Supporting Information Unique proline-benzoquinone pigment from the colored nectar of bird s Coca cola tree functions in bird attraction Shi-Hong Luo,, Yan Liu, Juan Hua, Xue-Mei Niu, Shu-Xi Jing, Xu Zhao, Bernd Schneider, Jonathan Gershenzon, Sheng-Hong Li *, State Key Laboratory of Phytochemistry and Plant Resources in West China, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, P. R. China, Key Laboratory for Conservation and Utilization of Bio-resources, Yunnan University, Cui Hubei Road 2, Kunming 650091, P. R. China, Max Planck Institute for Chemical Ecology, Beutenberg Campus, Hans-Knöll-Straße 8, D-07745 Jena, Germany and Graduate University of Chinese Academy of Sciences, Beijing 100039, P. R. China E-mail: shli@mail.kib.ac.cn S1
General experimental details: UV-visible spectral data were obtained on a Shimadzu-210A double-beam spectrophotometer. Optical rotations were measured on a Horiba-SEAP-300 spectropolarimeter. CD experiments were performed on a Chirascan circular dichroism spectrometer. The IR spectrum was recorded on a Bruker-Tensor-27 spectrometer with KBr pellets. NMR experiments were carried out on either a DRX-500 or an Avance III 600 spectrometer with TMS as internal standard. Assignments of each proton and carbon signals were made by two dimensional NMR ( 1 H- 1 H COSY, HSQC and HMBC) experiments. MS were recorded on a Waters Auto Spec Premier P776 spectrometer or Xevo TQ-s spectrometer or Api Qstar Pulsar I spectrometer. HPLC analyses were performed on an Agilent 1200 series instrument equipped with a quaternary pump, a vacuum degasser, an autosampler, a thermostated column compartment and a diode array detector. Column chromatography was performed on MCI gel (75-150 µm, Mitsubishi Chemical Corporation, Tokyo, Japan), and Lichroprep RP-C 18 gel (40-63 µm, Merck, Darmstadt, Germany). Medium-pressure liquid chromatography was performed on a Buchi C-605 series instrument. Material: Nectar of L. canum for the initial chemical analysis was collected from Dehong prefecture, one of the native distribution sites of the plant, and the Botanical Garden of the Kunming Institute of Botany, in December 2008. Because no obvious difference was observed between the nectar HPLC profiles of the two collection sites, and severe degradation of the major peak during storage at 4 C was observed, all samples for the following studies were collected with Eppendorf pipettes from the Botanical Garden of Kunming Institute of Botany in the morning (starting at 8:40) of the blooming season (December to March) between 2009 and 2011, and were immediately stored at -80 C until use. During the sample collection period, each inflorescence was bagged to avoid foraging by visitors. Isolation and purification of DPBQ: 645 ml of nectar was directly chromatographed on an open MCI gel column eluted with water and methanol. The water fraction (105.4 g) was repeatedly subjected to medium-pressure liquid chromatography on a MCI gel column using an eluent of methanol/water (from 0:1 to 1:10) to afford six subfractions A-F. Subfraction A (45.2 g) was purified on a RP-C 18 gel column eluted with water to yield nectar sugar (probably also containing some amino acids), which was used for the behavioral experiments. Subfraction B (13.1 g) was separated on a Sephadex LH-20 column chromatography using mixtures of methanol/water (from 0:1 to 3:7) and further purified by reversed-phase semi-preparative HPLC with a mobile phase of acetonitrile in 0.5% acetic acid/water (1:1, v/v) to yield the dark brown 2,5-Di-(N-prolyl)-para-benzoquinone (DPBQ) (27.5 mg). In the whole isolation procedure, temperature was maintained at <35 C to avoid degradation of DPBQ. Quantification of DPBQ in nectar: Nectar samples were separately collected from five inflorescences (each ten flowers) from 27-February to 3-March in 2010. After centrifugation at 8,000 g for 10 min, the supernatant of the nectar was analyzed by HPLC-DAD. At a flow rate of 1 ml/min, 10 µl of each sample was injected onto a ZORBAX SB-C 18 column (5 µm, 4.6 250 mm) (Agilent, USA). The mobile phase composed of solvent A (0.5%, acetic acid/water) and solvent B (acetonitrile) was used in a gradient mode (0-55 min: linear gradient of 5-95% B, 55-60 min: isocratic 95% B). The compound was detected at 369 nm, and its retention S2
