Several species of the genus Crataegus L. (Hawthorn) have been reported to possess a wide range of pharmacological actions on the cardiovascular system . Preparations of hawthorn including leaf, flower, and berry have been used traditionally in minor forms of coronary heart disease , heart failure  and cardiac arrhythmia , but the only approved monograph by German Commission E is hawthorn leaf with flower with the unique indication of decreasing cardiac output, classified by the New York Heart Association (NYHA) as stage II . Also, the approved monograph of hawthorn by the German Pharmacopoeia (DAB 1997) is hawthorn leaf with flower (Crataegi folium cum flore) which consists of dried approximately 7 cm long, flowering twig tops of C. monogyna Jacq. or C. laevigata (Poir.) DC. (syn.: C. oxyacantha L.) and, more seldom, of other Crataegus species, such as C. pentagyna Waldst. et Kit. ex Willd., C. nigra Waldst. et Kit. and C. azarolus L. . Hawthorn is known for itspolyphenolics among which the major active constituents are: Flavonoids (1-2%) including hyperoside and rutin as flavonol-O-glycosides (Figure 1A); vitexin (Figure 1B), vitexin-2”-O-rhamnoside and acetylvitexin-2”-O-rhamnoside as flavone-C-glycosides; proanthocyanidins (1-3%)including oligomeric procyanidins (n=2 to n=8 catechins and/or epicatechins), especially dimeric procyanidin B-2 (Figure 1C); and phenol carboxylic acids (PCAs) including caffeic acid and chlorogenic acid (Figure 1D and 1E) . Hawthorn species have different flavonoid compositions, and there are also qualitative and quantitative differences in the flavonoid compositions in the flowers, leaves and fruits of the same plants . According to the DAB 1997 monograph, the flowering tops of the official medicinal species should contain at least 0.7% flavonoids, calculated as hyperoside, determined by the aluminum chloride spectrophotometric method, and major polyphenolics of rutin and hyperoside as flavonoids and chlorogenic acid and caffeic acid as PCAs characterized by a thin-layer chromatography (TLC) method .
Figure 1. (A): Flavonol-O-glycoside: hyperoside (R=β D-galactosyl) and rutin (R=β-D rutinosyl);(B):Flavone-C-glycoside:Vitexin(R=βD-glucosyl);(C):ProcyanidinB-2; (D):Caffeicacid and (E): Chlorogenic acid (3-O-caffeoyl quinic acid).
In the present study, Iranian Crataegus spp. (Table 1) were studied analytically to evaluate the presence of two requirements mentioned by DAB 1997 in order to standardize them as official medicinal species. All of the species studied, except C. pseudo-heterophylla, contained flavonoid contentequal to or greater than 0.7 percent in flowering tops (Table 2). Also, all of the leaf samples, except leaves of C. pseudohetero-phylla, and only the flowers of C. curvicepala and C. pseudoheterophylla contained flavonoid contents greater than 0.7% (Table 2). High-performance thin-layer chromato-graphic (HPTLC) fingerprinting combining digital scanning profiling was developed to identify the major polyphenolics as flavonoids and PCAs in flowering tops, leaves and flowers samples, rather than conventional TLC as a standard pharmacopoeia method.
Figure 2. HPTLC fluorescence images under the excitation wavelength 366 nm of flowering tops (FT), leaves (L) and flowers (F) samples of Crataegus spp. (A): tracks 1-3 represents of reference compounds: rutin (1), chlorogenic acid (2), hyperoside (3); tracks 4-11 represents of FT samples of Crataegus spp.: C. atrosanguinea (4), C. pentagyna (f) (5), C. meyeri (f) (6), C. pentagyna (g) (7), C. curvicepala (8), C. meyeri (g) (9). (B): 1: FT, 2: L and 3: F of C. atrosanguinea; 4: FT, 5: L and 6: F of C. pentagyna (f); 7: FT, 8: L and 9: F of C. meyeri (f). (C): 1: FT, 2: L and 3: F of C. pentagyna (g); 4: FT, 5: L and 6: F of C.curvicepala; 7: FT, 8: L and 9: F of C. meyeri (g). (D): 1: FT, 2: L and 3: F of C. pseudoheterophylla; 4: FT, 5: L and 6: F of C. pseudoheterophylla (two identical species from one location but from diffrent source); 7: FT, 8: L and 9: F of C. microphylla.
