short, it is still challenging to find a method with high efficiency and throughput, noninvasiveness, cell type and cargo material independency, dosage controllability, and simple and low-cost operation.
ecent developments in nanoscience and nanotechnology have created a wide variety of functional nanomaterials and engineered biomolecules with applications in diverse biological scenarios, ranging from visualization of subcellular structures and dynamics, cell-based therapy, and gene editing to regenerative medicine and fundamental biology.1−5 To fully realize the promise from each nanomaterial, they should reach their biological targets with high efficiency and specificity.6 To achieve effective cytosolic delivery of nanomaterials, various methods and techniques have been reported.7−13 Current state-of-the-art intracellular delivery methods for nanomaterials can be categorized into either carrier-based or membrane-disruptionbased techniques. Carrier-based approaches are based on biochemically prepared cargo assemblies such as reconstituted viruses, vesicles, peptides, and functionalized nanostructures that are delivered mainly through endocytosis. While carrier-based approaches offer effective nucleic acid delivery (i.e., transfection) in vivo and in vitro, there are still concerns regarding safety, adverse immune response, slow delivery, high complexity, and cost of preparation for viral vectors in addition to inefficient and slow delivery as well as inconsistent results for nonviral carriers.1 © XXXX American Chemical Society
Received: February 18, 2018 Revised: March 19, 2018
A
DOI: 10.1021/acs.nanolett.8b00704 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 1. Inertial microfluidic cell hydroporator (iMCH) for intracellular delivery of nanomaterials. (a) Schematic illustrating the design and operating principles of iMCH. Cell-wall collision and fluid-shear create nanopores allowing nanocargos to diffuse through. (b) High-speed microscope images showing the cell collision process with the channel wall containing a sharp tip at the T-junction. (c) Delivery of FITC-dextran into MDA-MB-231 cells using iMCH. Fluorescent images showing delivery via (i) endocytosis and (ii) iMCH after 18 h. For quantitative analysis, flow cytometry data is also presented for (iii) endocytosis and (iv) iMCH. All scale bars represent 50 μm.
collide with the channel31 (see Figure 1a and b). An inertia-based approach is chosen here because it intrinsically allows nearperfect cell positioning in the channel midpoint (>99% focusing efficiency) and a high-throughput operation (>1 000 000 cells min−1 for a single channel). Unlike cell squeezing, the iMCH method is insensitive to cell size heterogeneity and free from channel clogging or cell lysis issues. Consistent molecule delivery is anticipated because uniform compressive forces and shear flow from fluid inertia are applied to create nanopores. In addition, through the created nanopores, freely suspended target materials are passively diffused into the cytosol; thus, the delivery mechanism does not largely depend on material properties. For highly efficient intracellular delivery, we employed a tip (Figure S1) instead of a flat wall at the T-junction, concentrating stress and thereby establishing enhanced nanomaterial delivery as shown in Figure S2 (near-zero clogging was seen). Note that we used standard single-layer PDMS-glass-based devices (see Figure S3) and that the entire delivery process was executed using a conventional low-cost infusion-only syringe pump excluding system complexity (no external apparatus). Notably, the iMCH platform does not require a costly microscope to operate because cells are passively manipulated in the fluidic channel, realizing a true lab-on-a-chip system and maximizing the potential utility of the system. We first aimed to deliver fluorescein isothiocyanate (FITC) conjugated dextran into an MDA-MB-231 cell line using iMCH. Dextran with molecular weights of 3 kDa (specifically the compound ranges from 3 to 5 kDa) and 70 kDa (approximately 12 nm