Food addiction, a condition recognized as overeating or eating disorder is related to mental health issue in which a person becomes addicted to food [1]. Terms like “chocoholic” and “craving” are used to describe man’s desire and fondness for food [2]. Certain foods have got addictive potential causing loss of control over food intake that may result in eating-related disorders (binge eating disorder, bulimia nervosa, weight gain and obesity) with alteration of behavior and neurological changes [3-5]. It has been explained that the brain response for food addiction is similar or as strong as addiction for drugs [6,7]. Though food items like coffee, bacon, milk, eggs, pizza, chocolate, cheesecake, maize have addictive potential [8,9], the craving for sugar is much stronger in comparison to cocaine [10]. High calorie food has been reported to be highly addictive [5]. Honey is one such saturated solution of sugar having higher calorific value than sugar. The consumption of honey in ancient age and the addiction of sugar in modern age have evolutionary connection [10].

      Honey has already been reported to have addictive properties [10]. Although honey is a saturated sugar solution, protein is present in honey in minute quantity along with other bioactive components. Thus, it was felt essential to know if the protein/ peptide present in honey has any role in addiction. therefore, the present article emphasizes on isolation and identification of protein present in honey along with its biochemical characterization. It has been successful in identifying a purified protein and is currently a fore runner of the future to address such issues.

      CART (Cocaine- and amphetamine-regulated transcript) peptides are novel putative brain/ gut neurotransmitter and co-transmittor that probably have a role in drug abuse, the control of feeding behavoir sensory processing, stress and development. They are abundant, processed and apparently released. On the other hand exogeneously applied peptide cause inhibition of feeding and have neurotropic properties. Besides CART, their are certain other peptides that may act as new putative neurotransmitter and appear to have an important role in different physiological processes including feeding, sensory processing, development, stress etc [11].

      In humans, increasing evidence suggests that individuals eating high calorie food have symptoms of addictive behavior [4,12]. However, understanding the addictive behavior of food necessitates its characterization [13]. Thus, it was felt essential to characterize ingredients present in honey that may be responsible for addictive nature of food. In the present article Litchi chinensis honey, abundantly available in India was collected and protein from the honey was isolated, purified and characterized.

      The structural configuration, shelf life, biological and chemical stability of proteins, its efficiency and recovery are directly influenced by the purification techniques used [14]. Prior purification is necessary to characterize protein [15,16]. Detailed study on purified protein is necessary to understand the characteristics of honey and therefore its role in addiction. Thus, based on this premise, isolation, purification and characterization of protein from Litchi chinensis honey have been considered in this article.

Sample

      Litchi honey ( Litchi chinensis ) was collected from colonies of Apis mellifera in Baruipur apiculture industrial co-operative society Ltd., Dakshin Gobindapur, Kolkata, West Bengal, India. The colonies were placed in litchi plantation following standard apicultural methods and honey was collected by beekeepers during the 2017-2018 harvest seasons (February to March 2017) using a stainless-steel honey extractor (Hi-Tech Natural Products Limited, India). The extracted honey was filtered through a sieve to remove unwanted debris and stored in sterilized sealed glass jars at 4°C.

Isolation of protein from honey

      Honey sample (200 g) was dissolved in 0.01M Tris-HCl buffer (pH 7.4) to a volume of 250 mL. Ultrafiltration process using a 10 kDa polyethersulfone membrane (Sartorious, India) was used to concentrate the solution. The obtained retentate was recirculated several times till the volume was reduced to approximately one-tenth of the initial (25 mL). The protein concentration was checked after each cycle. A fraction of the concentrated retentate (12 mL) was subjected to ultracentrifugation at 15,000 rpm for 15 min. The process was repeated 3-4 times until the dark pellet formed on the wall of the centrifuge tube was completely removed with the collection of supernatant.

Purification of protein by BioLogic LP ion exchange chromatography

      The supernatant (1.5 mL) from the ultrafiltration step was subjected to purification using a Q Sepharose (anion exchange) column (16/20 mm), attached to an FPLC system (Pharmacia). The cartridge was equilibrated with 0.01M Tris-HCl of pH 7.4 (buffer A), into which the concentrated protein was injected. Elution of bound proteins carried out using buffer B (0-50%, 0.5M NaCl in buffer A) at 1.5 mL/min flow rate. Absorbances of all fractions were detected at 280 nm by an online UV detector.

