1
Genetic mapping of sex and self-incompatibility determinants in the androdioecious
1
plant Phillyrea angustifolia
2
3
Authors:
4
Amélie Carré
1
, Sophie Gallina
1
, Sylvain Santoni
2
, Philippe Vernet
1
, Cécile Godé
1
, Vincent Castric
1
, Pierre
5
Saumitou-Laprade
1#
6
7
Affiliations :
8
1
CNRS, Univ. Lille, UMR 8198 – Evo-Eco-Paleo, F-59000 Lille, France
9
2
UMR DIAPC - Diversité et adaptation des plantes cultivées
10
11
12
#
Corresponding author: Pierre.Saumitou-Laprade@univ-lille.fr
13
14
15
16
.CC-BY-NC 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 6, 2021. ; https://doi.org/10.1101/2021.04.15.439943doi: bioRxiv preprint
2
Abstract
17
The diversity of mating and sexual systems in Angiosperms is spectacular, but the factors driving their
18
evolution remain poorly understood. In plants of the Oleaceae family, an unusual self-incompatibility (SI)
19
system has been discovered recently, whereby only two distinct homomorphic SI specificities segregate
20
stably. To understand the role of this peculiar SI system in preventing or promoting the diversity of sexual
21
phenotypes observed across the family, an essential first step is to characterize the genetic architecture
22
of these two traits. Here, we developed a high-density genetic map of the androdioecious shrub P.
23
angustifolia based on a F1 cross between a hermaphrodite and a male parent with distinct SI genotypes.
24
Using a double restriction-site associated digestion (ddRAD) sequencing approach, we obtained reliable
25
genotypes for 196 offspring and their two parents at 10,388 markers. The resulting map comprises 23
26
linkage groups totaling 1,855.13 cM on the sex-averaged map. We found strong signals of association for
27
the sex and SI phenotypes, that were each associated with a unique set of markers on linkage group 12
28
and 18 respectively, demonstrating inheritance of these traits as single, independent, mendelian factors.
29
The P. angustifolia linkage map shows robust synteny to the olive tree genome overall. Two of the six
30
markers strictly associated with SI in P. angustifolia have strong similarity with a recently identified 741kb
31
chromosomal region fully linked to the SI phenotype on chromosome 18 of the olive tree genome,
32
providing strong cross-validation support. The SI locus stands out as being markedly rearranged, while the
33
sex locus has remained relatively more collinear between the two species. This P. angustifolia linkage map
34
will be a useful resource to investigate the various ways by which the sex and SI determination systems
35
have co-evolved in the broader phylogenetic context of the Oleaceae family.
36
.CC-BY-NC 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 6, 2021. ; https://doi.org/10.1101/2021.04.15.439943doi: bioRxiv preprint
3
Introduction
37
Modes of sexual reproduction are strikingly diverse across Angiosperms, both in terms of the
38
proportion of autogamous vs. allogamous matings and in terms of the distribution of male and female
39
sexual functions within and among individuals (BARRETT 1998; SAKAI AND WELLER 1999; DIGGLE et al. 2011).
40
The conditions under which this diversity could arise under apparently similar ecological conditions and
41
have evolved rapidly -sometimes even within the same family- have been a topic of intense interest in
42
evolutionary biology (BARRETT 1998). The control of self-fertilization and the delicate balance between its
43
costs and benefits is considered to be a central force driving this diversity. Avoidance of self-fertilization is
44
sometimes associated with observable phenotypic variations among reciprocally compatible partners.
45
These variations can be morphological (e.g. distyly) or temporal (e.g. protandry, protogyny in the case of
46
heterodichogamy), but in many cases the flowers show no obvious morphological or phenological
47
variation, and self-fertilization avoidance relies on so-called “homomorphic” self-incompatibility (SI)
48
systems. These systems are defined as the inability of fertile hermaphrodite plants to produce zygotes
49
through self-fertilization (LUNDQVIST 1956; DE NETTANCOURT 1977), and typically rely on the segregation of
50
a finite number of recognition “specificities” whereby matings between individuals expressing cognate
51
specificities are not successful at producing zygotes. At the genetic level, the SI specificities most
52
commonly segregate as a single multi-allelic mendelian locus, the S locus. This locus contains at least two
53
genes, one encoding the male determinant expressed in pollen and the other encoding the female
54
determinant expressed in pistils, with the male specificity sometimes determined by a series of tandemly
55
arranged paralogs (KUBO et al. 2015). The male and female determinants are both highly polymorphic and
56
tightly linked, being inherited as a single non-recombining genetic unit. In cases where the molecular
57
mechanisms controlling SI could be studied in detail, they were found to be remarkably diverse, illustrating
58
their independent evolutionary origins across the flowering plants (IWANO AND TAKAYAMA 2012). Beyond
59
the diversity of the molecular functions employed, SI systems can also differ in their genetic architecture.
