HeLa human cervical cancer cells and L929 mouse fibroblast cells provided from the American Type Culture Collection (CCL-2) were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) medium with 10% fetal bovine serum, 1% penicillin/streptomycin containing L-glutamine and HEPES, in 25 cm² flasks, at 5% CO₂ and 37°C ambient conditions. When the cells filled the flasks 80%–90%, they were passaged into new 25 and 75 cm² flasks, and the stocks were multiplied for use in subsequent experiments (Erdogan et al., 2019).

The HeLa human cervical cancer cells were seeded into 75 cm² flasks. They were cultured at 5% CO₂ and in DMEM medium containing 10% fetal bovine serum and 1% penicillin-streptomycin. When the cells reached 90% occupancy, they were washed twice with Dulbecco’s phosphate buffered saline (DPBS). To induce exosome release, low glucose DMEM medium containing only 1% penicillin-streptomycin was added to the cells. This medium was collected every two days. To remove cell debris from the medium, the collected medium was first centrifuged at 4.000 rpm for 30 min and then at 30.000 rpm for 30–60 min. The supernatant was passed through a 0.22 µm diameter filter. Then, it was filled into ultracentrifuge tubes and the caps of the tubes were melted with a capping machine and closed. It was centrifuged at 90.000 rpm for 7 h in an ultracentrifuge. The supernatant was poured and the collapsed pellet was dissolved in 200 µL of phosphate buffer (pH 7.4). Total protein was measured by spectrophotometer in the UV region (260 nm) (Cenik et al., 2022).

An electroporator was used to load raloxifene into exosomes. For this, exosome and raloxifene at different concentrations (1–1000 µM) were added to 400 µL of PBS. This prepared mixture was loaded into a 400 µL electroporation cuvette. The protocol specific to HeLa cells was loaded into the electroporator (voltage = 400 V, capacitance = 125 µF, resistance = ∞Ω). The electroporation cuvette was placed in its chamber in the device and loading was performed after the given impulses. The samples taken from the cuvette were transferred to microfuge tubes (Isolab GmbH, Eschau, Germany). The samples were precipitated with ultracentrifuge at 90.000 rpm for 3 h to remove the drug that was not loaded on the exosome in the medium. The supernatant was poured and the pellet was dissolved in PBS. It was stored at −20°C until use (Cenik et al., 2022).

Loading doses of raloxifene on exosomes were calculated spectrophotometrically as earlier reported by Nagaraju et al. (2014). After measuring the optical density at 288 nm of 5, 10, 15, 20, and 25 µM doses of raloxifene, the standard graph was drawn. The entrapment efficiency of raloxifene was calculated from the standard graph by taking the optical density of the exosome-drug medium before and after loading (Demirbolat et al., 2022b).

For drug targeting, the Apt13 and Apt20 aptamers, which specifically bind to the L1 capsid protein found on the membrane of HeLa cervical cancer cells, were selected from the literature (Graham and Zarbl, 2012). The sequences of the aptamers are 5′GGGACAGACGGAAGATGAGAATTGTGGGCTTAGTATAGTGAGGTGCGGT-3′ for Apt13 and 5′GGGGAGGGAGACACA GTCATGGAGCAGTTATTAGGGTGTACCGGGTGTAGT-3′ for Apt20.

In aptamer binding to exosomes released from HeLa cells, aptamers were dissolved in tris- ethylenediaminetetraacetate (EDTA) buffer and a modified binding buffer (pH 7.4) was prepared in accordance with earlier protocols (Embark, 2016; Siddiqui and Yuan, 2021). For the binding experiment, optimization studies were performed using 1–500 µg protein-containing exosomes and 1–1000 nM aptamers. Eppendorf tubes were incubated at 37°C for 0–240 min. The amount of single-stranded DNA (ssDNA) in the samples was measured in the nanodrop and the amount of aptamer bound was calculated.

The binding affinities of aptamer-bound exosomes (Exo-Apt-Ral) to HeLa and L929 cells were demonstrated by fluorescence microscopy. For this, carboxyfluorescein-labeled aptamers were used. 1 × 10⁶ HeLa and L929 cells were seeded in 6-well plates. After 24 h of incubation for the cells to adhere to the plate base, FAM-labeled aptamer-bound exosomes (100 µg) were added to the medium. After 12 h of incubation, the cells were imaged under a fluorescent microscope (Graham and Zarbl, 2012).

HeLa cells were seeded in 12-well plates at 1 × 10⁵ cells per well. The cells were incubated in a carbon dioxide oven for 24 h to adhere to the bottom of the plate. The cells were incubated for 24 h with the half maximal inhibitory concentration (IC₅₀) dose of Exo-Apt-Ral formulation. After incubation, PKH67 exosome dye was added to the medium and after 60 min of incubation, images of cells were taken under a fluorescent microscope.

