Chapter 4 INFLUENCE OF VARIOUS BIOTIC AND ABIOTIC FACTORS ON THE SECONDARY METABOLITES PRODUCTION IN COFFEE BEANS

Chapter 4 INFLUENCE OF VARIOUS BIOTIC AND ABIOTIC FACTORS ON THE SECONDARY METABOLITES PRODUCTION IN COFFEE BEANS

4.1. Summary Plant growth and development is affected by various abiotic factors that include topography, soil, and climatic factors. Similarly biotic factors, determine the extent in which the genetic factor is expressed in the plant. Abiotic elements methyl jasmonate (MJ) and salicylic acid (SA) followed by biotic stress factors such as mycelial extracts of Rhizopus oligosporus and Aspergillus niger have shown significant augmentation of major secondary metabolites of green beans of robsuta coffee plants that were subjected to floral spray. Up to 42 % caffeine, 39 % theobromine, 46 % trigonelline along with 32 % cafestol and kahweol improvement were evident under respective elicitor treatments. Over all, surge in respective metabolites content depends on elicitor stress type and concentration. The elicitors MJ and SA are found to be efficient at 1 to 5 µm concentration in augmenting all the metabolites compared to R. oligopsorus and A. niger mycelial extracts spray at 0.5-2.0 % (w/v). To check the influence of altitude environment on Coffee beans metabolite profile, fruits were collected from C. canephora plants grown at different elevations to find out the influence of altitude on various metabolites of coffee beans. Caffeine content was found always more at all stages of coffee fruit ontogeny compared to the plants grown at low elevations. Maximum caffeine content of 1.868 ± 0.149 g. 100g -1 dry weight was evident in stage V seeds (9 months DAF) of low altitude collected samples. There was a 32% reduction in caffeine content of beans from plants grown at high elevations. Both nicotinic acid and trigonelline contents were always higher at all stages of coffee fruit ontogeny from plants grown at higher elevations. Maximum nicotinic acid and trigonelline content of 5.164 ± 0.131 mg.100g -1 and 975.8 ± 7.24 mg.100g -1 respectively was evident in stage V seeds (9 90

months DAF) of high altitude collected samples.the levels of free diterpenes cafestol and kahweol were found to be more in beans collected from low elevations (0.12% of cafestol and 0.045% kahweol) than that of high elevation grown plants (0.042% of cafestol and 0.014% of kahweol). Apart from these, some major cafestol and kahweol esters viz., cafestol linoleate, cafestol oleate, cafestol palmitate, cafestol stearate, kahweol linoleate, kahweol oleate and kahweol palmitate were identified. This, stress mediated augmentation of secondary metabolites of coffee would have a wider scope for studying the expression pattern of major biosynthetic pathway genes of caffeine, trigonelline, and diterpenes in coffee, under selected elicitor treatment, which would be helpful to develop a metabolomics and proteomics approach for improving the quality of the beans. 4.2. Introduction Secondary metabolites production in plants is influenced by various extrinsic and intrinsic factors (Ramakrishna and Ravishankar, 2011). Extrinsic factors such as environmental and biotic factors role on plant growth and important metabolites production was investigated under both in vitro and ex vitro conditions. The biotic factors such as fungi, bacteria and algae are reported to be good in eliciting production of wide range of chemical compounds such as alkaloids, flavonoids, phenyl propanoid intermediates (Prasad et al., 2006), as shown by using cell suspension cultures or root cultures (reviewed by Ramakrishna and Ravishankar, 2011).Recent studies have demonstrated the efficiency of this elicitor mediated approach to augment food value metabolites such as annatto pigment in Bixa orellana (Mahendranath et al., 2011; 91

Giridhar and Parimalan, 2010), phenyl propanoid intermediates and capsaicin in Chilli (Gururaj et al., 2012). Coffee is rich in bioactive secondary metabolites such as caffeine, trigonelline, nicotinic acid, and diterpenes cafestol and kahweol. Caffeine (1,3,7 trimethyl xanthine) is one of the major secondary metabolite in coffee and some other plants are reported to combat physical and biotic stress factors such as pathogens and predators (Nathanson, 1984; Kim et al., 2011). Coffee also contains trigonelline which is considered as the second abundant alkaloid compound and it is thermally converted to nicotinic acid and some flavour compounds during roasting (Taguchi, 1988). In addition it is considered important for taste and nutrition (Adrian and Fragne, 1991). Other important components that are associated with lipid fraction of coffee are diterpenes- cafestol (caf) and kahweol (kah) which are unique to coffee, and they are mainly pentacyclic diterpene alcohols based on kaurene skeleton (Urgert et al., 1995) and only a small portion is present in the free-form, and the remain is in esterified form. High levels of caffeine produced by coffee seedlings exhibit allelopathy there by inhibit the germination of other seeds in the vicinity of the growing plants (Peneva, 2007). Elicitation is an induction process which results in increased activities of enzymes of caffeine biosynthesis. The production of the purine alkaloid caffeine was shown to be stimulated by stressors such as high light intensity and high NaCl concentration (Peters et al., 2004). Moreover, exogenous calcium influences caffeine alkaloids production in C. arabica and C. canephora (Ramakrishna et al., 2011). Though sporadic reports are available on biotic elicitors influence on in vitro cultures of Coffea for caffeine production, no such reports are available on Coffea grown ex vitro and also during the 92

