Engineering Efficient CAR‐T Cells via Electroactive Nanoinjection

Chimeric antigen receptor (CAR)‐T cell therapy has emerged as a promising cell‐based immunotherapy approach for treating blood disorders and cancers, but genetically engineering CAR‐T cells is challenging due to primary T cells’ sensitivity to conventional gene delivery approaches. The current viral‐based method can typically involve significant operating costs and biosafety hurdles, while bulk electroporation (BEP) can lead to poor cell viability and functionality. Here, a non‐viral electroactive nanoinjection (ENI) platform is developed to efficiently negotiate the plasma membrane of primary human T cells via vertically configured electroactive nanotubes, enabling efficient delivery (68.7%) and expression (43.3%) of CAR genes in the T cells, with minimal cellular perturbation (>90% cell viability). Compared to conventional BEP, the ENI platform achieves an almost threefold higher CAR transfection efficiency, indicated by the significantly higher reporter GFP expression (43.3% compared to 16.3%). By co‐culturing with target lymphoma Raji cells, the ENI‐transfected CAR‐T cells’ ability to effectively suppress lymphoma cell growth (86.9% cytotoxicity) is proved. Taken together, the results demonstrate the platform's remarkable capacity to generate functional and effective anti‐lymphoma CAR‐T cells. Given the growing potential of cell‐based immunotherapies, such a platform holds great promise for ex vivo cell engineering, especially in CAR‐T cell therapy.


Introduction
Delivery of advanced functional biological effectors, such as nucleic acids, is a critical step in developing the next-generation of cell-based therapies. [1,2]One such therapy is based on engineering primary T cells to express chimeric antigen receptors (CARs), which has emerged as a promising approach for cancer immunotherapy, particularly in the treatment of blood-associated cancers. [3,4]In this therapy, the induced expression of CARs guides the T cells to bind and eradicate the targeted cancer cells. [5,6][12] One of the most common and well-established CAR-T cell manufacturing approaches is viral transduction, which offers high delivery efficiency and the ability to be integrated into the host genome, leading to permanent transgene expression. [13,14][17] An alternative route to viral transduction is conventional bulk electroporation (BEP) that provides a faster, simpler, and more versatile approach, enabling the delivery of various molecular cargos into different cell types via electrical stimulation. [18,19]Yet this approach has several limitations, including lack of fine control over membrane disruption and spatial resolution, inconsistent delivery outcomes, and substantial impact on cell recovery and functionality due to the use of high-voltage electric pulses (typically hundreds of volts) and the subsequent Joule heating. [1,10,20,21][24][25] These include high-aspect-ratio vertically configured nanostructure arrays-nanowires, [26][27][28][29][30][31][32][33][34] nanostraw, [35][36][37][38][39][40][41][42][43][44][45][46][47] and nanotubes [48][49][50][51] (NWs, NSs, and NTs).][54] This physical process is referred to as nanoinjection.When coupled with electrical stimulation (i.e., nanoscale electroporation or nanoscale EP), these electroactive high-aspect-ratio nanostructures can induce on-demand transient "holes" in the cell membrane for a more rapid and direct intracellular access compared to their non-electroactive counterpart (i.e., reliance on mechanical forces). [9,55]Such nanoscale EP technologies offer superior performance over conventional BEP by inducing a highly localized and uniform electric field at the cell-nanostructure interface; this dramatically lowers the applied voltage threshold required for efficient EP, from hundreds to only tens of volts. [37,56,57]But most reported nanoscale EP platforms have two inherent drawbacks: first, they lack precise control over nanostructure topological parameters such as diameter, position, and spacing, which can lead to inconsistent delivery outcomes; second, they require an integrated microfluidic reservoirs to deliver cargos into the cells during the nanoscale EP, complicating the platform's fabrication and operation. [56]Additionally, despite their recent success in the delivery of different bioactive cargos into a diverse range of cell types (predominantly adherent cell lines) with high efficiency, their progress has been slow in processing primary human T lymphocytesone of the most significant challenges in intracellular delivery.
