Phytochemical and Bioactivity Studies on Hedera helix L. (Ivy) Flower Pollen and Ivy Bee Pollen

Bee pollen, known as a ‘life-giving dust’, is a product of honeybees using flower pollen grains and combining them with their saliva secretions. Thus, flower pollen could be an indicator of the bee pollen botanical source. Identification of bee pollen sources is a highly crucial process for the evaluation of its health benefits, as chemical composition is directly related to its pharmacological activity. In this study, the chemical profiles, contents of phenolic marker compounds and pharmacological activities of Hedera helix L. (ivy) bee pollen samples from Türkiye and Slovenia, as well as ivy flower pollen grains, were compared. High-performance thin-layer chromatography (HPTLC) analyses revealed that pollen samples, regardless of where they were collected, have similar chemical profiles due to the fact that they have the same botanical origins. Marker compounds afzelin, platanoside and quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside, common to both bee pollen and flower pollen, were isolated from bee pollen, and their structures were elucidated by nuclear magnetic resonance (NMR) and mass spectrometry (MS). These three compounds, as well as chlorogenic acid and 3,5-dicaffeoylquinic acid (found in flower pollen), were quantified using high-performance liquid chromatography (HPLC) analyses. In vitro tests and effect-directed analyses were used to evaluate the xanthine oxidase inhibition and antioxidant activity of the marker compounds and extracts from flower pollen and bee pollen. This is the first report comparing chemical profiles and related bioactivities of the flower pollen and bee pollen of the same botanical origin, as well as the first report of the chemical profile and related bioactivities of ivy flower pollen.


Introduction
Hedera helix L. (ivy, Araliaceae) is an evergreen perennial plant that blooms from September to November in the northern hemisphere and is naturally grown or cultivated around the world. Its leaves are rich in saponins (e.g., hederacoside C) and phenolic compounds (chlorogenic acid, 3,5-caffeoylquinic acid, 3,4-caffeoylquinic acid, rutin, hyperoside, etc.), and they have a medical importance [1]. The pharmaceutical products containing standardized extract of ivy leaves are commonly sold over-the-counter as an expectorant for relieving respiratory tract infections [2]. Apart from the research on leaves, there is limited scientific research on the chemical composition and related bioactivity of other plant parts. There are only two reports on phenolic compounds in ivy flowers [3,4]. In addition to ivy flowers, phenolic compounds were also investigated in ivy fruits [3,4] and also in ivy leaves [4]. Chlorogenic acid [4], ferulic acid [3,4], p-coumaric acid [3], isoquercitrin [4], rutin (rutoside) [3,4], quercetin (quercetol) [3,4], and kaempferol [3] were found in flowers [3,4] and fruits [3,4], while quercitrin [3] was found only in flowers [3]. One of these reports [4] spectroscopy and mass spectrometry (MS)) of the marker compounds from bee pollen; (3) development and validation of the high-performance liquid chromatography (HPLC) method for the quantification of the marker compounds in flower pollen and bee pollen; (4) determination of the antioxidant activity and inhibition of xanthine oxidase (XO) by the extracts from flower pollen and bee pollen; (5) HPTLC-bioautographic analyses and evaluation of antioxidant activity and inhibition of XO by the marker compounds, as well as the extracts from flower pollen and bee pollen.

Samples
Ivy (H. helix) flowers were collected in the forest in Bayramiç (Türkiye), and androecia, containing anthers composed of pollen sacs, were separated from the flowers. Bee pollen samples were obtained from the professional beekeepers who placed their beehives in Ordu (Türkiye; BP-TR) and Hrastnik (Slovenia; BP-SI) districts. Palynological analysis was applied according to the standard methodology [25]. Investigated bee pollen samples were classified according to Barth [26] as dominant pollen (>45%), secondary pollen (15-45%), important minor pollen (3-15%) and minor pollen (<3%). Based on this classification ivy pollen was found in the bee pollen samples from Ordu and Hrastnik in the proportion of 88.7% and 92.3%, respectively. All samples were kept at −20 • C.

