2.1 Classification, Origin and Distribution of Coffee The genus Coffea belongs to the order Rubiales, family Rubiaceae, sub-family Ixoroideae and

Transcription

1 2.1 Classification, Origin and Distribution of Coffee The genus Coffea belongs to the order Rubiales, family Rubiaceae, sub-family Ixoroideae and tribe Coffeeae. Rubiales falls in the sub-class Asteridae that also include the most important family of plants comprising the horticultural vegetables called Solanaceae (including tomato, brinjal, potato, capsicum, chilli, tobacco etc.). Recent research on comparative EST analysis has indicated a gene to gene repertoire shared between coffee and tomato. Coffee also shares considerable synteny between Solanum lycopersicon (Lin et al., 2005). Conventional classification of coffee, which relied on morphology of plant specimens from botanical herbarium, was found to be very inadequate with the discovery of newer species (see, Charrier and Berthaud, 1985). The most widely used modern taxonomic classification of coffee was provided by Leroy, (1980) by using the type of gynoecium and placenta as the criteria to distinguish it from other genera in family Rubiaceae. The subdivisions of genera and subgenera were based on the length of the corolla tube (short or long) and the type of anther and style (extended or inserted) as well as the inflorescence style (monopodial or sympodial). This system classified coffee into two genera Coffea which have short corolla tube, inserted anthers and long style and Psilanthus having long corolla tube, non inserted anthers and short style. Recently, the Psilanthus genus has been subsumed into a single Coffea genus (Davis et al., 2006) with a total of 124 species (Davis et al., 2011). Africa, Madagascar and Mascarenes are the only three natural geographical areas where wild coffee exists and with no naturally existing species shared between them. Madagascar, Cameroon and Tanzania are the three recognized centre of diversities. Though coffee inhabits a diversity of forest types, most of them grow in humid, evergreen forests. In African subcontinent, coffee grows at the understory of the tropical forests. Out of the 124 species of coffee only two species is cultivated commercially, C. canephora (Robusta coffee) and C. arabica (Arabica coffee). C. liberica (Liberica, Liberan or Excelsa coffee) is grown mainly for local consumption and suffers a very poor cup quality. Nevertheless, it is an important crop with respect to introgression of disease resistance into the commercial varieties (Prakash et al., 2004). All the species in the genera Coffea, except C. arabica, are diploid carrying a diploid chromosomal number 2n=22. C. arabica is an allotetraploid that resulted from the natural cross of ancestors of the two present day diploids, C. canephora and C. 8

2 eugenioides with a recent origin of 650,000 years ago (Yu et al., 2011). Arabica coffee is a far more superior cup quality brew in terms of flavour and aroma. The caffeine content of Arabica coffee is also much lower than the more sturdy and a lower cup quality Robusta brew. The centre of diversity of C. arabica is agreed upon as the South- Western highlands of Ethiopia, whereas C. canephora are found along Ivory Coast and Central Africa. 2.2 Agronomic Cultivation of Coffee: Varieties, Cultivation Practices and Diseases/Pests of Coffee Coffee is cultivated in 72 producing countries with a global production of million bags (each is a 60kg bag) for the year Around 60% of the cultivated coffee is Arabica type (88.8 million bags) whereas, the remaining the Robusta type (56.2 million bags) (International Coffee Organization statistics; There are four coffee growing regions in the world: Africa, Asia and Oceania, Mexico and Central America and South America. South America alone contributes 47% of total production, 75% of which comes from Brazil. Countries like Burundi, Ethiopia, Rwanda, Honduras, Uganda, Nicargua and Guatemala depend on coffee for a major proportion of their Gross Domestic Product. The three cultivated species of coffee differ in their botanical aspects and even in their cultivation. C. arabica is generally a large bush with dark green oval leaves and oval fruits that take 7-9 months for complete maturation. The seeds, also called the beans, are flat. C. canephora are robust shrubs or trees (up to 10 metres high) with large oval leaves and round fruits that take 9-11 months for complete maturation. The fruits and seeds are oval and smaller than C. arabica. C. liberica are large trees (up to 18 metres high) with large leathery leaves and large fruits and seeds that take 12 months for maturity. C. arabica is highly susceptible to diseases and pest like leaf rust, berry borer and nematodes whereas, C. canephora is more resistant (Waller et al., 2007). The resistance to wilt disease is, however, lesser in C. canephora. C. arabica, which show deep root system, are grown at higher altitude ( metre above sea level or more), require lower optimal temperatures (15-24 C) and lower rainfall ( mm) and have lower yields ( Kg seeds/ha). Whereas, C. canephora have shallow root system, grow at lower altitude (0-700 metre above sea level), require higher optimal temperatures (24-30 C), higher rainfall ( mm) and give better yield ( kg seeds/ha). Arabica brew character is acidity whereas for Robusta it is bitterness. The chemical composition of the beans also varies with the species (see, 9

