A central goal of biology is to understand the evolutionary forces and molecular processes by which new traits and functions emerge in living organisms. A major challenge toward this goal is that many cellular processes rely on interactions between molecular partners (protein-protein interactions or regulatory interactions between for example transcription factors and their binding sites; Boyle et al., 2017; Courtier-Orgogozo et al., 2019) rather than on the action of individual components in isolation. At present, the overall structure of protein-protein or regulatory interaction networks is just starting to be understood at the genome level in a handful of model organisms, and a general understanding of how the diversification of these networks led to the emergence of new biological functions in distinct lineages is crucially lacking (Nooren and Thornton, 2003; Andreani and Guerois, 2014). In two-component genetic systems, specificity of the interaction can be very tight such as for example in receptor-ligand interactions (Laub and Goulian, 2007) or bacterial toxin-antitoxin systems (Aakre et al., 2015). In these systems, the emergence of novel traits involves evolutionary modification of the interacting partners, leading to potential disruption of the interaction or to detrimental cross-talk, at least transiently (Plach et al., 2017). A central aspect of the diversification process therefore concerns the functional nature of the evolutionary intermediate. Two main scenarios have been proposed (Aakre et al., 2015). The first scenario posits that a functional change results from introduction of a mutation that first modifies one of the two partners, initially disrupting the functional interaction and leading to a non-functional intermediate. If fitness of this non-functional intermediate is not impaired too drastically, it may remain in the population until a compensatory mutation hits the second partner and rescues the interaction in a modified state, resulting in the novel function. An alternative scenario though, is that the first mutation may broaden rather than inactivate specificity of one of the two interacting partners, transiently releasing constraint on the other before specificity of the first partner becomes restricted again, thus maintaining functionality of the system all along the process. Distinguishing between these two scenarios requires functional characterization of the ancestral evolutionary intermediates, which has been achieved in a very limited number of biological cases such as bacterial toxin-antitoxin systems (Aakre et al., 2015), acquisition of cortisol specificity of the mammalian glucocorticoid receptor (Bridgham et al., 2009) or mammalian retinoic acid receptors (Gutierrez-Mazariegos et al., 2016). The scenario of a promiscuous intermediate seems to have received better support, but whether this is a truly general pattern remains to be determined. A major limitation is that detailed population genetic models taking into account how natural selection is acting on variants of the two genes given their specific biological function are lacking for most genetic systems (Aakre et al., 2015; Aharoni et al., 2005; Bloom and Arnold, 2009; Bridgham et al., 2009; Matsumura and Ellington, 2001; Matton et al., 2000; Sayou et al., 2014).

Self-incompatibility (SI) in flowering plants is a prime biological system to investigate how functional diversification can proceed between two molecular partners over the course of evolution. SI has evolved as a strategy to prevent selfing and enforce outcrossing, thereby avoiding the deleterious effects of inbreeding depression (Nettancourt, 1977; Kitashiba and Nasrallah, 2014). In Brassicaceae, SI is genetically controlled by a single multiallelic non-recombining locus called the S-locus (Nettancourt, 1977). This locus encodes two highly polymorphic self-recognition determinant proteins: the female cell-surface ‘receptor’ S-LOCUS RECEPTOR KINASE (SRK) is expressed in stigmatic papillae cells (Takasaki et al., 2000) and the male ‘ligand’ S-LOCUS CYSTEIN RICH PROTEIN (SCR) is displayed at the pollen surface (Schopfer et al., 1999; Takasaki et al., 2000; Takayama et al., 2000). The SI response consists in pollen rejection by the pistil and is induced by allele-specific interaction between the SRK and SCR proteins when encoded by the same haplotype (S-haplotype, Kachroo et al., 2001; Nasrallah and Nasrallah, 1993; Takayama et al., 2001). These two interacting proteins exhibit a high degree of sequence variability and a large number of these highly diverged allelic variants segregate in natural populations of SI species (several dozens to nearly two hundred, Busch et al., 2014; Castric and Vekemans, 2004; Lawrence, 2000). This large allelic diversity of receptor-ligand combinations must have arisen through repeated diversification events, but the molecular process and evolutionary scenario involved in the diversification of an ancestral haplotype into distinct descendant specificities are still unresolved.

Three evolutionary scenarios for emergence of new self-incompatibility specificities have been proposed so far (Charlesworth et al., 2005). The ‘compensatory mutation’ scenario posits that diversification proceeds through self-compatible intermediates that is following transient disruption of the receptor-ligand interaction (Figure 1—figure supplement 1A; Gervais et al., 2011; Uyenoyama et al., 2001). Population genetic analysis confirmed that S-allele diversification is indeed possible through this scenario, but only under some combinations of model parameters (high inbreeding depression, high rate of self-pollination and low number of co-segregating S-alleles, Gervais et al., 2011). Under these restrictive conditions, the ancestral recognition specificity can be maintained in the long run along with the derived specificity, effectively resulting in allelic diversification. A second scenario was presented by Chookajorn et al. (2004), who proposed that diversification results from the progressive reinforcement of SRK/SCR recognition capacity among slight functional variants of SCR and SRK that might segregate within the population. At the limit of the very low level of intra-allelic polymorphism observed in natural populations (Castric et al., 2010) this model bears similarity to the model of Gervais et al. (2011), who also predicted the maintenance of an ancestral recognition specificity unchanged and the evolution of a new divergent specificity, except it assumes no SC intermediate. However, the progressive reinforcement process proposed by this model may rather result in the production of two descendant allelic specificities that are both functionally distinct from their ancestor (Figure 1—figure supplement 1B). A third scenario was proposed by Matton et al. (1999) and involves a dual-specificity intermediate (Figure 1—figure supplement 1C). This scenario was criticized on population genetics grounds because such a dual-specificity haplotype would recognize and reject more potential mates for reproduction than its progenitor haplotype resulting in lower reproductive success, and should therefore be disfavoured by natural selection and quickly eliminated from the populations (Charlesworth, 2000; Uyenoyama and Newbigin, 2000). A main unanswered question is whether functional specificities remain stable over time or are subject to frequent turnover. Detailed theoretical analysis of the model of Gervais et al. (2011) showed that under a large portion of the parameter space, the introduced self-compatible intermediate is predicted to exclude its functional ancestor from the population. Secondary introduction of the compensatory mutation then effectively results in turnover of recognition specificities along allelic lines rather than in diversification per se. A turnover of recognition specificities may therefore be expected along each allelic lineage, rather than their long-term maintenance over evolutionary times. A similar process could also be occurring according to the progressive reinforcement model of Chookajorn et al. (2004). Overall, while the large allelic diversity at the S-locus is one of the most striking features of this system, the evolutionary scenario and molecular route by which new S-alleles arise remain largely unknown. An essential limitation in the field is the crucial absence of direct experimental approaches.