Preparation of Standard Solutions
All standard solutions were prepared in methanol. Stock solution of standards (chlorogenic acid and 3,5-dicaffeoylquinic acid) and isolated compounds (quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside, afzelin, and platanoside) were prepared at a concentration of 200 µg/mL. Equal volumes (500 µL) of stock solutions of all five compounds were mixed to obtain a standard mixture (MIX) for HPTLC analyses.
Then, these seven working standard solutions were mixed together (from the lowest to highest concentration) to prepare standard mixtures for the calibration curves (0.5-50 µg/mL for the first three and 1-100 µg/mL for the last two compounds).
The following assignments were used to distinguish the STSs and DSTSs of the bee pollen samples based on the collection country (Türkiye-TR and Slovenia-SI): STS-TR, DSTS-TR, STS-SI and DSTS-SI. Assignments used for the STS and DSTS of flower pollen were: STS-FP and DSTS-FP.

Isolation and Structure Elucidation of the Marker Compounds from the Bee Pollen Sample
The bee pollen sample from Slovenia (15 g) was extracted with 80% ethanol (aq) (150 mL) for 30 min in an ultrasonic bath at 40 • C. After filtration, the solvent was evaporated under vacuum at 40 • C to yield crude hydroalcoholic extract (6.96 g). Then, the hydroalcoholic extract was suspended in distilled water (25 mL) and partitioned with ethyl acetate (25 mL × 3). The ethyl acetate fraction (293.8 mg) was applied onto Sephadex LH-20 (95 g) chromatographic column with methanol to afford three compounds ( Figure 1). The structures of these compounds were elucidated by NMR (1D and 2D) and MS.

MS/MS Analyses
The isolated quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside, afzelin and platanoside (1 mg) were first dissolved in methanol (1 mL), and the solutions obtained (1 mg/mL) were further diluted with methanol. Working solutions obtained (0.02 mg/mL) were analyzed by means of an LTQ Velos mass spectrometer with dual-pressure linear ion trap mass analyzer (Thermo Fisher Scientific, Waltham, MA, USA) using a heated electrospray ionization source (HESI) in negative ion mode to ionize the compounds. MS parameters were as follows: flow rate 10 µL/min, heater temperature 200 • C, capillary temperature 350 • C, sheath gas 60 a.u. (arbitrary units), auxiliary gas 10 a.u., sweep gas 0 a.u., spray voltage 2.5 kV, S-Lens RF level 69% and capillary voltage 38.8 V [16]. MS spectra were obtained in the range of 100-2000 m/z. The fragmentation of parent ions was performed with 35% collision energy and the isolation width of 1.0 m/z. Collected data were evaluated using Xcalibur software (version 2.1.0, Thermo).

HPTLC Analyses
MIX and SSTs (20 mg/mL, 2 μL and 5 μL) of bee pollen samples (STS-TR and STS-SI) and flower pollen sample were applied on 20 cm × 10 cm glass backed HPTLC silica gel 60 F254 plates (Merck, Art. No. 1.05642) with a semi-automatic applicator Linomat 5 (Camag, Muttenz, Switzerland) equipped with a 100 μL Hamilton syringe. Applications were performed as 8 mm bands (15.4 mm apart), 8 mm from the bottom of the plate and 15 mm from the left edge. The plate was developed up to 7 cm in a saturated (20 min) twin-trough chamber (20 cm × 10 cm, Camag) with a developing solvent, ethyl acetatedichloromethane-acetic acid-formic acid-water (100:25:10:10: [27]. After drying the plate in a stream of cold air, the plate was heated on a TLC plate heater (Camag) at 100 °C for 3 min and immersed into NP derivatization reagent and after drying also into PEG 400 derivatization reagent [28] by a Chromatogram Immersion Device III (Camag) for 3 s. Documentation of the plate images was performed using a Visualiser (Camag) after development (at 254 nm and 366 nm), after derivatization with NP reagent (at 366 nm) and after enhancement of the fluorescence by PEG 400 reagent (at 366 nm).

