*2.4.* SRO*, PCA, and LYC Bioaccessibility*

As previously mentioned, therapeutic applications of active principles from plants, such as SRO oil or curcumine, are hindered by their poor solubility in aqueous medium like the digestive fluids. Although their solubility during the digestive process is slightly improved by the emulsifying activity of bile salts, new technologies aimed at improving lipophilic active principle bioaccessibility are needed. To investigate Lipomatrix-based formulation (LBF) performance compared to the commercial formulations, we exposed a single dose of each formulation to in vitro digestion procedure mimicking human adulthood and evaluated the total amount of active principles and the apparent bioaccessible fraction released from its matrix. The apparent bioaccessible fraction includes the portion available to be absorbed. LBF has similar SRO apparent bioaccessibility of CF1 and CF2 formulations, with a good emulsifying effect of Lipomatrix technology (Table 3 and Figure 8). Lipomatrix emulsifying efficiency is similar to CF2, which contains soy lecithin emulsifier. CF1 has a lower emulsifying efficiency than the others do since it does not contain any specific emulsifier in the formulation. CF1 emulsion is only due to the presence of bile salts in the intestinal compartment. Bile acts, to some extent, as a surfactant, helping to emulsify lipids and lipophilic molecules and increasing their absorption.


**Table 3.** Total amount and apparent bioaccessible fraction of SRO for the three tested formulations, expressed as a percentage of the total SRO in a single dose (*n* = 3).

**Figure 8.** Apparent bioaccessibility of SRO contained in LBF and the two commercial formulations, CF1 and CF2. (*n* = 3).

In addition to SRO, LBF contains PCA and LYC as active principles. As expected from the gastric resistance test, Lipomatrix technology preserves PCA from degradation, as indicated from their total amount and apparent bioaccessible fraction after digestion (Table 4). Protective effect was also observed on LYC, which is preserved from digestion (Table 4).

**Table 4.** Total amount and bioaccessible fraction of PCA and LYC, expressed as a percentage of the total FlowensTM and total lycopene in a single dose of LBF. (*n* = 3).


#### *2.5. Impact of Digested Formulations on Intestinal Epithelium Viability*

Therapeutic formulations must respond to safety criteria. Taking into consideration their dose and posology, formulations should not negatively affect patients. In particular, damages to the intestinal epithelium must be carefully avoided. Before measuring apparent permeability of the three formulations, the impact of digested formulations on intestinal epithelium viability and integrity was assessed. To this aim, intestinal monolayers were exposed to increasing concentrations of the three formulations, and dose-response curves were obtained (Figure 9). From these curves, half maximal effective concentrations (EC50) were calculated (Table 5). As emerged from dose-responses curves and EC50 values, LBF is the safest formulation.

**Table 5.** Half maximal effective concentrations (EC50) from dose-response curves.


**Figure 9.** Impact of formulations on intestinal mucosa viability evaluated by determining dose-response curves on formulation concentrations. \* Highest non-toxic concentration. (*n* = 3).

#### *2.6.* SRO*, PCA, and LYC Absorption Rate*

Based on the impact of digested formulations on intestinal epithelium viability and posology (2 capsule/day LBF, 1 capsule/day CF1 and CF2), we set experiments for determining SRO, PCA, and LYC absorption rate. In vitro intestinal epithelia were exposed to the digested formulations for 3 h, and SRO (as fatty acids), PCA and LYC were measured in both apical (lumen) and basolateral (serosal) chambers. Absorption rate was then calculated and expressed as percentage of absorption.

PCA and LYC values in the basolateral compartment were below the detection limits (<0.005 mg/mL and <0.001 mg/mL respectively), as expected considering the low amount and titer of LYC (1.33%) and PCA (2.45%) in LBF. Concerning SRO, its absorption rate is higher in LBF than CF1 and CF2 (Table 6), though this difference is considered to be not statistically significant with respect to CF1.

**Table 6.** SRO absorption rate expressed as a percentage of absorption ± standard error (SE) (*n* = 3).


Despite apparent bioaccessibility values, the highest absorption rate of LBF suggests that the Lipomatrix technology supports the bioaccessible form of SRO more than the other two formulations (Figure 10).

**Figure 10.** Absorption rate of the three formulations: LBF, CF1, and CF2. (*n* = 3).

#### *2.7. Impact of Digested Formulations on Intestinal Mucosa Viability and Integrity*

After exposure of intestinal epithelia to digested formulations, Caco-2 monolayer viability and barrier integrity were analyzed. As expected, no viability reduction was observed during absorption rate experiments, while a slight but significant increase in apparent permeability (Papp) was observed for all the three digested formulations (Figure 11).

