Scientific Papers

Micro-syringe chip-guided intratumoral administration of lipid nanoparticles for targeted anticancer therapy | Biomaterials Research


Characterizations of implantable MSC and LNP candidates

To achieve the sustained and uniform intratumoral administration of LNPs, an implantable drug-infusing MSC was precisely fabricated via the photolithography method (Fig. S1). The reservoir-containing bottom and lid parts of the MSC were separately manufactured by photo-crosslinking PSA, and then attached to each other using the PSA precursor (Fig. 1a). The size of the fabricated MSC was 10.0 × 3.7 × 0.1 mm (height x width x thickness) and 1 μL of drug solution could be loaded stably in the drug reservoir of the MSC through a single inlet hole (500 μm in diameter) on the lid. When observed with TEM, the depth of the reservoir was accurately controlled to ~ 40 μm for securing a constant drug loading amount (Fig. 1b). The outer and inner widths of the needle part were uniformly measured as 540 and 200 μm, respectively. Thirteen micro-holes of 40 μm in diameter were arrayed at the tip of the needle, which enabled the gradual release of drug solutions out of the reservoir. Since the MSC was produced using the biocompatible PSA, it was harmless when implanted in bodies for a long time [38]. Moreover, its needle of over 5 mm in length enabled the deep tumor tissue delivery of LNPs. The MSC was directly implantable into tumor tissues after punching a pinhole onto the skin with 26 G stainless needles, smoothly discharging the loaded drug solution without any significant leakage or reflux (Fig. 1c). The MSC was tough and flex enough to endure the possible needle breakage, and it did not cause serious bleeding from the implant area during its 9 h administration.

Fig. 1
figure 1

Characterization of fabricated MSC and LNP candidates (a) The lid and reservoir-including bottom parts of MSC were separately fabricated and sealed together using the PSA precursor. The digital image of the drug-loaded MSC showed that the drug solution was stably charged inside the reservoir of the MSC. b The TEM images of the MSC showed its precise structure with an accurately shaped reservoir and needle, and drug-infusing pinholes on the tip of the needle. c The MSC was smoothly implanted directly into the tumor tissues, slowly infusing drugs without any leakage. d Six LNP candidates with different compositions and surface properties were prepared and e their hydrodynamic diameters and zeta potentials were measured. f The cryoTEM images of LNPs exhibited their uniform circular shapes and hollow structures. g The changes in sizes of LNPs over time were measured via DLS, demonstrating their high dispersion stability for 6 days

The selection of the most desirable LNPs for MSC-guided administration was essential since LNPs would considerably affect the drug delivery efficiency in the tumor microenvironment. Six LNP candidates with different compositions (100 mol% phosphatidylcholine (PC), 90 mol% PC and 10 mol% PEG2000-DSPE (10PEG), 90 mol% PC and 10 mol% phosphatidylserine (10PS), 80 mol% PC and 20 mol% PS (20PS), 90 mol% PC and 10 mol% DOTAP (10DOTAP), 80 mol% PC and 20 mol% DOTAP (20DOTAP)) were prepared through the film casting method to figure out the most preferable LNP for MSC-guided intratumoral drug delivery (Fig. 1d). Briefly, the solution of each lipid composition in 1 mL of chloroform was evaporated to cast a thin lipid film, and the film was hydrated with 1 mL of 3’DW and sonicated to obtain LNPs. All LNP candidates contained 0.5% Flamma 648-PE for their in vitro and in vivo fluorescent imaging analysis. First, the sizes of formulated LNP candidates in PBS were analyzed by DLS, exhibiting no significant difference depending on their compositions (Fig. 1e). Their sizes similarly ranged between 100 to 120 nm, and the size distributions were also monodisperse. In the zeta potential assay of LNPs, neutral PC showed a slightly positive value of + 7.67 ± 0.52 mV and 10PEG was also negatively charged to -24.43 ± 0.68 mV since PEG2000-DSPE has a phosphate group in its molecular structure. PS-containing LNPs (10PS and 20PS) showed negative surface charges (-40.23 ± 0.50 and -45.70 ± 0.79 mV, respectively) due to the negatively charged head group of PS, whereas 10DOTAP and 20DOTAP were highly positive (+ 39.97 ± 1.76 and + 46.50 ± 1.08 mV, respectively) due to the positively charged quaternary ammonium group in DOTAP [39,40,41]. All LNPs were furtherly observed with cryoTEM and determined to have similar sizes with spherical structures and uniform single-bilayer shells (Fig. 1f). When dispersed in PBS at a concentration of 1 mg/mL for up to 6 days, all LNPs stably secured their sizes without significant dissociation or aggregation (Fig. 1g).

