Preparation and characterization of RGD-TAT-CLPs/ATRA@miR-34a

The proposed RGD-TAT-CLPs/ATRA@miR-34a was prepared by a thin layer dispersion method. Firstly, a series of parameters were optimized to construct the blank liposomes with superior particle size and encapsulation efficiency (EE), including egg phospholipids (EP)/cholesterol (Chol) ratio, hydration volume, ultrasonic power and ultrasonic time (Figs. S1-S4). Then, the cationic lipid (2,3-Dioleoyloxy-propyl)-trimethylammonium-chloride (DOTAP) and helper lipid 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) with different adding ratios were involved in the preparation of CLPs (Table S1). The obtained CLPs formulations P1-P6 were subjected to Zeta Potential and safety evaluation. As shown in Fig. 1A, the cationic surface was shown in P4-P6 but not in P1-P3, indicating that DOTAP content greater than 3 mg is essential to construct CLPs. Nevertheless, the greater the DOTAP contents, the worse the safety for normal cells in the respiratory system. Thus, the effect of CLPs on the normal human bronchial epithelial cells Base-2b viabilities was determined to evaluate their safety (Fig. 1B). Considering that the encapsulated ATRA in 100 µg/mL CLPs could meet the safety requirements (Fig. S5), P5 was chosen as the optimized CLPs formulation. Then, the proposed CLPs were further modified with RGD peptide and TAT peptide to construct RGD-TAT-CLPs (Fig. 1C). As shown in Fig. 1D and Figs. S6-S7, the modification of the peptide did not obviously change the particle size, polydispersity index (PDI) and surface Zeta Potential. No significant change of ATRA EE was observed as well, demonstrating the satisfied drug loading capacity of RGD-TAT-CLPs (Fig. 1E). As shown in the transmission electron microscope (TEM) images, an evident shell structure was shown in the peptides-modified CLPs, in which RGD-TAT-CLPs showed the thickest shell, probably due to the dual peptide modifications (Fig. 1F). Then, the storage stability of RGD-TAT-CLPs was also studied based on the particle size, PDI and Zeta Potential measurements, which reflected the perfect stability within 15 days (Fig. 1G and Figs. S8-S9).

Having successfully established the RGD-TAT-CLPs, ATRA and miR-34a were loaded to construct the RGD-TAT-CLPs/ATRA@miR-34a. The ATRA encapsulation ratio was optimized as 1:20 according to the highest EE of about 88.37% and desired drug release profile (Table S2 and Fig. S10). Then, the miR-34a was adsorbed on the optimized formulation with different RGD-TAT-CLPs/ATRA:miR-34a ratios. As shown in Fig. 1H, the unloaded free miR-34a was measured by a gel electrophoresis analysis. Only when the ratio was higher than 10:1 could the miR-34a be completely adsorbed, and the formulation of 20:1 could also exhibit the desired particle size and positive Zeta Potential (Fig. 1I-1J). Thus, the RGD-TAT-CLPs/ATRA:miR-34a ratio of 20:1 was selected as the optimized formulation. Compared to free miR-34a, the stability of the encapsulated miR-34a in fetal bovine serum (FBS) was extraordinarily improved (Fig. 1K). After incubation of 12 h, the miR-34a could still maintain stability while the free miR-34a was completely degraded within 1 h. Taken together, the RGD-TAT-CLPs/ATRA@miR-34a was successfully prepared, which co-loaded ATRA and miR-34a and improved the stability of miR-34a.

Fig. 1

Preparation and characterization of RGD-TAT-CLPs/ATRA@miR-34a: (A-B) Zeta Potential (A) and cytotoxicity on Base-2b cells (B) of different CLPs formulations. (C) The schematical illustration of RGD-TAT-CLPs preparation. (D-F) The particle size (D), ARTA EE (E) and TEM morphology (F) of CLPs, RGD-PEG5000-CLPs, TAT-PEG2000-CLPs and RGD-TAT-CLPs. (G) The stability of RGD-TAT-CLPs. (H) The gel electrophoresis images of free miR-34a in the preparation of CLPs@miR-34a with different targeting peptides and different miR-34a loading. (I-J) The particle size (I) and Zeta Potential (J) of RGD-TAT-CLPs/ATRA@miR-34a with different miR-34a loading. (K) The gel electrophoresis images of miR-34a and RGD-TAT-CLPs/ATRA@miR-34a under FBS condition within 12 h. (n = 3)

