Abscisic acid controls embryo growth potential and endosperm cap weakening during coffee (Coffea arabica cv. Rubi) seed germination
- First Online:
- Cite this article as:
- da Silva, E.A.A., Toorop, P.E., van Aelst, A.C. et al. Planta (2004) 220: 251. doi:10.1007/s00425-004-1344-0
- 463 Views
The mechanism and regulation of coffee seed germination were studied in Coffea arabica L. cv. Rubi. The coffee embryo grew inside the endosperm prior to radicle protrusion and abscisic acid (ABA) inhibited the increase in its pressure potential. There were two steps of endosperm cap weakening. An increase in cellulase activity coincided with the first step and an increase in endo-β-mannanase (EBM) activity with the second step. ABA inhibited the second step of endosperm cap weakening, presumably by inhibiting the activities of at least two EBM isoforms and/or, indirectly, by inhibiting the pressure force of the radicle. The increase in the activities of EBM and cellulase coincided with the decrease in the force required to puncture the endosperm and with the appearance of porosity in the cell walls as observed by low-temperature scanning electronic microscopy. Tissue printing showed that EBM activity was spatially regulated in the endosperm. Activity was initiated in the endosperm cap whereas later during germination it could also be detected in the remainder of the endosperm. Tissue printing revealed that ABA inhibited most of the EBM activity in the endosperm cap, but not in the remainder of the endosperm. ABA did not inhibit cellulase activity. There was a transient rise in ABA content in the embryo during imbibition, which was likely to be responsible for slow germination, suggesting that endogenous ABA also may control embryo growth potential and the second step of endosperm cap weakening during coffee seed germination.
KeywordsAbscisic acidCellulaseCoffeaCoffee seedEndo-β-mannanaseEndosperm weakening
Scanning electron microscopy
The coffee (Coffea arabica L.) embryo is enveloped by an endosperm tissue (Krug and Carvalho 1939; Mendes 1941). The fully differentiated embryo lies inside an embryo cavity, is 3–4 mm long, and is composed of an axis and two cotyledons (Rena and Maestri 1986). The endosperm is surrounded by the endocarp, which resembles a seed coat (Chin and Roberts 1980). The coffee endosperm is composed of a hard greenish tissue with polyhedral cells, is isodiametrically divided into a hard external endosperm and a soft internal endosperm (Dedecca 1957), and belongs to the nuclear type (Mendes 1941). The endosperm cells have very thick walls that are crossed by plasmodesmata (Dentan 1985). These cell walls are composed of cellulose and hemicellulose (Wolfrom and Patin 1964). The main hemicellulose is an insoluble β-(1→4) d-mannan with 2% of galactose present in the side chains (Wolfrom et al. 1961). The galactose units are also found in the arabinogalactans in the coffee seed (Wolfrom and Patin 1965). The coffee seed belongs to the group of seeds that have a relatively high amount of mannans (Wolfrom et al. 1961).
Seed germination sensu stricto is completed and radicle protrusion starts when the expansive force of the embryonic radicle exceeds the mechanical restraint of the surrounding tissues (Hilhorst et al. 1998). The possible causes of embryo growth are lowering of its osmotic potential (ψπ), thus raising the pressure potential (ψp) in the radicle cells; relaxation of the radicle cell walls; and weakening of the tissues surrounding the embryo, or a combination of these causes (Bewley and Black 1994).
In celery the mature embryo grows inside the endosperm before radicle protrusion (Jacobsen and Pressman 1979; van der Toorn and Karssen 1992). In lettuce seed, Takeba (1980) found an accumulation of free amino acids in the embryonic axes that would decrease ψπ (more negative) and hence increase its growth potential to allow radicle protrusion. In Brassica napus embryos it was demonstrated that an increase of both turgor and cell wall extensibility were required for radicle protrusion (Schopfer and Plachy 1985).