ontogeny of fruits. Such studies pertaining to trigonelline, cafestol and kahweol production in green beans of C. canephora are lacking. From coffee quality point of view metabolite profiles of coffee beans is having paramount importance in global coffee market, hence understanding the influence of various extraneous factors on bean metabolite profiles of Coffea standing crop in field is warranted as it facilitate to understand the biochemical profiling and their biosynthetic pathways. Moreover, various microbes incidence on coffee berries is reported (Velmourougane et al., 2011; Velazquez et al., 2011) which may influence the quality of bean. The altitude at which coffee is grown plays a major role in determining the quality of the bean, because there is less oxygen, so coffees grown at higher altitudes take longer to mature than plants grown at lower altitudes. This allows the flavours to develop more fully and produces beans that are delicate and flavourful (Clifford, 1985). Arabica beans grow at altitudes between 600 and 1,800 metres above sea level and take six to nine months to mature. Robusta plants grow at low altitudes (sea level to 600 meters), more cold and moisture-tolerant and disease-resistant. Variation in the content of trigonelline, nicotinic acid and diterpenes in coffee beans is mainly based on geographical distribution of Coffea species (Lercker et al., 1995) and the altitude at which coffee is grown may play a major role in determining metabolites profiles (Decazy et al., 2003) and also quality of the bean (Muschler, 2001). As there are no reports available about the influence of elevation (altitudinal) variations on Coffee bean metabolite profiles from plants grown in India such studies are warranted. A study has been proposed to investigate the influence of abiotic elicitors (salicylic acid and methyl jasmonate) and biotic elicitors (Aspergillus niger and Rhizopus 93

oligosporus) on caffeine alkaloids, trigonelline and antinutritioal diterpenes cafestol and kahweol production in green beans of C. canephora under elicitor stress. Apart from this, analysis of major metabolites profile during the ontogeny of C. canephora fruit development in samples collected from plants grown at different altitudes was carried out. 4.3. Material and methods 4.3.1. Preparation of elicitors Biotic elicitors were prepared using two fungal cultures viz., Aspergillus niger and Rhizopus oligosporus that were procured from microbial culture facility of Food Microbiology Department of CFTRI, Mysore. Fresh cultures of A. niger and R. oligosporus were made on PDA slants and incubated for 7 days. Then the spores of the respective fungi were used to prepare spore suspension in 0.1 % sodium lauryl sulphate (w/v) and diluted with sterile distilled water under aseptic conditions to obtain a spore density of ~ 2.5 X 10 6 spore ml -1. Later the same was inoculated into the 40 ml of PDA broth contained in 150 ml Erlenmeyer conical flasks and the cultures were incubated in dark for 10days. After culture growth, the cultures were autoclaved and the mycelium was separated from the culture broth by filtration and its fresh weight was recorded. Aqueous extract of mycelium mat was made by homogenizing in mortar and pestle using neutralized sand. The extract was filtered through Whatman no. 1 filter paper. The filtrate was made up to the known concentration and kept as the stock solution from which the individual fungal mycelia extracts at a working concentration of 0.5, 1.0 and 2% % w/v (wet weight of fungal mycelium in 100 ml of distilled water) were prepared in sterile water and used for elicitation experiment. 94

Figure 4.1. A pictorial depiction of work component under objective 2 of the study 95

Abiotic elicitors viz. salicylic acid (SA) and methyl jasmonate (MJ) purchased from (Sigma, Bangalore) were dissolved in distilled H 2 O and diluted to obtain three concentrations (1.0, 2.5 and 5.0 µm) for the study. 4.3.2. Elicitor application Five years old plants of Coffea caenphora CxR were used for small-scale field study to find out influence of elicitors on annatto pigment on standing crop through elicitor mediated approach. Different plants were selected for spraying various elicitors in our study. All these plants were under cultivation in an area of 50 x40 ft at Plant Cell Biotechnology Department, of this institute, with a spacing of 4 x 4 ft between each plant. The prepared biotic (A. niger, R. oligosporus) and abiotic elicitors methyl jasmonates (MJ) and salicylic acid (SA) were sprayed to the flowers during anthesis (between 9-11 Hrs). The ripened red fruit from plants subjected to respective treatments were harvested at maturity (after 9 months), and used for extraction of respective metabolites and their analysis. 4.3.3. Extraction and analysis of caffeine and theobromine Fruit pulp of harvested fruits was removed and beans were taken out and dried to attain ~12 % of moisture content. To determine caffeine, 5g of fresh tissue (of green beans) was extracted using 80% ethanol using mortar and pestle, the resultant slurry was homogenized using neutralized sand, followed by centrifugation for 10min at 8000 rpm and supernatant was collected after centrifugation. The extract was flash evaporated to dryness and dissolved to 1ml of 80% ethanol prior to estimation of caffeine and theobromine by HPLC (Ashoor et al., 1983) as explained in chapter 3 material and methods 3.3.3. The compounds were identified by their retention times, chromatographic 96

comparisons with authentic caffeine and theobromine standard (Sigma, Bangalore), and their UV spectra. Quantification was based on the external standard method. 4.3.4. Extraction and analysis of trigonelline and nicotinic acid Extraction and HPLC analysis of trigonelline and nicotinic acid were carried out as explained earlier in chapter 3 material and methods section. 4.3.5. Extraction and analysis of diterpenes cafestol and kahweol. Known quantity ground coffee sample was weighed (5 g of beans/500 mg of pericarp) separately for extraction of each sample for free forms of diterpenes (cafestol and kahweol). The extraction and analysis of free forms of cafestol and kahweol was performed as explained under materials and methods section of Chapter 3. 4.3.6. Influence of altitude on metabolite profile of coffee beans Fruits of C. canephora P. ex Fr. CxR variety at different stages of development were collected from plantations grown at different elevations near Mudigere of Chikamagalur District of Karnataka during February 2009. Ripened fruits from plants grown at high (3700 ft MSL), medium (3300 ft MSL) and low altitudes (3000 ft MSL) were collected from different places viz., Devaramane, Guthi (Javali) and Tripura respectively (Lat 13 7' 60N, Long 75 37' 60E). Ten random plants were selected at each of the said altitude, and at least 250 g of fresh fruits were collected for caffeine trigonelline and nicotinic acid profiles and for diterpenes around 500 g of fresh fruits were collected from these plants. These samples were immediately taken into sterile polythene bags with small perforations and used for experiment within 24 hours. 97