To conclusively evaluate these models we used ancestral sequence reconstruction of S-haplotypes that are currently segregating in the plant Arabidopsis halleri. Ancestral sequence reconstruction is a new approach to 'resurrect' ancestral biological systems, whereby the sequence of an ancestral gene or allele is determined through phylogenetic analyses, then synthesized de novo and its functional properties (in terms of e.g. biochemical activity, conformation or binding capacity) are compared to those of its contemporary descendants. Although most S-haplotypes in A. halleri are typically so highly diverged that even sequence alignment can be challenging, previous work identified a pair of S-haplotypes (S03 and S28) with high phylogenetic proximity based on a fragment of the SRK gene (Castric et al., 2008). We took this opportunity to reconstruct, resurrect and phenotypically characterize their putative last common ancestral receptor SRKa (‘a’ for ancestor, Figure 1) into A. thaliana, which has previously been established as a model plant for mechanistic studies of SI (Nasrallah et al., 2002; Tsuchimatsu et al., 2010).

Figure 1 with 3 supplements see all

Download asset Open asset

Expression of S03 and S28 recognition specificities in A. thaliana by genetic transformation

Because of their relatively high sequence similarity, we first ensured that the S03 and S28 recognition specificities are indeed phenotypically distinct when expressed in A. thaliana. It was previously shown that the self-compatible model plant A. thaliana can mount a functional self-incompatibility response upon introduction of the S-locus genes SCR and SRK (Boggs et al., 2009; Durand et al., 2014; Nasrallah et al., 2002) but it is unclear whether transfer is possible for all SCR and SRK variants. We first assessed the activity of AhSRK03 and AhSRK28 promoters in transformed A. thaliana lines (Figure 2—figure supplement 1), which enabled us to identify the temporal window of SRK expression (Figure 2—figure supplement 2). Then, we used A. halleri pollen carrying either the S03 or S28 male determinant to validate that a proper SI response was indeed successfully transferred in transformed A. thaliana lines for each of our SRK receptors, and that this response was strictly specific toward either of their cognate ligand (two replicate transgenic lines for SRK03 and SRK28 each, Figure 2).

Figure 2 with 2 supplements see all

Download asset Open asset

Figure 2—source data 1

Number of germinated pollen grains per stigma after pollination.

Two SRK03 and SRK28 lines were pollinated with A. halleri pollen expressing either S03 or S28 specificities.

https://cdn.elifesciences.org/articles/50253/elife-50253-fig2-data1-v2.xlsx

Download elife-50253-fig2-data1-v2.xlsx

Abundant germination of pollen tubes occurred in crosses involving non-cognate SRK and SCR proteins, whereas crosses involving proteins from the same S-haplotype induced inhibition of pollen germination, which is characteristic of an incompatible response. Intensity of the incompatible response was then quantified by counting the number of germinated pollen grains at the pistil surface of two selected lines for each transgene (Figure 2A and B). Incompatible responses are characterized by less than 10 germinated pollen tubes whereas compatible crosses induced significant pollen germination, with more than 50 pollen tubes per stigma. This shows that cross-species pollinations of A. thaliana with A. halleri pollen can induce robust SI reactions, and also that S03 and S28 have distinct recognition specificities, despite their close phylogenetic proximity.

Ligand specificity of the ancestral SRK

We then evaluated the recognition phenotype conferred by these putative ancestral SRK sequences by performing controlled pollination assays for all remaining 23 SRKa lines with pollen from A. halleri S03, A. halleri S28 and A. thaliana C24 (used as a control to validate receptivity of the acceptor pistil). The SI reaction was scored as the number of germinated pollen grains per stigma. All 23 tested SRKa expressing lines showed intense pollen germination with C24 pollen, idid not impair pistil fertility or ability to receive pollen (Figure 3 and Figure 3—figure supplement 4).

Figure 3 with 4 supplements see all

Download asset Open asset

Figure 3—source data 1

Number of germinated pollen grains per stigma after pollination.

Each SRKa line was pollinated with A. thaliana C24, A. halleri S03 and S28 pollen.

https://cdn.elifesciences.org/articles/50253/elife-50253-fig3-data1-v2.xls

Download elife-50253-fig3-data1-v2.xls

With a single exception (line p03_SRKa^MA_k03_2), S28 pollen induced no consistent SI rejection in any of the lines expressing the different versions of the ancestral receptor, and was indistinguishable from the compatible control (C24, Figure 3). In stark contrast, SCR03 pollen induced robust incompatibility reactions in the vast majority (12 of 14) of the lines with amino acid I at position 208 (SRKa^IA and SRKa^IG, Figure 3). Hence, we conclude that the S03 haplotype has retained the ancestral recognition specificity largely unaltered, while the S28 haplotype underwent substantial modification and represents a novel recognition specificity, leading to functional diversification.

The incompatibility reactions involving our different ancestral SRK S-domain sequences were largely independent of the sequence context into which we introduced them (Figure 3, Figure 1—figure supplement 3). Similarly, the amino acid present at position 305 of the reconstructed sequence had no effect on the response: lines with either A or G produced rejection reactions equally. In contrast, only four of the lines with amino acid M at position 208 (SRKa^MA and SRKa^MG) displayed incompatibility reactions, and these were weak (Figure 3). Hence, the single change, from the ancestral Isoleucine amino acid (I) at position 208 to Methionine (M), seems to be sufficient to almost fully inactivate the ancestral receptor's ability to recognise SCR03.

Structural determinants of SRK specificity

We next sought to identify the structural determinants of the S03 and S28 specificities. We used the structural model of the interaction complex formed by SCR and SRK (Ma et al., 2016) to compare cognate and non cognate complexes in terms of number and nature of atomic contacts and predict how the ancestral receptor is expected to interact with either SCR03 or SCR28. As previously demonstrated by Ma et al. (2016), SCR binding induces homodimerization of the extracellular domain of SRK (eSRK, Figure 3—figure supplement 1), forming eSRK:SCR heterotetramers composed of two SCR molecules bound to two SRK molecules (Ma et al., 2016). The switch in recognition specificity occurred along the branch leading to SRK28, and therefore must involve some of the 29 amino acid differences along this branch. In contrast, the 31 amino acid differences that accumulated along the branch leading to SRK03 have not fully altered the recognition specificity. Overall, we find a slightly higher number of atomic contacts between the SCR and SRK molecules in the SRK03-SCR03 cognate complex (1543 contacts over 91 and 62 distinct SRK and SCR amino acids respectively) than in the SRK28-SCR28 cognate complex (1395 contacts over 89 and 57 distinct SRK and SCR amino acids respectively, Supplementary file 1). As already noted by Ma et al. (2016) in Brassica, and in line with the distribution of positively selected sites along the sequence (Castric and Vekemans, 2007), in both cases the interaction interface was mostly concentrated around the three ‘hypervariable’ portions of the SRK protein (Figure 4).

