Multi-Checkpoint Cellular Targeting of siRNAs Using Antibody-Directed Lipid-Polymer Hybrid Nanoparticles

Abstract

Despite advancements in tissue-specific gene therapy, current technologies struggle to target organs beyond the liver, spleen, and lungs. Passive approaches like selective organ targeting (SORT) lipids show potential but require time-intensive optimization. Active targeting, exemplified by antibody-drug conjugates (ADCs), offers a modular and effective alternative. Building on this, we developed antibody-functionalized hybrid lipid-polymer nanoparticles for siRNA delivery, targeting cancer-overexpressed receptors such as EGFR and Trop2, prevalent in aggressive cancers like triple-negative breast cancer (TNBC). In addition, to enhance therapeutic safety and efficacy, we integrated multi-siRNA delivery into a multi-checkpoint targeting strategy, minimizing reliance on single antigens and reducing off-target risks. Using TNBC cells as a model, this platform demonstrates potential for developing a robust and safe therapeutic approach. In this manuscript, we present our findings which lay the foundation for developing a multi checkpoint strategy to enhance target selectivity in nanomedicine.

Graphical Abstract

1. Introduction

Gene therapy has the potential to directly suppress the expression of diseased proteins, produce therapeutic proteins, or even edit DNA to correct genetic abnormalities, offering long-term solutions for many diseases.^1,2 The field gained significant momentum with the approval of Onpattro in 2018 and the development of COVID-19 vaccine candidates in 2021.^3–6 Despite these advances, non-viral methods for genetic medicine transport has predominantly targeted the liver.^7–10 This preference arises from liver’s large size and its tendency to accumulate nanoparticles there, facilitated by the formation of a “protein corona” that enhances passive localization in hepatic cells.^7,11–14 However, recent research has begun to explore targeting other organs such as the spleen, lungs, and heart.^15,16 The methodologies for achieving this remain largely empirical and unpredictable, often relying on trial and error to design effective targeting mediators. To improve targeting specificity, active targeting of therapeutics has emerged as a promising approach.^17–23 This strategy involves incorporating tissue-specific targeting abilities into nanoparticles, using motifs that can specifically bind to surface antigens on target tissues and cells. By utilizing antibodies, ligands, aptamers, and other targeting motifs, these interactions offer a customizable approach to fit various targets.^17–19,23

We are especially inspired by the recent clinical success of antibody-drug conjugates (ADCs) in the targeted delivery of chemotherapeutic small molecules.^24–26 However, the ADC-style strategy is not conveniently adaptable for delivering therapeutic proteins and nucleic acids, due to the susceptibility of such conjugates to degradation, RES (reticuloendothelial system) clearance, and potential immune responses. These shortcomings can be mitigated using a nanoparticle entrapment approach, encapsulating the biologic while covalently attaching the targeting antibody or peptide to its surface.^27–30 Such targeting strategies are relatively rare, and the proposed nanoparticles offer convenient access to these formats. In this context, we leverage our work on developing antibody nanogel conjugates and antibody polymer conjugates.^{26–28,30,31} This study focuses on developing methods to engineer hybrid lipid-polymer nanoparticles (hLPNs) conjugated with antibodies, allowing for antigen-specific targeting using a flexible modular scaffold. It is noteworthy that the hLPNs were utilized here owing to their fragmented design allowing for straight forward modifications on the lipid surface without sacrificing the nanoparticle integrity significantly.^32–37

In parallel, we aimed to assess the synergistic effects of combining distinct siRNAs within a multi-checkpoint targeting system. This approach is designed to enhance the therapeutic safety and efficacy of siRNA combinations in complex biological environments, addressing the potential limitations of relying solely on the overexpression of a single antigen in target tissues. By employing this strategy, we aim to reduce the risk of severe off-target effects that may arise in highly potent combination therapies. To investigate the described properties and assess their potential therapeutic effects, we utilize triple-negative breast cancer (TNBC) cells as a model system. TNBC was selected due to its complex biology and the significant clinical challenges associated with its treatment,³⁸ making it a relevant and critical focus for study. This proof-of-concept study is outlined in three parts: (i) optimization of the nanoparticle system for antibody-mediated active targeting in TNBC cell lines; (ii) demonstration of selective transfection in the target cells; (iii) exploration of a multi-checkpoint-based targeting strategy using an AND-gate approach with a dual-particle dual-targeting strategy to minimize off-target effects (Figure 1).

Figure 1.

Three major steps of developing a multi-checkpoint based active targeted nanoparticle system for siRNA delivery to select cells based on their antigen expression. 1) Antibody conjugation onto nanoparticle surface 2) Attaining selective and functional delivery of siRNA cargo to antigen presenting cells 3) Utilization of dual antigen as checkpoint to bias the flow of lethal siRNA combinations into tissue/ cells concurrently presenting both target antigens. (Created with BioRender.com)

2. Results and discussions:

Formulation of G0-C14/PLGA based hybrid lipid-polymer nanoparticles (hLPNs) for mutant P53 knockdown in vitro.

First, an initial set of nanoparticles was formulated with negative control scrambled siRNA using a modified nanoprecipitation process (see Materials and Methods section for details). Briefly, the siRNA was mixed with an ionizable lipidoid, G0-C14 (structure in Figure 2a) in acetone, followed by addition of a requisite amount PLGA solution (in acetone).^35,39–42 Next, this concoction was gradually added to an aqueous solution of lipids (in RNAse free DI water pH ~ 5.5) to encase the hydrophobic nanoparticle core (Figure 2a, b). To confirm the successful formulation, dynamic light scattering (DLS) was used to assess the size of the nanoparticles, which showed a size of ~150 nm and a good correlation function with a dispersity of 0.07 (Figure 2c). This size and the morphology of the nanoparticles were further evaluated with transmission electron microscopy (TEM), which revealed an average size of 130 nm −150 nm with uniform spherical shape and a core-shell type structure (Figure 2d).

Figure 2.