Cellular uptake and in vitro antitumor effect

After successful preparation of the RGD-TAT-CLPs/ATRA@miR-34a, the cellular uptake and in vitro antitumor effects were evaluated using human lung tumor cells A549. Equipped with the dual-peptide modifications, the proposed nanosystem was anticipated to efficiently target tumor cells and be internalized. To demonstrate delivery specificity, normal human bronchial epithelial cells (Beas-2b) were employed for comparative analysis. The liposomes were fluorescently labeled with coumarin 6 to facilitate tracking of their cellular distribution. As shown in Fig. 2A and 2C, in A549 cells, the intracellular green fluorescence in peptide-modified CLPs groups was much greater than CLPs groups, proving the higher CLPs uptake amounts induced by RGD and TAT modifications. As anticipated, the RGD-TAT-CLPs group demonstrated the most robust green fluorescence, exhibiting approximately 14.34 times higher intensity compared to the CLPs group. This observation highlights the synergistic effects of RGD’s tumor cell targeting capacity and TAT’s membrane permeation ability. However, it is noteworthy that in Beas-2b cells, while the RGD modification did not significantly enhance cellular uptake efficiency, a slight improvement was observed with the TAT modification in terms of CLPs’ cellular uptake level (Fig. 2B and 2D). These results indicated that the RGD could specifically identify the tumor cells while the TAT could enhance cellular internalization even in non-tumor cells. With this superior tumor cell targeting and uptake capacity, more ATRA and miR-34a were expected to be intracellularly delivered, boosting the antitumor gene therapy.

However, as another intracellular delivery challenge of nanosystems, the lysosome capture and sequestration might severely compromise their therapeutical efficiency [7]. Thus, the lysosome escape ability of the proposed RGD-TAT-CLPs@miR-34a was evaluated. As shown in Fig. 2E, the colocalization of miR-34a and lysosomes were imaged. Compared to free miR-34a, the co-localization of RGD-TAT-CLPs@miR-34a and lysosome sequentially decreased as time went by, indicating the lysosome escape ability of RGD-TAT-CLPs. It could be explained by the proton-sponge effect of cationic DOTAP in CLPs. After being captured by lysosomes, the DOTAP could be protonated and induce the chloride ions influx, which might lead to the osmotic swelling and the physical rupture of the lysosomal membrane.

Subsequently, the in vitro antitumor effect of the proposed RGD-TAT-CLPs/ATRA@miR-34a was determined. As shown in Fig. 2F, the free miR-34a could not induce an obvious antitumor effect, which may be attributed to their instability in the cell culture medium. The free ATRA could neither exhibit a cell-killing effect due to its low working concentration. When the A549 cells were treated with the combination of miR-34a and ATRA, a stronger cytotoxicity was shown, indicating that ATRA could enhance the antitumor effect of miR-34a by augmenting the GJ functions to boost the intercellular transport of miR-34a. Stunningly, when miR-34a was encapsulated in the RGD-TAT-CLPs, the cell cytotoxicity was significantly improved with cell viability decreasing from about 90.36% to about 51.53%. This could be ascribed to the enhanced stability and cellular uptake of RGD-TAT-CLPs@miR-34a. Expectably, the strongest cell-killing effect was observed in the RGD-TAT-CLPs/ATRA@miR-34a group, suggesting the gene therapy promotion effect of the GJ regulating strategy.

Then, the concentration-dependent and time-dependent in vitro tumor growth inhibition effect of RGD-TAT-CLPs@miR-34a (Fig. 2G) and RGD-TAT-CLPs/ATRA@miR-34a (Fig. 2H) was further investigated. As time went on, the cell viability in all the groups obviously decreased, indicating the time-dependent manner. Compared to that of RGD-TAT-CLPs@miR-34a, the cell growth inhibition effect of RGD-TAT-CLPs/ATRA@miR-34a group was much greater with about only 0.35% cell viability in 96 h under 200 nM, demonstrating the desired antitumor effect of RGD-TAT-CLPs/ATRA@miR-34 assisted by the GJ regulating strategy. These results collaboratively proved the enhanced antitumor effect by dual peptide-modified CLPs plus GJ function regulation.