To determine caffeine, trigonelline, nicotinic acid, cafestol and kahweol contents, fruits of five different stages were harvested i.e. Stage I (3 months DAF), Stage II (5 months DAF), Stage III (7 months DAF), Stage IV (8 months DAF) and Stage V (9 months DAF). These stages roughly represent pinhead stage immediately after dormancy period, rapid expansion stage, pericarp growth stage, efficient endosperm formation stage and lastly dry matter accumulation stage to harvesting stage respectively (Koshiro et al., 2006). Respective fruit tissues (whole fruit/pericarp/beans) of five different stages were used for metabolites analysis. The green beans were allowed to dry to attain ~12% moisture content. All the green beans were then grounded finely. The extraction and analysis of caffeine, trigonelline, nicotinic acid, was carried out as explained in materials and methods section of chapter 3. 4.3.7. Extraction and separation of both free form of cafestol and kahweol and their methyl esters by GPC and SPE The coffee oil obtained after soxhlet extraction as explained earlier in Chapter 3, was dissolved in dichloromethane and membrane filtered in a 5 ml measuring flask and the flask was made up to the mark with dichloromethane. Approximately 250 µl solution was subjected to Gel Permeation Chromatography (GPC) on Bio-Beads SX3 (BioRad, India) with dichloromethane as the mobile phase at a flow rate of 4ml/min. The fraction with the diterpenes esters was concentrated in a rotary evaporator and solvent residues were removed in a nitrogen steam. The diterepene esters were then separated by solid phase extraction (SPE) on the silica gel column with n-hexane (5ml) and washed the column with n-hexane:ethylacetate (2ml: 96+4,v/v) for elution of the 16-O-Methyl cafestol esters, with 2 ml n-hexane:ethylacetate (9:1) for cafestol/kahweol esters, with 98

4ml n-hexane:ethylacetate (8:2 v/v) for free diterpenes (Kurzrock and Speer, 2001). These extracts were dried in a nitrogen stream and redissolved in 1ml mobile phase (acetonitrile: isopropanol 70:30 v/v) and were used for chromatographic analysis. The sample injection volume was 20µl. As standards of cafestol and kahweol palmitate are not available the same were synthesized for reference standards purpose as suggested earlier (Kurzrock and Speer, 2001). In brief it is: The mixture of cafestol and kahweol standards ~ 60 mg was taken in 55:45 ratio (m/m) into round-bottom flask (100 ml) and dissolved in benzene and pyridine (2mL;3+1, v/v). Palmitoyl chloride (~ 120 mg) dissolved in benzene (0.5mL) and dichloromethane (3mL) were added to the mixture, followed by its stirring for one hour and then allowed to stand overnight in the dark. The reaction was stopped by addition of water (20mL) and then concentrated to about 10 ml on flash evaporator, and the residue was transferred to a 250mL separating funnel. To this 10% Na 2 CO 3 solution (60mL) was added and the aqueous solution was extracted with tert-butyl methyl ether (2+100mL) in order to remove the free fatty acid. The combined ether phases were dried for 15 min over Na 2 SO 4 (3g) filtered concentrated to about 5mL in rotary evaporator. To this added 25 ml acetone and concentrated again. The solvent residues were removed in the N 2 stream. The diterpenes esters and the free diterpenes were separated off by means of solid-phase extraction (SPE) on a silica-gel column as explained above for different fractions from samples. 4.3.8. Free diterpenes analysis by HPLC Free diterpenes fraction that obtained after GPC and SPE fractionation, was subjected to HPLC analysis (Waters Alliance 2695 HPLC, USA) by using Nucleosil 120-3, C18, 250/4 column (Macherey Nagel, Gmbh, Germany) to quantify cafestol and 99

kahweol (Kurzrock and Speer, 2001). The UV detector was set to 230 nm for cafestol and 290 nm for kahweol. Eluent solvent system used was acetonitrile: isopropanol (70:30 v/v) with a flow rate of 0.6 ml/min. Peaks were identified by comparing with the retention time of reference standards (Cafestol and Kahweol from Sigma, Bangalore) and by spiking. 4.3.9. LC-MS analysis of diterpene esters Analysis was performed (Kurzrock and Speer, 2001) on Waters Alliance 2695 HPLC, equipped with an auto sampler and coupled with a Waters 2696 photodiode array detector, using Nucleosil 120-3, C18, 250/4 column (Macherey Nagel, Gmbh, Germany) with UV detection in the range of 200-350 nm (optimal is 230 nm for cafestol esters and 290 for kahweol esters). The mobile phase used was acetonitrile: isopropanol (70:30 v/v) and flow rate of 0.6 ml/min for 30 min. Mass spectrum was acquired in a Q- TOF Ultima mass spectrometer (Waters Corporation, Micro Mass, and Manchester, UK). The adopted ionization mode was of positive type, APCI-temperature was 350 0 C, gas temperature (nitrogen) was 330 0 C, nebulizer gas pressure was 60 psi, and flow rate of the dry gas is 4.0 ml/min. The ionization energy (CID voltage) was 20V or 50 V as suggested earlier (Kurzrock and Speer, 2001). The confirmation of the occurrence of compounds is based on the molecular ions in the MS spectra as well as with the literature (Kurzrock and Speer, 2001). To get clear spectra for free forms of cafestol and kahweol low CID (<20v) was used. 100