Here, we introduce a non-viral, low-voltage (10 V), scalable, and reusable electroactive nanoinjection (ENI) platform that allows for targeted delivery of anti-CD19 CAR construct-encoded by a PCR expression cassette-into primary human T cells, with minimal impact on cell viability (>90%).The generated CAR-T cells show effective anti-cancer efficacy by suppressing the growth of target lymphoma Raji cells in vitro.The proposed ENI platform employs vertically aligned conductive NTs to a generate highly localized and uniform electric field at the NT-cell membrane interface, resulting in the formation of transient pores in the membrane at the point of contact between the membrane and each conductive NT; this allows for an influx of cargos from the NTs into the intracellular environment.Critically, the electric field that is emanating from the NTs dramatically lowers the required applied voltage, from hundreds of volts (for BEP) to only 10 volts, allowing for a considerably gentler approach to induce membrane poration.When interfacing the primary human T cells on conductive NT arrays for nanoinjection of CAR genes, the ENI platform enabled efficient delivery (68.7%, indicated by a fluorescence Cy5 tag) and expression (43.3%) of CAR genes in the T cells.Slightly higher delivery efficiency (72.1%) can be obtained by conventional BEP approach, but the CAR gene expression efficiency in T cells is nearly threefold lower (only 16.3%) than that transfected by the ENI platform.Notably, the ENI platform required 1000-fold lower CAR construct quantity compared to the BEP approach, which can be attributed to the ability to exploit the NTs' central cavity as a reservoir for loading the targeted cargos.Crucially, when co-culturing with lymphoma Raji cells (Target, or T) that express CD19 surface receptor, we demonstrate that the ENI-engendered anti-CD19 CAR-T cells (Effector, or E) exhibit a strong cancer suppression in vitro, indicated by the increase (by 20-fold) of detected E:T ratio from the original ratio (8:1).This anti-lymphoma capacity was further confirmed through a luciferase assay, which showed that the ENI-transfected CAR-T cells possess the highest cytotoxicity (86.9%) on Raji cells, leading to the most prominent reduction in total luciferase activity.We also observed that the use of such a low applied voltage (10 V) through ENI platform not only preserves gene expression but also enhances cellular proliferation, which are two crucial attributes in cell engineering.The outcomes strongly suggest that our engineered ENI platform is a promising non-viral, efficient, reusable, and safe tool for producing functional CAR-T cells that exhibit strong anti-lymphoma effects.

ENI Platform Fabrication and Operation
At the core of the ENI platform is the Au-coated vertically configured NT arrays, distributed evenly within a 3 mm × 3 mm region (ENI chip; Figure 1a).These highly ordered NT arrays were fabricated by means of a top-down approach, allowing for precise engineering of the NT geometry (Scheme S1, Supporting Information).We designed and fabricated the conductive NTs with pre-defined geometry: 2 μm height, 3 μm pitch, and 300/500 nm inner/outer diameter (Figure 1bi,ii).A crosssectional profile of the NT (Figure 1biii) was obtained using scanning electron microscopy (SEM) imaging after focusedion beam (FIB) milling, confirming the presence of a central inner cavity (0.12 μm 3 ) within each NT; this cavity allows for the direct cargo loading through the 300 nm diameter opening.
The operation of the ENI platform consisted of three main stages (Figure 1c).In stage 1, the NTs were loaded with the targeted cargos.The T cells were then cultured onto the NTs (Figure S1, Supporting Information).In stage 2, the ENI chips containing the cells were placed in a biocompatible holder (Figure S2, Supporting Information).Next, a train of low-voltage square pulses (10 V; 20 Hz; 400 μs; 600 cycles) was applied to the system, leading to the formation of transient pores on the plasma membrane at the NT-cell interface, and the subsequent influx of cargos from the NTs into the cells.In stage 3, before removing the cells from the platform, the cells were rested for 30 min on the NTs at 37 °C and 5% CO 2 , allowing sufficient time for cell recovery.The complete ENI platform's operation is illustrated in Scheme S2 (Supporting Information).