Fig. 2

Cellular uptake and in vitro antitumor effect of RGD-TAT-CLPs/ATRA@miR-34a: (A-B) Cellular uptake of different formulations in A549 (A) and Beas-2b cells (B), where green fluorescence represents different CLPs while blue fluorescence represents cell nucleus. (C-D) Semi-quantitative analysis of A (C) and B (D). (E The CLSM co-localization images of RGD-TAT-CLPs@miR-34a and miR-34a with A549 cells lysosomes at different time points. (F) Cell viability of A549 cells with different treatments. (G-H) In vitro A549 cell growth inhibition effect of RGD-TAT-CLPs@miR-34a (G) and RGD-TAT-CLPs/ATRA@miR-34a (H) with different miR-34a concentrations under different concentrations and time points. (n = 6)

Gene therapy enhancement mechanism by GJ regulating strategy

According to the abovementioned results, the proposed RGD-TAT-CLPs/ATRA@miR-34a exhibited an extraordinary antitumor effect in the A549 cell model, which was attributed to the outcome of the GJ regulating strategy. When the RGD-TAT-CLPs/ATRA@miR-34a was intracellularly delivered, the GJ-relative Cx was expected to be upregulated by ATRA, which in turn enhanced the GJ-mediated miR-34a intercellular transport by the “by-stander effect”. The widespread miR-34a in tumor cells could induce cell apoptosis (Fig. 3A). To consolidate this hypothesis, a series of experiments were carried out.

Firstly, the GJ existence and the function thereof were evaluated by a flow cytometry method, where the “donor cells” stained with Calcein-AM and Dil-CM were seeded into the blank “recipient cells”. Calcein-AM could only be transported between donor and recipient cells through GJ while Dil-CM could not be intercellularly transported. Thus, the percentage of the Calcein+Dil− cells could be employed to analyze the GJ functions. As shown in Fig. 3B and 3C, the ATRA treatment significantly increased the Calcein+Dil− cells from about 6.4% to about 19.5%, indicating the enhancement of GJ in A549 cells. Then, the expression of Cx43, a critical GJ constituting Cx, was investigated with different treatments to further validate the GJ-regulating effect of RGD-TAT-CLPs/ATRA@miR-34a (Fig. 3D and 3E). Surprisingly, the miR-34a and RGD-TAT-CLPs@miR-34a treatments slightly upregulate the expression of Cx43, which might be ascribed to the complicated signaling mechanism. Besides, the Cx43 expression in RGD-TAT-CLPs/ATRA and RGD-TAT-CLPs/ATRA@miR-34a group was much higher than other groups, indicating the payload could upregulate the Cx43 expression to enhance the GJ function. Subsequently, the effect of the enhanced GJ function on miR-34a intercellular transport was determined by treating the “recipient cells” A549 transfected with a green fluorescent protein (GFP-A549) plus “donor cells” A549 cells treated with different Cy3-labelled miR-34a formulations (Fig. 3F and 3G). Thus, the intercellular miR-34a transport could be revealed by the percentage of Cy3+GFP+ cells. As shown, the free miR-34a treatment could slightly increase the Cy3+GFP+ cells percentage from about 0.34% to about 3.98%, indicating the weak GJ functions of A549 cells and the instability of free miR-34a. However, the encapsulation of miR-34a into RGD-TAT-CLPs remarkably upregulated the miR-34a level in GFP-A549 cells, which might be the result of miR-34a stability enhancement. Expectedly, the highest Cy3+GFP+ cells percentage was exhibited in the RGD-TAT-CLPs/ARTA@miR-34a group, demonstrating the superior effect of GJ regulating strategy on miR-34a intercellular transport. These findings collectively support the notion that miR-34a can be transported between cells through GJ, a process further enhanced by ATRA co-delivery.