In vitro cellular uptake and cytotoxicity of LNPs

The formulated six different LNPs were subsequently applied to cancer cells to examine their cellular uptake profiles and cytotoxicity. To evaluate the cellular uptake of LNPs, 5 × 104 4T1 mouse breast cancer cells were seeded and treated with 0.1 mg/mL of each LNP and monitored using the CLSM for 24 h (Fig. 2a). The fluorescent signals by LNPs (red color) in the cytosol were gradually increased over time and reached their maxima at 24 h. The nano-sized PC, 10PEG, 10PS, and 20PS were moderately taken up by cancer cells regardless of their surface properties, showing a modest fluorescence level without a statistically significant difference (Fig. 2b). Their cellular uptake profile showed a linearly proportional correspondence to their treatment time. Notably, fluorescent signals of PC, 10PEG, 10PS, and 20PS were exclusively distributed in the cytoplasm rather than nuclei since LNPs are not permeable to the nuclear membrane [42]. Meanwhile, the fluorescent intensities of 10DOTAP and 20DOTAP were highest among the LNPs, showing 1.64 ~ 2.89-fold stronger signals than the other four LNPs. The positive surface charge of 10DOTAP and 20DOTAP was supposed to facilitate their interaction with negatively charged cell membranes and lead to their vigorous cellular uptake [43, 44]. However, their in vitro fluorescent signals were clumped and largely overlapped with the DAPI signal, which was not detected in other LNPs. It appeared that the low serum stability of 10DOTAP and 20DOTAP caused their partial agglomeration into submicron clusters inside the culture media and the clusters were attached to the cell membrane due to their strong charges (Fig. S2). Afterward, 4T1 cancer cells were treated again with LNPs at various concentrations ranging from 0.01 to 200 μM for their cytotoxicity assessment (Fig. 2c). All LNPs constantly showed negligible toxicities up to 100 μM, demonstrating their high biocompatibility as drug carriers. The endocytosis of LNPs mainly relied on their surface charges, which could not fully represent their accumulation and distribution inside tumor tissues since tumor microenvironments were not adequately implemented through the in vitro cellular assay. Although 10DOTAP and 20DOTAP expressed the most active interaction with cancer cells and intense endocytosis, their surface properties might not be appropriate for the intratumoral distribution. Therefore, actual delivery behaviors of LNPs with MSC guidance were investigated in the following in vivo test using tumor-bearing mice.

Fig. 2
figure 2

In vitro cellular uptake profile of LNP candidates (a) 4T1 cells were treated with six LNPs (0.1 mg/mL) and their endocytosis was observed via CLSM for 24 h, showing increasing intracellular fluorescent signals over time. b The fluorescent intensities of endocytosed LNPs were quantified and plotted as a function of time (n = 3). c 4T1 cells were treated with six LNPs at various concentrations from 0.01 to 200 μM and conducted with CCK-8 assay, expressing no significant toxicity. ns = not significant, ***p < 0.001

MSC-guided tumoral accumulation of LNPs in tumor-bearing mice

It was hypothesized that the surface properties of LNPs affect their accumulation and distribution inside the tumor tissue as they were administered using the MSC. To discover the most appropriate LNP for the maximal and uniform delivery of drugs throughout the tumor tissue, tumor-bearing mice were prepared and LNP-loaded MSCs were applied directly to tumor tissues to monitor the release and distribution of LNPs. The MSC-guided intratumoral administration of free Flamma 648 dyes was also conducted to compare the intratumoral accumulation behavior of LNPs with that of small molecules and determine the effect of LNP encapsulation on the intratumoral administration. Tumor-bearing mouse models were constructed by subcutaneously inoculating 1 × 106 4T1 cells in the left thighs of BALB/c mice. After the tumor volumes reached ~ 200 mm3, MSCs containing 1 μL of each LNP (30 mg/mL) or Flamma 648 solution (0.15 mg/mL) were directly implanted into the center of tumor tissues, and the release and localization of LNPs or free dye were non-invasively observed via fluorescence imaging device for 9 h. The implanted MSCs were protected by plastic caps to prevent any contamination or damage to the MSC (Fig. S3). When comparing the fluorescent signals of LNPs inside MSCs before and after their administration, the percentages of LNPs released out from MSCs were about 50 ~ 60%, similar to one another regardless of LNP compositions (Fig. S4). The LNPs were gradually released from MSC and diffused into tumor tissues over 9 h, which was detectable via in vivo fluorescence imaging of tumors (Fig. 3a). The fluorescences of administered LNPs from MSC were observed exclusively inside the tumor tissues, and the signals were maintained during the 9 h of infusion procedure. However, the intratumoral fluorescence of free Flamma 648 dye reached its highest at 3 h post-administration, and the intratumorally remaining dye was not observed after 6 h. The MSC-guided intratumorally administered free Flamma 648 dye was considered to rapidly diffuse through the tumor interstitium and washed out from tumor tissues due to the excessive diffusivity of small molecules [45]. LNPs were more advantageous for MSC-guided intratumoral administration than free dye due to their moderate intratumoral diffusivity and prolonged retention. It was notable that the intratumoral accumulation of LNPs was closely related to their surface properties. Neutral PC and negative 10PS and 20PS rapidly accumulated at targeted tumor tissues within 1 h post-administration, wherein the bright red fluorescent signals of LNPs were clearly observed in tumor tissues. Furthermore, most of their fluorescent signals were maintained up to 9 h post-administration. In particular, 10PS was freely diffused into the tumor and expressed the brightest fluorescence compared to other LNPs. On the contrary, the fluorescence of 10PEG gradually spread to the entire tumor tissue over time, but its intratumoral accumulation at 9 h was not high because of its extratumoral leakage. Moreover, positively charged 10DOTAP and 20DOTAP did not show perceptible fluorescent signals at tumor tissues for 9 h, indicating the lowest tumor targeting ability of positively charged LNPs. When quantifying the intratumoral fluorescent intensity over time, the tumor infusion rates of LNPs were most rapid in the first 1 h and saturated at 6 h (Fig. 3b). Noticeably, the fluorescent intensity of 10PS was the highest throughout the infusion time, which was 1.58 ~ 4.80-fold higher compared to those of other LNPs at 9 h post-administration. However, the positively charged 10DOTAP and 20DOTAP showed low fluorescent intensity over 9 h of MSC-guided administration, and that of free Flamma 648 dye was the minimum due to its fast diffuse-out.