3 section 2.6). Finally, C. arabica are self-pollinated and C. canephora are strictly selfincompatible leading to higher diversity of gene pool in the latter. In general, coffee exhibit high resistance to interspecific or cross hybridization making it difficult to develop new hybrid varieties or introgressed lines. C. arabica is cultivated in the form of two main varieties: Typica and Bourbon ( Many strains and cultivars have been developed from these two varieties for e.g., Caturra (a mutant compact form of Bourbon grown in Brazil, Colombia), Mundo Novo (a cross between typical and bourbon grown in Brazil), Kent (Indian-developed disease tolerant Arabica), Catuai (developed hybrid of mundo Novo and Caturra), Tico (Central America), the dwarf San Ramon and the Jamaican Blue Mountain. The main cultivated variety of C. canephora is Robusta, also called Conillon in Brazil. C. liberica is cultivated as var. liberica (Liberica coffee) or var. dewevrei (Excelsa coffee). Apart from these many introgressed and hybrid varieties have been developed, for e.g., Arabusta is a hybrid developed by cross between diploid Robusta and tetraploid Arabica, Hybrido de Timor (HDT) is a hardy natural cross between Robusta and Arabica and resembles Arabica as a tertaploid variety. Leaf rust resistant Catimor is a hybrid of HDT and Caturra, Sarchimor is a disease resistant Catimor, Ruiru Eleven, developed at the Coffee Research Station at Ruiru in Kenya, is resistant to coffee berry disease and to coffee leaf rust and Icatu hybrids are the result of repeated backcrossing of interspecific Arabica x Robusta hybrids to Arabica cultivars Mundo Novo and Caturra. Coffee is traditionally a shade grown plant growing under the canopy of trees like in a complex ecosystem that is very important for the biodiversity rich areas where it is cultivated (Marie-Vivien et al., 2014). However, the introduction of sun-tolerant varieties with substantial improvement in crop productivity is replacing the traditional cultivation practices (Gobbi, 2007). Many researchers have observed it as a threat to biodiversity in terms of soil and water deterioration due to heavy usage of fertilizers and pesticides, soil erosion and diminishing natural habitat of the bee and bird population (Elmqvist et al., 2013). Though the debate of benefits of shade grown coffee or the devastating predictions of full sun cultivation is very strong, agronomist believe that careful design of shade canopy and meticulous selection of the genotypes of plantation would benefit the traditional shade cultivation (Ricci et al., 2011; Schmitt-Harsh et al., 2012). 10

4 The most important agronomic qualities that need an address on sustainable coffee cultivation include traits like yield, quality, resistance to pest and disease, heat and drought tolerance, root structure, adaptability and early maturing. Tremendous efforts have been made for developing varieties with improved traits through both breeding and transgenic efforts. Concerning disease and pest resistance, the most devastating disease in coffee is the orange rust or coffee leaf rust (CLR) caused by the obligate biotrophic fungus Hemelia vastatrix Berk. & Broome (Basidiomycetes: Puccinales) and the most dangerous pest is the coffee berry borer (Hypothenemus hampei) (Ferrari) (Coleoptera: Curculionidae) (Silva et al., 2006). CLR begins with the germination of uredenospores of the fungus on the abaxial surface of the leaf, which later form the appresorium structures helping the pathogen to penetrate into mesophyll tissues through the stomatal opening and leading to differentiation of haustoria. Heavy colonization of tissues by the fungus leads to uredospore release in bouquet shaped uredinia (Silva et al., 1999). CLR causes premature leaf defoliation, leading to reduction in photosynthetic capacity and year-byyear loss of vigour and may cause as high as 20-40% yield loss. Resistance to CLR is explained by the classical Flor s gene for gene interaction hypothesis (Flor, 1971). There are nine virulent genes v1-v9 in 39 known rust races counteracted by nine host resistance genes SH1-SH9 in different varieties/species of coffee (Rodrigues et al., 1975). SH1, SH2, SH4, and SH5 are found in C. arabica of Ethopian origin, SH3 in C. liberica and SH6, SH7, SH8 and SH9 in HDT (Bettencourt and Rodrigues, 1988). At minimum all these nine genes are required for complete resistance to CLR, though partial resistance that have longer latent period and low spore formation may be of importance to counteract different races of pathogen leading to a much durable resistance in fields (Herrera et al., 2009). The host disease resistance genes carry a NBS domain (nucleotide binding site) that binds to ATD/ADP or GTP/GDP and a Leucine Rich Repeat (LRR) (Meyers et al., 2003). Many NBS domain-containing genes with unknown functions, also called the Resistance Gene Analogs (RGA) have also been reported in literature (Noir et al., 2001). HDT has been a very successful variety to combat CLR and has benefited a sustainable cultivation of coffee for four decades before cases of breakdown of resistance due to emergence of new races appeared. Susceptible genotypes culminate into a compatible reaction with the pathogen whereas; resistant genotypes undergo incompatible reaction leading to localized cell/tissue death also called a hypersensitive response (Silva et al., 11

5 2006). Various recent molecular studies on compatible and incompatible interactions have laid insights into pathogenesis and many candidate genes for combating the disease in a sustainable fashion has been postulated for e.g., the CaNDR1 (Non-Racespecific Disease Resistance) (Fernandez et al., 2004). Apart from CLR, coffee plantations also suffer from other diseases that are making greater global impact due to climatic change like the coffee berry disease (Colletotrichum kahawae JM Waller & PD Bridge) (Van der Vossen and Walyaro, 2009) and the coffee wilt disease (Fusarium xylarioides) (Rutherford, 2006). Among the pests of coffee, coffee berry borer is, globally, the most devastating in terms of prevalence, yield loss and containment. The coleopteran feeds exclusively on the seeds and reproduces within the endosperm causing huge yield losses (Rodríguez et al., 2013). Other pests like Leaf miner (Leucoptera coffeela, Guérin-Méneville) are also emerging as global threat to coffee plantations (Ghini et al., 2008). 2.3 Cultivation of Coffee in India India is one of the few countries known to cultivate coffee in the most eco-friendly manner and exclusively under shade cultivation (Marie-Vivien et al., 2014). In India, coffee is grown in the traditional areas (states of Karnataka, Tamil Nadu and Kerala), in the non-traditional areas (Andhra Pradesh, Odisha) and the North-Eastern regions. Karnataka by far has the highest share in production (around 70% produce of the country). More than 90% coffee in the country is grown in small holdings by small share-holders with area less than 10Ha. The final estimate for the crop year indicate that greater than 65% Indian cultivation relies on Robusta and the remaining on Arabica (Coffee Board of India, Many cultivars of C. arabica are cultivated in the country like Cauvery, Chandragiri, and sln9. However, there are only two released cultivars of C. canephora: S274 and CxR. Cultivar S274 has been developed by Central Coffee Research Institute, Chikmagalur-an independent institute under the Ministry of Commerce, Government of India-through a process of mass selection (Mishra et al., 2012). This genotype used in all the studies in this doctoral research have resemblance to pure Robusta genotype. CxR is an interspecific cross between C. congensis and S274 having high resemblance to C. canephora var. Robusta S274 though with slightly less caffeine content. CxR is fast replacing S274 from cultivation due to superior agronomic qualities. Neverthless, S274 is still widely used for breeding programs. The C. arabica variety selected for the work is a pure Ethiopian germplasm called tafarikela. 12