Initial formulation of the hybrid lipid-polymer nanoparticles (hLPNs). a) The cationic G0-C14 and hydrophobic PLGA together stabilize the nanoparticle core through electrostatic and hydrophobic interactions. b) Schematic of formulation procedure for the hybrid nanoparticles. c) Size measurement of the nanoparticles formed with negative control siRNA with a lipid coating consisting of lecithin and DSPE-PEG₅₀₀₀ at 85:15 molar ratio d) Negative stained TEM images showing uniform distribution of the nanoparticles and the core-shell morphology (scale bar 100 nm). e) Western blot and f) Quantitative analysis of p53 knockdown with sip53 containing hLPNs in T47D and MDA-MB-468 cells. (Fig. 2b Created with BioRender.com)

To test their efficiency in transfection, we opted for a functional readout in in vitro cell cultures. In this case, we targeted p53 as the protein of interest. Therefore, siRNA against p53 protein, sip53, was encapsulated into the nanoparticles using the same process. The transfection efficiency of the synthesized particle was then tested on mutant P53 expressing breast cancer cell lines, MDA-MB-468 and T47D. Upon treatment with the nanoparticles with an effective siRNA concentration ranging from 0–200 nM, we observed a 50% knockdown of p53 at only ~25 nM of siRNA treatment and >90% reduction in p53 levels at as low as 50 nM concentration in T47D cells whereas the same 90% knockdown was observed at 200 nM in MDA-MB-468 cells (Figure 2e, f).

Optimization of inverse electron demand Diels-Alder (IEDDA) reaction mediated antibody conjugation with the hLPNs.

To evaluate and optimize the conjugation efficiency of hLPNs with antibodies, we used surface lysines on the antibody as handles to install a clickable trans-cyclooctene (TCO) linker (Figure 3a, b). The functionalized antibody was characterized using a tetrazine-conjugated fluorophore (TAMRA-PEG4-tetrazine) to quantify the average linker-to-antibody ratio (LAR). The LAR was found to be ~3 across three different antibodies (Figure S5), viz., anti-EGFR, anti-TROP2, and isotype antibody (details provided in SI table 2). Additionally, the modification pattern of the antibodies was validated using reducing SDS-PAGE, which revealed that most modifications occurred in the heavy chain, irrespective of the antibody used (SI Figure S7). To install the antibody onto the nanoparticles, we synthesized a tetrazine-terminated PEG lipid (DSPE-PEG₅₀₀₀-Tz) using a simple NHS ester conjugation chemistry using the commercially available DSPE-PEG₅₀₀₀-NH2 and tetrazine-PEG5-NHS ester. The resultant tetrazine-terminated lipid was purified and characterized (details in SI). The synthesized lipids were incorporated into the hLPNs by progressively replacing the existing PEG lipids with tetrazine-terminated lipids. This was done by substituting 0 to 15 mol% (relative to total lipids) of PEG lipids with DSPE-PEG₅₀₀-Tz (referred to as Tz-lipid in subsequent descriptions below) while keeping the remaining PEG lipids as DSPE-PEG₅₀₀₀-OMe, within an overall PEG lipid concentration of 15 mol%. This process resulted in a tetrazine-decorated surface on the hLPNs. To further validate the success of tetrazine decoration of the hLPN surface, a TCO functionalized hydrophilic fluorescent probe, BDP FL-PEG₄-TCO has been utilized to verify the presence of tetrazine on the nanoparticle surface along with the degree of tetrazine incorporation as a relative measure (Figure S6).

Figure 3.

Optimization of antibody linker modification and antibody-hLPN conjugation process a) TCO-PEG₈-NHS ester modified antibody and tetrazine functionality modified DSPE-PEG₅₀₀₀ lipid for antibody-HLPN conjugates. b) Antibody conjugation with nanoparticle surface through TCO-tetrazine click chemistry. c) SDS-PAGE of antibody conjugation with tetrazine terminated nanoparticle at antibody to tetrazine molar ratio of 1:1, 1:2, 1:5, 1:10 and 1:20, the smeared nature of the protein band is characteristic of successful conjugation. d) Quantification of free antibody in the mixture post conjugation process carried out at 4 °C for 16 h. All tests are done in replicates of n = 3, statistical significance calculated by one-way ANOVA. *p < 0.1, **p < 0.01. (Fig. 3b created with BioRender.com)

To optimize the conditions to conjugate the functionalized antibodies with the prepared nanoparticles, we used 7.5 mol% DSPE-PEG₅₀₀₀-Tz incorporated nanoparticles (Tz-NPs) (overall lipid composition lecithin: DSPE-PEG₅₀₀₀-OMe: DSPE-PEG₅₀₀₀-Tz = 85: 7.5: 7.5 mol/mol). A gradually increasing ratio of antibody to tetrazine ranging from 1:1 to 1:20 was tested for complete antibody consumption (Figure 3c). Briefly, distinct amounts of Tz-NP were incubated with fixed amounts of TCO-functionalized antibodies for 12–18 h in PBS at 4 °C followed by estimation of free antibody using SDS-PAGE, which is characterized by an upward shifted smeared band and simultaneous disappearance of characteristic antibody band, representing successful antibody conjugation with the nanoparticles. We initially investigated the conjugation process by incubating a free TCO modified antibody with the tetrazine functionalized nanoparticles at a 1:1 to 1:20 molar ratio between the antibody and DSPE-PEG₅₀₀₀-tetrazine. Through this study, a relatively insignificant conjugation of antibodies was observed at 1:1, 1:2, and 1:5 molar ratios. However, a significant increase in antibody conjugation was observed with a 1:10 antibody: DSPE-PEG₅₀₀₀-tetrazine ratio with only ~15–20% free antibody left in the mixture, and progressively >95% antibody conjugation was observed at a 1:20 ratio (Figure 3c, d). The lack of free antibodies eliminates the necessity to undergo complicated purification procedures. Therefore, the next sets of studies were carried out without further purification of the antibody hLPN conjugates, prepared with a 1:20 antibody: DSPE-PEG₅₀₀₀-tetrazine ratio.

