Influence of supercritical fluid extraction parameters in preparation of black chokeberry extracts on total phenolic content and cellular viability

Abstract Black chokeberries (Aronia melanocarpa), deciduous shrubs of the Rosaceae family, are native to northeastern North America. Chokeberry fruits are cultivated to make jellies, juices, and wines. Black chokeberry pulp is rich in phenolics and other antioxidants and exhibits potential for health and food packaging benefits. Chokeberries’ in vitro antioxidant activity is among the highest values of all berries, though chokeberry extraction techniques frequently employ environmentally unfavorable solvents or are time‐inefficient. Batch extraction of antioxidants from chokeberry pomace using supercritical carbon dioxide with an ethanol modifier was used to examine the effects of plant loading, pressure, temperature, and percent ethanol by weight. Effects on total phenolic content (TPC) and the optimal conditions for extractions within these ranges are reported. Multivariate analyses reveal the following relationships of extraction conditions upon TPC: Temperature is directly proportional, percent ethanol by weight is inversely proportional, and chokeberry loads can be increased to enhance antioxidant activity, though not through a linear relationship. In studies involving 0.5 g plant load, the conditions 24.9MPa, 68°C, 90wt‐% CO2, and 10wt‐% ethanol generated the highest TPC value, 3.42 ± 0.20 mg gallic acid equivalents/gram chokeberry. Chokeberry extracts displayed antiproliferative effects on the SKBr3 breast cancer line and the 52KO MEF line, although TPC was not predictive of cellular responses. HPLC‐MS data suggest cyanidin hexose and cyanidin pentose compounds as well as quercetin deoxyhexose–hexose as components of the more favorable extraction product that reflected a significant decrease in viability for the extract in comparison with ethanol control in the SKBr3 breast cancer line.