To enhance the attachment of the non-adherent T cells, the Au-coated NTs were coated with poly(d-lysine) (PDL).To further promote cell attachment, an external force was applied via centrifugation to force the cells onto the NTs.By means of SEM imaging, we observed that most cells were interfacing with the NTs after only 30 min post-centrifugation (Figure 1di).The examination of the cross-sectional profile of the NT-cell interface, obtained via FIB-SEM imaging, revealed the cell membrane to be closely wrapping around the NTs (Figure 1dii).Such conformal membrane wrapping is crucial for a successful nanoscale EP, as tightly coupled and strengthened cellular interfacing can increase the efficacy of localized EP and the subsequent pore formation on the membrane for influx of exogenous cargos into cells from the NTs. [40]Additionally, the proximity between the NTs and the nucleus can increase the likelihood of direct transportation of the CAR construct into the nucleus, where it can be transcribed readily for gene expression. [50]

Generation of CAR PCR Expression Cassette and Its Loading onto the ENI Platform
We generated an exonuclease-resistant PCR expression cassette encoding for an anti-CD19 CAR sequence and a GFP tag by in vitro enzymatic reaction using the pVAX1PBat CAR19h28TM41BBz 2A GFP plasmid as the template and primers with phosphorthioate-modified bonds on the 5′ end (Figure 1e).Such PCR-amplified products are efficient and safe for transfection due to their smaller size, containing only the necessary transcription units such as the EF1 promoter and CAR gene.This reduces the potential of bacterial gene contamination, which can jeopardize clinical applications. [50]o visualize and validate the loading of CAR-expressing PCR cassette within the NTs' inner cavity, we first tagged the CAR constructs with a Cy5 fluorescent tag.Confocal microscopy was used to image the NTs after drop-casting of the Cy5-tagged CAR constructs onto the pattern area of the ENI chip.As shown in Figure 1f, the images showed bright fluorescent spots at each NT location, indicating successful loading of CAR constructs inside the NTs' central cavity.To verify that the signal originated from the true cargo loading, we induced an intentional quenching step by increasing the laser intensity to its maximum level in a defined small region within the sample, which caused photobleaching of the fluorescently tagged CAR constructs within the small region (Figure S3a,b, Supporting Information: before and after quenching, respectively).The capacity to directly load the targeted cargos into the NTs by exploiting their inner central cavity as reservoirs (0.12 μm 3 ) eliminates the need for microfluidic channel integration during the nanoscale EP.
The impact of fluorescence modification on the PCR fragment was investigated by running gel electrophoresis for both tagged and untagged PCR fragments.After imaging the gel, we did not observe any changes in physical characteristics of the Cy5-tagged PCR fragment: the size was identical to the untagged control on electrophoresis gel (Figure S4, Supporting Information).

Influence of ENI Platform on Cell Viability
After performing finite element simulations to map the electric field localization across the ENI platform during nanoscale EP, it was calculated that the generated electric field strength at the tip of the NTs is equivalent to ≈2.5 kV cm −1 for 10 V pulses (Figure S5, Supporting Information).This value is sufficient to surpass the electrical potential of the cell membrane and allow formation of transient/reversible pores at the cell membrane in mammalian cells for cargo uptake. [43,58]A detailed description of the theoretical simulations of electric field across the ENI platform is presented in the Supporting Information (page 14).
To study the effects of the nanoscale EP process using the ENI platform on primary human T cells, we first cultured the cells onto the NT arrays and then applied a train of monophasic square-wave pulses to the ENI chip (ENI-10 V; 20 Hz; 400 μs; 600 cycles).The cells cultured onto the NTs but without an applied voltage served as the non-electroporated cells (ENI-0 V).The cell viability post nanoscale EP was assessed using a live/dead staining with Hoechst 33342 (Hoechst), fluorescein diacetate (FDA), and propidium iodide (PI)-indicating the nucleus, live, and dead cells, respectively.The confocal images and their quantification (Figure 2a,b, respectively) showed a cell viability of 90.9% at 10 V pulses, only slightly lower than that obtained for ENI-0 V (95.4%); this indicated that the ENI platform causes minimal cell damage during the nanoscale EP, allowing most cells to recover post nanoscale EP.Remarkably, exposing the NTs to the chosen electric pulses did not cause any damage to the structural integrity and the conductive coating, making the platform reusable, potentially reducing the cost of operation (Figure S6, Supporting Information).

ENI-Mediated Nanoinjection of CAR Construct into Primary Human T Cells
While several studies have shown excellent performance in delivering various cargos into non-adherent human cell lines (e.g., Jurkat) and primary human immune cells (T cells and B cells) via vertically configured nanostructures, [36][37][38]47] to the best of our knowledge, delivery of CAR genes into primary human T cells via vertically configured nanostructures (nanoinjection) has not been reported. Her, we use the ENI platform to deliver CAR gene into primary human T cells and generate functional CAR-T cells that can suppress lymphoma cell growth.
