TREATMENT OF ISOLATED PISTILS WITH PROTEASE INHIBITORS OVERCOMES THE SELF-INCOMPATIBILITY RESPONSE IN BUCKWHEAT

Isolated pistils of distylous buckwheat (Fagopyrum esculentum Moench) were treated with protease inhibitors (PMSF, pepstatin A, and antipain). Pistils were crossor selfpollinated, and growth of pollen tubes was observed under a fluorescence microscope. Treatments with all inhibitors suppressed inhibition of self-pollen tube growth, suggesting that activity of proteases is involved in rejection of self-pollen during the SI response.


INTRODUCTION
Self-incompatibility (SI) is an outbreeding mechanism which enables plants to discriminate between self and non-self pollen grains.It is broadly distributed among the flowering plants: more than half of all flowering plant species display some of the SI types.There are two main well-defined types of SI, gametophytic and sporophytic, and the latter can be homomorphic or heteromorphic (for recent reviews, see D i x i t and N a s r a l l a h, 2001; Hiscockand Mc I n n i s, 2003; and T a k a y a m a and I s o g a i, 2005).While SI is considered advantageous from the evolutionary point of view, it nevertheless represents a drawback for plant breeders, since it prevents the creation and selection of homozygous lines.The heteromorphic SI system is characterized by different positions of stigmas and anthers in dimorphic or trimorphic flowers (de N e t t a n c o u r t, 1997).It is rare and occurs in 24 families of flowering plants.Among plants of this type, common buckwheat ( Fagopyrum esculentum Moench) is of the greatest economical importance.It posseses two flower morphs: pin (with long pistil and short anthers) and thrum (with short pistil and long anthers), which are equally distributed among the population.Legitimate pollination is possible only between these two types.In contrast to homomorphic systems, very little is known about the molecular mechanism of the SI response in heteromorphic systems (de Nettancourt, 1997) operating in buckwheat.
Our previous papers on SI in buckwheat (M i lj u š -Dj u k i c et al., 1998, 2003) showed that processes such as protein synthesis and glycosylation, together with protein phosphorylation, can be part of the SI response, such processes being common in plants which display one of the two SI systems(sporophytic in Brassicaceae, gametophytic inSolanaceae).On the other hand, the SI response in Papaveraceae belongs to quite a different system, where Ca 2+ serves as an important SI response mediator (F r a n k l i n -T o n g et al., 2002).
In the present paper, we describe effects of protease inhibitors on the self-incompatibility response in order to further "dissect" the SI response in buckwheat and analyze the possible role of proteases in this complex process.It could be predicted that proteases are involved in programmed cell death (PCD).Processes of PCD processes are very important for sexual plant reproduction, even when hidden from view (W u and C h e u n g, 2000).Cell death occurs in both pollen and the pistil.For example, in plants with a solid style, the transmitting tis-sue must pass through degeneration in order to let pollen tubes grow (W a n g et al., 1996).
Isolated buckwheat pistils, cross and self-pollinated, were treated with three different protease inhibitors: the serine protease inhibitor PMSF, the aspartic protease inhibitor pepstatin A, and the cysteine protease inhibitor antipain.

Plant material
Buckwheat plants (Fagopyrum esculentum Moench, cv.Darja) were grown in a greenhouse.Field-collected seeds gave rise to thrum plants (with short pistils) and pin plants (with pistils about three times longer).One day prior to experiments, all opened flowers were removed.Freshly opened flowers were collected the next morning and pistils were isolated under sterile conditions to minimize possible bacterial contamination.About 10 pistils of each morphwere put in a Petri dish containing the germination medium (B r e w b a c k e r and K w a c k, 1963).The medium consisted of 10% agar, 15% sucrose, and (in mg L -1 ): 100 H 3 BO 3, 300 Ca(NO 3 ) 2 .4H 2 O, 200 MgSO 4 .7H 2 O, and 100 KNO 3 .The pistils were pollinated with self and non-self pollen by touching their surfaces with dehisced anthers.Buckwheat pistils consist of three closely adhering styles, and about 10-20 pollen grains were visible on each stigma upon pollination.