4.3.10. Statistical Analysis The entire experiment was conducted in a randomized block design with five replicates. For each treatment fifty flowers (bunches) were sprayed. Five fruits of respective tagged flowers were taken for analysis of caffeine, theobromine, trigonelline, nicotinic acid, free diterpenes cafestol and kahweol content for each and every treatment. Regression Values from all five replicate determinations of each sample were averaged and represented as means with standard deviations. Data were analysed statistically by the SPSS 17.0 software by one-way ANOVA. Similarly, for studying influence of altitude on metabolites profile of developing fruits of C. canephora, five coffee bean samples, out of the ten samples each that were randomly collected from coffee plants grown at higher, medium and lower elevations were extracted and analysed for caffeine trigonelline, nicotinic acid, cafestol and kahweol. Values from all five replicate determinations of each sample were averaged and represented as means with standard deviations. Data were analysed statistically by the SPSS 17.0 software by Two-way ANOVA and homogenous subsets were determined to separate the mean values of the different altitudes and developmental stages. 4.4. Results and Discussion The increase in respective secondary metabolites of coffee beans was evident under elicitor treatments which vary with the elicitor and its concentration. Both biotic and abiotic elicitors responded in a similar way, however, the number of folds increase in metabolites was better in abiotic elicitors. The purine alkaloid caffeine content was maximum (2900 ± 244 mg.100g -1 ) at 2.5 µm MJ which was 42% more than that of 101

control. Even (Fig 4.2) under 2.5 µm SA treatment, 39% increase in caffeine content was noticed. On par with caffeine content, theobromine content too increased at respective elicitor treatments and maximum production under 2.5 µm SA spray was noticed which ~ 38% was more to control. When biotic elicitors (Fig.4.3) R. oligosporus and A.niger mycelial extracts were sprayed at 1% concentration as elicitors to flowers, there was a maximum enhancement of 20 % and 18 % for theobromine, 25 % and 21% for caffeine respectively was noticed in beans of harvested fruits, which was less than that of abiotic elicitors SA and MJ (Fig. 4.2). The content of pyrimidine alkaloid trigonelline and its precursor nicotinic acid too were influenced by respective elicitors (Fig 4.4). Trigonelline production was maximum at 1.0 µm MJ. Though there was an increase in nicotinic acid content it was in the range of 4-5 mg.100 g -1 dry wt., compared to control (3.78 ±1.17 mg.100g -1 ). At higher concentrations of SA (5 µm) and MJ (5µM) there was significant decrease in caffeine, theobromine, trigonelline content and also to moderately in case of nicotinic acid (Fig 4.5). 102

Figure 4.2. Caffeine and theobromine profile of harvested beans of C.canephora plants (values are mean ± SD of five analyses; significant at p< 0.05) under abiotic stress. Figure 4.3. Caffeine and theobromine profile of harvested beans of C.canephora plants under biotic stress (values are mean ± SD of five analyses; significant at p< 0.05; RO: R. oligosporus; AS: A. niger). Figure 4.4. Trigonelline and nicotinic acid profiles of harvested beans of C.canephora plants under abiotic stress (values are mean ± SD of five analyses; significant at p< 0.05). 103

Figure 4.5. Trigonelline and nicotinic acid profiles of harvested beans of C.canephora plants under biotic stress (values are mean ± SD of five analyses; significant at p< 0.05; RO: R. oligosporus; AS: A. niger). The free form of cafestol and kahweol profiles were less influenced by elicitor treatments (Fig 4.6) compared to other metabolites of the study viz., caffeine, theobromine, trigonelline and nicotinic acid. There was 12, 26 and 18% increase in cafestol content under lower to high concentration of SA treatments compared to control (156 ± 4.43 mg.100g -1 ). But under 1.0, 2.5, 5.0 µm MJ spray, the increase in cafestol content was more to SA treatments, and it was 17, 32 and 22%. Similarly, both R. oligosporus and A. niger triggered moderate enhancement of cafestol compared to abiotic elicitors and it was 4%, 14% and 10% for R. oligosporus and 6, 15 and 18% for A. niger respectively compared to control (Fig.4.7). A similar trend was evident for kahweol content enhancement with maximum kahweol production at 2.5 µm SA treatment. 104

Figure 4.6. Cafestol and kahweol profiles of harvested beans of C.canephora plants under abiotic stress (values are mean ± SD of five analyses; significant at p< 0.05) Figure 4.7. Cafestol and kahweol profiles of harvested beans of C.canephora plants under abiotic stress (values are mean ± SD of five analyses; significant at p< 0.05; RO: R. oligosporus; AS: A. niger). The significant achievement in this study is the augmentation of major secondary metabolites under ex vitro conditions through floral spray of respective biotic or abiotic elicitors which is a novel method with reference to natural compounds of any category. The bioactive metabolites from beans of coffee that 105

investigated in the present study are compounds that not only are connected with important traits of the plant itself, with reference to coffee quality, resistance to pests and diseases, but also proved to be functionally important at physiological level in consumers (Fredholm et al., 1999; Van Dam et al., 2006). Elicitors of both microbial origin and abiotic elicitors are known for triggering varied responses in plants including augmented levels of secondary metabolites for value addition (reviewed by Ramakrishna and Ravishankar, 2011), which upon contact with higher plant cells, trigger the increased production of phytoalexins (Giridhar and Komor, 2002), and other defense related compounds (Robbins et al., 1985; Eilert et al., 1984; Eilert et al., 1986; Flores and Curtis, 1992; Sim et al., 1994; Singh 1999). Also they trigger pigments (Mahendranath et al., 2011) and isoflavones (Saini et al., 2013). Similarly the role of biotic elicitors on increased production of alkaloid in Brugmansia candida was reported (Pitta-Alvarez et al., 2000). In general plants produce jasmonic acid and methyl jasmonate in response to biotic and abiotic stresses, which accumulates in the damaged parts of the plant. The same MJ is also demonstrated as best abiotic elicitor. Treatment with jasmonates can elicit the accumulation of several classes of alkaloids (Zabetakis et al., 1999), phenolics (Lee et al., 1997) activation of phenylalanine ammonia-lyase, accumulation of scopoletin and scopolin in tobacco (Sharan et al., 1998), betacyanin synthesis in Portuluca (Bhuiyan and Adachi, 2003) and capsaicin in Capsicum (Prasad et al., 2006). Moreover, a mechanism by which MJ induced-gene expression involved in plant secondary metabolites biosynthesis at molecular level was demonstrated 106