Estimation of protein concentration by Bradford assay

      Total protein content at each step of purification was checked by the Bradford method [17]. Bovine serum albumin (BSA) was used as standard. Buffer A was used as a blank.

Characterization of purified protein

Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE)

      The extent of homogeneity at each step of purification was detected by SDS-PAGE performed on 12% resolving gel and 4% stacking gel following the protocol of Laemmli [18]. Molecular weight (Mw) of the purified protein was determined by comparing the relative mobility of standard protein marker of 10-250 kilodaltons (kDa) (Precision Plus Protein Standard, Bio-Rad).

Determination of molecular weight by MALDI-TOF mass spectrometry

      The molecular weight of the unknown protein was confirmed by Matrix-Assisted Laser Desorption/Ionization Time-Of-Flight Mass Spectrometric (MALDI-TOF/MS) analysis. Sinapic acid was used as a matrix for the analysis.

Protein sequencing

      Automated Edman degradation was carried out on a protein sequencer (Model PPSQ-31A; Shimadzu Scientific, Kyoto, Japan) to determine the N-terminal amino acid sequence of the protein. The purified protein was loaded onto a polyvinylidene difluoride (PVDF) membrane (Millipore) by electrophoresis which was further stained with Coomassie Brilliant Blue R-250 dye (Thermo Fisher Scientific‎), destained and washed thoroughly. Stained spots were cut off and sequence analysis was done. Homology search of the obtained sequence was carried out using BLAST (Basic Local Alignment Search Tool).

Matrix-Assisted Laser Desorption/Ionization Time-Of-Flight Mass Spectrometry (MALDI-TOF/MS)

      MALDI-TOF/MS analysis and peptide masses were determined on a mass spectrometer (UltrafleXtremeTM, Bruker, Germany) using α-cyano-4-hydroxycinnamic acid as matrix. The desired purified protein fraction obtained from ion-exchange chromatography was subjected to hydrolysis using a proteolytic enzyme. The reaction was carried out using sequencing grade trypsin (Promega, Madison, WI) at enzyme: purified protein ratio 1:5 (w/w). Samples after overnight incubation (37°C) were boiled for 5 min and centrifuged at 8000 rpm in a microcentrifuge. The supernatant was then collected for mass spectrometric analysis. MASCOT search program was performed for database search for peptide mass fingerprinting (PMF).

Isoelectric Focusing (IEF)

      Rotofor system (Bio-Rad, USA), with a mini focussing chamber (18 mL) equipped with 20 fractionation compartments was employed to check the isoelectric point (pI) of the purified protein. 0.1 M sodium hydroxide and 0.1 M phosphoric acid were used as electrolytes in cathode and anode assembly respectively. A pH gradient was created using ampholyte (BioLyte 3/10, BioRad, USA) of range 3.0-10.0. Sample solution (18 mL distilled water, 1 mL ampholyte and 0.5 mL purified protein) was prepared and loaded into the rotofor chamber. Focusing was performed at a constant power of 10W for 4 h. After the complete run, 20 fractions were collected and evaluated for pH and protein concentration. The pI value was further confirmed by MALDI-TOF mass spectrometric analysis.

Enzymatic hydrolysis

      The purified protein was digested using sequencing grade trypsin (Promega, USA) to produce protein hydrolysate or peptides [19]. Trypsin (0.03%, w/w) was added to purified protein fraction for hydrolysis at 37°C and pH 7.4 for 24 h. The proteolytic mixture was then boiled for 5 min and subjected to 15 min centrifugation at 8000 rpm. The supernatant was then assayed for antioxidant activities.

Bioactive properties

DPPH (1, 1-diphenyl-2-picrylhydrazyl) assay

      Crude honey, purified protein, and protein hydrolysate/peptides were used for further analysis of the antioxidant property. Scavenging of free radicals by the samples was determined [20]. Samples (0.5 mL) was mixed with a solution of DPPH (4 mL, 0.5mM) in methanol and incubated for 30 min in dark. Methanol was used as a blank to measure the absorbance at 515 nm and results were calculated in triplicates as percent inhibition of DPPH radical using the formula:

      DPPH activity (%) = [(D_control - D_sample)/D_control] × 100

      Where, D_control is the absorbance of solution without sample and, D_sample is the absorbance of sample solution.