Weakening of the tissues opposing the radicle tip has been proposed as a prerequisite for radicle protrusion in tomato (Haigh and Barlow 1987; Groot and Karssen 1987), muskmelon (Welbaum et al. 1995) and Datura ferox (de Miguel and Sánchez 1992). In tomato and muskmelon seeds, water uptake by the embryo is restricted by the endosperm during germination and lowering of the osmotic potential or an increase in embryo turgor have never been observed prior to radicle protrusion (Haigh and Barlow 1987; Welbaum and Bradford 1990). In Datura ferox the increase in embryo growth potential was insufficient to allow germination (de Miguel and Sánchez 1992). In tomato seed, endo-β-mannanase (EBM; EC 188.8.131.52) activity correlated with weakening of the endosperm cap (Groot et al. 1988; Toorop et al. 2000). EBM activity also correlated with porosity in the endosperm cap cell walls, as observed by cryo-scanning electron microscopy (cryo-SEM) and with a decrease in the force required to puncture the endosperm cap (Toorop et al. 2000). Other hydrolytic enzymes, such as polygalacturonase (Sitrit et al. 1999), cellulase (Leviatov et al. 1995) and arabinosidase (Bradford et al. 2000) have also been shown to increase in activity during tomato seed germination. Also, in muskmelon seed, cellular degradation and weakening occurred concomitantly with the decrease in puncture force (Welbaum et al. 1995). In Datura spp., scanning electron micrographs and analysis of endosperm cell wall polysaccharide composition showed morphological and compositional changes in the micropylar endosperm cell walls, prior to radicle protrusion (Sánchez et al. 1990). In pepper seeds the endosperm cap displayed compressed cells and loss of integrity before radicle protrusion, as well as a decrease in the required puncture force (Watkins et al. 1985). However, EBM activity was only detected after radicle protrusion (Watkins et al. 1985).
Cell wall hydrolytic enzymes have previously been studied in coffee seed. These include α-galactosidase (EC 184.108.40.206; Shadaksharaswamy and Ramachandra 1968), cellulase (EC 220.127.116.11; Takaki and Dietrich 1980; Giorgini 1992) and EBM (Giorgini and Comoli 1996; Marraccini et al. 2001). However, there is little information about enzyme activity in relation to the germination mechanism and its regulation.
Abscisic acid (ABA) is known to induce dormancy and inhibit seed germination (Bewley and Black 1994). In lettuce seed, endogenous ABA inhibited both EBM activity (Dulson et al. 1988) and cellulase activity (Bewley 1997). In fenugreek and carob seeds ABA suppressed the activity of EBM in the endosperm (Kontos et al. 1996). In tobacco, β-1,3-glucanase (EC 18.104.22.168) correlated with endosperm rupture and ABA delayed this rupture (Leubner-Metzger et al. 1995). In the endosperm cap of tomato, ABA did not inhibit cellulase (Toorop 1998; Bradford et al. 2000) or EBM activity (Toorop et al. 1996; Still and Bradford 1997) but radicle protrusion was prevented. Schopfer and Plachy (1985) have shown that ABA inhibited cell wall loosening in the embryo of Brassica napus. Valio (1976) found that endogenous ABA-like substances and exogenous ABA caused inhibition of coffee seed germination through inhibition of embryo growth. However, the role of ABA during coffee seed germination has not been described in clear detail. The aim of the present work is to determine the targets and mechanism of the ABA-controlled inhibition of coffee seed germination.
Materials and methods
Coffee (Coffea arabica L., cv. Rubi) seeds were harvested in 1998 in Lavras-MG-Brazil, depulped mechanically, fermented and dried to 12% moisture content (fresh-weight basis), and shipped to The Netherlands where they were stored at 10°C. Germination tests executed immediately before shipping and upon arrival showed that during the 48-h transportation period no appreciable damage to the seeds had occurred.