| INTRODUC TI ON
Nutraceuticals, substances at the junction of "nutrients" and "pharmaceuticals" (DeFelice, 1995), have been used for years in treating disease. There has been a resurgence in popularity of dietary supplements, herbal remedies, and natural treatments due to factors such as a mainstream shift in preference toward natural over synthetic remedies, a desire to find more cost-efficient treatments, and a need to find alternatives to medications with undesirable side effects (Nicoletti, 2012). Indeed, many proven medical treatments arose from natural plant-derived products such as digoxin in treating atrial fibrillation, Hypericum perforatum,in treating depression, quinine in treating malaria, salicylates in treating fevers, and taxol in treating cancers (Aronson, 2017). Commercial production and distribution of nutraceuticals is a growing industry (Nicoletti, 2012), but there is growing inquiry for testing of natural compounds at all levels of drug development (Santini et al., 2018).
Thus, there is an expanding market for cost-efficient, environmentally friendly means of harvesting vast amounts of plant materials for nutraceutical benefit. Vital to this process is proper scientific assessment of true benefits and risks of supplements and plant-derived treatments.
Plant polyphenols are a diverse group of compounds presenting a variety of human health benefits. Indeed, many health benefits of plants come from phenolic compounds including anthocyanins such as cyanidins, flavonoids such as quercetin, catechin, and resveratrol, proanathocyanidins, and phenolic acids. In the mixtures of which they are found in nature, there can be synergistic benefits of antioxidant, antimutagenic, antimicrobial, anticarcinogenic, and anti-inflammatory properties, which are greater than those of the constituent parts (Juranić & Žižak, 2005;Katalinić et al., 2010;Rasouli, Farzaei, & Khodarahmi, 2017). The most prevalent subtype of polyphenols found in plants is flavonoids. Flavonoids, commonly attributed to the bitter taste and the yellow, orange, and red hues of fruits and vegetables, possess medical benefits such as anti-inflammatory, antioxidant, anticancer, antiplatelet, antiviral, antiallergic, cardio-protective, and cancer-protective properties (Tanwar & Modgil, 2012).
Berries are particularly enriched in flavonoids and tend to have some of the highest overall amounts of phenolic compounds of all fruits. In berries, phenolic contents correlate highly with overall antioxidant activity. Anthocyanins, catechins, flavonols, and proanthocyanins are the predominant flavonoids in berries (Macheix, Fleuriet, & Billot, 1990). In a comparison of chokeberry, blueberry, cranberry, and lingonberry extracted with 80% acetone and 2% formic acid, chokeberries had the highest antioxidant values as assessed by three different estimators of antioxidant capacity: the oxygen radical absorbance capacity assay (ORAC), anthocyanin content through the pH differentiation method, and total phenolic content assays (Zheng & Wang, 2003). ORAC measurements reviewed by Kulling and Rawel suggested that chokeberries had the highest antioxidant capacities as measured by that assay in comparison with thirteen other berries, oranges, red and white grapes, and apples (Kulling & Rawel, 2008).
The Aronia berry, commonly called chokeberry, black apple berry, and rowanberry, is native to northeastern North America and the Great Lakes Region, and in the 1900s, this berry was introduced to Europe and Russia. The genus Aronia can be further categorized into the species melanocarpa, arbutifolia, and the hybrid prunifollia.
Chokeberries are used as a component of fruit juice blends, jellies, teas, and wines, and as food coloring (Kulling & Rawel, 2008).
Chokeberry's antioxidant activity has been attributed to a variety of in vivo mechanisms, including not only the traditional radical scavenging but also recharging antioxidant enzymes, inhibiting oxidant enzymes, preventing the formation of reactive oxygen and nitrogen compounds (ROS and NOS), and participating in signal transduction in response to oxidative stress (Denev et al., 2012).
Antioxidants hold great potential as nutraceuticals. Cellular tests suggest that antioxidants may help to combat destructive processes that result from accumulation of oxidative compounds because they block free-radical damage and engage in signaling cascades as studied in various cell lines. Indeed, animal models and human clinical trials show multiple medical benefits of chokeberry administration in various forms (Chrubasik, Li, & Chrubasik, 2010;Denev et al., 2012;Kulling & Rawel, 2008). First, chokeberry juice may have value against autoimmune and inflammatory processes as it appears to decrease reactive oxygen species production and induced apoptosis in human neutrophils (Zielińska-Przyjemska, Olejnik, Dobrowolska-Zachwieja, & Grajek, 2007). Second, cellular studies involving complete or enriched chokeberry extracts show promising cellular death responses in several cancer cell lines including HT-29 colon cancer cells (Olsson, Gustavsson, Andersson, Nilsson, & Duan, 2004;Zhao, Giusti, Malik, Moyer, & Magnuson, 2004), HeLa cervical cancer cells (Rugina et al., 2012), and MCF 7 breast cancer cells (Olsson et al., 2004). Third, chokeberry research suggests potential benefits to the cardiovascular system of rats and humans where diet supplementation with chokeberry products led to benefits in cholesterol profiles, blood pressure, and in cardiovascular endothelial cell restoration (Skoczynska et al., 2007). Fourth, anthocyanins isolated from chokeberries were able to decrease toxicity due to cadmium and carbon tetrachloride exposure and reduced amounts of heavy metals in the kidney and liver of rats (Kowalczyk et al., 2003;Valcheva-Kuzmanova, Borisova, Galunska, Krasnaliev, & Belcheva, 2004 (Dai & Mumper, 2010). In the ultrasonic-assisted extractions (UAE) of dried chokeberries performed by d' Alessandro et al., it was reported that the phenolic yield of black chokeberry dramatically increased within the first hour of extraction with increasing temperature from 20 to 80°C (D'Alessandro, Dimitrov, Vauchel, & Nikov, 2014;d'Alessandro, Kriaa, Nikov, & Dimitrov, 2012 Supercritical fluid extractions present an alternative method to extract medicinally relevant materials from plants. Supercritical fluids exhibit characteristics of liquids and gases. They exist above both the pressure and temperature conditions required for a substance to have a distinct phase boundary between the liquid and gas, and they are able to extract compounds faster than traditional methods (Sairam, Ghosh, Jena, Rao, & Banji, 2012). Recently, an extraction was employed on chokeberries using supercritical carbon dioxide with an ethanol modifier (Wozniak, Marszalek, Skapska, & Jedrzejczak, 2017), which used a partial factorial design where temperature, pressure, and ethanol concentration were varied, but solvent density was allowed to change with operating conditions. In contrast with the Wozniak studies, the parameters tested in the study described herein employed lower ethanol concentrations where solvent density was held constant, and pressure was allowed to vary with operating conditions. In addition for this paper, for some of the conditions selected, the solvent system was a binary supercritical fluid mixture of carbon dioxide with ethanol.
It is important to explore the best extraction conditions to harvest medicinal compounds from chokeberries. In this paper, we investigate a relatively nontoxic, batch extraction method to extract compounds from chokeberries by employing a solvent of supercritical carbon dioxide and an ethanol modifier (used to increase the dielectric constant (Schmidt & Moldover, 2003)) with different extraction parameters than previously employed by Wozniak et al. (2017). In this study, the variables temperature, percent ethanol, and mass of plant load are examined to determine the most ideal extraction conditions. The combination of temperature and pressure conditions that are above the carbon dioxide and ethanol mixture critical point may be identified using published experimental data and correlations (Pohler & Kiran, 1997). The total phenolic content (TPC) assay was employed as a preliminary screening method to compare concentration of probable phenolic antioxidant compounds obtained under varying conditions. Antiproliferative effects were determined for extracts of high, medium, and low TPC values on the SKBr3 breast cancer and fkbp52-deficient mouse embryonic fibroblast (52KO MEF) control cell lines. HPLC-MS analysis was performed to profile the most probable major components of the most antiproliferative extraction products toward SKBr3 breast cancer cells.

