The complex hybrid origins of the Root Knot Nematodes

COMPLEX HYBRID ORIGINS OF ROOT KNOT NEMATODES

Dave LuntEvolutionary Biology Group, University of Hull

Institute of Evolutionary Biology, University of Edinburgh

Georgios KoutsovoulosMark Blaxter

Sujai Kumar


AcknowledgementsAfrica Gómez, Richard Ennos, Amir Szitenberg, Karim Gharbi, Chris Mitchell, Steve Moss, Tom Powers, Janete Brito, Etienne Danchin, Marian

Thomson & GenePool

FundingNERC, BBSRC, Yorkshire Agricultural Society,

Nuffield Foundation, University of Hull, University of Edinburgh



http://www.slideshare.net/davelunt/presentations

davelunt.net

@davelunt

[email protected]

@EvoHull +EvoHull

+davelunt

http://www.github.com/davelunt

davelunt.net

@davelunt

[email protected]

@EvoHull +EvoHull

+davelunt


THE MELOIDOGYNE RKN SYSTEM

Meloidogyne Root Knot Nematodes• Globally important agricultural pest species

• Enormous plant host range

• parasitise all main crop plants

• ~5% loss of world agriculture

JD Eisenback

RKN juveniles enter root tip

infected uninfectedSEM Meloidogyne female

JD Eisenback


Meloidogyne ReproductionWide variety of reproductive modes in a single genus

• Many species are mitotic parthenogens without chromosome pairs

Asexuals

• Other species are meiotic parthenogens• automixis

• Some species are obligatory outbreeding sexuals with males & females

• amphimixis

Sexuals


Meloidogyne ReproductionWide variety of reproductive modes in a single genus

MELOIDOGYNE HYBRIDIZATION

Hybrid Speciation

Once thought that hybrid speciation was rare and inconsequential in animals

Genome biology is revealing a very different view

MELOIDOGYNE HYBRIDIZATION

Hybrid Speciation in Meloidogyne?

Have the mitotic parthenogen root knot nematodes arisen by interspecific hybridization?

Is M. floridensis the parent of the asexuals?

M. floridensis is found within the phylogenetic diversity of asexual species

It reproduces sexually by automixis

Could it be a parent of the asexual lineages via interspecific hybridization?

MELOIDOGYNE HYBRIDIZATION GENOMICS

M.floridensis M. ???

M. incognitaM. javanica

M. arenaria

x

apomicts

parental species

automict


Meloidogyne comparative genomics

We have sequenced M. floridensis genome and are able to compare to 2 other Meloidogyne genomes published by other groups

M.floridensis M. ???

M. incognitaM. javanica

M. arenaria

x

apomicts

parental species

automict

asexual, hybrid?

sexual, parental?

sexual, outgroup

100MB, 100x coverage, 15.3k protein coding loci


1. look at the within-genome patterns of diversity to determine hybrid nature of genomes

!

2. look at phylogenetic relationships of all genes to study origins and parents


1: Intra-genomic diversity

2: Phylogenomics

Investigated using whole genome sequences and 2 distinct approaches;

Extreme Hybrid Allelic Sequence Divergence1. INTRA-GENOMIC ANALYSES: ALLELIC SEQUENCE DIVERGENCE

1: Intra-genomic diversity look at the within-genome patterns of diversity to determine hybrid nature of genomes

A B A D CAlleles

Taxa

Recent

Ancient

Sexual parental species

Divergence between alleles of parental species

Divergence between hybrid species alleles

A C D

Ancestor of sexual parental

species

Hybridization event


hybrid apomict

hybrid apomict

mitotic

Extreme Hybrid Allelic Sequence Divergence1. INTRA-GENOMIC ANALYSES: ALLELIC SEQUENCE DIVERGENCE

‘Alleles’ (homeologues) may date to the divergence of the parental species which hybridized

A B A D CAlleles

Taxa

Recent

Ancient


Divergence between alleles of parental species

Divergence between hybrid species alleles

A C D

Ancestor of sexual parental

species

Hybridization event


hybrid apomict

hybrid apomict

mitotic

1. INTRA-GENOMIC ANALYSES

Divergence of protein-coding allelesLunt et al arXiv 2013 http://arxiv.org/abs/1306.6163