(Suzuki et al., 2005). But such reports are scanty with reference to augmentation of coffee metabolites especially trigonelline, diterpenes cafestol and kahweol. Hence, the present study is first of its kind, wherein, floral application of MJ could increase the major secondary metabolites content in robusta coffee beans as discussed. In the present study along with caffeine, its precursor theobromine was also analysed. Through it is not necessary for theobromine profiles in this study, the data pertains to this theobromine further substantiates caffeine content in beans. Moreover extraction and analysis is same as for caffeine. Similarly salicylic acid is known as essential component of the plant resistance to pathogens and plays a part in the plant response to adverse environmental conditions. Recent studies have demonstrated about the importance of SA in isoprenoids (precursors of carotenoids pathway) production and accumulation (Czerpak et al., 2003; Moharekar et al., 2003; Çag et al., 2009), alkaloids production in Stemona curtisii hairy root cultures (Chotikdachanarong et al., 2011) and trigonelline production in Trigonella foenum-graceum cell cultures (Mathur and Yadav, 2011). The production of caffeine was shown to be stimulated by stress factors such as high light intensity and high NaCl concentration (Peters et al, 1985). Exogenous calcium too influences alkaloids in C. arabica (Ramakrishna et al., 2011). In the present study, biotic elicitors A. niger and R. oligosporus at 0.1% (w/v) enhanced ~ 15% of cafestol and kahweol which is more to control. But studies on elicitor mediated enhancement of trigonelline, and diterpenes cafestol and kahweol are 107

lacking and for the first time the same is reported in the present study. Under elicitor stress, the activity of the N-methyltransferases will be enhanced that leads to enhanced production of caffeine. The levels of caffeine in developing seeds are important as there are no reports about the actual role of this in developing fruits unlike polyamines (Sridevi et al., 2009). Caffeine that initially synthesized in pericarp tissues subsequently shifts to endosperm of seed and it accumulates apart from its own caffeine content. Once caffeine synthesis stops in pericarp in ripened fruits, caffeine content rather stabilizes in matured seeds of harvested fruits (Baumann and Wanner, 1972). So the actual elicitor response triggers during the initial developmental stage and it possibly lasts till fruit maturity stage. Nicotinic acid being the precursor for trigonelline production, the trigonelline data in the present study at all elicitor treatments was in direct relation to nicotinic acid levels. As C. canephora fruit development requires 8 to 9 months there is a scope for variation in bioactive content in beans during ontogeny of fruit as they get exposed to various microbes and pathogens. Though trigonelline is normally synthesized in almost all parts of coffee plant (Zheng et al., 2004), its accumulation is high in young tissues (Zheng and Ashihara, 2004). Generally the diterpenes cafestol and kahweol are associated with creamy fat part of the coffee brew. Though there are no reports on direct or indirect influence of elicitor stress on lipid content and diterpenes of beans, the increase in cafestol and kahweol content may be due to the higher activity of gibberellic acid pathway enzymes wherein the entekaurene is the precursor for both cafestol and kahweol (De Roos et al., 2006). A part from elicitor stress, other associated factors such as 108

climate, temperature, and availability of light and water during the ripening stage of fruit play crucial role on metabolite profiles of beans (Clifford, 1985; Decazy et al,. 2003; Bertrand et al., 2006). The data obtained in the present study clearly showed the potential of elicitor mediated augmentation of major bioactive metabolites of coffee which is important from quality aspect and also have a wider scope for studying the expression pattern of major biosynthetic pathway genes of caffeine, trigonelline, and diterpenes in coffee, which would be helpful to develop a metabolomics and proteomics approach for deciphering the quality aspects of coffee. 4.4.1. Influence of altitude on caffeine content There were significant variations in caffeine levels of green beans (dry wt. basis) of fruits collected from plants grown at different altitudes. The caffeine levels were positively influenced by altitude variation, with 18.5, 1.5 and 1.25 mg g -1 dry wt. at low, middle and high altitude samples respectively (Fig 4.8). Similarly there was significant variation observed in caffeine levels during the ontogeny of coffee bean development (Fig 4.8). In stage I samples (3 months DAF) i.e. immediately after dormant period, wherein very small fruits appear, the caffeine content was 0.381 ± 0.047 to 0.287± 0.009 g.100g -1. Maximum content of caffeine was detected in harvested beans (Stage V, 9 M DAF). The proportion of increase in caffeine content was more from the stage II (5M DAF) to stage III (7M DAF). 109

Figure 4.8. Caffeine profiles in beans of C.canephora CxR variety during ontogeny of fruit (values are mean ± SD of five analyses); Different alphabet letters indicate the statistical significant difference within the different developmental stages (p< 0.05), and different symbols (α, β and γ) represent statistical significant difference within the altitude (p< 0.05). In seeds of stage III fruit there was 14% and 20% increase in caffeine content in samples from medium and high altitudes compared to low altitude samples. A progressive increase in caffeine content of stage V seeds was evident from low altitude collected samples (1.868 ± 0.149 mg.100g -l ) and there was 32% increase in high altitudes samples. To validate the efficiency of extraction procedure one sample was extracted six times and the coefficient of variation was 1.95%. Similarly a linear relationship between 110

Figure 4.9. HPLC chromatogram of caffeine a) standard (RT 13.56 min); b) samples from low altitude (RT 13.49 min); c) samples from middle altitude (RT 13.58 min) and d) samples from high altitude (RT 13.47 min). 111