Investigation of an ideal degree of surface decoration with tetrazine-functionalized PEG lipid.

Although we carried out our initial optimizations of the antibody conjugation process with 7.5% tetrazine-terminated lipids (Tz-7.5%), it is essential to assess the surface antibody density that maximizes antigen binding to enhance selectivity towards the specific cells. To examine the direct impact on antibody conjugation, we investigated varying degrees of Tz-lipid incorporation on the nanoparticle surface. We formulated nanoparticles with Tz-lipids ranging from 0–15 mol% of overall lipid composition by gradually replacing DSPE-PEG₅₀₀₀-OMe in the initial lipid compositions (Figure 4a). Surprisingly, we observed a gradual increase in size from ~120 nm to ~190 nm of the nanoparticles with increasing amount of Tz-lipid functionalization on the nanoparticle surface, possibly due to the hydrophobicity of the tetrazine units (Figure 4b). This size change persists even after conjugating with antibodies up to Tz-7.5%, whereas HLPNs showed a steep change in size to ~230 nm for Tz-15% after antibody conjugation, which alternatively indicates the gradual perturbation of the hydrophilic-lipophilic balance owing to increasing numbers of antibody conjugation on the HLPN surface.

Figure 4.

Optimization of surface coverage of tetrazine functionalized PEG lipids. a) Schematic view of varying degrees of antibody functionalization on nanoparticles because of variance in the incorporation of DSPE-PEG₅₀₀₀-Tz ranging from 0 % −15% relative to total mols of lipid. b) Increasing trend is the resultant size of hLPNs with increasing degree of DSPE-PEG₅₀₀₀-Tz pre and post conjugation with isotype antibody. c) Flow cytometric analysis of anti-EGFR antibody conjugated nanoparticles binding to EGFR expressing MDA-MB-468 cells were monitored in concentrations from 0– 200 nM of Alexa fluor 647 labeled negative control siRNA. d) Mean fluorescence intensity obtained from flow cytometry analysis at a 200 nM effective siRNA dose. All tests are done in replicates of n = 3, statistical significance calculated by one-way ANOVA. *p < 0.1, **p < 0.01, ***p < 0.001, ****p <0.0001. (Fig. 4a created with BioRender.com)

The interdependence of the surface antibody density with the propensity of the nanoparticles binding to the surface exposed receptor in vitro was then investigated. We chose epidermal growth factor receptor (EGFR) as the target antigen due to its wide applicability in various cancer types and direct correlation with many of the breast cancer subtypes.^43,44 We used cetuximab antibody for targeting EGFR and Alexa Fluor 647 (AF647) conjugated negative control siRNA to monitor the antigen-directed uptake in EGFR-overexpressing MDA-MB-468 cell lines. Briefly, cetuximab-conjugated hLPNs were incubated with cells for 30 mins. Subsequently, the initial treatment media was removed and replaced with fresh media, followed by an additional 3 h of incubation at 37 °C. Flow cytometry analysis showed a gradual concentration-dependent increase in fluorescence signal in all cases, except for the nanoparticles lacking antibody decoration. The slope of this increase was directly proportional to the antibody density, except at the highest density (15%) (Figure 4c, d). This may result from poorly formed hLPNs, as discussed earlier, or due to antibody overcrowding, which could hinder proper orientation for effective epitope recognition. Realizing that 7.5% modification is a ‘sweet spot’, we carried subsequent experiments with Tz-7.5% HLPNs.

Formulation and target selective delivery with anti-EGFR antibody conjugated hLPNs.

We investigated the target selective uptake and transfection with EGFR-targeting hLPNs. Prior to the assays, we characterized the antibody conjugation process and the resultant hLPNs using SDS-PAGE, dynamic light scattering (DLS), and zeta potential measurements, as outlined in Figure 5a–c. SDS-PAGE confirmed complete antibody conjugation, as indicated by the near complete disappearance of the free antibody band (Figure 5a). DLS results showed a minimal size increase from 163 nm to 166 nm with unchanged polydispersity index (Figure 5b). Zeta potential measurements further validated conjugation, revealing a shift from −7 mV to −18 mV (Figure 5c), consistent with antibody conjugation. Subsequently, we assessed the cargo encapsulation of both modified and unmodified nanoparticles and found no discernible difference (Figure 5d), underscoring the structural integrity of the hLPNs post-antibody conjugation.

Figure 5.

Characterization of anti-EGFR antibody conjugated Tz-7.5% hLPNs. a) SDS-PAGE verification of successful conjugation of anti-EGFR antibody with the nanoparticles. b) The nanoparticle does not undergo a significant alteration in size before and after antibody conjugation. c) Antibody conjugation was verified by the change in zeta potential of the nanoparticle’s characteristic of antibody decoration. d) Encapsulation efficiency of the nanoparticles remained unaltered with antibody conjugation. e) Immunocytochemistry based quantification of receptor on cell surface f) relative expression of EGFR in MDA-MB-468 and T47D cells. g) Relative uptake/binding with EGFR overexpressing MDA-MB-468 cell using anti-EGFR, isotype, and unmodified nanoparticles. Flow cytometric evaluation was carried out with Alexa Fluor 647 labeled siRNA as cargo. Relative uptake of anti-EGFR-Tz-7.5% nanoparticles in MDA-MB-468 (EGFR+) and T47D (EGFR-) cells h) using flow cytometry with siRNA concentration from 0 nM to 200 nM and i) with confocal microscopy at 200 nM siRNA concentration. j) Relative uptake of anti-EGFR-Tz-7.5% particles in MDA-MB-468 and T47D cell lines with varying degrees of serum in the media from 0–50% (v/v). (Fig. 5e created with BioRender.com)

hLPNs were then evaluated for selectivity towards EGFR-overexpressing MDA-MB-468 cells. We had previously established the dose-dependent uptake in these cells using an AF647 siRNA conjugate as the cargo (Figure 4c). Following a similar strategy, we assessed the efficiency of cellular uptake of cetuximab-conjugated hLPNs and compared it with unconjugated as well as isotype antibody-conjugated hLPNs (Figure 5g). We observed dose-dependence in the uptake of the cetuximab-conjugated hLPNs (anti-EGFR-Tz-7.5 %), whereas both the isotype control and bare hLPN showed minimal uptake over the same period. This supports the possibility of targeting specific cells using specific antibody-conjugated hLPNs for siRNA delivery.

