Transcript Document

Validation of a novel mutation
screening strategy for Familial
Hypercholesterolaemia:
LIPOchip®, a DNA-array based
system
Jonathan Callaway
Wessex Regional Genetics Laboratory
Familial Hypercholesterolaemia (FH)
• Autosomal dominant disorder of lipid metabolism
• Heterozygous prevalence in UK population of 1 in 500
• Characterised by:
– Raised serum LDL-c (low density lipoprotein cholesterol)
– Tendon and skin xanthomata (due to cholesterol deposits)
– Premature coronary heart disease
• Early diagnosis is beneficial to patients since treatment with lipidlowering therapy (e.g. statins) can result in a near-normal life
expectancy by lowering the risk of coronary heart disease
• Homozygous FH exists but is rare
– Prevalence of 1 in a million
– Symptoms more severe: appear in childhood and often lead to early
death from coronary heart disease
Genes implicated in FH
• FH is a genetically heterogeneous disorder
• Mutations which cosegregate with the disease have
been found in at least three genes:
– LDLR (low density lipoprotein receptor)
• Over 1000 mutations spread throughout gene
• Exonic deletions and duplications (5-10% FH cases)
– APOB (apolipoprotein B)
• 9 mutations
– PCSK9 (proprotein convertase subtilisin/kexin type 9)
• 6 mutations
• Most mutations identified are in LDLR (~79%) with lower
proportion in APOB (~5.5%) and PCSK9 (~1.5%)
LDL-receptor pathway
ApoB-100 protein
LDL-c
Lipid core of cholesterol esters
LDLR
? PCSK9
Cholesterol synthesis
LDLR synthesis
LDL-receptor pathway maintains intracellular cholesterol homeostasis
Current FH testing strategy
• FH20 Elucigene ARMS kit (Tepnel Diagnostics)
– Identifies 20 most common mutations in UK
population
– Sensitivity of only 40%
• NICE guidelines recommend DNA testing be
used to confirm a diagnosis of FH (March 2009)
• Need for an increase in testing sensitivity
LIPOchip®
• A DNA array-based system designed by Progenika
• Used as the primary testing strategy for FH in Spain
• Detection of 251 common FH point mutations
– 242 LDLR, 3 APOB, 6 PCSK9
• Copy number variation detection in LDLR
• Currently targeted towards Spanish population within which the
manufacturers claim a sensitivity of 80%
• 4 FH20 mutations are not detected by Spanish version
– FS206, K290RfsX20, Q363X, C656R
• British version is under development
– June 2010 availability
– Sensitivity of 80-85%
– Probe sets for the 4 missing FH20 mutations
LIPOchip®: DNA-array technology
1.
2.
3.
4.
Multiplex PCR amplification
Product fragmentation
Labelling with biotin
Hybridization and washing
-
Using Tecan 4800 HS Pro station
Addition of Cy3-streptavidin (fluorochrome)
5. Results analysis
-
Using Agilent scanner and customised software
Light
Laser
Cy3-streptavidin
Biotin
Fragmented
PCR product
A scanned LIPOchip® slide
• 2 pairs of oligonucleotides per mutation:
– Each pair consists of a WT probe and a mutant probe
– Signal intensity ratios calculated for WT / (WT+Mut)
• Controls for hybridization process and for
measuring background signal noise
• Copy number variation detection controls
Graphical display of results
generated by LIPOchip® software
Normal
Heterozygous WT/mutant
Homozygous mutant
Validation Strategy
•
48 LIPOchip® slides were provided by Progenika to validate the
technology
•
Maximum of 12 samples per run - in order to perform copy number
detection two of these must be normal male & female controls
•
Samples selected for validation:
–
–
–
•
10 normal controls
6 FH20 positive controls
22 FH20 negative patients
Criteria for selection of FH20 negative patients:
–
–
‘Definite FH’ on referral card; or
High cholesterol level (over 8 mmol/L) plus either
(i) Family history of high cholesterol; or
(ii) Family history of cardiovascular disease
Results from Validation
• Normal controls:
– 9/10 slides passed quality control
– No point mutations were detected
• FH20 positive controls:
– 6/6 mutations correctly called by LIPOchip®
• FH20 negative patients:
– 2 pathogenic LDLR missense mutations:
• c.1796T>C (p.Leu599Ser)
• c.1618G>A (p.Ala540Thr)
– 1 unclassified LDLR variant:
• c.2177C>T (p.Thr726Ile)…likely non-pathogenic by in-silico
analysis
– Mutations were confirmed by direct sequencing
Problem 1: The M064 probe set
Normal
• c.91G>T (p.Glu31X)
• ‘No Call’ result was frequently
obtained:
– 7/9 normal controls
– 3/6 positive controls
– 20/22 FH20 negative patients
• Signal intensity values
extended beyond the normal
distribution parameters
although they were still distinct
from the mutation range
• Progenika are aware of this
problem and hope to resolve it
in the forthcoming British
version
Problem 2: Copy number variation
detection
7/9 normal controls appeared to have a deletion
of the LDLR promoter and exon 1
Normal LDLR gene dosage
Apparent deletion of promoter + exon 1
Problem 2: Copy number variation
detection
Also, poor quality dosage data was often generated…
These issues raised the question as to whether
LIPOchip® could be used in our laboratory for reliable
copy number variation detection
Traditional Full Screen
• Testing strategy
– Combination of dHPLC and direct sequencing
of LDLR gene
– MLPA for dosage analysis of LDLR gene
(MRC-Holland kit P062-C1)
• Samples:
– 10 normals from the validation
– 22 FH20 negative patients tested using
LIPOchip®
Results from Traditional Full Screen
• The 2 pathogenic LDLR mutations and the unclassified
LDLR variant identified by LIPOchip® were confirmed
• 2 further pathogenic LDLR mutations and an additional
unclassified variant were detected:
– c.1061A>T (p.Asp354Val)
– c.1067delA (p.Ala356ValfsX14)
– c.2479G>A (p.Val827Ile)…undecided pathogenicity by in-silico
analysis
• MLPA did not detect any deletions or duplications in the
LDLR promoter or exonic regions of patients or controls
Possible Stratified Testing
Approach for FH
FH20 – 40% sensitivity
Full screen with dHPLC and
sequencing plus MLPA – 99.9%
sensitivity
LIPOchip® - possibly 80-85%
sensitivity (currently need to use
MLPA in addition)
?
Full screen with dHPLC and
sequencing
Implementation of LIPOchip®?
Full screen
LIPOchip®
Sensitivity : Cost
Ratio
Conclusions
• LIPOchip® can be reliably used to detect common FH
point mutations with an increase in testing sensitivity
• Currently MLPA is required as a necessary complement
to LIPOchip® testing
• Some mutations detected by LIPOchip® require further
investigation regarding their pathogenicity using in-silico
analysis
• Additional validation work is needed on the British
version of LIPOchip®, when available
• Costing is an issue for LIPOchip® and will influence the
decision on whether or not to use the technology in a
diagnostic setting
Acknowledgements
• Wessex Regional Genetics Laboratory:
– Oliver Wood
– Esta Cross
– Alison Skinner
– Dr John Harvey
• Progenika:
– Dr Xabier Abad Lloret
– Maximiliano Crosetti