| Chemicals
The following reagent grade or greater chemicals were utilized:

| Plant material
Fresh, organically grown black chokeberries, Aronia melanocarpa, were obtained in western New York United States, during the 2017 growing season. Immediately following retrieval, the chokeberries were refrigerated. Subsequently, the chokeberries were destemmed, pressed, and the resulting pomace was frozen. The frozen pomace was then ground in a coffee grinder and stored in a nitrogen-purged container at −20°C until time of use, up to one year later. The ground chokeberry pomace particle size ranged from 20 to 50 mesh.

| Extraction method
Chokeberry pomace was extracted batch-wise in a 24 ml, 2.54 cm outer diameter, 1.93 cm inner diameter, 316 stainless steel test cell capped with Swagelok® fittings using a previously described custom-built batch extraction system (Wenzel et al., 2017). The extraction solvent was supercritical carbon dioxide with ethanol modifier. Temperature, ethanol weight fraction, and chokeberry loading were varied, with an extraction time of 60 min and a target solvent density of 0.76 g/ml. Since solvent density was held constant, the pressure must vary with temperature. Solvent loading was determined using the Peng-Robinson equation of state with Wong-Sandler mixing rules. Depending upon the experimental condition, for each experiment, between 0.25 and 1.5 grams of chokeberry pomace was weighed and placed into the test cell.
Then, a predetermined amount of nitrogen-purged ethanol was placed into the test cell. The test cell was then sealed, connected to the batch extraction system, and heated. Upon reaching the target temperature, carbon dioxide was fed into the test cell to the target pressure; then, the temperature was held at a constant for 60 min. Following this, the test cell was allowed to cool, the gaseous carbon dioxide depressurized, the vessel was opened, and the liquid extract suctioned out. The liquid extract was stored in a nitrogen-purged, double-sealed glass vial at 4°C in the dark.
Prior to use in any assay, extracts were first centrifuged for 5 min at 1,690 g to remove solid residuals without concentrating the extract. Next, each extract was filtered with a vented Millex® 0.22um PVDF filter to remove any remaining suspended particulates for downstream applications of antioxidant and antiproliferative testing.

