2022 TSGs/oncogenes

Question 1

Templeton et al 2014

Answer 1

• Meta-analysis.
• Looked at 16 studies comprising 11, 056 pts.
• There was sig association between overexpression of RTKs and decreased 5 year OS in women with breast cancer.
• Worst overall survival was seen with overexpression f FGFR and EGFR/HER1.
• Before this study, the HER2 overexpression was known to have an impact on prognosis but other GF receptors were not.
• Limitation - no correction could be made for other prognostic factors, such as tumour grade - so actually, it remains unknown whether overexpression of the RTK described here are truly independent adverse prognostic features

Question 2

Konduri et al 2016

Answer 2

• Identified EGFR fusions as novel therapeutic targets in lung cancer.
• To determined the frequency of EGFR mutations in lung cancer, they analysed data from ~10,000 clinical cases.
• Fusion events were detected in 5 patients with metastatic lung cancer.
• NGS revealed this fusionwa swas commonly a fusion of the EGFR-RAD51 fusion (where RAD51 is a protein involved in DNA damage responses).
• This could have just been a passenger mutation.
• They confirmed this fusion was oncogenic in vitro - They transfected NR6 cells (which lack the endogenous EGFR) with EGFR-RAD51 and this significantly increased colony formation of NR6 cells in soft agar when compared to control transfected cells.
• In cell culture - went on to show that a variety of existing EGFR inhibitors were able to target EGFR-RAD51 - inhibiting proliferation of EGFR-expressing cells.
• However, these effects were just seen in vitro - still could be a passenger mutation and not actually contribute to oncogenesis. Colony formation and proliferation also does not necessarily mean cancer

Question 3

1984 Weinberg

Answer 3

isolate Neu, Her2 homologue from carcinogen induced neuroblastomas in rats

Question 4

Margetuximab

Answer 4

Anti-HER2 antibody with increased affinity for both the low-affinity and high-affinit forms of CD16A, and Fc receptor that is important for the engagement of ADCC (antibody-dependent cellular cytotoxicity) against tumour cells

Question 5

Bang et al 2017

Answer 5

• phase I study involving patients with HER2-positive advanced-stage solid tumours
• In this study, 50% (30 of 60) had stable disease.
• The majority of those who had stable disease or better had disease progression on previous HER2-targeted therapies.
• Ex vivo analyses of peripheral blood mononuclear cell samples confirmed the ability of margetuximab to support an enhanced level of ADCC compared with trastuzumab.

Question 6

1964 Jennifer Harvey

Answer 6

observed that a preparation of a murine leukaemia virus, taken from a leukaemic rat, induced sarcomas in new-born rodents

Question 7

Mulcahy 1985

Question 8

Filmus et al 1994

Answer 8

demonstrated this when a relationship was seen between activated RAS and overexpression of cyclin D1 in epithelial cells from the rat intestine.
o In their study, they transfected intestinal cells with an inducible RAS expression vector.
o They treated these with anti-sense cyclin D1 oligonucleotides and this led to a reduction in rate of cell proliferation.
o This led to the conclusion that RAS-induced cyclin D1 contributed to the higher rate of proliferation caused by the RAS oncogene in intestinal epithelial cells

Question 9

Mituhashi et al 1995

Answer 9

o They observed a marked upregulation of VEGF mRNA and secreted functional protein in human colon cancer that expressed mutated KRAS
o Furthermore, in two cell lines where KRAS was rendered non-functional, ELISA assay showed a 4-fold and a 5-fold decrease in VEGF protein; a similar result was seen when pharmacological disruption of the mutant HRAS allele led to suppression of VEGF

Mainly correlative

Question 10

AMG510

Answer 10

an inhibitor targeting KRAS G12C for non-small cell lung cancer. A phase I trial published in May 2019 showed response rates of around 50%. Although this is promising, further research is needed to evaluate drug resistance, evaluate drug efficacy and target other RAS mutants

Question 11

Davies 2002

Answer 11

Did a genome-wide screen in cancer cell lines - DNA seq - found that missense mut in BRAF occur in 66% of malignant melanoma. They used a kinase assay to demonstrate that all four mutant found had elevated basal kinase activity compared to WT BRAF.