**Figure 11.** Cell vitality (**A**) and apparent permeability (Papp) (**B**) of intestinal epithelium exposed to digestive fluids (DF; control) and digested formulations (*n* = 3). LBF is composed of SRO, PAC, and lycopene bioaccessible fractions as reported in Tables 5 and 6. CF1 and CF2 represent the bioaccessible fractions as reported in Table 3.

Increase of the absorption rate parallels to a reduction of intestinal epithelia trans-epithelial electrical resistance (TEER) after exposure (Figure 12). Both digested formulations and digestive fluids reduced TEER transiently, since they fully recovered to pre-treatment values.

**Figure 12.** TEER values recorded before the treatment (Pre-treatment), after exposure to digested fluids (DF) or formulations (Post-treatment) at the bioaccessible concentrations, and upon 24 and 48 h recovery. Values are expressed as percentage of the pre-treatment TEER value (*n* = 3).

#### *2.8. Cytotoxic Effect of Diclofenac and Permeable Fractions on Prostatic Epithelium Model*

SRO is one of the most popular natural treatments for treating HPH as mentioned. The permeable fractions were tested on prostatic epithelium in vitro model to evaluate their efficacy against inflammation process. Before performing efficacy tests, cytotoxicity of the anti-inflammatory positive control Diclofenac and bioaccessible fractions on LNCaP was measured.

As shown in Figure 13, diclofenac significantly decreases prostatic cell viability at a concentration of 80 μg/mL after 6 h exposure, with a highest non-toxic concentration of 32 μg/mL. Conversely, no effect on cell vitality was observed on LNCaP cells after treatment with bioaccessible fractions (Figure 14).

**Figure 13.** Effect of diclofenac on LNCaP prostatic cells vitality. \* *p* < 0.05 (*n* = 3).

**Figure 14.** Effect on LNCaP prostatic cells vitality after 6 h exposure to SRO. (*n* = 3).

#### *2.9. Prostate-Specific Anti-Inflammatory Activity of Permeable Fractions*

In recent years, inflammation has been recognized as the main phenomenon responsible for the onset of HPH, a noncancerous increase in size of the prostate, leading to the appearance of bothersome symptoms, such as frequent urination, difficult urination, weak stream, inability to urinate, and loss of bladder control. Prostatic-specific anti-inflammatory activity of bioaccessible fractions was evaluated on the in vitro prostatic epithelium model based on LNCaP cells, by measuring expression levels of pro-inflammatory cytokines IL-1β and TNF-α.

As shown in Figure 15, bioaccessible fractions significantly reduce the production of IL-1β compared to the control. In particular, the higher amount of SRO of LBF compared to the two commercially available formulations, presents the strongest effect, with an 85% reduction in IL-1β production.

**Figure 15.** Production profile of IL-1β following treatment of LNCaP prostatic cells with bioavailable fractions: LBF: 22% SRO; CF1: 18.8% SRO; CF2: 3.2% SRO (Table 6). Diclofenac was used at 32 μg/mL. CM: Conditioned medium. \* *p* < 0.05 (*n* = 3).

Interestingly, no TNF-α reduction was observed for the other considered pro-inflammatory cytokine, TNF-α (Figure 16).

**Figure 16.** Production profile of TNF-α following treatment of LNCaP prostatic cells with bioavailable fractions: LBF: 22% SRO; CF1: 18.8% SRO; CF2: 3.2% SRO (Table 6). Diclofenac was used at 32 μg/mL. CM: Conditioned medium. \* *p* < 0.05 (*n* = 3).

Conversely, to IL-1β, the production of TNF-α significantly increases upon exposure to LBF and CF1 bioaccessible fractions. No effect was observed for CF2. This peculiar trend could be explained considering the pro-apoptotic activity of SRO. Indeed, TNF-α is a cytokine known to be involved in the apoptotic process. Silvestri and colleagues [30] demonstrated that SRO extract induces apoptosis in LNCaP cells. To test this hypothesis, we evaluated the pro-apoptotic effect of SRO absorbable fractions on the prostatic epithelium model in both uninflamed and inflamed conditions (Figure 17).

**Figure 17.** Pro-apoptotic activity of bioavailable fractions on LNCaP prostatic cells: LBF: 22% SRO; CF1: 18.8% SRO; CF2: 3.2% SRO (Table 6). Diclofenac was used at 32 μg/mL. CM: Conditioned medium. Samples are normalized on no CM control. \* *p* < 0.05 (*n* = 3).

As expected, pro-apoptotic activity induced by bioaccessible fractions well correlates with TNF-α expression, meaning a role of SRO in inducing apoptosis against inflamed and tumor prostatic cells. The apparent contrast between LNCaP prostatic cells vitality (Figure 14) and apoptosis (Figure 17) results could be explained by an early-apoptosis phenomenon. During early apoptosis phenomenon, indeed, cells retain their vitality. Activation of early apoptosis cascade leads to cell dead at later times, suggesting a decrease in LNCaP prostatic cells viability following prolonged exposure to SRO. While no effects were observed after 6 h incubation (Figure 18A), cell viability significantly decreases after 24 h exposure (Figure 18B).

