Download 05 October 2000

05 October 2000
Nature 407, 649 - 651 (2000) © Macmillan Publishers Ltd.
<>
S-RNase uptake by compatible pollen tubes in
gametophytic self-incompatibility
DOAN-TRUNG LUU, XIKE QIN, DAVID MORSE & MARIO CAPPADOCIA
Biology Department, University of Montreal, Montreal , Quebec H1X 2B2, Canada
Correspondence and requests for materials should be addressed to M.C. (e-mail: [email protected]).
Many flowering plants avoid inbreeding through a genetic mechanism termed selfincompatibility. An extremely polymorphic S-locus1 controls the gametophytic selfincompatibility system that causes pollen rejection (that is, active arrest of pollen tube
growth inside the style) when an S-allele carried by haploid pollen matches one of the
S-alleles present in the diploid style. The only known product of the S-locus is an SRNase expressed in the mature style2. The pollen component to this cell–cell
recognition system is unknown and current models3, 4 propose that it either acts as a
gatekeeper allowing only its cognate S-RNase to enter the pollen tube, or as an inhibitor
of non-cognate S-RNases. In the latter case, all S-RNases are presumed to enter pollen
tubes; thus, the two models make diametrically opposed predictions concerning the
entry of S-RNases into compatible pollen. Here we use immunocytochemical labelling
of pollen tubes growing in styles to show accumulation of an S-RNase in the cytoplasm
of all pollen-tube haplotypes, thus providing experimental support for the inhibitor
model.
We have used a monospecific polyclonal antibody directed against a specific S-RNase
to label pollen tubes growing in incompatible styles. S-RNases contain an RNase
activity domain required for pollen rejection5, 6 and a hypervariable region involved in
allele-specific recognition7, 8. The 15-amino-acid peptide used to prepare the antibody
corresponded to the hypervariable region of the S11-RNase, and the antibody recognizes
the S11-RNase but does not recognize the S13-RNase, which differs from the former by
only 3 amino acids in this region8. In self-pollinated styles of plant V22 (genotype
S11S13; expressing both S11 and S13 RNases), the arrest of pollen-tube growth occurs
between the upper third and upper half of the style and is complete by 48 hours postpollination. Styles collected 18 hours post-pollination (Fig. 1a), which contain stillgrowing pollen tubes, were thus selected for the immunolabelling. Light microscopy
was used to identify the stylar regions containing the apical tips of pollen tubes, where
elongation occurs, and transverse sections from these regions were then stained for
transmission electron microscopy (TEM) with the anti-S11 and gold-labelled secondary
antibodies. Pollen tubes appear as small, irregularly shaped cells amid the larger,
rounder cells of the transmitting tissue, and their identity was determined by comparing
the morphology of unpollinated and pollinated styles9, 10 and by in situ hybridization
with a pollen-specific gene probe (data not shown). In these samples (not stained with
osmium, in order to preserve their antigenicity), apical regions of pollen tubes contain
darker-staining cytoplasm at their peripheral regions surrounding either a paler electronlucent vacuole or numerous electron-lucent secretory vesicles, depending on the
position of the section relative to the pollen tip11.
Figure 1 S11-RNase entry into incompatible pollen tubes.
Full legend
High resolution image and legend (47k)
Pollen-tube labelling (Fig. 1b and c) occurs principally in the cytoplasm. More distal
regions of pollen tubes (not shown) appear collapsed, devoid of cytoplasm, and are not
labelled. In contrast to the pollen tubes, cells of the transmitting tissue generally show
labelling over the lighter-staining regions. The presence of S-RNases inside membranebounded compartments is expected for a secreted protein and is consistent with the label
found in the extracellular matrix ( Fig. 1b and c). For the experiments shown in Fig. 1,
label density in the extracellular matrix (6 2 gold particles per µm2) is twice that
found inside the transmitting tissue cells (3 0.6 gold particles per µm 2) and
substantially less than that inside the pollen tube cytoplasm (32 10 gold particles
per µm2). All label densities are greater than the background (0.5 0.3 gold particles
per µm2), measured either in the absence of the primary antibody (not shown) or in
stylar epidermal tissues stained with the anti-S11 antibody (Fig. 1d). Epidermal tissues
are known not to express S-RNases12, 13, and are not expected to cross-react with the
antibody.