Experiments with inhibitors
Inhibitors were added by wetting a piece of filter paper, which was put on the agar surface under the pistils.In the controls, this filter paper was wetted with distilled water.Care was taken to prevent any contact of stigma surface with inhibitors.Incubation lasted 24 h.The serine protease inhibitor PMSF (phenylmethylsulfonylfluoride), aspartic protease inhibitor pepstatin A, and cysteine protease inhibitor antipain were added in the following concentrations: 1 and 10 mM PMSF; and 1 and 10 μM pepstatin or antipain.All experiments were repeated three times, with 10 pistils in each sample.

Microscopy
After incubation, the pistils were fixed overnight in ethanol:acetic acid (3:1, v:v), then washed in distilled water, macerated in 7 N NaOH for 24 h,and washed again.Finally, they were stained with 0.1% decolorized aniline blue in 0.1 N K 3 PO 4 and mounted on a microscope slide in a drop of 50% glycerol.Pollen tubes were then observed in UV light using a Leitz ARISTOPLAN fluorescence microscope (12 V 100 W halogen lamp) and an excitation block filter (I3-blue, Leica).

RESULTS AND DISCUSSION
We examined the effects of different protease inhibitors to establish possible protease involvement in the SI response of buckwheat.
In isolated control pistils of the thrum morph, selfpollen tubes were arrested at the junction between stigmatic tissue and the style (Fig. 1E).Treatment of isolated pistils with PMSF resulted in overcoming of the SI response at both applied concentrations (Fig. 1F), and self-pollen tubes elongated down the style as in the compatible control.The effect of pepstatin A (Fig. 1G) was similar to that of PMSF, i.e., at both concentrations (1 and 10 μM), the SI response was abolished and self-pollen tubes elongated down the style.The same results were obtained after treatment of isolated pistils with both applied antipain concentrations (Fig. 1H).
In isolated control pistils of the pin morph, selfpollen tubes were arrested at two thirds of the style's length (Fig. 1A).At both concentrations (1 and 10 mM), PMSF produced a breakdown of the SI response, and self-pollen tubes elongated beyond two thirds of the style's length, as in the compatible control (Fig. 1B).Treatment with pepstatin A and antipain abolished the SI response and self-pollen tubes reached the bottom of the style, as in compatible crosses (Figs.1C and 1D).
In experiments with protease inhibitors in buckwheat, the SI response was overcome and self-pollen tubes elongated down the style with all three inhibitors, while compatible crosses were not affected (Table 1).This finding indicates that proteases could be important players in the SI system operating in buckwheat.The SI response is complex: besides proteins at the stigma surface directly involved in self-pollen discrimination (thrum morph), there must also be others involved in processes leading to interruption of pollen tube elongation (pin morph).
It could be expected that proteases induced PCD during the SI response to prevent self-pollen tube elongation in buckwheat.There are recent data indicating a role for proteases in programmed cell death occuring in the presence of incompatible pollen tubes during the SI response in other plants.Aspartic proteases were found in reproductive tissues, including pistils, of many plants, although their exact roles are not clear (Wuand Cheung, 2000).This was shown for cardosin, an aspartic proteinase from cardoon that is involved in adhesion-mediated proteolytic mechanisms in pollen recognition and growth (F a r o et al., 1999).The question which remains to be answered is what mechanism activates proteases in the buckwheat two morphs, where the site of self-pollen tube arrest differs.
In poppy, a caspase inhibitor appears to reduce the level of DNA fragmentation (a marker for PCD) that occurs in pollen tubes elongating in the presence of incompatible S-proteins (T h o m a s and F r a n k l i n -T o n g, 2004).The authors showed that Ca 2+ acts as a second messenger in the SI response, and that increase of intracellular Ca 2+ concentration can be linked with cytochrome c leakage, the classical marker for PCD.To find out if the situation in buckwheat is similar, the phenomenon of PCD during the SI response in buckwheat should be futher examined in the light of our previous results with Ca 2+ antagonists (M i lj u š -Dj u k i ć et al., 2003).