Question 12

Chapman 2011

Question 13

Bar-Peled et al 2013

Answer 13

• First, they did coimmunoprecipitation experiemnts human embryonic kidney cells expressing FLAG-tagged RagB with a chemical crosslinker.
  Allowed them to identify GATOR complex (with two subcomplexes GATOR1 and GATOR2) that were coexpressed with Rags.

Then used RNAi in HEK cells and Drosophila S2 cells to examine the function of each GATOR component in amino acid sensing by mTORC1 and dTORC1 respectively.
Depletion of GATOR1 or their Drosophila orthologues prevented the inactivation of mTORC1 and dTORC1 normally caused by aa deprivation.

They then invetsigated the role GATOR1 plays in human tumours.
They looked at pubically available data from TCGA - found a subset of glioblastomas and ovarian cancers with nons or frameshift mutations or truncating deletions in genes encoding GATOR1 proteins (DEPDC5 and NPRL2).
THey used Cosmic resources to identify human cancer lines with deletions in DEPDC5, NPRL2 and NPRL3. They then found htat in these lines mTORC1 signalling was insensitive to aa deprivation.
When DEPDC5 and Nprl2 were reintroduced into the cancer cell lines lacking them, the mTORC1 pathway regained sensitivity to aa regulation.

Question 14

Colman et al 2009

Question 15

Descamps et al 2005

Answer 15

A previous study of alternate-day fasting, which was performed in middle-aged mice for a total of 4 months, also found that fasting reduced the incidence of lymphoma, bringing it from 33% (for control mice) to 0% (in fasted animals). There was no difference in overall food consumption between the ADF and AL mice - this effect was not due to CR. 25 mice total

Short duration of study - did fasting prevent or simply delay tumour inset?

Question 16

Kerr, Wyllie and Currie in 1972

Answer 16

Noticed it was a morphologically distinct form of cell death - used electron microscopy - the apoptosis they observed was characterised by nuclear condensation and cellular fragmentation, followed by phagocytosis of the fragments by nearby cells

Question 17

Sulston 1976

Question 18

Horvitz et al 1983

Answer 18

Identified two genes in C.elegans, which impaired this programmed cell death - ced-3 and ced-4. The showed that functional ced-3 and ced-4 genes are a prerquiste for cell death to be executed.This was the first evidence that programmed cell death was genetically encoded. We now know that most genes involved in controlling cell death in C.elegans have counterparts in humans

Question 19

Evan et al 1992

Answer 19

They looked at Rat-1 fibroblasts (these have the property of being dependent on mitogens for their signalling). Overexpressed c-myc in Rat-1 fibroblasts constitutively via preparation of recombinant retroviruses. They found that the growth curves of all cells was very similar, irrespective of whether the cells were constiutively expressing c-myc or in they were controls. This indicated that constituive c-myc expression has no effect on the ability of Rat-1 cells to slow growth in low serum levels.

However, Evan then looked at the cell cycle distribution of these cells following withdrawal from serum for 48h. They examined the growth of WT Rat-1 cells and Rat-1/myc cells cells by DNA content BrdU incorporation by flow cytometry. They found that the control cells were not proliferating and had DNA content that was equivalent to G1 phase. In contrast, they found that the c-myc cells, althoguh not increasing in cell number - 45% of these were in S phase. Therefore, the cells were proliferating but without increase in cell number. Time-lapse video microscopy revealed cells were dying by apoptosis - this was rather unexpected. Showed correlatively using 3 different myc-expressing clones that apoptosis was dependent on Myc expression

Perceptive authors already made the link with p53 as an inducer of apoptosis. But they do get wrong that c-Myc is in the final common death pathway - predict this because they saw it overexpressed in a lot of cancers

That such an observation was seminal, but doesn’t quite capture the specificity and comprehensiveness of the cellular anti-cancer pathway

Question 20

Serrano et al 1997

Answer 20

One of the mutant alleles of HRAS is G12V, Transduced human, mouse embryonic and rat embryonic fibroblasts with a retrovirus encoding HRAS G12V or a control virus. They observed that introducing G12V abolished the increase in cell number that we see in the control culture - i.e. the cells stopped proliferating. They then used a propidium iodide-BrdU flow cytometry method to determine DNA content per cell as well as newly synthesized DNA in the fibroblasts. It was seen that there was a significant decrease in the S phase fraction following introduction of the mutant RAS allele. Even on day 1 after tarnsductoin we can see a decrease frim 31% ti 5.4% and by 3 days we can see the S phase fraction is almost completely gone. Therefore, mutant RAS in these cells led to cell cycle arrest, rather than proliferation