Figure 4

Download asset Open asset

Contrary to our expectation, however, the non-cognate complexes did not differ substantially from the cognate complexes on the basis of their total number of atomic contacts (Figure 4). Specifically, the SRK03/SCR28 and SRK28/SCR03 non-cognate complexes were predicted to establish a total of 1546 and 1328 contacts, respectively, between atoms of their SCR and SRK molecules, very close to the 1543 and 1395 predicted contacts for the SRK03/SCR03 and SRK28/SCR28 cognate complexes (Supplementary file 1). They were also not characterized by any obvious steric clashes that would prevent the proper docking of SCR and SRK. Hence, the overall stability of the complex does not seem to be the primary determinant of recognition specificity. Consequently, the specificity of the interaction must be a function of some qualitative features rather than of the quantitative strength of the interaction.

While the majority of amino acids retained unchanged atomic contacts in cognate vs. non-cognate interactions, using a 5% threshold we identified six amino acid positions of the eSRK protein that displayed important differences in terms of contacts they establish with SCR in the cognate vs. the non-cognate complexes (Figure 5—figure supplement 1B, C, E and F). Four of these amino acid positions also displayed very contrasted interactions in the S03 vs. the S28 complexes (residues S201, R220, R279, W296, Figure 5—figure supplement 1A and D). For instance, the R residue at position 279 of SRK establishes three times more contacts with SCR in the SRK03/SCR03 than in the SRK28/SCR28 complex (Figure 5, Figure 5—figure supplement 1A). These four residues are therefore prime candidates for the determination of binding specificity. Intriguingly, three of the six amino acids were identical between SRK03 and SRK28, and yet interacted in sharply different ways with their respective SCR ligands. For instance, while identical between the SRK03 and SRK28 sequences, residues W296 and R279 established three to four times more contacts in the S03 than in the S28 complex, and the R220 residue is involved in twice as many contacts in the S28 than in the S03 complex (Figure 5, Figure 5—figure supplement 1A). This suggests that involvement of these residues in the activity of the complex is mediated by substitutions at other positions along the protein sequence, for instance by displacing them spatially (Figure 5).

Figure 5 with 1 supplement see all

Download asset Open asset

We then compared the predicted binding features of the functional form of the putative ancestral receptor (SRKa^IA) with both SRK03 and SRK28. When complexed with SCR03, 46 SRKa^IA interacting pairs of amino acids were identical with SRK03, whereas only 24 were identical with SRK28. Similarly, when complexed with SCR28, 37 SRKa^IA interacting pairs of amino acid residues were identical with SRK03, but only 18 were identical with SRK28 (Figure 6). This trend is also visible at the level of individual atomic contacts and at the level of involved amino acids both in SRK and SCR molecules (Figure 6—figure supplement 1). Hence, even though there is no quantitative difference in how strongly SRK03 and SRK28 bind SCR03 and SCR28 (Supplementary file 1), the amino acids that come into contact in the two complexes are not identical, and the contacts established by SRKa^IA with SCR qualitatively resemble more those established by SRK03 than those established by SRK28.

Figure 6 with 1 supplement see all

Download asset Open asset

Overall, asymmetrical diversification of the ancestor into the phenotypically distinct S03 and S28 specificities is therefore associated with structural changes of the SRK receptor and the way it interacts with its SCR ligand. Despite similar levels of overall sequence divergence between SRKa and either SRK03 or SRK28, the binding properties of SRK03 have remained similar to those of SRKa^IA, while the way SRK28 is predicted to interact with the SCR ligand has diverged from its ancestral state, consistent with the phenotypic switch that has occurred along this lineage.

All data generated or analysed during this study are included in the manuscript and supporting files.

1. 1. Aakre CD
   2. Herrou J
   3. Phung TN
   4. Perchuk BS
   5. Crosson S
   6. Laub MT

   (2015) Evolving new protein-protein interaction specificity through promiscuous intermediates

   Cell 163:594–606.

   https://doi.org/10.1016/j.cell.2015.09.055

   • PubMed
   • Google Scholar
2. 1. Aharoni A
   2. Gaidukov L
   3. Khersonsky O
   4. McQ Gould S
   5. Roodveldt C
   6. Tawfik DS

   (2005) The 'evolvability' of promiscuous protein functions

   Nature Genetics 37:73–76.

   https://doi.org/10.1038/ng1482

   • PubMed
   • Google Scholar
3. 1. Andreani J
   2. Guerois R

   (2014) Evolution of protein interactions: from interactomes to interfaces

   Archives of Biochemistry and Biophysics 554:65–75.

   https://doi.org/10.1016/j.abb.2014.05.010

   • Google Scholar
4. 1. Bechsgaard JS
   2. Castric V
   3. Charlesworth D
   4. Vekemans X
   5. Schierup MH

   (2006) The transition to self-compatibility in Arabidopsis thaliana and evolution within S-haplotypes over 10 myr

   Molecular Biology and Evolution 23:1741–1750.

   https://doi.org/10.1093/molbev/msl042

   • PubMed
   • Google Scholar
5. 1. Bloom JD
   2. Arnold FH

   (2009) In the light of directed evolution: pathways of adaptive protein evolution

   PNAS 106:9995–10000.

   https://doi.org/10.1073/pnas.0901522106

   • PubMed
   • Google Scholar
6. 1. Boggs NA
   2. Dwyer KG
   3. Shah P
   4. McCulloch AA
   5. Bechsgaard J
   6. Schierup MH
   7. Nasrallah ME
   8. Nasrallah JB

   (2009) Expression of distinct self-incompatibility specificities in Arabidopsis thaliana

   Genetics 182:1313–1321.

   https://doi.org/10.1534/genetics.109.102442

   • PubMed
   • Google Scholar
7. 1. Boyle EA
   2. Li YI
   3. Pritchard JK

   (2017) An expanded view of complex traits: from polygenic to omnigenic

   Cell 169:1177–1186.

   https://doi.org/10.1016/j.cell.2017.05.038

   • PubMed
   • Google Scholar
8. 1. Bridgham JT
   2. Ortlund EA
   3. Thornton JW

   (2009) An epistatic ratchet constrains the direction of glucocorticoid receptor evolution