The seed coat was removed by hand and seeds were surface-sterilised in 1% sodium hypochlorite for 2 min. Seeds were then rinsed in water and imbibed in demineralised water or ABA solution (racemic mixture; Sigma, St. Louis, MO, USA). Seeds were placed in 94-mm Petri dishes on filter paper (No. 860; Schleicher & Schuell, Dassel, Germany) in 10 ml of the imbibing solution. During imbibition, seeds were kept at 30±1°C in the dark (Valio 1976). The germination percentage was recorded daily.
ABA solutions were prepared by dissolving the compound in 1 N KOH followed by neutralisation with 1 N HCl. Fluridone solution was prepared by dissolving the compound in 0.1% acetone until complete dissolution. Control experiments showed that this acetone concentration did not affect germination.
Intact seeds were imbibed as described above and the fresh weight of individual seeds was measured daily.
Twenty embryos from water-imbibed seeds were isolated by cutting the endosperm with a razor blade. Embryo length was measured by using calipers. After length measurement the embryos were separated into embryonic axes and cotyledons and these were measured again.
Water and osmotic potential measurements
The water potential (ψ) and osmotic potential (ψπ) of coffee embryos from seeds imbibed in water or ABA solutions were measured by using a calibrated thermocouple psychrometer (Model HR-33T; Wescor, Logan, UT, USA) with a C-52 sample chamber (Wescor). Samples were equilibrated for 40 min and two readings were taken before starting the experiments to ensure that equilibrium had been attained. Cooling time was 45 s. The C-52 chamber was placed in an airtight glove box kept at 100% relative humidity by a stream of water-saturated air at a constant temperature of 25±1°C. Embryos were isolated as described above and placed in the C-52 chamber for measurements. Three replications of five embryos were used for the measurements. After measurement of the water potential the embryos were put in liquid nitrogen for determination of the ψπ. After 2 h in liquid nitrogen the embryos were left to thaw and ψπ was determined. The pressure potential (ψp) was calculated from the equation: ψp=ψ−ψπ.
Puncture force measurement
The force required to puncture individual endosperm caps was measured as described by Toorop et al. (2000). An S 100 material tester (Overload Dynamics, Schiedam, The Netherlands) was used with a JP 50 load cell (Data Instruments, Lexington, MA, USA) at a range of up to 10 lbs. A probe with a hemispherical tip and a diameter of 0.7 mm was placed on the load cell. The endosperm cap was cut from the seed and the embryo removed. The endosperm cap was placed on the probe and perforated by moving the probe down into a polyvinyl chloride block with a conical hole with a minimum diameter of 1 mm. The force required to puncture the endosperm cap was used as a parameter for the mechanical restraint of this tissue. All data points represent the average of 30 measurements.
Diffusion assay for EBM activity
Extracts from 30 endosperm caps were made from water- and ABA-imbibed seeds. EBM was extracted in McIlvaine buffer (0.05 M citric acid/0.1 M Na2HPO4, pH 5.0) with 0.5 M NaCl, and assayed in an activity gel (0.5 mm thick) containing 0.5% (w/v) locust bean gum (Sigma) in McIlvaine buffer (pH 5.0) and 0.8% type III-A agarose (Sigma) on gelbond film (Pharmacia). 2-μl aliquots of the extracts were applied to holes that were punched in the gel with a 2-mm paper punch. Gels were incubated for 21 h at 25°C, and then washed in McIlvaine buffer (pH 5.0) for 30 min, stained with 0.5% (w/v) Congo Red (Sigma) for 30 min, washed with 80% (v/v) ethanol for 10 min, and de-stained in 1 M NaCl for 5 h. All staining steps were performed on a rotating platform. Commercial EBM from Aspergillus niger (Megazyme, North Rocks, Sydney, Australia) was used to generate a standard curve. The gel was scanned and printed out for calculation of enzyme activity in the samples. Calculations were done according to Downie et al. (1994).
