3.3.3. Liposome Size Change

Flow cytometry was used to analyze the liposomal size change during the stability study (Section 3.3.1), because the large number of natural particles in biological solutions prevents reliable measurements with DLS. The large-angle light scattering (LALS) measurements of the liposome populations in vitreous and plasma samples with Apogee flow cytometry were converted into hydrodynamic diameters using a calibration curve (Figure 4B; Figure S4A, Supplementary Materials). Only the liposome population was gated based on the fluorescence, removing the naturally occurring particle aggregates from the signal (Figure S4B–E, Supplementary Materials). The coefficients of variance were 70–100 for all samples throughout the experiment without showing a clear tendency for increase or decrease. *Pharmaceutics* **2020**, *12*, x FOR PEER REVIEW 11 of 25

**Figure 4.** (**A**) Liposome formulations and their color-coded groups in corona experiments, combined with data from earlier studies [28,29]. (**B**) Large-angle light scattering measurements (LALS) for size change during storage in plasma and vitreous samples. (**C**) Surface plasmon resonance (SPR) measurements for hard (HC) and soft corona (SC) thickness in undiluted plasma with individual data points, mean, and SD**.** (**D**) Corona thicknesses in undiluted vitreous. **Figure 4.** (**A**) Liposome formulations and their color-coded groups in corona experiments, combined with data from earlier studies [28,29]. (**B**) Large-angle light scattering measurements (LALS) for size change during storage in plasma and vitreous samples. (**C**) Surface plasmon resonance (SPR) measurements for hard (HC) and soft corona (SC) thickness in undiluted plasma with individual data points, mean, and SD. (**D**) Corona thicknesses in undiluted vitreous.

3.4.2. Hard and Soft Corona Composition The HC and SC corona fractions from Figure 4C,D were eluted for high-resolution proteomics analysis with nanoliquid chromatography tandem mass spectrometry. Sufficient chromatographic separation was obtained with a one-hour linear gradient for plasma, while the vitreous samples required three-hour gradients. Figure 5 shows the relative enrichment and hierarchical clustering of The diameter of the PEG-coated liposomes remained unchanged throughout the one-week exposure period in vitreous. In plasma, a gradual size increase up to a diameter of 155 nm was observed. The relative size increase was even higher for the HA-coated liposomes in plasma (270 nm at two days, 318 nm at one week), whereas, in the vitreous sample, the HA-coated liposomes reached a diameter of 150 nm in the beginning and remained at that level until the end of the experiment.

#### the top 60 most abundant proteins in human plasma after median normalization. A list and a heatmap of the 178 proteins identified in at least three samples are provided in Figure S5 (Supplementary *3.4. Protein Corona Formation*

### Materials). 3.4.1. Hard and Soft Corona Thickness

Surface plasmon resonance (SPR) measurements under dynamic conditions in undiluted human plasma and porcine vitreous samples were carried out to determine the hard (HC) and soft corona (SC) thicknesses and separate these fractions for proteomics analysis of their composition [28]. Due to the higher number of the samples as a result of data integration from earlier studies, additional formulation numbers (F1–7) and color codes based on charge and size are used (Figure 4A). In plasma, the HA liposomes formed a 5.4 ± 1.2 nm hard corona and a 9.3 ± 1.6 nm soft corona, which is on par with the 4.9 ± 0.7 and 10.6 ± 1.6 nm SC obtained for the PEG liposome (Figure 4C). The uncoated 50 nm anionic liposome without PEG (F6) formed a very thin HC and SC in a consistent trend with the uncoated

100 nm neutral liposome (F1). The HA liposome without ICG (F5) formed a thinner corona than its ICG-containing counterpart.

In the vitreous sample, the HA liposome formed a thicker 1.6 ± 0.3 nm HC and 2.7 ± 0.5 nm SC than the PEG liposome (1.1 ± 0.2 nm and 1.6 ± 0.3 nm, respectively). Both 100 nm formulations formed thinner coronas than the 50 nm anionic liposomes regardless of surface charge or polymer coating, and the SC thickness suggests that it comprised only a few additional loosely interacting proteins in addition to the HC. The protein concentrations of the undiluted porcine vitreous and human plasma samples determined using the bicinchonic acid (BCA) assay were 1.5 mg/mL and 75.8 mg/mL, respectively. These correspond to dynamic exposures of 0.25 mL at 375 µg of protein per replicate in the vitreous sample (1.4 pg/nm<sup>2</sup> /min) and at least 0.75 mL at 57 mg if protein per replicate in plasma (57 pg/nm<sup>2</sup> /min). The unfunctional formulations without ICG were not tested in the vitreous sample.