60
.CC-BY-NC 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 6, 2021. ; https://doi.org/10.1101/2021.04.15.439943doi: bioRxiv preprint
4
In the Poaceae family for example, two independent loci (named S and Z) control SI (Yang, et al., 2008). In
61
other cases, the alternate allelic specificities can be determined by presence-absence variants rather than
62
nucleotide sequence variants of a given gene, such as e.g. in Primula vulgaris, where one of the two
63
reproductive phenotypes is hemizygous rather than heterozygous for the SI locus (LI et al. 2016).
64
In spite of this diversity of molecular mechanisms and genetic architectures, a common feature of
65
SI phenotypes is that they are all expected to evolve under negative frequency-dependent selection, a
66
form of natural selection favoring the long-term maintenance of high levels of allelic diversity (WRIGHT
67
1939). Accordingly, large numbers of distinct SI alleles are commonly observed to segregate within natural
68
and cultivated SI species (reviewed in CASTRIC AND VEKEMANS 2004). There are notable exceptions to this
69
general rule, however, and in some species only two SI specificities seem to segregate stably. Most often
70
in such diallelic SI systems, the two SI specificities are in perfect association with morphologically
71
distinguishable floral phenotypes. In distylous species, for instance, two floral morphs called “pin” (L-
72
morph) and “thrum” (S-morph) coexist (BARRETT 1992; BARRETT 2019). In each morph, the anthers and
73
stigma are spatially separated within the flowers, but located at corresponding, reciprocal positions
74
between the two morphs. Additional morphological differences exist, with S-morph flowers producing
75
fewer but larger pollen grains than L-morph flowers (DULBERGER 1992). These morphological differences
76
are believed to enhance the selfing avoidance conferred by the SI system but also to increase both male
77
and female fitnesses (BARRETT 1990; BARRETT 2002; KELLER et al. 2014), although it is not clear which of SI
78
or floral morphs became established in the first place (CHARLESWORTH AND CHARLESWORTH 1979).
79
The Oleacea family is another intriguing exception, where a diallelic SI system was recently found
80
to be shared across the entire family (VERNET et al. 2016). In this family of trees, the genera Jasminum (2n
81
= 26), Fontanesia (2n = 26) and Forsythia (2n= 28) are all heterostylous and are therefore all expected to
82
possess a heteromorphic diallelic SI system; in Jasminum fruticans self- and within-morph fertilization are
83
unsuccessful (DOMMÉE et al. 1992). The ancestral heterostyly gave rise to species with hermaphrodite (e.g.
84
.CC-BY-NC 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 6, 2021. ; https://doi.org/10.1101/2021.04.15.439943doi: bioRxiv preprint
5
Ligustrum vulgare, Olea europaea), androdioecious (e.g. P. angustifolia, Fraxinus ornus), polygamous (e.g.
85
Fraxinus excelsior) and even dioecious (e.g. Fraxinus chinensis) sexual systems, possibly in association with
86
a doubling of the number of chromosomes (2n= 46 in the Oleeae tribe) (TAYLOR 1945; WALLANDER AND ALBERT
87
2000). Evaluation of pollen germination success in controlled in vitro crossing experiments (whereby
88
fluorescence microscopy is used to score the growth of pollen tubes reaching the style through the stigma;
89
referred to below as the “stigma test”) revealed the existence of a previously unsuspected homomorphic
90
diallelic SI in one of these species, P. angustifolia (SAUMITOU-LAPRADE et al. 2010). In this androdioecious
91
species (i.e. in which male and hermaphrodite individuals coexist in the same populations), hermaphrodite
92
individuals form two morphologically indistinguishable groups of SI specificities that are reciprocally
93
compatible but incompatible within groups, whereas males show compatibility with hermaphrodites of
94
both groups (SAUMITOU-LAPRADE et al. 2010). This “universal” compatibility of males offsets the
95
reproductive disadvantage they suffer from lack of their female function, such that the existence of the
96
diallelic SI system provides a powerful explanation to the long-standing evolutionary puzzle represented
97
by the maintenance of high frequencies of males in this species (PANNELL AND KORBECKA 2010; SAUMITOU-
98
LAPRADE et al. 2010; BILLIARD et al. 2015; PANNELL AND VOILLEMOT 2015). Extension of the stigma test
99
developed in P. angustifolia to other species of the same tribe including L. vulgaris (DE CAUWER et al. 2020),
100
F. ornus (VERNET et al. 2016) and O. europaea (SAUMITOU-LAPRADE et al. 2017; DUPIN et al. 2020),
101
demonstrated that all species exhibited some form of the diallelic SI system, but with no consistent
102
association with floral morphology. Cross-species pollination experiments even showed that pollen from
103
P. angustifolia was able to trigger a robust SI response on O. europaea and the more distant F. ornus and
104
F. excelsior stigmas (the reciprocal is also true). This opens the question of whether the homomorphic
105
diallelic SI determinants are orthologs across the Oleeae tribe, even in the face of the variety of sexual
106
polymorphisms present in the different species. More broadly, the link between determinant of the
107
homomorphic diallelic SI in the Oleeae tribe and those of the heteromorphic diallelic SI in the ancestral
108
.CC-BY-NC 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 6, 2021. ; https://doi.org/10.1101/2021.04.15.439943doi: bioRxiv preprint
