An optimized extraction technique for acetylcholinesterase inhibitors from the Camellia japonica seed cake by using response surface methodology

Abstract

The response surface methodology (RSM) was used to optimize the extraction conditions for the acetylcholinesterase (AchE) inhibitory activity and extraction yield from Camellia japonica seed cake. Predicted values for AchE inhibition and extraction yield were 19.41 and 13.35%, respectively, which are in good agreement with the experimental values from validation, suggesting that RSM may provide a useful tool to optimization processes.

Key words:

• Camellia japonica
• camellia seed cake
• acetylcholinesterase
• Alzheimer’s disease
• response surface methodology

Alzheimer’s disease (AD) is one of the most common neurodegenerative disorders, affecting more than 24 million people worldwide. This devastating dementia is predicted to double every 20 years through 2040.¹⁾ The brains of patients with AD show multiple pathological changes which include neuritic plaques, neurofibrillary tangles, and abnormal neurochemical status.²⁾ Notably, the abnormal neurochemical status, which results from the degeneration of the basal forebrain cholinergic neurons, is known to play an important role in the AD etiology. The resulting lowered levels of acetylcholine (Ach) in the synaptic cleft have led to the formulation of the cholinergic hypothesis of memory dysfunction in AD.³⁾ Despite intensive research effort, there is a dearth of effective therapeutic treatments for AD. Recently, bio-active compounds derived from food sources (e.g. fruits and vegetables) have gained increasing interest because of their beneficial effects on chronic disease prevention (e.g. anti-oxidative, anti-viral, and anti-cancer) as well as their low toxicity.⁴⁻⁶⁾ For instance, many studies have demonstrated that phenolic compounds from plants are potential preventive agents against neurodegenerative disorders including AD.⁷⁾

Camellia japonica is a tea seed plant that is widely utilized as an ornamental plant in many Asian countries.⁸⁾ Derived from the plant’s seed, camellia oil is very similar to olive oil and is characterized by a high content of oleic acid (more than 85%).⁹⁾ However, the camellia seed hull, a residual product from camellia oil processing accounts for more than 60% of the seed (total weight basis), suggesting that fruit hull of camellia constitutes major resource of C. japonica. In addition, it was shown that this seed hull contains various bio-active compounds (e.g. lignin), warranting further investigations for this residual by-product.¹⁰⁾ In order to better utilize this abundant resource and to elucidate its health promoting effects, the response surface methodology (RSM) was used to establish the optimal extraction conditions for C. japonica that exert the highest activity in acetylcholinesterase (AchE) inhibition. Specifically, in the study, extraction temperature (°C), extraction time (min), and ethanol concentration (%), as independent variables, were optimized using RSM.

Defatted seed cake was freeze-dried for the extraction of AchE inhibitors. Briefly, samples were frozen at −80 °C (Model 958, Forma Scientific Inc., Marietta, OH, USA) and then, freeze-dried using a lyophilzer (PVTFD10R, Ilshin Bio Co., Seoul, Republic of Korea). Freeze-dried C. japonica seed cake was finely grounded using a blade-type food grinder (model 203; Krups Co., New York, NY, USA) for 0.5 min and sieved through No. 35 mesh sieve (0.50 mm/inch; Newark wire cloth Co., Newark, NJ, USA) in order to obtaining test materials with maximum uniformity of particle size (<0.50 mm opening). Specifically, once ground, samples were passed through the sieve for 1 min with tapping. All materials retained on the top of the sieve were ground again to produce the smallest particle size as possible. Above procedures were repeated until nearly all material passed through the sieve. In a 125 mL flask, 2 g of sample and 40 mL of ethanol were added and stirred at 150 rpm (C-Ski-II, YS-Tech Co., Seoul, Republic of Korea). Based on the preliminary experiment results, levels of each condition (i.e. extraction temperature (°C), extraction time (min), and ethanol concentration) were determined are summarized in Supplemental Table 1 (see http://dx.doi.org/10.1080/09168451.2014.915723).

The rat pheochromocytoma (PC12) cell line was obtained from the American Type Culture Collection (Manassas, VA, USA). The Roswell Park Memorial Institute 1640 medium, donor horse serum, fetal bovine serum, and antibiotic–antimycotic were purchased from Gibco (Grand Island, NY, USA). All other chemicals used were of analytical grade.

