Genetic Toxicity and Cancer
Dec 22, 2015
Genetic Toxicity and Cancer
From:Mitchell L. Cohen, 2000. Nature 406, 762-767. Changing patterns of infectious disease.
cancer
cancer
Molecular Biology of Cancer
Events in ToxicosisBiochemical Developmental Genetic
Access to the organism
Absorption into bloodstream
Activation
Transport to target Various tissues Embryo DNA
Binding to target /Alteration of target
Various cell constituents
Embryonic cells and tissues
DNA
Inactivation No additional damage
Too late? Prevents damage to additional DNA molecules
Excretion No additional damage
Too late? Damage propagates regardless
Repair Significant Limited -- includes cell death. Damage often irreversible
In 1st replication or not at all.
Events in Genetic Toxicity
• Access to the organism• Absorption into the bloodstream• Transport in bloodstream• Entry to target cell
– All cells are potential targets• Binding to (altering) target
– DNA for genotoxic compounds– Variable for nongenotoxic cpds
• Activation• Inactivation
– May be irrelevant if DNA damage has occurred
• Excretion• Repair
– DNA repair ordinarily very efficient
• Critical Difference:
– DNA is self-replicating.
– Damage to DNA is propagated to all daughter cells.
– Damage can increase for decades after toxicant is gone
Reversible illness
Irreversible effects:(Neuro, Developmental, Organ failure,Diabetes ….)
Genetic Toxicity
Functional effects (e.g.: Enzyme inhibition)
Death
DNA Damage
Germcells
Somaticcells
Cancer
Cytoplasmic toxicity
Inheriteddisease
Alteredmutationrate
Cell death
Comparison of cytoplasmic and genetic toxicities
Birth defects
Premature aging
Evidence for Genetic Origins of Cancer • Historical arguments:
– Inbred mouse lines– High-cancer families
• p53: Li-Fraumeni families• BRCA2: breast/ovarian cancer
– Genetic diseases carrying a high risk of cancer• Xeroderma pigmentosa • Downs’ syndrome• Ataxia telangiectasia• Bloom’s syndrome• Fanconi’s anemia
– Correlation between carcinogenic and mutagenic chemicals
• Molecular evidence– Beginning in 1950s– Overwhelming by 1980
Cancer is a Genetic Disease• Origins
– Errors of DNA replication to somatic cells• = somatic mutations
• Nature of mutations leading to cancer– Cell must survive– Initial mutation enhances rate of cell division
• Turns off (inactivates) gene that blocks cell division– Tumor suppressor gene
• Turns on (activates) gene that increases cell division– Oncogene
• Multiple changes needed to produce most malignant tumors.– And, because the probability that both of 2 independent events
occur is the product of their individual probabilities,– It takes a long time to accumulate all the changes that make a cell
a cancer cell.
Digression: Tumors
• Tumor– Overgrowth of cells– Benign tumors
• Cells – Replicate autonomously– Do not invade neighboring tissue– Do not migrate via bloodstream or lymph
• Chromosome complement usually looks normal
– Malignant tumors• Cells
– Replicate autonomously– Invade neighboring tissues– May migrate via bloodstream or lymph to
distant parts of body– Lose characteristics of parent tissue (look
primitive or dedifferentiated)• Chromosome complement often abnormal
Progression of tumor formation
Cancer
• Cancer consists of cells that have escaped growth regulation, and that have the additional properties of– Invasiveness– Metastasis
• Tumors are clonal– They originate from a single cell– That has suffered DNA damage, causing it to divide
more rapidly.– Because DNA is self-replicating, this damage is
propagated to all progeny of the original cell.
Most Errors Occur During DNA Replication
• Semiconservative replication allows for checking of accuracy.• Uncorrected error rate ~ 1x10-6.• After proofreading, error rate ~ 1x10-9.
• Why do we not all succumb to consequences of errors of DNA replication?– Code is degenerate.– Not all genes are turned on in all cells.– Many errors lead to cell death ± replacement.