Although the principles of TLC and HPTLC methods are identical, because of the use of kinetically optimized fine-particle layers in HPTLC, separation is faster and more efficient and the results are more reliable and reproducible. In combination with digital scanning profiling, HPTLC also provides accurate and precise RF values and quantitative analysis of sample constituents by in situ scanning densitometry aided by theformation of easily detected derivatives by post-chromatographic chemical reactions as required, as well as a record of the separation in the form of a chromatogram with fractions represented as peaks with defined parameters including absorbance (intensity), RF, height and area . Furthermore, the feature of a pictorial fluorescence image of HPTLC coupled with a digital scanning profile is more and more attractive to herbal analysts for constructing a herbal chromatographic fingerprint by means of HPTLC . The main objective of this study was to evaluate and optimize the HPTLC fingerprint method in standardization of Crataegus spp. These HPTLC fluorescence images coupled with the scanning profiles provided adequate information and parameters for comprehensive identification, assessment and comparison of major active constituent fingerprints in the samples studied to serve as a basis for their use in medicinal preparations for cardiovascular diseases.
2. Materials and methods
Eight flowering tops samples from six species of the genus Crataegus were collected from the north, north-west and west of Iran on May 2006 (At first, two flowering tops samples were collected from Sanandaj, Kordestan, Iran and then authenticated as C. pseudoheterophylla and the HPTLCchromatograms of them (Figure 1D) were shown to be identical. Therefore, one of the samples was studied) (Table 1). In each sample, flowering tops and separated leaves and flowers were dried at the room temperature at a relative humidity of 40%. All samples were authenticated by Dr. Gh. Amin (Herbarium of the Faculty of Pharmacy, Tehran University of Medical Sciences, Tehran, Iran, where voucher specimens are deposited) according to their morphological characteristics. Rutin, hyperoside and chlorogenic acid were used as reference substances, diphenylboryloxyethylamine (Natural Product, NP) and polyethylene glycol-4000 (PEG-4000) as spray reagents. Solvents of analytical grade were purchased from Merck (Merck, Darmstadt, Germany).
Figure 3. Typical HPTLC images and corresponding digital scanning profiles of flowering tops samples of Crataegus spp.(A): C. atrosanguinea; (B): C. pentagyna (f); (C): C. meyeri (f); (D): C. pentagyna (g); (E): C. curvicepala; (F): C. meyeri (g); (G): C. pseudoheterophylla and (H): C. microphylla.
2.2. Preparation of reference solution
Reference solutions were: rutin, hyperoside (2.5 mg of each separately) and chlorogenic acid (1 mg) were dissolved in methanol (10 ml).
2.3. Preparation of sample solution
Powdered samples (1 g of flowering tops, leaves, or flowers, separately) were extracted with methanol (10 ml) for 5 min. on a water bath at about 60 °C and then filtered. A total of the 5 ml of the methanolic extract was concentrated to about 2 ml under vacuum; water (1 ml) and ethyl acetate (10 ml) were added and shaken several times. The ethyl acetate phase was separated and reduced in volume (to 1 ml) under vacuum.
Aliquots (2 μl) of the reference and sample solutions were applied separately bandwise by Camag Linomat IV (Camag, Muttenz, Switzerland) with parameters of bandlength: 6 mm, track distance: 3 mm, distance from left edge: 11 mm and low edge: 8 mm, and delivery speed: 15 sec/μl to silica gel 60 F254-precoated HPTLC plates, 10×10 cm (Merck, Darmstadt, Germany). After saturation for 20 min. with the mobile phase vapor, the plates were developed horizontally in Camag horizontal developing chamber (10×10 cm) for 80 mm at the room temperature, i.e. 18-22 °C with ethyl acetate-formic acid-glacial acetic acid-water (100:11:11:26, v/v) as the mobile phase. After heating the plate at 100 °C for 5 min., derivatization of the chromatogram was performed by Camag glass reagent spray by spraying still-hot plate with 1% methanolic NP, followed by 5% methanolic PEG-4000. The plate was observed after 30 min. under UV-366 nm light in Camag UV cabinet and the HPTLC fluorescence image documented. The corresponding digital scanning profiling was carried out with a Camag TLC scanner 3 fitted with winCATS V1.2.3 software at a single-wavelength UV-366 nm. Documen-tation of chromatograms was carried out with a Sony digital camera, 5 Mpixel (Japan).
Figure 4. The quantifiable comparison of 3D graphs of HPTLC fingerprints of flowering tops (A), leaves (B), and flowers (C) samples of Crataegus spp.
Table 1. A summary of tested samples.