Fig. 3
figure 3

In vivo intratumoral accumulation analysis of LNP candidates administered via MSC guidance (a) The IVIS of 4T1 tumor-bearing mice treated with six LNPs (30 mg/mL, 1 μL) or free Flamma 648 dye (0.15 mg/mL, 1 μL) via MSC guidance showed the time-dependent intratumoral accumulation behaviors of LNPs and Flamma 648 dye. b The in vivo intratumoral fluorescent intensity was measured over the administration time, showing the highest accumulation of 10PS. c The ex vivo fluorescence images of excised tumor tissues 9 h after the MSC-guided LNP treatment were obtained and (d) their fluorescent intensities were quantified. e The intratumoral delivery efficiencies of six LNPs and Flamma 648 dye were calculated by the ratios of tumor-remaining LNP or dye concentrations over those released from the MSC

After 9 h post-administration, tumor tissues were excised for a more precise analysis of intratumoral LNP and free Flamma 648 localization. The strongest intratumoral fluorescent signal of 10PS was clearly observed in tumor tissues, while free Flamma 648 rarely remained inside the tumor tissues (Fig. 3c). The fluorescent intensity of 10PS was 2.2 ~ 4.9-fold higher than those of other LNPs, whereas 10DOTAP and 20DOTAP showed the lowest fluorescent signals among the LNPs (Fig. 3d). To determine the actual delivery efficiency of LNPs from MSC, each tumor was homogenized and the amount of each LNP in the entire tumor tissues was measured using the high-performance liquid chromatography (HPLC). As expected, 10PS exhibited a high delivery efficiency of 93%, which was calculated by the ratios of LNP concentrations inside tumor tissues to those released out from MSCs (Fig. 3e). However, other LNPs showed decreased delivery efficiencies from 32 to 68% in the HPLC analysis, and the delivery efficiency of free Flamma 648 was merely less than 2%. Based on these data, the negatively charged 10PS was determined as the most desirable LNP for the MSC-guided intratumoral administration, while positively charged or PEGylated LNPs can hinder their efficient localization inside tumors.