6 2.4 Botanical Description and Ontogeny of Coffee Fruit Coffee has a clustered inflorescence type and the fruit is botanically a drupe or resembling cherry, although the misnomer berry is oftenly used (Charrier and Berthaud, 1985). The fruit encompasses two seeds that are majorly endosperms and called bean. Fruits with single bean (peaberry) or three beans are also observed in considerable numbers. The exterior-most layer is the epicarp which is leathery and green during immature stages and turns yellow to pinkish during ripening. Epicarp is followed by a fleshy mesocarp and a hard lignified endocarp that becomes membranous parchment layer after drying. Figure 2.1: Ontogeny of coffee berry and the major physiological changes The epi-, meso- and endocarp are together called the pericarp and enclose two seeds in mature fruits, each covered with a thin membranous layer called spermoderm or silver skin which actually is the remnant of perisperm that is consumed by endosperm during seed development. After flowers open, the anthesis may last for one to a few days or 13

7 until pollination after which, the flower dies and the developing fruit undergo a period of minimal growth in a latent state for 2-3 months (2 months for C. arabica and 3 for C. canephora). At this stage, the fruit is also referred as pinhead stage. Cellularization of endosperm leads to a discoid shaped translucent white endosperm encompassing an embryo. The initial growth of the fruit is due to expansion of perisperm tissue that is derived from the maternal nucellus or the integument tissue. The next phase is the cellular expansion of endosperm which uses up most of the perispermic tissue for its metabolism reducing the perisperm into a thin layer which can be seen as a silver skin in the dried mature bean. Once the endosperm has occupied the entire perispermic phase, dry matter in the form of dehydrins, late embryogenic proteins (LEA) (Hinniger et al., 2006) and seed storage proteins like 11s globulin (Rogers et al., 1999; Marraccini et al., 1999) begins. By the time of ripening, the mesoderm turns to mucilaginous pulp and the seed is hardened. The colour of the fruits also transit to a yellow-red or pink to reddish colour. There are detailed reviews on the morphology and anatomy of fruit development and on the physiology of the ontogeny of fruits (de Castro and Marraccini, 2006; Eira et al., 2006). Figure 2.1 shows the ontogeny of the coffee fruits and the relative time of development from anthesis onwards. Recent research has shown various phenological markers for the developing endosperms like the α- galactosidase for the green stage, caffeine synthase for the yellow green stage and iso-citrate lyase and ethylene receptor 3 for late maturation stages (de Gaspari-Pezzopane et al., 2012). 2.5 Diversity and Genome Evolution of Coffee Different methods like RAPD (Lashermes et al., 1996a), SSR (Maluf et al., 2005), RFLP (Lashermes et al., 1996b; Spaniolas et al., 2006), AFLP (Pearl et al., 2004), ISSR (Aga et al., 2005), ISTR (Aga et al., 2006), microsatellite or EST-SSR (Aggrawal et al., 2007; Poncet et al., 2006) and plastid intergenic regions (Maurin et al., 2007) have been employed to study the diversity of coffee in different germplasm accessions of cultivated species or wild collections. ISSR and analysis of chloroplast genome has been the most accurate methods in the analysis of the diversity dendrograms for both interspecies and intraspecies comparisons. A high density genetic map for C. arabica was developed by the use of 288 AFLP markers (Pearl et al., 2004). Before the publication of the coffee genome draft (Denoeud et al., 2014), several analysis of BACend libraries generated for the sequencing project provided important insights into the 14

8 study of the evolution of coffee genome. The tetraploid C. arabica which by original studies on diversity was believed to have originated by allotetraploidization between the C. canephora and the C. eugeneoides genome was calculated to have a recent origin dating back a mere 650,000 years from the parent genomes (Yu et al., 2011). Furthermore, comparative synteny analysis of BAC libraries with the Arabidopsis genome and the then ongoing tomato (Lycopersicon esculentum) and grapes (Vitis vinefra) genome showed high synteny between these species of the asterideae sub-class. The synteny is closest in the order of Vitis followed by tomato and then by Arabidopsis (Guyot et al., 2012; Dereeper et al., 2013). 2.6 Chemical Profile of Coffee Beans Coffee beans are harvested when the fruits are bright red and glossy and dried to 10-14% moisture levels prior to dehusking. The seeds, commonly called green beans, are then marketed for roasting. The chemical components of green beans largely depend on the species and variety of the coffee used whereas the brew properties largely depend on the procedure adopted for roasting. Since the coffee aroma is dependant largely on the roasting process, the amount and type of volatiles follow temperature-time relation during the roasting process. Certain speciality coffee from India involve a procedure of natural curing for 3-4 days in moisture after drying (Monsoon coffee). However, an approximation of various components and amount in the Arabica and the Robusta beans and also in an average cup of instant coffee as discussed in the Proceedings of International Seminar on Coffee and Health at the 40 th anniversary of ICO held at Cartegena, Columbia, 2003 may be found in Table 2.1 and in description below. Roasting of coffee beans may lead to detoriation of many of the components like amino acids, saccharides, and chlorogenic acids that are found in the green bean and may even give rise to different components like for example the formation of aromatic compounds by maillard s reaction between reducing sugars and trigonelline. 15