| Total Phenolic Content/Folin-Ciocalteau Assay
The total phenolic content (TPC) assay or Folin-Ciocalteau assay is an electron transfer-based colorimetric assay which quantifies reduction of a molybdotungstate indicator reagent in response to antioxidant activity of primarily phenolic compounds (Folin & Ciocalteu, 1927;Singleton & Rossi, 1965). Since the majority of antioxidant compounds in plants are phenolics, this assay is often used to estimate electron transfer-based antioxidant activity in plants (Singleton & Rossi, 1965). The modernized version that was employed in this paper uses a 96-well plate format (Ainsworth & Gillespie, 2007). The TPC assay was performed using the same methods as previously published by the team (Wenzel et al., 2017), though for the chokeberry studies, all extracts were diluted twentyfold prior to analysis and assays were performed in duplicate for three to four independent trials. Following the assays, statistical tests and analyses were performed as described in the Extraction Experimental Design section.

| Extraction experimental design and statistical analysis
Two factors, temperature and ethanol content, were evaluated using a 2 2 factorial design with randomization, with 2 replicates for corner points and 3 replicates for the center point. Temperature was varied from 50 to 68°C and ethanol content from 10 to 20 wt-%.
Chokeberry pomace loading for the factorial design was held constant at 0.5 g, and total solvent density was held constant at 0.76 g/ ml. At least four independent analyses of duplicates for TPC were performed for each extraction sample. All experimental results were reported as mean values with corresponding standard deviations of assay measurements. The response for the factorial design, TPC, was evaluated by ANOVA analysis, where a p-value less than 0.05 was considered statistically significant. Statistical analysis was performed using Minitab® version 16.2.0 statistical analysis software.
Additionally, the effect of pomace loading was evaluated at 60°C, a solvent loading of 15 wt-% ethanol, and total solvent density of 0.76 g/ml. Pomace loading was varied from 0.25 to 1.5 g, with each extraction performed in duplicate.

| Alamar Blue® Assay
The alamar Blue® assay, first cited in 1993 for use in mammalian cells (Fields & Lancaster, 1993), is a highly sensitive, inexpensive, relatively nontoxic fluorescent quantification method often used in estimating cellular proliferation. It utilizes the weakly fluorescent property of the primary chemical resazurin and its conversion to highly fluorescent resorufin as an indicator of cellular metabolism by actively respiring cells. (Ahmed, Gogal, & Walsh, 1994;O'Brien, Wilson, Orton, & Pognan, 2000).
SKBr3 or 52KO MEF cells were counted using Trypan blue per manufacturer's instructions and the BioRadTC20 Automated Cell Counter, followed by plating at 10,000 cells/well on Corning® Costar CLS3603 96-well assay plates (black plate, clear bottom with lid, tissue culturedtreated polystyrene). Following cell seeding, plates were incubated for approximately 16 hr at 37°C with 5% CO 2 . Then, media were replaced with new media containing 5% chokeberry extraction treatments, solvent, or media control. Cells incubated for 24 hr at 37°C with 5% CO 2 followed by a wash in phosphate-buffered saline. A 10% alamarBlue® solution was made directly with fresh media and was then added to the wells. Following a four-hour incubation, plate fluorescence was detected

| HPLC-ESI MS Analysis of Chokeberry Extract
Separation and mass/charge analysis of phenol were performed using high performance liquid chromatography (HPLC)-mass spec-  ied from 10 to 20 wt-%. Solvent density was held constant at 0.76 g/ ml whereby the pressure was allowed to vary to maintain constant solvent density. The antioxidant potential for both sets of conditions was estimated with the total phenolic content (TPC) assay. Prior studies of berries show that the total phenolic concentrations correlate highly with overall antioxidant activity (Ga̧siorowski et al., 1997;Kähkönen et al., 1999;Prior et al., 1998;Wu, Gu, Prior, & McKay, 2004 Table 2, as the temperature increases, the total phenolic content assay value increases, which is expected. In addition, the total phenolic content also increases as the ethanol weight percentage decreases. Additionally, the effect of the amount of chokeberry pomace in relation to ethanol loaded in the extraction cell upon total phenolic content assay value was evaluated in Figure 2.