   Nature 461:515–519.

   https://doi.org/10.1038/nature08249

   • PubMed
   • Google Scholar
9. 1. Busch JW
   2. Witthuhn T
   3. Joseph M

   (2014) Fewer S-alleles are maintained in plant populations with sporophytic as opposed to gametophytic self-incompatibility

   Plant Species Biology 29:34–46.

   https://doi.org/10.1111/j.1442-1984.2012.00386.x

   • Google Scholar
10. 1. Castric V
    2. Bechsgaard J
    3. Schierup MH
    4. Vekemans X

    (2008) Repeated adaptive introgression at a gene under multiallelic balancing selection

    PLOS Genetics 4:e1000168.

    https://doi.org/10.1371/journal.pgen.1000168

    • PubMed
    • Google Scholar
11. 1. Castric V
    2. Bechsgaard JS
    3. Grenier S
    4. Noureddine R
    5. Schierup MH
    6. Vekemans X

    (2010) Molecular evolution within and between self-incompatibility specificities

    Molecular Biology and Evolution 27:11–20.

    https://doi.org/10.1093/molbev/msp224

    • PubMed
    • Google Scholar
12. 1. Castric V
    2. Vekemans X

    (2004) Plant self-incompatibility in natural populations: a critical assessment of recent theoretical and empirical advances

    Molecular Ecology 13:2873–2889.

    https://doi.org/10.1111/j.1365-294X.2004.02267.x

    • PubMed
    • Google Scholar
13. 1. Castric V
    2. Vekemans X

    (2007) Evolution under strong balancing selection: how many codons determine specificity at the female self-incompatibility gene SRK in Brassicaceae?

    BMC Evolutionary Biology 7:132.

    https://doi.org/10.1186/1471-2148-7-132

    • PubMed
    • Google Scholar
14. 1. Chang BS
    2. Jönsson K
    3. Kazmi MA
    4. Donoghue MJ
    5. Sakmar TP

    (2002) Recreating a functional ancestral archosaur visual pigment

    Molecular Biology and Evolution 19:1483–1489.

    https://doi.org/10.1093/oxfordjournals.molbev.a004211

    • PubMed
    • Google Scholar
15. 1. Charlesworth D

    (2000) How can Two-Gene models of Self-Incompatibility generate new specificities ?

    Evolutionary Dynamics of Dual-Specificity Self-Incompatibility Alleles 12:309–317.

    https://doi.org/10.1105/tpc.12.3.309

    • Google Scholar
16. 1. Charlesworth D
    2. Vekemans X
    3. Castric V
    4. Glémin S

    (2005) Plant self-incompatibility systems: a molecular evolutionary perspective

    New Phytologist 168:61–69.

    https://doi.org/10.1111/j.1469-8137.2005.01443.x

    • PubMed
    • Google Scholar
17. 1. Chookajorn T
    2. Kachroo A
    3. Ripoll DR
    4. Clark AG
    5. Nasrallah JB

    (2004) Specificity determinants and diversification of the Brassica self-incompatibility pollen ligand

    PNAS 101:911–917.

    https://doi.org/10.1073/pnas.2637116100

    • PubMed
    • Google Scholar
18. Preprint

    1. Courtier-Orgogozo V
    2. Arnoult L
    3. Prigent SR
    4. Wiltgen S
    5. Martin A

    (2019) Gephebase, a database of Genotype-Phenotype relationships for natural and domesticated variation in eukaryotes

    bioRxiv.

    https://doi.org/10.1101/618371

    • Google Scholar
19. 1. Darriba D
    2. Taboada GL
    3. Doallo R
    4. Posada D

    (2012) jModelTest 2: more models, new heuristics and parallel computing - Supplementary information

    Nature Methods 9:772.

    https://doi.org/10.1038/nmeth.2109

    • Google Scholar
20. 1. Durand E
    2. Méheust R
    3. Soucaze M
    4. Goubet PM
    5. Gallina S
    6. Poux C
    7. Fobis-Loisy I
    8. Guillon E
    9. Gaude T
    10. Sarazin A
    11. Figeac M
    12. Prat E
    13. Marande W
    14. Bergès H
    15. Vekemans X
    16. Billiard S
    17. Castric V

    (2014) Dominance hierarchy arising from the evolution of a complex small RNA regulatory network

    Science 346:1200–1205.

    https://doi.org/10.1126/science.1259442

    • PubMed
    • Google Scholar
21. 1. Dwyer KG
    2. Berger MT
    3. Ahmed R
    4. Hritzo MK
    5. McCulloch AA
    6. Price MJ
    7. Serniak NJ
    8. Walsh LT
    9. Nasrallah JB
    10. Nasrallah ME

    (2013) Molecular characterization and evolution of self-incompatibility genes in Arabidopsis thaliana: the case of the sc haplotype

    Genetics 193:985–994.

    https://doi.org/10.1534/genetics.112.146787

    • PubMed
    • Google Scholar
22. 1. Felsenstein J

    (1985) Confidence limits on phylogenies: an approach using the bootstrap

    Evolution 39:783–791.

    https://doi.org/10.1111/j.1558-5646.1985.tb00420.x

    • Google Scholar
23. 1. Fiser A
    2. Do RK
    3. Sali A

    (2000) Modeling of loops in protein structures

    Protein Science 9:1753–1773.

    https://doi.org/10.1110/ps.9.9.1753

    • PubMed
    • Google Scholar
24. 1. Gervais CE
    2. Castric V
    3. Ressayre A
    4. Billiard S

    (2011) Origin and diversification dynamics of self-incompatibility haplotypes

    Genetics 188:625–636.

    https://doi.org/10.1534/genetics.111.127399

    • PubMed
    • Google Scholar
25. 1. Goldman N
    2. Yang Z

    (1994) A codon-based model of nucleotide substitution for protein-coding DNA sequences

    Molecular Biology and Evolution 11:725–736.

    https://doi.org/10.1093/oxfordjournals.molbev.a040153

    • PubMed
    • Google Scholar
26. 1. Goubet PM
    2. Bergès H
    3. Bellec A
    4. Prat E
    5. Helmstetter N
    6. Mangenot S
    7. Gallina S
    8. Holl AC
    9. Fobis-Loisy I
    10. Vekemans X
    11. Castric V

    (2012) Contrasted patterns of molecular evolution in dominant and recessive self-incompatibility haplotypes in Arabidopsis

    PLOS Genetics 8:e1002495.

    https://doi.org/10.1371/journal.pgen.1002495

    • PubMed
    • Google Scholar
27. 1. Gronemeyer H
    2. Bourguet W

    (2009) Allosteric effects govern nuclear receptor action: dna appears as a player