**Figure 18.** Effect on LNCaP prostatic cells vitality after 6 h (**A**) and 24 h (**B**) exposure to SRO. \* *p* < 0.05 (*n* = 3).

#### *2.10. Activity of Bioaccessible Fractions on PSA Secretion*

Prostate-Specific Antigen (PSA) is considered the main serum marker for the progression of prostate cancer [31]. A decrease in PSA secretion, following treatment with bioaccessible fractions, indicates a potential therapeutic effect of the formulations. To verify the effect of the bioaccessible formulations on PSA secretion, LNCaP hormone-sensitive cell line was used. Compared to control conditions, PSA secretion decreased in cells treated with LBF and CF1 bioaccessible fractions, while no differences were observed in cells treated with CF2 formulation (Figure 19).

**Figure 19.** PSA secretion in LNCaP prostatic cells treated with bio-accessible fraction of the different formulations. \* *p* < 0.05 (*n* = 3).

#### *2.11. Smooth Muscle Myorelaxing Activity*

The decrease in epithelial-to-stromatic tissue ratio is a well-known marker of benign prostatic hypeplasia (BPH) development. In the prostate, the stromal component is mainly composed of smooth muscle tissue, which is normally contracted in response to adrenergic stimulation. As a result, urethra lumen is reduced and urination made difficult. To explore the potential effect of tested formulations on this BPH symptom, the myorelaxing activity of the bioaccessible fractions on smooth muscles was evaluated using WPMY-1 myofibroblast in vitro model.

As shown in Figure 20, no muscle relaxation was induced by CF2 bioaccessible fraction. Conversely, the bioaccessible fractions of LBF and CF1 showed a significant myorelaxing activity on smooth muscles, with LBF presenting the highest myorelaxing activity.

**Figure 20.** Myorelaxing (**A**) and contraction inhibitory (**B**) activity of the bioaccessible fraction of the different formulations on an in vitro model of smooth muscle. \* *p* < 0.05 (*n* = 3).

#### **3. Materials and Methods**

#### *3.1. Materials*

*Serenoa repens* oil 85% fatty acids GC was purchased from Naturex S.p.a. (Caronno Pertusella, VA, Italy). Ascorbyl palmitate and lecithin were purchased from A.C.E.F. (Fiorenzuola D'Arda, PC, Italy). Mono- and diglycerides of fatty acids were from BASF Italia S.p.a. (Cesano Maderno, MB, Italy). Mannitol, synthetic amorphous silica, magnesium stearate and sodium bicarbonate were purchased from Giusto Faravelli S.p.a. (Milano, Italy). Sodium chloride was purchased from Fagron Italia S.r.l. (Quarto Inferiore, BO, Italy). Thaurocholic acid was purchased from Shangai T and W Pharmaceuticals Co. (Shangai, China), maleic acid and lipase from porcine pancreas were purchased from Sigma-Aldrich (St Louis, MO, USA), FlowensTM and lycopene (powder titrated at min. 6% *w*/*w*) were purchased from Naturex.

Caco-2 human colon adenocarcinoma cell line (ATCC® HTB-37™), LNCaP androgen-sensitive human prostate adenocarcinoma cell line (ATCC® CRL-1740™), WPMY-1 human myofibroblast stromal cell line (ATCC® CRL-2854™) and THP-1 (ATCC® TIB-202™) were purchased from ATCC (Manassas, VA, USA). High glucose Dulbecco's Modified Eagle Medium (DMEM), Roswell Park Memorial Institute (RPMI) 1640 Medium, Hanks' Balanced Salt Saline (HBSS), non-essential amino acids (NEAA), L-glutamine, penicillin-streptomycin mix, lipolysaccharide (LPS), diclofenac, dihydrotestosterone (DHT), phorbol 12-myristate 13-acetate (PMA), proantocyanidine A, B1, and B2 standards and Lucifer Yellow (LY) were purchased from Sigma-Aldrich (St Louis, MO, USA). Foetal bovine serum (FBS) was purchased from Euroclone (Milan, IT). Interleukin 1β (IL-1β), Tumor Necrosis Factor α (TNF-α) and prostate-specific antigen (PSA) ELISA kit were purchased was purchased from R

and D Systems, PeproTech (London, UK) and Abcam (Cambridge, UK), respectively. Cell contraction assay was purchased from Cell Biolabs (San Diego, CA, USA). Transwell® insert were purchased from Millipore (Burlington, MA, USA). CellTiter 96® AQueous One Solution Cell Proliferation Assay (MTS) and Apo-ONE® Homogeneous Caspase-3/7 Assay were purchased from Promega (Madison, WI, USA). C18 cromatographic columns were purchased from Agilent (Santa Clara, CA, USA). Cyclo-oxygenase (COX) activity assay kit was purchased from Cayman Chemicals (Ann Arbor, MI, USA).