FRAP (Ferric reducing antioxidant power) assay

      Crude honey, purified protein, and protein hydrolysate/peptides were assayed for reducing power [21]. Sample (0.5 mL) was added to 1.5 mL FRAP reagent [acetate buffer (300mM/L, pH 3.6): TPTZ solution (10mM in 40mM/L HCl): ferric chloride (20mM FeCl³ .6H² O) at 10:1:1 ratio] and incubated at 37°C for 30 min. Distilled water was used as a blank for measuring absorbance at 593 nm. Calibrations were performed using ferrous sulphate solutions (0-100μM). Results were expressed as μmol of Fe (II) in triplicate values.

ABTS [2, 2’-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid)] antioxidant assay

      Crude honey, purified protein, and protein hydrolysate/peptides were assayed for antioxidant activity in reaction with ABTS radical [22]. ABTS cation was produced by reacting ABTS (50 mL) in phosphate‐buffered saline (2 mM, pH 7.4) with 0.2 mL of potassium persulphate (70 mm) in dark followed by incubation for 12-16 h. ABTS•+ solution was diluted with buffer to obtain an absorbance of 0.700 ± 0.05 at 734 nm. Sample (0.5 mL) was then added to 3.5 ml ABTS^•+ solution, homogenized and incubated for 10 min. Absorbance was then measured at 734 nm against an artificial honey sample as blank. The percentage decrease in the absorbance was calculated using the formula:

      ABTS^•+ inhibition (%) = [(A_blank - A_test)/A_blank] × 100

      Where, A_blank is the absorbance of blank sample (t=0 min) and A_test is the absorbance of test sample at the end of the reaction (t=10 min).

Concentration of protein through ultrafiltration and ultracentrifugation

      To isolate the desired protein from honey solution, ultrafiltration (10 kDa cutoff membrane) process was adopted. Concentration and fractionation of protein were carried out by passing the entire solution for several cycles. Molecular weight compounds more than 10 kDa present in honey solution was collected in the retentate whereas the low molecular weight compounds were collected in the filtrate. After several runs, ultrafiltration followed by ultracentrifugation resulted in honey protein concentration that was next employed to purification. The protein concentration at each step of purification is shown in Table 1.

Purification of protein

      The concentrated protein obtained through ultrafiltration and ultracentrifugation was subjected to purification through anion exchange chromatography where a graph showing two peaks were observed, a minor hump followed by a single major peak as shown in Figure 2(A). Fractions 32 to 41 represents peak 1 while peak 2 is represented by fractions 42 to 53. Q Sepharose (Quaternary sepharose) being a strong anion exchanger, elution of protein was done through a continuous gradient of salt (NaCl). The protein concentration of these fractions were measured by Bradford method and the following fractions yielded the highest protein content (Table I).

      SDS-PAGE of the above fractions showed clear bands in fractions 46, 47 and 48 proving the fact that the desired honey protein has been effectively purified. Figure 2(B) represents a thick band of retentate (lane 1) which got reduced to thin bands in fraction 44 (lane 3), fraction 45 (lane 6), fraction 46 (lane 5), and fraction 47 (lane 7) with subsequent purification steps. Hence the protein present in the highest concentration in the honey sample was purified having a protein content of 0.56 mg/mL in fraction 47. The major peak was isolated and subjected to further characterization.

Molecular weight of the protein

      The fraction of protein showing highest concentration in ionexchange chromatography was collected and subjected to SDSPAGE analysis and the molecular weight was found to be 55 kDa as shown in Figure 3(A). The obtained result had similarity with the honey samples reported previously [23,24]. The molecular weight of the unknown protein was found to be 53-54 kDa by MALDI-TOF mass spectrometry technique as shown in Figure 3(B) which was almost similar to that inferred by gel electrophoretic technique.

N-terminal sequence of the protein

      The protein sequence obtained contains a putative leader sequence and a long segment that contains several pairs of basic amino acids which are potential cleavage sites. The obtained sequence, N-I-L-R-G-E-S-L-N-K-S-L-P-I-L was nearly identical to N(S)-I-L-R-G-E-S-L-D-K deduced from MRJP1 of Apis cerena [25] and was also identical to the previously reported N-terminal sequence of MRJP1 deduced from honeybee Apis mellifera (AmMRJP1) [26,27]. Obtained sequence on homology search exhibited 86% identity with MJRP1 of Apis cerana and Apis florea .

      The analysis showed that the number of nonpolar amino acids were higher than the polar amino acids indicating hydrophobicity of the protein which has an important role in determining the tertiary structure of the proteins shaping the molecule and its active sites.