The excised tumors were subsequently chopped into three tissues with a 1 mm-thickness interval to observe the intratumoral distribution of LNPs. Each section was stained with DAPI (nuclei staining, blue color) and analyzed with CLSM, showing the specific localization of Flamma 648-PE-labeled LNP fluorescent signals (red color) (Fig. 4a). In the confocal whole-section images of Sect. 1, the central region of MSC administration, the tumor tissues treated with PS, 10PS, 20PS, 10DOTAP, and 20DOTAP showed the concentrated LNP accumulation in Sect. 1, whereas 10PEG-treated tumor tissue showed the minimum fluorescent signal, indicating the lowest localization in tumor tissues. Notably, only 10PS showed bright and uniform fluorescence distribution throughout Sect. 2 and Sect. 3, indicating that negatively charged LNPs with 10 mol% PS might freely diffuse from the center of MSC administration to whole tumor tissues and then be robustly taken up by cancer cells before eluding out at targeted tumor tissues. In contrast, LNP-administered tumors except 10PS showed relatively dim fluorescence in Sect. 2, marginally apart from the administration site, and the intensity was further decreased at the distal Sect. 3. The diffusivity of PC, 10DOTAP, and 20DOTAP was considered not adequate for their intratumoral distribution. Notably, all three sections of 10PEG-treated tumors were the lowest in fluorescence level since 10PEGs were diffused out from the tumor tissue rather than remaining inside it due to the PEGylation effect. When magnifying the confocal images of Sect. 2, tumor tissues treated with other LNPs than 10PS had mottled fluorescence patterns due to the skewed distribution of LNPs (Fig. 4b). Considerable differences in fluorescent signals between LNP-concentrated and sparse regions were observed, indicating their inhomogeneous localization. The partial accumulation of 10DOTAP and 20DOTAP was clearly noticeable (white square), being attributed to their aggregation in the physiological condition and thereby abnormal diffusion in whole tumor tissues. In the case of 10PS-treated tumor tissue, however, the fluorescent signal was uniformly distributed and the regional fluorescent intensity gap between the highest and the lowest was ignorable. As a result, 10PS was finally chosen as the most adequate LNPs for MSC-guided drug delivery since it could be efficiently taken up by cancer cells, highly accumulated inside tumors, and uniformly diffuse throughout whole tumor tissues.

Fig. 4
figure 4

Ex vivo intratumoral distribution assay of MSC-guided administered LNP candidates (a) The CLSM images of tumor sections from three different regions were acquired, determining that the 10PS-treated tumor tissue expressed the highest fluorescent signals throughout all sections. b The brightest and darkest areas in the CLSM images of Sect. 2 were magnified to precisely compare the regularity of LNP distribution inside the tissues. 10PS was found to be most evenly distributed inside the tumor. c The extracted organs and tumors from all recipient mice were examined with IVIS and (d) their fluorescent intensities were quantified

Fluorescent signals from normal organs were additionally analyzed ex vivo to confirm any undesirable delivery of LNPs in normal tissues (Fig. 4c). The fluorescent signals of all LNPs were barely detectable from all organs regardless of the MSC-guided administered LNP types, indicating that the non-specific accumulation of LNPs in normal tissues was ignorable. The quantified fluorescent intensities of the organs also exhibited no statistical difference from one another (Fig. 4d). The utilization of MSC for the intratumoral LNP delivery allowed LNPs to be slowly infused into tumor tissues via their physiological diffusion mechanism, providing them sufficient time for 9 h to spread throughout the tumor tissues. Moreover, LNPs could be administered precisely into the center of tumors through the microchannel of MSC, which promoted their homogenous diffusion to the tumor periphery without any inappropriate delivery to normal tissues. In addition to the infusion control by MSC guidance, the surface chemistry of LNPs further affected their intratumoral distribution, wherein highly cationic LNPs were unevenly dispersed and PEGylated LNPs tend to rapidly elude out from tumors. These data indicate that MSC-guided intratumoral administration of LNPs can solve the serious problems of intravenously administered LNPs such as interactions with blood components and entrapment to the reticuloendothelial system (RES) that greatly disturb their passive tumor targeting at targeted tumor tissues.

