evolution of morphological diversity in Heliconius

Question 1

heliconius butterfly properties

Answer 1

larvae eat cyanide containing passion vines
sequester it to become toxic
adults remain toxic
patterned brightly as warning colouration

Question 2

heliconius phylogeny

Answer 2

-whole divesity originated from common ancestor w 2 major branches

-most species genetically polymorphic (not polyphenism) - various pattern forms are due to difference in genetics

-same basic form of pattern occurs again and againin different heliconius species
variation is geographic - same pattern found in different species in the same region of S. America
different patterns in same species across region borders

Question 3

reason for geographical districutions of heliconius patterns

Answer 3

advantage to looking the same as the other butterflies in that region
erato in same areas look like melpomene in same areas e.g.

because naive predators need to sample heliconius individuals before associating bad taste with pattern
more of same pattern in same area = less likely for individual to be sampled
so advantageous to look like others in area
MULLERIAN MIMICRY

hence different species having same pattern in same area
but same species in different areas look different

Question 4

testing advantage of mullerian mimicry of other butterflies in same area

Answer 4

H. melpomene
H. erato

one side of area border both species dennis ray
other side of border - both postman pattern

moved individuals
different location on same side (control)
different location on same side

individuals moved to same side did a lot better than ones of same species moves to other side

benifitted from learning of predators on same side to avoid their pattern
didnt get sampled/injured as much

Question 5

genetic control of patterning across H species?

Answer 5

found various genes
different versions of these determine phenotype patterning

number of unlinked loci separated in chromosome or on different ones
different combinations of which (alleles?) were producing different wing patterns

matching seen in different species in same area confers protection

Question 6

hybridisation of different patterns in same species effect

Answer 6

hybridise postman and dennis ray H. melpomene
genes controlling patterning are unlinked
so messed up mix of alleles in progeny
end up with messed up pattern
no protection in either region as pattern matches neither

Question 7

H. Cydno morphs

Answer 7

black/white or balck/yellow
controlled by one gene
coexist w diff species

H. eleuchia in Cydno yellow morph area - resembles yellow cydno
H. sapho and white cydno same area resemble - other side of border to yellow

white did much better than yellow when moved in sapho region
other way round for moving into eleuchia region

sapho and eleuchia are much more abundant than cydno in either region
in eleuchia site
released yellow do better than white at low density of released
-however if you release 4x higher density of white and yellow cydno:
not mcuh difference
so many more butterflies that dont match others so predators learn much quicker to avoid them - much less detrimental effect on non matchers

Question 8

selection against hybridisation of border “races”

Answer 8

two same species individuals of different pattern at border mate
get hybrid offspring that dont match anyone else
more likely to be predated
waste of reproductive resources

expect pattern races to preferentially not mate with each other to develop

Question 9

mate preference tests for melpomene

Answer 9

look at mate preference in males for different forms of females

postman melpomene much more likely to mate similarly appearing females - no interest in other morphs with different wing patterns

cydno same result - white cydno went for white cydno - ignored others

however - banded melpomene from part far away from where cydno and rosina melpomene - had much less preference (none?? idk)

strong mating preference evolved in region where many other morphs were nearby

strong preference males had same preference for dead females and even paper models

so appearance of butterflies are MOST important cue (not pheromones, behaviour…)

Question 10

Heliconius heurippa - hybridisation?

Answer 10

purple region relatively small

find distinctive heliconius eurippa species

red and yellow band

other butterflies in area (heurippa region within red melpomene region and right next to yellow cydno region)

heurippa looks like simply putting together melpomene and cydno patterns

but is separate species (closely related but separate) by classical identification

molecular analysis shows that it shares a lot of molecular variation with two other species (particularly cydno)

heurippa arose by hybridisation between melpomene and cydno???

could this hybridisation give this pattern form

Question 11

replication of heurippa pattern morph by hybridisation of cydno and melpomene morphs

Answer 11

bred together cydno and melpomene

found that male offspring of F1 were viable and fertile

BUT females were not (viable but sterile)

didnt resemble heurippa

BUT backcrossed them with cydno for many generations - gave many differently appearing butterfly wing patterns

HOWEVER -produced some offspring that looked like the wild Heurippa

• were also true breeding - meaning that the allelic combination that gives this pattern is likely homozygous, and interbreeding these hybrids would maintain the heurippa like pattern

so -captured wild heurippa males

test mating preference like last one

took heurippa males and challenged them with heurippa, melpomene, cydno, OR heurippa that had been blacked out (had either red or yellow bar vlacked out)

heurippa males preferred only the heurippa females with the normal double bar pattern (not even the blacked out ones as behaviour is not really main cue - own wing pattern preference is main cue)

this may have taken heurippa 1000s of years in the wild to evolve this mating preference

do the experimental hybridisation generated Heurippa like morphs have preference?

Question 12

novel heurippa hybrid preferences?