The release profile of raloxifene was carried out by dialysis bag diffusion procedure. After the dialysis bag (MW 12 kDa) was absorbed in the phosphate buffered saline (PBS), pH 7.4 overnight, a specific amount of Exo-Apt-Ral was put into the dialysis bag. The dialysis bags were closed from both sides, and were placed in bakers including 100 mL of PBS of pH 7.4, and incubated in a water bath set at 37°C ± 0.1°C. The system was repeatedly stirred at 100 rpm. At prearranged time periods, 1 mL of the media was withdrawn and refilled with fresh release medium at the same temperature. Samples were measured with a UV spectrophotometer, and the concentration of each sample was assayed. All the release studies were carried through in triplicate and the mean values with standard deviation were calculated (Demirbolat et al., 2022a).

HeLa cells were added in 96-well plates at a density of 1 × 10⁴. After the cells adhered to the bottom of the plate, 1,10,100 and 1000 µM doses of raloxifene and Exo-Apt-Ral formulations were added to the wells. After 24 h of incubation, 10 µL of 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT) dye was added to each well. After 4 h of incubation, the medium was discharged and 100 µL of DMSO was added to all wells. The absorbance was read at 570 nm (Erdoğan et al., 2021).

HeLa cells were seeded in 12-well plates at a density of 1 × 10⁵. After 24 h of incubation, a line was drawn in the middle of each well with a pipette tip. The IC₅₀ doses of raloxifene and Exo-Apt-Ral were added to each well. Images of cells were taken at 0, 24, 48, and 72 h. The migration amount was measured with the ImageJ program (Paşa et al., 2021).

HeLa cells were seeded in 12-well plates at 1 × 10⁵ cells per well. The cells were incubated in a carbon dioxide oven for 24 h to adhere to the bottom of the plate. The cells were incubated for 24 h with the IC₅₀ dose of the raloxifene and the Exo-Apt-Ral formulation. After incubation, lysotracker DND-26 dye (50 nM) was supplied to the medium and after 30 min of incubation, images of cells were taken under a fluorescent microscope and changes in lysosomal functions were observed (Ye et al., 2018).

HeLa cells were seeded in 6-well plates with the 1 × 10⁶ cell density. They were incubated for 24 h for cells to adhere. The IC₅₀ doses of raloxifene and Exo-Apt-Ral were added to each well. After 24 h of incubation, cells were removed with 200 µL of trypsin and transferred to microfuge tubes. After centrifugation at 1200 rpm for 5 min, the supernatant was discarded. The cells were dissolved in 100 µL of medium and 100 µL of annexin V reagent was added. After 20 min of incubation in the dark, apoptosis was measured with the Muse Cell Analyzer (Türk et al., 2020).

This experiment was performed to investigate the safety of raloxifene and Exo-Apt-Ral formulations for intravenous administration. Freshly collected venous blood was taken into tubes containing EDTA. Tubes were centrifuged at 1000 × g for 15 min. Precipitated erythrocytes were washed 3 times with pH 7.4 phosphate buffer. To prepare a 10% erythrocyte solution, 100 µL of red blood cells were taken and 900 µL of phosphate buffer (pH 7.4) was added. 50 µL of this prepared erythrocyte solution was taken and 450 µL of raloxifene and Exo-Apt-Ral were added at concentrations of (1, 10, 100, and 1000 µM). NaCl (0.9%) was used as the negative control and distilled water was used as a positive control. Microfuge tubes were incubated at 37°C for 30 min and then centrifuged at 1000 × g for 15 min. The absorbance of the color formed by taking 100 µL from the supernatant was measured at 540 nm (Demirbolat et al., 2022b).

HeLa cells were seeded in 6-well plates at 1 × 10⁶ cells per well. The cells were incubated in a carbon dioxide oven for 24 h to adhere to the bottom of the plate. Cells were incubated with IC₅₀ doses of raloxifene and Exo-Apt-Ral formulations for 24 h. After 24 h of incubation, the cells were removed and the cells were lysed by adding 350 µL of 2X sample loading buffer. Using solutions of pH 6.8, 0.5 M Tris, pH 8.8 1.5 M Tris, 30% acrylamide-bisacrylamide, 10% ammonium persulfate, 10% sodium dodecyl sulfate (SDS) and TEMED, 4% stacking, and 12% separation gel were poured. The samples containing 5–10 µg of protein were loaded onto a gel and run with a running buffer (5 mM Tris, 38.4 mM glycine, 1% SDS) at 100 V constant current for 1–2 h. After electrophoresis, immunoblotting was performed with a semi-dry system to transfer the protein bands in the gel to the polyvinylidene fluoride (PVDF) membrane. The gel was carefully placed on the membrane, and a nitrocellulose layer wetted with a transfer buffer was placed on it again. In a semi-dry transfer device (Biorad Transblot Turbo), blotting was performed at 25 V and 1 Amp constant current for 30 min. After transferring the proteins to the PVDF membrane, the membrane was washed three times with TBST (20 mM Tris, 154 mM NaCl, 0.1% Tween 20) and blocked with 2.5% BSA for 2 h. After washing three times with TBST, the PVDF membrane was incubated with the following primary antibodies BCL-2-associated X protein (Bax, Santa Cruz SC7480), BCL-2 (Santa Cruz SC7382), LC3AB (Cell Signaling 12741), Beclin-1 (Cell Signaling 3495), and B-actin (Santa Cruz SC47778) at 4°C overnight on an orbital shaker. After incubation, the membrane was washed three times with TBST for 10 min. Subsequently, the membrane was incubated with horse radish peroxidase-conjugated secondary antibody for 2 h at room temperature, the membrane was incubated with chemiluminescence reagent (ECL, Santa Cruz) for 1 min in the dark. The images of the bands were taken with the imaging system (Syngene GBOX). Densitometric analysis of the bands was done with the ImageJ program (Paşa et al., 2021).