Figure 4.10. Regression coefficients for effect of elevation on a) variation in caffeine content in stage V (9 months DAF) beans of C. canephora fruit, b) Variation in average caffeine content during ontogeny of C. canephora fruit. caffeine concentration and UV absorbance (270 nm) was perceived, wherein the linearity was managed over the concentration range of 10-100 ppm and the correlation coefficient for the caffeine standard curve was 0.9999. The retention time of caffeine was 13.56 min (Fig 4.9) with a relative standard deviation of RSD = 0.54%.The method of detection was 0.030 ppm and the precision was 1.95% at 50 ppm caffeine concentration. The spiked recoveries of caffeine were 1.02% for tested coffee bean extract. The variation in retention time for caffeine of test samples was insignificant. Caffeine concentration at advanced stages of fruit development decreased significantly with increasing elevation (Fig 4.8). The coefficient of correlation (R 2 ) was recorded as 0.5938 at stage V beans (Fig 4.10a). Similarly the average caffeine content of all stages also decreased and inversely proportional to elevation, wherein the coefficient of correlation (R 2 ) was recorded as 0.8553 (Fig 4.10b). In the present study, there was significant variation in caffeine content in developing seeds during ontogeny of C. canephora fruits. A decreasing trend in caffeine 112

content of pericarp was evident when the fruit becomes mature and the rate of reduction was more (58%) between stage IV to stage V fruits. The trend was same in samples collected from plants grown at different altitudes. In C. canephora in general the fruit development requires 8-9 months or even up to 10 months and it is to some extent asynchronous and could be responsible for variation in major secondary metabolites in beans during ontogeny of coffee fruit. However, a tendency for synchrony was observed during the later stages of maturation when a significantly high proportion of fruits entered the largest sized ripe cherry stage as opined by De Castro et al., (2005). Caffeine was well documented from over 63 plant species worldwide (Neigishi et al., 1998; Jin et al., 2005; Ali et al., 2012) and also from different Coffea species (Ashihara et al., 1998; Zheng et al., 2002; Koyama et al., 2003). Caffeine is normally synthesized in almost all parts of coffee plant, especially in young tissues such as leaves, even in floral parts (Baumann 2006), and highly concentrated initially in pericarp Baumann (2006). Our results too indicate that though caffeine was found in significant amounts in pericarp at initial growth stages of fruit, its accumulation was less in beans at maturation stage of fruit, which was supported by earlier studies. The high levels of caffeine in harvesting stage beans (Stage V) is mainly due to acquisition from the perisperm, import from the pericarp (Baumann and Wanner, 1972) and also due to high activity of S- adenosine methionine (SAM) dependent N-methyl transferases involved in caffeine biosynthesis in developing endosperm (Mazzafera, 1994; Mizuno et al., 2003). Apart from these metabolism aspects of caffeine in developing fruits of Coffea, catabolism of caffeine too influence the caffeine content of beans. The ratio between biosynthesis and 113

biodegradation controls the difference in caffeine levels during ontogeny especially at fruit ripening stage as reported in C. arabica and C. dewevrei (Lopes and Monaco 1979). High-grown coffee beans usually have a high density than low-grown beans. Longer maturation times and increased bean sizes are usually observed for plants grown at high altitude or in shade conditions (Guyot et al., 1996; Muschler, 2001; Vaast et al., 2006). In C. arabica the effects of shade on the development and sugar metabolism was investigated and also showed that shade grown plants produce less caffeine (Geromel et al., 2008). Similarly, post-harvest processing conditions and brewing methods too influence the metabolic profile of beans (Sridevi et al., 2011).Though caffeine production takes place in pericarp of young fruits, the same will decline in pericarp of ripened fruits (Zheng et al., 2004). Similarly the caffeine content in seeds during fruit development was varied and there was a progression in caffeine content with the advancement of age of fruit. The increase in caffeine content was more significant from fruit development stage III to stage V. Though caffeine biosynthesis actively takes place during leaf let emergence (Frischknecht et al., 1986), also the same takes place in developing fruit, the pericarp and perisperm (Keller et al., 1972). Mazzafera and Goncalves (1999) reported that, absolute amounts of caffeine parallel the dry weight of bean, with more or less constant caffeine content during maturation, nearing 1% on a dry matter basis at the time of harvest. The increase in caffeine content during stage III (6 months) to stage IV (8 months) in our study is further supported by similar observations in C. arabica (Keller et al., 1972). The levels of caffeine in developing seeds are important though we don t have any information about the actual role of this in developing fruits unlike polyamines (Sridevi et al., 2009). Most of the caffeine that synthesized in pericarp shifts to 114

endosperm of seed where, it accumulates apart from its own caffeine content, and once further caffeine synthesis stops in pericarp in ripened fruits, caffeine content rather stabilizes in matured seeds of harvested fruits (Baumann and Wanner, 1972). This study indicates that the altitudes at which coffee plants grow had influence on caffeine content. 4.4.2. Influence of altitude on trigonelline and nicotinic acid content Significant variations in trigonelline (Trig) and a nicotinic acid (NA) level of green beans of fruits collected from plants grown at different altitudes was noticed. The Trig levels were positively influenced by altitude variation, with a maximum of (975.8±7.24 mg.100g -l ) in beans of high altitude samples, which was 20% and 25% more than that of medium and lower altitudes respectively (Fig 4.11). (Fig 4.12). Figure 4.11. Trigonelline (mg.100g -1 ) profile in beans of C.canephoara CxR variety during ontogeny of fruit (values are mean ± SD of five analyses); Different alphabet letters indicate the statistical significant difference within the different developmental stages (p< 0.05), and different symbols (α, β and γ) represent statistical significant difference within the altitude (p< 0.05). 115

