An appropriate ratio of unsaturated fatty acids is the constituent of hickory nut extract for neurite outgrowth in human SH‐SY5Y cells

Abstract Hickory nuts (Carya cathayensis Sarg, CCS), a well‐known Chinese medicinal nut, is thought to improve memory in Chinese folks. However, functional constituents have not been scientifically identified. In this study, human SH‐SY5Y cells, combined with Q‐TOF mass spectrometry (Q‐TOF‐MS) and standard substances, were used to evaluate the function in neuronal development and to identify constituents of CCS hydrophobic extracts (CCS‐HE). Data showed that CCS‐HE but not the control induced neurite outgrowth of SH‐SY5Y cells in a dose‐dependent manner, supported by which CCS‐HE induced the expression of nerve growth factor (NGF), neurofilament 160 (NF160), and neuronal peptide Y (NPY) mRNA. Q‐TOF‐MS analysis with standard substances indicated that linolenic acid (LNA), linoleic acid (LA), and oleic acid (OA) were the main constituents in CCS‐HE. Furthermore, mixtures of these unsaturated fatty acids (UFAs) at the natural ratio (1:8:16) significantly induced neurite outgrowth and gene expression of NGF, NF160, and NPY in a dose‐dependent manner. However, the individual and alternative ratios were not effective to induce the neurite outgrowth and gene expression of NGF, NF160, and NPY. These data implicate that an appropriate ratio of UFAs is the main constituent for the neurite outgrowth.

. Nerve growth factor (NGF) is a classic growth factor in family of neurotrophic factors, a group of proteins that are mainly synthesized and secreted by neurons and astrocytes and that are crucial for neuronal survival, growth, and differentiation (Sofroniew, Howe, & Mobley, 2001).
Studies have shown that NGF causes axonal branching and elongation (Madduri, Papaloizos, & Gander, 2009) and NGF could induce the expression of NF proteins (Huang et al., 2010;Schimmelpfeng, Weibezahn, & Dertinger, 2004). In contrast, NGF defect results in an undergoing apoptosis in neurons (Freeman et al., 2004). NGF is also reported to play a potential role in depression and schizophrenia (Martino et al., 2013;Martinotti et al., 2012). Neuropeptide Y (NPY) is a 36 amino acid peptide and extensively distributes in neurons of the central and peripheral nervous systems (Eipper, Stoffers, & Mains, 1992). It is found that NPY has a higher level than do all other peptides studied in mammalian brain (Gray & Morley, 1986). It acts as another neurotransmitter and/or another modulator of several neuroendocrine functions (Colton & Vitek, 2006). Studies have proved that NPY could exert neuroprotection against Aβ toxicity in both neuroblastoma and primary cells, function as a neuroprotective agent against AD, and indirectly induce neurite outgrowth (White & Mansfield, 1996). Studies reported that NPY exerted its neuroprotective roles by influencing the gene expression of neurotrophins (Angelucci et al., 2014;Croce et al., 2011) and by inducing the neurite outgrowth (White, 1998).
In this study, we used SH-SY5Y cell line as a neuronal model to evaluate the function of bioactive constituents of CCS hydrophobic extracts (CCS-HE). The neurotrophic property of CCS-HE in stimulation of neurite outgrowth and in gene expression of NGF and NFs was validated. The possible involvement of NPY was also determined to further reveal the mechanism of neurites outgrowth by CCS-HE. In addition, an appropriate ratio of LNA, LA, and OA was identified as the main bioactive constituent of CCS-HE in the induction of neurite outgrowth. This study provides the information on which human SH-SY5Y cells would be a desired model to evaluate the function of CCS-HE in neuronal development by measurement of neurite outgrowth.

| Preparation of CCS hydrophobic extracts (CCS-HE)
Fresh hickory nuts that had been frozen in liquid nitrogen were ground into powder. Ten grams (10 g) of powder were used to extract hydrophobic constituents with 150 ml of petroleum ether in Soxhlet apparatus for 6 hr. After extraction, the petroleum ether was thoroughly removed on a rotary evaporator and CCS hydrophobic extracts (CCS-HE), approximately 5 g, were stored at −80°C.
CCS-HE was dissolved in fresh DMSO to prepare a stock concentration of 160 mg CCS-HE/ml and filtrated with a 0.45 µm membrane. To minimize the cytotoxicity of DMSO to SH-SY5Y cells and maximize CCS-HE solubility in cell culture medium, CCS-HE stock solution was diluted in 1:400 or larger dilution in the medium and ultrasonic was used to promote CCS-HE solubility, generating a concentration of 0.25% (v/v) DMSO that had no influence on the growth of SH-SY5Y cells in pre-experiment and a highest final concentration of 0.4 mg/ml CCS-HE in the medium.