We then evaluated the target selectivity of hLPNs in T47D cells as a control, due to their low EGFR expression (Figure 5f). Both MDA-MB-468 (EGFR+) and T47D (EGFR-) cell lines were dosed with equal amounts of AF647-siRNA containing hLPNs. While MDA-MB-468 cells exhibited significant uptake, we observed lower uptake in T47D cells, even at an siRNA dosage of 200 nM. Flow cytometry analysis further corroborated the confocal microscopy findings (Figure 5h– j).

Target selective P53 knockdown using EGFR targeted hLPNs.

Next, we utilized antibody-conjugated hLPNs to deliver therapeutically relevant siRNAs with target selectivity. We focused on p53, because the P53 gene is a crucial tumor suppressor that regulates the cell cycle and maintains cellular homeostasis but is frequently mutated in human cancers.^45–47 Many mutant p53 (mut-p53) proteins exhibit gain-of-function properties that contribute to tumorigenesis, making them a promising target. However, due to the diverse range of p53 mutations, designing small-molecule inhibitors, antibodies, or genetic alteration techniques that selectively target mutant p53 while sparing wild-type p53 remains challenging. Given this limitation, our approach does not aim to restore p53 function but rather to leverage siRNA-mediated knockdown as a potential strategy for tumor suppression. While this results in the depletion of both mutant and wild-type p53, in cancers where mutant p53 drives oncogenic activity, broad p53 knockdown may still be therapeutically beneficial by reducing tumor-promoting effects. Therefore, instead of specificity in the therapeutic modality, enhancing selectivity in the delivery method may provide a modular approach to targeting mutant p53-driven diseases.

To this end, we evaluated the transfection efficiency of these cetuximab-conjugated hLPNs toward cancer cells with varied EGFR expression levels (Figure 6a). A significantly higher potency of anti-EGFR-Tz-7.5 was observed in EGFR positive MDA-MB-468 cell line (DC50 ~ 25–50 nM) compared to T47D cells (DC50 was not obtained within the concentration range) (Figure 6b, c). With bare nanoparticles no observable loss of p53 expression was observed in both cell lines (Figure 6d, e), suggesting that the efficient uptake of cetuximab-conjugated hLPNs is due to receptor-mediated uptake. This establishes the effectiveness of the targeting technology in knocking down selective genes in target cells/tissues based on their receptor expression.

Figure 6.

EGFR selective delivery of siRNA against P53 gene in p53 mutant breast cancer cell lines a) schematic representation of target selective knockdown of the target gene in EGFR overexpressing cells b) western blot analysis of p53 expression with vinculin as housekeeping protein reference post treatment with anti-EGFR-TZ-7.5% particles encapsulating P53 siRNA c) quantified expression of p53 protein in MDA-MB-468 cells (EGFR+) vs T47D cells (EGFR-) d) western blot analysis of p53 expression with unmodified TZ-7.5% particles encapsulating P53 siRNA e) quantitative p53 protein expression in MDA-MB-468 and T47D cells post-treatment with control non targeted hLPN. All tests are done in replicates of n = 3, statistical significance calculated by two-way ANOVA. *p < 0.1, **p < 0.01, ***p < 0.001, ****p <0.0001. (Fig. 6a created with BioRender.com)

Expanding the scope of technology with TROP2 targeted hLPNs.

We subsequently showcased the broad applicability of the platform by employing another disease-relevant target, trophoblast cell surface antigen 2 (TROP2), given its overexpression in many TNBCs.^48,49 Employing identical linker equivalents and conjugation conditions, we successfully generated the desired LAR of 3 (SI figure S5) in the anti-TROP2 antibody, Sacituzumab, and the corresponding hLPN (Figure 7a–c). Additionally, we verified the specific binding of anti-TROP2-siAF647 in MDA-MB-468 cells (TROP2+) through flow cytometry analysis (Figure 7d, e). To investigate targeting, we also tested these hLPNs on HEK293 as a non-cancerous and low TROP2 expressing cell line.

Figure 7.

Expanding the scope of technology with TROP2 targeted hLPNs a) SDS-PAGE verification of successful conjugation of anti-TROP2 antibody with the nanoparticles. b) The nanoparticles do not undergo a significant alteration in size before and after conjugation with anti-TROP2 antibody. c) Encapsulation efficiency of the nanoparticles remained unaltered with antibody conjugation. d) Immunocytochemistry based quantification of TROP2 receptors on cell surface of MDA-MB-468 and HEK293 cells. e) Relative uptake of anti-TROP2-Tz-7.5% nanoparticles in MDA-MB-468 and HEK293 cells using flow cytometry with siRNA concentration from 0 nM to 200 nM. f) Western blot analysis of p53 expression with anti-TROP2-Tz-7.5% particles encapsulating P53 siRNA. g) Quantitative p53 protein expression in MDA-MB-468 and HEK293 cells post treatment with control TROP2 targeted hLPN. All tests are done in replicates of n = 3, and statistical significance is calculated by two-way ANOVA. *p < 0.1, **p < 0.01, ***p < 0.001, ****p <0.0001.

The TROP2-targeting hLPNs showed ~50% p53 knockdown in MDA-MB-468 cells at only 50 nM siRNA concentration, whereas no significant amount of p53 knockdown was observed with HEK293 even at high siRNA concentrations (Figure 7f, g). Also, the hLPN were screened for potential cytotoxicity in all cell lines using isotype antibody and negative control siRNA, and no significant cytotoxicity was observed across the concentration range (0–200 nM) (Figure S9). Overall, these results demonstrate the potential for translating the conjugation protocols broadly to different antibodies and siRNA cargoes.