Similarly there was significant variation observed in both Trig and NA during the ontogeny of coffee bean development (Fig 4.11, 4.12). The levels of NA too were positively influenced by altitude variation, with a maximum of (5.16 mg 100 g-l) in beans of high altitude samples. In stage I samples (3 months DAF) i.e. immediately after dormant period, wherein tiny fruits appear, the NA and Trig contents were (0.686 ± 0.078 mg.100g -l ) to 1.25 ± 0.068 mg.100g -1 ) and (20.86 ± 3.23 to 32.4 ± 4.75 mg.100 g -l ) respectively at low to high altitudes and in subsequent stages of bean development there was a gradient increase in both NA and Trig content and it reached to (3.92 ± 0.127 to 5.164 ± 0.131 mg.100g -l ) and (734.8 ± 7.39 to 975 ± 7.24 mg.100 g -l ) at harvesting stage of samples collected from low and high altitudes respectively (Fig. 4.11, 4.12). Trigonelline concentration at all stages of fruit development increased significantly with increasing elevation (Fig 4.11). The regression was significant (y=0.1739x-217.17), with high coefficient of determination as R 2 = 0.9774. Similarly, NA concentration too increased significantly with increasing elevation, wherein regression was y=0.0015x- 2.6174 with coefficient of determination as R 2 = 0.8526 which is slightly less than that of trigonelline (Fig 4.13). In stage V fully ripened fruit s bean Trig concentration regression too was significant (y=0.3509x-336.53) with high coefficient of determination as R 2 =0.9434 which is almost same to R 2 values of average trigonelline content of at all stages of ontogeny. Even for NA of stage V fruit beans (Fig. 4.13), the coefficient of determination (R 2 = 0.8729) was similar to that of ontogeny stages (R 2 = 0.8526). So in the present study, the way altitude (quantitative variable) affects the secondary metabolite Trig and 116

also its precursor nicotinic acid was interpreted through linear correlation. A glance at NA profile during ontogeny shows that, there was significant variation at stage II fruit, stage III, IV and V seed of medium altitude (mean = 2.35, 3.964, 4.108, 4.634; standard deviation = 0.367, 0.149, 0.291, 0.268 respectively), compared to low and high altitudes, which indicates the influence medium altitude was prominent for nicotinic acid. Similar observations were noticed for trigonelline too. Figure 4.12.Nicotinic acid mg.100g -1 dry wt) profiles in beans of C.canephoara CxR variety during ontogeny of fruit (values are mean ± SD of five analyses); Different alphabet letters indicate the statistical significant difference within the different developmental stages (p< 0.05), and different symbols (α, β and γ) represent statistical significant difference within the altitude (p<0.05). 117

Figure 4.13. Regression coefficients for effect of elevation on a) variation in trigonelline content during ontogeny of C. canephora fruit, b) variation in nicotinic acid content during ontogeny of C. canephora fruit, c) variation in stage V seed trigonelline content, d) variation in stage V seed nicotinic acid content 118

Figure 4.14. HPLC chromatogram of nicotinic acid (RT 1.990) and trigonelline (RT 2.663) in seed extracts of C. canephora fruit grown at different altitudes; a) standards b) Lower altitude sample c) medium altitude sample d) higher altitude sample 119

In the present study, the extraction and HPLC analysis of both Trig and NA facilitated to obtain the data significantly (p < 0.05). To validate the efficiency of extraction procedure one sample was extracted three times and the coefficient of variation was 1.54% and 1.12% for Trigonelline (Trig) and nicotinic acid (NA). Similarly a linear relationship between Trig / NA concentration and UV absorbance was perceived, wherein the linearity was managed over the concentration range of (0.15-400 µg ml and 0.10-400 µg ml) and the correlation coefficient for the Trig/NA standard curve invariably exceeded 0.999. In HPLC the retention time (RT) for NA and Trig were 1.990 and 2.663 respectively (Figure 4.14). In case of seed sample extracts another major peak was evident at 4.54 RT. This peak is nothing but caffeine of seed extracts and data of caffeine is not quantified as it requires different extraction method and also absorption maxima (274 nm). In fact caffeine was quantified separately as explained in material and methods of this chapter. In the present study, there was significant variation in Trig and NA content in developing seeds during ontogeny of C. canephora fruits. The pyridine alkaloid Trig accumulates in seeds, however its biosynthetic activity is reported to be higher in the pericarp (Koshiro et al., 2006). Nicotinic acid being the precursor for Trig production, the Trig data in the present study at all stages of ontogeny of fruit was in direct relation to NA levels. A decreasing trend in Trig content of pericarp was evident when the fruit becomes mature and the rate of reduction was more from stage III to stage V fruits. The trend was almost same in samples collected from plants grown at different altitudes. In stage III fruit pericarp more trigonelline was detected compared to earlier stage of growth. In C. canephora in general the fruit development requires 8 to 9 months and it is to some extent asynchronous. This may be one of the reasons for variation in bioactive 120

content in beans during ontogeny of fruit. However, a tendency for synchrony was observed during the later stages of maturation when a significantly higher proportion of fruits entered the largest sized ripe cherry stage as opined (De Castro et al., 2005). Trigonelline was well documented from various plants (Poulton, 1981; Barz, 1979; 1985) and trigonelline is normally synthesized in almost all parts of coffee plant (Zheng et al., 2004), and its accumulation is higher in young tissues as reported by Zheng et al (2004). Zheng et al (2004) also reported trigonelline activity in endosperm stage. Our results indicate that though Trig was found in significant amounts in pericarp at initial growth stages of fruit, its accumulation was more evident in beans at maturation stage of fruit, which was supported by earlier studies of (Zheng et al., 2004 ) in C. arabica, wherein, high net biosynthetic activity of Trig was demonstrated in dry matter accumulation stage of seeds i.e. stage III - IV, which subsequently decreases markedly, however, Trig transportation takes place from pericarp to seeds. A similar observation for caffeine too was showed (Baumann and Wanner, 1972) for C. arabica. Longer maturation duration and increased bean sizes are usually observed for plants grown at high altitude or in shade conditions (Muschler, 2001; Guyot et al., 1996). A positive relation between altitude and taster preferences of Arabica coffee from different territory of Costa Rica was reported (Avelino et al., 2005). The increase in trigonelline content of robusta beans from high altitude in the present study was further supported by earlier report of Avelino et al (2005). Though the slope exposure where in, the light or shade availability too have some influence on Arabica coffee quality as opined by Avelino et al (2005) in the present study, altitude influence only was investigated as it mainly influence biochemical profile of robusta beans (Guyot et al., 121