How this was working at a molecular level. They did Western blotting (without a loading control - caveat - could have had unequal amounts of protein loaded) of lysates from fibroblasts transudced with retrovirus encoding HRAS G12V or control retrovirus. They found a pretty dramatic increase in p53 levels in all three types of fibroblasts following G12V transfection. Conversely, there is also a decrease in cyclin A abundance. Rb decreases and cyclin A decreases - interpretted as dephos of Rb. This suggests that induced of mutant RAS is halting the cell cycle through a combination of activation of p53 and dephos of Rb.

Question 21

Patton 2005

Answer 21

Found that BRAF cooperates with p53 in the genesis of melanoma. Generated zebrafish expressing the most common BRAF mutant form under the control of a melanocyte promoter. Expression of the mutation, and not the WT lead to dramatic patches of ectopic melanocytes (nevi). In p53 deficient fish, activated BRAF induced formation of melanoma lesions that rapidly developed into invasive melanomas.

Question 22

Montes de Oca Luna et al 1995

Answer 22

It was already known that p53 heterozygote mice have elevated cancer incidence. You can breed these p53-/+ mice to generate p53 -/- mice - these are v cancer prone and die mostly from lymphomas by 6 months. In contrast, if we knock out Mdm-2 - we find that this is an embryonic lethal mutation. The surprising result ehre however is that if we combine the Mdm-2 KO and the p53 KO mice - these are viable. Suggests action of Mdm2 is due to action of p53. This astonishing result is a clear piece of evidence that the major function of Mdm2 is to limit the activity of p53 (whats happening in the Mdm2 KO mice alone is that they are dying of massive p53 induced apoptosis)

Question 23

Bond et al 2004

Answer 23

Genomic DNA from 50 healthy volunteers PCR amp and seq the MDM2 promoter. They found a SNP in the region. They used a computer algorithm to reveal several putative binding sites for TF Sp1. Analysis in subsequent blotting experiments revealed this SNP increases affinity for transcriptional activator Sp1, resulting in higher levels of MDM2 RNA and protein and subsequent p53 attenutation (from ChIP data with Sp1 antibody). Presence SNP was then shown to correlate with accelerated tumors in humans (both hereditary and sporadic)

Why didn’t they look at DNA from volunteers with cancer instead? Maybe this process is different between the two populations. They have only shown this SNP correlates with cancer.

Question 24

Dannenberg et al 2000

Answer 24

Generated ES carrying single or compound loss of differentiation mutations in Rb gene family
- In TKO (Rb -/-, p10 -/-, p130 -/-) fibroblasts lacking all three pocket proteins, there was a loss of R point control
o WT withdraw from cell cycle in low serum (0.1%)
o TKO > healthy abundance of cells in s phase
o Lacked G1 control, increased turnover in growth-inhibiting conditions. Shows the R point monitors the presence or absence of mitogens

Question 25

Bartkova et al 2005

Answer 25

They transfected osteosarcoma cells with inducible cyclin E or E2F1 (overexpression drives cells into S phase and so is a way of mimicking downstream effects of oncogene activation). They used a tet inducible system to do this for cyclin E and tamoxifen for E2F1. In both cases, immunofluorescence microscopy showed that there was upregulation of DNA damage markers stained with antibodies - these were gamma-H2AX, Ser15-phosphorylated p53 or phosphorylated Chk1. Induction of DNA damage associated markers as a consequence of induction of cyclin E and E2F1.

They used an osteosarcoma cell line - this is already a tumour cell

Question 26

Foijer et al 2005

Answer 26

Found that, in conditions of serum starvation, triple knockout (TKO) mouse embryonic fibroblasts of all three pocket proteins, were able to pass through the R checkpoint. Showed using time lapse videomicroscopy that these cells became very sensitive to apoptosis. (SImilar to Evan 97). Then showed they were subsequently arrested at G2. This was even more dramatic if we suppress the apoptosis in those populations by overexpressing BCL2 oncoprotein (which is antiapoptotic). This suggested that loss of pocket protein function is insufficient to allow cell cycle progression and something else that permit evasion of G2 and apoptosis, are also needed for tumorigenesis. What was that something else that is leading to the arrest of these cells in G2? - p53! They then used RNAi to knock down p53 - saw saw that these cells were undergoing proliferation in the absence of serum. Shows loss of restriction point controls needs to collaborate w p53 function. Also shows there is a second safety net in G2.