AND-gated approach for targeted delivery of siRNA combinations.

So far, we have demonstrated that hLPNs can be successfully decorated with antibodies and that these antibody-functionalized hLPNs can selectively and effectively deliver siRNA to achieve gene suppression. This methodology offers a convenient modular ‘plug and play’ type approach that is compatible with multiple siRNA and antibodies of choice. With this in hand, we posited enhancing target specificity by converging two different cellular epitopes, which deliver two complementary therapeutic cargos to specific cells. This idea of AND-gate affording higher specificity arises from the fact that while one of the disease-specific epitopes might also be present at an off-target location, the likelihood that two different epitopes would be present in the same off-target location is very low. Therefore, converging two cargos that synergistically impact the fate of a diseased cell based on two overexpressed epitopes on that cell offers high specificity.

Here, for an initial test of concept, two distinct siRNAs targeting a pair of genes are independently incorporated into two distinct hLPNs, one with Cetuximab and the other with Sacituzumab as the surface antibody, targeting EGFR and TROP2 respectively (Figure 8a). To investigate the validity of the concept, we utilized MDA-MB-468 cells that overexpress both EGFR and TROP2, T47D cells that lack EGFR but overexpress TROP2, and HEK293 cells that have low or no expression of both antigens (SI figure S8b). We encapsulated two unique dye-tagged negative control siRNAs, labeled with Alexa Fluor 647 and Alexa Fluor 488 (siAF647 and siAF488), separately within Sacituzumab- and Cetuximab-decorated hLPNs respectively.

Figure 8.

a) Concurrent delivery of two distinct siRNA cargos, with enhanced cellular targeting via AND gating, which requires the recognition of two different cell surface antigens. b-d) Flowcytometry analysis of relative binding/ uptake of anti-EGFR-hLPN containing Alexa Fluor 488, anti-TROP2-hLPN containing Alexa Fluor 647 and 1:1 combination of both at effective total siRNA concentration normalized at 200 nM following for b) MDA-MB-468 c) T47D and d) HEK293 cells. e) Depiction of cell populations with either of AF488 or AF647 signal and containing both signal and neither of the signals. f) Quantification of the distinctively fluorescent cell populations for MDA-MB-468, T47D, and HEK293 cells. (Fig. 8a created with BioRender.com)

Uptake profiles of the two hLPNs individually and their combination at 1:1 molar ratio with respect to their encapsulated siRNA content (tested using RiboGreen assay) co-treated in the three cell lines reveal the propensity of anti-TROP2-siAF647 hLPNs are taken up by MDA-MB-468 (~62% cells with Alexa Fluor 647 signal) and T47D (~74% cells) cell lines both of which have high TROP2, compared to only ~3% in HEK293 cells that lack TROP2 expression. Similarly, anti-EGFR-siAF488 hLPNs were taken up only by MDA-MB-468 cells, with approximately 44% of the cell population showing fluorescence from Alexa Fluor 488, compared to ~7% in T47D and ~2% in HEK293 cells. Notably, MDA-MB-468 cells also showed significantly higher accumulation of both anti-EGFR-siAF488 and anti-TROP2-siAF647 nanoparticles, with about 45% of the population containing both siAF488 and siAF647. This is considerably higher compared to T47D and HEK293 cells, where only about 13% and 1% of cells have shown uptake respectively (Figure 8b– f).

Concurrent delivery of TNF and P53 siRNAs towards knocking down protein pairs.

Next, we identified a pair of proteins that synergistically impact cellular fate, when suppressed concurrently. Briefly, alterations in the expression of mutant p53 and tumor necrosis factor α (TNF-α) individually have been demonstrated to impact the fate of the treated cells, but a combined downregulation of both proteins leads to a significantly pronounced detrimental effect.^45,50–57 Our goal was to achieve concurrent knockdown of mutant p53 and TNF-α in a target cell based on the presentation of two distinct surface antigens (Figure 9a).

Figure 9.

a) Optimization of nanoparticle combination dosage in MDA-MB-468 cells. b) Schematic of cell viability testing on treatment with nanoparticles individually and their combinations. Obtained results have been presented with respect to sip53 concentration (Left) and siTNF concentration (right) as individually treated or in combination. The selected dosage for sip53 and siTNF is marked through red square c) Concurrent and target selective knockdown of p53 and TNF-α in MDA-MB-468 (EGFR+/TROP2+), T47D (EGFR-/TROP2+), and HEK293 (EGFR-/TROP2-) cells, P.S. green arrow signifies affinity towards cells and red arrow signifies the lack of affinity. d-f) Cytotoxic effects of combinatorial treatment of both sip53 and siTNF targeting nanoparticles (200 nM at 1:2 molar ratio with respect to siRNA) compared to individual nanoparticles with the same effective siRNA concentrations for 72 h and 96 h in d) T47D e) MDA-MB-468 f) HEK293. g-i) Caspase-8 activation assay in g) T47D h) MDA-MB-468 i) HEK293. All tests are done in replicates of n = 3, and statistical significance calculated by one-way or two-way ANOVA. *p < 0.1, **p < 0.01, ***p < 0.001, ****p <0.0001 (more detailed statistical analysis has been provided in SI figure S13). (Fig. 9a–c partly created with BioRender.com)

The effect of concurrent knockdown of both genes was first assessed using Lipofectamine^™ RNAiMAX transfection reagent in three selected cell lines, as previously described. Briefly, sip53 and siTNF individually and at 1:1 ratio was dosed to the cells for 4 h followed by replenishing with fresh media and additional incubation for 72 h. Indeed, high loss of cell viability compared to the individual siRNAs and untreated controls was observed for all three cell lines (see SI Figure S10), which establishes the validility of the siRNA combination selected here.

