Dispersal mode of species influences the trypsin inhibitor levels in fruits

Proteinase inhibitors (PI) occur ubiquitously across microorganisms, plants, and animals [1-3] and have been shown to inhibit "unwanted or foreign" proteolysis and to defend the host tissues against herbivore predation [4, 5]. In plants, PIs are found at very high levels in storage tissues such as tubers and fruits and are argued to serve as defense against herbivores [3, 5]. Recently, Pearce et al. [6] showed that PI accumulation in wild tomato followed a parabolic pattern with age of the fruit; the PI level was low at early (2-3 weeks after anthesis) and late (ripened) stages of fruit growth, but very high at peak fruit growth. They suggested that such a pattern of accumulation may help deter premature dispersal of seeds by dispersers. In this paper, we extend this idea and argue that the dispersal mode of a spedes may strongly influence both the level and the temporal pattern of PI accumulation in its fruit. Analyzing the trypsin inhibitors, a major class of PI in 47 plant species representing various dispersal modes, we offer a test of this hypothesis. Fruits of animal-dispersed compared to those of wind/water or passively (explosively) dispersed species are mostly modified into rich storage organs containing a relatively high level of soluble compounds. A characteristic feature of fruits is the temporal separation of the growth of the mesocarp/pericarp (fleshy part) and the embryo [7]. In most species the embryo initiates growth much later than that of the mesocarp. Such a pattern has important implications for the successful dispersal of fl'uit in animaldispersed species. If dispersers remove fruits when much of the mesocarp growth is complete, but the embryo is still immature (usually coinciding with the peak growth stage of the fruit), plants do not gain any dispersal advantage. Selection in these species should favor the preferential removal of fruits by dispersers at a time when the embryo is mature (usually at the ripened stage of fruit) to gain effective dispersal of their seeds. Accordingly, in species where fruits are dispersed by animals, selection can be expected to favor a high accumulation of PI at the peak growth stage of the fruit (to deter dispersers from removing premature fruits) and low or negligible accumulation at the ripened stage of fruit (to encourage dispersers to remove the fruits). In contrast to fleshy fruits, nonfleshy fruits dispersed passively or by wind, do not run the risk of being removed prematurely by dispersers because, unlike fleshy fruits, their pericarp is not generally modified into rich storage organs. Thus, in these species, selection should not favor the accumulation of PI in fruits. Further, for the same reason, though the temporal pattern of growth of pericarp and embryo in nonfleshy fruit is also separated as in the fleshy fruits, selection should not favor any temporal pattern of PI accumulation in fruits dispersed passively or by wind. Randomly collected fruits of species representing the three dispersal modes, animal, wind/water, and passive (Table 1), available at the Botanical Garden, University of Agricultural Sciences, Bangalore (12 ° 58'N, 77 ° 35'E), were assessed for a major class of PI, namely, trypsin inhibitors (TI). The TI levels were determined in the rues©or pericarp during the peak growth phase of the fruits, when the growth of the fruit tissue was nearly complete. To examine the temporal pattern of TI accumulation for two randomly selected species of each dispersal mode, the TI levels of the fruit tissue were estimated at three stages of growth, namely, immediately following fertilization; at the peak growth stage; and at the ripened stage. Five hundred mg of the fruit tissue (mesocarp or pericarp) from randomly collected fruits of each species was macerated in 2.5 ml of 0.01 M potassium phosphate buffer (pH 7) containing 1 M KC1, and centrifuged at 18000 g for 10 min at 4 °C. The supernatant was depigmented by gel filtering through a Sephadex G-25 column (1 x 10 cm), equilibrated with 0.01 M phosphate buffer containing 0.15 M NaC1 (pH 7.6); 1-ml fractions were eluted with the same buffer. The fractious containing proteins, as detected by the dye-binding method [8], were pooled and analyzed for the presence of trypsin inhibitors. The trypsin inhibitors in the samples were estimated according to [9]. The data were expressed as trypsin inhibitory units (TIU) per gram fresh weight as