To confirm genotype-independent S-RNase uptake by pollen tubes, we labelled sections
from V22 styles following a fully compatible cross. Pollen was obtained from the Shomozygous plant VF60, so that all have an identical S12-allele constitution. All pollen
tubes are stained with the anti-S11 antibody, and the pattern of label accumulation inside
the cytoplasm ( Fig. 2a) is indistinguishable from that found in fully incompatible
crosses (Fig. 1). The label density inside S 12-pollen tubes (31 15 gold particles
per µm 2) is higher than that found in the extracellular matrix (3.8 0.8 gold particles
per µm2) or inside the transmitting tissue cells (2.6 0.8 gold particles per µm2 ). Thus,
the S11-RNase enters S12 pollen tubes. In another experiment, styles of V22 were
pollinated with S13 and S14 pollen (a semi-compatible cross). Once again, the S11 -RNase
was found in the cytoplasm of all pollen tubes ( Fig. 2b) even though none of the pollen
was S11. Despite a lower overall intensity of label, the label density in pollen cytoplasm
(12 4 gold particles per µm2) was again fivefold higher than that found in the
extracellular matrix (2.2 0.9 gold particles per µm2) and tenfold higher than the
transmitting tissue cells (1.4 0.7 gold particles per µm 2). As a last control for the
specificity of the reaction, sections containing compatible S11 and S13 pollen tubes
growing in S12S14 styles were stained with the anti-S11 antibody (Fig. 2c). As expected,
given the absence of an S11-RNase, only background label density was measured (0.6
0.3 gold particles per µm2). We note that the lack of label in the pollen tube cytoplasm
indicates that the antibody does not cross-react with a pollen-specific antigen. Taken
together, these results show an active uptake or accumulation of the S11-RNase in pollen
tube cytoplasm. We also note that, as all pollen tubes are similarly labelled, S-RNase
accumulation in pollen tubes is genotype-independent. All these observations are
inconsistent with a gatekeeper model requiring specific entry of an S-RNase only into
pollen tubes carrying the corresponding S-allele.
Figure 2 S11-RNase entry into compatible pollen tubes.
Full legend
High resolution image and legend (15k)
To place in perspective our findings showing S-RNase uptake in both compatible and
incompatible pollen tubes growing in vivo, we note that genotype-independent
accumulation of S-RNases in pollen tubes germinated in vitro in solutions of purified SRNases has been previously reported14. However, in these studies, both pollen-tube
RNA degradation14 and growth arrest15 were genotype-independent, and thus the
genotype-independent S-RNase uptake was attributed to an artefact of in vitro
germinated pollen16.
Our studies support a model for self-incompatibility in which an RNase inhibitor must
be present inside pollen tubes. The current inhibitor model of gametophytic selfincompatibility3 identifies pollen-S with this RNase inhibitor. Pollen-S, like stylar SRNase, would contain two domains, one binding in an allele-specific way to the
recognition domain of its corresponding S-RNase, the other binding to (and inhibiting)
the activity domain of any other S-RNase. For the self-incompatibility system to work,
one must assume that binding to the two domains is mutually exclusive, and that
binding to the recognition domain is thermodynamically favoured over binding to the
RNase activity domain. Thus, both binding of pollen-S to the recognition domain (for
incompatible crosses) or to the RNase activity domain (for compatible crosses) would
result in S-RNase accumulation in pollen tubes, as we show here.