    Science Signaling 2:pe34.

    https://doi.org/10.1126/scisignal.273pe34

    • PubMed
    • Google Scholar
28. 1. Guindon S
    2. Dufayard JF
    3. Lefort V
    4. Anisimova M
    5. Hordijk W
    6. Gascuel O

    (2010) New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0

    Systematic Biology 59:307–321.

    https://doi.org/10.1093/sysbio/syq010

    • PubMed
    • Google Scholar
29. 1. Gutierrez-Mazariegos J
    2. Nadendla EK
    3. Studer RA
    4. Alvarez S
    5. de Lera AR
    6. Kuraku S
    7. Bourguet W
    8. Schubert M
    9. Laudet V

    (2016) Evolutionary diversification of retinoic acid receptor ligand-binding pocket structure by molecular tinkering

    Royal Society Open Science 3:150484.

    https://doi.org/10.1098/rsos.150484

    • PubMed
    • Google Scholar
30. 1. Heo L
    2. Lee H
    3. Seok C

    (2016) GalaxyRefineComplex: refinement of protein-protein complex model structures driven by interface repacking

    Scientific Reports 6:32153.

    https://doi.org/10.1038/srep32153

    • PubMed
    • Google Scholar
31. 1. Hohmann N
    2. Wolf EM
    3. Lysak MA
    4. Koch MA

    (2015) A Time-Calibrated road map of Brassicaceae species radiation and evolutionary history

    The Plant Cell tpc.15.00482.

    https://doi.org/10.1105/tpc.15.00482

    • Google Scholar
32. 1. Kachroo A
    2. Schopfer CR
    3. Nasrallah ME
    4. Nasrallah JB

    (2001) Allele-specific receptor-ligand interactions in Brassica self-incompatibility

    Science 293:1824–1826.

    https://doi.org/10.1126/science.1062509

    • PubMed
    • Google Scholar
33. 1. Kang HJ
    2. Wilkins AD
    3. Lichtarge O
    4. Wensel TG

    (2015) Determinants of endogenous ligand specificity divergence among metabotropic glutamate receptors

    Journal of Biological Chemistry 290:2870–2878.

    https://doi.org/10.1074/jbc.M114.622233

    • PubMed
    • Google Scholar
34. 1. Kitashiba H
    2. Nasrallah JB

    (2014) Self-incompatibility in Brassicaceae crops: lessons for interspecific incompatibility

    Breeding Science 64:23–37.

    https://doi.org/10.1270/jsbbs.64.23

    • PubMed
    • Google Scholar
35. 1. Kusaba M
    2. Nishio T
    3. Satta Y
    4. Hinata K
    5. Ockendon D

    (1997) Striking sequence similarity in inter- and intra-specific comparisons of class I SLG alleles from Brassica oleracea and Brassica campestris: implications for the evolution and recognition mechanism

    PNAS 94:7673–7678.

    https://doi.org/10.1073/pnas.94.14.7673

    • PubMed
    • Google Scholar
36. 1. Laub MT
    2. Goulian M

    (2007) Specificity in Two-Component signal transduction pathways

    Annual Review of Genetics 41:121–145.

    https://doi.org/10.1146/annurev.genet.41.042007.170548

    • Google Scholar
37. 1. Lawrence M

    (2000) Population genetics of the homomorphic Self-incompatibility polymorphisms in flowering plants

    Annals of Botany 85:221–226.

    https://doi.org/10.1006/anbo.1999.1044

    • Google Scholar
38. 1. Lensink MF
    2. Velankar S
    3. Wodak SJ

    (2017) Modeling protein-protein and protein-peptide complexes: capri 6th edition

    Proteins: Structure, Function, and Bioinformatics 85:359–377.

    https://doi.org/10.1002/prot.25215

    • PubMed
    • Google Scholar
39. 1. Liu P
    2. Sherman-Broyles S
    3. Nasrallah ME
    4. Nasrallah JB

    (2007) A cryptic modifier causing transient self-incompatibility in Arabidopsis thaliana

    Current Biology 17:734–740.

    https://doi.org/10.1016/j.cub.2007.03.022

    • PubMed
    • Google Scholar
40. 1. Logemann E
    2. Birkenbihl RP
    3. Ülker B
    4. Somssich IE

    (2006) An improved method for preparing Agrobacterium cells that simplifies the Arabidopsis transformation protocol

    Plant Methods 2:16.

    https://doi.org/10.1186/1746-4811-2-16

    • PubMed
    • Google Scholar
41. 1. Ma B
    2. Tsai CJ
    3. Pan Y
    4. Nussinov R

    (2010) Why does binding of proteins to DNA or proteins to proteins not necessarily spell function?

    ACS Chemical Biology 5:265–272.

    https://doi.org/10.1021/cb900293a

    • PubMed
    • Google Scholar
42. 1. Ma R
    2. Han Z
    3. Hu Z
    4. Lin G
    5. Gong X
    6. Zhang H
    7. Nasrallah JB
    8. Chai J

    (2016) Structural basis for specific self-incompatibility response in Brassica

    Cell Research 26:1320–1329.

    https://doi.org/10.1038/cr.2016.129

    • PubMed
    • Google Scholar
43. 1. Matsumura I
    2. Ellington AD

    (2001) In vitro evolution of beta-glucuronidase into a beta-galactosidase proceeds through non-specific intermediates

    Journal of Molecular Biology 305:331–339.

    https://doi.org/10.1006/jmbi.2000.4259

    • PubMed
    • Google Scholar
44. 1. Matton DP
    2. Luu DT
    3. Xike Q
    4. Laublin G
    5. O'Brien M
    6. Maes O
    7. Morse D
    8. Cappadocia M

    (1999) Production of an S RNase with dual specificity suggests a novel hypothesis for the generation of new S alleles

    The Plant Cell 11:2087–2097.

    https://doi.org/10.1105/tpc.11.11.2087

    • PubMed
    • Google Scholar
45. 1. Matton DP
    2. Luu DT
    3. Morse D
    4. Cappadocia M

    (2000) Reply: establishing a paradigm for the generation of new S alleles

    The Plant Cell 12:313–315.

    https://doi.org/10.1105/tpc.12.3.313

    • PubMed
    • Google Scholar
46. 1. Nasrallah ME
    2. Liu P
    3. Nasrallah JB

    (2002) Generation of self-incompatible Arabidopsis thaliana by transfer of two S locus genes from A. lyrata

    Science 297:247–249.