Then, the relative miR-34a expression of the abovementioned A549-GFP recipient cells was investigated. As shown in Fig. 3H, similar to the results of miR-34a intercellular transport analysis, the improved miR-34a expression levels were observed in RGD-TAT-CLPs@miR-34a and RGD-TAT-CLPs/ATRA@miR-34a groups, and the highest expression level was shown in the latter. These results further validated the feasibility of dual-peptide modified CLPs and the GJ regulating strategy in the enhanced delivery of miR-34a. Afterward, the effects of miR-34a expression on cell apoptosis were studied. It was reported that the p53-related cell apoptosis signaling pathway might be triggered by the miR-34a expression. The Annexin V-FITC/PI flow cytometry method was employed to analyze the percentage of viable apoptotic cells (Annexin V-FITC+PI− cells) and death cells (Annexin V-FITC+PI+ cells). As shown in Fig. 3I-K, miR-34a treatment enhanced the viable apoptotic cell percentage from 8.6 to 17.9%, indicating its apoptosis-inducing function. Nevertheless, much significant cell apoptosis (about 29.2%) and death (about 11.7%) were induced by the RGD-TAT-CLPs@miR-34a, demonstrating the enhanced stability and delivery of miR-34a. The most effective cell apoptosis and cell death were caused by the RGD-TAT-CLPs/ATRA@miR-34a group, suggesting the boosted cell-killing efficiency induced by GJ regulation mediated miR-34a expression. Taken together, the miR-34a-induced gene therapy was validated to be boosted by the ATRA causing GJ regulation and dual-peptide modified CLPs delivery.

Furthermore, in accordance with the gap junction regulatory strategy, it is noteworthy that a strategically designed sequential release system could potentially exhibit a profound antitumor effect, wherein the administration of ARTA would precede that of miR-34a. In this scenario, the initial release of ARTA would effectively upregulate Cx43 expression to facilitate gap junction opening, subsequently enabling the subsequent release of miR-34a to traverse through tumor cells and induce apoptosis. This innovative design will be subject to forthcoming investigation.

Fig. 3

Gene therapy enhancement mechanism by GJ regulation strategy: (A) Schematical illustration of the mechanism of GJ regulation enhanced miR-34a transport. (B-C) The flow cytometry results of donor and recipient A549 cells labeled with Calcein-AM and Dil-CM staining (B) and its semi-quantitative analysis of Calcein+Dil− cells (C). (D-E) The Western blot results of Cx43 in A549 cell after different treatments (D) and its semi-quantitative analysis (E), (1) Control, (2) RGD-TAT-CLPs, (3) miR-34a, (4) RGD-TAT-CLPs@miR-34a, (5) RGD-TAT-CLPs/ATRA, (6) RGD-TAT-CLPs/ATRA@miR-34a. (F-G) The intercellular transport of miR-34a in A549 cells analyzed with flow cytometry (F) and its semi-quantitative analysis of GFP+Cy3+ cells (G). (H) The miR-34a expression level in A549 cells with different treatments. (I-K) The Annexin V-FITC/PI flow cytometry results of A549 cells treated with different treatments (I) and its semi-quantitative results (J and K) (n = 3)

In vivo anti-tumor effect and mechanism on transplant subcutaneous Tumor model

The desired in vitro anti-tumor effect of RGD-TAT-CLPs/ATRA@miR-34a inspired us to evaluate its in vivo antitumor effect, which was first validated on a transplant subcutaneous A549 tumor model. Before evaluating the tumor growth inhibition efficiency, the in vivo tumor-targeting ability of dual-peptide modifications was investigated. The fluorescence probe DID was labeled on the RGD-TAT-CLPs to trace their in vivo accumulation, and the free DID group was set as a comparative group (Fig. 4). After intravenous injection, the RGD-TAT-CLPs/DID quickly accumulated in the tumor site within 2 h while the tumor fluorescence in the free DID group was not obvious (Fig. 4A). The high tumor accumulation in RGD-TAT-CLPs/DID group was maintained until 8 h, indicating its superior tumor targeting capacity. A similar conclusion could be drawn from the ex vivo results, where the highest DID accumulation was recorded in tumors except for the liver in RGD-TAT-CLPs/DID groups while no distinct tumor targeting ability was found in the free DID group (Fig. 4B and 4C). In addition, the CLPs/DID group was further added to evaluate the function of dual-peptide modification (Fig. S11). A significantly enhanced tumor drug accumulation was also exhibited in the RGD-TAT-CLPs/DID group compared to the CLPs/DID group. These results could be explained by the tumor cell targeting capacity of RGD and the cellular transmembrane capacity of TAT. Taken together, RGD-TAT-CLPs could serve as an effective carrier to deliver drugs to tumor lesions, facilitating the antitumor effect.