9 Table 2.1: Approximation of green and roasted beans of Arabica and Robusta coffee and instant coffee powder on percentage dry weight basis. Component Arabica Robusta Instant Green Roasted Green Roasted Minerals Caffeine ~ ~ Trigonelline Lipids Total Chlorogenic acids Aliphatic acids Oligosaccharides Total polysaccharides ~6.5 Amino acids Proteins Humic acids Source: (ICO sponsored conference, 2003) Non-Volatiles in Coffee Apart from the saccharides, proteins, free amino acids, minerals and lipids, the nonvolatiles in green coffee bean include important bioactives like chlorogenic acid, caffeine and minor amounts of other methylxanthines, trigonelline, soluble fiber, and diterpenes. Methylxanthines: Methylxanthines like caffeine, theobromine and theophylline are purine alkaloid type of bioactives having neurostimulatory, vaso-dilatory and diuretic properties. Caffeine (1,3,7-trimethylxanthine) by far exhibits the strongest physiological and pharmocological activity amongst other methylxanthines (Clifford and Wilson 1985). The base structure of methylxanthines involves a purine ring which is methylated in different nitrogen positions. (Figure 2.2). Caffeine under low consumption enhance the mood and performance and enhance the cognitive behaviour. Caffeine is a non-specific antagonist of the A1 and A2 adenosine receptors in the 16

10 central nervous system (CNS), thus preventing adenosine from inhibiting the release of neurotransmitters, glutamate, GABA, acethylcholine and monoamines. This leads to stimulation of the nervous system. However, higher doses may lead to insomnia, and anxiety. Trigonelline: Trigonelline is a pyrimidine alkaloid derived by enzymatic methylation of Nicotinic acid (Niacin). Trigonelline combines with reducing sugars during roasting to produce the aroma of coffee (de Castro and Marraccini, 2006). During roasting a portion of trigonelline is demethylated to form niacin or nicotinamide which is a component of vitamin B complex. Trigonelline has potent inhibitory activities against invasiveness of carcinoma cell lines. Trigonelline is present in higher amounts in the Arabica coffee (Ky et al., 2000). Chlorogenic acids: Chlorogenic acids (CGA) are phenolic compounds formed by esterification of trans-cinnamic acids like ferulic acid, caffeic acid and p-coumaric acid with quinic acid (Clifford and Wilson 1985, Sridevi and Giridhar 2008). They impart astringency, acidity and bitterness to the brew.the major chlorogenic acids present in coffee are caffeoylquinic acids, dicaffeoylquinic acids, feruloylquinic acids and, less abundantly, p-coumaroylquinic acids and caffeoyl-feruloylquinic acids each of which exist in many different isomeric forms (de Castro and Marraccini, 2006). Caffeoylquinic acids comprise 80% of the total CGA and the most abundant isomer being 5-caffeolyquinic acid (60% of the total CGA). Cafestol and Kahweol: Cafestol and kahweol are the two major pentacyclic diterpenes alcohols present in green beans. They are shown to have potent hepatoprotective and anti-carcinogenic activities in vitro, but at the same time known to elevate levels of homocysteine and low density lipoproteins in plasma increasing the risk of cardiovascular diseases (reviewed by de Castro and Marraccini, 2006). The diterpenes have high affinity to adsorb to paper filters and can be removed from brew by use of candle filters. Soluble dietery fibers: The soluble dietery fibers in coffee majorly consists of galactomannan and type II arabinoglycans. Lipids: Lipids are a major component of green bean (as high as 14% dry weight in Arabica coffee). The fatty acids are majorly esterified to glycerol and are of unsaturated nature. Linoleic acid, oleic acid and linolenic acids are the major fatty acids present in coffee (reviewed by de Castro and Marraccini, 2006). 17

11 2.6.2 Volatiles in Coffee There are 100 volatile compounds present in green beans-mostly formed during maturation phase and after the roasting process and include pyridines, pyrrols, pyrazine and methyl-nicotinates Composition of an Average Cup of Brew The compositions of a cup of brew may vary due to a number of factors like the variety, the blend (mixture of Robusta and Arabica), the post-harvest processes as well as the roasting procedures used. However, the average cup of instant coffee is approximated as in Table Health Benefits Imparted by Coffee Consumption Though the position of coffee in functional food groups is debatable, there are many reports of beneficial effects of drinking moderate amounts of coffee. Coffee consumption, in general, improve mood and performance of the brain. Coffee consumption is proven to have potent hepatoprotective effects on the consumer by lowering the incidence of liver cirhossis and reversing the ill effects of alcohol on hepatic tissue. Coffee drinking is also associated with lower risks of Parkinson s and Alzheimer s diseases and the incidence of type II diabetes Methylxanthines and Health Implications Methylxanthines in general and more potently caffeine, have neurostimulatory roles. Nevertheless their participation in neuro-psychological responses like reward, motivation and addiction is generally disapproven (Nehlig, 1999). However, it is known that high consumption lead to disturbed sleep pattern, dehydration and altered heart beat rhythm. Caffeine also stimulate the secretion of gastric acids leading to severe heartburn in sensitive individuals. In rat models, caffeine was shown to be risky during pregnancy. A few deaths have also been registered due to overdoses of caffeine tablets. Nevertheless, caffeine has found a place in the pharmacology sector as a synergist to anti-pyretic and analgesics and in the cosmetic sector as aging reversal effect. Methylxanthines are also used in surgical process as diuretic and heart rhythm activation Caffeine Caffeine is most identifiable compounds of coffee in literature searches. It is present in a very few phylogenetically-distinct plant species like coffee (Coffea sp.) tea (Camellia sp), cocoa (Theobroma sp), mate (Ilex sp), cola, guarana and the floral parts of citrus (Kretschmar and Baumann, 1999). However, xanthosine, the committed precursor of 18