First, to demonstrate the ENI platform's capability to electroactively inject impermeable payloads into the pre-activated primary human T cells, we chose to deliver PI-a membraneimpermeable fluorescent dye. [59]PI was used in Figure 2a to demonstrate lack of cytotoxicity of the ENI approach.But it can also be loaded in the NTs as cargos and delivered into cells during the ENI process.The analysis of ENI-treated cells showed successful insertion of PI into the T cells with 95.4% efficiency (Figure S7, Supporting Information), and negligible impact on cell viability (>90%).In comparison, the ENI-0 V cells (i.e., cells interfacing with PI-loaded NTs without the applied voltage) showed only 4.4% PI + cells, which confirms that the applied low voltage effectively induces transient pores at the NT-membrane interface, facilitating the rapid uptake of cargos into the intracellular environment.
Next, as our main goal, we used the ENI platform as a nonviral platform to create functional CAR-T cells by delivering the Cy5-tagged CAR constructs into the primary human T cells.The detection of Cy5 will indicate successful delivery of CAR constructs into T cells, while transfection of the CAR gene can be assessed based on the reporter GFP expression. [60]To benchmark the performance of the ENI platform against an existing nonviral method, we used the standard BEP via a commercially available electroporator, using the optimized parameters for delivery of Cy5-tagged CAR constructs into the primary human T cells (Figure S8, Supporting Information).
After 24 h, the cells were analyzed via flow cytometry to quantify the percentage of Cy5 + and GFP + cells within each group.The schematic of the workflow for the entire flow cytometry selection process is presented in Figure S9 (Supporting Information), providing an overview of the gating and compensation strategies.The quantification of flow cytometry showed a significantly higher percentage of Cy5 + population and therefore a higher CAR gene delivery in the ENI-10 V transfected cells (68.7%) compared to the ENI-0 V transfected group (34.3%; Figure 2c; Figure S10a,b, Supporting Information).Cargo uptake by ENI-0 V transfected cells could be explained by passive diffusion of cargos through enhanced endocytic pathways (caveolae-mediated), caused by the tight cell membrane-NT interfacing and the subsequent membrane deformation. [49,50]Expectedly, due to a higher level of CAR gene insertion, the ENI-10 V transfected cells exhibited a significantly higher CAR expression (comprising of gene transcription and protein translation) compared to the ENI-0 V transfected group; this was indicated by the twofold higher percentage of GFP + cells within the entire population (43.3% via ENI-10 V compared to 21.3% via ENI-0 V; Figure 2d).This outcome further confirms that an applied induces transient pores at the NT-membrane interface, facilitating the rapid uptake of cargos into the intracellular environment from the NTs.Similarly, these results demonstrate an improvement over our previous published study, which relied on mechanical cues (i.e., non-electroactive silicon nanotubes, SiNTs) to deliver the CAR genes into primary mouse T cells.Though direct comparison is difficult given the different cell source, but we showed that the use of nanoscale EP via our new ENI platform achieves a significant enhancement over the SiNT method, indicated by greater than twofold increase in CAR gene delivery (68.7% via ENI compared to 20-37% via SiNTs) and twofold increase in CAR transfection and GFP expression (43.3% via ENI compared to 18-24% via SiNTs). [50]nexpectedly, despite showing a comparable percentage of Cy5 + population (72.1%) to that obtained by ENI-10 V, the BEPtransfected cells exhibited a significantly lower GFP + population (16.3%) than cells treated by the ENI-10 V platform (43.3%; Figure 2d; Figure S10a,c, Supporting Information).This corresponds to an almost threefold higher CAR expression efficiency in T cells transfected by the ENI-10 V than that by the BEP.One contributing factor for this lower gene expression with BEP is the inability of the delivered Cy5-tagged CAR constructs to travel into the nucleus of the cell to be transcribed.As we have shown recently using primary mouse T cells, the close contact between the cells and the NTs and the resulting proximity between NTs and the nucleus, even without nanoscale EP, can augment the transportation of CAR construct into the nucleus where they can be readily transcribed for gene expression. [50]Another contributing factor is high level cellular stress caused during BEP, which can lead to retarded cell recovery and function. [20,61]In comparison, the ENI platform is a much gentler approach, using 20-fold lower voltage value; this can significantly minimize the impact on the gene expression of treated cells, as well as cellular stress. [44]It is also important to note that in addition to achieving a significantly higher gene expression efficiency compared to BEP, the ENI platform (10 V) required 1000-fold lower CAR construct quantity due to the ability to exploit the NTs' central cavity as reservoir for loading the targeted cargos.