How does the inhibitor model deal with the breakdown of self-incompatibility in pollenpart mutations? In most cases1, 17 the breakdown has been associated with the presence
of two different S-alleles in the same pollen grain, a situation similar to the
compatibility of tetraploids derived from self-incompatible diploids (also termed
competitive interaction). To explain the competitive interaction, the inhibitor model
requires a mechanism enabling pollen-S to bind to the RNase activity domain even in
the presence of its cognate S-RNase recognition domain. In other cases, where
competitive interaction had been ruled out18, 19, pollen compatibility was presumed to
result from deletion or inactivation of the pollen S-allele. However, the inhibitor model
for pollen-S predicts that deletions of pollen-S should be lethal17. To reconcile the
inhibitor model with the compatibility of pollen-S deletions, we propose that the two
functions of RNase inhibition and allele-specific recognition are carried by separate
proteins, RNase inhibitor and pollen-S, respectively. In our view, all pollen contain a
general RNase inhibitor that can potentially bind and inactivate any S-RNase. This
binding can only be prevented when pollen-S binds to the recognition domain of its
cognate S-RNase, which would lead to the self-incompatibility reaction by activation of
the RNase activity. Our version of the inhibitor model predicts that deletion of pollen-S
would produce fully compatible pollen, and could best be confirmed by recovery of
self-compatible pollen-part mutants generated by mutagenesis of S-homozygous plants.
The compatibility of S-heteroallelic pollen could now be due either to a mechanism
impeding pollen-S binding to the recognition domain of its cognate S-RNase or to a
simple lack of pollen-S expression.
Methods
Plant material Two fully self-incompatible but cross-compatible diploid Solanum
chacoense Bitt. lines were obtained from the Potato Introduction Station (Sturgeon Bay,
Wisconsin). The self-incompatibility alleles in the parental line PI458314 were
identified by genetic crosses and gene sequence as S11 and S12 (ref. 20 ) and the alleles
in the parental line PI230582 identified as S 13 and S14 (ref. 21). The F 1 hybrids used in
the present study were obtained by genetic crosses of the parental lines, and include G4
(S12S14) (ref. 22) and V22 (S11S13) (ref. 23), whereas VF60 (S12S12) was a selfed line
from the parental line PI458314 (ref. 24).
Microscopy and immunolabelling Styles from pollinated flowers were harvested
18 hours post-pollination and either fixed immediately in 0.5% glutaraldehyde in 0.1 M
PBS (pH 7.4) as preparatory for immunogold labelling25, or squashed and stained with a
0.2% solution of aniline blue to estimate the location of the pollen tube tips by
ultraviolet microscopy26. Samples fixed with glutaraldehyde for 1.5 hours at room
temperature were dehydrated and embedded in LR White (Electron Microscopy
Services). Regions of the blocks containing the tips of growing pollen tubes were
selected by observation of sections stained with toluidine blue. Regions behind the
growing pollen tips do not contain cytoplasm and are not stained with the antibody. The
selected regions were then thin-sectioned (150–170 nm 'purple' sections) and placed on
Formvar-coated grids. Sections were incubated with a 1/50 dilution of a rabbit anti-S11
antibody for 1 hour at room temperature, washed with PBST (PBS containing 0.05%
w/v Tween 20) then incubated with a secondary goat anti-rabbit conjugated to 20-nm
gold particles (Ted Pella Inc.). To confirm identification, some samples were hybridized
to an antisense pollen-specific Solanum gene probe homologous to the Petunia hybrida
PGP35 (Accession number AF161330). The single-strand-RNA probe was prepared
using T7 RNA polymerase and Digoxigenin (DIG)-labelled UTP (Boehringer
Mannheim). The probe was hybridized to the sections for 2 hours at 42 °C in a solution
containing 40% formamide as described27. The DIG label was detected using a sheep
anti-DIG conjugated to 10-nm gold particles (Ted Pella Inc). The sequence of
antibodies for double labelling subsequent to the in situ hybridization was either (1) 10nm anti-DIG, anti-S11, 20-nm anti-rabbit or (2) anti-S11 followed by a mixture of 10-nm
anti-DIG and 20-nm anti-rabbit. Results were identical with both protocols. Sections
were observed using a JEOL JEM 100S operating at 80 kV. The rabbit anti-S11
antibody was prepared by Cocalico Biological Inc. against the peptide
KPKLTYNYFSDKMLN synthesized as a branched multiple-antigen peptide (Research
Genetics). On standard western blots, the antibody reacts only with the S11-RNase8.