diameter near-spherical molecule) was chosen to mimic the typical protein delivery.25 Figure 1c panels ii and iv show the delivery of 3 kDa FITC-dextran with iMCH, and for our control, the identical material was delivered purely based on endocytosis (Figure 1c panels i and iii). As can be seen, the iMCH method demonstrated a drastically higher efficiency of dextran delivery compared to endocytosis. Note that it is reported that, upon the creation of transient pores, the cell repairs its membrane within a few minutes, not allowing further inward or outward material transfer.32 To provide enough time for exogenous materials to diffuse into the cytoplasm, we reduced the concentration of calcium in the solution. Calcium inrush mediates cell-membrane repair;33 thus, when the calcium concentration is controlled, we were able to achieve improved delivery results while maintaining
To address the challenges of current methodologies, recent advancements in microfluidics with nanoengineering have allowed a wide variety of new emerging solutions. 1,20 Representative methods include microfluidic nanostraws,9,21 electroporation,18,22 cavitation,23,24 and cell squeezing.25,26 Among them, considering the simplicity and scalability of design and operation, the cell squeezing method has gained a lot of interest. Briefly, the technique forces cells to pass through microfluidic constrictions of half to one-third of the cell’s diameter, in which rapid mechanical deformation of the cell induces membrane disruptions, allowing for the passive diffusion of nanomaterial(s) into the cell cytosol.25 Via cell squeezing, diverse nanomaterials were delivered to different cell types in a high-throughput manner;26 however, clogging is an inevitable issue, and due to the heterogeneity in cell sizes, inconsistent delivery is another critical drawback.27 Additionally, large molecules such as plasmid DNAs cannot be delivered without the aid of an electric field,28 adding system complexity and undesirable cell damage. Here, we present a novel nanomaterial delivery platform termed inertial microfluidic cell hydroporator (iMCH) as a potential next generation intracellular delivery method. As shown in Figure 1a, the platform is composed of three sequential steps: (1) cell and target nanomaterial injection, in which the cell suspension mixed with the desired delivery nanomaterial(s) is injected into a standard polydimethylsiloxane (PDMS)-glass channel using a syringe pump; (2) inertial cell alignment, in which fluid inertia guides randomly aligned incoming cells into position in the middle of the channel; and (3) cell-wall collision and fluid-shear induced membrane disruption, in which a stream of cells collides with a channel wall with a sharp tip (Movie S1), producing transient nanopores that facilitate the transport of effective exogenous materials into the cell cytoplasm. Through iMCH, we have successfully delivered a wide range of nanoscopic cargos (e.g., protein, siRNA, CRISPR-Cas9, plasmid DNA, and DNA nanomaterials) to different cell types with high efficiency while maintaining high cell viability. The presented device is designed to operate at moderate Reynolds numbers (Re ≈ O(100−102)) to take advantages of inertial effects.29 Inertial lift forces first direct cells to laterally migrate across streamlines to the channel center line30 and cells’ high momentum (vc̅ ell ≈ 10 m s−1 at Re = 325) guides them to B
DOI: 10.1021/acs.nanolett.8b00704 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. iMCH device characterization. (a) Fluorescent (top row) and bright field (bottom row) images of MDA-MB-231 cells 18 h after delivery of 3 kDa FITC-dextran at different Reynolds numbers. Delivery efficiency of (b) 3 kDa and (c) 70 kDa FITC-dextran into MDA-MB-231 cells after 18 h as a function of flow rate. (d) Fluorescent images of MDA-MB-231 cells assessed using calcein-AM (live) and ethidium homodimer-1 (EthD-1; dead) after 18 h. (e) Cellular viability vs Reynolds numbers. All scale bars represent 100 μm. All error bars indicate standard deviations (N = 3).