Limitation - could not tell if they were in G2 or M using their flow cytometry approach- now we know they were in G2. These cells were also heavily modified - could maybe stress the cell

Question 27

Hockenbery et al 1990

Answer 27

Provided a direct link between BCL-2 and apoptosis with the obervation that BCL-2 did not simply have a prosurival function, but rather specifically prevented apoptotic death in cells deprived of GF. Transfected BCL-2 into a murine lymphocyte line. Found that this overexpression significantly extended the lifespan of these cells when deprived of GFs, in comparison to those that those that were no overexpressing it. Examination of the dead cells revealed death was characteristic of apoptosis

Question 28

Strasser et al 1990

Answer 28

They showed that mice overexpressing bcl-2 and myc show hyperproliferation of pre-B and B cells and develop tumours much faster that only myc-overexpressing mice

Question 29

Beroukhim et al 2010

Answer 29

Studied copy number alterations using an Affymetrix array - comparing signal intensities from each array to >1000 paired normal tissue specimens. Looked at 3131 cancer specimens across more than two dozen cancer types. Found amplification of MCL1 and BCL2L1 among the most frequent chromosomal gains - these are anti-apoptotic.

Doesn’t distinguish between the alterations that drive cancer growth from those accumulating during tumorigenesis

Question 30

Blombery et al 2019

Question 31

He et al 2007

Answer 31

Examined miRNA expression profiles in WT and p53-deficient mouse embryonic fibroblasts. Used semi-quantitative RT-qPCR to measure expression of a panel of 150 mouse miRNA in WT of p53 KO MEFs. The expression of miR-34 was found to be precisely correlated with p53 status. THen did CHiP - in WT MEFs follwoing induction of p53 activity through DNA damage,regions of the genes encoding miR-34 that contained putative p53 binding sites were enriched in p53 immunoprecipitates. This enrichment was absent from p53 KO MEFs. So miR34 is a durect transcriptional target of p53. Althoguh the data was not show, authors state that delivery of miR34 sensitised MEFs to apoptosis in response to genotoxic stress - however they did not show a role on Bcl-2 directly. However, later that year another group showed that miR-34 induction in SW480 cells, this was correlated with downregulation of transcript levels of BCL2, though again this was only correlative

Question 32

Daige et al 2014

Answer 32

Systemic (IV) delivery of MRX34 (miR-34 mimetic) in mice bearing liver tumour xenografts resulted in about 1000-fold increased expression of miR-34a and the inhibition of tumour growth. Furthermore, MRX43 induced tumour regression in more than 1/3 of mice. They did not observed toxicity in at any dose tested.

They used an orthotopic liver cancer cell model - used immortalised human cancer cell lines - in vitro passaging of cells and lead to unrepresentative tumour histology and heterogeneity - poor predictive relevance in later clinical development

Question 33

MRX34

Answer 33

Results of a phase 1 study published in 2020. Enrolled 85 patients with solid tumours. The trial recruitment closed early due to serious immune mediated adverse effects - RESULTED IN 4 PATIENT DEATHS (different from animal study). Despite this, they found a 4% overall respoonse rate - clinical development of has halted. They dont know what caused these adverse effects.

Question 34

APR-246

Answer 34

FDA has recently granted breakthrough therapy designation for the treatment of patients with TP53 mutant myelodysplastic syndromes (MDS) with the comvination of APR-246 and azacitidine and the randomised phase III study of APR-246 and azacitidine versus azacitidine is ongoing in MDS patients. Trial published May 2021. 44% pts in complete remission.

