A review of the cancer gene cloning strategies:
pre-genomic era
STONE AGE
post-genomic eras
MODERN AGE
www.interlog.com/~lenore/ art/pix/stone.jpg
fmwww.bc.edu/JSPMA/
Pre-genomic era startegy are basic
Knowledge about human genome is a landmark that gave us a chance to get closer to goal, but not solve all problems in goal attaining itself (in cancer gene cloning)
Typical strategy to find a cancer-related gene contain many steps that take time
There are many variations in the strategy,
and many different approaches to prove that gene is cancer-related can be chosen HUMAN GENOME SEQUENCE CAN HELP US TO CHOOSE CANDIDATE GENES but have nothing to do with answers “YES”/”NO”
In a modern time a search for a cancer related gene
IS NOT A TASK OF CLASSICAL GENETICS
You ask why?
1. Most changes in cancer-related genes occurs in somatic cells, not in germ-line 2. Almost all Mendelian germ-line affected genes that are underlie cancer syndromes are already described 3. All “associations” and “influences” of polymorhic allele to cancer are true weak to moderate; Up to now most attempts to locate cancer polymorphism with methods of classical genetics are inconclusive.
In this course of lecture we will review cancer syndromes itself
BUT
We will omit “classical genetics” strategies to clone cancer genes
Linkage disequilibrium, log of odds of linkage (LOD) scores….
“Classical genetics” strategies to clone genes involved in inherited cancer syndromes are similar to “classical genetics” strategies to clone any non-cancer syndrome involved gene, and subject of Human Genetics
Cancer genetics = SOMATIC CELL GENETICS
HOW TO FIND A GENE
Positional cloning – based strategies
Strategies not connected to certain position in genome
Goal of approach: Goal of approach: choose a list of candidates In a whole genome-based search located in place of interest, make list of candidates located in different places in genome;
Remove false-positives;
Prove that a particular candidate is a true cancer related gene
HOW TO FIND A GENE
Locate a chromosomal area deleted in particular tumor type Screen hundreds of patients to find a “lucky” small deletion Make a cosmid (PAC) contig Retrieve a genomic sequence You Can Believe In The Strategy
Map all genes-candidates (experimentally or in silico)
Check all candidates (try mutations, expression in cell lines, or knockout in mice) The Pure Science Starts HERE
Honestly study gene functions in the tissues and tumors
Locate a chromosomal area deleted in particular tumor type
(Deletion mapping)
(Study a sites of translocations) often accompanied by deletion
Chromosome-based screenings
Gross chromosomal abnormalities
Polymorhism-based screenings
LOH (loss of heterozygozity) of polymorphic makers with known locations
FISH Analysis
CGH analysis
Gross chromosomal abnormalities
(Conventional cytogenetics testing)
t(9:22) CML
Microscopically visible
Trained eye can recognize chromosomes After conventional Band-revealing staining
LOCUSES revealed by Gross chromosomal abnormality staining are include hundreds of genes…. Problem…
www.cmlsupport.com/cyto.jpg
FISH analysis (Fluorescent in situ hybridization) -- translocation
DNA probes are hybridized to fixed cells on microscope slides.
metaphase
The hybridization to target loci is visualized by the detection of fluorescent signals on metaphase chromosomes or interphase nuclei.
Pre-metaphase
FITC or Fluorescein (green), Texas-red (red), Rhodamine (red), DAPI (blue)
COMMON FLUORO DYES Green + RED = YELLOW
mti-n.mti.uni-jena.de/~huwww/ MOL_ZYTO/imageAU9.J
FISH analysis (Fluorescent in situ hybridization) -- DELETION
Interphase FISH, relaxed chromatin
Two green, two reds on different chromosomes – no deletion Two green, one red – One red is deleted.
Size of deletion can not be estimated. Deletion size is smaller than distance to green signal (probe)
GREEN SIGNAL SERVE AS A CONTROL PROBE ON A SAME CHROMOSOME.
http://lambertlab.uams.edu/images/cell.jpg
Aneuploidy revealed by FISH
8 copies of chromosome 13 in pancreatic carcinoma
Chromosome13-specific probe painting
http://68.33.28.8/geneticsweb/fish.htm
RESOLUTION of FISH methods
Classical Cytogenetics can see aberrations approximately one chromosomal band in size A. High resolution FISH mapping on interphase chromatin.