Identification of protein

      The purified protein of 55 kDa had significant similarity with the reported literature as Major Royal Jelly Protein 1 (MRJP1) or royalactin of Apis mellifera . This protein was further analyzed through MALDI-TOF mass spectrometric analysis. MASCOT search program showed a significant-top score of 77 as depicted in Figure 4(A) with protein sequence coverage of 25% and 14 peptide matches showed in Figure 4(B).

      Major Royal Jelly Proteins (MRJPs) or yellow protein family are proteins available in honey after removing pollen protein [28]. These proteins in honey are a family of nine members MRJP 1-9 (Tamura et al., 2009). MRJP 1-5, also known as apalbumins are the most prominent and has a molecular weight ranging from 49-87 kDa [26]. The most abundant among them is MRJP1, comprising of an oligomer of 280-420 kDa or a monomer of 55 kDa [26]. Thus, the identified purified protein was a monomeric form (MRJP1) also known as apalbumin-1 or royalactin as depicted in Figure 4(B) where the sequence of the major peptides of the protein has been shown.

Isoelectric point (pI) of protein

      Protein concentration was observed in the 7th fraction whereas the remaining fractions revealed no protein content. Thus, the pH 5.5 of the 7th fraction was the respective pI of the protein. The result was identical to the estimated pI value of MRJP1 of Apis cerena (AcMRJP1) [25]. The pI value obtained by isoelectric focusing was different from the pI (5.1) estimated by MALDI-TOF/MS analysis. The difference in the pI value may be because of the lack of true separation barriers between the rotofor chambers, salt and/or buffers transferred with the sample, or minor leakage within the core of anodic and cathodic solutions affecting the gradient linearity [29]. Post-translational modifications during protein separation and characterization may also have altered the pI of protein [29].

Bioactivities

      The crude honey, purified protein and protein hydrolysate/ peptides obtained after tryptic digestion was subjected to biochemical analysis for the evaluation of their biological activities. The results obtained have been shown in Table 2.

      The present study shows protein hydrolysate/peptides to have higher DPPH activity compared to that reported by Guo et al. (2005). The result revealed peptides to have a % inhibition value of 68.21 ± 4.01, which is higher than in the case of purified protein (Table 2). The higher % inhibition value of peptides might be because of the increased solvent accessibility of amino acids due to disruption of the tertiary structure of protein leading to free radical scavenging and metal chelation.

      Reducing power is an important parameter to estimate the reductants present in a biological sample. The ability of a biological sample to reduce the ferric ion to ferrous ion acting as a reducing agent is determined by FRAP assay (Alzahrani et al., 2012). The reducing capacity of peptides was noted to be higher than the purified protein (Table 2). The difference in reducing capacity may be due to the specific composition of amino acid and the smaller size of peptides than the high molecular weight of protein. The results reveal that peptides act as good electron donors and are strong reducing agents.

      The ABTS assay is one of the most frequently used analytical strategies for antioxidant activity. The present study reported ABTS scavenging activity of peptides to be higher than protein (Table 2), which may be because of the amino acid side chain, chain length, and hydrophobicity. The amino acid composition of protein hydrolysate is also an important factor contributing to its antioxidant activities (Ulagesan et al., 2018).

      However, the bioactive properties of crude honey were found to be much higher than the purified protein and the peptides. The higher antioxidant property of crude honey may be contributions of other bioactive molecules in honey such as phenolics, flavonoids, ascorbic acid, enzymes such as catalase and peroxidase (Habib et al., 2014).

      The protein extracted from Litchi chinensis honey (monofloral) was a monomer of the major protein (MRJP) present in honey. The isolated major protein, 55 kDa was identified as MRJP1, SDS-PAGE examination of which confirmed it to be a monomer. The purified protein had pI of 5.5 and the N-terminal sequence suggested the protein to be hydrophilic in nature. Moreover, the protein, upon digestion with trypsin yielded hydrolysates or peptides with significant antioxidant activity. The small size, hydrophobicity, specific amino acid composition, molecular weight and chain length are factors responsible for the antioxidant activity of peptides, which if orally available can be used for preventing and treating chronic diseases resulting due to oxidative stress. The isolated protein confirms its non addictive nature. Thus, honey can be recommended as one of the food ingredients for regular consumption.

      The authors gratefully acknowledge the Council of Scientific & Industrial Research [Grant no. 38(1355)/13/EMR-II dated 14.02.2013], Govt. of India, New Delhi, India; for the financial support provided to Ms. Debalina Bose for the grant of her fellowship.