time, UV-visible spectrum and peak area were compared with those of authentic compound. To prepare a calibration curve for DPBQ, triplicate injections were made at six concentrations (50, 100, 200, 500, 1000 and 2000 µg/ml). Then a plot of concentration vs. peak area was obtained which had an equation and correlation coefficient of y = 0.1983x + 20.257 (R 2 = 0.9983). Synthesis of (-)-DPBQ and (+)-DPBQ: (S)-L-(-)-Proline (41 mg, 0.36 mmol) and Na 2 CO 3 (22 mg, 0.27 mmol) were mixed in MeOH (2 ml), after that 23 mg (0.21 mmol) of para-benzoquinone was added to the solution. The reaction mixture was stirred for an hour, then diluted with suitable amount of water and directly chromatographed on MCI gel column with an eluent of water and methanol. The water fraction was condensed and subjected to reversed-phase semi-preparative HPLC (methanol : 0.4% acetic acid-water; 46.5 : 53.5) to yield (-)-DPBQ (16.5 mg, 27.4% yield). In the same way, 100.5 mg (0.87 mmol) of (R)-D-(+)-proline reacted with 49 mg (0.45 mmol) of para-benzoquinone under the condition of Na 2 CO 3 (50 mg, 0.61 mmol) and MeOH (4 ml) to afford (+)-DPBQ (45.2 mg, 31.0% yield). Behavioral experiments: Behavioral responses of birds to nectar sugar (purified from the nectar as described above) and nectar, and DPBQ, (-)-DPBQ and (+)-DPBQ of L. canum were respectively determined in experiments conducted in bird cages (70 45 45 cm 3 ) in the laboratory, using a method described in the literature with slight modifications (S. D. Johnson, A. L. Hargreaves, M. Brown, Ecology 2006, 87, 2709-2716.). Pure distilled water was used as a control. Naïve Japanese White-eyes (Zosterops japonicus, n = 5) were the test organisms. Preference of birds among nectar, nectar sugar and water were tested. Each bird was simultaneously offered 100 µl of nectar, nectar sugar and water in three separate small glass dishes (300 µl capacity) which were placed evenly spaced in a larger petri dish (12 cm in diameter), and then offered to a single bird simultaneously. Each trial was terminated after the bird had probed one of the glass dishes. Then the first choice and amount consumed were recorded, and the glass dishes were replenished. The positions of the three glass dishes were randomized after each trial. Each bird was tested 50 times with an interval of 30 min after every 10 trials. Subsequently, a similar experiment was conducted on nectar, with the glass dishes of nectar sugar and water being painted outside to the color of nectar solution to give the same appearance, and each bird was tested 30 times with an interval of 30 min after every 10 trials. The randomization of glass dishes in these experiments was also carried out after each trial. Preference of birds between DPBQ, (-)-DPBQ, (+)-DPBQ and water were separately tested. Each bird was simultaneously offered 100 µl of DPBQ or (-)-DPBQ or (+)-DPBQ solution (500 µg/ml) and water in two separate small glass dishes (300 µl capacity) which were placed evenly spaced in a larger petri dish (12 cm in diameter), and then offered to a single bird simultaneously. Each trial was terminated after the bird had probed one of the glass dishes. Then the first choice and amount consumed were recorded, and the glass dishes were replenished. The positions of the two glass dishes were randomized after each trial. Each bird was tested 50 times with an interval of 30 min after every 10 trials. Subsequently, a similar experiment was conducted on DPBQ only, with the water-containing glass dish being painted outside to the color of DPBQ solution to give the same appearance, and each bird was tested 30 times with an interval of 30 min after S3
every 10 trials. The randomization of glass dishes in these experiments was also carried out after each trial. Bird behaviors in all of the above experiments before and after feeding were also compared, and abnormal behaviors were recorded. Statistical analysis. Paired values were subjected to the t-test to determine the significance with SPSS 17.0 for windows. Fig. S1 Representative HPLC profile (monitored at 400 nm) of nectar of L. canum showing the dominant peak DPBQ, the UV-visible spectrum of DPBQ and its chemical structure. Fig. S2 Changes in the concentration of DPBQ in the nectar of L. canum flowers over the typical 5-day life of a flower. Nectar was removed each day. Each of the five curves represents samples from 10 flowers of a different inflorescence. S4
Fig. S3 1 H NMR spectrum of DPBQ recorded at 600 MHz in DMSO-d 6. Fig. S4 13 C NMR, DEPT-90 and DEPT-135 spectra of DPBQ recorded at 150 MHz in DMSO-d 6. S5
Fig. S5 1 H- 1 H COSY spectrum of DPBQ recorded in DMSO-d 6. Fig. S6 HSQC spectrum of DPBQ recorded in DMSO-d 6. S6
Fig. S7 HMBC spectrum of DPBQ recorded in DMSO-d 6. Fig. S8 ROESY spectrum of DPBQ recorded in DMSO-d 6. S7
mau 400 (a) 300 200 16.333 100 0 mau 400 5 10 15 20 25 16.337 (b) min 300 200 100 0 mau 400 5 10 15 20 25 16.356 (c) min 300 200 100 0 5 10 15 20 25 min Fig. S9 Overlay of HPLC chromatograms (monitored at 400 nm) of DPBQ (a) with (-)-DPBQ (b) and (+)-DPBQ (c). The mobile phase composed of solvent A (0.5% acetic acid-water) and solvent B (MeOH) was used in a gradient mode (0-30 min: linear gradient of 10-95% B, 30-35 min: isocratic 95% B). S8
Fig. S10 Overlay of 1 H NMR spectra of DPBQ (a, 600 MHz, DMSO-d 6 ) with (-)-DPBQ (b, 500 MHz, DMSO-d 6 ) and (+)-DPBQ (c, 500 MHz, DMSO-d 6 ). S9
Fig. S11 Overlay of 13 C NMR spectra of DPBQ (a, 150 MHz, DMSO-d 6 ) with (-)-DPBQ (b, 125 MHz, DMSO-d 6 ) and (+)-DPBQ (c, 125 MHz, DMSO-d 6 ). S10