    https://doi.org/10.1126/science.1072205

    • PubMed
    • Google Scholar
47. 1. Nasrallah JB
    2. Nasrallah ME

    (1993) Pollen-Stigma Signaling in the Sporophytic Self-Incompatibility Response

    The Plant Cell 5:1325–1335.

    https://doi.org/10.2307/3869785

    • Google Scholar
48. Book

    1. Nettancourt DD

    (1977) Incompatibility in Angiosperms

    Springer Verlag.

    https://doi.org/10.1007/978-3-662-12051-4

    • Google Scholar
49. 1. Nooren IM
    2. Thornton JM

    (2003) Diversity of protein-protein interactions

    The EMBO Journal 22:3486–3492.

    https://doi.org/10.1093/emboj/cdg359

    • PubMed
    • Google Scholar
50. 1. Plach MG
    2. Semmelmann F
    3. Busch F
    4. Busch M
    5. Heizinger L
    6. Wysocki VH
    7. Merkl R
    8. Sterner R

    (2017) Evolutionary diversification of protein-protein interactions by interface add-ons

    PNAS 114:E8333–E8342.

    https://doi.org/10.1073/pnas.1707335114

    • PubMed
    • Google Scholar
51. 1. Pougach K
    2. Voet A
    3. Kondrashov FA
    4. Voordeckers K
    5. Christiaens JF
    6. Baying B
    7. Benes V
    8. Sakai R
    9. Aerts J
    10. Zhu B
    11. Van Dijck P
    12. Verstrepen KJ

    (2014) Duplication of a promiscuous transcription factor drives the emergence of a new regulatory network

    Nature Communications 5:4868.

    https://doi.org/10.1038/ncomms5868

    • PubMed
    • Google Scholar
52. 1. Ranwez V
    2. Harispe S
    3. Delsuc F
    4. Douzery EJ

    (2011) MACSE: multiple alignment of coding SEquences accounting for frameshifts and stop codons

    PLOS ONE 6:e22594.

    https://doi.org/10.1371/journal.pone.0022594

    • PubMed
    • Google Scholar
53. 1. Ronquist F
    2. Huelsenbeck JP

    (2003) MrBayes 3: bayesian phylogenetic inference under mixed models

    Bioinformatics 19:1572–1574.

    https://doi.org/10.1093/bioinformatics/btg180

    • PubMed
    • Google Scholar
54. 1. Sali A
    2. Blundell TL

    (1993) Comparative protein modelling by satisfaction of spatial restraints

    Journal of Molecular Biology 234:779–815.

    https://doi.org/10.1006/jmbi.1993.1626

    • PubMed
    • Google Scholar
55. 1. Sayou C
    2. Monniaux M
    3. Nanao MH
    4. Moyroud E
    5. Brockington SF
    6. Thévenon E
    7. Chahtane H
    8. Warthmann N
    9. Melkonian M
    10. Zhang Y
    11. Wong GK
    12. Weigel D
    13. Parcy F
    14. Dumas R

    (2014) A promiscuous intermediate underlies the evolution of LEAFY DNA binding specificity

    Science 343:645–648.

    https://doi.org/10.1126/science.1248229

    • PubMed
    • Google Scholar
56. 1. Schopfer CR
    2. Nasrallah ME
    3. Nasrallah JB

    (1999) The male determinant of self-incompatibility in Brassica

    Science 286:1697–1700.

    https://doi.org/10.1126/science.286.5445.1697

    • PubMed
    • Google Scholar
57. 1. Shimizu KK
    2. Tsuchimatsu T

    (2015) Evolution of selfing: recurrent patterns in molecular adaptation

    Annual Review of Ecology, Evolution, and Systematics 46:593–622.

    https://doi.org/10.1146/annurev-ecolsys-112414-054249

    • Google Scholar
58. 1. Siddiq MA
    2. Loehlin DW
    3. Montooth KL
    4. Thornton JW

    (2017) Experimental test and refutation of a classic case of molecular adaptation in Drosophila Melanogaster

    Nature Ecology & Evolution 1:0025.

    https://doi.org/10.1038/s41559-016-0025

    • Google Scholar
59. 1. Smyth DR
    2. Bowman JL
    3. Meyerowitz EM

    (1990) Early flower development in Arabidopsis

    The Plant Cell 2:755–767.

    https://doi.org/10.1105/tpc.2.8.755

    • PubMed
    • Google Scholar
60. 1. Starr TN
    2. Picton LK
    3. Thornton JW

    (2017) Alternative evolutionary histories in the sequence space of an ancient protein

    Nature 549:409–413.

    https://doi.org/10.1038/nature23902

    • PubMed
    • Google Scholar
61. 1. Takasaki T
    2. Hatakeyama K
    3. Suzuki G
    4. Watanabe M
    5. Isogai A
    6. Hinata K

    (2000) The S receptor kinase determines self-incompatibility in Brassica stigma

    Nature 403:913–916.

    https://doi.org/10.1038/35002628

    • Google Scholar
62. 1. Takayama S
    2. Shiba H
    3. Iwano M
    4. Asano K
    5. Hara M
    6. Che FS
    7. Watanabe M
    8. Hinata K
    9. Isogai A

    (2000) Isolation and characterization of pollen coat proteins of Brassica campestris that interact with S locus-related glycoprotein 1 involved in pollen-stigma adhesion

    PNAS 97:3765–3770.

    https://doi.org/10.1073/pnas.97.7.3765

    • PubMed
    • Google Scholar
63. 1. Takayama S
    2. Shimosato H
    3. Shiba H
    4. Funato M
    5. Che FS
    6. Watanabe M
    7. Iwano M
    8. Isogai A

    (2001) Direct ligand-receptor complex interaction controls Brassica self-incompatibility

    Nature 413:534–538.

    https://doi.org/10.1038/35097104

    • PubMed
    • Google Scholar
64. 1. Tsuchimatsu T
    2. Suwabe K
    3. Shimizu-Inatsugi R
    4. Isokawa S
    5. Pavlidis P
    6. Städler T
    7. Suzuki G
    8. Takayama S
    9. Watanabe M
    10. Shimizu KK

    (2010) Evolution of self-compatibility in Arabidopsis by a mutation in the male specificity gene

    Nature 464:1342–1346.

    https://doi.org/10.1038/nature08927

    • PubMed
    • Google Scholar
65. 1. Tufts DM
    2. Natarajan C
    3. Revsbech IG
    4. Projecto-Garcia J
    5. Hoffmann FG
    6. Weber RE
    7. Fago A
    8. Moriyama H
    9. Storz JF

    (2015) Epistasis constrains mutational pathways of hemoglobin adaptation in high-altitude pikas

    Molecular Biology and Evolution 32:287–298.

    https://doi.org/10.1093/molbev/msu311

    • PubMed
    • Google Scholar
66. 1. Uyenoyama MK
    2. Zhang Y
    3. Newbigin E

    (2001)

    On the origin of self-incompatibility haplotypes: transition through self-compatible intermediates

    Genetics 157:1805–1817.