The resolution (distance between probes) of such mapping is approximately 20 Kb while the coverage is 20Kb -- 5,000Kb. Coverage is a size of the probe itself.
B. High resolution FISH mapping on released DNA fiber.
The resolution of DNA fiber mapping is 1-2 Kb.
The figure illustrates three probes ( a, b and c) simultaneously located within 50 kb region
www.seedna.com/gif/fiber1.jpg
FIBER FISH – down to exon level!
Fiber-FISH of HER-2/neu (GREEN) and topoisomerase II-alpha (RED)
in DNA from UACC-812 cell line
You can see exons!
To be honest, not always….
It is a difficult method. Most researchers stop on interphase FISH stage
www.uta.fi/laitokset/imt/ sgy/jorma/uaccfiber.jpg
Comparative Genomics Hybridizations (CGH)
CGH is a genome-wide scanning of differences in DNA sequence copy number
CGH is based on a modified in situ hybridisation, where differentially labelled test DNA (cancer, green) and reference DNA (normal, red) are co-hybridised to normal metaphase spreads
Both test (tumor) and reference (normal) are whole genome DNA preparations
Digital imaging system
DNA gains and in test DNA are seen as chromosomal regions with an increased fluorescence ratio (> 1.25), while losses result in a reduced ratio (<0.75).
Typical pictures of CGH
Of course, CGH data should coincide with FISH and LOH data
Pitfalls of CGH
1. A technically demanding technique, requiring expensive hardware and software.
2. Cannot detect balanced translocations or rearrangements. (Only net gain or loss of material).
3. There is a room for standardization, more objective and uniform interpretation and statistical evaluation of the CGH profiles. Common database has just been set. 4. The technique is limited by the resolution of the hybridisation target, e.g. the metaphase chromosomes. The current resolution is 5-10Mb. Changes affecting smaller regions are only detectable in the case of high level amplifications (e.g. 5-10 fold amplification of 1Mb). 5. Sensitivity can also be hampered by contamination of tumor cells with normal cells, Mosaicism is difficult for detection.
LOH (loss of heterozygosity)
Reminder:
Informative microsatellite ( = polymorphic in this particular normal sample) TWO alleles in normal tissue versus ONE allele in tumor tissue RESOLUTION of LOH deletion mapping depends on a density of polymorphic markers on the map. Average density on good whole genome LOH scan is 10 Mb.
The total genetic map length of the human genome is about 3,000 Mb. We have to run 300 polymorphic markers (500 to be sure)
A hell of work to 10 Mb resolution!
Typical result of LOH mapping
Allelotyping of ductal carcinoma in situ of the breast: deletion of loci on 8p, 13q, 16q,17p, and 17q.
[clear box], p arm; [black-filled box], q arm
www.genlink.wustl.edu/.../figures/ figure1_radford1.html
How to avoid huge amount of work and poor resolution of LOH?
1. Start from particular chromosome.
If previous analysis suggests that chromosome 16 is involved in breast cancer, work only with chromosome 16 – derived marker panel. You can achieve a good resolution by choosing 300 markers form the same chromosome.
2. Use polymorphic markers residing in the introns of candidate genes.
In such case you can be sure that candidate gene is deleted without extensive checking of neighbouring genes.
Typical result of LOH mapping
Smallest regions of overlap of LOH patterns on 16q in breast cancer
PISSLRE, copine VII, CTCF, SIAH1 are excluded
(1999-2002)
Those smallest regions are ready to go in screening for genes residing in them (1994)!
WWOX, a gene that maps to the common fragile site FRA16D region in chromosome 16q23.3-24.1
So, resolution is still poor….. CBFA2T3, 2002
Screen hundreds of patients to find a “lucky” small deletion
(Small-scale deletion mapping)
FiberFish (as was described before)
Genomic Southern with densely scattered probes
Electrophoresis of digested genomicDNA (band sizes 1-10 kb, EcoRI, HindIII or both A serie of consequent hybridizations with short, non-repetitive genomic probes with known locations
Band that is absent on autograph is homozygously deleted
Band that is twice weaker than control is hemizygously deleted
How genomic Southern looks
Housekeeping gene Serve as a control Potential Pitfall: presence of normal cells makes 2-allele deleted probe looks like 1-allele deleted; 1-allele deletion can be undistinguishable from normal
Screen hundreds of patients to find a “lucky” small deletion
(Small-scale deletion mapping)
FiberFish (as was described before)
Genomic Southern with densely scattered probes
Why not detect by simple PCR?