HPTLC Analyses
MIX and SSTs (20 mg/mL, 2 µL and 5 µL) of bee pollen samples (STS-TR and STS-SI) and flower pollen sample were applied on 20 cm × 10 cm glass backed HPTLC silica gel 60 F 254 plates (Merck, Art. No. 1.05642) with a semi-automatic applicator Linomat 5 (Camag, Muttenz, Switzerland) equipped with a 100 µL Hamilton syringe. Applications were performed as 8 mm bands (15.4 mm apart), 8 mm from the bottom of the plate and 15 mm from the left edge. The plate was developed up to 7 cm in a saturated (20 min) twin-trough chamber (20 cm × 10 cm, Camag) with a developing solvent, ethyl acetate-dichloromethane-acetic acid-formic acid-water (100:25:10:10:11, v/v/v/v/v) [27]. After drying the plate in a stream of cold air, the plate was heated on a TLC plate heater (Camag) at 100 • C for 3 min and immersed into NP derivatization reagent and after drying also into PEG 400 derivatization reagent [28] by a Chromatogram Immersion Device III (Camag) for 3 s. Documentation of the plate images was performed using a Visualiser (Camag) after development (at 254 nm and 366 nm), after derivatization with NP reagent (at 366 nm) and after enhancement of the fluorescence by PEG 400 reagent (at 366 nm). The winCATS program was used to operate all the instruments (Camag, Version 128 1.4.8.2031).

HPLC Analyses
HPLC analysis was performed using the 1260 Infinity HPLC system (Agilent, Waldbronn, Germany), consisting of a quaternary pump, an autosampler, a thermostatted column compartment and a diode array detector (DAD). The HPLC system was operated by ChemStation software. HPLC analysis was carried out on a Zorbax RP18 Column (4.6 mm × 250 mm I.D., 5 µm particle size, Agilent). The column temperature was set to 25  [29]. The validated method was then applied for the quantification of all five compounds investigated in DSTSs.

DPPH• Assay
Equal volumes (20 µL) of the DSTSs, trolox standard solutions (3.125-100 µg/mL) and methanol (as a blank) were placed in separate wells of a 96-well microplate, which was followed by the addition of DPPH solution (0.1 mM, 280 µL) to each well. After incubation in the dark at room temperature for 30 min, the absorbance was measured at 530 nm [30].
2.9.4. ABTS Assay ABTS assay [33] was slightly modified and then applied. Equal volumes (20 µL) of the DSTSs, trolox standard solutions (6.25-100 µg/mL) or methanol (as a blank) were placed in separate wells of a 96-well microplate plate, and ABTS reagent (280 µL) was added. The absorbance was measured at 734 nm, after incubating the well plate for 6 min at room temperature.
2.9.5. Xanthine Oxidase Inhibitory Activity XO inhibitory activity was carried out as described in Ref. [34] with minor changes. Sodium phosphate buffer (50 mM, pH 7.5, 75 µL) was placed in the wells of a 96-well microplate; this was followed by the DSTSs in different concentrations (25 µL); followed by freshly prepared XO solution (0.2 U/mL in phosphate buffer solution, 25 µL); followed by water (25 µL). Incubation of the reaction mixture at 37 • C for 15 min was followed by the addition of a substrate solution (0.15 mM xanthine, 50 µL) into the mixture and incubation at 37 • C for 30 min until the reaction was terminated by the addition of hydrochloric acid (0.5 M, 50 µL). Finally, the absorbance was measured at 290 nm against a blank, containing all reagents, except the enzyme solution. Allopurinol at different concentrations in the range from 15 to 200 µg/mL was used as a positive control.
2.9.6. Superoxide Radical Scavenging (SOD) Activity SOD activity was performed as described in Ref. [27]. DSTSs (20 µL) and allopurinol (10-200 µg/mL; used as the positive control) were added in separate wells of a 96-well microplate; this was followed by the substrate solution (mixture of 0.4 mM xanthine and 0.24 mM NBT, 80 µL); followed by XO solution (50 mU/mL in sodium phosphate buffer (pH 7.5), 80 µL). After incubation at 37 • C for 20 min the reaction was terminated by hydrochloric acid (0.6 M, 80 µL) and absorption was measured at 560 nm against a blank solution containing all reagents except XO. The HPTLC plates developed as described in Section 2.7. HPTLC analyses were dipped into DPPH solution (0.1%) for 3 s using a Chromatogram Immersion Device III (Camag). The plates were left to dry in the air in the dark and were documented under white light after 30 min. The compounds having antioxidant activity appeared as yellow bands on the purple background.
The image of the HPTLC plate captured at white light illumination after HPTLC-DPPH analyses was for image analyses converted to a different format using WinCATS software (Camag) and then converted to videodensitograms in fluorescence mode using VideoScan TLC/HPTLC Evaluation Software (Version 1.02.00) (Camag).