Fig. 4

In vivo targeting ability evaluation of RGD-TAT-CLPs: (A) The fluorescence images of subcutaneous tumor-bearing mice after the intravenous injection of free DID or RGD-TAT-CLPs/DID at different times. (B-C) The fluorescence image (B) and its semi-quantitative analysis results (C) of major organs and tumors derived from tumor-bearing mice at 8 h. (n = 3)

Then, the in vivo antitumor effects were evaluated (Fig. 5A). There were five groups: Model, RGD-TAT-CLPs, RGD-TAT-CLPs/ARTA, RGD-TAT-CLPs@miR-34a and RGD-TAT-CLPs/ATRA@miR-34a. The tumor volume curves are shown in Fig. 5B and 5C. As shown, the tumor in the model group was enlarged in an exponential manner with the tumor volume exceeding 1000 mm3 on the 14th day. No obvious inhibition effect was exhibited in the RGD-TAT-CLPs group, indicating the negligible antitumor effect of the nanocarrier. However, tumor growth was significantly inhibited in the other three groups. The loading of ATRA or miR-34a into RGD-TAT-CLPs could to some extent exert some antitumor effect. The tumor inhibition effect of ATRA might be ascribed to its tumor plasticity-regulating ability while that of miR-34a was the result of its tumor apoptosis-inducing ability. Limited by the poor tumor cells penetration of miR-34a, the tumor inhibition effect of RGD-TAT-CLPs@miR-34a was not satisfied. In comparison, the strongest tumor-killing effect was shown in the RGD-TAT-CLPs/ARTA@miR-34a group, with the slowest increase in tumor volume in all mice and an average tumor volume of about 370 mm3 after 14 days, which might be attributed to the enhanced miR-34a delivery into deep tumors by the GJ regulating strategy. These results were confirmed by the tumor inhibition rate results (Fig. S12). To further validate the conclusion, the tumors were excised on the 14th day, which were then imaged and weighed. As shown in Fig. 5D and 5E, the lowest average tumor weight (about 235.00 mg) was recorded in the RGD-TATPCLPs/ARTA@miR-34a group, which was much lower than that in the model group (about 729.40 mg), suggesting a 2.78-time higher therapeutical efficiency than RGD-TAT-CLPs@miR-34a group. From the abovementioned results, the effectiveness of the GJ regulating strategy on gene therapy enhancement in vivo was strongly supported.

Afterward, the excised tumors were sliced for pathological examination to analyze their antitumor mechanism. According to the results of H&E staining, the highest tumor damaged area and the lowest tumor cells density were shown in the RGD-TAT-CLPs/ARTA@miR-34a group, indicating its strongest tumor cell-killing effect (Fig. 5F). To validate the delivery of GJ regulating strategy mediated miR-34a to deep tumor tissues, the Cx43 staining was employed to detect the GJ function in tumor tissues (Fig. 5G). As expected, the model group with low Cx43 expression exhibited impaired GJ function. However, treatment with RGD-TAT-CLPs/ARTA and RGD-TAT-CLPs/ARTA@miR34a significantly enhanced green fluorescence intensity, indicating upregulated Cx43 expression in tumor tissues. This observation further supports the notion of improved GJ functions. The augmented GJ function is likely to facilitate intercellular transport of miR-34a and subsequently induce tumor apoptosis. As revealed in the staining images of Bcl-2, a miR-34a-related apoptosis-inhibition protein, the apoptosis rate was remarkably enhanced in the RGD-TAT-CLPs@miR-34a and RGD-TAT-CLPs/ATRA@miR-34a groups, indicating the apoptosis-inducing effect of miR-34a (Fig. 5H). Finally, the miR-34a expression level in tumor tissues was quantitatively measured by qRT-PCR (Fig. 5I). Consistent with the in vitro results, the highest miR-34a expression level was revealed in the RGD-TAT-CLPs/ARTA@miR-34a group, which was about 4.49 times higher than the model group. The miR-34a expression was also higher than the RGD-TAT-CLPs@miR-34a group, suggesting the robustness of the GJ regulating strategy. These results collaboratively proved the great antitumor effects of the proposed RGD-TAT-CLPs/ARTA@miR-34a.