Why not detect by simple PCR?
Speaking theoretically, we can design primers for every probe (the same probe as for genomic Southern) And check it easily without any time-consuming hybridization
In reality,
It will be very difficult to do
Because of cross-contamination due to the extreme sensitivity of PCR
We will have a mess of
PCR-amplified traces of air-born DNA contamination + sample to sample difference of normal DNA contamination, That is why PCR detection of deletions method
is not very practical
Screen hundreds of patients to find a “lucky” small deletion
(Small-scale deletion mapping)
FiberFish (as was described before)
Genomic Southern with densely scattered probes
Why not detect by simple PCR?
Representation Difference Analysis (RDA) {“very special” PCR}
Representation Difference Analysis (RDA)
Introduced by Nikolay Lisitsyn (Science 259: 946, 1993)
It is a PCR procedure that analyzes the difference between two complex sets of genomes RDA detects deletions, rearrangements, or exogenous viral DNA sequences
RDA enriches for unique DNA sequences from one of the genomes by removing shared sequences. It takes advantage of PCR to selectively amplify the regions of difference
DRIVER
NORMAL
TESTER
CANCER
1. FRAGMENTED TESTER DNA IS LIGATED WITH ADAPTER PRIMERS FOR PCR
2. TESTER WITH PRIMERS ATTACHED is melted and re-anneled with DRIVER (IN EXCESS) CANCER 3. FILL IN ENDS OF FRAGMENTS to create double-stranded PCR priming sequences, then PCR AMPLIFY 4. Only sequences unique to tester WILL AMPLIFY EXPONENTIALLY
CANCER
5. ADD MORE DRIVER to PCR products and SUBTRACT DRIVER AGAIN.
X
X
www.ich.ucl.ac.uk/.../academicunits/
HOW RDA LOOKS
1. 2. 3. 4. 5.
Lane 1 Molecular weight markers
Lane 2 TESTER
(digested tumor genomic DNA)
Lanes 3 to 5
progressive enrichment for sequences
amplified in tumor DNA
following three successive rounds of subtractive hybridisation.
www.tv.slu.se/birch/ image1.htm
PITFALLS OF RDA
1. Amplifications are more easily detected by RDA than deletions, (as we always should add an excess of driver). Normal driver DNA is usually more available than cancer driver DNA 2. If studied tester sample have more than one region of deletion/amplification, found genomic probes will belong to all rearranged regions (will be scattered across the genome)
3. All disadvantages of PCR (all contaminations will be found as false-positives!)
+
Make a cosmid or PAC contig
That overlap an area of interest As we have now complete sequences of human and mouse genomes, in many cases we no need to fulfill this goal anymore
However
This step still can not be skipped for mapping of cancer-related genes in model animals and in livestock
ACTUALLY, MOUSE IS NOT THAT GOOD MODEL FOR MANY HUMAN CANCERS…..
Typical representation of contig assembly
Polmorphic markers or mapped probes used as anchors
Contig can be made from cosmids, PACs or YACs
Most reliable contigs are 4 to 7x coverage
http://www.mrc-lmb.cam.ac.uk/happy/gifs/contig.gif
HOW TO BUILD A CONTIG
Every cosmid presumably belonging to the region of interest is hybridized to all cosmids from chromosome-specific library with 5-7x coverage Results of all hybridisations are subject to analysis by contig-building software
Weak (less covered) points need re-hybridization to libraries made by different method
http://www.roche-applied-science.com/dig/images/dot_blot.jpg
Pitfalls of contig construction in the areas deleted in tumors
1. Areas rearranged in tumors
often contain a lot of repeats
(that is why they are prone to rearrangements). Repetitive areas are difficult to clone (therefore they are under-represented in libraries).