HPTLC-Xanthine Oxidase Inhibitory Activity
The XO bioautography assay was carried out as described in Ref. [35]. The HPTLC plate developed as described in Section 2.7. HPTLC analyses was immersed into a derivatization chamber with a mixture of phosphate buffer solution (pH 7.6) containing EDTA (1 mM), NBT (1 mM) and XO (0.1 U/mL), and incubated at 37 • C for 30 min in the dark. After drying in the air, the HPTLC plate was dipped into a phosphate buffer solution containing xanthine (1.5 mM) and incubated at 37 • C for 30 min in the dark. After drying in the air, the HPTLC plate was documented at white light in RT mode immediately (0 min), as well as at 5, 10, 15, 20, 30, 45, 60, 90, and 120 min, subsequently. The compounds having XO inhibitory activity were detected as white/yellow zones on a purple background. Allopurinol (8 µg on the plate) was used as the positive control.

Statistical Analyses
Each assay for bioactivity and quantitative analysis was repeated thrice. The average values and standard deviations (SD) were calculated by using Microsoft Excel 2013, and the results were expressed as average value ± SD. The analytical data obtained from tests were evaluated by using one-way analysis of variance (ANOVA). Tukey's test was applied to appraise the differences (p < 0.05) by Minitab 17.

Results and Discussion
Chemical profiling of the STS-TR and STS-SI prepared from ivy bee pollen samples collected in Türkiye and Slovenia was first performed on the HPTLC silica gel F 254 plates using NP detection reagent followed by PEG detection reagent. The most intense bands obtained at the same R F values in both STS-TR and STS-SI presenting unidentified marker compounds were selected for isolation and evaluation of their structures.

HPTLC Chemical Profiling
The chemical fingerprintings of the STS-TR and STS-SI prepared from the bee pollen were comparatively investigated with the STS prepared from the ivy flower, standards and isolated compounds on HPTLC silica gel F 254 plates before (at 254 nm ( Figure 2A) and at 366 nm ( Figure 2B)) and after derivatization with NP detection reagent ( Figure 2C) followed by enhancement and stabilization of the fluorescence with PEG detection reagent ( Figure 2D). As shown in Figure 2D, STS-TR (tracks 8 and 11) and STS-SI (tracks 9 and 12) from bee pollen showed similar profiles with the most intensive chromatographic zones at the R F values of the isolated compounds: quercetin-3-O-β-glucopyranosyl-(1→2)-βgalactopyranoside (track 1, orange-colored zone; R F ≈ 0.06), afzelin (track 3, green-colored zone; R F ≈ 0.52) and platanoside (track 5, green-colored zone; R F ≈ 0.84). At the same R F values chromatographic zones with the same color were detected also in the chemical profile of the STS of flower pollen ( Figure 2D, tracks 7 and 10), but with higher intensity than in the profiles of the STS-TR and STS-SI. The most intensive chromatographic zones in the STS of flower pollen profile ( Figure 2D, tracks 7 and 10) were found at the R F values of chlorogenic acid ( Figure 2D, track 2, light blue-colored zone; R F ≈ 0.22) and 3,5-dicaffeoylquinic acid ( Figure 2D, track 4, light blue-colored zone; R F ≈ 0.58), which were used as the standards. Although chlorogenic acid and 3,5-dicaffeoylquinic acid were dominant in the flower pollen, they were not detected in bee pollen samples. The qualitative differences between ivy flower pollen and ivy bee pollen chemical profiles may be due to the enzymes present in the bee saliva, resulting in alteration of the compounds. Afzelin, platanoside and quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside were detected in ivy flower pollen and ivy bee pollen for the first time and can be considered as marker compounds.
inant in the flower pollen, they were not detected in bee pollen samples. The qualitative differences between ivy flower pollen and ivy bee pollen chemical profiles may be due to the enzymes present in the bee saliva, resulting in alteration of the compounds. Afzelin, platanoside and quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside were detected in ivy flower pollen and ivy bee pollen for the first time and can be considered as marker compounds.
Although bee pollen samples were from different countries, their similar chemical profiles indicate that phenolic compounds are directly related with the plant source used by the honeybee. As a consequence, HPTLC chemical profiling in this study may encourage and support future studies focusing on botanical identification of bee pollen samples.