Ten imbibing seeds were rinsed in deionized water and cut in half with a razor blade. Seed parts were blotted dry on filter paper, and laid on top of an activity gel, cut side down. Activity gels were incubated for 1 h at room temperature. Seed parts were then removed from the gel with tweezers. After incubation the gels were stained with Congo Red, as described above.
Isoelectric focussing (IEF)
Fifty endosperm caps and 15 endosperm remainders from 8-day-imbibed seeds were extracted in 0.1 M Hepes/0.5 M NaCl buffer (pH 8.0). EBM activity was assayed by using the gel diffusion assay as described above and the same protein amount for all treatments was applied to the IEF gel. The extracts were desalted using a Microcon-50 concentrator (Amicon, Beverly, MA, USA) and 5 μl (water-imbibed) or 10 μl (ABA-imbibed) of the extracts from the endosperm cap, and 20 μl of extracts from the endosperm remainders were applied to an IEF gel with a pH range of 3–10. The gels were run for 1 h and 20 min at 2,000 V. The gels were stained and the isoelectric points determined according to Toorop et al. (1996).
Diffusion assay for cellulase activity
Gels (0.75 mm thick) were used containing 0.15% (w/v) carboxymethylcellulose (BDH, Poole, UK), 0.05 M acetate/0.1M NaCl buffer (pH 5.5) and 0.8% type III-A agarose (Sigma) on gelbond film (Pharmacia). Extracts of endosperm caps were made prior to radicle protrusion, according to Giorgini (1992). Samples of 4 μl were applied to holes that were punched in the gel with a 2-mm paper punch. Gels were incubated for 22 h at 25°C, and then washed in acetate buffer (pH 5.5) for 30 min, stained with 0.5% (w/v) Congo Red (Sigma) for 30 min, washed with 96% (v/v) ethanol for 10 min, de-stained in 1 M NaCl for 24 h and washed again in acetate buffer (pH 5.5). All staining steps were performed on a rotating platform. Commercial cellulase (Serva) was used to generate a standard curve. Calculation of enzyme activity in the samples was similar to the calculation of EBM activity.
Cryo-scanning electron microscopy
Water- and ABA-imbibed seeds were longitudinally sectioned with a razor blade and mounted on a cup-shaped holder with tissue-freezing medium. After mounting, the samples were plunge-frozen and stored in liquid nitrogen for subsequent cryo-planing and observations. Cryo-planing, which attempts to produce flat surfaces for observation by cryo-SEM, was performed using a cryo-ultramicrotome with a diamond knife, according to Nijsse and van Aelst (1999). The specimens were sputter-coated with platinum and placed in the cryostat of a Field emission scanning electron microscope (JEOL 6300). Observations were made at −180°C using an accelerating voltage of 2.5–5 kV. Digital images were taken and printed.
ABA extraction and quantification
ABA was extracted from embryos that were isolated during and following germination, from water-imbibed seeds. Three replications, containing ten embryos each, were used for ABA extraction and quantification. The embryos were stored at −80°C. The embryos were lyophilised and ground to a powder in liquid nitrogen. ABA was extracted according to Raikhel et al. (1987) and Berry and Bewley (1992). ABA quantification was carried out by using an ABA immunoassay detection kit (Sigma).
Statistical analysis was performed with the general linear model (SPSS 10.0.5).
Water relations of the embryo
In ABA the embryo water potential increased from −4.31 MPa to −1.53 MPa at 5 days of imbibition and the osmotic potential from −4.50 MPa to −1.85 MPa at 5 days of imbibition. At 6 days of imbibition there was a decrease in water potential from −1.53 MPa to −3.63 MPa. The osmotic potential also decreased, from −1.85 MPa to −3.84 MPa. No change in pressure potential in ABA-imbibed seeds was observed (Fig. 4b). Values were always slightly above zero.