Fig. 5

In vivo anti-tumor effect of RGD-TAT-CLPs/ARTA@miR-34a: (A) Schematical illustration of in vivo antitumor effect evaluation in transplant subcutaneous A549 tumor model. (B-C) A549 tumor volume curves of mice in 14 days with different treatments. (D-E) Tumors images (D) and average weights (E) of tumors dissected from tumor-bearing mice on the 14th day after different treatments. (F-H) H&E staining (F), Cx43 staining (G) and Bcl-2 (H) staining images of tumors from mice with different treatments. (I) The relative expression of miR-34a in tumors after different treatments on the 14th day. (n = 5)

The safety of the proposed RGD-TAT-CLPs/ARTA@miR-34a was also evaluated during the therapeutical period. No obvious body weight change was recorded (Fig. S13). According to the organ weight efficiency analysis, the main organs including heart, liver, spleen, lung and kidney were not obviously damaged, which could also be found in the appearance (Figs. S14 and S15). Then, the main organs were excised for H&E analysis. As shown in Fig. S16, no noticeable abnormalities, such as fibrosis, infiltration, or inflammation, were found in the major organs of treated mice, indicating the good safety of the nanosystem. Finally, revealed by the hemogram examination, no significant impact of the proposed RGD-TAT-CLPs/ARTA@miR-34a on blood cells and functions was found (Fig. S17). Taken together, the presented RGD-TAT-CLPs/ARTA@miR-34a did not cause any observable systematical toxicity.

Construction and characterization of RGD-TAT-CLPs/ARTA@miR-34a-DPIs

Given its desirable lung accumulation ability and low enzyme activity in the respiratory system, pulmonary delivery of gene therapeutic agents holds promise for exerting a more potent antitumor effect compared to other delivery routes for lung tumor treatment. In light of their solid-state form that enhances drug loading and maintains gene stability, DPIs have been recognized as an ideal pulmonary delivery system. Therefore, we aimed to develop DPI formulations of the synthesized RGD-TAT-CLPs/ARTA@miR-34a complex. For this purpose, lactose (Lac), an FDA-approved DPI carrier material, along with hydroxypropyl-β-cyclodextrin (HP-β-CD), a generally recognized as safe (GRAS) material by FDA, were selected as carrier materials for RGD-TAT-CLPs/ARTA@miR-34a-DPIs.

To obtain desired DPIs with satisfied aerosolization properties, a series of Lac- HP-β-CD ratios was set to prepare formulations L0-L7 by a spray drying (SD) method (Table 1). Then, the basic physicochemical properties of the prepared DPIs were investigated to determine the optimized formulation. As shown in Fig. 6A, the particle size of all the formulations was similar, with a d0.1 of about 1.5 μm and a d0.9 of about 5 μm, suggesting the great potential for delivery to the lower airway. It was considered that the appropriate particle size for inhalation should be between 0.5 and 5 μm, which would guarantee particles deposition in the deep lung after multiple deposition mechanisms like collision with the airway surface, gravitational sedimentation and diffusional deposition [39]. Then, the bulk density (ρb) and tap density (ρt) were measured (Fig. 6B). No obvious difference in ρt was observed while ρb decreased from L0 to L3 but increased from L4 to L7, indicating the addition of HP-β-CD might exert significant effects. Among them, the smallest ρb was recorded in L3, implying the best dispersibility and flowability. Further, the morphology of DPIs particles was observed by scanning an electronic microscope (SEM), and the results are shown in Fig. 6C. A perfect spherical and smooth particle was observed in L0 without HP-β-CD addition. With the addition of HP-β-CD, the particle surface became rougher and corrugated in L1-L6, which was more similar to the date stone-like morphology of the HP-β-CD carrier (L7). However, the surface morphology of L4-L7 was too corrugated to maintain the spherical shape, which might induce the particles to cross-link to each other, resulting in worse pulmonary delivery performance. Then, the crystallinity of DPIs was analyzed by Powder X-Ray Diffraction (PXRD). As shown in Fig. 6D, a typical α-crystallinity was found in Lac while HP-β-CD exhibited an amorphous state. For DPIs, no obvious diffraction peaks were found in all the formulations, indicating that the DPIs particles were in the amorphous state. Besides, the hygroscopicity of DPIs was determined as water adsorption might significantly impact the DPIs particles stability, in turn influencing their aerosolization performance. The weight gain of L0 was about 50.7% at 90% RH, suggesting the great hygroscopicity of Lac carrier (Fig. 6E). Nevertheless, with the addition of HP-β-CD, the water adsorption was remarkably inhibited with less than 5% in L1-L7, which might be attributed to the moisture-resistance nature of HP-β-CD. It was revealed that the hygroscopicity was negatively correlated with the HP-β-CD addition amount (Fig. 6F). The above results collaboratively demonstrated that the proper addition of the HP-β-CD could enhance the potential pulmonary delivery performance of DPIs by increasing the surface roughness and inhibiting water adsorption, while L1-L3 may possess the best performance.

