*2.2. Delivery Barriers for IP Administration*

An emerging drug administration strategy for ovarian cancer is the IP route. Thus, it might feasible to develop an IP delivery system for siRNAs as well. Considering that ovarian cancer metastasis very frequently targets the peritoneum/omentum, knowledge of the biology of this serous membrane is crucial to designing optimal siRNA delivery systems. The surface of the peritoneum contains a single layer of mesothelial cells (Figure 2) with a cuboidal shape and adipocytes (reviewed as reported by Sarfarazi et al [39]). Tight junctions and plasmalemma interdigitations connect mesothelial cells. Underneath the mesothelial cells there is basal lamina, which separates the cells from the connective tissue layer. This last layer, composed of collagen fibers, elastin, and fibroblasts, is characterized by the presence of irregular oval openings with diameters ranging from 3 to 15 μm, distributed irregularly in the layer. Dispersed on the peritoneum surface are lymphatic lacunae, lined by endothelial cells. The lacunae are in connection with lymphatic capillaries. In lacunae, MS and endothelial cells are close to each other and share a common basal lamina. The diameter of lacuna aperture is 1.8–6 μm in most

animal species, and 3.2–3.6 μm in humans. Blood vessels are abundantly scattered in the connective tissue layer [39].

Following IP administration, delivery systems with a size similar to the diameter of the lymphatic lacuna are preferentially drained from the peritoneum via lacunae. Smaller delivery systems (<10 nm diameter) are mostly cleared via the blood, as they can more efficiently penetrate blood capillary walls. Moreover, the higher flow rate (100–500-fold) in blood vessels compared to lymphatic capillaries favors the diffusion of small molecules into the blood [40]. Thus, the retention of delivery systems in the peritoneal cavity can be modulated by their size. Delivery systems larger than the diameter of the aperture of the lymphatic lacunae will persist longer in the peritoneal cavity; those inferior in size will be predominantly cleared via the lacunar route, while the very small ones (nm range) will be mostly drained by blood vessels. Peritoneal persistence can be also improved by using positively charged delivery systems that can easily interact with the negatively charged membrane of MS [41]. Prolonged peritoneal persistence can present an advantage, as ovarian cancer metastasis typically colonizes the mesothelial cell layer without affecting the basal lamina [42]. Obviously, tumor cell membrane crossing, the presence of cytosolic nucleases, and endosome entrapment still remain obstacles for siRNA action in the case of IP administration.

#### **3. Strategies to Optimize siRNA Delivery**

The fragile nature of siRNAs in the cellular and extracellular environment prevents their use as naked molecules. To improve their stability and thus effectiveness, it is possible to chemically modify their structure to make them more resistant to degradation [2,33]. An additional and complementary strategy consists of the combination of siRNAs with delivery vectors able to preserve their integrity and drive them to the target cell [2,33]. Additionally, electroporation can be also used [43]. Often, but not always, the two strategies are combined together. However, as chemical modifications can impair siRNA effectiveness, their employment should be carefully considered. Here, we focus on a brief description of the most commonly used delivery materials, i.e., those based on polymers and lipids (Figure 5).

#### *3.1. Polymer-Based Delivery Vectors*

Polymers have been frequently used as siRNA delivery systems due to their versatility, low cost of production/isolation, and biocompatibility. The most commonly used polymers for siRNA delivery are polyethylenimine (PEI), polyethylenglycole (PEG), polycaprolactone (PCL), poly(lactic-*co*-glycolic acid) (PLGA), chitosan (CH), and hyaluronic acid (HA). These polymers are frequently combined with each other and/or with other molecules to create delivery particles with optimized delivery features. Regardless of the chemical nature of the polymer, they need to have positively charged resides to bind the negatively charged siRNAs via electrostatic interactions. Moreover, the resulting polymer–siRNA complex should have an appropriate surface charge: excessively strong cationic complexes can lead to nonspecific cellular uptake [44–46] and nonspecific membrane disruption. On the other hand, anionic particles that are too strong can have problems interacting with the negatively charged cellular membrane, thus resulting in inefficient intracellular delivery of siRNA.

Repeated units composed of amine groups and two carbon aliphatic CH2–CH2 spacers (Figure 5A) form the structure of PEI. It can exist in both linear and branched forms: in the first form, it contains all secondary amines, while in the second form it contains primary, secondary, and tertiary amino groups. The amino groups confer a cationic nature to PEI, suitable for binding siRNAs. High-molecular-weight PEI displays a superior ability to carry and deliver siRNAs compared to low-molecular-weight [47]. However, high-molecular-weight PEI is more cytotoxic than low-molecular-weight PEI, a problem that can be attenuated by the addition of hydrophilic and hydrophobic moieties or cell-/tissue-specific ligands [48]. Finally, due to its ability to promote the "proton sponge effect", PEI favors siRNA endosomal escape [49].

