DNA-barcoding on a different scale: challenges within
Coccoidea
Lyn Cook, The University of Queensland, Australia
SIBIScale Insect Barcode Initiative
(Stellenbosch, Feb 2007)
Outline
• What are scale insects?
• Why have they been a challenge?
• Prospects for a universal scale
insect barcode
APHIDS SCALE INSECTS
10 MY210 - 160 Myr ago
1st scale insect “barcode”• 1995 - DNA extracted from new
morphotype. Some ovarian tissue
frozen at -70C. Cuticle slide mounted as
voucher. Other specimens from same
collection frozen or stored in 100%
ethanol
• 1999 - PCR and sequencing
• 2003 - paper published including
sequences and stored DNA as part of
holotype material
Apiomorpha gullanae
A. munita
A. munita18S rDNA (nuclear SSU rRNA)
18S variation within a morphospecies
PCR product for 5’ end of nuclear ssu rDNA (18S)
360 bp
480 bp
720 bp
980 bp1160 bp
*single genus
18S helix E10-1
18S NJ K2P+G
Cryptococcus
COI 5’
barcode
COI 3’ tRNA numts
*mealybugs few many + yes?
*armored scales
no many +/-
*soft scales no some
felt scales few some +/- yes?
*archaeococcoids
few few yes?
*stictococcoids no no
other scales no some* major pest groups
results of COI barcoding attempts
Why such low success with COI?
AT A C G T
COI 69-85 23-42 8-17 6-19 28-47
3rds ≤ 97.5
30-58 2-19 0-5 27-54
COII 72-86 28-52 8-23 4-15 23-43
3rds ≤ 100 38-65 0-26 0-16 22-55
mtDNA nucleotide frequencies (%)
COI barcode region
• Taxa that amplify appear to have higher CG
in 3rd positions than failures– not direct comparison
• Amino acid divergence of up to 8% in aphid
crown
• Amino acid divergence of up to 55% within
scale crown
18S var. 28SD2
var.
*mealybugs most many most most
*armored scales
most x most most?
*soft scales most most many
most
felt scales most most many
most
archaeococcoids
most most many
most
*stictococcoids most most
other scales most most yes in many
most
Alternatives
Florida lac scale
(= lobate lac scale)
Case study
Paratachardina lobata
28S
conserved regions only
COI - 3’ region
28S D2 region for lac scales
length of the region varies from 703-1114
28S D2
Aligned using
muscle -no
adjustment
Analysed using NJ
K2P
India
India
Florida
Christmas Is.
Bermuda
&, as of 8 weeks ago, Cuba
Conclusions• Proceed with non-COI barcode - 28S when successful
• 18S might be an alternative in “decision tree”
-18S to assign specimen to group or direct ID
- next step depends on grouping (e.g. COI barcode or COI 3’, 12S etc with group-specific
primers)
• Attempt to find solution to COI barcode problem. Options include:
• more primer design
• mini COI barcode
• cocktails
• whole mt genome
• other?
SIBI, Stellenbosch, February, 2007Participants: BB Normark, W Pieterse, PJ Gullan, I Millar, J Giliomee, D Downie, F Haas, C Eardley, LG Cook, A Rung, D Masinga, A Venter, D Muller, DR Miller, R Hanna
Acknowledgements
CBOLSAFRANETNSFNSF-PEETARCABRS