1996). Apart from this, in the present investigation, the range of altitude difference from where samples were collected was 3000-3700 ft and this mesoclimate conditions influence was found to be significant on both Trig and NA content of beans. In contrast to the study of Avelino et al (2005) negative relation between elevation and triogonelline content was found by Bertrand et al (2006), but this study is based on Near-Infrared Spectra (NIR) of some Arabica hybrids involving Sudanese-Ethiopian origins with traditional varieties. Such variations in metabolic profiles trend is a common phenomenon as it is indirectly influenced by some environmental and coffee agroforestry systems (Vaast et al., 2006). Similarly, roasting conditions and brewing methods too influence the metabolic profile of beans (Sridevi et al., 2011). Unlike other metabolites of Coffee, Trig is important from post-harvest processing aspects that imparts quality to Coffee. Because Trig alkaloids gives rise to many flavour (aroma) compounds, like alkyl-pyradines and pyrroles (De Maria et al., 1994). In view of this, the levels of Trig in developing seeds are important though we don t have any information about the actual role of this in developing fruits unlike polyamines (Sridevi et al., 2009). Shimizu and Mazzafera (2000) showed the role of trigonelline during inhibition and germination of coffee seeds. Though Trig production takes place in pericarp of young fruits, the same will decline in pericarp of ripened fruits (Koshiro et al., 2006). However, our data suggests that trigonelline accumulation takes place in seeds from stage III to stage V fruits. This was in concurrence with earlier studies (Zheng et al., 2004; Koshiro et al., 2006). All these substantiate the trend in Trig content and also its precursor nicotinic acid content in the present study. Moreover the content of Trig in harvested beans is on par with its trend in C. canephora as reported earlier which would be in the range of (0.75 to 122

1.24%) as reviewed by De Castro and Marracccini (2006). Certainly, the altitudes at which coffee plants grow had influence on Trig and NA content. 4.4.3. Influence of altitude on free diterpenes content Free diterpenes levels of beans in fruits collected from high elevation grown coffee plants were less (cafestol 420 ±9.5 µg.g -1 and kahweol 140 ±4.5 µg.g -1 ) compared to samples collected from low elevation (cafestol 1200 ±18.4 µg.g -1 and kahweol 450 ±7.5 µg.g -1 ) which were analysed by HPLC (Fig. 4.15, 4.16, 4.17). Though cafestol was detected right from the early stage of fruit development, kahweol was not found to be at detectable levels till stage III of fruit development. Similarly, cafestol was not detected in pericarp of stage III fruits. In case of stage IV fruits, both cafestol and kahweol were detected at significant levels but in the subsequent stages both these diterepenes were reduced drastically (Fig. 4.15, 4.16). This indicates that stage IV of fruit development is the most crucial stage, because at this stage both diterpenes were found in detectable levels. This trend was more or less same as in case of other metabolites such as caffeine and trigonelline. 123

Figure 4.15. Cafestol profiles (µg.g -1 dry weight) during ontogeny of C.canephoara CxR variety fruit (values are mean ± SD of five analyses); Different alphabet letters indicate the statistical significant between the altitude (p< 0.05). Figure 4.16. Kahweol profiles (µg.g -1 dry weight) during ontogeny of C.canephoara CxR variety fruit (values are mean ± SD of five analyses); Different alphabet letters indicate the statistical significant between the altitude (p< 0.05). In stage I & II fruits, stage III & IV (pericarp) kahweol not detected 124

Figure 4.17. HPLC analysis of kahweol (peak 1-RT 7.85) and cafestol (peak 2-RT 12.92): a) Standards, b) high altitude, c) medium altitude, d) low altitude. 125

Figure 4.18. Regression coefficients for effect of elevation on a) Variation in cafestol content in C. canephora b) variation in kahweol content in C. canephora This HPLC analysis of free diterpenes facilitated to obtain the data significantly (p<0.05). To validate the efficiency of extraction procedure one sample was extracted thrice and the coefficient of variation was 1.15 %. The linear relationship between cafestol / kahweol concentration and UV absorbance was perceived, wherein, the linearity was managed over the concentration range of 10 µg -150 µg ml -l. In HPLC the retention time (RT) for kahweol and cafestol were 7.85 and 12.92 (Figure 4.17). Unlike the method of Campanha et al (2010) that followed in Chapter 3, some changes were made to HPLC analysis of free form of cafestol and kahweol in fractions that obtained through GPC and SPE as explained in materials and methods part of this Chapter 4. In place of Nucleosil 100-5 C- 18 column of size 5μ X 250mm (HPLC Trennsaule, - Macherey-Nagel Gmbh & Co. KG., Duren, Germany), another column i.e. Nucleosil 120-3, C18, 250/4 column (Macherey Nagel, Gmbh, Germany) was used (Kurzock and Speer, 2001) to detect cafestol and kahweol at 230 nm and 290 nm by using acetonitrile: isopropanol (70:30) in place of acetonitrile: water: glacial acetic acid 126