The proximate composition of the camellia seed cake was analyzed using the Association of Official Analytical Chemists method described elsewhere.¹¹⁾ Moisture content was determined as the difference in mass before and after the samples were freeze dried. Freeze-dried samples were used to determine protein, lipid, and ash content. Protein content was determined using the Kjeldahl method with a protein factor of 6.25. Lipid content was measured using a Soxhlet extractor for 16 h. Ash value was determined using a muffle furnace (Model FP189, Dongyang Inc., Seoul, Republic of Korea) at 600 °C until a white ash was obtained (approximately 6 h).

PC12 cells were cultured in 100 mm tissue culture dishes in Roswell Park Memorial Institute 1640 medium supplemented with 10% heat-inactivated horse serum, 5% fetal bovine serum, and 1% antibiotic–antimycotic. An incubator for cell culture was maintained at 37 °C with water saturation and 5% CO₂. Cell culture dishes were passaged when each dish was approximately 80–90% confluent. The media was changed 2–3 times a week. In order to prepare the enzyme source, PC12 cells were homogenized using a homogenizer (Daihan Scientific, Seoul, Republic of Korea) with five volumes of a buffer [10 mM Tris–HCl (pH 7.2) containing 1 M NaCl, 50 mM MgCl₂, and 1% Triton X-100]. Homogenates were then centrifuged at 10,000 × g for 30 min. The supernatant was used as the source of enzyme. All extraction steps were carried out at 4 °C.

The activity of AchE was measured using the slightly modified method, previously reported elsewhere.^12,13) Briefly, the AchE activity was spectrophotometrically monitored; each sample (10 μL) was mixed with the enzyme solution (10 μL) and incubated with an enzyme reaction mixture (70 μL; 0.5 mM acetylthiocholine and 1 mM of 5,5′-dithio-bis(2-nitrobenzoic acid) in a 50 mM sodium phosphate buffer (pH 8.0)) at 37 °C for 15 min. Absorbance was measured at 405 nm using a microtiter plate reader (Model 550, Bio-Rad, Hercules, CA, USA).

The RSM was used to optimize the extraction conditions for the AchE inhibitory activity from C. japonica seed cake. The Box and Behnken design was applied. A fractional three-level, three-factor experimental design with three replicates at the center point was used to investigate the effects of three independent variables (i.e. extraction time (X₁), ethanol concentration (X₂), and temperature (X₃)) on the dependent variables (i.e. AchE inhibition and extraction yield). Each factor was coded at three levels (–1, 0, and 1), and each level was selected based on the results of our preliminary experiments (Supplemental Table 1). The RSM experimental design is summarized in Supplemental Table 2. The complete experimental design consisted of 15 points. Data analysis was used to predict the following second-order polynomial model through the response surface regression procedure of the SAS 9.2 (SAS Institute, Cary, NC, USA):

where Y is a response; β₀, β_i, β_ii, and β_ij are constant coefficients; and X_i are uncoded independent valuables. Regression analysis and analysis of variance were used to assess the model. The Maple software version 7 (Waterloo Maple, Waterloo, ON, Canada) was utilized to create response plots by holding constant one variable of the second-order polynomial equation. The three-dimensional representation of the response surface is the graphical representation of the regression equation, showing the optimum values of the variables at which response is maximized. Data were expressed as the mean ± standard deviation. The statistical significance between groups was calculated and grouped using one-way analysis of variance, followed by the Tukey’s range test (SAS Institute, Cary, NC, USA).

In order to characterize C. japonica, the proximate compositions of camellia seed cake were analyzed. The contents of moisture, protein, fat, ash, and carbohydrates were found to be 4.28 ± 0.02, 10.45 ± 0.07, 9.29 ± 0.05, 2.04 ± 0.02, and 73.94 ± 0.05%, respectively.

The AchE inhibition and extraction yield were measured from 15 sets of variable combinations (Supplemental Table 2) and the data were fitted to a second-order polynomial equation using the response surface regression procedure for all responses investigated, including linear (X₁, X₂, and X₃), interaction (X₁X₂, X₁X₃, and X₂X₃), and quadratic terms (X₁², X₂², and X₃²). The estimated constant coefficient values and the regression models are given in Supplemental Table 3 and Table 1, respectively. It was found that X₁, X₂, and X₃ (i.e. constant coefficient values of extraction time, ethanol concentration, and temperature) are linear (p = 0.0118 for X₁, p = 0.0101 for X₂, and p = 0.0289 for X₃) except for X₃ (p = 0.2584) at Y₂. These were all positive, which indicates a trend of the independent variables on the dependent variables, Y₁ and Y₂. Furthermore, significant interactions between variables were also noted (Supplemental Table 3).