12 caffeine biosynthetic pathway, is ubiquitous to all plants (Stoychev et al., 2002). The caffeine content of wild and cultivated species of coffee is known to show a discontinuous variation (Campa et al., 2005). The biological role of caffeine was earliest described by the Chemical defense theory where these secondary metabolites are proposed to provide protection from pestivory and herbivory. Genetically altered tobacco plants able to biosynthesize caffeine showed lower preference as food for lepidopteran pest (Uefuji et al., 2005). Caffiene was also shown to be inhibitory for many insects in transgenic Chrysanthemum (Kim et al., 2011). However, the antagonism of caffeine to coffee pests and pathogens like CLR, berry borers and other coffee pests is indefinite. In another theory caffeine may impart allelopathic reactions by leaching into the soil from water imbibed seeds and preventing germination of other plants in vicinity (Anaya et al., 2002). Furthermore, of more important biological role, the caffeine in nectar of coffee flowers imparts added advantage of memory to visiting bees for revisiting the same plantations (Wright et al., 2013). Since, bee pollination leads to upto 40% increase in seed set in coffee plantations (Klein et al., 2003), caffeine has a very prominent ecological role. 2.7 Caffeine Biosynthetic Pathway The research on biochemistry of caffeine synthesis dates back to the mid 20 th century. The hypothesis that caffeine derives its xanthine ring from purines was dated to the year 1962 (Anderson and Gibbs, 1962). Tea extracts purified using polyclar columns identified SAM dependant N 1 and N 7 methyltranferase activies in the partially purified protein extracts resulting in the formation of 7-methylxanthosine and theobromine, respectively (Suzuki and Takahashi, 1975). These results and many others still relate to a state of dilemma in the caffeine biosynthetic route since many papers did not indicate involvement of 7-methylxanthine or xanthosine since these compunds were not detected in the radiolabelled fraction of methylxanthine in in vitro or in situ radio-tracer experiments (see, Baumann, 2006). However, as debated the in vitro and in situ radiolabel feeding experiments are not indicative of the nautral mechanism occurring in the plant. Nevertheless, the first time purification of tea caffeine synthase enzyme and N-terminal sequencing defined the core caffeine biosynthetic pathway to follow the 7- methylxanthine 7-methylxanthine theobromine caffeine route (Kato et al., 2000). The major followed biochemical pathways for caffeine biosynthesis is depicted in the Figure

13 Figure 2.2: The caffeine biosynthetic pathways in tea and coffee. The core pathway from xanthosine to caffeine is indicated in the red box. 2.8 Sources of Xanthine Ring The possiblie source of the purine ring derived in caffeine was probed by in situ radiotracer feeding experimetns (leaf disk or tissue cultured cells) or in vitro radio-tracer incorporation to added precursors by using enzyme extracts prepared from tea and coffee enzyme tissues. Ogutuga and Northcote (1970) with their initial work on tea callus showed that xanthine ring from adenine derived from breakdown pool of nucleotides and not from the purine pools is the source of the xanthine ring. Suzuki and Takahashi (1976), however debate that the main xanthine ring of caffeine in tea was derived from the nucleotide pool and not by the degradation pathways of nucleotides. Pulse-chase experiments using 8-14 C adeninine fed to stamens of tea indicated 25% label incorporation into theobromine and caffeine whereas, 50% was salvaged into DNA and RNA nucleotides (Fujimori and Ashihara, 1993). IMP (inosine monophosphate) also was shown to be precursors for xanthosine formation. Radiotracer feeding of 14 C-theobromine to leaf segements incorporated the label only to caffeine 20