Specificity
The identity of each investigated compound in the samples was verified by comparing its retention time (tR) with the tR of the corresponding standard ( Figure 3) and Although bee pollen samples were from different countries, their similar chemical profiles indicate that phenolic compounds are directly related with the plant source used by the honeybee. As a consequence, HPTLC chemical profiling in this study may encourage and support future studies focusing on botanical identification of bee pollen samples.

HPLC Method Validation Specificity
The identity of each investigated compound in the samples was verified by comparing its retention time (t R ) with the t R of the corresponding standard ( Figure 3) and overlaying the UV spectrum of each of the compounds investigated with the spectrum of the corresponding standard. The retention times were as follows: 4.6 for chlorogenic acid; 6.8 for quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside; 12.1 for 3,5-dicaffeoylquinic acid; 13.2 for afzelin; 24.3 for platanoside. The specificity of the method was evaluated by comparison of the chromatogram of each of the compounds investigated with the chro-matogram of the blank. Chromatographic peaks for the compounds investigated were not detected in the blank chromatogram, which confirmed the specificity of the method.
overlaying the UV spectrum of each of the compounds investigated with the spectrum of the corresponding standard. The retention times were as follows: 4.6 for chlorogenic acid; 6.8 for quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside; 12.1 for 3,5dicaffeoylquinic acid; 13.2 for afzelin; 24.3 for platanoside. The specificity of the method was evaluated by comparison of the chromatogram of each of the compounds investigated with the chromatogram of the blank. Chromatographic peaks for the compounds investigated were not detected in the blank chromatogram, which confirmed the specificity of the method.

Linearity of the Calibration Curve, Limit of Detection (LOD), and Limit of Quantification (LOQ)
To establish the calibration curve, seven concentration levels for each standard compound were analyzed in triplicate. The detailed data for the linearity of the calibration curves are presented in Table 2. The obtained correlation coefficient values (r 2 ) for the calibration curves are greater than 0.990. LOD and LOQ values ( Table 2) were calculated as 3 × (SD/S) and 10 × (SD/S), respectively.  To establish the calibration curve, seven concentration levels for each standard compound were analyzed in triplicate. The detailed data for the linearity of the calibration curves are presented in Table 2. The obtained correlation coefficient values (r 2 ) for the calibration curves are greater than 0.990. LOD and LOQ values ( Table 2) were calculated as 3 × (SD/S) and 10 × (SD/S), respectively.

Precision
The intraday precision of the HPLC method was evaluated by using a standard mixture solution containing: 5 µg/mL of chlorogenic acid, quercetin-3-O-β-glucopyranosyl-(1→2)β-galactopyranoside and platanoside; 10 µg/mL of 3,5-dicaffeoylquinic acid and afzelin. The mixture was analyzed in three consecutive runs at three different times during the same day. Relative standard deviation (RSD) values are indicated in Table 3. The interday precision was evaluated by repeating the analysis of a standard mixture solution used for examining intraday precision three times on three consecutive days. The intraday precision and interday precision should not exceed 5%. The relative standard deviation values were found in the ranges of 0.019-0.752 for intraday precision and 0.045-0.430 for interday precision ( Table 3), indicating that the values found fit the criteria.