The coefficients of determination (R²) for Y₁ (AchE inhibition) and Y₂ (extraction yield) were 0.902 (p < 0.05) and 0.963 (p < 0.05), respectively. The analysis of variance indicates that both of the predicted models were significant at the 5% level (Supplemental Table 4). Optimum conditions for AchE inhibition and extraction yield and their predicted dependent values are shown in Table 1. Optimum conditions were as follows: 41.17 min of extraction time, ethanol concentration of 71.98%, and extraction temperature of 51.99 °C for AchE inhibition; and 61.35 min of extraction time, ethanol concentration of 61.06%, and extraction temperature of 45.80 °C for extraction yield. Predicted optimized values for AchE inhibition and extraction yield were 19.41 and 13.35%, respectively.

Table 1. Optimized extraction conditions and predicted RSM model.

Response	Optimum conditions			Predicted value (%)	Experimental value (%)
	Time (min)	Ethanol (%)	Temperature (°C)
AchE inhibition	41.17	71.98	51.99	19.41	18.85 ± 1.79
Extraction yield	61.35	61.06	45.80	13.35	13.27 ± 1.48
	Quadratic polynomial model				R^2
AchE inhibition	Y1 = 0.3810 + 0.1688X1 + 0.2557X2 + 0.2444X3 − 0.0006X1^2 − 0.0017X2^2 − 0.0020X3^2 − 0.0007X1X2 − 0.0014X1X3 + 0.0002X2X3				0.9249
Extraction yield	Y2 = 1.9811 + 0.0935X1 + 0.2262X2 + 0.0697X3 − 0.0007X1^2 − 0.0014X2^2 − 0.0004X3^2 − 0.0003X1X2 − 0.0002X1X3 − 0.0007X2X3				0.9067

The second-order polynomial model allowed for prediction of the effects of extraction conditions on the AchE inhibitory activity as well as the extraction yield of the C. japonica seed cake. The response surface plot of the extraction conditions on AchE inhibition and extraction yield are shown in Figs. 1 and 2. As shown in Fig. 1, increased ethanol concentration and extraction temperature led to increased AchE inhibitory activity, up until a threshold level. Beyond this threshold, the inhibition decreased, suggesting that moderate ethanol concentrations and extraction temperatures yield greater AchE inhibitory potential from C. japonica. A similar trend was also observed for the extraction yield, where the maximal yield of C. japonica seed cake was increased up to a threshold level of ethanol concentration and extraction time (Fig. 2).

Fig. 1. The response surface plot of the effect of extraction time (min), ethanol concentration (%), and temperature (°C).

Note: On AchE inhibition under the fixed optimal conditions of (A) extraction time, 41.17 min; (B) ethanol concentration, 71.98%; and (C) temperature, 51.99 °C.

Fig. 2. The response surface plot of the effect of extraction time (min), ethanol concentration (%), and extraction temperature (°C).

Note: On extraction yield (%) under the fixed optimal conditions of (A) extraction time, 61.35 min; (B) ethanol concentration, 61.06%; and (C) temperature, 45.80 °C.

To validate the predicted values, a verification experiment was performed under the optimization conditions obtained from the RSM. The AchE inhibition and extraction yield from the C. japonica seed cake were 18.85 ± 1.79% and 13.27 ± 1.48%, which was in good agreement with the predicted values from the RSM, indicating that the model demonstrated in the present study can be used for optimization of C. japonica seed cake extractions and activity assays.

As mentioned earlier, there are several pathological changes in AD patients, including abnormal neurochemical status.²⁾ In particular, diminished Ach levels in AD brains are often observed; therefore, utilizing AchEIs to replenish Ach levels has been considered as an effective approach to improve hippocampal function.¹⁴⁾ In this regard, exposure to bio-active compounds from food stuffs could provide several benefits for disease prevention, including low toxicity, chronic consumption, and lowered economical cost.¹⁵⁾ In a recent meta-analysis, it was shown that dietary intake of anti-oxidants induces protective effects against AD, supporting the notion that AD can be prevented by means of chronic exposure to bio-active compounds via diet.¹⁶⁾ However, the potency of active compounds from edible plants, including C. japonica seed cake, and their long-term effects are questionable.