Question 35

Xin Lu et al 2001 (Oxford)

Answer 35

ASPP1 was identified in a genome database search for proteins with homology to the previously isolated protein called 53BP2/Bbp - which had previously been shown to interact with p53. They UV-irradiated breast cancer cells and immunoprecipitated ASPP1 and ASPP2.Western blot analysis of the immunocomplexes showed that p53 interacts with ASPP1 and ASPP2. Osteosarcoma cells were transfected with a p53 expression plasmid or a control plasmid together with cell surface marker CD20. FACS analysis showed that the expression of p53 caused about 17% apoptosis among the transfected cell population (Figures 2A and 2B). Expression of ASPP1 or ASPP2 in the absence of p53 caused a small fraction of the cells to die by apoptosis. In contrast, coexpression of p53 with ASPP1 or ASPP2 caused 50% of the transfected cells to die of apoptosis. ASPPs seemed to enhance the apoptotic functions of p53.

In 2003, they reported the identification of the third member of the p53-regulating protein family-iASPP (inhibitory member of the ASPP family). They knocked down the gene function using antisense RNA in human cell lines and saw an increase in apoptosis. This finding indicated that the newly identified ASPP protein had an inhibitory effect on apoptosis, earning it the name iASPP

Question 36

Nunney et al 2018

Answer 36

Nunney analyzed data from four large-scale studies totaling hundreds of thousands of cancer patients. He found that every additional 10 centimeters (4 inches) in height was associated with a 10 percent increase in cancer risk. Nunney states that tall people are thought to have slightly higher circulating level of IGF-1

Question 37

Martincorena et al 2015 (in Cambridge)

Answer 37

Addresses the number of cancer mutations in phenotypically normal skin. They source of skin was eyelids from pts who had undergone surgery. They then did ultra-deep (NGS) sequencing of 74 cancer relevant genes from 234 biopsies of these phenotypically normal eyelid epidermis speciments from four individuals. The burden of mutations in normal skin in just 74 genes is about five mutations per megabase - this is only a factor of 10 less than seen in squamous cell carcinoma and is comparable to mutation burden seen in cancers of the breast, head and neck.. To determine if these mutations were giving a selective advantage to the normal skin, Martincorena used the ratio of mutant sequence reads that do or do not alter the amino acid. . Six genes—NOTCH1, 2, and 3, P53, FAT1, and RBM10—had an excess of nonsilent mutations, indicating that these genes had been selected for and are therefore presumptive drivers. PThe first five are known drivers for keratinocyte tumors and the last is mutated in other cancers. A seventh gene, FGFR3, had recurrent activating mutations. In total, one-quarter of middle-aged skin contains a mutation in one of these drivers.

They only looked at 74 genes in particular - there could be even more

Question 38

Martincorena et al 2018

Answer 38

Skin on the eyelids was exposed to UV light. They wanted to look at tissue that had not been exposed to such powerful mutagens. They performed targetted gene sequencing of normal oesophageal epithelial from nine human donors of varying age. The mutation rate was lower in the oesophagus than in the skin but there was a strong positive selection of clones carrying mutations in 14 cancer-associated genes. Interestingly, mutations in the cancer driver gene NOTCH1 were more common in normal oesophageal epithelium THAN IN NORMAL CANCER. They also noted that the number of mutations increases with age, mainly owing to errors in intrinsic processes such as aberrant DNA replication or repair

Only 9 indivduals

Future

• entire genome
• other tissues (particularly those associated w tumours e.g. lung)
• sample individuals over time - how do these mutations lead to cancer progression?

Question 39

Nicolas et al 2003

Answer 39

They used mouse genetics to see how Notch signalling ablation affects development of SCC. They used the DMBA-TPA mouse model of cutaneous chemical carincogenesis system. They found application of DMBA-TPA to Notch-/- skin results in a dramatic increase in tumour burden. They used a tamoxifen-inducible model of Notch activation as it is embryonic lethal. Only one in 10 of the littermate controls developed a single papilloma. The induced Notch1-/- had an average of 14 tumours in 25 weeks after starting treatment.

Question 40

Wodarz et al 2007

Answer 40

Describe a relation between cell death and generation of mutant cells within a population after a furst wave of clonal expansion. In their mathematical model, they find that less cell death correlates with fewer divisions during clonal expansion, thus leading to a less variable cell population. In contrast, hgih death rates correlate with more cell divisions during expansion causing appearence of many different mutants. With increasing sub-clonal variability, the risk that individual cells overcome selective barriers and progress towards malignancy increases. Do these in silico models on cell-death stimulated tumour progression find counterparts in vivo?