Quantitative Analyses
The developed and validated HPLC method was used for the quantitative analyses of the investigated compounds in bee pollen and flower pollen samples. As can be seen in the HPLC chromatograms (Figure 3), all compounds investigated were found in the flower pollen sample ( Figure 3B), while chlorogenic acid and 3,5-dicaffeoylquinic acid were not detected in the bee pollen samples ( Figure 3C,D). This confirmed the results of the HPTLC analyses ( Figure 2). The results of quantitative analyses are presented in Table 5. Quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside was detected as the dominant compound in the bee pollen samples (≈29 mg/g for BP-TR, ≈24 mg/g for BP-SI), however, this was not the case for the flower pollen sample (Table 5), where much lower concentration (≈4 mg/g) was determined. The dominant compound in the flower pollen sample was 3,5-dicaffeoylquinic acid (≈22 mg/g), which was not detected in the bee pollen samples (Table 5). Both bee pollen samples contained similar concentrations (≈2 mg/g) of afzelin, which concentration was about three times higher in the flower pollen sample (Table 5). In terms of platanoside content a remarkable difference was found between bee pollen samples (≈2 mg/g for BP-TR, ≈7 mg/g for BP-SI).  Table 5. Contents of compounds investigated in bee pollen (BP-TR and BP-SI) and flower pollen (FP) samples. This is the first report on the content of the phenolic compounds in ivy flower pollen. This report compliments the only two existing reports on the contents of phenolic compounds in ivy flowers [3,4], which include the following phenolic compounds: chlorogenic acid [3], ferulic acid [3], p-coumaric acid [3,4], kaempferol, quercetin (quercetol) [3,4], rutin (rutoside) [3,4], quercitrin [3] and isoquercitrin [3].

In Vitro Bioactivity Analyses
3.4.1. Antioxidant Activity Determined by DPPH, FRAP, ABTS, and CUPRAC Assays DPPH, FRAP, ABTS and CUPRAC assays revealed that phenolic acids (chlorogenic acid and 3,5-dicaffeoylquinic acid) have much higher antioxidant activity than the other compounds investigated (Table 6). Of these four assays the highest antioxidant activity was determined for chlorogenic acid, which was followed by 3,5-dicaffeoylquinic acid and quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside (Table 6). Afzelin and platanoside had much lower antioxidant activity. As expected, considering the content of phenolic compounds investigated (Table 5), the DSTS of the flower pollen sample had a higher content of investigated phenolic compounds than the DSTSs of the bee pollen samples (Table 5). It also had a higher antioxidant activity than the DSTSs of the bee pollen samples (Table 6). DPPH, FRAP, ABTS and CUPRAC assays showed that the DSTS-SI had higher antioxidant activity than the DSTS-TR (Table 6).

Xanthine Oxidase Inhibitory Activity and Superoxide Radicals Scavenging Activity
Among the compounds investigated, chlorogenic acid, afzelin and 3,5-dicaffeoylquinic acid showed mild-to-moderate XO inhibitory activities, with IC 50 values ranging from 12.7 to 17.5 µg/mL (Table 7). XO inhibitory activities of chlorogenic acid and 3,5-dicaffeoylquinic acid were also reported by other authors [42][43][44]. Quercetin-3-O-β-glucopyranosyl-(1→2)β-galactopyranoside exhibited low XO inhibitory activity in this study, while platanoside and hydroalcoholic extracts of bee pollen samples had no XO inhibitory activity at the applied concentrations (Table 7). Although both afzelin and platanoside are kaempferol derivatives, afzelin showed XO inhibitory activity, while platanoside showed no activity, which could be due to the bulk groups of coumaric acid esters on platanoside tending to reduce the affinity towards XO. The enzymatic method was also used to determine SOD activity. Compared to allopurinol as the positive control (IC 50 3.0 µg/mL), a significant SOD activity was observed for chlorogenic acid (IC 50 1.6 µg/mL) ( Table 7). Platanoside showed no SOD activity. The rest of the compounds investigated, as well as hydroalcoholic extracts of the flower and bee pollen samples, showed mild to moderate SOD activity with IC 50 values ranging from 5 to 63 µg/mL (Table 7).