Puncture force measurements
Endosperm structure during germination
Localisation of EBM activity
Little work has been done with the aim of understanding the behaviour of coffee seeds during germination, and there is a lack of information concerning the regulation of their germination. Under optimal conditions, coffee seed germination is still relatively slow and asynchronous (Fig. 1). Under field conditions, coffee seeds germinate even slower (Rena and Maestri 1986). Seedling emergence from the soil starts after 50–60 days in the warmer periods of the year (Maestri and Vieira 1961). Under natural conditions the dispersal unit is a berry and germination may even be slower because of the fruit tissues surrounding the seeds. The water uptake during imbibition followed the common triphasic pattern as described by Bewley and Black (1994). Phase II of germination was not attained until day 3 of imbibition. The coffee endosperm has very thick cell walls that may slow down water uptake.
Although no detectable change in water content occurred during phase II, the pressure potential in the embryo increased prior to radicle protrusion. However, the osmotic potential in the embryo did not decrease (become more negative) during the first 5 days of imbibition. The coffee embryo lies inside a cavity (Maestri and Vieira 1961) where there is sufficient space for expansion. Apparently, a gradient in water potential between endosperm and embryo allowed a continuous influx of water into the embryonic cells and the build-up of a pressure potential. This situation is different from that observed in tomato (Haigh and Barlow 1987) and muskmelon seeds (Welbaum and Bradford 1990), which did not display an increase in pressure potential prior to radicle protrusion. After 5 days of imbibition the pressure potential was released, indicating cell wall relaxation and, obviously, a new gradient in water potential was generated, allowing more influx of water into the embryo cells and, consequently, embryo growth. This coincided with the overall decrease in the required puncture force and an increase of EBM and cellulase activities and cell wall porosity in the endosperm cap. We hypothesise that the decrease in osmotic potential after day 5 of imbibition is caused by the accumulation of solutes in the embryo, originating from endosperm cell wall degradation during imbibition. Evidently, this process can only occur if the extensibility of the embryonic cell walls increases.
Direct measurement of embryo length confirmed that the embryo grew inside the endosperm before radicle protrusion and that both axis and cotyledons contributed to the total embryo growth to a similar extent, although the cotyledons only grew during the first 5 days of imbibition. Therefore, pressure potential and cell wall extensibility are responsible for embryo growth inside the coffee endosperm prior to radicle protrusion. In Brassica napus, cell wall loosening and decline in turgor in the embryo control germination (Schopfer and Plachy 1985), although in this species the seeds do not possess a thick-walled endosperm. Embryo growth before radicle protrusion is common in seeds with an immature embryo, e.g. celery (Jacobsen and Pressman 1979; van der Toorn and Karssen 1992). However, the mature coffee seed contains a fully differentiated mature embryo (Mendes 1941) but still displays growth before completion of germination.
Valio (1976) proposed that ABA inhibits germination of the coffee seed through inhibition of embryo growth. Here we showed that ABA inhibited the transient increase in turgor of the embryo that occurred around day 5 in the water control. Cessation of growth of the embryonic axis in ABA-imbibed seeds was also observed at day 5 of imbibition (Fig. 3b). The second transient rise in endogenous ABA content in the embryos, around day 5, coincided with the cessation of growth, whereas the lowering in ABA content to near zero values at day 8 of imbibition coincided with radicle protrusion (Fig. 5). These results show that ABA is synthesized de novo in the embryo during coffee seed imbibition and is degraded or leached out thereafter. Fluridone, an inhibitor of ABA biosynthesis (Li and Walton 1990), significantly advanced radicle protrusion (Fig. 1). Therefore, ABA biosynthesis during imbibition may contribute to the rate of radicle protrusion in coffee seeds. ABA synthesis during seed germination has been demonstrated in imbibing seeds of Nicotiana plumbaginifolia (Grappin et al. 2000). In those seeds, fluridone also shortened the lag time for germination. Schopfer and Plachy (1985) proposed that ABA inhibited embryo expansion in Brassica napus by inhibiting cell wall loosening and, thus, limiting the growth potential of the embryo. Our results indicate that this may be the case in coffee. Hilhorst (1995) suggested that ABA might suppress the enzymatic hydrolysis of load-bearing bonds in the embryonic cell walls, thereby inhibiting tissue elongation.