Received 5 June 2000;
accepted 19 July 2000
References
1. de Nettancourt, D. Incompatibility in Angiosperms (Springer, New York, 1977).
2. McClure, B. et al. Style self-incompatibility products of Nicotiana alata are ribonucleases.
Nature 32, 955-957 (1989).
3. Kao, T.-H. & McCubbin, A. How flowering plants discriminate between self and non-self
pollen to prevent inbreeding. Proc. Natl Acad. Sci. USA 93, 12059-12065 (1996).
4. Dodds, P., Clarke, A. & Newbigin, E. A molecular perspective on pollination in flowering
plants. Cell 85, 141-144 (1996). Links
5. Huang, S., Lee, H.-S., Karunanandaa, B. & Kao, T.-H. Ribonuclease activity of Petunia
inflata S proteins is essential for rejection of self pollen. Plant Cell 6, 1021-1028 (1994).
Links
6. McClure, B., Gray, J., Anderson, M. & Clarke, A. Self-incompatibility in Nicotiana alata
involves degradation of pollen RNA. Nature 347, 757-760 (1990).
7. Matton, D. P. et al. Hypervariable domains of self-incompatibility RNases mediate allelespecific pollen recognition. Plant Cell 9, 1757-1766 (1997).
8. Matton, D. P. et al. The production of an S-RNase with dual specificity suggests a novel
hypothesis for the generation of new S-alleles. Plant Cell 11, 2087-2097 (1999). Links
9. Cheung, A. The pollen tube growth pathway: its molecular and biochemical contributions
and responses to pollination. Sex. Plant Reprod. 9, 330-336 (1996).
10. de Nettancourt, D. et al. Ultrastructural aspects of the self-incompatibility mechanism in
Lycopersicum peruvianum Mill. J. Cell Sci. 12, 403-419 (1973). Links
11. Steer, M. W. & Steer, J. M. Pollen tube tip growth. New Phytol. 111, 323-358 (1989).
12. Anderson, M. et al. Cloning of cDNA for a stylar glycoprotein associated with expression of
self-incompatibility in Nicotiana alata. Nature 321, 38-44 (1986).
13. Certal, A. et al. S-RNases in apple are expressed in the pistil along the pollen tube growth
path. Sex. Plant Reprod. 12, 94-98 (1999).
14. Gray, J., McClure, B., Bönig, I., Anderson, M. & Clarke, A. Action of the style product of the
self-incompatibility gene of Nicotiana alata (S-RNase) on in vitro grown pollen tubes. Plant
Cell 3, 271-283 (1991).
15. Jahnen, W., Lush, W. & Clarke, A. Inhibition of in vitro pollen tube growth by isolated Sglycoproteins of Nicotiana alata. Plant Cell 1, 501-510 (1989).
16. Hess, D., Gresshoff, P., Fielitz, U. & Gleiss, D. Uptake of protein and bacteriophage into
swelling and germinating pollen of Petunia hybrida. Z. Pflanzenphysiol. 74, 371-376 (1974).
17. Golz, J., Clarke, A. E. & Newbegin, E. Mutational approaches to the study of selfincompatibility: revisiting the pollen-part mutants. Ann. Botany 85, 95-103 (2000).
18. Lewis, D. Chromosome fragments and mutation of the incompatibility gene. Nature 190,
990-991 (1961).