Note: * North-West of Iran; ** West of Iran; *** North of Iran
3. Results and discussion
3.1. Interpretation of HPTLC fingerprint chromatograms
The HPTLC images shown in Figure 2 indicate that all reference substances and sample constituents were clearly separated without any tailing and diffuseness. The RF values and fluorescence colors of the reference substances according to Wagner's Atlas of thin-layer chromatography are: Orange fluorescence band with RF 0.3-0.4 as rutin, blue fluorescence band with RF 0.4-0.5 as chlorogenic acid, orange fluorescence band with RF 0.5-0.6 as hyperoside, yellow-green fluorescence band with RF 0.6-0.7 as vitexin and blue fluorescence band with RF 0.8-0.9 as caffeic acid . By comparison of sample constituents with reference substances on the same plate, constituents of sample extracts were identified. As seen in Figure 3, all of the flowering tops samples contained rutin, hyperoside, chlorogenic acid and caffeic acid, except of C. pentagyna (g) in which no rutin was detected on the chromatogram and digital scanning profile (Figures 2C and 3D). Also, all of the leaf and flower samples, except for flowers of C. pentagyna (g) and C. curvicepala, contained four majorpolyphenolics (Figure 2). Rutin in flowering tops and leaf samples of C. curvicepala were hardly detected on the chromatogram (Figure 2C), related to the low concentration in the sample extract, but the corresponding peak was detected on digital scanning profiles (Figure 3E). The fluorescence band corresponding to hyperoside in the flower sample of C. pseudoheterophylla showed much great intensity than the corresponding band in flowering top and leaf samples (Figure 2D) related to much higher concentration as shown by quantitative data in Table 2. However, in other samples, the hyperoside percentages in flower samples were much less than in flowering tops and leaf samples, except for C. curvicepala (Table 2). The highest intensity corresponding to the hyperoside band was detected in samples of C. curvicepala (Figure 2C) supported byquantitative data in Table 2 (about 1.2%). The highest intensity for chlorogenic acid and caffeic acid bands was detected in the flowering top sample of C. microphylla corresponding to their high concentrations in the sample extract (Figure 2D). Vitexin was only detected in flowers and because of that in flowering tops samples (Figures 2A, B and C). The yellow fluorescence bands at greater RF than the hyperoside band were related to other flavonoids and blue fluorescence bands at greater RF than the chlorogenic acid and caffeic acid bands were related to other phenol carboxylic acids (Figure 2).
3.2. Interpretation of digital scanning profiles of HPTLC fingerprint chromatograms
Figure 3 shows HPTLC fluorescence images coupled with digital scanning profiles of flowering top samples of the species studied. Every fraction in each track is represented as a specific peak with defined values of absorbance, RF, height and area. By digital scanning of the chromatogram, the trace quantity of fractions was readily determined as peaks, whereas could not be visually detected on the plate (rutin on Figures 2C and 3E). Since the digital scanning profile was intuitively converted from the HPTLC image, all the peak intensities were in accord with fluorescence bands and their brightness. Thus, it can be easily evaluated by comparison of the polyphenolic content of samples by quantitative comparison of the peak intensities (maximum peak heights) (Figure 4). The hyperoside peak-to-peak ratios expressed by the fingerprint patterns and their integrated peak area values obtained from digital scanning profiling of sample chromatograms (Figure 5) underscored the precision of quantitative data obtained by the spectropho-tometric method (Table 2).
Table 2. Flavonoid content calculated as hyperoside (g/100g dry matter).
Note: (f): Fandogh-Loo; (g): Gardane-Heyran
Although TLC is a conventional method used generally in pharmacopoeias as one of the standardization methods, the HPTLC method is more practical. HPTLC is feasible for development of chromatographic fingerprints to determine major active constituents of medicinal plants. The separation and resolution are much better, and the results are much more reliable and reproducible than TLC. Combined with digital scanning profiling, it has the main advantage of in situ quantitative measurement by scanning densitometry. Furthermore, the colorful pictorial HPTLC image provides extra, intuitive visible color and/or fluorescence parameters for parallel assessment on the same plate. In the present study, the proposed HPTLC fingerprint method combined with digital scanning profiling was used for standardization. In conclusion, the results obtained from flavonoid content determination (Table 2), qualitative evaluation of HPTLC fingerprint images (Figures 2 and 3) and quantitative comparison of hyperoside peak areas (Figure 5) of flowering tops samples of species studied showed that according to the DAB 1997 monograph on hawthorn leaf with flower, C. atrosanguinea, C. pentagyna (f), C. meyeri (f,g), C. curvicepala and C. microphylla contained two Pharmacopoeial requirements for standardization and can be introduced as new official medicinal species for the preparation of effective herbal cardiovascular medicines. C. curvicepala is of special note with the highest hyperoside content (about 1.2%) which means that it could serve as a suitable alternative to the conventional official species, i.e. C. monogyna and C. laevigata. Also, all of the leaf samples, except for C. pseudoheterophylla, and the flowers of C. pseudoheterophylla met the two Pharma-copoeial requirements for standardization and can be used as medicinal parts.
Figure 5. Comparison of hyperoside integrated max peak area values of flowering tops (A), leaves (B), and flowers (C) samples between Crataegus spp. HPTLC fingerprints.
This research was funded by Soha Research and Development Laboratory dependent to Red Crescent Organization of the Islamic Republic of Iran. The authors also would like to thank Prof. K.T. Douglas from School of Pharmacy, Manchester University, for editing the manuscript.