Characterization and in vitro cellular assay of ApoLNPs

For the MSC-mediated cancer therapy, cathepsin B-cleavable and pro-apoptotic prodrugs (SMAC-P-FRRG-DOXs) were encapsulated into 10PS, which had been determined as the optimal LNP for intratumoral administration, resulting in ApoLNPs (Fig. 5a). The SMAC-P-FRRG-DOX prodrug is a conjugate of second mitochondria-derived activator of caspases mimetic peptides (SMAC-P; AVPIAQ) and DOX with a cathepsin B-cleavable peptide linker (FRRG). The SMAC-P-FRRG-DOX is subsequently cleaved to SMAC-P and DOX in cathepsin B-overexpressing cancer cells. It has been already reported that SMAC-P-FRRG-DOX showed a significant antitumor efficacy owing to the synergistic activity of the inhibitor of apoptosis proteins (IAPs) antagonism with chemotherapy in drug-resistant breast tumor models [46]. The SMAC-P-FRRG-DOX was successfully synthesized via the simple esterification coupling of SMAC-P-FRRG and DOX, which was confirmed through HPLC and LC–MS analysis (m/z = 841.7, [M]/2 + 1H+ and 1683.2, [M] + 2H+) (Fig. S5). Subsequently, 3 mg of SMAC-P-FRRG-DOX was loaded into 30 mg/mL 10PS through the film casting method, and the loading efficiency of SMAC-P-FRRG-DOX was 98% [47]. The freshly prepared ApoLNPs were dispersed in PBS at 1 mg/mL and analyzed via DLS, determined to have a mean diameter of 78.95 ± 0.96 nm (Fig. 5b). The cryoTEM image of ApoLNPs exhibited its spherical and single-bilayered structure, similar to that of 10PS before the SMAC-P-FRRG-DOX encapsulation. In addition, the size and polydispersity of ApoLNPs in PBS were stably maintained for up to 6 days (Fig. 5c). The release profile of SMAC-P-FRRG-DOX from ApoLNPs was monitored for 10 days and 40% of SMAC-P-FRRG-DOX slowly released for 4 days (Fig. 5d). The cathepsin B-responsive degradation of SMAC-P-FRRG-DOX was further confirmed in bench condition (Fig. 5e). When 0.1 mM SMAC-P-FRRG-DOX was incubated with 10 μg/mL cathepsin B for 24 h and analyzed via HPLC, the peak corresponding to SMAC-P-FRRG-DOX completely disappeared and the new peak of G-DOX newly emerged. The cleaved G-DOX was analyzed via LC–MS again, confirming the consistency of m/z to its theoretical value (m/z = 623.3, [M] + Na+) (Fig. S6). The G-DOX cleaved from SMAC-P-FRRG-DOX was successfully metabolized into free DOX by intracellular proteases in lysosomes [48, 49]. Next, 4T1 cancer cells were treated with ApoLNPs to investigate their cellular uptake behavior and cytotoxicity. Firstly, 5 × 104 4T1 cells were seeded and cultured with ApoLNPs containing 5 μM of SMAC-P-FRRG-DOX and observed via CLSM at 1, 6, and 24 h. As expected, 4T1 cancer cells expressed 5.6-fold higher cathepsin B compared to normal cells of rat cardiomyocytes (H9C2), indicating that the prodrug of SMAC-P-FRRG-DOX in ApoLNPs can be specifically cleaved via cancer cell-overexpressed cathepsin B (Fig. 5f) [46]. The fluorescent signals of DAPI (nuclei, blue color), Flamma 648-PE (10PS LNPs, red color), and DOX (SMAC-P-FRRG-DOX or free DOX, green color) were separately obtained to evaluate time-dependent cellular uptake and intracellular localization of ApoLNPs (Fig. 5g). The fluorescent signal of ApoLNPs in cytosols continuously increased over time and was maximized at 24 h, showing the similar cellular uptake behavior to that of 10PS. Notably, the fluorescence of DOX mainly existed in cytosols rather than nuclei at 1 and 6 h, but some of it could be also detected inside nuclei at 24 h. The ratio of DOXs inside the nuclei and cytosol at 24 h was 0.41 ± 0.01, 3.4-fold higher than that at 6 h (Fig. 5h). It was considered that SMAC-P-FRRG-DOXs were slowly released from ApoLNPs and cleaved by cathepsin B to SMAC-P and free DOX [46]. On the contrary, the fluorescent signal of 10PS (green color) was rarely found in nuclei since LNPs could not penetrate the nuclear membrane. The ratio of intranuclear and cytosolic 10PS during the cellular uptake assay did not exceed 0.07 and it was 9.2-fold lower than that of DOX at 24 h. The cell-internalized ApoLNPs induced cancer-specific and pro-apoptotic cell death due to cancer cell-overexpressed cathepsin B cleavage mechanism of SMAC-P-FRRG-DOX (Fig. 5i). The IC50 value of ApoLNPs against 4T1 cells was 6.6 μM, with a slight difference from that of free DOX (3.9 μM) due to the delayed drug release from ApoLNPs. In contrast, ApoLNPs showed greatly reduced toxicity to H9C2 rat normal heart cells with 230.7 μM IC50 value, whereas free DOX was substantially toxic to normal cells (IC50 = 1.1 μM). Moreover, ApoLNPs efficiently suppressed the DOX-induced drug resistance through the IAPs antagonism by SMAC-P released from ApoLNPs. In the western blot assay, 4T1 cells treated only with DOX expressed a 1.2-fold excessive level of inhibitors of IAPs compared to the saline-treated ones, which regulates programmed cell death of cancer cells (Fig. 5j) [50]. Meanwhile, ApoLNP-treated cells had a relatively low IAPs expression level since the co-delivered SMAC-P directly blocked IAPs and led them to be degraded [51]. The level of IAPs of ApoLNP-treated cancer cells was 3.1-fold lower than that of DOX-treated 4T1 cells. Therefore, ApoLNPs demonstrated their highly specific pro-apoptotic cell death induction in cathepsin B-overexpressing cancer cells.