**Figure 5.** Structure of polymer- and lipid-based delivery materials, (**A**) polyethylenimine (PEI), (**B**) polyethylene glycol (PEG), (**C**) polycaprolactone (PCL), (**D**) poly(lactic-*co*-glycolic acid) (PLGA), (**E**) chitosan (CH), (**F**) hyaluronic acid (HA), and (**G**) lipid particles.

The structure of PEG is represented by the formula H−(O−CH2−CH2)*n*−OH (Figure 5B). Being substantially non-toxic [50] and soluble in water, PEG can be used in many biological applications. PEG can reduce the toxicity of other delivery materials and, because of this property, it is often added (PEGylation) to various delivery particles [51]. Moreover, it can be also used as a linker for specific ligands on the surface of delivery particles [52].

PCL is obtained by the ring opening polymerization of the cyclic monomer ε-caprolactone (Figure 5C) [53]. It is biodegradable and, in the field of drug delivery, it is typically used to tune the physical, chemical, and mechanical properties of different materials when co-polymerized with polymers such as PEG. Its main limitation in siRNA delivery is the lack of cationic groups able to interact with the negative siRNA backbone [54]. Thus, PCL is often combined with PEI and/or PEG.

PLGA is a copolymer of poly(d, l-lactic acid; LA) and poly(glycolic acid; GA) linked together through ester linkages [54] (Figure 5D). PLGA, biodegradable into its original monomers LA and GA, has been approved for human use by the US Food and Drug administration. PLGA gained attention as an siRNA delivery material due to its favorable safety profile and sustained-release characteristics. However, one of the weaknesses of PLGA is the modest siRNA loading capacity due to the repulsion between the anionic acid groups in PLGA and the negative phosphate backbone of siRNA. To minimize this problem, PLGA is often combined with cationic molecules such as dioleyltrimethylammoniumpropane (DOTAP) or PEI.

It is possible to commercially obtain CH by deacetylation of chitin, a substance contained in the exoskeletons of crustaceans and the cell walls of fungi, [55] (Figure 5E). Chitosan is a linear polysaccharide with a carbohydrate backbone with two types of repeating residues, 2-amino-2-deoxy-glucose (glucosamine) and 2-*N*-acetyl-2-deoxy-glucose (*N*-glucosamine), linked by a (1-4)-β-glycosidic linkage. The amino groups confer to CH a positive charge, which allows binding with the negatively charged phosphate groups of siRNAs. CH conjugation with PEI and PEG can bypass the problems of low solubility and delivery effectiveness. Alternatively, it is possible to tune the degree of acetylation and modify the molecular weight [56].

HA, a polymer of natural origin, is a linear polysaccharide formed by glucuronic acid and N-acetylglucosamine units linked via alternating –1,4 and –1,3 glycosidic bonds (Figure 5F). HA is biocompatible, biodegradable, and characterized by low toxicity. In the field of siRNA delivery, it is often used in combination with other cationic polymers. Indeed, its anionic nature attenuates positive charges, thus resulting in reduced toxicity and improved stability of the complex [57]. Finally, the presence of hyaluronic acid receptors (CD44, see Section 4) in many healthy and diseased human (cancer) tissues makes HA particularly suited as a targeting agent [58].

#### *3.2. Lipid-Based Delivery Vectors*

Liposomes are spherical vesicles formed by concentric lipid bilayers (Figure 5G) with a hydrophilic region localized in the core [59,60]. Because of this structure, liposomes can entrap hydrophilic molecules in the water compartment and hydrophobic molecules in the lipid layers. Frequently, liposomes are made up of neutral (cholesterol; Chol) or zwitterionic (dioleoylphosphatidylethanolamine; DOPE) lipids, giving origin at the physiological pH to neutral liposomes [60]. While neutral liposomes have good biocompatibility and favorable pharmacokinetics, they display poor interactions with negatively charged siRNA. An alternative to the use of neutral lipids is the employment of positively charged lipids, such as 3β[*N*-(*N*',*N*'-dimethylaminoethane)-carbamoyl]DC-cholesterol, 1,2-dioleoyl-3-trimethylammonium propane (DOTMA) [60]. Cationic liposomes are composed by positively charged lipids within a matrix of neutral lipids, required for cellular internalization. Cationic liposomes allow easy and effective loading of negative molecules such as siRNA. Moreover, they facilitate the endosomal escape of siRNA, promoting the fusion between liposomal and endosomal bi-layers. Despite being suitable for siRNA delivery to the cells, cationic liposomes have some drawbacks [59]. For example, they are more toxic than neutral liposomes. They also have a low "fusogenicity" attitude, i.e., they are less effective in interacting with cell membranes to allow cell uptake. To attenuate this limitation, cationic lipids are often combined with non-cationic lipids such as DOPE, which display enhanced fusogenic activity. Another limitation of cationic liposomes is represented by the fact that, when applied to the blood stream, they are cleared by the reticulo-endothelial system. Thus, their effectiveness in siRNA delivery can be significantly diminished. To minimize this problem, cationic liposomes are often equipped with shielding molecules on their surface such as a hydrophilic polymer like PEG. However, once they reach the target cell, the PEG has to be removed to allow efficient siRNA delivery and to permit efficient endosome escape.