#### *3.2. Methods*

#### 3.2.1. Preparation of Lipomatrix Powder

#### 3.2.1.1. Preparation of Lipomatrix Powder with SRO, Flowens, and Lycopene

MDGFA and ASP were dry mixed in a beaker and the mixture was melted at temperature of 80 ◦C under mechanical stirring (IKA® RTC Basic Staufen, Germany). The resulting oily liquid phase was added of SRO 85% fatty acids GC and Lycopene (powder titrated at min 6% *w*/*w*), maintaining the temperature at 75 ◦C. Mannitol and FlowensTM, previously mixed and cooled down in fridge at 15 ◦C, were put in a planetary mixer (Kenwood KMX750RD, De Longhi Treviso, Italia, Kenwood group) and the hot oily phase was added on it in small continuous additions under mixing so to induce an instant and uniform solidification of the lipid phase onto the cold powder. The formed composite powder was cooled down at room temperature for 24 h. Finally, the granules were passed through a 1.5 mm mesh net connected to the mechanical sieve (Vasquali, Marchesini group, Cerro Maggiore (MI), ITALY). The resulting powder was added of synthetic amorphous silica and magnesium stearate to improve powder flowability. Type 0 animal gelatin capsules were filled with the powder using a manual encapsulator (MultiGel) so that the content of any capsule corresponded to 160 mg of SRO, 125 mg of Flowens and 5 mg of lycopene (powder titrated at 6% *w*/*w*).

#### 3.2.1.2. Preparation of Lipomatrix Powder with SRO Alone

The same method of Section 3.2.1.1. in which no FlowensTM and lycopene powder have been introduced.

#### 3.2.2. Flow Property

Flow through an orifice was measured according to Ph. Eur. Chapter 2.9.36. The test was performed using a metal truncated cone Flowability Tester (Flowability Tester BEP Auto Copley Scientific, Colwick, Nottingham) with different size of the orifice. The time it took for 100 g of powder to pass through 25–15–10 mm diameter orifices was measured in triplicate.

#### 3.2.3. In Vitro Emulsification Test

In vitro emulsification in FaSSIF-V2 was performed as follows: a 100 mL glass beaker was filled with 50 mL of simulated gastric fluid (GSF, according to Eur. Pharmcopea) and placed at 37 ◦C ± 0.5 ◦C (IKA® RTC Basic). FaSSIF-V2 was chosen as model of enteric fluid to assess the real capability of Lipomatrix to create soluble micellar forms in the duodenum without any interfering molecules coming from ingested foods such as higher Lecithin concentration than FaSSIF-V2 and oleic acid (OA) esters and salts (FeSSIF-V2) [32]. GSF was realized from demineralized water (inverse osmosis process) added of dilute solution of HCl up to pH = 1. The quantity of Lipomatrix powder corresponding to 4 capsules as described in Section 3.2.1.2. containing a total amount of 640 mg of SRO, has been dispersed in the afore described GSF at 37 ◦C under moderate magnetic stirring (200 rpm, IKA® RTC Basic) for 90 min simulating gastric transit time. After that, the suspended Lipomatrix powder was decanted, the GSF was removed and replaced with 50 mL of FaSSIF-V2. FaSSIF-V2, composition was the following: sodium chloride 68.62 mM, thaurocholic acid 3 mM, lecithin 0.2 mM, maleic acid 19.12 mM, lipase from porcine pancreas 100 units/mL, pH 7.20 adjusted with sodium bicarbonate. We preferred to use sodium bicarbonate instead of phosphate buffer, as described in the consolidated FaSSIF-V2 model [33], to simulate the secretion of sodium bicarbonate by pancreas [34]. This model showed furthermore to be useful to verify the pH lowering behavior of Lipomatrix ascribable to ASP ionization, so confirming the postulated mechanism of action. The temperature of the media was set at 37 ± 0.5 ◦C by means of a thermostatic probe. The powder was maintained in the dissolution medium for 60 min under magnetic stirring (200 rpm, IKA® RTC Basic). During the test, pH probe was introduced in the FaSSIF-V2 dispersion and pH was monitored continuously (Sension + PH3, Hach). Since ASP embedded in Lipoamtrix ionizes in FaSSIF-V2 (pKa < pH), ionization leads to a pH reduction so each 15 min pH was adjusted with sodium bicarbonate to maintain pH not less than 7.00 and simulate the continuous pancreatic secretion of sodium bicarbonate. The test has been assessed with 50 mL of both GSF and FaSSIF-V2 to set the models close to the average volumes, of physiologic fasted condition of gastro-enteric tract in humans [35,36].
