Day 6: MSI, serrated pathway, CRC subtypes Flashcards
HC15, 16
Most DNA mutations are…
Due to polymerases: makes mistakes
> DNA replication
> rate mistakes 1 in 10^5
First line DNA mutations repair
Polymerase itself repairs it: proofreading
> recognize wobble structure because wrong matching of the nucleotides
> Proofreading goes right in 99%
> error rate 1 in 10^7
What happens if DNA cannot reverse anymore to repair DNA
No proofreading, quick death because cancer development
Second line DNA mutation repair
Mismatch repair (MMR)
> recognize mistakes (wobble structure / nick in DNA) by MMR machinery
MMR in E coli
Proteins which recognize mismatches recruit other proteins for unwinding and cleavage of the wrong DNA region and resynthesis
MMR in eukaryotes
More proteins involved than E. coli
> first: recognition wobble, recruiting proteins, unwinding and excision and resynthesis
Validity MMR
99%
> error rate 1 in 10^9
> 3 per cell cycle
> accumulate mutations, but most are silent
DNA MMR is essential to prevent …
cancer
In CRC, sequences can be observed as ….(length)
Shorter
> regions in DNA are missing
> deletions often in short repeat sequences
Microsatellites
Short repeat DNA sequences
Are all microsatellites the same length
No, various lengths and amounts of repeats per individual
Why are microsatellites more mutated and shortened in CRC?
Prone to replication mistakes
> coding and matching sequence if one of the repeats is looped out because mutation
> still wobble, but not that distorted in DNA helix, not recognized by DNA polymerase as mistake
> in next replication, extension or shortening of microsatellite
When MSI (microsatellite instability)
dMMR > deficient mismatch repair
DNA replication error in microsatellite when pMMR
Repaired
Lynch syndrome
Familial syndrome distinct from FAP
> CRC early age
> Hereditary Non-polyposis CRC
> Lynch tumors are right sided mostly (proximal)
> Often mucinous
> mutation in mismatch repair in the germline
> less aggressive than FAP (germline APC loss)
How many MSI patients have Lynch syndrome, and the rest?
Only 10%
> the rest: spontaneous MSI patients > inactivation of MLH1 by CpG island methylation (CIMP+)
Lynch patients often carry germline mutations in …
-MLH1
-MSH2
-MSH6
-PMS2
> dMMR
> only explains 20% of dMMR CRC
Spontaneous MSI
Inactivation MLH1 by CpG island methyaltion
> epigenetic silencing
> Condensed DNA
> No transcription of MLH1
> Both alleles
> Deficient for MLH1 and MMR
CpG sequences in MSI CRC
Heavily methylated
> genes for DNA methylation upregulated
When in sequence of development cancer dMMR in MSI tumors?
First step
> unknown what mediates the changed CpG methylation profile > maybe inflammation or microbiota
Which mutations are also common in MSI tumors
KRAS and BRAF mutations
> induced by dMMR > accumulate mutations
MSI tumors create big .. within the tumor and the involved mutations
Heterogeneity
Wnt activation has important role in transformation to hyperplasia in MSI tumors, how?
dMMR induces Wnt overactivation of pathway which is essential for hyperplasia
> Mostly b-catenin mutations instead of APC loss
Third hit in MSI cancers
First: MLH1
Second: b-catenin
Third: KRAS/BRAF
> activation pathway in almost all MSI tumors
> BRAF in spontaneous MSI
> KRAS mostly in Lynch Sydrome
dMMR when microsatellites in coding sequence
Creates deletion of part
> mostly results in frameshift > whole protein changed
> for example in Bax: insensitivity to intrinsic apoptosis pathway (one of Hallmarks of cancer, rest is slowly acquired due accumulation mutations in MSI)
> other MSI affected genes
- Bax, Fas, Casp-5, Axin2, IGF2R, PTEN
Pathways in cancer development
APC loss pathway
MSI pathway
Serrated pathway
Serrated adenoma
Has serrated (kartel achtig) structure instead of cystic
Serrated pathway character
Fast growth, progression from first mutation to adenoma and cancer is faster than in FAP (APC conventional pathway)
> more frequent in women
> mostly right sided
Serrated Polyposis Syndrome (SPS)
- Epigenetically driven
- At least five serrated polyps with 2 larger than 10 mm diameter
- not familial mutations generally
- first degree relative
Common familial mutation for serrated pathway
RNF43 (inhibitor Wnt signalling via Frizzled)
> overactive Wnt pathway
> RNF43 is E3 ubiquitin ligase which ubiquitinates Frizzled to remove it from membrane
> after mutation, more sensitive for Wnt ligands (which are still required to thrive)
Common mutations in serrated pathway
-First BRAF mutation mostly > hyperactive MAPK (Mek/Erk) parthay > survival and proliferation
-Association with CIMP profile (CpG island methylator phenotype)
Senescence in serrated pathway
-Telomeres too short: senescence induced
> Hyperactivation oncogenic pathways: senescence
> or via DNA damage
> via TP53
> CIMP: escape senescence via tumor suppressors
Nevi (moedervlekken) often carry .. mutation
BRAF, but senescence induced