#### **4. Targeting OC: Delivery Systems and siRNA Targets**

The ideal therapeutic approach to OC should be able to selectively deliver the drug (siRNA in our case) to tumor cells, leaving the normal tissues as untouched as possible. This goal can be approached with two complementary strategies: the first is to equip the siRNA delivery system with molecules able to specifically target tumor cells, and the second is to use siRNAs against OC-specific oncogenes. The combination of these two strategies should give the highest specificity level possible.

#### *4.1. OC-Targeting Molecules*

So far, different antigens present on the surface of OC cells have been considered suitable for OC targeting (Table 1).


**Table 1.** Specific surface antigens on OC cells.

Epidermal growth factor receptor (EGFR) is highly expressed in OC [61], as well as in other tumor cells. An extracellular ligand-binding domain, a transmembrane domain, and a cytoplasmic domain containing the tyrosine kinase region [62] characterize this surface antigen. Following ligand binding to EGFR, receptor auto-transphosphorylation in the tyrosine kinase region occurs, triggering a series of signaling events. These, in turn, promote cell proliferation, apoptosis inhibition, cell invasion, and stimulation of neovascularization [63]. Because of its pro-proliferative effects, EGFR has also been considered as a drug target [64].

Erythropoietin-producing hepatocellular receptor A2 (EphA2) belongs to a large family of receptor tyrosine kinases mainly involved in the regulation of cell proliferation and migration [65]. The extracellular domain contains a ligand-binding domain, a Sushi domain, an epidermal growth factor (EGF)-like domain, and two fibronectin type-III repeats. A transmembrane domain and a tyrosine kinase domain follow the extracellular domain. EphA2 is overexpressed in ovarian cancer [66] as well as in various other solid tumors, such as breast, prostate, pancreas, glioblastoma, neck, renal, lung, melanoma, bladder, gastric esophageal, colorectal, and cervical cancers. In OC, the N-terminals of EphA2 are processed by membrane-type 1 matrix metalloproteinase, triggering ligand-independent signal activation, which in turn promotes cell motility, invasion, and metastasis [66].

The folic acid (FA) receptor (FR) is a 38 kDa glycosyl-phosphatidylinositol membrane-anchored glycoprotein [67]. Under normal conditions, FR is present only in some polarized epithelia and it is localized to the apical/luminal cell surface. In tumor cells of epithelial origin, it is often overexpressed and it loses its polarized nature, covering the entire cell surface. FR acts as a transporter of folate into the cells. The fact that folates are essential for the biosynthesis of purines and thymidine, necessary for DNA synthesis, methylation, and repair, explains why actively growing tumor cells need higher FR compared to normal cells. Notably, FR overexpression is present in more than 80% of HGSOCs [68]. Moreover, increased expression level of FR correlates with OC stage, poor response to chemotherapy, and worse survival.

In addition to the above-mentioned OC-targeting antigens, other antigens associated with ovarian cancer stem cells (OCSCs) are gaining great interest. The discovery of quiescent OCSCs with the ability to escape chemotherapy and to regenerate OCs has directed attention towards the possibility of specifically targeting OCSCs [69]. CD44 is a surface transmembrane glycoprotein that acts as a receptor for different molecules, including hyaluronic acid (HA). CD44 influences the expression levels of genes related to cellular differentiation and cell–matrix adhesion. It is one of the most common OCSC surface markers, used to identify this kind of cell. Despite this, CD44 does not seem to have relevant prognostic value in OC [70]. Monoclonal antibodies and siRNAs [71] directed against CD44 have already been tested. Another OCSC-related antigen is CD133, a glycosylated transmembrane protein with prognostic value in OC. CD133 modulates several pathways involved in the control of cancer stemness and metastasis. Notably, even though CD133-positive cells represent a minority of OC cells, its toxin-mediated targeting effectively downregulated OC growth in in vitro and in vivo models [72]. Further potential targeting antigens for OC are CD117 and CD24. CD117, also known as c-kit, is a receptor tyrosine kinase devoted to the promotion of cell survival, metabolism, and differentiation. In OC, poor disease-free survival is directly associated with CD117 levels [73]. While CD117 targeting

in vitro successfully downregulated OC cell survival [74], only modest effects were observed in a phase II clinical trial [75]. Finally, another potential targeting antigen is represented by CD24, highly expressed in many cancers with a percent positivity in OC of about 70% [76].

#### *4.2. siRNA Targets in OC*

Due to their mechanism of action, siRNAs are particularly suited to targeting gene products of which the exuberant expression is responsible for increased cell growth, cell survival, migration, angiogenesis, and drug resistance (Tables 2–4). Notably, a single gene product often controls many of the above cellular phenotypes.


**Table 2.** Molecular targets implicated in OC cell growth/migration.