To test the potential for selectively converging these two siRNA cargos using hLPNs, anti-EGFR-sip53 and anti-TROP2-siTNF hLPNs were generated using Cetuximab and Sacituzumab antibodies respectively. To determine the optimal dosage, net siRNA concentrations ranging from 50 to 200 nM with varying molar ratios of siP53 to siTNF (1:0, 2:1, 1:1, 1:2, and 0:1) were tested in MDA-MB-468 cells, owing to their concurrent expression of TROP2 and EGFR antigens at significant levels. Notably, anti-TROP2-siTNF hLPNs showed ~65% viability at 200 nM concentration of the siRNA, whereas anti-EGFR-siP53 hLPNs showed ~55% viability at the same concentration. Interestingly, the combined effect of anti-TROP2-siTNF and anti-EGFR-sip53 hLPNs was significantly more potent than the equivalent concentrations of the individual siRNAs (Figure 9b and Figure S11). Among the combinations tested, the 1:2 molar ratio (sip53: siTNF) demonstrated the most significant improvement in cytotoxicity (highlighted with red square in Figure 9b), showing only 40–45% cellular viability at a total siRNA concentration of 200 nM (66.6 nM siP53 and 133.3 nM siTNF. Consequently, this combination was carried forward to the subsequent studies.

To test the impact of the AND-gate strategy, selected cell lines were treated with individual- and combination-hLPNs for 1 h, followed by a washing step and incubation for 72 h and 96 h. For this purpose, the siRNA dosages were fixed to the previously optimized dose, i.e., 66.6 nM anti-EGFR-sip53 hLPN and 133.3 nM anti-TROP2-siTNF (Figure 9b). The combination-treated MDA-MB-468 group (EGFR+/TROP2+) exhibited a significant reduction in viability (~46%) with respect to the untreated control, and the groups treated with anti-EGFR-siP53 (~59%) and anti-TROP2-siTNF (~65%) nanoparticles individually. Remarkably, after 96 h, cell viability in the combination-treated group dropped to ~21%, while further decrease in cell viability was not observed for the individual-hLPN treatment groups.

To test for specificity, we next carried out similar experiments in T47D cells that overexpresses TROP2 but not EGFR. At similar treatment conditions as above, the anti-TROP2-siTNF hLPN exhibited cellular viability of ~56% and ~57% at 72 h and 96 h respectively, while the combination treatments exhibited ~68% and ~58% viabilities at 72 h and 96 h respectively. In these cells, the anti-EGFR-siP53 treatment showed minimal effect with ~87% and ~96% viabilities at 72 h and 96 h respectively, attributable to low or absent EGFR expression. This selectivity was further confirmed in HEK293 cells, which do not express TROP2 or EGFR, where none of the treatments significantly hindered cell growth. Briefly the anti-TROP2-siTNF hLPNs demonstrated ~75–80% viability at both 72 and 96 h timepoints whereas both groups treated with anti-EGFR-siP53 hLPN and the combination of both hLPNs showed >80% viable cells during the time course of the study (Figure 9d – f). These differences are not due to the inherent difference in sensitivities of these cells to the cargos, as discerned by the lipofectamine transfection studies (Figure S10). Together, these results further corroborate the target selective nature of the designed hLPNs to simultaneously deliver a combination of sip53 and siTNF and attain a higher cell killing activity in cells with concurrent expression of both target antigens compared to the cells expressing one or none of the target antigens.

As an additional confirmation, we assessed pro-apoptotic caspase-8 activity for all treatment conditions. We observed a significant ~1.5-fold increase in caspase-8 activity in the combination-treated MDA-MB-468 cells, while no significant changes in caspase-8 levels were detected in the other cell lines, corroborating the observed cell death increase in the combination treatment (Figure 9g – i). To elucidate the underlying mechanisms of enhanced viability loss in the combination-treated cells, we monitored the relative p53 and TNF-α protein expression across all treatment conditions and cell lines. In MDA-MB-468 cells, a combined loss of both p53 and TNF-α protein expression was observed, whereas in T47D cells, only the TROP2-targeted siTNF was effective, significantly reducing TNF-α expression in treatments containing anti-TROP2-siTNF nanoparticles. No loss of expression for either targeted protein was observed in HEK293 cells (Figure S12). These findings further underscore the potential of the ‘AND’ gated strategy to selectively target and effectively reduce the viability of cells expressing both TROP2 and EGFR receptors, offering a promising approach for targeted combination therapy.

3.0. Conclusions:

In summary, we have developed an antibody-decorated lipid-polymer nanoparticles (hLPNs) to selectively home siRNA cargos in cells. Key requirements for targeted hLPNs include a modular nanoparticle platform with tunable surface chemistry and a fast, efficient antibody conjugation process. To achieve this, we employed hLPNs for their modular “plug-and-play” design, enabling rapid antibody functionalization without extensive optimization. Conjugation was facilitated by a highly efficient inverse electron-demand Diels–Alder (IEDDA) reaction using trans-cyclooctene and tetrazine-based click chemistry, allowing antibody decoration while maintaining particle stability. Using breast cancer cells as the model, we optimized tumor-selective hLPN localization by targeting overexpressed antigens such as EGFR and TROP2 and delivering siRNA against mutant P53. Additionally, we report an AND-gated delivery strategy, employing disparate hLPNs to deliver siRNAs (sip53 and siTNF) exclusively to cells expressing both EGFR and TROP2, achieving maximum efficacy. While this strategy enhances selectivity, it introduces trade-offs in efficiency due to the complexity of multi-checkpoint targeting. Future efforts will focus on refining the platform to improve both selectivity and efficiency in vivo, with potential applications in cancers that exhibit multi-antigen profiles.