We can overcome it by extensive hybridizations (enhancing the coverage) or by using special cloning systems durable to loss of repetitive inserts
2. Contig branching (because of region specific low copy repeats)
Retrieve a genomic sequence
than computationally search for candidates
Currently we able to retrieve free of charge:
1. Human Genomic Sequence located between two known markers 3. BLAST Human Genomic Sequence to mouse genome and draft genome and retrieve corresponding genomic sequences
2. BLAST Human Genomic Sequence to human, mouse and other subdivisions of dbEST database and retrieve corresponding expressed sequences
Comparative genomics
Direct comparizon of two synthenic genome areas (often human vs mouse)
Helps to find all conservative ( important) region.
1. genes that absent in dbEST because of restricted pattern of expression or low-level expression 2. Exons that absent in common mRNA, but important for certain (ideally, cancer-related) gene function 3. Regulatory regions (promoters, enchancers, silencers, SAR/MAR =scaffold-associated /matrix-associated regions
VISTA is a best tool for comparative genomic study
Lawrence Berkeley National Laboratory, 2002
It was designed to visualize long sequence alignments of DNA from two or more species with annotation information
mVISTA (main VISTA) A program for visualizing alignments of an arbitrary number of genomic sequences from different species rVISTA (regulatory VISTA) combines transcription factor binding sites database search with a comparative sequence analysis.
www.lbl.gov
HOW VISTA OUTPUT LOOKS
www.lbl.gov/
EST-based reconstruction of mRNA isoforms
ESTs also can be aligned in contigs.
Automatically built EST contigs are stored in UNIGENE and in TIGR database
Manual alignment of ESTs in contig is time-consuming, but gave to researcher a deeper understanding of gene structure, (…presumably…) error-free and allows to make a multi-species ESTs contigs.
Some area of the gene of interest can be overlapped by Mouse-derived clones, and reconstructed by conservative human genome sequence
EST-based semi-automatic scheme of gene
http://www.mad-cow.org/contig.gif
Experimental search for cancer-related genes in defined chromosomal regions
Hybridization screening of cDNA libraries with probes made from whole cosmids (PACs, YACs). 1. Before hybridization, the complexity of the probes should be reduced via pre-hybridization with repetitive fraction of DNA
2. Choice of library is dictated by tissue of tumor origin. So, to find a leukemia-involved gene it will be clever to screen lymphocyte-derived library. Not muscle, brain… (By the way, testis is a clever choice for any screening)
Strategies of cancer gene search not connected to certain locus in the genome
Differential display
Tumor markers and phage display
Microarrays
SAGE (Serial Analysis of Gene Expression)
Kinetics of paper publishing
RDA and Subtr Hybr is almost the same technic
http://www.genhunter.com/support/comparison.graphic.jpg
Differential Displaying (DD) - PCR
A set of different primers for DDRT-PCR are complimentary to poly (A), and contain two additional 3’ nucleotides like GG, GG, GA…. All separately
5’ xxxxxxxxxxxxxCTAAAAAAAAAA 3’ GATTTTTTTTTTT 5’
12 primers = 12 aliquotes
First strand reverse transcription to cDNAs in 12 aliquotes
2nd PCR with the same anchor primers (separately) and a randomly designed decamer in the presence of labeled dATP for visualization
2nd PCR
A
Arbitrary primers can be gene-specific by chance
Gene-specific primer
B
oligo dT primer
C D
Possible primer combinations (we can not control them)
ALL 12 PCR reaction ARE MADE with mRNA from tumor and normal sample, then separated in parallel lanes on gel
Differentially expressed products are seen after comparison
http://biology.uark.edu/drhoads/ddrt_pcr.gif
Verification is extremely important
www.jenapharm.de/eng/research/endo/bilder/genomics.jpg
Most common issues with the DDRT method
A. High Frequency of false positives
1. Fail to re-amplify with the same primers
2. Fail to be differentially expressed on Northern
B. Low Sensitivity for transcripts that are expressed at (very) low copy numbers
more than 70 technical articles have been published on improvements and powerful derivatives of the differential display method (mostly on specificity improvement)
SAGE
(Serial Analysis of Gene Expression)
Invented in 1995 Velculescu V.E, Zhang L., Vogelstein B. and Kinzler K. W. (Science 270:484-487, 1995)
Includes tons of sequencing
SAGE Scheme
1. The restriction enzyme Nla III, = "anchoring enzyme" (AE), cleaves at the sequence 5'-GTAC, leaving a four nucleotide 3' overhang.