    • PubMed
    • Google Scholar
67. 1. Uyenoyama MK
    2. Newbigin E

    (2000) Evolutionary dynamics of dual-specificity self-incompatibility alleles

    The Plant Cell 12:310–312.

    https://doi.org/10.1105/tpc.12.3.310

    • PubMed
    • Google Scholar
68. 1. Yang Z
    2. Nielsen R
    3. Goldman N
    4. Pedersen AM

    (2000)

    Codonsubstitution models for heterogeneous selection pressure at amino acid sites

    Genetics 155:431–449.

    • PubMed
    • Google Scholar
69. 1. Yang Z

    (2007) PAML 4: phylogenetic analysis by maximum likelihood

    Molecular Biology and Evolution 24:1586–1591.

    https://doi.org/10.1093/molbev/msm088

    • PubMed
    • Google Scholar
70. 1. Yang Z
    2. Nielsen R

    (2008) Mutation-selection models of Codon substitution and their use to estimate selective strengths on Codon usage

    Molecular Biology and Evolution 25:568–579.

    https://doi.org/10.1093/molbev/msm284

    • PubMed
    • Google Scholar
71. 1. Zhang X
    2. Henriques R
    3. Lin SS
    4. Niu QW
    5. Chua NH

    (2006) Agrobacterium-mediated transformation of Arabidopsis thaliana using the floral dip method

    Nature Protocols 1:641–646.

    https://doi.org/10.1038/nprot.2006.97

    • PubMed
    • Google Scholar

Acceptance summary:

Self-incompatibility in the Brassicaceae is mediated by the S locus, which encodes two interacting proteins, a female-expressed S-locus Receptor Kinase and a male-expressed ligand, S-Locus Cysteine-rich Protein (SCR). Pollen is rejected when SRK and SCR are encoded by the same haplotype. Highly diverged allelic variants segregate in natural populations, but how this allelic diversity arose from an ancestral haplotype is unresolved. The authors use ancestral reconstruction to investigate how different S haplotypes might have arisen – which is interesting in itself but also as a model for the general question of how genes diversify despite constraints due to specific interactions. A main strength of this work is that four plausible ancestral alleles are reconstructed (and replicated). They then tested 3 proposed models, by introducing various constructs of A. halleri SRK into A. thaliana. Their results provide a convincing answer: favoring compensatory mutations and SC intermediates, and rejecting 2 other proposed models.

Decision letter after peer review:

Thank you for submitting your article "Asymmetrical diversification of the receptor-ligand interaction controlling self-incompatibility in Arabidopsis" for consideration by eLife. Your article has been reviewed by three peer reviewers, including Sheila McCormick as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by a Reviewing Editor and Christian Hardtke as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Nick H Barton (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

This impressive paper sets out to distinguish how different S haplotypes have arisen. The authors use ancestral reconstruction to investigate S locus evolution – which is interesting in itself but also, as the authors explain, a model for the general question of how genes diversify despite constraints due to specific interactions. A key issue in such studies is the reliability of the reconstruction. However, four plausible ancestral alleles are reconstructed (and replicated), which is a strength.

There are 3 ideas in the field, and to test them the authors introduce various constructs of A. halleri SRK into A. thaliana. Their results provide a convincing answer (i.e. they favor the one of compensatory mutations and SC intermediates (Gervais).

Essential revisions:

The experiments are easy to follow and the discussion is thoughtful, but could be shortened by removing unnecessary repeats. For instance, the first paragraph just repeats what has already been stated perfectly clearly. Discussion subsection “Asymmetrical structural and functional divergence” can also be shortened considerably, and these shortenings will help readers come to the most important questions that they will have in mind at this point – i.e., which of the models outlined in the Introduction are excluded. Given the value of the work, it is worthwhile to make the text as clear as possible.

This section needs slight revision, to something like the following: "Overall, our results thus unambiguously reject two previously proposed models of SI specificity diversification. The model of gradual divergence of pairwise SRK-SCR affinities along lineages (Chookajorn et al., 2004) is rejected because it predicts that both descendent alleles would have been functionally distinct from the ancestor. The model of promiscuous dual-specificity intermediates (Matton et al., 1999) is also rejected, because Uyenoyama and Newbigin, 2000, have shown that a dual-specificity haplotype can be maintained in the population only in the absence of the ancestral specificity, while we show here that the ancestral specificity (S03) remained present. In contrast, the remaining scenario (in which diversification proceeds through transient self-compatible intermediates; Gervais et al., 2011; Uyenoyama et al., 2001) is compatible with our observation of long-term maintenance of functional specificities".

Please consider stating putative ancestral throughout or "ancestral", as you only surmise that this could be the scenario. Similarly, the terms "resurrected" and "now extinct" are suppositions.

In later experiments they analyze potential amino acid contacts in the synthesized SRKs and SCRs. Using the number of contacts seems somewhat crude; are there other cases that can be cited where this sort of metric has succeeded?

Subsection “Asymmetrical structural and functional divergence”: – One cannot conclude that specificities evolve through cladogenetic change rather than anagenetic. The latter refers to the question of whether changes are associated with speciation events. Here, the observation is that function has been preserved along one lineage but not the other. Does this really address whether speciation is associated with functional divergence?

Discussion section: It may be that "the specificity of the receptor/ligand interaction depends on qualitative features rather than quantitative", but that seems more general than " allosteric activation", which refers to a change in specificity caused by a change in the shape of the protein. If the argument is that allostery is involved, it needs to be made more clearly.

Please give more information (i.e. add a figure) about the amino acid sequence differences, including the total number of differences between the SRK sequences used in the study, and their locations in relation to the recognized functional domains of the protein. Subsection “Phylogeny-based ancestral SRK reconstruction” mentions "the vast majority of amino acid residues where SRK03 and SRK28 differ", but doesn't tell us the actual number or any other information. Subsection “Ligand specificity of the ancestral SRK” says that the study examined the S-domain, but the domains have not been clearly outlined or explained. It would also have helped understanding to be told about the differences from the inferred ancestral protein sequence. Without these details, it is difficult to understand the naming of the sequences that were tested. A brief description of the phylogenetic approach used to reconstruct the sequence of SRK03 and SRK28's most recent common ancestor would also be helpful.

Can Figure 1A be revised a bit? The shapes are an over-simplification of the experiments, as there were 4 different versions tested, the figure implies only 1.