The total RNA content was isolated from 5 × 10⁶ of HeLa cells as previously described (Paşa et al., 2021). 1 μg total RNA was reverse transcribed as the template for cDNA synthesis using a high-capacity cDNA Reverse Transcription Kit (Applied Biosytems Waltham, USA). Quantitative real-time PCR (qRT-PCR) was carried out with Bcl-2, Bax, LC3B, Beclin-1, and GAPDH primers. The primer sequences are given here: Bcl-2: 5′-GACAGAAGATCATGCCGTCC-3′ (forward), 5′-GGTACCAATGGCACTTCAAG-3′ (reverse); Bax: 5′-GCCCTTTTGCTTCAGGGTTT-3′ (forward), 5′-TCCAATGTCCAGCCCATGAT-3′ (reverse); LC3B: 5′-ACTTTGTTGTTTGGCAGAAGC-3′ (forward), 5′-TTTGTCCCGAGCCTTCATT-3′ (reverse); Beclin1: 5′-CGGAAACCATTCATATCTGG AG-3′ (forward), 5′-TCCCAGAAAAACCGCAAC-3′ (reverse); GAPDH: 5′-TGCACCACC AACTGCTTAGC-3′ (forward), 5′-GGCATGGACTGTGGTCATGAG-3′ (reverse). 100 nanograms of cDNA were amplified by Sybr Green PCR Master Mix (Applied Biosystem) on the ABI StepOne Plus detection system. The program parameter for amplification was 95°C for 10 min, then 40 cycles of 95°C for 15 s, and 60°C for 1 min. The qPCR data were resolved using StepOne Software v2.3 (Applied Biosystems, Foster City, CA), and the genes of interest were normalized to the corresponding GAPDH results. Data plotted as fold change vs. control (Erdogan et al., 2019).

Data from three independent experiments are presented as mean ± SD. Variations between groups were assessed by the One Way ANOVA test. Statistical analysis was carried out with GraphPad Prism version 7.0 software. Statistical importance was considered as follows: *, p < 0.05; **, p < 0.01; ***, p < 0.001.

The expression of CD9, CD63 and CD81 proteins expressed in the membrane of the HeLa cervical cancer cells exosomes isolated by the ultracentrifugation method was demonstrated by western blotting. The expression of CD9, CD63, and CD81 proteins in cell lysates was less pronounced than the expression of CD9, CD63, and CD81 proteins in the exosome, according to the amount of equally loaded protein (Fig. 1A). Dimensional analysis of exosomes isolated from HeLa cervical cancer cells was performed with transmission electron microscopy (TEM) and zeta sizer. Dimensions of the Exo-Apt-Ral formulation were analyzed with a zeta-sizer. According to the zeta results, the size of the pure exosome was measured as 66 ± 12 nm, while the size of the Exo-Apt-Ral was 120 ± 21 nm (Fig. 1B). When the TEM image was examined, spherical structures of approximately 52.34 and 116.1 nm for empty exosome and raloxifene loaded exosome were observed, respectively (Figs. 1C and 1D).

The entrapment efficiency of raloxifene was analyzed by the spectrophotometric method. The linear regression equation of raloxifene was y = 43.03x − 1.6011, r2 = 0.998 (Suppl. Fig. S1). The entrapment efficiency of raloxifene was calculated to be 72% ± 2.15%.

The release profile of raloxifene in the Exo-Apt-Ral formulation in PBS at different pH is shown in Fig. 2D. The prepared formulation released more raloxifene at pH = 5.0. In PBS at pH = 5.0, pH = 7.4, and pH = 8.2, raloxifene was found to be maximally released at approximately 40 h.