Answer 12

just generated from hybridisation
do pattern preferences need to develop over long time ??

heurippa males generated already had a preference for own wing form compared to other wing forms you can challenge them with

• if this novel genotype WAS actually generated in wild by these hybridisation
  due to the fact it is true breeding pattern - this instantaneous mating preference will have kept the pattern allowing novel heurippa species to diverge
• how are the patterns and prefernces so intrinsically linkes?????

Question 13

refugia theory of heliconius pattern race regions

Answer 13

in central and southern america

with advance of glaciers from north

whole region became cooler and drier

could’ve had conseqeuences for butterflies inhabiting this region

if they were not so diverse back then

diversity may have arose in glecial refugia

most of land became uninhabitable to butterflies and they were restricted to small isolated restricted habitable regions - refugia

if kept that way for long while - refugia populations would diverge from each other

small pops - different mutations - high probability in small loction that they are fixed in pop - different selection pressures in different refugia

if you have different speceis (melpomene, erato e.g.) in same refugia - even if they look different, they will convergeb on same pattern (resembling others = good)

conditions get better

populations spread out

boundaries form between pattern morphs form

predation selection, and reproductive preference isolation keeps that border distinct

happens all over gives rise to patchwork of different colour races

though not a lot of direct evidence

theory sounds nice but trying to locate what it was like in past and where refugia were is hard

Question 14

refugia allopatric pattern development example problems

Answer 14

if allopatric like refugia example: would expect:

different populations do have genetic differences that match the geographical differences (populations in same areas have same differences to others)

due to allopatric divergence

so would expect phylogenies in pattern races in different species in different areas would line up - as they evolved allopatrically together

results - counter to refugia hypithesis

1. pattern races are not real races - are not genetically distinct from otehr pattern races within a species
   if different patterns cam from different refugia would expect distinct genetically
   different pattern races very genetically similar to nearby of same species
   phylogeny reflects geographical difference much more than pattern differences
   i.e. different same species patterns near nearby across border - similar
   whereas same pattern same species across other side of area is more different
   some species would have radiated out through regions occupied by ohter butterflies - and would change pattern to match those in that area
   difference would be much younger than refugia differences too

so don’t have exact answer for how they diverged

Question 15

breeding experiments to identify genes that control patterning on heliconius

Answer 15

D/R melpomene locus - controls red patterning elements

D in erato - controls red patterning elements

are these actually homologous genes doing the similat jobs in different species

linkage mapping - cross experiments and looking at recombination freuquency with specific markers

can show whether particular gene is long way from marker (gets separated)

or if it is at or very near same site (won’t recombine away in reasonable timescale)

D/R and D loci correspond very closely in location on its chromosome

managed to approximate location on chromosome down to stretch of DNA

small portion of whole genome

but still long stretch of DNA that contains many genes

postman - red in mid wing

dennis ray - not red in mid wing

took pupal stage wings and cut out mid area of wings

extracted RNA

difference in gene expression in that tissue

RNA from optix gene region of genome present in one erato pattern morph and not in the other

fluorescentyl tagged the optix gene with reporter

get OPTIX in midwing of postman form forewing - the red part

in dennis ray form - it is in closer to inside of wing and hindwing stripes

the patterning differnce is REGULATORY

not difference in protien

differences in enhancers not protein coding sequence

coding region sequence has no significant difference between postman optix allele and dennis ray optix allele

regulatory difference

optix expressed in different locations

Question 16

optix gene action

Answer 16

OPTIX is a TF - where OPTIX is expressed - red pigment later on will be produced

the optix gene of that particular species is expressed where red will be

causal relationship - turning on optix → red pigment in area

shows in hybrids too

gene editing - inactivated optix by mutating coding region → no red where there should be - proof of causal relationship

differences in optix enhancers - different expression patterns of optix in wings

→ different pattern in different pattern morphs

note// optic role in colour patterning is restrictes only to HELICONIUS butterflies

optix TF has other roles in different places - optix has role in drosophila eye development

in heliconius optix has been recruited to additional eing patterning role

(similar to crystallins in eye development)

Question 17

wrong theory for dennis ray pattern area spread

Answer 17

(The wrong part is the butterflies themselves spreading out - butterfly phylogeny showing geographical closeness is more important for relation than pattern

orange region - occupied by very similar pattern forms - show slight variations in dennis ray form

yellow - same situation but w/ postman

grey - same w simple banded form

big contiguous area of D ray in middle

large disjunct regions around periphery that express other pattern

suggestion - orange regions contain butterflies closely related to each ohter that spread to form orange region

this didnt happen - molecular phylogeny - great distance butterflies of one species less related than closer ones - nearer ones closer even if different pattern

Question 18

looking at specifically Optic gene phylogeny

Answer 18

looked specifically at optix sequence phylogeny

sequence diversity - optix genes more similar to each other - if compare dennis ray populations than if comparing banded or postman

suggests that dennis ray version of optix originated much more recently as it is still much more similar

so dennis ray patterns have v close optix to each other

however if you look at neutral genes in rest of genome
in species - neutral gene similarity much more between proximal individuals than further same pattern

so dennis ray on opposite sides of region

• similar optix
• more different neutral genes (where geography is more important)

Question 19

How did erato optix gene spread out to make D ray region?