Confocal imaging further confirmed both Cy5 and GFP signals in ENI-10 V-treated T cells (Figure 2e; Figure S11, Supporting Information), suggesting successful delivery and expression of the CAR construct in the primary human T cells.

Influence of ENI Platform on T Cell Proliferation
We also explored the influence of ENI-mediated transfection on the proliferation of primary human T cells.This is particularly important for generating CAR-T cells for therapeutic purposes, ensuring that the employed gene delivery approach allows for efficient expansion of transfected T cells. [62]For this investigation, CellTrace Violet (CTV) was used as the proliferation tracking dye. [63]The CTV dye is designed to covalently bind to intracellular proteins in the cytoplasm and nucleus.After each round of cell division, the amount of CTV is halved, resulting in a decrease in fluorescence intensity over time that can be used to quantify cell proliferation. [50]n Day 0 (D0), the T cells were stained with CTV followed by treatment with the ENI platform (0 V and 10 V) or BEP.The non-treated T cells that were CTV-stained served as the control.The T cells from each group were then transferred to an anti-CD3/anti-CD28 coated well-plate and incubated at 37 °C and 5% CO 2 to allow for expansion.For the analysis, CTV intensity was measured daily by means of flow cytometry for 4 consecutive days (0-96 h; Figure S12, Supporting Information).
Statistical quantification of CTV intensity of T cells showed that the overall proliferation in all the treatment groups was progressively higher than that in the control non-treated group by the end of the 96-h period (Figure S12a, Supporting Information).After normalization to the control, we observed that the BEP group had a progressively slower decrease in CTV intensity compared with ENI-treated cells (0 and 10 V), indicating a lower proliferation rate in BEP-treated cells (Figure S12b, Supporting Information).On Day 4, T cells from the BEP group showed a lower proliferation index (normalized to the control) compared with that from ENI-transfected groups (Figure S12c, Supporting Information).This lower proliferation rate in BEP group could be caused by the higher cellular stress, which can lead to retarded cell recovery. [62,64]

Lymphoma Cell Suppression by ENI-Generated CAR-T Cells In Vitro
Inspired by the successful transfection of the CAR gene into primary human T cells, we explored the anti-cancer efficacy of these CAR-T cells in vitro.Due to the anti-CD19 specificity of the CAR construct, we selected CD19-expressing Raji cells (a Burkitt's lymphoma cell line) as the target. [65]To examine lymphoma suppression by CAR-T cells generated by the ENI platform (0 V and 10 V) and BEP, we co-cultured the transfected T cells (containing CAR-T cells; Effector, E) with Raji cells (Target, T) at an E:T ratio of 8:1 (Figure 3a).Non-transfected T cells cocultured with Raji cells served as control (Ctrl).The cells treated with ENI-10 V but without any cargo loading (ENI-10 V (No cargo)) was also considered as control to account for the influence of ENI-10 V treatment on T cells' ability for lymphoma suppression.
By means of flow cytometry, the changes in E:T ratios after 0 (immediately after co-culture), 4, 24, 48, and 72 h of co-culture were measured.The schematic of the workflow for the entire flow cytometry selection process is presented in Figure S13 (Supporting Information), providing an overview for the gating strategy used to distinguish between T cells (E) and Raji cells (T) for the quantification of the E:T ratios.By quantifying the flow cytometry results (Figure 3b), it is evident that the ENI-10 V group demonstrated progressively greater E:T ratio changes compared with the other groups over the 72 h period.In particular, the E:T ratio was detected as 159.4 at 72 h for the ENI-10 V group (Figure 3c), corresponding to a 20-fold increase from the original one (8:1); this is significantly higher than the other groups, suggesting that the CAR-T cells engineered via ENI-10 V performed the strongest suppression on lymphoma Raji cell growth over the 72 h period.In contrast, the CAR-T cells obtained via BEP were less successful in suppressing Raji cell growth, indicated by the lowest E:T ratio increase (by 4.4-fold) from the starting ratio; this change in E:T ratio was even lower than that of the nontransfected control group (Ctrl).The poor performance of the BEP-generated CAR-T cells could be due to the high level of cellular stress on the T cells during the BEP transfection process, leading to retarded cell recovery and impaired CAR expression.