19. Pandey, K. Elements of the S-gene complex. II. Mutations and complementation at the SI
locus in Nicotiana alata. Heredity 22, 255-284 (1967).
20. Saba-El-Leil, M., Rivard, S., Morse, D. & Cappadocia, M. The S11 and S13 self
incompatibility alleles in Solanum chacoense Bitt. are remarkably similar. Plant Mol. Biol. 24,
571-583 (1994). Links
21. Despres, C., Saba-El-Leil, M., Rivard, S., Morse, D. & Cappadocia, M. Molecular cloning of
two Solanum chacoense S-alleles and a hypothesis concerning their evolution. Sex. Plant
Reprod. 7, 169-176 (1994).
22. Van Sint Jan, V., Laublin, G., Birhman, R. & Cappadocia, M. Genetic analysis of leaf explant
regenerability in Solanum chacoense. Plant Cell Tissue Org. Cult. 47, 9-13 (1996).
23. Veronneau, H., Lavoie, G. & Cappadocia, M. Genetic analysis of anther and leaf disk culture
in two clones of Solanum chacoense Bitt and their reciprocal hybrids. Plant Cell Tissue Org.
Cult. 30, 199-209 (1992).
24. Rivard, S., Saba-El-Leil, M., Landry, B. & Cappadocia, M. RFLP analyses and segregation
of molecular markers in plants produced by in vitro anther culture, selfing, and reciprocal
crosses of two lines of self-incompatible Solanum chacoense. Genome 37, 775-783 (1994).
25. Vandenbosch, K. in Electron Microscopy of Plant Cells (eds Hall, J. & Hawes, C.) 181-218
(Academic, New York, 1991).
26. Martin, F. Staining and observing pollen tubes in the style by means of fluorescence. Stain
Technol. 34, 125-128 (1959).
27. McFadden, G. in Electron Microscopy of Plant Cells (eds Hall, J. & Hawes, C.) 220-255
(Academic, New York, 1991).
Acknowledgements. We thank L. Pelletier and N. Nassoury for technical assistance, G.
Teodorescu for plant care, and S. McCormick, A. Cheung, T.-H. Kao and V. de Luca
for helpful discussions. The work was supported by a fellowship from Programme
Québecois de Bourses d'Excellence, Québec (D.-T.L.) and by grants from NSERC
(M.C.) and Fonds pour la Formation de Chercheurs et l'Aide à la Recherche (D.M.,
M.C.).
Figure 1 S11-RNase entry into incompatible pollen tubes. a, Self-pollinated styles of an S11S13 plant
were fixed 18 hours post-pollination, sectioned in region II, and stained sequentially with the antiS11 antibody and a 20-nm gold-labelled secondary antibody. b, c, The apical tips of all growing
pollen tubes (P) are labelled to a greater extent than either the transmitting tissue cells (TT) or the
extracellular matrix (ECM) separating the cells. Label inside pollen tubes appears associated with
the cytoplasm (Pc) rather than the vacuolar spaces (Pv), unlike the labelling found in transmitting
tissue cells. d, Only background labelling is present in the cytoplasm, the surrounding cell wall or
the cuticle (Cu) of epidermal cells (EP), which do not express S-RNases. All scale bars are 1 µm.
Figure 2 S11-RNase entry into compatible pollen tubes. Pollinated styles were treated as described
in Fig. 1. a, A fully compatible cross of S11S13 styles pollinated with S12 pollen. b, A semicompatible cross of S11 S13 plants pollinated with S13 and S14 pollen. The region observed is close to
the middle of the style and is deduced to contain only fast-growing compatible S14 pollen tubes.
The several electron-lucent regions correspond to secretory vesicles (Psv). c, A fully compatible
cross of S12S14 styles pollinated with S11 and S13 pollen, showing the background labelling of the
pollen tubes in the absence of an S11-RNase. Cells are labelled as in Fig. 1 and all scale bars are
1 µm