4. Materials and Characterizations

Synthesis of cationic lipid G0-C14

The cationic lipid G0-C14 was synthesized using a ring opening reaction between 1,2-epoxytetradecane and G0 poly (amido amine) (PAMAM) dendrimer, a detailed reaction scheme has been provided in SI.^39,40 The G0 dendrimer was chosen owing to its siRNA complexation ability along with lower cytotoxicity than the higher generation analogues. The selection of 1,2-epoxytetradecane was based on its optimal 14-carbon structure for siRNA delivery. The reaction involves seven equivalents of 1,2-epoxytetradecane with one equivalent of G0 to form G0-C14The chemical structure was confirmed by ¹H NMR and MALDI-TOF mass spectrometry and compared with the literature reported characterization of G0-C14 (See SI for details).^{39, 42,58}

Synthesis of tetrazine functionalized DSPE-PEG₅₀₀₀

The tetrazine modified lipid–PEG (DSPE–PEG₅₀₀₀–Tz) was prepared using a simple reaction between an activated ester in the form of tetrazine-PEG5-NHS ester and amine-terminated PEG lipid DSPE-PEG₅₀₀₀ amine in presence of a non-nucleophilic base diisopropylethylamine.³⁰ In this reaction, the tetrazine-PEG₅-NHS ester was used in 5 molar excess to ensure complete conversion. The resulting DSPE–PEG₅₀₀₀–Tz was purified using cold methanol washing. The successful formation of DSPE–PEG₅₀₀₀–Tz and the product purity was verified using ¹H NMR and MALDI-TOF mass spectrometry (see SI for details).

Antibody modification with NHS-PEG₈-TCO linker

In this study, anti-EGFR, anti-TROP2, and isotype-IgG antibodies were utilized, with the same protocol followed for linker conjugation for all antibodies. A total of 500 μg of antibody was diluted in 500 μL of reaction buffer, composed of 0.2 M Na₂HPO₄ and 0.1 M NaCl at pH 8.5, followed by buffer-exchange 3 times using a 50 kDa Amicon ultracentrifugation filter. The solution was concentrated to a final concentration of 10 mg/mL. The antibody solution was then transferred to a 500 μL Eppendorf tube, NHS-PEG8-TCO linker (in DMSO) was added, and the mixture was incubated at 4 °C overnight with moderate shaking. For all the antibodies, the target linker-to-antibody ratios (LAR) were fixed at 3 to 4. According to our previously optimized protocol, 7 molar equivalents of the linker were added to the solution.^59,60 Finally, the unreacted linker was removed from the reaction mixture by centrifugation using a 50 kDa Amicon ultracentrifugation filter. The LAR and antibody concentrations were assessed via a fluorescent reporter molecule TAMRA-PEG4-tetrazine alongside UV–vis spectroscopy absorbances at 280 nm (Details of LAR calculations in S5).

Preparation of nucleic acid-loaded NPs

To synthesize the hybrid lipid-polymer nanoparticles, cationic lipid G0-C14 and PLGA were formulated together with either functional or non-functional scrambled siRNA with lecithin and DSPE-PEG¬5000 (85:15 mol: mol) as the external shell. Briefly, the siRNAs were mixed with G0-C14 at 1:10 weight ratio in acetone to create an emulsion where we added 10 weight equivalents of PLGA with respect to G0-C14 to support the core formation of the nanoparticles based on the electrostatic interaction from G0-C14 and enhanced structural integrity provided by the hydrophobic PLGA matrix, which supports encapsulation efficiency, nanoparticle stability, and sustained release. This concoction was added slowly and gradually into a lipid solution at 0.1 mg/mL concentration in 5% (v/v) ethanol /water (pH ~ 5.5) and stirred at moderate to high rotation for a minimum of 2 h to generate the hybrid nanoparticle via nanoprecipitation. The resultant nanoparticle is further concentrated with Amicon Ultra centrifugal filters (MWCO 100 kDa) to remove solvents, followed by buffer exchange with phosphate buffered saline (PBS, pH 7.4).

TEM characterization of siRNA NPs

In this design, the core-shell structure of the nanoparticles is an important attribute that ensures the encapsulation of siRNA and protection from nucleases, as well as the decoupled nature of the nanoparticles, provides modularity in terms of an alteration of the lipid components without sacrificing the encapsulation efficiency. To identify the resultant morphology of the nanoparticles, TEM was utilized to confirm the structure of the siRNA NPs. An organotungsten based negative stain Nano-W^™ was used to provide negative stain contrast in TEM images. The TEM images of the siRNA NPs were produced using a FEI Tecnai-T12 electron microscope.

SDS-PAGE analysis of antibody conjugation

To investigate the conjugation of TCO-modified antibodies with tetrazine-coated hLPNs SDS-PAGE gel was utilized. The smeared nature of the band appearing with an upward shift was considered as an indicator of successful conjugation of the antibodies onto the nanoparticle surface. For the gel run, 5 μg of antibody or the equivalent amount (containing 5 μg of antibody) of antibody-hLPN conjugates were mixed with 10 μl gel loading buffer; then this mixture was loaded in 8% Bolt^™ Bis-Tris plus mini protein gels. The gel was run at 220 V for 25 mins in 1X MES buffer. After running, the gel was stained using SimplyBlue^™ SafeStain reagent to visualize and analyze the protein band using a Bio-Rad ChemiDoc imaging system.

Cell Culture

HEK293 (kidney), T47D (mammary gland/breast cancer), MDA-MB-468 (mammary gland/breast cancer), procured from ATCC, cell lines were cultured in a T-75 cell culture flasks containing Dulbecco’s modified Eagle’s medium/nutrient mixture F-12 (DMEM/F-12) in a humidified incubator with 5% CO₂ at 37 °C. Culture media was supplemented with 10% fetal bovine serum (FBS), 1% L-glutamine, and 1% antibiotic–antimycotic (100 units/mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL amphotericin B).

Cellular uptake studies through confocal flow cytometry and confocal microscopy

For flowcytometric analysis of nanoparticle uptake in cells, fluorescence dye (Alexa fluor 647, Alexa fluor 488) conjugated siRNA encapsulated nanoparticles were incubated with 3 × 10⁴ cells/ well in a 96 well plate in serum free media. The treated cells were incubated with nanoassemblies for 30 min at 37 °C in a 5% CO₂ incubator followed by washing with PBS and subjected to further 4 h incubation at 37 °C in complete media. Afterwards, the cells were trypsinized and pelleted by centrifuging at 3000 rpm for 5 mins. The pelleted cells were redissolved in 35 μL eBioscience^™ flow cytometry staining buffer and analyzed using BD LSRFortessa 3 laser flow cytometer.