Summary
• Cancer is a genetic disease.• Cancer results from one or more genetic alterations that
– Increase rate of cell division– Allosw tumor to invade adjacent tissues– Allow cells to migrate through blood or lymph to new sites
• If a tumor requires only 1 or 2 “hits” (genetic alterations)– It will often occur in children– Example: retinoblastoma
• Sporadic = 2-hit• Familial = 1-hit
• If a tumor requires multiple ‘hits’– It will occur in adults– Incidence will increase with increasing age
Cancer is a disease of aging
• Most adult cancers require more than one alteration to DNA – All changes must occur in a single cell or its progeny
• It takes time to accumulate multiple changes in DNA• Cancers that require only a single ‘hit’ to DNA tend to
occur in children– Retinoblastoma– Leukemia
Multistage Carcinogenesis: Colon Cancer
Normal colon cell
Increased cell growth
Early Adenoma
Adenoma 2
Adenoma III
Carcinoma
Metastasis
Initiation: loss of APC on 5q
DNA demethylation
Ras gene mutationon 12p
Chromosome 18 loss
Chromosome 17 loss: p53
Other chromosome
losses
after Fearson and Vogelstein, Cell 61:759-767, 1990
How DNA Damage Leads to Cancer• 1. Translocation
– Burkitt’s lymphoma• c-myc from chromosome 8 to chromosome 14• antibody genes from chromosome 2 or 22 to chromsome 8
– Mantle cell lymphoma• t(11;14)(q13;q32) translocation• Moves bcl-1 to Ig controller• Deregulation of cyclin D1• compare to CLL (chronic lymphocytic leukemia)
• 2. Amplification of oncogene– By transfer from viral genome– c-ras in culture
• 3. Deletion, insertion or point mutation leading to
– a. Overexpression of oncogene– b. Inactivation of tumor suppressor gene
• Retinoblastoma in children– Loss of segment of chromosome 13 on both copies of the chromosome
There are also 240 subtypes of leukemia
Immunophenotype of B-cell chronic lymphocytic leukemia (B-CLL) and lymphomas that resemble it.
Antigen B-CLL Mantle Cell SLVL Follicular Lymphoma
slg Weak ++ ++ ++
CD5 ++ ++ - -
CD19 ++ ++ ++ ++
CD20 + ++ ++ ++
CD22 Weak or - ++ ++ ++
CD23 ++ - - -
CD79b Weak or - ++ ++ ++
FMC7 - ++ ++ ++
CD10 - - - ++
low-grade aggressive low-grade low-grade
Summary
• Cancer is a genetic disease.• Cancer results from one or more genetic alterations that
– Increase rate of cell division– Allosw tumor to invade adjacent tissues– Allow cells to migrate through blood or lymph to new sites
• If a tumor requires only 1 or 2 “hits” (genetic alterations)– It will often occur in children– Example: retinoblastoma
• Sporadic = 2-hit• Familial = 1-hit
• If a tumor requires multiple ‘hits’– It will occur in adults– Incidence will increase with increasing age
Cancer Is An Environmental Disease:• Differences between countries
– could be genetic or cultural– Look at immigrants
• Israel– Ashkenazi (European) Jews– Sephardic (African and Middle Eastern) Jews– Gentiles
• USA– Immigrants, their children and grandchildren
• Diffferences within countries– Hawaii
• Japanese/nonJapanese– occupation – ambient environment– cutural habits
Cancer in Israelis: 1961-1965
Group Population
Annual Incidence per 100,000 in population:Males
Annual Incidence per 100,000 in population:Females
non-Jews (Mostly Arab ethnicity) 140,000 179 93
Jews born in Israel 425,000 193 195
Jews born in Afirca or Asia(Sephardic)
291,000 208 167
Jews born in the US or Europe(Ashkenazi)
352,000 294 313
U.S. Population300-400 300-400
Cancer in Japanese: Japan and California
Cancer Site Japanese in Japan Japanese Immigrants to California
Sons of Japanese Immigrants
Stomach 6.5 4.6 3.0
Liver 3.7 2.1 2.2
Colon 0.2 0.8 0.9
Prostate 0.1 0.5 1.0
Richard Doll: Environmental Causes of Cancer(JNCI 66:123,1981)
Occupational Medical Life-Style Ambient
Aromatic amines Alkylating agents Smoking Aflatoxin
Benzene Chlornaphazine Alcohol Asbestos
Asbestos Immunosuppressants Anabolic steroids Benzene
Bis(chloromethyl) ether Busulfan Chewing tobacco Asbestos
Cadmium Estrogens Chewing lime Schistosomiasis
Chromium Phenacetin Chewing betel Parasites
Furniture-making Ionizing radiation Contraceptives?