Concomitant with embryo growth inside the endosperm, the required puncture force showed a significant decrease before radicle protrusion. Tissue printing confirmed the presence of EBM in the endosperm, which showed activity first in the endosperm cap and only later in the remainder of the endosperm. The decrease in puncture force followed a biphasic pattern with the first phase occurring between days 0 and 4 and the second between days 6 and 9. A similar biphasic pattern was found for the increase in both EBM and cellulase activities, resulting in a high negative linear correlation between required puncture force and enzyme activities (r2=0.86 for EBM and r2=0.83 for cellulase). Endosperm weakening prior to radicle protrusion has also been demonstrated to occur in muskmelon (Welbaum et al. 1995), Capsicum annuum (Watkins et al. 1985) and Datura ferox (de Miguel and Sánchez 1992), which also coincided with the occurrence of enzyme activity in the endosperm. However, we cannot exclude the possibility that embryo growth during coffee seed germination also contributes to the second phase of the endosperm cap weakening.
Cryo-SEM studies showed that the endosperm cap cells were compressed and lost their integrity before radicle protrusion. Evidently, growth of the embryo inside the endosperm caused the occurrence of the protuberance, as well as the compression of cells in the endosperm cap and loss of cell integrity. Cryo-SEM also showed porosity in the endosperm cap and in the remainder of the endosperm before radicle protrusion. These pores are caused by evaporation of water during the freeze-drying process and indicate absence of cell wall material, presumably as a result of hydrolytic enzyme activity (Toorop et al. 2000). There was a progressive increase in porosity before radicle protrusion in the endosperm cap and in the remainder of the endosperm. The same trend in porosity, albeit at lower levels, was observed in ABA-imbibed seeds. The porosity in the endosperm cap coincided with the decrease in required puncture force, an increase in cellulase and EBM activities, and with the occurrence of specific EBM isoforms in the endosperm cap and in the remainder of endosperm. In tomato seeds the development of porosity in the endosperm cap coincided with the increase in EBM activity and the overall decrease in required puncture force (Toorop et al. 2000). Also, in Datura spp. eroded cell walls were present in the micropylar endosperm before radicle protrusion (Sánchez et al. 1990). Moreover, the cell walls in the coffee endosperm cap are thinner than the cell walls in the remainder of the endosperm. The same structural difference has been described in the endosperm cap of tomato (Hilhorst et al. 1998), muskmelon (Welbaum et al. 1995) and in Datura species (Sánchez et al. 1990).
ABA reduced EBM activity to approximately 10% or less of the water control throughout the germination process but inhibited only the second step of endosperm cap weakening. The remaining EBM activity could not be attributed to enzyme present in tissue remainders from the rest of the endosperm attached to the cap tissue. Dissection of the cap tissue could be done with high precision since the protuberance (Fig. 7a) indicated its exact location. Isoelectric focussing showed that there were three different isoforms of EBM in the endosperm cap and that ABA inhibited two of them (pI 4.5 and pI 6.5). This suggests that there are EBM isoforms that are controlled by ABA and isoforms that are not. In the present case, the pI-4.5 and pI-6.5 isoforms do not appear to contribute to endosperm weakening during the first phase but the pI-7.0 isoform may still be involved. Apparently, the occasional presence of EBM activity in the cap region, as shown by tissue printing, must be attributed to this isoform. Since the second phase of endosperm cap weakening is under control of ABA, the pI-4.5 and pI-6.5 isoforms may be essential for this phase of the germination process.