14 and not any other metabolite indicating theobromine as the immediate precursor of caffeine (Ashihara et al., 1996). However, only traces of 14 C-xanthine was incorporated into caffeine while most went to the catabolic route until inhibitors of catabolism was added to the reaction. In a more recent research it was shown that significant amounts of radioactivity was incorporated from methyl 14 C-SAM and methyl 14 C-methionine into caffeine and theobromine indicating a new entry pathway for the source of xanthine ring as SAM SAH adenosine adenine AMP IMP Xanthosine to caffeine pathway (Koshishi et al., 2001) as shown in Figure 2.3. It was shown in coffee that theobromine but not caffeine has separate routes of entry for xanthine ring (Nazario and Lovatt, 1993). 14 C-xanthine incorporated into caffeine but not theobromine indicating that theobromine may not be the intermediate in caffeine biosynthesis through xanthine. Detailed review of biochemical experiments carried out on tea and coffee for the search of the the source of xanthine is provided by Suzuki et al., (1992). Though Baumann (2005) hypothesizes more twist in the caffeine biosynthetic pathway, the most popular pathway is adenosine adenine adenosine mono-phosphate (AMP) inosine mono-phosphate (IMP) xanthosine mono-phosphate (XMP) Xanthosine 7-methylxanthosine 7-methylxanthine 3,7-dimethylxanthine or theobromine 1,3,7-trimethylxanthine or caffeine. The different entry sources for xanthine ring from the above cited literature are listed as a) Inosine 5 -mono-phosphate derived from de novo purine nucleotide synthesis. b) Different nucleosides and nucleotides (GMP and AMP) derived from salvage of nucleic acids. c) Adenosine release from the deamination of SAM. 2.9 Caffeine Degradation and Caffeine Turnover Feeding experiments using methyl labelled and ring labelled caffeine indicated low incorporation of the radioactivity into ureides, allantoin and allantoic acid whereas high incorporation of label from radiolabelled theophylline to ureides (Kalberer, 1965; Suzuki and Waller, 1984). It was concluded that the rates of caffeine catabolism is higher in young fruits than in mature fruits. In contrary to the earlier notion, it was later shown that caffeine degradation to theophylline was higher in mature tissue than in young tissues (Mazzaferra et al., 1991). Detailed studies revealed clearly that the route of caffeine degradation is through Caffeine theophylline 3-methylxanthine xanthine uric acid allantoin allantoic acid urea NH 3 and CO 2 (Ashihara et al., 1996). 21

15 Caffeine dynamics is largely believed to be through a ratio between the biosynthetic activity and the caffeine turnover, since theobromine and 7-methylxanthine are degraded at a very low rate (Mazzafera et al., 1994). This ratio has been found to be higher in young fruits as well as in higher caffeine accumulating species like C. canephora compared to low caffeine containing species like C. liberica var. dewerei (Mazzaferra et al., 1994). Based on the studies so far, the consensus main catabolic pathway in coffee plants is: caffeine theophylline 3-methylxanthine xanthine uric acid allantoin allantoic acid glyoxylic acid + urea NH 3 +CO 2 (Mazzafera, 2004). Depending on the plant species, other minor routes may operate with the formation of theobromine and 7-methylxanthine, which are salvaged for caffeine formation since they also appear in the biosynthetic pathway. Conversion of caffeine to theophylline is the rate-limiting step in caffeine biodegradation in coffee plants N-Methyltransferases Involved in Caffeine Biosynthesis The first caffeine synthase isolated was from tea, called the Tea caffeine synthase 1 (TCS1) which was identified in the purified enzyme extracts of leaves of Camellia sp (Kato et al., 1999). N-terminal sequencing lead to the sequencing of 1438bp fragment of the complete coding sequence with the recombinant enzymes showing methyltransferase activities on the N 1 and N 7 position in the substrates 7- methylxanthine and theobromine, respectively (Kato et al., 2000). Based on the TCS1 sequence and Arabidopsis O-methyltransferase, seven homologues namely CtCS1- CtCS7 were obtained (Mizuno et al., 2001) from a RACE library of C. arabica immature seeds and young leaves. Whereas, CtCS3 and CtCS4 did not show any activity for the intermediates tested, CtCS1 showed N 7 methylation to xanthosine (Mizuno et al., 2003a). This sequence named as CmXRS1 (Coffee methyl xanthosine synthase) was specific to xanthosine substrate and not to XMP, 7 methylxanthine and theobromine. Degenerate primers designed on the conserved regions of TCS1 were also used to amplify a probe for cdna library screening, partial sequencing and subsequently physical mapping leading to cloning of CaMXMT1 having 7- methylxanthine to theobromine and paraxanthine to caffeine converting ability, the former to a 5-fold higher activity (Ogawa et al., 2001). Other clones namely CaMTL1, 2 and 3 showed no activities on the substrates tested. Further efforts for RACE library screening identified sequences CtCS6 and 7 as homologues of earlier identified but 22

16 unpublished theobromine synthase CTS1 and CTS2. CtCS6 was named as caffeine synthase (Mizuno et al., 2003b) and showed dual activity for conversion of 7- methylxanthine to theobromine and theobromine to caffeine. It also had a high affinity for paraxanthine. In a separate study using primers designed on the CaMXMT1 sequence Uefuji et al., (2003) amplified two novel sequences other than CaMTL3 (renamed as CaXMT1 or xanthosine methyltransferase) namely CaMXMT2 and CaDXMT1. Recombinant CaXMT1 formed 7-methylxanthosine from xanthosine substrate, recombinant CaMXMT2 showed activity towards conversion of 7- methylxanthine to theobromine and to a lesser extent paraxanthine to caffeine whereas, recombinant CaDXMT1 had activity towards conversion of 7-methylxanthine to theobromine and theobromine to caffeine and a much higher activity for conversion of paraxanthine to caffeine (for review see Kato and Mizuno, 2004). The crystal structures of two C. canephora NMTs-CaXMT1 and CaDXMT1 was resolved at high resolution in X-ray diffraction (McCarthy and McCarthy, 2007) and the structure were found to be similar in the base structure to Salicylic acid methyltransferase (SAMT). Crystal structure for CcDXMT1 and CcXMT1 almost superimposed with each other and the SAM/SAH binding pocket at similar conformation and position in SAMT, CcDXMT and CcXMT1 with extensive hydrogen bondings and vander walls interactions. Substrate binding pockets also map to similar regions in SAMT (Zubieta et al., 2003), CcDXMT and CcXMT1 and the authors noted subtle changes in or near the active site leading to substrate discrimination. The enzymes were predicted and later proved to form homo and heterodimers within themselves and with each other respectively (Kodama et al., 2008) Sequence analysis of N-Methyltransferases Sequence analysis of NMTs indicate similarities to the carboxy-methyl transferase, jasmonic methyl transferaes and the O-methyltransfersases all that fall under the SABATH superfamily. These genes are characterized by the absence of motif B but contain motif B' instead (for review see, Kato and Mizuno, 2004). The general motif structure is represented in Figure 2.3. Motifs A, B' and C are the SAM binding motifs. 23