Fig. 5
figure 5

In vitro characterization of ApoLNP (a) ApoLNP was prepared by encapsulating the SMAC-P-FRRG-DOX prodrug inside the 10PS. b The formed ApoLNP (1 mg/mL) was assessed via DLS and cryoTEM, confirming a similar size and morphology to that before SMAC-P-FRRG-DOX loading. c The dispersion stability of ApoLNP in PBS buffer (1 mg/mL) was observed for up to 6 days using DLS. d The ApoLNP (15 mg/mL) was placed into a dialysis bag (MWCO = 12–14 kDa) and dialyzed with 10 mL PBS for 10 days to examine the release behavior of SMAC-P-FRRG-DOX. e The HPLC analysis of SMAC-P-FRRG-DOX (0.1 mM) before and after the incubation with cathepsin B (10 μg/mL) displayed its cathepsin B-specific cleavage into SMAC-P-FRR and G-DOX (f) The excessive cathepsin B expression level of 4T1 cancer cells compared to that of H9C2 normal cells was analyzed via western blot assay (n = 3). g 4T1 cells were treated with ApoLNP (5 μM, based on SMAC-P-FRRG-DOX content) and their CLSM analysis was carried out at 1, 6, and 24 h after the treatment to determine its endocytosis and intracellular distribution. h The ratios of LNPs (Flamma 648-PE, red color) and DOXs (green color) localized inside the cytosol and nuclei were compared to confirm the increased intranuclear distribution of DOXs over time (n = 3). i 4T1 cancer and H9C2 normal cells were exposed to various concentrations of ApoLNPs, DOXs, and 10PSs from 0.01 to 200 μM and their CCK-8 assay was conducted to prove the selective toxicity of ApoLNP to cathepsin B-overexpressing cancer cells. j The reduced IAP expression level of ApoLNP-treated 4T1 cells over DOX-treated ones was determined via the western blot (n = 3). * p < 0.05, **** p < 0.0001

In vivo tumor accumulation of ApoLNPs in 4T1 tumor-bearing mice

The tumor accumulation of ApoLNPs depending on their administration routes was investigated using 4T1 tumor-bearing mice. 1 × 106 4T1 cells were subcutaneously administered into BALB/c mouse models, then as tumor tissues grew to ~ 200 mm3, ApoLNPs (0.15 mg/kg, based on DOX content) were administered to the mouse models through different routes (intravenous (IV), intratumoral (IT), or MSC-guided). The different accumulations of ApoLNPs inside tumor tissues were monitored in vivo for 48 h via an IVIS imaging system, visualizing the fluorescence signals by LNPs (Flamma 648-PE) (Fig. 6a). In the case of IV administration, ApoLNP was hardly delivered to tumor tissues thereby the intratumoral fluorescence was not different from that with saline administration, indicating the poor delivery efficiency of the systemic administration. In the meantime, the direct IT administration exhibited significantly higher tumor accumulation of ApoLNP than its IV administration, but a large proportion of administered ApoLNP rapidly diffused out within 6 h, and no fluorescent signal was detected inside the tumor tissue at 24 h after administration. In particular, the MSC-guided administration for 6 h successfully facilitated the enhanced and long-lasting tumor accumulation of ApoLNP. The quantified data of time-dependent changes in tumor fluorescent intensity showed that the area under the curve (AUC) of MSC-guided administration was 584.8 ± 22.6, which was 5.1- and 2.3-fold larger than those of IV and IT administration, respectively (Fig. 6b). Although the peak fluorescent intensity of the MSC-guided administration was similar to that of the IT administration, it allowed the ApoLNPs to remain in the tumor more persistently. The immediate IT administration of ApoLNPs caused them to be excessively localized inside the tumor tissues at once, resulting in their leakage through the blood and lymphatic vessels before being uniformly dispersed and endocytosed. In contrast, the MSC-guided administered ApoLNPs could sufficiently diffuse all around the tumor tissues and be taken up by cancer cells during 6 h of infusion time, thereby achieving a higher and persistent tumor targeting efficiency.

Fig. 6
figure 6

In vivo investigation for the intratumoral accumulation of ApoLNP with different administration routes (a) 4T1 tumor-bearing mice were administered with ApoLNPs (0.15 or 3 mg/kg, based on DOX content) via IV, IT, and MSC-guided routes and their IVIS images were collected for 48 h (fluorescence signals of Flamma 648-PE in 10PS). b The in vivo tumor fluorescent intensities were calculated and described as a function of time. c After 48 h of intravenous, intratumoral, and MSC-guided ApoLNP treatment, tumor tissues were extracted to take their ex vivo IVIS images, and (d) their fluorescent intensities were measured (n = 3). e The distribution of ApoLNPs inside IV, IT, and MSC-guided administered tumors was assessed using the CLSM. f The pharmacokinetics of ApoLNPs administered via different routes were analyzed by quantifying the changes in blood plasma concentrations of ApoLNPs over 48 h through the HPLC. ns = not significant, *** p < 0.001, **** p < 0.0001