BIOTIN
overhangs
Biotinylated fragments are isolated using streptavidin beads.
2. Beads with DNA divide in two tubes and DNA is ligated to linkers A or B. The linkers contain unique primer binding sites, the recognition sequence (5'-GGGAC) for a tagging enzyme (TE), in this case BsmFI, and an Nla III compatible sticky end.
Cleave with BsmFI tagging enzyme (TE) and blunt-end-fill in.
The enzyme BsmFI cuts 10 and 14 bases in the 3' direction from its recognition site, thus leaving the 9-11 nt "Tag” connected to linkers.
BsmFI rec site BsmFI rec site
9-11 nt in size
9-11 nt in size
mRNA-specific TAGs
All pieces with beads ligated all together and amplified using primers A and B.
Restriction with anchoring enzyme
The we isolate ditags,
concatenate them and clone.
X and O represent nucleotides from different transcripts.
SEQUENCING OF concatemers
SAGE software searches GenBank for matches to each tag
Count their relative abundance in tumor and control sample
SAGE analysis of genes that change their expression during p53-induced apoptosis
www.sagenet.org/ p53sage/ MAINSAGE.HTM
Colon carcinoma SAGE project
290,394 tags were analyzed from normal, adenomatous, and cancerous colonic epithelium 21,343 different transcripts observed
957 were found to be differentially expressed
45 transcripts were elevated 20-fold in adenomas, 40 transcripts were elevated 20-fold in cancers, 9 transcripts were elevated 20-fold in both.
Cancer Res 2001 Oct 1;61(19):6996-7001
Diagram of genes overexepressed and underexpressed in Colon Cancer
UNDEREXPRESSED OVEREXPRESSED
Major conclusion: primary tumors and tumor cell lines
are very different object to study
Experimental proof of SAGE data
H769020 (TGF-beta induced gene beta-IGH3, M77349) , increased
H516402 (CL 100protein tyrosine phosphotase, X68277) , decreased
EXPRESSION MICROARRAY
High-density arranged gene- representing spots on surface suitable for hybridization Nylon Surface Probe labeling Radioactive
Glass or plastic
Fluorescent
PCR fragments
Genes can be represented by
Oligonucleotides
www.dkfz-heidelberg.de/funct_genome/ fig-cancer-chip.jpg
Working cycle of fluorescently labelling microarray
Healthy lung Lung tumor Excitation
Emission Green Red Fluorescent scanning Normal in red Tumor in green
Hybridization to microarray
Comparizon of gene profiles
www.omrf.org/OMRF/Images/PRPics/ 2001SprSchematic.jpg
Representation of genes differentially expressed in normal mucosa and colon tumors
(upper and lower boundaries represent a 4-fold difference).
2% of 4000 genes changed expression significantly
microarray.princeton.edu/oncology/ images/figure1big.jpg
Adenoma vs. Carcinoma (differential diagnosis of this states by small subset of genes)
microarray.princeton.edu/oncology/ images/figure1big.jpg
Golub et al., Molecular Classification of Cancer: Class Discovery and Class Prediction by gene expression monitoring. Science 1999 286: 531-537
For the most effective treatment in a specific cancer type.
Acute Lymphoblastic Leukemia vs Actue Myeloid Leukemia RNA from 38 Bone Marrow Samples (27 ALL, 11 AML) hybridized to high-density oligonucleotide microarrays of 6817 human genes
50 Genes identified as best predictors of cancer class;
so real diagnostic array can be cut to 50 spots
TUMOR MARKERS AND PHAGE DISPLAY
Tumor markers are proteins that can often be detected in higher-than-normal amounts in the blood, urine or body tissues of some patients with certain types of cancer.
Tumor markers are produced either by the tumor itself or by the body in response to the presence of cancer or certain benign conditions.