| Effects of supercritical extraction parameters on antioxidant potential of black chokeberry
For the range of 50-68°C, it is notable that higher temperatures favored increased TPC assay yield. This is most likely due to influences of diffusivity; when temperature increases, so does the diffusivity of the extraction solvent. Additionally, as temperature increases, the matrix of materials that make up the chokeberry pomace relaxes, also enabling increased diffusion of compounds. In distinction from the conditions tested by Wozniak et al., density is held constant in the studies of this paper; therefore, pressures are equal to or higher than previously published, and the percent ethanol employed is only in the range of 10%-20% as compared with the range of 20%-80% from the other study (Wozniak et al., 2017).
Due to maintaining a constant density in the studies, pressure did increase with increasing temperature, though it is not a direct variable in the study; the most optimal TPC value was obtained when the pressure was at its highest for the study, 24.9MPa. It is indeed notable that in supercritical fluid extraction of grape seeds and pomace, as well as with chokeberry, typically antioxidant potential will increase with increasing pressure, between 10 and 30 MPa, when total solvent density, carbon dioxide with ethanol, is not held constant. (Ghafoor, Al-Juhaimi, & Choi, 2012;Murga, Ruiz, Beltran, & Cabezas, 2000;Pinelo et al., 2007;Wozniak et al., 2017).
In Figure 1 and From Table 4 and Figure 2, the phenolic content assay value increased with increasing chokeberry load. The proportion of chokeberry to ethanol ranged from 0.071 to 0.429 on a mass basis. These experiments were performed using duplicate extractions, with the exception of the load of 0.5 g, which was in triplicate since it was the center point of the response surface analysis. The antioxidant potential was measured using the total phenolic content assay for three trials in duplicate.
As chokeberry loading increased from 0.25 to 1.5 g, total phenolic content assay value increased from an average of 1.72 ± 0.32 mg GAE/g to 2.78 ± 0.23 mg GAE/g, though the relationship is not linear. While the relationship of these factors was not linear, it is important to note that for the conditions tested, it is possible to produce higher TPC values in chokeberry extracts by increasing the amount of plant matter added to the extraction cell for batch extraction.