The ex vivo images of tumors with different administration routes after 48 h also proved the effectiveness of MSC-guided administration (Fig. 6c). The fluorescent signal in the tumor with low-dose IV administration was the lowest, and the signal gradually became brighter in the order of IV, IT, and MSC-guided administration. The fluorescence of the tumor tissue with MSC-guided administration was 4.6- and 1.7-fold brighter than those of IV and IT administration, respectively. This was deduced that a considerable amount of intratumorally administered ApoLNPs leaked out from the tumor whereas those administered by MSC guidance still remained inside the tumor tissue (Fig. 6d). The intratumoral accumulation of ApoLNPs was further confirmed via the CLSM analysis of tumor tissues (Fig. 6e and S7). After 48 h post-administration, the tumor tissue with MSC-guided ApoLNP administration expressed the strongest fluorescence which was 2.2-fold higher than the IT-treated one, whereas the fluorescence was not detectable in the IV-treated tumor (Fig. S8). In addition, ApoLNPs were most uniformly distributed inside the whole tumor tissue when they were administered using the MSC. The MSC-guided administration of ApoLNPs showed the promoted tumor targeting efficiency at targeted tumor tissues compared to IV and IT administration.

To investigate the effect of LNPs on MSC-guided intratumoral administration, free DOXs, SMAC-P-FRRG-DOXs, and ApoLNPs were intratumorally administered to 4T1 tumor-bearing mice at their doses of 0.15 mg/kg (based on DOX content) through the MSC guidance and their intratumoral accumulation was observed for 48 h via the IVIS imaging. In the case of the ApoLNP administration, the fluorescence signals by DOXs could be detected for a prolonged period of up to 48 h, demonstrating the enhanced intratumoral retention of ApoLNPs due to the moderate diffusivity of 10PSs (Fig. S9a). In contrast, MSC-guided intratumorally administered free DOXs smoothly were released into tumor tissues for 3 h but they were rapidly diffused out from tumor tissues, thereby no fluorescence signal by DOX was visible after 6 h of administration. SMAC-P-FRRG-DOXs could not remain inside the tumor tissues longer than 6 h either, indicating that the LNP encapsulation is essential for the persistent intratumoral localization of drugs in their MSC-guided administration. The intratumoral fluorescence intensity of ApoLNPs at 48 h post-administration was 1.9- and 3.2-fold higher than those of SMAC-P-FRRG-DOXs and free DOXs, respectively (Fig. S9b). The intratumorally remaining concentration of SMAC-P-FRRG-DOXs was considered slightly higher than that of free DOXs due to their higher molecular weight and lower diffusivity, but not sufficient compared to ApoLNPs. The utilization of LNPs in the MSC-guided intratumoral administration was proven to improve the long-term tumor-specific localization of drugs, which would directly affect their therapeutic efficacy and systemic toxicity.

After determining the intratumoral accumulation of ApoLNPs depending on their administration routes, their pharmacokinetic behaviors and distribution in normal organs were analyzed over time. When tracking their plasma concentration for 48 h, the concentrations of ApoLNPs with IV, IT, and MSC-guided administration at 0.15 mg/kg doses were significantly low in blood samples of recipient mice since their doses were about 20-fold lower than the typical intravenous administration (Fig. 6f). The AUC of IV, IT, and MSC-guided administered ApoLNPs at 0.15 mg/kg doses were 0.52, 0.27, and 0.25, respectively, 204.8 ~ 426.0-fold lower than that of IV-administered ones at a 5 mg/kg dose. Further, the AUC of MSC-guided administered ApoLNPs was 2.1-fold lower than that of IV-administered ones at the same doses, confirming the prolonged tumor retention and diminished extratumoral leakage of ApoLNPs via their MSC-guided administration. In the ex vivo fluorescence images of normal organs, the fluorescence signals by MSC-guided administered ApoLNPs were rarely detected, thereby their accumulation to non-target tissues was considered ignorable (Fig. S10). Those results in pharmacokinetic profiles and biodistribution assessment of ApoLNPs indicated that the MSC-guided administration can effectively prohibit the undesirable distribution of ApoLNPs in normal tissues by delivering them tumor-exclusively and reducing their required minimal doses.