Then, the in vitro aerosolization performance of RGD-TAT-CLPs/ARTA@miR-34a-DPIs was investigated by the next generation impactor (NGI), and the results are shown in Fig. 6G. The small geometric standard deviation (GSD) values of all formulations suggested a narrow particle size distribution (Fig. S18). Compared to other formulations, a much higher lower airway deposition (S3 - S7) and a decreased deposition in the device and adaptor, induction port, pre-separator and S1 - S2 were revealed in L3. Further, the highest (about 61.17%) fine particle fraction (FPF) and appropriate (about 2.16 μm) mass median aerodynamic diameter (MMAD) was also validated in L3. Herein, FPF was defined as the fraction of drug effectively deposited in the lung, while MMAD was regarded as the aerodynamic size of particles with a mass cumulative percentage of 50% (Fig. 6H and I) [40]. These results strongly supported that the optimized pulmonary delivery performance would be achieved by L3. And a negative correlation between the FPF value and the ρb of DPIs was also revealed (R2 = 0.9032), which was consistent with other reports (Fig. 6J) [40].

Having constructed the optimized RGD-TAT-CLPs/ARTA@miR-34a-DPIs, the drug loading capacity and stability were investigated before and after the SD process. The size and surface Zeta Potential of loaded RGD-TAT-CLPs/ARTA@miR-34a were determined before and after SD (Fig. 6K and Fig. S19). It was revealed that no obvious change in size and Zeta Potential was observed before and after SD, indicating the great recoverability of RGD-TAT-CLPs/ARTA@miR-34a when delivered into the lung. Then, the stability of the encapsulated drugs was analyzed before and after the SD process. As shown in Fig. 6L, the activity of loaded miR-34a was not impacted by the SD process with similar electrophoresis images. Further, the qRT-PCR results revealed that the relative expression level of miR-34a was not influenced by the SD process, which was still much higher than the control group (Fig. 6M). In addition, the ATRA loading and its release profile were also maintained after the DPIs construction process (Fig. S20). Taken together, a negligible impact of the SD process on the stability and drug loading of RGD-TAT-CLPs/ARTA@miR-34a was found, indicating its high feasibility to serve as a pulmonary delivery platform for antitumor gene therapy.

Based on what we discussed above, the proposed RGD-TAT-CLPs-DPIs were equipped with great pulmonary delivery performance and desired gene delivery stability. Thus, it was believed that the presented DPIs system could serve as a pulmonary gene delivery platform. For further in vitro and in vivo studies, more gene therapy would be supported by the proposed DPIs platform for the treatment of other lung local diseases.

Table 1 Formulation of carrier solution with different ratios of Lac-HP-β-CDFig. 6

Construction and characterization of RGD-TAT-CLPs/ARTA@miR-34a-DPIs: (A-E) Particle size measurements (A), bulk density and tap density (B), surface morphology images (C), PXRD patterns (D) and water adsorption curves (E) of RGD-TAT-CLPs/ARTA@miR-34a-DPIs of different formulations. (F) The correlation between moisture uptake and the HP-β-CD content in RGD-TAT-CLPs/ARTA@miR-34a-DPIs. (G) The in vitro deposition distribution of RGD-TAT-CLPs/ARTA@miR-34a-DPIs of different formulations. (H-I) The FPF (H) and MMAD (I) of different formulations. (J) The correlation between FPF and ρb value of DPIs. (K) The particle size of RGD-TAT-CLPs/ARTA@miR-34a before and after SD. (L-M) The gel electrophoresis images (L) and qRT-PCR detection results (M) of RGD-TAT-CLPs/ATRA@miR-34a before and after SD