17 Figure 2.3: General structure of motif B' N-methyltransferases:. The conserved regions are shown as I to VII and the SAM binding motifs A, B' and C are represented closed boxed. The conserved proline and cysteine residues are also shown Phylogenetic relation of N-Methyltransferase The sporadic occurrence of caffeine in taxonomically unrelated plants has led to the idea that the origin and evolution of caffeine in these species occurred by mere chance, a process leading to convergent evolution (Pichersky and Lewinsohn, 2011) wherein two plants evolve same biochemical pathways just by co-incidence. Hence, the mechanisms of evolution of caffeine in these plants may differ from each other. Phylogenetic trees indicate that caffeine biosynthetic NMTs might have evolved from a repertoire of similar genes involved in transfer of methyl group from SAM to nitrogen or carboxyl group of plant small molecules especially those involved in scent (Scalliet et al., 2008) and distant signalling (Seo et al., 2001). These genes in the present date form a large super-family of SAM-dependant methyltransferases called the SABATH super-family (Salycilic Acid methyltransferase-benzoic Acid methyltransferase- THeobromine synthase) (Kato and Mizuno, 2004). The SABATH members share sequence homology as well as structural homology amongst themselves as evident from the X-ray crystal structures of Clarkia breweri salicylic acid methyltransferase (SAMT) (Zubieta et al., 2003), Arabidopsis thaliana indole 3-acetic acid methyltransferase (IAMT) (Zhao et al., 2008), C. canephora xanthosine methyltransferase (XMT1) (McCarthy and McCarthy, 2007), and C. canephora di-methylxanthine methyltransferase (DXMT1) (McCarthy and McCarthy, 2007). SABATH family comprise of 24 members in A. thaliana (D Auria et al. 2003) and 41 members in Oryza sativa (Zhao et al. 2008). In O. sativa the SABATH members (OsSABATH) are localized in clusters and organized in tandem repeats indicating local duplication process in the genes. Since the members of SABATH play an important role in plant 24

18 ecological interactions by way of altering the volatile compounds, the gene family has been extensively studied with an evolutionary biology perspective Mechanism of Gene Duplications and Fixation Gene duplication is a mechanism for generating novelty of function in the sister genes. Gene duplication may occur by various mechanisms (Chain and Evans, 2006) involving duplication of a small segment of genome in tandem array fashion (Brown et al., 1972) or by transposition or even by whole genome duplication and polyploidization. In any population, gene duplication is the major source of generating novelty by allowing one sister counterpart to perform the primitive functions whereas the other to change free of evolutionary constraints. Hence as soon as a duplication event occur the sister paralogue (homologues descended by duplication) experience a selective pressure depending on the evolutionary fitness of the gene and the duplication event. However, in most cases the duplicate gene undergoes pseudogenization (Walsh, 1995) as a result of reduction of fitness due to overburden of gene copies and are slowly lost from the population. However, at low frequency, the evolutionary forces like positive selection (Emes and Yang, 2008), negative selection (Eirín-López et al., 2004) as well as neutral evolution (for review see Nei et al., 2010), after several generations, may lead to retention of the sister paralogues and fixation (high allele frequency) in the population (Nowak et al., 1997). Over a long evolutionary time scale, the orthologous genes (homologues descended by speciation) in different species may still carry evidences of the nature of evolutionary forces that shaped up their evolution. The rate (ω) of nonsynonymous mutations or Kn or dn (mutation in nucleotide leading to change in amino acid) to that of synonymous mutations or Ks or ds (silent mutations not causing amino acid changes) in the present day proteins, as well as the position of the effected amino acid could be used to depict the nature of the evolutionary force. In general, ω>1 is indicative of positive selection ω<1 of negative selection and ω=1 of neutral evolution. Accordingly, nine different models apart from pseudogenization have been proposed until date to describe the mechanisms of fixation and retention of the duplicated paralogues (for review see Innan and Kondrashov, 2010). a) Pseudogenization: The sister homologue gathers deleterious mutations and is lost from the population over time. b) Neo-functionalization: The original gene is sufficient to carry out the ancestral function and the sister homologue is free to mutate and develop new function, thus leading to positive selection in one branch. 25

19 c) Duplication-Degeneration-Complementation: Degenerate neutral mutations may occur in the sister paralogues such that the efficacy of the pairs of genes is reduced individually. d) Sub-functionalization: In this model, the ancestral function is carried out in parts by the two sister paralogue. Usually applicable to promoter regions and effect spatiotemporal gene expression. e) Escape from adaptive conflict (EAC): Escape from adaptive conflict states that when a multifunctional ancestral duplicates then both the sister paralogue may undergo positive selection over the course of further generations to improve one or more of the ancestral functions which, together in the ancestral gene would have been impossible. The adaptive conflict is escaped by both the paralogues as one protein performs one task better than the other and the other duplicate performing the other task better than the first one. f) Positive dosage: As the copy number of a gene doubles it may add up to fitness due to dosage benefit. g) Sheild against deleterious mutation (purifying solution): The sister paralogues strictly control the rate of deleterious mutations in each other such that they are preserved over long period of time. h) Adaptive radiation: Certain gene copies in the population pre-adapt to new environmental conditons while still performing the ancestral function. i) Permanent heterozygote: Hetrozygote advantage leads to preservance of both the sister paralogues. j) Diversifying selection: When gene duplication favours polymorphism then many sister polymorphic copies would be retained. Finally, duplicated genes located in tandem repeats on the chromosomes are prone to homogenization of sequence information through either Birth-and-Death evolution (Nei et al., 1997; Nei et al., 2000; Nei and Rooney, 2005; Piontkivska et al., 2002) or gene conversion (concerted evolution) (Chen et al., 2009). Gene conversion (Aguileta et al., 2004) leads to non-reciprocal exchange of genetic material between homologous genes arranged in a tandem array leading to difficulties in estimation of dn/ds values. As a result of gene conversion over the entire gene or small stretches of the gene, the sequence divergence is greatly reduced between the paralogues as compared to orthologues. The proportion of synonymous (Ps) and non-synonymous (Pn) differences per site, which is calculated as the sum of synonymous or non-synonymous transition 26