| Cell culture
Human SH-SY5Y cells were purchased from ATCC, and SH-SY5Y-EGFP cells were gifted from NUPTEC, respectively. Both cells were maintained in T-25 flasks (Falcon) with DMEM/F12 medium (Sigma) containing 10% fetal bovine serum (FBS, Invitrogen) under conditions of 5% CO 2 and 37°C. The medium was changed every other day. For experiments, cells were cultured in 60-mm culture dishes (Corning) at an initial density of 50,000 cells/ml.

| Treatment with CCS-HE or standard substances
In order to measure effects of CCS-HE or standard substances including linolenic acid, linoleic acid, and oleic acid on neurite outgrowth, CCS-HE at the final concentrations of 0, 0.1, 0.2, and 0.4 mg/ml or standard substances in an appropriate ratio of linolenic acid, linoleic acid, and oleic acid were added to the cell culture medium for 6 days and the concentration of DMSO in each treatment were adjusted to 0.25% (v/v). Twenty ng/ml of bFGF was used as a positive control for neurite outgrowth (Boku et al., 2013). To determine gene expression of neurofilament 160 (NF160), NGF, and NPY, CCS-HE was added to cell culture medium as same concentrations as mentioned above for 24-48 hr. To better visualize the neurite outgrowth under a fluorescent microscope, the ratio of SH-SY5Y cells to SH-SY5Y-EGFP cells was 6:1.

| Measurement of neurite length
Cells were cultured as described above in 6-well plates, and six images of live cell morphology were randomly captured for each treatment on a Leica fluorescent microscope equipped with Compix Imaging System. The images were labeled with a scale in proportion to the magnification, leading to a calculation of μm/ pixel. Each cell had several neurites with different lengths. Usually, the neurite length was measured for total length of all neurites in single cell or for longest neurite length of each cell. In this study, the longest neurite of each cell was randomly selected to measure the neurite length and 50 longest neurites were measured from six images of each treatment. The actual length of neurite outgrowth was calculated as follows: measured pixel × μm/pixel.
Alternatively, the neurite lengths from different treatments were compared by percentages of cells with a longer than 40 μm neurite length over total 50 selected cells.

| Real-time PCR
Total RNAs were extracted from cells using the TRIzol method

| Identification of CCS-HE structures
Twenty microliter of CCS-HE at 160 mg/ml was chromatographied on an Agilent RRLC system (Agilent Technologies, Inc.) consisting of a ZORBAX SB-C8 column (5 µm, 4.6 × 250 mm) in a G1316A incubator, a G1312B Dual Unit pump, a G1322A degasser and a G1367D Auto Sample injector with a flow phase consisting of water (95%) and acetonitrile (5%) for 7 min followed by a gradient of acetonitrile (5%-80%) for 5 min at a flow rate of 1 ml/min.
Twenty micro liters of each standard sample was chromatographied on an Agilent RRLC system under same conditions described as above. Signal peak area was linearly regressed with the concentration of standard unsaturated fatty acid to make the standard curve for each unsaturated fatty acid. Same volume (20 µl) of CCS-HE sample was chromatographied as same as standard UFAs on an Agilent RRLC system, and the area of identical signal peak to each standard unsaturated fatty acid was measured with software on Agilent RRLC system. Individual unsaturated fatty acid content and the ratio in mass of three UFAs were calculated according to the prepared standard curves.

| Statistical analysis
Means of measured parameters from each group were analyzed, and standard errors were produced by using statistical software SPSS version 6.1 (StatSoft Inc.) according to manufacturer's instruction.
Probability levels of p < .05 were considered as statistically significant. Error bars in the graphs indicated standard error (SE) of the mean.

| CCS-HE induced the neurite outgrowth of SH-SY5Y cells
To verify the activity of CCS-HE in neuronal development, CCS-HE was used to test its effect on the neurite outgrowth of SH-SY5Y cells. CCS-HE was added to SH-SY5Y cell medium for a prolonged 6 days. As shown in Figure 1A, basic fibroblast growth factor (bFGF) or fibroblast growth factor 2 (FGF-2), a positive control for the induction of neurite outgrowth, induced the neurite outgrowth of SH-SY5Y-EGFP cells. As expected, CCS-HE also induced the neurite outgrowth. Both CCS-HE and bFGF groups had longer neurite outgrowth than blank group. To quantify the length of neurite outgrowth, six randomly captured images from each group were used to measure the length by the software on the microscope, as described in "Section 2." As shown in Figure 1B

| CCS-HE induced NGF, NF160, and NPY expression
In this study, NF160 and NGF were measured as molecular indicators of CCS-HE effects on neurite outgrowth. As shown in Figure

| Silencing NPY gene expression attenuated CCS-HE-induced NF160 and NGF gene expression
To further confirm whether CCS-HE increased gene expression of NF160 and NGF via NPY, siRNA was used to reduce NPY gene expression. As shown in Figure 2D, consistent with Figure 2A

| Bioactive constituents were identified as UFAs
To identify the constituents of CCS-HE, which can induce the neurite outgrowth of SH-SY5Y cells and upregulate gene expression in the process of neurite outgrowth, CCS-HE was chromatographied by HPLC analysis. As shown in Figure 3A, CCS-HE contains three featured signal peaks in the box, peak 1 at 18.6 min, peak 2 at 20.0 min, and peak 3 at 21.8 min of retention time ( Figure 3A, a), which are absent in petroleum ether used for CCS-HE extraction ( Figure 3A, b) and DMSO used for CCS-HE dissolvability ( Figure 3A, c).
In order to identify the properties of these three signals, next methods were applied for this purpose. First, three signal peaks of CCS-HE were collected, respectively, and their masses were identified by Q-TOF mass spectrometry analysis. As shown in Figure 3C, According to the standard curve of UFA content versus signal peak area, the mass ratio of LNA, LA, and OA was calculated at approximately 1:8:16 and individual contents of three UFAs were 0.096, 0.754, and 1.46 µg/mg, respectively, in CCS-HE samples. Second, based on Q-TOF mass spectrometry analysis, the chromatography of these signals was further compared to that of standard substances including LNA, LA, and OA. As shown in Figure 3B, LNA, LA, and OA

F I G U R E 2 Induction of NF160, NGF, and NPY gene expression by CCS-HE. (A) NF160 gene expression was induced by CCS-HE at different doses for 24
and 48 hr; (B) NGF gene expression was induced by CCS-HE at different doses for 24 and 48 hr; (C) Cells were treated with different doses of CCS-HE for 24 and 48 hr; (D) Attenuation of CCS-HE-induced NGF and NF160 expression by silencing NPY expression. Cells were treated with CCS-HE at 0.4 mg/ml for 24 hr after NPY siRNA treatment as described in "Section 2." Data represents mean ± SE from three independent experiments (n = 3, *p < .05, **p < .01) F I G U R E 3 Identification of functional constituents in CCS-HE. (A) CCS-HE sample, a, was chromatographied in a flow phase consisting of water and acetonitrile (see "Section 2"). Petroleum ether, b, and DMSO, c, were used as blank controls. Three specific signal peaks at 18.6 min (1), 20.0 min (2) and 21.8 min (3) FIGURE 3A), was used as a blank control. Three specific signals, peak 1, peak 2, and peak 3, in CCS-HE were identical to LNA, LA, and OA, respectively. (C) CCS-HE sample was analyzed by Q-TOF mass spectrometry, and three masses (277, 281 and 282) were detected in anion mode, which were identical to the masses of LNA (a, 278), LA  (Table 1).

| Mixture of UFAs induced the neurite outgrowth of SH-SY5Y cells
Based on the mass ratio of three UFAs in CCS-HE, a reconstruction of three standard UFAs was prepared to mimic the bioactive constituent of CCS-HE. As expected, the mixture of UFAs at the mass ratio showing a best appropriate concentration of total UFAs ( Figure 4B). However, we recognized the difference in the percentage of neurite outgrowth with longer than 40 μm in all 50 cells between low and high concentration ranges of UFAs. Also, the individual of each UFA did not show a dose-dependent manner in the induction of neurite outgrowth and was not effective as observed in the mixture of UFAs even though the concentration of individual UFA was higher than the mixture of UFAs ( Figure 5). Further test observed that a natural ratio of LNA:LA:OA, 1:8:16, was a best appropriate ratio to induce the neurite outgrowth ( Figure 6).

| D ISCUSS I ON
Nuts are an important category of food resources for UFA supplies in human nutrition. However, it is rare to investigate that the different ratios of UFAs are relative of their nutritional values in the nervous system. In this study, we demonstrated that CCS-HE had activity to promote the neurite outgrowth in human SH-SY5Y cells (Figure 1), supporting CCS as a conventional folk medicine in China for brain health. This activity for the neurite outgrowth of neuronal morphology was supported by which CCS-HE upregulated the gene expression of NF160 and NGF (Figure 2A,B), one cytoskeleton protein for the neurite outgrowth (Al-Chalabi & Miller, 2003;Madduri et al., 2009;Reuss, Dono, & Unsicker, 2003) and one neurotrophic factor for stimulating neurite outgrowth (Huang et al., 2010;Schimmelpfeng et al., 2004), respectively. The activity was further supported by which CCS-HE also upregulated gene expression of NPY ( Figure 2C), a modulator of NGF and NF proteins (Angelucci et al., 2014;Croce et al., 2011;White, 1998), Both NGF and NF160 expression were upregulated by CCS-HE; however, they showed different patterns. NGF upregulation by CCS-HE showed a transient pattern and NF160 did show a prolonged upregulation by CCS-HE ( Figure 2). This is consistent with literature reports that NGF can regulate NF160 expression (Huang et al., 2010;Schimmelpfeng et al., 2004) and NF160 is one of three neurofilaments for the core structures of the neurite outgrowth and axonal elongation (Angelucci et al., 2014;Barnes & Polleux, 2009;Cheng & Poo, 2012). This is also consistent with morphologic change of human SH-SY5Y cells induced by CCS-HE even though we did not measure both NGF and NF160 expression at an earlier stage (e.g., 12 hr), suggesting that CCS-HE stimulated neurite outgrowth by upregulation of NGF expression and NGF further turned on NF160 expression for neurite outgrowth.
Our further study indicated that NPY was also involved in the induction of the neurite outgrowth by CCS-HE and probably an upstream modulator of NGF and NF160 because CCS-HE upregulated gene expression of NF160, NGF, and NPY whereas NPY depletion by siRNA attenuated the induced expression of NF160 and NGF by CCS-HE ( Figure 2C,D). Although the upregulated patterns of NPY, These results were consistent with literature reports that NPY is a neurotransmitter and/or a modulator of several neuroendocrine functions to regulate gene expression of NGF and NFs and plays a role in neuroprotection (Colton & Vitek, 2006). More recently, it has been shown that NPY was involved in Alzheimer's disease (AD) and NPY exerted neuroprotective action associated with changes in intracellular production of NGF (Angelucci et al., 2014;Colton & Vitek, 2006;White & Mansfield, 1996). Thus, these data support that CCS consumption is beneficial to brain health, described in ancient Chinese herbal book.
This study further identified that LNA, LA, and OA in an appropriate ratio was a bioactive constituent of CCS-HE by analysis of HPLC chromatography combined with standard UFAs Our study mainly focused on in vitro cell model and whether CCS-HE could perform similar effects as that in vivo is unclear. This study provided the information on which human SH-SY5Y cells could be used as in vitro cell model to reliably evaluate the bioactivity of hickory nut extracts for natural products. It also encourages us to use this in vitro model to continue our work. Future study will further identify the mechanism of UFAs mixture-induced neurite outgrowth. Furthermore, optimized ratio of different UFAs in in vitro cell model will be verified in model animals for future clinical trial.
The long-term goal for clinical practice of UFAs mixture will be the manipulation of neuronal development and the recovery of neuronal repair for neurodegenerative disorders.

CO N FLI C T O F I NTE R E S T
We declare that we have no conflict of interest.