There is no study to demonstrate anti-AchE potency of C. japonica seed cake. Recently, Ye et al. reported that sasanqua saponins from defatted seeds of Camellia exhibited anti-inflammatory and analgesic activities.¹⁷⁾ This report is possibly interesting because the extraction conditions for the sasanqua saponins fraction were very similar to ours (ethanol concentration: 71.98 vs. 70.00%; extraction temperature: 51.99 vs. 50.00 °C). Further, even though it was not an identical compound, Lee et al. have showed that structurally similar saponins (i.e. steroidal saponins from Anemarrhena asphodeloides) dose-dependently inhibited AchE activity (IC50 value, 35.4 μM),¹⁸⁾ warranting further investigations to isolate the responsible constituent(s) and to elucidate their identity, using bioassay guided fractionation and analytical chemistry techniques (e.g. MS/MS and NMR), respectively. In addition, bioavailability and permeability of putative compounds through the blood brain barrier should be also characterized in future investigations.

Optimization, by its definition, refers to advancing the performance of a process (e.g. extraction) in order to achieve maximum benefits. Conventionally, optimization experiments involved observing the influence of one variable on an experimental response over time. However, this approach is unable to capture interactions between variables, leading to the development of alternative multivariate statistical techniques, such as RSM. RSM is a collection of statistical and mathematical techniques that can be utilized for the modeling and analysis of problems in which a response is influenced by multiple variables simultaneously.¹⁹⁾ Several other studies have also applied the same statistical design of RSM to optimize experimental conditions (e.g. derivatization and extraction conditions) of biological matrices.^20,21) Recently, Ebrahimi et al. utilized RSM to optimize the experimental conditions for AchE immobilization on ceramic packaging.²²⁾ However, this is the first report regarding the AchE inhibitory potency of the byproduct from C. japonica processing and optimization of these processing conditions, adding significant experimental detail for future industrial purposes.

On the other hand, there is a caveat to interpret RSM data. For instance, when it comes to the extraction yield, it would be generally increased with longer exposure to solvent and then eventually reached to a plateau phase. In fact, even though most data points are showing that the extraction yield is increased with longer extraction time whereas one data point indicating that the extraction yield was slightly “decreased” (i.e. run 3 and 4; Supplemental Table 2) which might contributed to the threshold value of extraction time in the model. However, the actual difference between these time points was very small (i.e. 0.28%) and definitely fell into a range of experimental deviation. The overall deviations for the extraction yield were approximately ±0.85% throughout our data-set. Thus, although it seems somewhat illogical that the extraction yield was slightly decreased in response to extraction time, once reached to the threshold, it could be a pitfall of a RSM approach since it does not count experimental deviations but only mean.

Taken altogether, in this study, we explored the in vitro inhibitory potential of AchE extracted from C. japonica seed cake. Furthermore, we demonstrated that the efficacy of C. japonica seed cake extraction on AchE activity was influenced by the extraction conditions. We used extraction time, ethanol concentration, and temperature as variables to calculate the highest efficacy and extraction yield for this process. The results from this study are potentially interesting in two aspects. First, although a few studies showed health-promoting effects (e.g. anti-oxidative activity) of C. japonica oil,¹⁰⁾ the beneficial activity of its industrial byproduct (i.e. the seed cake) is still relatively unexplored. As this residual product from camellia oil processing accounts for more than 60% of total weight, this provides a reasonable rational to utilizing them in future studies. Second, we demonstrated that extraction yield, as well as AchE inhibitory activity, of the C. japonica seed cake can be optimized utilizing the RSM statistical model. This model was validated by means of direct comparison between the predicted values and experimental values, suggesting that RSM may provide a useful tool to optimize conditions, not only for experimental purposes, but also for industrial applications.

Supplemental material

The supplemental material for this paper is available at http://dx.doi.org/10.1080/09168451.2014.915723.

Funding

This work was supported by the Gyeongnam National University of Science and Technology [grant number 2013].

Notes

Abbreviations: Ach, acetylcholine; AchE, acetylcholinesterase; AchEI, acetylcholinesterase inhibitor; AD, Alzheimer’s disease; PC12, rat pheochromocytoma cells; RSM, response surface methodology.