HPTLC-DPPH Analyses of Antioxidant Activity
HPTLC-DPPH analyses of the compounds investigated and the STSs of flower pollen and bee pollen samples were performed to obtain additional information about the antioxidant activity of the separated compounds on the HPTLC plate. Such information could not be obtained for the STSs of the flower pollen and bee pollen samples by spectrophotometric assays. HPTLC-DPPH analyses and all spectrophotometric analyses performed in this study showed that phenolic acids have a much higher antioxidant activity than flavonoids. This observation is evident in Figure 4, where the highest intensity of the yellow-colored chromatographic zone was obtained for chlorogenic acid (Figure 4, track 2) and slightly lower intensity was obtained for 3,5-dicaffeoylquinic acid (Figure 4, track 4) at the same amount (0.4 µg) applied on the plate. The yellow zones of the STS of flower pollen (Figure 4, tracks 7 and 10) were ranked on intensity: the most intense zone was at R F ≈ 0.58 for 3,5-dicaffeoylquinic acid (Figure 4, track 4), followed by the zone at the R F ≈ 0.22 for chlorogenic acid (Figure 4, track 2), followed by other less intense zones at R F ≈ 0.   The image of the HPTLC plate ( Figure 4) was converted to videodensitograms ( Figure 5) in fluorescence mode. The videodensitograms of the STSs of bee pollen and flower pollen samples ( Figure 5B) show peaks at R F values of the compounds isolated ( Figure 5A, R F ≈ 0.06 for quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside, R F ≈ 0.52 for afzelin and R F ≈ 0.83 for platanoside). The videodensitograms of the STS from the flower pollen showed additional peaks at R F values of chlorogenic acid (R F ≈ 0.22) and 3,5dicaffeoylquinic acid (R F ≈ 0.58) ( Figure 5B).    (Figure 4, track 10) samples. The total peak area obtained for the STS of flower pollen was significantly higher than the total peak areas obtained for the STSs of bee pollen ( Figure 6). The lowest total peak area was obtained for the STS of bee pollen from Türkiye.

HPTLC-XO Inhibitory Activity
Evaluation of the XO inhibitory activity was performed on HPTLC plates for all compounds investigated and sample test solutions (50 mg/mL) of flower pollen and bee pollen samples, which were all applied in higher concentrations than for HPTLC-NP-PEG and HPTLC-DPPH analyses. This could be because XO inhibitory activity is directly related to enzymatic activation. In HPTLC-XO inhibitory activity assay, the developed plate was dipped into a phosphate buffer solution containing xanthine and after incubation and drying was documented immediately (t = 0 min) and after 5,10,15,20,30,45, 60, 90 and 120 min. Within this documentation interval, the plate was stored in the dark. The XO inhibitors were detected as white/yellow zones on a purple background (Figure 7). The positive control allopurinol appeared on the HPTLC plate immediately as a white zone, which lasted for 120 min (Figure 7, track 1).
The videodensitograms ( Figure 5) were used for the image analysis of the tracks of the compounds investigated and the tracks with the highest application of the STSs bee pollen (Figure 4, tracks 11 and 12) and flower pollen (Figure 4, track 10) samples. The total peak area obtained for the STS of flower pollen was significantly higher than the total peak areas obtained for the STSs of bee pollen ( Figure 6). The lowest total peak area was obtained for the STS of bee pollen from Türkiye.