In parallel, the anti-lymphoma efficacy of the generated CAR-T cells was evaluated by quantifying the elimination of target Raji cells by CAR-T cells in the co-culture (8:1) via a luciferase assay after 0 (immediately after co-culture), 4, 24, 48, and 72 h incubation.The target CD19-expressing Raji cells were geneti-cally modified to constitutively express a Firefly luciferase (Fluc)-GFP reporter.Hence, examining the total luciferase activity via the luciferase assay could be used to determine the elimination of Raji cells by the engineered CAR-T cells.Raji cells cultured on their own with the same seeding density served as the luminescent positive control (Pos -Raji).The non-transfected T cells co-cultured with Raji cells served as the background control (Ctrl), measuring the background killing ability of the T cells.
As shown in Figure 3d, the luminescence measurements over the 72 h period showed that the CAR-T cells produced by ENI-10 V progressively showed the lowest luminescence, compared to the other co-culture groups.After 72 h (Figure 3e), the CAR-T cells generated by ENI-10 V exhibited a significant drop in Raji cells' total luciferase activity (4242 relative light units, RLU) compared to the Raji cells' total luciferase activity in the positive control group (32 484 RLU); this corresponds to a cytotoxicity of 86.9% toward Raji cells.In comparison, the luciferase activity of Raji cells in the other three co-culture groups after 72 h showed a higher level to that observed in ENI-10 V group, with luminescence measured as 8786 RLU in control (Ctrl), 8905 RLU in ENI-0 V, and 8791 RLU in BEP groups, suggesting a similar cytotoxicity on target Raji cells (72.9%) across these three groups.This observation agreed with our E:T ratio measurements, suggesting that CAR-T cells generated by the ENI-10 V demonstrated the strongest inhibition of the target lymphoma Raji cells (i.e., highest anti-lymphoma efficacy).We were surprised to see the control group (containing T cells not transfected with CAR) exhibiting comparable luciferase activity to that produced by BEP and ENI (0 V) groups, as indicated by their similar assay luminescence.This high level of background killing suggests that the non-transfected T cells naturally possess the ability to recognize and attack any cells that they perceive as foreign or abnormal (i.e., CD19-expressing Raji cells); yet this background killing can vary donor-to-donor-one of the major challenges when investigating primary immune cells. [66]Nevertheless, when considering both changes in E:T ratio and the luciferase assay, we can infer with confidence that the CAR-T cells produced by the ENI-10 V showed the highest anti-lymphoma efficacy among all the experimental groups, indicating the superiority of the ENI platform to produce functional and effective anti-CD19 CAR-T cells.
Encouraged by the successful generation of functional anti-CD19 CAR-T cells with high level of anti-lymphoma efficacy using the ENI-10 V setup, we further investigated the interaction between ENI-engineered CAR-T cells and CD19-expressing Raji cells in real time.To distinguish between the Raji cells and CAR-T cells, we stained the Raji cells with CTV dye and T cells with anti-CD3 conjugated with APC-Cy7 before the co-culturing.The low magnification view time-lapse recording (Video S1, Supporting Information), obtained via confocal microscopy, showed that the ENI-engineered CAR-T cells were able to rapidly recognize and bind to the target Raji cells, forming clusters around these target Raji cells within 1 h of co-culturing.Such clustering effect confirms the ability of ENI-generated (10 V) CAR-T cells to rapidly recognize and bind to the specific antigen (i.e., CD19) on the target cells.This rapid target recognition, combined with the potential cytokine secretion to enhance the recruitment of T cells, can lead to the ENI-generated CAR-T cells' overall effective suppression of the target lymphoma cells. [67]Predictably, the nontransfected control T cells showed a significantly lower tendency to target and form clusters around the targeted Raji cells (Video S2, Supporting Information).The real-time imaging of the coculturing further corroborates the findings from flow cytometry and luciferase assay.In a higher magnification time-lapse recording (Video S3, Supporting Information) and confocal images obtained at the beginning and then end of the recording (1 h 30 min apart; Figure S14, Supporting Information), it is clear that the clusters of CAR-T cells around the Raji cells induce the apoptosis of Raji cells, as evidenced by cell shrinkage, chromatin condensation, and fragmentation. [68]

Conclusion
We have demonstrated that the ENI platform can generate functional primary human CAR-T cells, giving them the ability to suppress lymphoma cell growth in vitro.We showed that the ENI platform enabled efficient delivery (68.7%) and expression (43.3%) of CAR genes in the primary human T cells, with minimal cellular perturbation (>90% cell viability).This represents a significant improvement over our previously reported nanoinjection into mouse T cells using non-electroactive SiNTs, with greater than twofold increase in CAR gene delivery and twofold higher transfection efficiency.The use of primary human T cells in our current work allows for more relevant insights for clinical translation.When compared to BEP, the ENI platform achieved an almost threefold higher CAR transfection efficiency, indicated by the significantly higher expression of reporter GFP (43.3% in ENI compared to 16.3% in BEP), despite their highly comparable delivery efficiency (68.7% in ENI and 72.1% in BEP).This can be attributed to the ENI platform's rather gentle treatment of cells (low applied-voltage, 10 V), allowing for preservation of gene expression and cellular proliferation-two important aspects in cell engineering.Notably, the ENI platform (10 V) required 1000-fold lower CAR construct quantity compared to the BEP approach, which can be attributed to the ability to exploit the NT's central cavity as reservoir for loading the targeted cargos.By co-culturing with lymphoma Raji cells (Target, or T) that express CD19 surface receptor, we demonstrated that the ENI-generated anti-CD19 CAR-T cells (Effector, or E) exhibit a strong cancer suppression in vitro, indicated by the increase (by 20-fold) of detected E:T ratio from the original ratio (8:1).This was further confirmed through a luciferase assay, which showed that the ENI-transfected CAR-T cells possess higher cytotoxicity (86.9%) toward Raji cells compared to BEP-transfected cells, leading to a more prominent reduction in total luciferase activity.We anticipate that the ENI platform will offer an exciting new path to more versatile cellular delivery, enabling a paradigm shift at the interface of nanotechnology and immuno-oncology, with strong potential of offering huge benefits for patients with the many blood cancers in which CAR-T cells are involved.

Figure 1 .
Figure 1.The ENI platform and its operation.a) The ENI chip, consisting of a 3 mm × 3 mm pattern area at the centre.b) SEM images showing the Au-coated NTs at tilted (45°) zoom-out (i), top (ii), and cross-sectional (resin embedded (iii)) views.c) The steps of the ENI platform's operation: The Au-coated NTs are loaded with CAR construct and primary human T cells are seeded onto the loaded NTs with centrifugation applied (stage 1); localized on-demand nanoscale EP performed by applying a series of low voltage electric pulses and the subsequent intracellular delivery (stage 2); and post nanoscale EP membrane recovery (stage 3).d) False-colored SEM images showing the interfacial interactions between primary human T cells and Au-coated NTs arrays (i) and the cross-sectional profile of NT-membrane interface (ii).e) Schematic of the PCR expression cassette of the CAR construct, encoding anti-CD19-CAR and a reporter GFP gene.f) Confocal microscopy 3D view image of the loading of Cy5-tagged CAR constructs inside the Au-coated NTs.

Figure 2 .
Figure 2. The influence of applied voltage through ENI platform on viability of primary human T cells post nanoscale EP and nanoinjection of the Cy5-tagged CAR construct into these T cells.a) Confocal microscopy images of T cells before and after nanoscale EP are compared using 0 V (nonelectroporated) and 10 V. b) Quantification of the percentage of live cells post nanoscale EP.The error bars indicate ± stadrard deviations (SDs), n = 3, unpaired t-test.c) Quantification of delivery efficiency as the percentages of Cy5 + populations within T cells transfected via ENI platform (0 and 10 V) and BEP at 24 h, representing delivery efficiency of the CAR construct.The error bars indicate ± SDs, n = 3, ****P < 0.0001, one-way analysis of variance (ANOVA).d) Quantification of the gene expression efficiency within T cells delivered with the Cy5-tagged CAR construct using the ENI platform (0 and 10 V) and BEP.Error bars indicate ± SDs, n = 3, ***P = 0.0001, ****P < 0.0001, one-way ANOVA.e) Confocal microscopy images demonstrating the insertion of the Cy5-tagged CAR construct into the T cells (magenta) and the subsequent GFP expression (green) using the ENI platform (10 V).The cells were fixed and stained with Hoechst (blue).