Ionizing radiation Hormones (DES) Overeating/Obesity
Isopropanol Mfg Chemotherapy Reproductive history
Leather Mfg Sexual promiscuity
Mustard gas Dietary ingredients
Nickel (some forms)
Vinyl chloride
PAHs
UV light PAHs PAHs PAHs
Arsenic UV light UV light UV light
Arsenic Arsenic
Hepatitis b Hepatitis b Hepatitis b
Occupations with High Cancer IncidenceOccupation Type of Cancer Probable Chemical Carcinogen
Chemist brain, pancreas, bladder, lymphatic system
Coal miner stomach
Foundry worker lung benzpyrene, metals
Petrochemicals industry brain, blood, lung, bone, stomach, esophagus, blood
benzene
Painters blood solvents
Printers lung, mouth, pharynx metals? solvents?
Rubber industry bladder, blood, lung, prostate, stomach
organic volatiles
Textile industry nasal cavity, sinuses formaldehyde
Woodworkers lymphatic system terpenes, wood oils, formaldehyde
Farmers skin, lymphatic system UV light, pesticides?
Sailors skin UV light
Radiologist bone marrow gamma radiation
True or False ??
• “If you get cancer, it’s your own fault. You smoked, or ate too much, or drank alcohol.”– speaker from the American Cancer Society at a toxicology
meeting in Florida, ca 1978.
Cancer Epidemiologyand
Identifying Carcinogensand
A Digression Into Statistics
Contrast between cancer incidence and mortality, by state, 2011
http://www.cdc.gov/cancer/dcpc/data/state.htm
http://www3.cancer.gov/atlasplus/charts.html
ALL SITES:1950-1994
US, black lines
IL, tan, left;LA, tan, right
US by race & sex: all sites US by race and sex:non-Hodgkins lymphoma
colon cancer
melanoma
Brain & nervous systemStomach
Cervix Testis
Cancer Statistics
• For an idea of how convoluted the arguments can (and do!) become, see– http://katatrepsis.com/2012/10/30/why-does-breast-cancer-
research-receive-more-research-funding-than-prostate-cancer/
– (optional!!!)
Assays for carcinogenicity
• In vitro or cell-based– Ames assay in S. typhimurium– CHO cells– Other mammalian cells– Drosophila melanogaster mutagenesis– COMET assay
• In vivo or animal bioassays– Rats – Mice
• Epidemiology– Prospective– Retrospective– Ecological
COMET Assay (Single Cell Electrophoresis)
• Expose cells to mutagen – In vivo or in vitro– Can use bacteria, mammalian cells, human cells
• White blood cells from smokers vs nonsmokers
• Fix cells appropriately• Electrophoresis of single cells
– Observe migration of DNA in single exposed cells– Compare to unexposed cells
Control
Low Dose
High Dose
Example of COMET Assay Results
Mammalian Bioassays
• 2 species• 2 sexes per species• 50 animals per dose• 3 doses + vehicle control• Begin dosing animals soon after weaning• Dose for 18 months• Hold for 6 months• Examine every tissue and organ for signs of tumors
– Benign– Malignant
• Interpretation– Not without problems
Mammalian Bioassays
• Advantages:– Reliability
• Adequate data exist on 400 chemicals– Carried out by NTP
• Problems:– Discordance between species or sexes
• Fluoride– Bone cancer in male rats
• Disadvantages:– Cost
• $5,000,000 ??– Time
• 2-3 years– Manpower
• Histopathology!• Animal care• Record keeping
– Extrapolation • Between species• To lower doses
Extrapolation Outside of Dosing Levels
• Humans are exposed to many chemicals at low levels– ppm – ppb
• There are 300,000,000 people in the US today– An incidence of “1 in a million”
= 300 cases• How many mice and rats can we test?