Essential revisions:

The experiments are easy to follow and the discussion is thoughtful, but could be shortened by removing unnecessary repeats. For instance, the first paragraph just repeats what has already been stated perfectly clearly.

We have removed this first paragraph.

Discussion subsection “Asymmetrical structural and functional divergence” can also be shortened considerably, and these shortenings will help readers come to the most important questions that they will have in mind at this point – i.e., which of the models outlined in the Introduction are excluded. Given the value of the work, it is worthwhile to make the text as clear as possible.

We have shortened this paragraph by one half and now come more directly to the core questions.

This section needs slight revision, to something like the following: "Overall, our results thus unambiguously reject two previously proposed models of SI specificity diversification. The model of gradual divergence of pairwise SRK-SCR affinities along lineages (Chookajorn et al., 2004) is rejected because it predicts that both descendent alleles would have been functionally distinct from the ancestor. The model of promiscuous dual-specificity intermediates (Matton et al., 1999) is also rejected, because Uyenoyama and Newbigin, 2000, have shown that a dual-specificity haplotype can be maintained in the population only in the absence of the ancestral specificity, while we show here that the ancestral specificity (S03) remained present. In contrast, the remaining scenario (in which diversification proceeds through transient self-compatible intermediates; Gervais et al., 2011; Uyenoyama et al., 2001) is compatible with our observation of long-term maintenance of functional specificities".

We have revised this paragraph (subsection “Asymmetrical structural and functional divergence”) to better clarify which models can be rejected, which led us to expand a bit the paragraph where we present the models in the Introduction. In particular we placed a stronger focus on the persistence of the ancestral specificity after diversification, which is predicted by Gervais et al., 2011 and possibly also by Chookajorn et al., 2004. We note that functional turnover is another possible outcome for both models, but the lack of formal theoretical analysis of the latter prevents more precise distinction at this stage. We feel that this more precise presentation of the models will make the implications of our results more evident to the readers.

Please consider stating putative ancestral throughout or "ancestral", as you only surmise that this could be the scenario. Similarly, the terms "resurrected" and "now extinct" are suppositions.

We have changed the wording to “putative ancestral” at most places in the text.

As requested, we have changed the term “resurrected” to “reconstructed” in the Abstract. At other places in the text we still refer to a “resurrection” experiment, as this wording is now well established in the literature.

We have retained in the Abstract the mention to the fact that the putative ancestor is “now extinct”, as this is indeed reflecting the reality and is not a supposition.

In later experiments they analyze potential amino acid contacts in the synthesized SRKs and SCRs. Using the number of contacts seems somewhat crude; are there other cases that can be cited where this sort of metric has succeeded?

Computational predictions of protein-protein interaction use several quantities to estimate the quality of the predicted model, including the fraction of native contacts. This quantity is purely based on interatomic and/or inter-residue distances criteria, as we have used here and is common practice in the field. We have added a reference to the CAPRI experiment (Lensink et al., 2017) that evaluates the performance of different metrics used to evaluate the quality of protein docking (including the number of atomic contact).

Subsection “Asymmetrical structural and functional divergence”: One cannot conclude that specificities evolve through cladogenetic change rather than anagenetic. The latter refers to the question of whether changes are associated with speciation events. Here, the observation is that function has been preserved along one lineage but not the other. Does this really address whether speciation is associated with functional divergence?

We have removed this sentence.

Discussion section: It may be that "the specificity of the receptor/ligand interaction depends on qualitative features rather than quantitative", but that seems more general than " allosteric activation", which refers to a change in specificity caused by a change in the shape of the protein. If the argument is that allostery is involved, it needs to be made more clearly.

We have modified the sentence to better explain the allosteric activation model.

Please give more information (i.e. add a figure) about the amino acid sequence differences, including the total number of differences between the SRK sequences used in the study, and their locations in relation to the recognized functional domains of the protein. Subsection “Phylogeny-based ancestral SRK reconstruction” mentions "the vast majority of amino acid residues where SRK03 and SRK28 differ", but doesn't tell us the actual number or any other information. Subsection “Ligand specificity of the ancestral SRK” says that the study examined the S-domain, but the domains have not been clearly outlined or explained. It would also have helped understanding to be told about the differences from the inferred ancestral protein sequence.

We have modified Figure 3—figure supplement 1 to make this information more apparent (number of amino acid sequence differences between SRK03 and SRK28 and with SRKa, functional domain annotation of the SRK protein). We also report these numbers directly in the text.

Without these details, it is difficult to understand the naming of the sequences that were tested. A brief description of the phylogenetic approach used to reconstruct the sequence of SRK03 and SRK28's most recent common ancestor would also be helpful.

We now explain directly the naming of the sequences: “This most likely ancestral amino acid sequence had an Isoleucine (I) at position 208 and an Alanine (A) at position 305 and is noted SRKa^IA”.

Can Figure 1A be revised a bit? The shapes are an over-simplification of the experiments, as there were 4 different versions tested, the figure implies only 1.

We have improved this figure to represent the four reconstructed ancestors.

Author details

1. Maxime Chantreau

   CNRS, Univ. Lille, UMR 8198—Evo-Eco-Paléo, F-59000, Lille, France

   Contribution

   Data curation, Formal analysis, Investigation, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2844-1989

2. Céline Poux

   CNRS, Univ. Lille, UMR 8198—Evo-Eco-Paléo, F-59000, Lille, France

   Contribution

   Data curation, Software, Formal analysis, Investigation, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9379-2769

3. Marc F Lensink

   Univ. Lille, CNRS, UMR 8576 - UGSF - Unité de Glycobiologie Structurale et Fonctionnelle, F-59000, Lille, France

   Contribution

   Software, Formal analysis, Investigation, Methodology

   Competing interests

   No competing interests declared

4. Guillaume Brysbaert

   Univ. Lille, CNRS, UMR 8576 - UGSF - Unité de Glycobiologie Structurale et Fonctionnelle, F-59000, Lille, France

   Contribution

   Software, Formal analysis, Investigation, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-6807-6621

5. Xavier Vekemans

   CNRS, Univ. Lille, UMR 8198—Evo-Eco-Paléo, F-59000, Lille, France

   Contribution

   Supervision

   Competing interests

   No competing interests declared

6. Vincent Castric

   CNRS, Univ. Lille, UMR 8198—Evo-Eco-Paléo, F-59000, Lille, France

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Validation, Investigation, Project administration

   For correspondence

   vincent.castric@univ-lille.fr

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-4461-4915

• 1,785

  Page views

• 245

  Downloads

• 9

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.