20 and transversions divided by number of synonymous and number of non-synonymous sites respectively, is used to define sequence divergence. Another characteristic feature of segments undergoing gene conversion is that these regions have a higher GC content compared to the flanking regions. On the other hand, Birth-and-Death evolution maintains low sequence divergence by strong purifying selection that selects against deleterious mutations. Concerted evolution would subsequently lead to sequences being more similar within species than between. Concerted evolution and birth-and-death evolution have different consequences with respect to the observed synonymous differences (Ps) within and between species (Chen et al. 2009, Piontkivska et al. 2002). In case of purifying selection without concerted evolution, DNA differences are biased towards synonymous sites. Thus, the intra-specific and inter-specific synonymous differences (Ps) are similar. On the other hand, if concerted evolution drives the evolution of the genes the intra-specific synonymous differences become much lower than the inter-specific synonymous differences Isolation of Theobromine Synthase Gene from C. canephora Using homology based PCR techniques a number of partial genomic clones similar to CaMXMT1 were isolated from the SspI digested library of C. canephora var. robusta cv. S274 genomic DNA. These partial clones namely CS2A, CS2B, CX8 and CX10 ranged from sizes 1729bp, 1710bp, 2003bp and 1943bp, respectively (Satyanarayana, 2005a). Simultaneously a 728bp upstream genome walking product (clone petsspi) was identified as the promoter of the putative theobromine synthase gene (Satyanarayana et al., 2005b; Satyanarayana et al., 2006). This promoter was capable of driving the expression of uida gene as observed by transient gus assay in transformed tobacco, leaf and coffee endosperms (Marraccini et al., 2006). Amplification with the forward primer designed based on 5' end of the petsspi clone sequence and the reverse primer designed based on the 3' UTR of CaMXMT1, three clones stretching from promoter to gene were isolated namely PG-1 (2452bp), PG-4 (2650bp) and PG-5 (2621bp). Clone CX8 and PG-4 represent pseudogenes and have premature stop codons in their open reading frames (ORFs). This was the first documentation of the genomic structure of caffeine biosynthetic NMTs in any plant system. The theobromine synthase-like gene has four exons and three introns. The intron/exon splice site is very conserved except in the clone CS2A and all of them belong to phase 0 based on the different positions on codon (0=intron insertion before first base of code, 1= intron insertion before second base of code and 2= intron insertion before third base of code). 27

21 The first exon was found to be relatively small (75bp). It was noted that the four exons shared a high similarity to each other and to NMTs isolated from C. arabica (>85%). The first two introns have considerable similarity among the C. canephora sequences but the third intron exhibited length polymorphism. Introns are no longer considered as completely junk DNA and have been assigned many roles in increasing the mrna accumulation (Nott et al., 2003) transcriptional regulation like binding to transcription factors, splicing factors for the efficient transcription and normal or alternate splicing (Simard and Cahbot, 2000) and interacting with the cap binding complex and spliceosomal complex leading to an integration of various RNA processing mechanisms to transcription (see review by Proudfoot et al., 2002) and ensuring the posttranscriptional movement of hnrna into the cytoplasm where they are translated. An analysis of the intron sequences using molecular evolutionary tools that compare the rate of evolution with that of the individual sites of the codons (Hoffman and Birney, 2007) and to the rates of insertion/deleletion (Chen et al., 2009) would be a first step to find if the nucleotide positions are mutating freely or are having constraints in mutation. Since the third nucleotide portion has more freedom to mutate due to the codon redundancy, a comparision of the observed rate of change in the third nucleotide of the codon to that of the intron would be an ideal preliminary experiment to find if the intron might have been evolving freely. Recently through sirna mediated post transcriptional gene silencing, the putative NMTs were confirmed to be involved in caffeine biosynthetic pathway. However, due to the conserved nature of the sirna designing, it was unclear which gene was targeted in the transgenic coffee (Mohanan et al., 2013a). It was also shown that the second intron contain precursor sequences for mirna that target other NMT like genes thus bringing in high level of complexity in the regulation of caffeine biosynthesis (Mohanan et al., 2013b). The coffee gneome sequence has revealed a total of 23 NMT sequences annoptated to be related to caffeine biosynthesis (Deonaud et al., 2014) Regulation of Caffeine Biosynthesis Due to complexity in the varied entry points in caffeine biosynthesis as well as in the degradation dynamics, regulation of caffeine biosynthesis has been an alluring topic till the present date. Differences in spatio-temporal gene expression as well as species to species variation (Campa et al., 2005) in the mechanism of caffeine biosynthetic NMT gene regulation have been noted in the literature. This has added to the complexity of the subject ofis. Hence the identification of transcription factors would revive the 28