For confocal microscopy analyses, 1 × 10⁵ cells were plated in 35 mm glass-bottomed petri dishes and incubated for 24 h at 37 °C in a 5% CO₂ incubator for proper adhesion. Afterward, cells were treated with respective nanoassemblies at preselected concentrations in serum-free media, incubated for 30 mins, washed, and subjected to a further 4 h incubation at 37 °C in complete media. Next, the media was removed, washed three times with PBS, and incubated with NucBlue Live ReadyProbes reagent in FBS containing media for 30 min to stain the cell nucleus. Live cell imaging was performed using a Nikon Spinning Disk Crest V2 confocal microscope. All images were analyzed with ImageJ software.

Knockdown of P53 and TNF-α in MDA-MB-468, T47D and HEK293 cells

To study gene silencing, MDA-MB-468, T47D, and HEK293 cells were plated in a 24-well plate (2 × 10⁵ cells in each well) and incubated for 24 h at 37 °C. After that, cells were transfected with P53 or TNF-α targeted siRNA-loaded nanoassemblies (0–200 nM siRNA concentrations measured by RiboGreen assay, see SI for more details) and incubated for 30 mins for targeted transfection experiments and 24 h for non-targeted experiments. Next, the culture media was replaced with fresh one and incubated for another 48 h at 37 °C. The cells were then harvested and subjected to Western blot analysis of the target proteins.

Western Blot Analyses

Western blot analyses were performed to identify p53 and TNF-α protein levels in cells. Mouse and rabbit monoclonal antibodies (P53, TNF-α, and vinculin mAbs, ThermoFisher scientific / Cell Signaling) were used to detect target proteins and loading control. To isolate total proteins, cells were washed with cold PBS buffer once. Next, the RIPA lysis buffer containing protease/phosphatase inhibitor mix was added to the cell pellet, keeping it in ice and incubated for 30 min, followed by 3 × 30 s sonication to ensure complete lysis of cells. Finally, lysed cells were centrifuged at 14000 rpm at 4 °C to collect soluble protein extracts and quantified with the BCA protein assay, followed by the protein transfer to a poly (vinylidene difluoride) (PVDF) membrane using the Invitrogen^™ iBlot^™ 3 dry transfer system at low molecular weight setting. Next, the membrane was blocked with 5% milk in 1× TBS Tween-20 for 1 h at room temperature, followed by incubation with the designated primary antibody at manufacturer suggested dilution in 5% milk at 4 °C overnight. The membrane was washed with 0.1% Tween-20 three times with a 15 min gentle shaking each time and incubated with a secondary antibody for an hour at room temperature. Finally, the membrane was washed again with 0.1% Tween-20 and incubated with horseradish peroxidase-linked anti-rabbit/mouse IgG was used as the secondary antibody, and protein bands were detected by an enhanced chemiluminescence reagent (Luminol, coumaric acid, and H2O2) using Bio-Rad ChemiDoc imaging system.

CellTiter-Glo^® assay for Cellular Viability

HEK, T47D, and MDA-MB-468 cells were seeded (15 × 10³ cells in 0.1 mL per well) into 96-well tissue culture plates and incubated at 37 °C. After 24 h, cell culture media was replaced with serum-free media containing hLPN nanoassemblies bearing different concentrations of negative control siRNA (25, 50, 100, and 200 nM). Identical control isotype antibody conjugated HLPN nanoassemblies were also tested for toxicity. After 1 h / 24 h of incubation, depending on the treatment condition to mimic, the media was replaced with fresh media and incubated for another 2 days (72 h in total). After that 100 μL of CellTiter-Glo^® (Promega) assay reagent was mixed with equal parts of complete media in the plate. Finally, cells were incubated for 10 min at 37 °C, and solutions were transferred to a white 96-well flat-bottom plate for luminescence intensity measurement with a SpectraMax^® ID5 microplate reader as per manufacturer guidelines.

Caspase 8 activity assay

HEK, T47D, and MDA-MB-468 cells were seeded (15 × 10³ cells in 0.1 mL per well) into 96-well tissue culture plates and incubated at 37 °C. After 24 h, cell culture media was replaced with serum-free media containing hLPN nanoassemblies. After 1 h of incubation, the media was replaced with fresh media and incubated for another 2 days (72 h in total). After that 100 μL of Caspase-Glo^® 8 (Promega) assay reagent was mixed with equal parts of complete media and mixed in a plate shaker. Finally, cells were incubated for 1 h at room temperature, and solutions were transferred to a white 96-well flat-bottomed plate for luminescence intensity measurement with a SpectraMax^® ID5 microplate reader as per manufacturer guidelines.

Supplementary Material

The Supporting Information includes detailed synthetic schemes and characterization data for G0-C14 (Scheme 1; Figs S1, S2) and DSPE-PEG₅₀₀₀-tetrazine (Scheme 2; Figs S3, S4), method for estimating the linker-to-antibody ratio using TAMRA-PEG₄-tetrazine (Fig S5), and quantification of DSPE-PEG₅₀₀₀-Tz incorporation on the hLPN surface using a fluorophore-tagged probe (Fig S6). siRNA encapsulation was assessed using the Quant-iT^™ RiboGreen assay, and receptor expression profiling was performed via flow cytometry. Details of statistical analyses (ANOVA with Tukey’s post hoc test) are also provided. Supplementary data include a list of siRNAs and antibodies used (Tables 1 and 2), verification of TCO-antibody conjugation by SDS-PAGE (Fig S7), demonstration of dual siRNA delivery and AND-gated uptake (Fig S8), cytotoxicity assays in various cell lines (Fig S9), Lipofectamine-mediated dose-response curves (Fig S10), optimization of hLPN dosage and siRNA ratios (Fig S11), gene silencing efficacy for p53 and TNF-α (Fig S12), and caspase-8 activation assays supporting combination effects (Fig S13).