Therapeutic efficacy of MSC-guided administered ApoLNPs against tumor-bearing mice

The therapeutic efficacy of ApoLNPs with MSC-guided administration was assessed via their repeated application to tumor-bearing mice and compared to that with the IT and IV administration. When 4T1 tumor tissues grew to ~ 100 mm3, ApoLNPs (0.15 mg/kg, based on DOX content) were administered to the mouse models once per three days (three times in total), wherein MSC-guided administration was carried out for 6 h. The repeated ApoLNP administration was continually observed through fluorescence imaging. As expected, ApoLNPs could be smoothly administered into tumor tissues three times through both IT and MSC-guided administration (Fig. 7a and S11). The fluorescent intensity of the MSC-administered tumor reached peak points 6 h after each administration and gradually increased in tumor tissues for 10 days, which was similar to the single administration profile (Fig. 7b). In contrast, the fluorescent intensity of the IT-administered tumor was decreased immediately after the administration, thereby the AUC of IT-administered tumors was 1.7-fold lower than the MSC-guided administered ones. Notably, MSC-guided administered tumors at 7–10 days exhibited 2.8 ~ 47.8-fold brighter fluorescent signals than those of IT administration, confirming the persistent intratumoral remaining of MSC-guided administered ApoLNPs.

Fig. 7
figure 7

In vivo therapeutic efficacy comparison of ApoLNP with different administration routes (a) During the MSC-guided ApoLNP treatment (0.15 mg/kg, based on DOX content, once per 3 days), the IVIS images of recipient mice were acquired and (b) the intratumoral fluorescent intensities were calculated. c The tumor sizes were measured for 12 days to compare the difference in therapeutic effect according to the ApoLNP administration routes, and (d) their digital images were also obtained (black dashed line = tumor) (n = 3). e Upon completing the treatment procedure, tumor tissues were collected and their H&E histology was performed. f The tumor tissues were further stained with TUNEL to determine the degree of tumor apoptosis. g The tumoral IAP expression levels depending on the ApoLNP administration routes were evaluated via western blot (n = 3). * p < 0.05, ** p < 0.01, **** p < 0.0001

To evaluate the therapeutic efficacy of ApoLNPs, the average tumor sizes were measured every two days and compared to those with saline treatment (black dashed line = tumor) (Fig. 7c and d). The MSC-guided ApoLNP administration successfully suppressed the tumor growth with 195.02 ± 75.65 mm3 of tumor sizes after 12 days. The tumors of the MSC-guided ApoLNP administration group were 4.7- and 2.2-fold smaller than those of saline-treated (1036.50 ± 313.44 mm3) and IT-administered (430.02 ± 91.80 mm3) groups, respectively, confirming the therapeutic effectiveness of MSC-guided administration. Noticeably, IV administration did not show any therapeutic efficacy, thereby the growth of IV-administered tumors was similar to that of the saline-treated ones. This was attributed to that the 0.15 mg/kg was too low compared to the conventional DOX dose for its intravenous administration, which is generally 1–3 mg/kg [46]. After the treatment, all experimental groups were sacrificed and their normal organs and tumor tissues were harvested for further analysis. The ex vivo fluorescence images of organs from MSC-guided ApoLNP administration groups showed not much difference in their fluorescent intensities compared to those from saline-treated groups even after three times repeated doses, indicating that no significant leakage and off-target accumulation of ApoLNPs were observed (Fig. S12). The IT-administered ApoLNPs were slightly detectable in the liver, which was not statistically significant. During the therapeutic procedure, all treatment groups did not show any notable weight loss (Fig. S13). This was attributed to the inactiveness of SMAC-P-FRRG-DOX against normal tissues that exhibit low cathepsin B expression levels. In addition, the histological images of normal organs from the three-times ApoLNP-treated groups did not exhibit significant damage regardless of the administration route (Fig. S14). The organs were preserved with their integrities similar to those from saline-treated groups, demonstrating the systemic safety of ApoLNPs.

The excised tumor tissues were further precisely analyzed by staining them with H&E, TUNEL, and IAPs antibodies. In the histological assay, MSC-guided ApoLNP-treated tumor tissues were determined to be seriously damaged due to the therapeutic effect of localized ApoLNPs (Fig. 7e). Moderate damages were also found in the tumors treated with IT administration of ApoLNPs, but not so intense as MSC-guided administration since the tumor accumulation of ApoLNP was not sufficient. The ApoLNP-induced tumor apoptosis was visible more obviously in the TUNEL images (Fig. 7f). The free DOX and SMAC-P released from ApoLNPs were thought to synergistically act in tumor tissues, leading to their severe apoptosis. Moreover, the IAPs expression was remarkably suppressed in the tissues with MSC-guided ApoLNP administration, which was attributed to the activity of released SMAC-P (Fig. 7g and S15). The IAP expression level of MSC-guided ApoLNP-administered tumors was 5.4-fold lower than those with saline treatment. The level was also 1.7-fold reduced to that of the IT administration, expecting that the MSC-guided ApoLNP administration would be highly effective in drug-resistant tumor species as well. Finally, it was fully confirmed that the MSC-guided administration of ApoLNPs could maximize the tumor-specific therapeutic efficacy, far reducing the off-target toxicity in normal organs.



Source link