Analysis and interpretation of unclassified variants UVs

1
Analysis and interpretation of unclassified
variants (UVs)

• Aberdeen
• November 26th 2008

2
What is a UV?

• Novel sequence change detected in a gene of
  interest
• Clinical significance of the variant is unknown
• Can occur either in the coding region of the
  gene, or in the intronic, 5UTR or 3UTR regions
• Usually substitutions but may also include
  in-frame deletions/insertions

3
How common are they?

• In BRCA1 represent 27 of the sequence
  changes detected (excluding common neutral
  variants)
• no cases to date of UV in both BRCA1/2 genes
• In Long QT represent 29 (minimum) of the
  sequence changes detected (excluding common
  variants)
• no cases to date of UV in gt1 gene
• Problems arise when reporting these variants
  are they responsible for the phenotype or are
  they novel, rare variants of no clinical
  significance??

4
Guidelines for analysis and reporting

• Best Practice Meeting (CMGS) April 2007
• Formed an agreed set of standards that
  laboratories should follow to assist in
  determination of the clinical significance of
  these variants
• http//cmgsweb.shared.hosting.zen.co.uk/BPGs/pdfs
  20current20bpgs/UV20GUIDELINES20ratified.pdf

5
Lines of evidence

• Literature, mutation databases (curated?), SNP
  databases
• Testing controls
• Co occurrence in trans with deleterious mutations
• Co-segregation with phenotype in family
• De novo variant occurrence
• Species conservation
• In silico prediction- missense and splicing
  effects
• RNA/splicing studies
• Functional studies
• Not all possible for each variant
• Summarise results and decide on likelihood of
  pathogenicity

6
Intronic variants

• In silico analysis programmes available Splicing
  sequences finder EMBL
• alternative splicing site Fruitfly etc
• Approximately 1/3 of UVs in BRCA1 and LQT are
  intronic

7
Intronic considerations

• Dependent on region of intronic variant only
  some programmes available- consensus of results
• Nonsense mediated decay - site of variant in gene
  
• Establishing both copies present in RNA studies
  if no effect other variants in region of
  interest
• Level of gene expression in tissue available for
  RNA work
• Are we identifying all intronic variants of
  interest dependent on position of primers in
  screen

8
BRCA1 c.5194-12GgtA

• Not detected previously in over 1000 control
  chromosomes
• Not in SNP databases
• Recorded several times in BIC database unknown
  clinical significance
• In silico analysis (3/5) suggests putative
  cryptic splice site created
• Three affected family members (3 generations)
  demonstrate sequence change
• Likely to be pathogenic????

9
BRCA1 c.5194-12GgtA
Sequencing revealed new splice site created at
site of variantled to inclusion of 10bp
intronic sequence in transcript to create
frameshift mutation

exon19/ exon20
AAA ATG CTG AAT GAG /CAT GAT
TTT GAA (WT) A
ATG AG T TTC TTT CAG CAT GAT TTT GAA (NEW)

fsx5
Lower level of mutant sequence detected
efficiency of new site/degradation of
mutant? Likely to be cause of phenotype
pathogenic?? More family members.. unaffecteds?
10
KCNQ1 c.478-56TgtG

• Sudden death case - 12 year old boy (playing
  football)
• Not detected previously in over 200 control
  chromosomes
• Not in SNP databases, or inherited arrhythmias
  mutation databases
• No other variants in KCNQ1, KCNH2 (to date)
• In silico analysis (3/3) suggests branch point
  destroyed
• But no other tissue available .
• No other affected family member
• Have requested parental samples

11
Missense variants
In silico analysis programmes available Clustal,
SIFT, Polyphen, Align GVGD etc
12
Missense considerations

• Availability of suitable sequences for
  conservation analysis
• Range of available sequences wide spectrum of
  species (human to pufferfish ideal)
• Alignments (own vs programme)
• Basis of analysis residue conservation vs
  structural vs chemical difference
• Functional studies dependent on gene and region
  of gene that variant resides
• Effect on splicing if at exon boundary
• ESE, ESS etc

13
BRCA1 c.181TgtG p.cys61gly
Not detected in gt1000 control chromosomes Publishe
d variant Recorded numerous times in BIC
clinically significant Very well conserved in
protein in a functional domain (ring finger) In
silico analysis suggests not tolerated Multiple
papers of functional studies indicate
deleterious .pathogenic
14
KCNQ1 c.458CgtT p.thr153met

• Sudden death at 6 weeks (fiscal case)
• Not detected in 200 control chromosomes
• No other variants detected to date in other LQT
  genes
• Not in SNP databases nor in inherited arrhythmias
  mutation database
• Occurs in functional domain conserved in 8
  available sequences (human-zebrafish)
• In silico analysis suggests variant not
  tolerated/tolerated
• No functional studies published
• Likelihood of
  pathogenicity?????
• Next step??? awaiting parental samples

15
BRCA1 c.5207TgtC p.val1736ala

• Detected in 4 individuals (apparently unrelated
  but 3 from Orkney?)
• Published variant
• Recorded 17x in BIC unknown significance
• Recorded 1x in DMuDB
• Recorded in SNP database but no frequency data
• Conserved in protein in a functional domain
• In silico analysis suggests not tolerated
• One paper of functional studies suggest
  deleterious
• Likelihood of pathogenicity??
• Next step to confirm - additional family
  members??

16
Reporting considerations

• Missense variants more difficult to interpret (if
  RNA available, ex vivo splicing assays becoming
  more routine)
• Ethnicity
• Time taken to analyse (manual vs automated)
• Complete screening
• gt1 mutation causative in disorder
• Penetrance/ expressivity/ phenocopies
• Lack of clinical information with referral (test
  only affecteds??)
• Small family sizes limited family members
  available

• Test for presence of variant in family
• ? Affected