*2.3. Harvest of EVs*

To harvest EVs, human breast carcinoma BT-474 cells and human neuroblastoma SK-N-MC cells were selected as the EV secreting cell lines and purchased from the Korean Cell Line Bank (Seoul, Korea). BT-474 cells were cultured in an RPMI-1640 medium containing 1% penicillin-streptomycin and 10% EV-depleted FBS in a 5% CO<sup>2</sup> atmosphere at 37 ◦C. SK-N-MC cells were cultured in the DMEM medium containing 1% penicillin-streptomycin and 10% EV-depleted FBS in a 5% CO<sup>2</sup> atmosphere at 37 ◦C. The cell culture medium was centrifuged at 2000 *g* for 20 min at 4 ◦C and additionally centrifuged at 10,000 *g* for 30 min to eliminate large cell debris and dead cells. Subsequently, the supernatant was again ultracentrifuged at 100,000 *g* at 4 ◦C for 70 min to separate EVs pellets. The obtained pellets were washed using fresh PBS (150 mM, pH 7.4) and then ultracentrifuged at 100,000 *g* at 4 ◦C for 70 min. These purified EVs were stored at −80 ◦C, after being suspended in fresh PBS (150 mM, pH 7.4) [3,30,32–34]. In addition, the suspended EVs concentration was analyzed using Nanosight (LM10, Malvern Instruments, Malvern, UK) with NTA 2.3 software [3,35].

#### *2.4. Preparation of EV Samples*

Sonication was performed to incorporate HDEA (or HDOC) and DOX to EVs [3,30,36]. HDEA (300 µg, or Ce6-labeled HDEA) or HDOC (200 µg, or Ce6-labeled HDOC) dissolved in DMSO (0.1 mL) containing DOX (400 µg) were mixed with EVs (200 µg) suspended in PBS (10 mL, 150 mM, pH 7.4) at 25 ◦C, and the solution was sonicated using a tip sonicator, vcx-130 with cv-18 (Sonics, Newtown, CT, USA) with a 30% amplitude for 30 s. This sonication process was repeated 6 times at 3 min intervals. The obtained solution was incubated at 37 ◦C for 60 min to recover the EVs. The filtration method using 0.22 µm membranes was performed to remove HDEA (or HDOC) or DOX aggregates. Ultracentrifugation at 100,000 *g* at 4 ◦C for 70 min was performed to remove free HDEA (or HDOC) and DOX to yield HDEA@EVs or HDOC@EVs [3,30,36]. In addition, DOX@EVs with DOX and without HDEA and HDOC were prepared following the same procedure as described above.

#### *2.5. Measurement of Loading Contents*

The concentration of HDEA or HDOC (with fluorescent Ce6 dye) in the EVs was measured using a fluorescence spectrofluorometer (RF-5301PC, Shimadzu, Kyoto, Japan) at λex of 450 nm and λem of 670 nm using a DMSO/PBS solution (90/10 vol.%) [3,30,31]. The HDEA or HDOC loading content (%) was calculated as the weight percentage of HDEA or HDOC in the EVs. The concentration of DOX entrapped in the EVs was measured using a fluorescence spectrofluorometer at λex of 470 nm and λem of 592 nm using a DMSO/PBS solution (90/10 vol.%). The DOX loading content (%) was calculated as the weight percentage of DOX in the EVs [3,30,31].

#### *2.6. Characterization of EV Samples*

The particle size and zeta potential of the EV samples at pH 7.4 or 6.5 were measured using a Zetasizer 3000 instrument (Malvern Instruments, Malvern, UK) [3,30,37–40]. The morphologies of the EV samples at pH 7.4 or 6.5 were analyzed using a transmission electron microscope (Talos L120C, FEI, Hillsboro, OR, USA) [3,31,37,38].

#### *2.7. In Vitro DOX Release Test*

The EV samples (equivalent to DOX of 100 µg/mL) in PBS (3 mL, 150 mM, pH 7.4) were added to a dialysis membrane (Spectra/Por® MWCO 10 K) and immersed in fresh PBS (15 mL, 150 mM, pH 7.4 or 6.5) [3,30,31,40]. A DOX release test was conducted using a mechanical shaker (100 rpm) at 37 ◦C [3,30,31,40]. The external PBS of the dialysis membrane was extracted and replaced with fresh PBS at the specified time point. The amount of DOX released from the EV samples was measured using a fluorescence spectrofluorometer at λex of 470 nm and λem of 592 [3,27].
