OMIC techniques on degraded tissue samples Pim French Dept - - PowerPoint PPT Presentation

omic techniques on degraded tissue samples
SMART_READER_LITE
LIVE PREVIEW

OMIC techniques on degraded tissue samples Pim French Dept - - PowerPoint PPT Presentation

OMIC techniques on degraded tissue samples Pim French Dept Neurology FFPE FFPE: formalin fixed, paraffin embedded The choice of storage for pathology labs FF tissues: 1000s FFPE tissues millions worldwide, ~20 million each


slide-1
SLIDE 1

OMIC techniques on degraded tissue samples

Pim French Dept Neurology

slide-2
SLIDE 2

FFPE

  • FFPE: formalin fixed, paraffin embedded
  • The choice of storage for pathology labs

– FF tissues: 1000’s – FFPE tissues millions worldwide, ~20 million each year in US – Well characterized samples

  • Histological
  • Pathological
  • Clinical
  • Clinical trial
slide-3
SLIDE 3

Formaldehyde

  • Formalin=Formaldehyde in solution
  • Crosslinks primary amino groups through CH2 linkage
slide-4
SLIDE 4

Primary amino group

slide-5
SLIDE 5

FFPE and DNA: Protein-DNA and interstrand crosslinks

Huang, J Am Chem Soc 1993; Lu, J. Am. Chem. Soc 2010

slide-6
SLIDE 6

FFPE fixation comes at a price

1. Degradation: Nicks and Double strand breaks 2. Hydrolysis of N-glycosyl bond 3. Deamination 4. Oxidation

1. 8-hydroxyguanine formation 2. Thymine or cytosine glycol formation

5. Thymine dimers

slide-7
SLIDE 7

Storage in FFPE compromizes DNA/RNA quality 1: Degradation

  • Formalin causes hydrolysis of

phosphodiester bonds

  • Dependent on time in paraffin

– (extra variable)

– 100 b fragments

slide-8
SLIDE 8

Storage in FFPE compromizes DNA/RNA quality 2: Hydrolysis of N-glycosyl bond

– Depurination: 10.000/cell/day – Depyrimidination: 500/cell/day

slide-9
SLIDE 9

Storage in FFPE compromizes DNA/RNA quality 3: Deamination

Rate ~100/cell/day

slide-10
SLIDE 10

Storage in FFPE compromizes DNA/RNA quality 4: Oxidation (Glycol formation)

  • Glycol: two hydroxyl (−OH) groups attached to

different carbon atoms.

  • Thymidine glycol: 100/cell/day

Chaw, Biochemistry 1980)

slide-11
SLIDE 11

Solving FFPE problem: 1: Adopt method (DNA)

  • Uracil-DNA glycosylase

Do, Oncotarget 2012, Clin Chem 2013

slide-12
SLIDE 12

Solving FFPE problem: Repair damage

  • Mono-methylol group additions (-CH2 OH) can be

reversed by heat

  • Enzymatic FFPE restoration

– mix of various DNA repair enzymes – cocktail of (proofreading) DNA polymerase, DNA repair enzyme and ligase

Arutyunyan, Izvestiya Akademii Nauk SSSR, Seriya Khimicheskiya, 1968.

slide-13
SLIDE 13

Problems working with FFPE tissue samples: RNA

  • RNA: Most arrays/methods are 3’ based and no

longer work

slide-14
SLIDE 14

Solving FFPE problem: Adopt method (RNA)

  • Revert chemical modification
  • Use random primers for cDNA synthesis
  • But:

– only mRNA has poly A tail – mRNA is only ~2% of all RNA in a cell

  • Requires rRNA depletion

– Ribominus – Ribo zero

  • LNA to rRNA specific sequences

mRNA: proteins (~2%) rRNA: ribosomes (~80%) tRNA: codons (~15%)

slide-15
SLIDE 15
slide-16
SLIDE 16

Examples of high throughput methods using FFPE tissues

  • Array based

– DNA

  • Oncoscan DX (Affymetrix)
  • Methylation arrays (Illumina)

– RNA

  • Exon arrays (Affymetrix)
  • DASL (Illumina)
  • Sequencing

– DNA – RNA

slide-17
SLIDE 17

DNA-arrays: Oncoscan

slide-18
SLIDE 18

DNA methylation

slide-19
SLIDE 19
slide-20
SLIDE 20

DNA methylation on FFPE tissues

slide-21
SLIDE 21

Molecular diagnostics on FFPE

slide-22
SLIDE 22

Targeted sequencing: PCR

  • Small PCR amplicons

Histology Molecular

slide-23
SLIDE 23

Targeted sequencing: Baits

Illumina TruSeq, Agilent Sureselect, Roche/Nimblegen etc

slide-24
SLIDE 24

Exome sequencing: Mutations

Gad Getz, Kristin Ardlie

slide-25
SLIDE 25

Exome sequencing: CNV

Gad Getz, Kristin Ardlie

slide-26
SLIDE 26

RNA expression: Exon array

slide-27
SLIDE 27

RNA-Seq

slide-28
SLIDE 28

Agilent sureselect

slide-29
SLIDE 29

Conclusions

  • Analysis of FFPE tissues is feasible

(RNA/DNA/Methylation)

  • Be aware of noise in the data