conventional cationic lipofection was tested. Specifically, we used lipofectamine 2000, and as shown in Figure 3a,b, iMCH showed much-higher and robust delivery efficiency. It should be mentioned that the delivery protocol is much simpler for iMCH, and the time required for transfection is much shorter than lipofectamine (a few hours versus 1 min). Similarly, we cultured ES2 ovarian cancer cells that express high endogenous levels of ITGA1, the α-1 subunit of the integrin receptors involved in ovarian cancer cell binding to extracellular matrix proteins in the tumor microenvironment, and delivered siRNA against ITGA1. As expected, iMCH exhibited much-improved knockdown efficiency (>95%) compared to the lipofectamine (Figure 3c,d). To further assess the ability to deliver different target nanomaterials, we tried programmable RNA recognition and cleavage using CRISPR-Cas9 for genome engineering.39 We aimed to knockdown the COL11A1 gene in A2780cis cells (similar to Figure 3a) in such a way as two gRNAs were designed and delivered via iMCH to target different loci of exon 2 (referred to as COL11A1 gRNA nos. 1 and 2; see Note S2 for details). As presented in Figure 3e,f, the results showed the clear absence of COL11A1 expression compared to control cells, and for gRNA no. 2, a knockdown efficiency of over 80% was accomplished. This successful delivery of the CRISPR-Cas9 system is crucial because the majority of CRISPR-Cas9 delivery is still based on lipid-mediated transfection, implying the time- and laborintensive and costly procedures. Using iMCH, however, highly efficient but facile, low-cost, and rapid delivery can be achieved. Besides relatively small molecule deliveries, we also investigated the cellular uptake of macromolecules through iMCH. Nuclear transfection of DNA into mammalian cells still remains a
equivalent cell viability (Figure S4). After the processing of cells using iMCH, images were taken after 18 h to ensure delivery (Note S1). For quantitative characterization of the delivery, flow cytometry was used for measuring fluorescence from individual cells.34,35 We first set our baseline for endocytosis that includes autofluorescence and surface binding and thresholded the top 5%, finding live control cells that fall into this region (shaded in green in Figure 1c panel iii). The delivery efficiency was then defined as the fraction of live cells that receive the delivery material above the threshold. As can be seen in Figure 2b,c, highly efficient delivery is achieved. Our results show that the delivery efficiency increased monotonically as Reynolds numbers become larger (Figure 2b,c). At Re = 325, the MDA-MB-231 cell line exhibited high delivery efficiency while maintaining high cell viability (Figures 2d,e and S5), but based on applications, a different operational condition may be needed. Note that these results indicate that the iMCH platform can control the delivery dosage simply by changing the flow rate. This feature is particularly important because the ability to control the maximum intracellular concentration becomes crucial in reducing off-target effects.36 To expand the method’s applicability, we assessed its ability to deliver a wide range of target nanomaterials to different cell types. We used iMCH for gene silencing in ovarian cancer cell lines by delivering siRNA. We first designed the experiment to deliver siRNA against COL11A1, a disease progression associated gene linked to ovarian cancer recurrence and poor survival,37 to A2780cis cells, cisplatin-resistant ovarian cancer cells that express high endogenous levels of COL11A1.38 Using iMCH, >90% knockdown efficiency was achieved (Figure 3a). To compare these results to the carrier-mediated transfection approach, a C
DOI: 10.1021/acs.nanolett.8b00704 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 3. Cell transfection. (a, b) COL11A1 gene knockdown in A2780cis and (c, d) ITGA1 gene knockdown in ES2 cell line using Lipofectamine and iMCH. (e, f) CRISPR-Cas9 delivery into A2780cis cell line using iMCH for COL11A1 gene knockdown. (g, h) Plasmid DNA delivery into HEK293 cell line via iMCH and consequential GFP expression after 24 h. The scale bar represents 100 μm. Triple asterisks indicate a P value of less than 0.001 when compared to the control. All error bars indicate standard deviations (N = 3).
an interested cell line, a large number of such cells need to be transformed for the downstream analysis before animal tests. Instead of using ineffective and time-consuming methods such as passive DNA uptake,41,42 lipofection,42,43 or viral carriers,44 here, we delivered large DNA nanostructures using iMCH. As shown in Figure 4, we have successfully delivered large DNA nanostructures robustly (∼40−50% efficiency) and rapidly (