– 50 per dose per sex– 2 sexes x 2 species x 4 doses = 16 x 50 – = 800 animals / bioassay
xx
x
% Response
Log Dose
Laboratory studies maximize dose because small numbers of animals are treated.Extrapolations have to made outside of these dosing levels to the actual exposure of 300,000,000 people.
x
How do you extend the line from dosing levels to control?
experimental doses
Control
human exposurelevel
xx
x
% Response
Log Dose
Linear response; no threshold
xx
x
% Response
Log Dose
Linear response with threshold(a linear interpolation using the best fit to the 3 points.)
xx
x
% Response
Log Dose
Nonlinear response, no threshold
xx
x
% Response
Log Dose
Nonlinear response with threshold
Human Epidemiology Studies
• Types– Retrospective– Prospective– Ecological
• Advantages of each– cost– time– reliability
• Disadvantages– Cost
• financial• human
– time– reliability
Lies, Damned Lies, and………..…
Statistics
Statistics are like a bikini: what is revealed is interesting; what is concealed is crucial.
A. R. Feinstein, Science 242:1257-1263, 1988
…. Some Data
Article: case reports– 6 cases of probable DEET-induced convulsions – all in girls under 6 years of age– 3 girls died
• What conclusions can we draw?
…Some More Data …
• Bendectin was sold by the same company that wanted to market thalidomide in the US.
• Bendectin was reportedly effective against nausea of pregnancy.
• In the U.S. in 1980, numerous children whose mothers took bendectin during pregnancy were born with limb defects.
• The lawsuits argued that bendectin caused limb defects, and eventually Merrill took it off the market.
• What is wrong here?
Bendectin and Birth Defects
• Births in the US in 1980 = 3,612,258• Prescriptions for bendectin = 903,064• Incidence of limb defects (reduction):
– 5 per 10,000 livebirths– = ~ 451 limb reduction defects in 1980
• If 1/4 of pregnant women took bendectin• And if bendectin does NOT cause limb defects:
– then an estimated 100 women who took bendectin had children with limb defects.
And it might be true…. ?
• My friend Beverly has compiled a list of 20 people in her neighborhood who have cancer.
• When she was young, she knew of only 1 case of cancer in her entire town, so clearly cancer is on the rise.
• Pesticides came in as she was growing up, and everyone in the neighborhood uses pesticides on their lawns.
• Beverly is convinced that lawn pesticides cause cancer.
Tricky Percentages: Absolute vs Relative Risk
• The headline says:• “TAMOXIFEN REDUCES RISK OF BREAST
CANCER RECURRENCE BY ALMOST 50%”
• A woman with an invasive tumor < 1 cm in diameter, ER+PR+ HR-, treated with either mastectomy or with lumpectomy plus radiation, has a recurrence risk of 17% over 10 years.
• If she takes tamoxifen for 5 years, her recurrence risk drops to 9%.
• What is the risk reduction: 8% or 47% ???
And what about the controls?
• At Love Canal, the incidence of chromosome abnormalities was significantly higher than the U.S. “background” level.
• The data were not allowed because the investigators did not use matched controls from the area around the affected Love Canal neighorhood.
• Question: are people living in somewhat-less-polluted neighborhoods of a very polluted city really the right controls to see if a subset of neighborhoods in that city are unfit for human habitation?
One Last Example …
• When computers were still quite new, and scientists reveled in their new ability to test multiple factors at once, an ecologist looked at 500 factors affecting tree growth.
• He gathered data for 10 years, put all the numbers into the computer, and found – 5 factors that were significant at the
0.01 level - which is highly significant.
• Should we be impressed?
And the moral is ….?
• Don’t believe anything you read.
• There are lies, damned lies, and statistics.
• Environmentalists are all crazy.
• Really good results are obvious enough that you don’t need statistics.
Critical Reviews of Health Data
• How do you know if one treatment is better than another?• How do you know if a test will do more harm than good?
• Cochrane Reviews– http://community.cochrane.org/cochrane-reviews
• Bandolier– http://www.medicine.ox.ac.uk/bandolier/band71/b71-2.html
ASSIGNMENT – DUE Monday 3/2/15Select one of the following topics:
Breast Cancer Prostate Cancer
• Discuss the benefits and costs of the PSA assay to identify early prostate cancer.
• Under what circumstances is this test advisable?
• Discuss the benefits and costs of yearly mammograms to identify early breast cancer.
• Under what circumstances are yearly mammograms advisable?