The binding levels of Apt13 and Apt20 aptamers to the exosomes, which specifically bind to the L1 capsid protein in the membrane of HeLa cervical cancer cells, were measured spectrophotometrically. While Apt13 and Apt20 aptamers seemed to bind maximally for up to 1 h depending on the concentration, the Apt20 aptamer retained its binding over time. For this reason, Apt20 aptamer was preferentially selected in formulation studies of raloxifene because it is more stable than Apt13 (Figs. 2A and 2B).

To demonstrate that Exo-Apt-Ral specifically binds to HeLa cells, a raloxifene-loaded exosome coupled with fluorescently labeled Apt20 aptamer was used. The Exo-Apt-Ral formulation entering the cells after incubation with Exo-Apt-Ral HeLa cells has been shown (Fig. 2C) with a fluorescent microscope. L929 cells were used as a negative control. Fluorescently labeled formulations in HeLa cells are indicated by the tip of the arrows.

The exosome uptake behavior of HeLa cells was determined by PKH67 exosome staining (Fig. 3A). Observations of an increase in green fluorescence and lumen formation supported the uptake of external exosomes into cells.

The changes in cell morphology after the treatment of raloxifene and Exo-Apt-Ral formulations to HeLa cells were examined with an inverted microscope. It was determined that the cells in the Exo-Apt-Ral treated groups moved away from each other and their cell size decreased compared to the cells in the control group (Fig. 3B). The 24-h cytotoxic effects of 1, 10, 100, and 1000 µM doses of raloxifene on HeLa and L929 cells were investigated by the MTT method. The IC₅₀ values of cells treated with raloxifene were determined as 22.19 ± 1.37 µM and 107.40 ± 8.81 µM for HeLa and L929, respectively (Fig. 3C). It was determined that when Exo-Apt-Ral was incubated with 22.19 µM of raloxifene, cell viability was significantly reduced compared to cells treated with empty exosomes (p < 0.001) (Fig. 3D).

The formed wounds at 0 h and the healing progression at 24, 48, and 72 h in HeLa cells are shown in Fig. 4A. In cells treated with raloxifene, the migration rate decreased to 55.84% after 72 h (p < 0.001). In cells treated with Exo-Apt-Ral, the migration rate decreased to 46.14% after 72 h (p < 0.001) (Fig. 4B).

An in vitro hemolysis experiment was carried out to investigate whether the prepared formulation is safe for intravenous administration. As a result of the hemolysis experiment (Suppl. Fig. S2A), raloxifene and Exo-Apt-Ral (200 μg/mL) significantly reduced the hemolysis rates of red blood cells compared to the positive control (p < 0.001) (Suppl. Fig. S2B).

The apoptotic cell rates in HeLa cells were determined with annexin-V binding. The percentages of apoptotic cell rates were increased while the percentages of live cell rates were decreased after treatment with Exo-Apt-Ral formulations (Figs. 5A and 5B).

To determine the apoptotic effect of the Exo-Apt-Ral formulation on HeLa cells, the expression of anti-apoptotic BCL-2 and pro-apoptotic Bax proteins were examined by western blotting (Fig. 5C). The expression level of the pro-apoptotic protein Bax was dramatically increased (Fig. 5D), whereas the expression of the anti-apoptotic protein Bcl-2 was reduced (Fig. 5E) after treating HeLa cells with raloxifene and Exo-Apt-Ral. Further qPCR assessment revealed that the expression of the anti-apoptotic BCL-2 gene in HeLa cells decreased while that of the pro-apoptotic Bax increased compared to the control group after raloxifene and Exo-Apt-Ral treatments (Figs. 5F and 5G).

The tendency of raloxifene and Exo-Apt-Ral formulation to accumulate in the lysosomes of HeLa cervical cancer cells was examined by lysotracker DND-26 staining. When the images taken at 40× magnification were examined under a fluorescent microscope, it was seen that the cells treated with Exo-Apt-Ral appeared to have more fluorescence than the cells treated with raloxifene (Fig. 6A). Also, the intensity of the lysotracker green color of HeLa cells treated with exosomes, raloxifene, and Exo-Apt-Ral formulations was more compared with the control group cells. The lysotracker green color intensity increased in cells treated with only exosomes compared to the control group (p < 0.01). Likewise, it was determined that the intensity of lysotracker increased in cells treated with raloxifene and Exo-Apt-Ral (p < 0.001) (Fig. 6B).

LC3AB and Beclin-1 protein expression was examined by western blotting to determine the autophagic effects of the Exo-Apt-Ral formulation on HeLa cells (Fig. 6C). The expression level of Beclin-1 and LC3AB increased in HeLa cells treated with raloxifene and Exo-Apt-Ral compared to the control group (Figs. 6D and 6E).

In addition, there was an increase in the expressions of autophagic genes Beclin-1 and LC3B in HeLa cells treated with raloxifene and Exo-Apt-Ral when compared to the control group (Figs. 6F and 6G).