senescence when BRAF mutated
-Cell cycle inactivated in BRAF mutated cells
-p16ink4a > activated through oncogenic signals: tumor suppressor
> inhibits cells going from G1 to S phase by inhibiting cyclins
> pRB inactivated by cyclin/CDK (pRB inhibits cell cycle G1/S)
> senescence
Escape senescence in serrated BRAF mutated tumor
CIMP genes include p16 (CDKN2A) and MLH1
> system escapes from senescence
> methylation p16INK4a after BRAF mutation
Two fates with MLH1 after BRAF mutation and p16 silencing
-Mutated MLH1: MSI serrated CRC tumor
-Not mutated: MSS serrated CRC tumor
> later TP53 mutation as well
Which one is more aggressive: MSI/MSS serrated CRC
MSS > less neo-antigens made > resistance against ICI
Serrated adenomas formed when
BRAF mutation and p16 mutation
BRAF mutants are resistant against the chemo and targeted:
Cetuximab (anti-EGFR) > upstream of BRAF
5-FU (when MSI)
Serrated pathway
Normal mucosa
> CIMP+ and BRAF mutation
Adenoma
> Methylation p16INK4a (and or not MLH1)
> Loss TP53
MSI/MSS serrated CRC
BRAF mutation but no CIMP
No spontaneous p16 inactivation > senescence
> traditional serrated adenoma
> CIMP-H: sessile serrated adenoma
Serrated tumors and prognosis
Poor, MSI better than MSS
> tumor budding, invasive growth pattern
HC16: MSS tumors and metastasis
Increased chance, less recognition by immune system, less neo-antigens
Treatment stage 3 CRC
Surgery and adjuvant therapy like FOLFOX (5-FU + oxaliplatin)
Toxicity Oxaliplatin
Neuropathy side effects which can become rationale to stop therapy
Problem treatment CRC
One size fits all strategy, not all patients respond well to therapy > heterogeneity
Stage 3 CRC has increased chance of … after surgery
Recurrence
50% of patients get relapse with CRC stage … > …
stage 3 which shows heterogeneity after therapy
> who need adjuvant therapy to prevent relapse, because chemotherapy is long, expensive and has side effects!
Stage 4 CRC treatment
Not well treatable, chemotherapy and targeted therapy used
Adjuvant therapy response for recurrence in stage 3 CRC
Improvement
50% > 30%
Personalized medicine in cancer: detect profiles
- Look for mutations in cancer: DNA profiles
- Look for MSI features
- Look for gene expression differences: RNA seq, immense differences between subtypes in gene expression > coupled to routes in development CRC
CCS classification CRC
Based on gene expression profiles
> linked to development routes CRC
> Unsupervised classification > find best seperation in subgroups > generate classifier
> CCS1, CCS2, CCS3
Molecular characterization subtypes
CCS1: KRAS and TP53 mutated > Chromosomal instable tumors
CCS2: MSI and CIMP and BRAF > MSI tumors
CCS3: KRAS, BRAF, TP53 > ??
Gene set enrichment analysis with CCS and adenomas
Take top X genes expressed in serrated adenomas and FAP adenomas (APC loss route) and compare with gene expression in CRC
> CCS1: Tubular adenomas (FAP, APC route)
> CCS3: Serrated adenomas (BRAF loss first)
Survival CCS1 and CCS3 in stage 2 CRC
Better CCS1 and worst CCS3
> more recurrence in CCS3
Different researchers came with own subtype classification: towards consensus subtyping
Different methods used
> 4 consensus molecular subtypes made
> Network analysis of all patients and subtype calls per research group used
> each group classified each dataset to make big matrix
> CMS1-4
Molecular features CMS
- No mutation specific associations
- CMS1: MSI tumors (MSI-H and CIMP-H)
- CMS1 least copy number counts but most mutations
- CMS3 most KRAS mutant
CMS features
- CMS1: MSI, immune infilitrated (serrated)
- CMS2: canonical (tubular adenoma, CIN+)
- CMS3: Metabolic (tubular, KRAS mutated)
- CMS4: mesenchymal (CCS3, serrated adenoma, high density stromal cells)
CMS clinical features
CMS4 has worst survival rate (serrated with MSS likely)
> recurrence and metastasis: bad prognosis
Goal with subtyping
Stratification based on gene expression based markers to improve treatment
Which (CMS2 or CMS4) has the worst survival in cell line models
CMS2 > from the APC vogelgram pathway
(CMS4 is mesenchymal with a lot of stroma)
Which CMS (2/4) reacts best to chemotherapy when transplanted to mice?
CMS2
Why is immunotherapy a good option for MSI cancer
Mutated proteins (frameshifts induced due dMMR) > more neo-antigens made
> due extension or shortening of microsatellites
> proteins made which should not exist: on MHC1 > recognized by immune cells
> neo-antigens perfect targets for immunotherapy: T-cells which recognize them are present but tumor has developed immune escape for example through immune checkpoint
CD8+ marker higher in MSI/MSS?
MSI, more immune infiltration
> immune system in alarm phase but cannot clear tumor: immune checkpoint
> immunotherapy: immune checkpoint inhibitors
Personalized neo-adjuvant therapy
Give immunotherapy before surgery in MSI patients
> 97% have no tumor when surgeon should operate
> tumor disappears completely