| Effects of black chokeberry extracts on cell proliferation
Berries and antioxidant components of berries have strong anticancer properties toward breast cancer cell lines (Aiyer, Warri, Woode, Hilakivi-Clarke, & Clarke, 2012;Olsson et al., 2004). Some of the primary mechanisms of berry antioxidants in hormone-dependent and hormone-independent breast cancer cell lines may include targeting estrogen receptor signaling, targeting receptor tyrosine-protein kinase erbβ-2 [HER-2] signaling, activating apoptosis, interacting with autophagy cascades, and modulating cell cycle regulation (Aiyer et al., 2012). The SKBr3 breast cancer cell line is classified as a human epidermal growth factor receptor 2 (HER2+) expressing cell line; it is deficient in expression of estrogen and progesterone receptors (Mota et al., 2017). Features of HER2 + cell lines are intermediate between luminal cell lines (representative of hormone-responsive, less aggressive cancers) and basal cell lines (representative of hormoneindependent, more aggressive cancers) (Carey et al., 2006;Mota et al., 2017). The primary treatment used for patients expressing a similar molecular background to SKBr3 employs the monoclonal antibody Trastuzumab, but in cases of Trastuzumab resistance, there are few medicinal alternatives other than traditional chemotherapy (Carey et al., 2006;Maher, 2014). Ideally, compounds that show potent anticancer activity will have a higher IC 50 (half maximal inhibitory concentration) in noncancer cells and will show minimal toxicity at the doses used to induce death in cancer cells.
To this end, three chokeberry extracts were tested on both the SBKBr3 breast cancer cell line and the fibroblast cell line 52KO MEF.
A fibroblast line was chosen as a control since fibroblasts are an active component of connective tissue found in the breast. Figure 3 shows the viability of SKBr3 and 52KO MEF cells using the alamar-Blue® assay. It was surprising to note that for the three extracts tested in Figure 3, total phenolic content of chokeberry extracts was not necessarily predictive of antiproliferative activity in SKBr3 and 52KO MEF cells. To test this theory further, extracts could be concentrated down relative to their TPC ratios prior to profiling in cellular studies, so that all extracts could be normalized to their TPC values and direct comparisons could be made. Concentrating extracts could also result in lowered ethanol content used in cellular assays,   Table 6. There is no apparent trend between any of the extraction variables (TPC value, extraction temperature, extraction pressure, and chokeberry load or ethanol percentage) and cellular proliferation.
In summary, it appears that Extract 3 exhibits a profile that warrants further investigation: toxicity to the breast cancer cell line and cyanidin xyloside (a cyanidin pentose) by extracting chokeberry with acetone containing 0.2% formic acid (Zheng & Wang, 2003).
In comparison with the supercritical extraction techniques used by Wozniak et al., which particularly probed anthocyanins, cyanidin-3-galactoside was the major compound, and cyanidin-3-glucoside was a minor component (Wozniak et al., 2017). This supports our finding of a cyanidin hexose. The fragmentation pattern of the mass spectra of the well-resolved major peak at about 20 min indicates a high likelihood that it is a quercetin deoxyhexose-hexose conjugate, which is consistent with the finding by Häkkinen and Auriola (1998 (Häkkinen & Auriola, 1998). Thus, our overall finding of anthocyanins cyanidin hexose and pentose as well as the flavonoid quercetin deoxyhexose-hexose are consistent with findings from others. Our HPLC findings seem to be consistent with literature but do reflect a more simplistic mixture of compounds (Oszmiański & Wojdylo, 2005;Wozniak et al., 2017). This contrast is likely due to the differences in extraction and HPLC preparation methods.
Considering the HPLC profile of Extract 3, it is of no surprise that there were anti-proliferative results for the chokeberry extracts in the breast cancer cell line SKBr3 as seen in Figure 3. The suggested phenolic components of Extract 3, the most potent extract, were cyanidin hexose, cyanidin pentose, and quercetin deoxyhexose-hexose ( of HER2, in breast cancer cells (Aiyer et al., 2012). This information is particularly interesting because the SKBr3 line for which chokeberry exerts major effects relies on HER2 in part for growth; perhaps it is affecting HER2 directly or indirectly.

| CONCLUSIONS
Supercritical carbon dioxide extraction, using an ethanol modifier, is an effective means of extracting antioxidant compounds from chokeberry pomace. In this study, the extraction solvent density was held constant, while the ethanol content and temperature were varied, with pressure varying correspondingly according the relationships between pressure, temperature, and density for compressible fluids. The highest TPC value, 3.38 mg GAE/g chokeberry, was obtained at 68°C and 10% ethanol by weight, representing the highest temperature data point and the lowest percent ethanol by weight employed. Likewise, as chokeberry pomace loading in the extractor was increased, TPC values increased, reaching a maxima at 0.286 g chokeberry/g ethanol. In the alamar Blue®-based cellular profiling of the chokeberry extracts, the extract prepared with 15% ethanol and 62°C caused a 72% decrease in SKBR3 breast cancer cellular proliferation and only a 49% decrease in proliferation in the control fibroblast line compared with live cell control. Comparative HPLC analysis suggested the antiproliferative agents cyanidin hexose, cyandin pentose, and quercetin deoxyhexose-hexose may contribute to the toxicity seen in the cell lines. In the future, mechanistic studies to assess means of cell death and profiling in a broader range of breast and control cell lines could be carried out to further characterize cellular responses.

| E THI C AL RE VIE W
This study does not involve any human or animal testing; there were no human participants from which to receive informed consent while conducting the study. assistance of Erin Kissick, Allison Seeley, and Elijah Ward for their assistance with this work.

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest.