Beyond conventional biomass valorisation: pyrolysis-derived products for biomedical applications

Biomass valorisation is conventionally associated with the production of green biofuels. However, this could extend beyond the conventional perception of biomass application into other domains such as medical sciences. Acid condensate (AC) obtained from pyrolysis promises a good potential for biomedical applications, notably for its antimicrobial, antioxidant, and anti- inflammatory properties. In this study, concentrated AC extract (CACE) obtained from microwave-assisted pyrolysis of palm kernel shells was fractionated, and the resulting fractions were pooled according to similar thin layer chromatography profiles into combined fractions (CFACs). CFACs were evaluated for total phenolic content, antioxidant level, cytotoxicity, and wound healing activities toward human skin fibroblast cells (HSF 1184). CFAC-3 showed the highest total phenolic content (624.98 ± 8.70 µg GAE/mg of sample) and antioxidant activities (DPPH IC 50 of 29.47 ± 0.74 µ g/mL, ABTS of 1247.13 ± 27.89 μg TE/mg sample, FRAP of 24.26 ± 0.71 mmol Fe(II)/mg sample, HFRS of 257.74 ± 1.74 µg/mL) compared to CACE (DPPH IC 50 of 81.76 ± 2.81 µg/mL, ABTS of 816.95 ± 30.49 μg TE/mg sample, FRAP of 9.22 ± 0.66 mmol Fe(II)/mg sample, HF RS of 689.30 ± 36.00 µg/mL), no cytotoxic properties at ≤50 µg/mL, and significantly faster wound closure (at 1.25 µg/mL) compared to the control 12 h after treatment. The phosphorylation of the phosphatidylinositol 3-kinase (PI3K) and protein kinase B (AKT) were upregulated, thus indicating that wound healing of CFAC-3 followed through this signalling pathway. To conclude, phenolic- rich CFAC-3 obtained from the pyrolysis of palm kernel shells demonstrated potential biomedical application as an alternative wound healing agent with high antioxidant and wound-healing activity. To the best of our knowledge, this was the first study to report on the wound healing activity of AC and its wound healing mechanism.


Introduction
Malaysia produces around 19 million tonnes of palm oil annually, the second largest in the world. However, at the same time, a huge amount of solid and liquid oil palm biomass is also generated, which amounts to more than 140 million tonnes (Malaysian-German Chamber of Commerce and Industry, 2017). Improper disposal of the high volume of biomass waste through open burning or being left for natural degradation poses serious environmental problems as it may lead to the emission of greenhouse gases such as carbon dioxide and methane and cause air pollution (Purnomo et al., 2018).
Biomass valorisation into biofuels and value-added products offers a highly attractive solution for waste management, simultaneously reducing fossil fuel usage and the production of greenhouse gases . The advantages of biomass include carbon-neutrality, renewability, and sustainability in terms of not interfering with food and feed supplies. Biomass valorisation can be achieved via thermochemical conversions such as microwave-assisted pyrolysis, which yields biochar, bio-oil, acid condensate (AC), and syngas. Many research works focus on producing biofuel through biomass pyrolysis to achieve a low carbon-intensive environment. This offers the opportunity to utilise the other value-added products generated through this biomass valorisation pathway, i.e., biochar and AC.
AC is a reddish-brown pyrolytic liquid condensate obtained from the pyrolysis of highly lignocellulosic biomass such as palm kernel shells and is rich in phenol and its derivatives (Zulkifli et al., 2021). AC has been reported for various chemical and biological properties such as antioxidant (Zulkifli et al., 2021), antibacterial (Mohd Hamzah et al., 2022), antifungal (Ibrahim et al., 2013), and anti-inflammatory (Rabiu et al., 2021). In addition, AC has low cytotoxicity at a 100-fold dilution (Kimura et al., 2002), while some reported the range was between 0.14 to 2% v/v (Filippelli et al., 2021;Ho et al., 2021). It also does not pose a severe environmental hazard (Tiilikkala et al., 2010). Currently, the market size for AC is relatively small and valued at USD 4.5 million in 2019 and projected to grow to USD 6.4 million by 2027 as it is commercially used in agriculture such as pesticide and fertiliser and animal feed as a feed supplement. A recent study has incorporated AC in an oral application to prevent biofilm formation related to dental caries (de Souza et al., 2021).
Chronic wounds are a big burden on the health care system due to their prevalence and high-cost projections, with recent estimates at USD 96.8 billion (Sen, 2021). In Malaysia, surgical site infection incidence at public hospitals was 11.7% which was higher than published figures from India (5%) and Greece (5.3%) (Wong and Holloway, 2019). Bad management of microbial infection can, in turn, lead to prolonged healing and, worse, becoming non-healing wounds due to a longer inflammation phase (Rowan et al., 2015). In recent years, one-third of the drugs intended for wound healing have been obtained or derived from plants (Lordani et al., 2018) due to their potent antimicrobial and wound-healing properties while exhibiting fewer side effects.   In light of the above, this work aimed to evaluate the cell cytotoxicity and wound-healing activities of phenolic-rich fractions of AC extract obtained through the pyrolysis of palm kernel shells towards human skin fibroblast (HSF 1184). Moreover, to the best of our knowledge, the in vitro wound-healing activity and wound-healing mechanism of AC compounds were also investigated for the first time.

Sample preparation of palm kernel shell
The palm kernel shell samples were obtained from a local palm oil mill located in Johor, Malaysia. The sample was washed using tap water, sun-dried for 3 d, and grounded (Df-25 automatic herb grinder, DA DE Brand) to 1-3 mm before use (Zulkifli et al., 2021).

Production of acid condensate
Concentrated AC extract (CACE) was produced using a laboratory-scale microwave-assisted pyrolysis reactor setup previously described by Abas et al. (2018). AC production was performed based on the optimisation study using response surface methodology via Design Expert Software Version 7 on three different factors. From the study, the highest yield of AC was recorded at 29.1 wt% and was achieved at the following optimized conditions; nitrogen flow rate of 3 L/min, microwave power of 575 W, and final temperature of 450 ºC (Zulkifli et al., 2021). The obtained AC was collected, filtered, and extracted using ethyl acetate (EA) AR grade at a 1:1 ratio (Loo et al., 2008). The AC extracts were then concentrated using a rotary evaporator (120 mBar, 80ºC, Heidolph, Germany), dried in a desiccator, and termed CACE.

Fractionation of acid condensate
The CACE was fractionated using column chromatography (5 cm i.d.

Total phenolic content
The total phenolic content (TPC) in AC was determined as follows (Ma et al., 2014); 1 mL of the CFACs and 1 mL of 50% of Folin Ciocalteu reagent were mixed, followed by the addition of 1 mL of 10% sodium carbonate (105.99 g/mol, QRec). The mixture was left to stand for 2 h at room temperature, and the absorbance was measured at 765 nm using a UV-vis spectrophotometer (Shimadzu UV-1800, Japan). Similar procedures were repeated for gallic acid as standard. The determined TPC was expressed as μg gallic acid equivalent/mL of dried sample (μg GAE/mL).

Antioxidant activity
The 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay was performed with minor modifications to the method described by Brand-Williams et al. (1995), where 1 mL of CFAC-3 was mixed with 2 mL of methanolic DPPH reagent. The mixture was shaken at 100 rpm and allowed to stand at room temperature for 30 min. The absorbance was measured at 517 nm with methanol as blank. The ferric reducing antioxidant power (FRAP) assay was conducted according to Ma et al. (2014) by adding 100 μL of 30 μg/mL CFAC-3 or standard L(+)-Ascorbic acid and butylated hydroxyanisole (BHA) 96%, respectively, into 3 mL of freshly mixed FRAP reagent (300 mM acetate buffer (pH 3.6), 10 mM 2,4,6-tripyridyl-s-triazine dissolved in 40 mM hydrochloric acid, and 20 mM ferric chloride in the ratio of 10:1:1) and shaken thoroughly before being left upright to react for 90 min at 37 ºC in the dark. The absorbance of the mixture was recorded at 593 nm using a UV-vis spectrophotometer. The results were expressed as mmol Fe(II) being reduced by per milligram of the sample (mmol Fe(II)/mg sample). The 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) radical scavenging (ABTS) assay was carried out as described by Re et al. (1999). ABTS radical cation was prepared by mixing both 7 mM ABTS solution and 4.9 mM potassium persulphate solution in a 1:1 ratio (v/v). The radical stock solution of ABTS•+ was diluted using ethanol to an absorbance of 0.8 ± 0.005 at 734 nm. The Trolox solution was used to generate the Trolox standard curve. CFAC-3 or standards (L(+)-ascorbic acid and butylated hydroxyanisole) with a volume of 0.4 mL (20 μg/mL) was mixed with 3.6 mL of the ABTS•+ solution and incubated at 37 °C for 7 min followed by measurement at 734 nm.

Cell cytotoxicity study
Cell cytotoxicity assay was performed using MTT (3-(4,5-Dimethyl-2thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide) assay procedure described by Mustaffa et al. (2015) with a slight modification. The fibroblast cells were seeded in 96-well microplates at a density of 5×10 5 cells/well with 100 μL of Dulbecco's Modified Eagle Medium (DMEM) containing 10% v/v foetal bovine serum (complete DMEM) and incubated overnight at 37 °C in a humidified atmosphere containing 5% CO2 to allow the cells to confluence in the wells. Then, the medium was discarded and replaced with 100 μL complete DMEM containing 0.1% v/v DMSO of different concentrations of 100, 50, 25, 12.5, 6.25, and 3.125 μg/mL of CFAC-3 in respective wells. The wells without treatment served as a negative control. The plate was incubated for 24 and 48 h at 37 °C in a humidified atmosphere containing 5% CO2. After incubation, the cell was washed with 100 μL phosphate buffer saline. Sterile MTT solution of 5 mg/mL was added by 20 μL to each well, incubated for 3 h, replaced with 200 μL of DMSO to each well, and left at room temperature for 30 min to allow the insoluble formazan to dissolve. The absorbance of the supernatant was measured at 570 nm using an ELISA microplate reader. The percentage cell viability was calculated using Equation 1.

Scratch wound healing assay
The scratch wound healing assay was carried out using the method by Mustaffa et al. (2015). The fibroblast cells were grown at a density of 5×10 5 cells/well in a 6-well plate and cultured in complete DMEM for 24 h. The fibroblast cells with 90% confluency underwent starvation by replacing the current media with DMEM without foetal bovine serum for 24 h. After 24 h, the media was removed, the confluence cells were scratched using the sterilized pipette tip (yellow tips; volume 100 μL), washed with 1 mL sterile phosphate buffer saline, and treated with 2 mL of 0.1% v/v DMSO in complete DMEM media (as negative control) and CFAC-3 with a concentration of 1.25 µg/mL and 12.5 µg/mL for 30 h. The pictures of wound closure were taken at intervals of 0, 12, 24, and 30 h. The percentage of wound closure was calculated using Equation 2.

Molecular docking
Autodock Vina (Trott and Olson, 2010) was used to evaluate the possible binding mode between CFAC-3 compounds and the specified binding site of target proteins (AKT, 3QKK and ERK2, 6NBS). The molecular docking was performed according to the method described by Rabiu

Western blot
The phosphorylation level of phosphatidylinositol 3-kinase (PI3K) and protein kinase B (AKT) in HSF 1884 was determined using the method proposed by Abate  Phosphorylated band ratio = phosphorylated band / Eq. 3 unphosphorylated band × 100

Gas chromatography-mass spectrometry analysis
CFAC-3 were analysed for chemical compositions using a gas chromatograph-mass spectrometer (GC-MS, QP500, Shimadzu) based on the method suggested by Zhai et al. (2015) with slight modification. Briefly, a sample volume of 20 µL was dissolved in 2 mL of 95% methanol HPLC grade and filtered using a 0.2 µm membrane syringe filter. About 1 µL of the filtered sample was injected with a split rate of 20:1 into the capillary column (HP-5) with a length × diameter of 29.4 m × 0.25 mm. The injector pressure and split flow rate were 10.97 psi and 23.8 mL/min, respectively. Helium gas was used as a carrier gas at a flow rate of 2 mL/min, and the temperature of the injector was at 300 °C; 50 °C for 2 min with a heating rate of 10 °C/min up to 300 °C. The final temperature of around 325 °C was held for 10 min, and each sample was run for around 37 min. As for mass spectrometry (MS), the electron ionization with 70eV was used to detect the mass fragment at a scan range between 50 to 550 m/z. The ion source temperature and transfer line were set at 200 °C and 300 °C, respectively. The GC peak areas were integrated, and the component identification was made by comparing the MS with standards and with a library search (National Institute of Standard and Technology (NIST), USA.

Statistical analysis
All experiments were carried out in triplicates. Quantitative data were analysed using Microsoft Office Excel and GraphPad Prism 7.0 (GraphPad Software, Inc.), and all the results were expressed as a mean ± standard deviation.

Total phenolic content
The total phenolic content of CACE and CFACs is presented in Figure 1. CFAC-3 exhibited the highest TPC of 625 ± 9 µg GAE/mg of sample followed by CFAC-2 (434 ± 2 µg GAE/mg) and CFAC-1 (344 ± 15 µg GAE/mg). All these combined fractions displayed significantly higher TPC (p < 0.0001) than CACE, standing at 296 ± 6 µg GAE/mg. The other combined fractions (CFAC-4 to CFAC-9) exhibited significantly lower TPC than CACE. Most of the phenolic compounds were eluted early due to the low polarity solvent system of n-hexane and ethyl acetate, which had a good elution effect on monophenols and derivatives (Wang et al., 2016). The high amount of TPC in the CACE was due to the high lignin content of the feedstock used, as phenol and derivatives were mainly generated from thermal degradation of lignin fraction, which is very high in palm kernel shells (Collard and Blin, 2014). The obtained TPC value of CACE was more favourable than the AC obtained from oil palm fibre (Abas et al., 2018) and pineapple waste biomass, i.e., 95.0 ± 1.1 µg GAE/mg (Mathew et al., 2015).

Antioxidant activity
Reactive oxygen species (ROS) are closely related to wound healing, particularly inflammation-and oxidative stress-induced cellular damage, which is the main cause of delayed wound healing (Sanchez et al., 2018). Therefore, experimental studies on the regulation of ROS through antioxidant assays could be an important strategy for chronic wound healing. The results for the antioxidant assays of CACE and CFAC-3 are summarised in Table 1. CFAC-3 with IC50 of 29.5 ± 0.7 μg/mL exhibited significantly better DPPH radical scavenging activity than CACE (p < 0.0001) while showing similar scavenging activity to both standards of ascorbic acid and BHA with no statistical difference between them (p > 0.05). AC from walnut was reported to exhibit 1.5 times more DPPH radical scavenging activity than ascorbic acid (Wei et al., 2010). AC from Litchi chinensis and rice hull showed lower or at least comparable scavenging activity to commercial antioxidant BHT Yang et al., 2016). CFAC-3 displayed significantly higher ABTS radical scavenging activity (p < 0.0001) than CACE and ascorbic acid but not as good as BHA. Loo et al. (2008) reported that ABTS radical scavenging activity values of successfully isolated syringol, catechol, and 3-methylcatechol from Rhizophora pyroligneous acid were in the range of 956 ± 40 and 1039 ± 51 μg Trolox/mg sample.
CFAC-3 displayed the highest reducing ability towards TPTZ-Fe (III) with a FRAP value of 24.2 ± 0.7 mmol Fe(II)/mg sample (p < 0.01), followed by BHA, CACE, and ascorbic acid. Similar results were reported by Ma et al. (2014) and Wei et al. (2010), as AC from Rosmarinus officinalis and walnut had higher FRAP values than standard BHA and ascorbic acid. CFAC-3 exhibited better hydroxyl free radical scavenging (HFRS) activity with its IC50 value of 258 ± 2 μg/mL compared to CACE but lower activity compared to ascorbic acid as its IC50 was 1.69 times larger than that of ascorbic acid (152 ± 11 μg/mL). This finding aligned with the result reported by Wei et al. (2010), indicating that AC had a lower scavenging rate of hydroxyl free radical than ascorbic acid. All the findings showed a direct correlation between the TPC of AC and its antioxidant activities, strongly suggesting that the antioxidant activity was mainly attributed to the phenolic compounds as major compounds of AC. Theapparat et al. (2019) reported a similar correlation between the TPC of the AC obtained from the brushwood biomass waste of mangosteen, durian, rambutan, and langsat and DPPH radical scavenging activity.

Cell cytotoxicity
Cell cytotoxicity was determined through MTT dye colourimetric assay, which utilises cell metabolic activity as MTT is reduced by NAD(P)Hdependent cellular oxidoreductase and dehydrogenase enzymes in viable cells into water-insoluble purple formazan (Ghasemi et al., 2021). The cytotoxicity of CFAC-3 towards human skin fibroblast (HSF 1184) was rated based on the following guidelines suggested by Kanaparthy and Kanaparthy (2016); cell viability > 90% is non-cytotoxic, 60-90% is slightly cytotoxic, 30-60% is moderately toxic while less than 30% is strongly cytotoxic. The morphological alteration was observed when the concentration was increased to 50 μg/mL for CFAC-3. The cells exhibited shrinkage, became almost spherical and lost their ability to attach to the wall surface after 24 and 48 h incubation. These round shape cells were no longer living as they were detached, floated, and lifted from the material surface (  cytotoxic at a concentration of ≤50 μg/mL after 24 h. The percentage of cell viability significantly increased when treated with 12.5 μg/mL of CFAC-3 after 24 h, which displayed its ability to enhance cell proliferation. When the treatment period increased beyond 24 h, a concentration of 50 μg/mL started to pose slight cytotoxicity to the HSF 1184 after 48 h with a percentage cell viability of 67.7 ± 2.8% (p < 0.001), respectively. CFAC-3 became moderately cytotoxic at concentration of 100 μg/mL for 24 h (44.2 ± 13.6%, p < 0.001) and 75 μg/mL for 48 h (30.6 ± 10.0%, p < 0.0001). It was strongly cytotoxic to HSF 1184 at a concentration of 100 μg/mL after treatment for 48 h as the percentage cell viability significantly dropped (p < 0.0001) to 25.4 ± 7.9%.
The fractionation process led to a lower range for safe and nontoxic concentration than CACE (≤100 μg/ml), which might be due to higher TPC in CFAC-3. CFAC-3 was observed to enhance proliferation of fibroblast cells at the concentration of 12.5 µg/mL (p < 0.05). At low concentrations, phenolic compounds can stimulate several signalling events, namely mitogenactivated protein kinases (MAPKs) and PI3K/AKT, which regulate cells functions, including proliferation, gene expression, differentiation, mitosis, cell survival, and apoptosis, while at high concentrations, ICE/Ced-3 proteases (caspases) is triggered, where this pathway plays a key role in apoptotic cell death (Kong et al., 2000;Abate et al., 2020). In short, AC can be biphasic, where at low concentration, it can cause cell proliferation, while at high concentration, it can be antiproliferative.

In vitro wound healing activity
The ability of CFAC-3 to stimulate fibroblast cell migration was observed at two concentrations with 10-fold differences (1.25 µg/mL and 12.5 µg/mL) under an inverted light microscope at time intervals of 0, 12, 24, and 30 h, as depicted in Figure 3a. Treatment with CFAC-3 (1.25 µg/mL) exhibited significantly fastest wound closure (p < 0.01 at 12 h, p < 0.0001 at 24 h, and p < 0.001 at 30 h) compared to the negative control after 30 h of treatment as shown in Figure 3b. This result opens a new perspective on applying AC as a wound healing agent. This finding also contradicted the previous report by Lee et al. (2011), indicating that oak wood vinegar exhibited antiproliferative activity against keratinocytes at a dose-dependent concentration (Lee et al., 2011). The difference in the results might be due to the high concentration of AC used in their study, which ranged up to 1.6%, while in this study, lower concentrations were used (1.25 and 12.5 μg/mL). This view was further supported by the experimental results where 12.5 μg/mL of CFAC-3 showed no cell migration, suggesting that low concentrations of CFAC-3 could promote wound healing activity while high concentrations could cause an antiproliferative effect instead. Phenolic derivatives from Calendula arvensis L., Lavandula stoechas L., and Helichrysum italicum extracts also showed enhanced wound healing activity at low concentrations of 1, 5, and

Wound healing mechanism compounds via molecular docking verification
Molecular docking was performed to elucidate the potential molecular interaction of CFAC-3 towards AKT and extracellular signal-regulated kinase (ERK) enzymes. The hydrophobic motif (HFPQFpSYSAS) of AKT was targeted as it is the binding site of PH domain leucine-rich repeat protein phosphatase (PHLPP) on activated AKT and causes subsequent dephosphorylation on Ser473 leading to deactivation of AKT (Sierecki et al., 2010). The molecular docking target was specified to its D domain with a sequence of LEQYYDPSDEPVAEA of ERK2, thus preventing the dephosphorylation of ERK2 by MAPK phosphatases which downregulate the level of phosphorylated ERK2, thus reducing cell migration and proliferation (Tanoue et al., 2000;Bardwell et al., 2003) The binding energy, as well as the formation of hydrogen bonds and other physical interactions towards AKT and ERK2, were determined for each ligand, as shown in Table 2. All the possible docking poses of selected CFAC-3 compounds showed negative binding energy indicating favourable binding affinity with the target enzymes. For AKT, most compounds displayed similar lowest binding energy, ranging from -4.0 to -5.5 kcal/mol. 1-butanone, 3-methyl-1-(2,4,6-trihydroxy-3-methylphenyl)-exhibited the lowest binding energy (-5.5 kcal/mol) while phenol, 2,6-dimethoxy-(-4.0 kcal/mol) showed the highest binding energy among the CFAC-3 compounds. The 3D mode and the 2D interaction residues of CFAC-3 compounds, i.e., 1butanone, 3-methyl-1-(2,4,6-trihydroxy-3-methylphenyl)-with AKT at the hydrophobic motif (HFPQFpSYSAS) are illustrated in Figure 4a. All the ligands were observed to anchor in the binding pocket of the hydrophobic motif by mainly interacting with Arg144, Val145, Phe217, and Asp473. Besides that, other amino acid residues of the hydrophobic motif, including His468, Phe469, Pro470, Gln471, Phe472, Tyr474, and Ser475, noncovalently interacted with the ligands through a conventional hydrogen bond, carbon-hydrogen bond, pi-donor hydrogen bond, pi-alkyl, amide-pi stacked, alkyl and van der Walls as listed in Table 2.
Peptides derived from sea cucumber reportedly accelerated wound healing by upregulating ERK/AKT pathway by inhibiting the hydrophobic motif of AKT (Zheng et al., 2022). The highly important residue of Asp473 was generally targeted by the ligands as this residue is phosphomimetics, the phosphorylated Ser473 of activated AKT, which is the main target of the PHLPP for dephosphorylation of AKT (Balasuriya et al., 2018). The interaction between these amino acids and ligands would competitively inhibit PHLPP binding with AKT and interrupt the dephosphorylation. These results were in line with the western blotting assay as the p-AKT level remained high after being treated with CFAC-3 for 24 h.

Wound healing mechanism via PI3K/AKT pathway signallingin vitro study
Western blotting experiments were performed to investigate the effects of CFAC-3 treatment (1.25 and 12.5 µg/mL) on the wound healing mechanism focusing on the protein expression and phosphorylation of PI3K and AKT of fibroblast cells. PI3K/AKT signalling pathway is a recognised pathway strongly associated with the formation of an epidermis barrier and wound healing as it regulates cell proliferation, differentiation, and migration, along with angiogenesis and metabolism (Hou et al., 2019;Qu et al., 2021). AKT is located as one of the target proteins downstream of PI3K. The activation of PI3K leads to the formation of phosphatidylinositol (3,4,5)-trisphosphate (PIP3) from phosphorylation of phosphatidylinositol (3,4,5)-bisphosphate (PIP2), which in turn leads to activation of AKT by phosphorylating serine residue of AKT. In this study, the total protein of unphosphorylated PI3K (t-PI3K), phosphorylated PI3K (p-PI3K), unphosphorylated AKT protein (t-AKT), and phosphorylated AKT protein (p-AKT) were detected and analysed using western blot in the presence of housekeeping protein β-Actin as a loading control at 4, 6, and 24 h intervals as shown in Figure 5a and Figure 6a. The relative protein quantification of p-PI3K/t-PI3K and p-AKT1/t-AKT1 as normalized to β-Actin was presented as a bar chart in Figure 5b and Figure 6b based on the grey value detection using ImageJ software.

Pi-Alkyl Tyr126
Abbreviations: BE: Binding energy; H Bond: hydrogen bond.   The protein expression of t-PI3K and t-AKT remained relatively the same compared to the negative control after treatment at any time intervals, as shown in Figure 5a and Figure 6a. There was neither upregulation nor downregulation of protein expression after the treatment. To prevent the inaccuracy of the amount of protein loading, housekeeping protein β-Actin was used as an internal control for protein loading as well as a reference (western blot normalisation) in the western blotting analysis. The housekeeping protein β-Actin depicted a similar size and intensity of bands which showed that the amount of protein loaded onto SDS-PAGE gel was relatively the same, and the change in the band of p-PI3K and p-AKT was caused by the amount of protein phosphorylation. Meanwhile, the bands of p-PI3K and p-AKT showed a difference in terms of size and band intensity as the cells treated with 1.25 µg/mL CFAC-3 and positive control displayed darker and slightly thicker bands compared to the negative control. CFAC-3 treatment showed two opposite results. At the lower concentration of 1.25 µg/mL, relative p-PI3K/PI3K level showed significant increase compared to the negative control for all time intervals of 4 h (p < 0.01), 6 h (p < 0.01), and 24 h (p < 0.0001) while p-AKT/AKT level of 1.25 µg/mL CFAC-3 showed significant increase compared to negative control for all time intervals of 4 h (p < 0.05), 6 h (p < 0.001), and 24 h (p < 0.0001). Contrastingly, the treatment with 12.5 µg/mL of CFAC-3 yielded a decrease of p-PI3K/PI3K in HSF 1184 fibroblast cells compared with those in control cells at 6 h. Treatment with 12.5 µg/mL of CFAC-3 yielded no significant change in p-AKT/AKT level in fibroblast cells compared with those in negative control at other time intervals. Positive control showed significant increase of p-PI3K/PI3K level after 6 h (p < 0.0001) and p-AKT/t-AKT level after 4 h and 6 h (p < 0.05).
Thus, in this study, the treatment of CFAC-3 was able to upregulate the protein expression of p-PI3K and p-AKT as early as 4 h and maintained it until 24 h while there was no change to the protein expression of t-PI3K and t-AKT. This result is consistent with previous studies, which suggested that the phosphorylation of PI3K and Akt signalling pathways plays an important role in the promotion of proliferation and migration of fibroblasts, keratinocytes, and endothelial cells (Pericacho et al., 2013;Sepe et al., 2013). Treatment of dracorhodin perchlorate, a polyphenol, activated PI3K/AKT, ERK, p38/MAPK, and Wnt/β-catenin signalling cascades as there was a markedly increase in the phosphorylation level of p-AKT, p-ERK, and p-p38 in keratinocytes cells (HaCaT) (Lu et al., 2021). 4-hydroxylbenzylaldehyde treatment significantly increased phosphorylation of ERK and AKT in keratinocytes, leading to keratinocyte cell migration (Kang et al., 2017). Buxuahuayu decoction, which is a mixture of herbs containing polyphenols, was reported to regulate the level of nitric oxide (NO) in local wounds by activating the PI3K/Akt/eNOS signalling pathway when tested in vivo on rats with diabetic ulcers (Qu et al., 2021).
The treatment of CFAC-3 yielded two contrasting results at two different concentrations; low concentration (1.25 µg/mL) yielded upregulation of protein expression level of p-PI3K and p-AKT while high concentration (12.5 µg/mL) caused downregulation of the protein expression level. This finding shows that CFAC-3 exhibited a biphasic effect which means the increase or decrease of concentration of the drugs/chemicals could result in opposite effects (Kong et al., 2000). 3-hydroxytyrosol, the main component in olive oil, has been reported to be pro-angiogenesis at concentrations lower than 10 µM or approximately 1.5 µg/mL, which is similar to the concentration used in this study (Abate et al., 2020). Resveratrol has been reported to be anti-angiogenic as it inhibited AKT and ERK1/2 signalling pathways as the concentration increased from 25 to 100 µM to treat renal cell carcinoma cells (Zhao et al., 2018).

Limitations of the present study
The study was performed only using fractionated samples of AC extract that exhibited the highest TPC and antioxidant activities. The mass concentration of the samples used was based on the consistent dry weight of the sample. Molecular docking was performed based on a single compound to target enzyme interaction. Moreover, the in vitro wound healing mechanism was limited to phosphorylation of the PI3K/AKT signalling pathway at the molecular level.

Conclusions
CFAC-3 exhibited higher antioxidant activity compared to CACE and standards, which, based on the TPC and GC-MS analysis, could be associated with the presence of many phenolic compounds and their derivatives. It also showed no cytotoxicity at the concentration of ≤50 µg/mL against HSF 1184 after 24 h. CFAC-3 at a 1.25 µg/mL concentration led to the fastest wound closure. Molecular docking analysis suggested favourable binding energy for all chemical compounds present in CFAC-3, notably 7,9-di-tert-butyl-1oxaspiro(4,5)deca-6,9-diene-2,8-dione, ethanone, and 1,2-benzenediol, 3methyl-towards AKT and ERK2. This was further supported by the Western blot analysis as the treatment with 1.25 µg/mL CFAC-3 caused an upregulation of PI3K and AKT phosphorylation as soon as 4 h.

Practical implications of the study
To the best of the authors` knowledge, this is the first report to determine the wound healing activity of phenolic-rich fraction of AC from palm kernel shell and its potential wound healing mechanism through the PI3K/AKT signalling pathway. These findings broadened the application of bioproducts obtained from biomass valorization (through pyrolysis herein) beyond the conventional applications in biomedical science, particularly for wound healing.

Future perspectives
Further study using specific inhibitors of PI3K and AKT such as wortmannin is required to further confirm that CFAC-3 directly targeted these proteins. In vivo study using rats would allow a more in-depth understanding of the wound healing effect of CFAC-3 in a more complex system which mimics the human system.