Examples of tumor markers:
PSA and PAP
Prostate Specific Antigen (PSA) and Prostatic Acid Phosphatase (PAP) are elevated in benign prostate conditions, such as prostatitis and benign prostatic hyperplasia (BPH), or in a prostatic carcimomas. CA125 Many women with ovarian cancer have elevated CA 125 levels. A falling CA 125 level generally indicates that the cancer is responding to treatment. It is also used for monitoring of recurrence.
Carcinoembryonic Antigen (CEA) Used for monitoring of colorectal and lung cancer
Alpha-fetoprotein (AFP)
AFP is normally produced by a developing fetus. An elevated level of AFP strongly suggests the presence of either primary liver cancer or germ cell cancer
Changes in tumor marker concentration during the course of disease
No response to treatment
No treatemnt Relapse
Good response
Second remission
Remission
: www.dpcweb.com/medical/cancer/ cancer_tm_overview_lg.html
Phage display
Is a method of a search for a new tumor markers
Antibody fragments can nowadays be isolated from naive phage display libraries without immunisation, by-passing hybridoma technology human recombinant antibodies can be presented in phage
libraries in scFv format. They consist of a single polypeptide chain:
-- antibody heavy chain variable domain (VH) -- a flexible polypeptide linker -- a light chain variable domain (VL)
How scFv antibodies looks :
Ig G
Bispecific IgG
Fab
Fv
scFv
Phage display usually utilize only 3 of 50 antibody germline gene segments. Large repertoire produced by appending short variable complementarity-determining regions 3 (CDR3) onto them. (random loops of 4, 5 or 6 amino acids. ) ....... C92 A R (X)4-6 F D Y ......
Scheme of phagemid library cloning scFv antibodies are presented on coat of phages
We can generate a millions of random scFv plaques representing a variety of potential Abs on bacterial lawn
www.pharma.ethz.ch/bmm /protocols/schema.jpg
Selection of antibodies from phage display libraries:
the enrichment of an antigen-specific phage-antibody (circle) on a background of non specific phage-antibodies (square).
www.pharma.ethz.ch/bmm/ protocols/schema.jpg
Antigen comes from serum of cancer patients
HOW to prove that our candidate is real cancer-related gene
Detection of mutations in tumor samples
Experiments on cells in culture
Promoter methylation in tumor samples
Knockout/ transgenic mice
Detection of mutations in tumor samples
DGGE (Denaturing Gradeint Gel Electrophoresis)
Both discovery and diagnostics
SSCP (Single Strand Conformation Polymorphism)
SEQUENCING – a king of methods. It is already cheap enough Mutation detection chips (more diagnostics than discovery) Allele-specific oligonucleotides (ASO) and modifications… (Diagnostics only)
SSCP (Single Strand Conformation Polymorphism)
PCR product is denatured, chilled quickly and run on non-denaturing gels. Any sequence change causes formation a different conformer resulting in variation in strand mobility. Needs sequencing to prove a significance of the change www.uct.ac.za/depts/liver/ images/sscp.jpg
A differences in conformation depends on conditions of gel running (% of buffer and acrilamide, glycine in buffer, temperature etc….)
Often it is important to check 3-5 different conditions
before undoubtful resolution can be achieved
www.hci.utah.edu/groups/genomics/ images/APC%20SSCP.gif
DGGE (Denaturing Gradeint Gel Electrophoresis)
The small (200-700 bp) genomic fragments are run on a low to high denaturant GRADIENT acrylamide gel Each fragments move according to molecular weight, but as they progress into more denaturing conditions, each (depending on its sequence composition) reaches
A POINT where the DNA BEGINS TO MELT
They retard, and we will see shift in mobility
We will see different shifts in mobility for differing products
Here is just one sample !
hwww.enbiogen.com/product/ image/tgge1.gif
“mirroring” of two curves indicative of mutation
Separated homoand heterodimeres
Urea/formamide gradient 0%100%
Many samples in one gel – we will not see any curves
1- normal
3- Homozygous mutations will yield one band on a different position 2, 4, 5, 6 – heterozygous mutations will yield 4 bands (2 homozygous and 2 heterozygous) NOT ALL BANDS ARE SEEN !!!!!
www.leveninc.com/cftr_ex.gif
Sequencing is direct method of mutation detection
Mutations Deletion
www.infobiogen.fr/doc/STADENdoc/ manual/
Mutation detection chips
photolithography or piezoelectric printing of Prefabricated Oligonucleotides OR directed oligonucleotide synthesis on chip
Example:
Chip to detect all possible heterozygous mutations in the 3.4 kb BRCA1 exon 11 (~60% of coding region).