Answer 19

in erato:

dennis ray optix gene newer origin

spread out by gene flow

selective advantage to hybrid progeny to having dennis ray phenotype

so spread out further through gene flow

dennis ray optix was “superimposed” onto other pattern phenotypes

D ray optix gene newer origin
reason why butterflies of similar pattern on opposite sides of D ray region have:
-more different neutral genes
-more similar Optix genes

Question 20

denis ray Optix control of patterning

Answer 20

some species pattern (of melpomene and timareta) morphs are not Dennis ray despite being in dennis ray area

top one (melpomene example) has one part of dennis ray - has forewing red pattern - but not rays on hindwing just proximal part

bottom (H timareta) has hindwing rays but not dennnis front part

so aspects of the dennis ray pattern can be separated?

due to enhancer structures in dennis ray optix

Question 21

dennis ray enhancer stuff

Answer 21

looked at what molecular marker differences they have vs what red patterning they have

association for having particular differences at left enhancer and having particular “dennis” components of the patter

association for having particular differences at right enhancer and having particular “ray” components of the patter

so these separate parts of genome control whether individual will have dennis and ray components separately

both function by regulating optix coding region next to them

these enhancers are much more similar between similar patterns in different related species than expected

enhancer diversity may have moved through these species by “enhancer flow” (like the gene floe of dennis ray Optix) (hybrid between two close species - introduces dennis ray in that hybrid from dennis ray parent species, dennis ray patterning confers some advantage in progeny, these dennis and ray enhancers are retained through introgression - hybrid backcrossing with non DR parent species except selected to retain D and R enhancers - why they are similar between diff species)

basically through introgression of these dennis and ray regulatory elements in hybrids of these species as they conferred some advantage

this introgression can happen as they are close species so can interbreed

Question 22

dennis and ray enhancer phylogenies

Answer 22

separate phylogenies for these separate enhancers

dennis enhancer region of optix - dennis forms of melpomene, timareta - all grouped with H elevatus (their enhancers are similar)

ray enhancer refion of optix phylogeny:

melpomene and timareta both group with elevatus

dennis and ray enhancers have separate origins in different species as ray forms have more similar ray enhancers - same for dennis

Question 23

more recent research in heliconius patterning stuff

Answer 23

dont think too important

• other enhancer regions interact w dennin=s and ray regions
• enhancer function - D mutants not expressing - no red
• very similar regulatory enhancers in Erato to Melpomene even though they are very dissimilar in ways where hybrids cant form (genetics, mating…) - so no introgression couldve taken place to give them that similarity
  these species diverged 12mya - before the dennis/ray enhancers arose
  may be down to independent but similar changes to ancient enhancers
• what TFs are the dennis and ray enhancers binding - theyre being activated by cues in certain spatial regions but not others - positional cues on wing binding dennis and ray
  -potentially could be by other genes linked to wing patterning (the ones found in the earlier stuff) that are binding to these enhancers
  WNT-A - WNT signalling has roles all over development
  in larva - wing disc stage
  in metamorphosis - pupal wing on pupal surface - pigmented in adult
  red pigment on banded wing
  go ealier - optic expression on that part
  earlier to larval disc - WNT - A completely absent from this region - proven to be causal
  earliest signal for this wing patterning
  so optix repressed by WNT-A signal outside region
  allowed on inside region so get red

Question 24

WNT patterning in heliconius

Answer 24

WNT-A pattern distinguishes the Optix expression borders in melpomene

in erato - WNT-A expressed in proximal part to pattern proximal border but not on distal part of wing so something else does it - similarities and differences in species - KO of WNT-A in erato - still get distal border of red

WNT-A and these WNT-A enhancers pattern in ither distant butterfly species - pattern different things though

Question 25

how could different pattern phenotypes have arisen?

Answer 25

how particular phenotypes may have arisen
like the heurippa example - hybridisation
or dennis ray phenotype spreading out

to establish and maintain new phenotype:

establish - relaxed predation needed

enough individuals can appear for predators to learn to avoid that pattern

spread - needs to confer some other advantage to be that phenotype (e.g. fot that dennis ray version of optix to gene flow across amazon)

• could actually be by adopting phenotype - those butterflies get protection as they are adopting new phenotype that is already present in high frequency in that region (e.g. melpomene dennis ray phenotype spread out later - mimicking other dennis ray that had already spread)
• though this mimicking one doesn’t really work with ^^
  if it is really a novel pattern - what underpinned its ability to spread
  could be more visible
• banded forms found in brighter drier regions, dennis ray forms in darker wetter
  so different forms have different visibility - better visibility = easier learning to avoid
  however there are exceptions that disrupt this
  could be many possible aspects