**5. Conclusions**

We compared ultrastructure of epithelial cells, which possess hepatocyte-type of polarization (HepG2) and columnar polarization (HEK 293) cultured in 2D- and 3D-forms (monolayer and spheroids). Monolayer HepG2 cells showed hepatic epithelia-specific morphological features, while HEK293 cells in monolayer did not show signs of epithelial tissue.

Cultivation of HepG2 and HEK293 cells on non-adhesive conditions led to formation of spheroids, a common feature of which was formation of spheroid outer surface by basal cell plasma membrane. This finding should be taken in account in experiments on drug delivery.

To examine interaction of different NPs with cells in monolayer and spheroids, we synthesized AuNPs (12.0 ± 0.1 nm in diameter, TEM data) and covered them with BSA and PEI. Values of hydrodynamic diameter were 17.4 ± 0.4; 35.9 ± 0.5 and 125.9 ± 2.8 nm for AuNPs, AuBSA-NPs and AuPEI-NPs, and Z-potential (net charge) values were −33.6 ± 2.0; −35.7 ± 1.8 and 39.9 ± 1.3 mV, respectively.

TEM study revealed inability of AuNPs uptake by HepG2 cells both in monolayer and spheroid form, while AuPEI-NPs and AuBSA-NPs were actively internalized via clathrin-mediated endocytosis, and this is an evidence for selectivity of HepG2 cells in respect to different NPs. At the same time, AuNPs actively penetrated the HEK293 cells in monolayer and spheroids.

Our data showed that the presence of a protein or polymer corona affects the behavior of NPs in endosomes after their endocytosis: AuBSA-NPs remained dispersed, while AuPEI-NPs were fused and formed aggregates. These features could influence drug release from an endosome, and so deserve attention when studying efficacy of drug delivery.

We did not observe appreciable distinctions in mechanisms of all studied NPs interaction with HepG2 and HEK293 cells in monolayer and spheroids. This observation is important for planning of different experiments because allow choosing cell monolayers or spheroids as experimental model, depending on a task. Thus, to know a depth of drug penetration, only spheroids are suitable, and for studies of drug-cell interaction mechanisms monolayers can be used. Undoubtedly, the use of multicellular spheroids as a 3D model of tumors and organs provides additional and "useful" opportunities for studying nanomedicine preparations in comparison with the cell monolayer.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2079-4991/10/10/2040/s1, Figure S1: "Openings" on surface of HepG2 spheroids in SEM and semithin sections, Figure S2: Pseudosinusoids in ultrathin sections of HepG2 spheroids, Figure S3: Periphery of HepG2 spheroid, Figure S4: Accumulation of AuBSA-NPs in HepG2 cells.

**Author Contributions:** Conceptualization, E.R.; methodology, B.C. and A.T.; validation, I.P. and A.E.; investigation, B.C., J.P., A.E. and A.T.; data curation, E.R.; writing, E.R., B.C. and J.P.; original draft preparation, E.R.; writing—review and editing, E.R.; visualization, B.C., J.P. and A.T.; supervision, E.R.; project administration, I.P.; funding acquisition, E.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Russian Science Foundation, grant number 19-15-00217; preparation of the nanoparticles were funded by Russian State Funded Budget Project of ICBFM SB RAS # VI.62.1.4, 0309-2016-0004.

**Conflicts of Interest:** The authors declare no conflict of interest.