HPTLC-XO Inhibitory Activity
Evaluation of the XO inhibitory activity was performed on HPTLC plates for all compounds investigated and sample test solutions (50 mg/mL) of flower pollen and bee pollen samples, which were all applied in higher concentrations than for HPTLC-NP-PEG and HPTLC-DPPH analyses. This could be because XO inhibitory activity is directly related to enzymatic activation. In HPTLC-XO inhibitory activity assay, the developed plate was dipped into a phosphate buffer solution containing xanthine and after incubation and drying was documented immediately (t = 0 min) and after 5, 10, 15, 20, 30, 45, 60, 90 and 120 min. Within this documentation interval, the plate was stored in the dark. The XO inhibitors were detected as white/yellow zones on a purple background (Figure 7). The positive control allopurinol appeared on the HPTLC plate immediately as a white zone, which lasted for 120 min (Figure 7, track 1).
Several yellow zones were detected in the sample test solution of flower pollen (Figure 7, tracks 7 and 11) and were ranked on intensity and the time of their detection: the most intense zone at R F ≈ 0.57 of 3,5-dicaffeoylquinic acid (Figure 7, track 5) was detected at 0 min; this was followed by the zone of an unknown compound at R F ≈ 0.25 at 90 min, which was not yellow earlier; followed by the zone at R F ≈ 0.65 that appeared as a pale yellow at 15 min and its intensity increased after 30 min. The zones in-between R F ≈ 0.3 and R F ≈ 0.55 were not detected intensely until after 30 min (Figure 7, tracks 7 and 11). In the case of bee pollen samples ( Figure 7, tracks 8, 9, 12 and 13), weak yellow zones appeared immediately (0 min) at R F ≈ 0.05 for quercetin-3-O-β-glucopyranosyl-(1→2)-βgalactopyranoside (Figure 7, track 2). At the R F value of platanoside (Figure 7, track 6) yellow zones appeared in all tracks of the sample test solutions (Figure 7, tracks 7-13). However, the occurrence of these yellow zones appearing at the same R F value as platanoside could have resulted from other unknown compounds contributing to the activity.
It should be highlighted that enzymatic reactions are time-dependent, so documentation of the plate in certain time intervals is significant for the evaluation of the bioactive compounds. This is the first study emphasizing the importance of the time when the plate images in HPTLC-XO inhibitor analyses are captured. It should be noted that the time of capture of the plate images was not provided in the literature [35].  Spectrophotometric assays (Section 3.4.2. Xanthine Oxidase Inhibitory Activity and Superoxide Radicals Scavenging Activity) of the samples showed no XO inhibitory activity (Table 7). However, in HPTLC-XO analyses the contribution of both the extracts and the compounds in the extracts to the bioactivity was displayed. The reason for this difference is directly related to the sample concentration applied for analysis, which was much lower in the spectrophotometric analyses than in HPTLC-XO analyses. Testing samples in high concentration is the advantage of HPTLC-bioautography compared to spectrophotometric assays.
XO inhibitory activity of the compounds investigated obtained by HPTLC-XO was found to be comparable to the results of spectrophotometric tests (Table 7). Consequently, the compounds investigated appeared as intense yellow-colored zones on a purple background, except platanoside, which had a weak yellow zone.

Conclusions
In this study, bee pollen samples collected from Türkiye and Slovenia were evaluated. Palynological analysis was first used to identify their botanical source, which was ivy flower pollen. Then, the chemical profiles and pharmacological activities of the bee pollen samples, together with their botanical source, ivy flower pollen grains, were further investigated. HPTLC profiles of bee pollen samples revealed that samples of the same botanical origin exert similar chemical composition independent of where they were collected. Quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside, afzelin and platanoside were found as marker compounds for the identification of bee pollens from ivy flower. This study showed that HPTLC profiling could be an alternative to palynological analysis, and this approach can be applied to future studies of bee pollen. Among the marker compounds, quercetin-3-O-β-glucopyranosyl-(1→2)-β-galactopyranoside was found with the highest concentration in bee pollen samples, but not in the flower pollen. This difference may be the result of how honeybees produce bee pollen. It is interesting that quercetin-3-Oβ-glucopyranosyl-(1→2)-β-galactopyranoside was the most bioactive one among marker compounds. It showed the highest antioxidant activity by in vitro DPPH, CUPRAC, FRAP and SOD activity tests. Additionally, it showed the highest XO inhibitory activity. HPTLCbioautography (HPTLC-DPPH and HPTLC-XO) also confirmed its contribution to the bioactivity of the extract. To the best of our knowledge, this is the first report comparing chemical profiles and related bioactivities of the flower pollen and bee pollen of the same botanical origin, as well as the first report of the chemical profile and related bioactivities of ivy flower pollen. This study is important as the determination of the botanical sources of bee pollen should be taken into consideration for the standardization of bee pollen extracts. Therefore, more research on marker compounds in bee pollen, as well as on flower pollen and bee pollen of the same biological origin, is needed. The findings of this study can be applied to apitherapy and manufacturing of bee pollen-based food supplements.