> 96,600 oligonucleotides designed to detect all possible single-base substitutions, single-base insertions and 1-5 bp deletions on both strands. 28 separate oligonucleotides (sense and antisense strands) for each nt position
Allele-specific probes (tons of modifications of principle)
Just one variant (fluorescent) Principle: specific annealing of probes to one but not to another alelle that differ by one nt
Suitable only for detection of known mutations
www.idahotech.com/.../RAPID/ images/ hybridizationprobes.jpg
Promoter methylation in tumor samples
c
Best methods are based on bisulfite modification of genomic DNA
CU C m unmodified
www.bmskorea.co.kr/
Note the disappearance of bands in the "C" lane of the unmethylated DNA demonstrating that the unmethylated cytosine nucleotides were converted to uracil by the bisulfite treatment.
G
www.intergenco.com/pictures/ cpgenome_seq.jpg
Classical representation of promoter methylation in papers
otir.cancer.gov/tech/slides-brock/ 19-01_sequence.gif
Experiments on cells in culture
Transfection a TSG to cancer cells makes changes of cell phenotype
Apoptosis Growth became slow
Transfection an oncogene to normal cells makes them transformed
STOP make tumors in nude mice
Rapidly divide
grow in soft agar; Start growing in multilayer
Stop grow in soft agar; grow only in monolayer
Morphology changes (more ECM and attachment)
Make tumors in nude mice
Morphology changes (less ECM and attachment)
Morphology changes after TSG transfection (MAD = Myc antagonist)
Less cells; shape not “round”
Morphology comparison of parent cells (A), control BEL-7404M0 cells (B) and mad overexpressed BEL-7404-M1 cells (C).
http://www.cell-research.com/991/991-xyh.htm
Mad inhibit growth of cells in culture
Mock–transfected and control cells
Mad –transfected cells
http://www.cell-research.com/991/991-xyh.htm
Influence of MAD transfection on growth in soft agar
Cells BEL-7404 cell BEL-7404-M0 cell BEL-7404-M1 cell Mad -transfected Number of colonies in soft agar
1104
1031 151
http://www.cell-research.com/991/991-xyh.htm
Cell cycle analysis of Mad-overexpressing BEL-7404-M1 cells
Cells Percentage in cell cycle G0/G1 BEL-7404 cells BEL-7404-M1 cells 48.50% 63.37% S 26.95% 18.40% G2/M 24.45% 18.23%
BEL-7404-M0 cells
50.15%
24.45%
25.40%
Conclusion: MAD-overexpressing cells are in quiescence more often than control cells
NUDE MICE with tumors grown from transfected cell
Tumor cells transfected with MAD (suppressor gene) produce tumor of smaller size research.dfci.harvard.edu/ coe/project_03/
KNOCKOUT MICE
Normal (+) gene X
Genome
Isolate gene X and insert it into vector.
Defective (-) Gene X
VECTOR MARKER GENE
Inactivate the gene by inserting a marker gene that make cell resistont to antibiotic (e.g. puromycin) Transfer vector with (-) gene X into ES cells (embryonic stem)
Vector and genome will recombine via homologous sequences
Grow ES cells in antibiotic containing media; Only cell with marker gene ( without target gene) will survive
Inject ES cells with (-) gene X into early mouse embryo Transfer embryos to surrogate mothers
Resulting chimaras have some cells with (+) gene X and (-) gene X.
Mate them with normal mice
Lucky you, if germline contain (-) gene X
Screen pups to find -/+ and mate them
Next generation will split as 3:1 (Mendelian)
Evaluation of KO mice features
Cancer-involved genes often produce organismal changes.
1.
(-/-) embryos can die on early gestational stages (like it happens with BRCA 1 and 2 embryos)
In this case we will see only +/+ and +/- pups (changes in Mendelian frequency)
2. 3.
(-/-) pups can have profound defects in different systems, often problems with immunity or behavior (-/-) mice can develop more spontaneous tumors than normal ones, or are more susceptible to chemical induction of tumorigenesis