Coding sequences from each of the three target genomes were compared to the set of genes from the same species The percent identity of the best matching (non-self) coding sequence was calculated, and is plotted as a frequency histogram

Both M. incognita and M. floridensis show evidence of presence of many duplicates, while M. hapla does not

Self identity comparisons

http://arxiv.org/abs/1306.6163



Coding sequences from each of the three target genomes (M. hapla, M. incognita and M. floridensis) were compared to the set of genes from the same species

The percent identity of the best matching (non-self) coding sequence was calculated, and is plotted as a frequency histogram Both M. incognita and M. floridensis show evidence of presence of many duplicates, while M. hapla does not






The percent identity of the best matching (non-self) coding sequence was calculated, and is plotted as a frequency histogram







The percent identity of the best matching (non-self) coding sequence was calculated, and is plotted as a frequency histogram


This is exactly the pattern expected for hybrid genomes




!

look at phylogenetic relationships of all genes to study origins and parents


1: Intra-genomic diversity2: Phylogenomics

2. PHYLOGENOMIC ANALYSES

Hybridization HypothesesLunt et al arXiv 2013 http://arxiv.org/abs/1306.6163

There are very many ways species could hybridize, duplicate genes, lose genes

We have selected a broad range of possibilities informed by prior knowledge

We have tested their predictions phylogenetically

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

A

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

B

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3M

. hap

la

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

C

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

D


��20

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5M

. hap

la

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X ZM

. flo

riden

sis

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

Hybridization hypothesesA B

C D

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5M

. hap

la

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X ZM

. flo

riden

sis

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

(A) Whole genome duplication(s)

��22

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

(B) M. incognita is an

interspecific hybrid with M. floridensis as one

parent

M. h

apla

X Y ZM

. flo

riden

sis

M. i

ncog

nita

X+Y Y+Z

C Scenario 4M

. hap

la

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

(C) M. incognita and M.

floridensis are independent hybrids sharing one parent

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

(D) M. floridensis is a hybrid

and M. incognita is a secondary hybrid

between M. floridensis and a 3rd parent


Testing by PhylogenomicsLunt et al arXiv 2013 http://arxiv.org/abs/1306.6163

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

A

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

B

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3M

. hap

la

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

C

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

D

• Coding sequences from 3 genomes were placed into orthologous groups (InParanoid)

• 4018 ortholog clusters included all 3 species • We retained those with a single copy in the

outgroup M. hapla • Phylogenies of relationships between Mi and

Mf gene copies (RAxML) • Trees were parsed and pooled to represent

frequencies of different relationships


��26

Each tree contains a single M. hapla sequence as outgroup (black square)

Grey square indicates relative

frequency of those

topologies

Trees are pooled within squares into different patterns of relationships

Grid squares represent different numbers of gene copies



M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

A

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

B

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3M

. hap

la

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

C

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

D

• We assess the fit of the tree topologies to our hypotheses

• Five out of seven cluster sets, and 95% of all trees, support hybrid origins for both M. floridensis and M. incognita • ie exclude hypotheses A and B

• Hypothesis C best explains 17 trees • Hypothesis D best explains 1335 trees




M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

A

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

B

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y Y+Z

C Scenario 4

M. h

apla

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

C

M. floridensis is a parental species of “double hybrid” M. incognita with other parent unknownM

. hap

la

X Y ZM

. flo

riden

sis

M. i

ncog

nita

X+Y Y+Z

C Scenario 4M

. hap

la

X Y Z

M. f

lorid

ensi

s

M. i

ncog

nita

X+Y

(X+Y)+Z

D Scenario 5

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X X+Z

B Scenario 3

M. h

apla

X Z

M. f

lorid

ensi

s

M. i

ncog

nita

X Z+Z

A Scenario 1 & 2

X+Y

Hypothesis D


MELOIDOGYNE COMPARATIVE GENOMICS

Meloidogyne hybrid species formation

• Suggestions that hybrid speciation may be common in Meloidogyne

• Do asexual agricultural pathogens have a single (hybrid) origin

• What are the common features of hybrid genome architecture? • Ongoing work...



http://www.slideshare.net/davelunt/presentations

davelunt.net

@davelunt

[email protected]

@EvoHull +EvoHull

+davelunt


davelunt.net

@davelunt

[email protected]

@EvoHull +EvoHull

+davelunt


Education

The complex hybrid origins of the Root Knot Nematodes