Of the four different isoforms of EBM found in the remainder of the endosperm, ABA inhibited one isoform (pI 4.5). However, since this isoform always occurred as a small spot of relatively low intensity, we cannot exclude the possibility that it originates from the endosperm cap. The isoforms of pI 5.5, 6.5 and 7.0 were unaffected by ABA. Marraccini et al. (2001) observed more EBM isoforms in germinating coffee seeds. This difference may be due to the fact that in the present study we used seeds prior to radicle protrusion, whereas these authors used 28-day-imbibed seeds, i.e. well after radicle protrusion.
These results suggest that the different isoforms of EBM may have different functions during coffee seed germination. Tomato seeds also display different isoforms of EBM during germination (Toorop et al. 1996; Nonogaki et al. 1998). Two EBM genes were cloned from tomato, of which one, LeMan2, was specifically expressed in the endosperm cap before radicle protrusion, whereas the other gene, LeMan1, was expressed in the lateral endosperm after radicle protrusion (Nonogaki et al. 2000).
The increase in cellulase activity coincided with the first phase of decrease in puncture force. ABA neither inhibited endosperm weakening nor cellulase activity throughout the germination process. This suggests that cellulase may be decisive in the first phase of endosperm weakening but not in the second. The presence of cellulase has previously been demonstrated in coffee seeds (Takaki and Dietrich 1980; Giorgini 1992). Tissue printing demonstrated that cellulase activity was present throughout the endosperm (including the cap) during imbibition and no differences were observed with and without ABA. Also, ABA did not inhibit cellulase activity in tomato seeds (Toorop 1998). A cDNA having high homology with known β-1,4 glucanases was isolated from radicle and endosperm cap tissue of tomato seeds prior to radicle protrusion, and ABA had no effect on its expression (Bradford et al. 2000). However, class-I β-1,3-glucanase (βGlu1) activity is under ABA control in germinating tobacco and other solanaceous seeds, and is involved in endosperm rupture, preceding radicle protrusion (Leubner-Metzger 2003; Petruzzelli et al. 2003).
Tomato seeds also show a biphasic endosperm cap weakening (Toorop et al. 2000). During the first phase the decrease in required puncture force correlated with an increase in EBM activity and the occurrence of ice crystal-induced porosity in the endosperm cells, but during the second phase the EBM activity and required puncture force were uncoupled in ABA-imbibed seeds in that EBM activity was similar to control levels but endosperm weakening was inhibited. Thus, coffee seeds show a similar behaviour in endosperm weakening but a dissimilar pattern of EBM activity as compared with tomato seeds (Toorop et al. 2000).
The combination of increased porosity of endosperm cap cell walls and the increase in turgor of the radicle opens the possibility of a second endosperm weakening mechanism, apart from enzymatic hydrolysis of cell walls. This second mechanism may be operational during, or may represent, the second phase of endosperm weakening. It is feasible that the increasing pressure force or ‘thrust’ of the radicle causes mechanical disruption of the already weakened endosperm cell walls. The occurrence of the protuberance after day 5 of imbibition coincided with the beginning of the second phase of endosperm weakening. The inhibitory action of ABA may then be explained by its suppression of the generation of the embryonic pressure potential, as demonstrated in the present study.
We hypothesise that in coffee seeds ABA controls both the embryo growth potential during germination and the second step of endosperm cap weakening by inhibiting two isoforms of EBM.
The seed laboratory at Lavras Federal University-MG-Brazil (UFLA) is acknowledged for handling and shipping the coffee seeds to The Netherlands. We are grateful to Richard Bourgault of the Department of Botany, University of Guelph, Ontario, Canada for his help with the IEF gels. We also thank Ms Katja Grolle at the Department of Food Science for her technical advice on the material tester. Plant Research International-The Netherlands is acknowledged for the use of the psychrometer. We thank CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nivel Superior-Brazil), for financial support of the studies of E.A. Amaral da Silva (project number: 1485/96-2).