Oncogenic Subversion and Signal Transduction

Question 1

How can cells sense their extracellular surroundings?

Answer 1

• Via molecules dissolved in water or in air that bind to cell surface receptors or enter cells through channels or transporters
• Via extracellular solid substrates with which they interact
• Via mechanical interactions
• Via monitoring light, temperature, pressure, movement etc.

Question 2

How can cells sense neighboring cells?

Answer 2

• Via direct cell to cell junctions through which molecules are exchanged; tight junctions between epithelial cells which can exchange molecules
• Via exchange of diffusible molecules that bind to cell surface receptors or enter cells through channels or transporters
• Via mechanical interactions

Question 3

What are the three layers of a signalling network?

Answer 3

Input layer that corresponds to receptors, core process results in signal transmission, with the output layer effecting the response directly.

Question 4

What are the three ways of looking a signalling networks?

Answer 4

Nodes: Focus on a single signal transduction element and links to other elements e.g. Ras

Modules: Focus on a group of signal transduction elements, often linked in different cellular responses e.g. PLC

Pathways: Focus on transduction elements that link input signal to cellular machinery generating a response e.g. MAPK

Question 5

How is the great degree of diversity that enable signalling networks’ capabilities produced?

Answer 5

A huge number of genes are involved in signalling;

Number of human genes ~20,000 
Some genes for signal transduction components 
•	Receptors ~1,500 (GPCRs ~350) 
•	Protein Kinases ~520 
•	Protein Phosphatases ~150 
•	Transcription Factors ~1,800 

Due to alternative splicing and post-translational modifications, number of components in human proteome is 10-100 fold higher compared to genome

Combinatorial considerations - interactions between components allow for a great degree of combinatorial control and functional states, many would be involved in many different scenarios

Contextual specificity - cell type and state limits the content of signal transduction components

Question 6

What is the timescale of individual signalling events?

Answer 6

Signalling through the network (s):

Kinase/phosphatase reactions………… 10^–3
Protein conformational changes……… 10^–3
Cell-scale protein diffusion (passive)..100–10^1
Cell-scale protein diffusion (active)….. 10

Question 7

What is the timescale of cellular scale responses to signalling?

Answer 7

Response mediated by cell machinery (s):

Cell migration 10^0–10^2
Transcriptional control 10^2
Cellular growth 10^4

Question 8

What is Oncogene Addiction?

Answer 8

This describes the phenomenon by which some cancers that contain multiple (genetic and epigenetic) abnormalities remain dependent on/addicted to one or a few genes for both maintenance of the malignant phenotype and cell survival.

Question 9

What lines of evidence are there for oncogene addiction?

Answer 9

Evidence that supports the concept of oncogene addiction has now been obtained in three diverse systems:

1) Genetically engineered mouse models of human cancer
2) Mechanistic studies in human cancer cell lines
3) Clinical trials with specific molecular targeted agents

Question 10

What is implied by the study of oncogene addiction in human cancer cell lines?

Answer 10

That human cancer cells in culture retain the oncogene addiction of their source tumour is notable given the significant genetic drift caused by continuous culture - addiction may be beneficial or even essential for tumour survival.

Question 11

How can oncogene addiction be exploited?

Answer 11

Inactivation of said gene can lead to apoptosis.

Question 12

How do oncomirs relate to addiction?

Answer 12

Oncomirs have emerged recently as important players in cancer (Esquela-Kerscher and Slack 2006). Their role is highlighted by the demonstration that antisense inhibition of miR-17-5p and miR-20a leads to apoptosis of lung cancer cells overexpressing the microRNA miR-17-92 (Matsubara et al. 2007).

Thus, “oncomir addiction” may be yet another manifestation of oncogene addiction. However, the potentially pleiotropic action of oncomirs may prove to substantially limit the therapeutic opportunity that such addiction might otherwise present.

Question 13

What are examples of oncogene addictions in specific cancers?

Answer 13

• Small GTPase Ras in lung cancers and melanom
• Transcription factor Myc in pancreatic cancer and osteogenic sarcoma
• Ser/Thr Kinase B-Raf in melanoma
• Tyrosine Kinase Receptor EGFR in head and neck, colorectal and pancreatic cancers
• Tyrosine Kinase Receptor HER-2 in breast cancer (approved treatment - trastuzumab)
• BRAF oncogene in BRAF Mutant Melanoma (PLX4032 treatment in trials)

Question 14

What is the main limitation of therapies exploiting oncogene addiction?

Answer 14

It quickly selects for cells within the tumour that are less vulnerable, causing the tumour to quickly evolve and rewire its signalling pathways (often by bypassing nodes) to no longer be dependent upon the oncogene.

Combination therapy is a potential way to overcome this, but the interaction of the drugs and their side-effects may be problematic.

Question 15

What receptor types are important in cancer?

Answer 15

Serpentine receptors (GPCRs) - Eg Wnt signalling via frizzled, chemokine receptors

Integrin receptors

Protein tyrosine kinase receptors (RTKs and Receptor Ser/Thr Kinases)

Question 16

What are some examples of RTKs and Receptor Ser/Thr Kinases?

Answer 16

RTKs - growth factor receptors such as EGFR, PDGFR, FGFR

Receptor Ser/Thr Kinases - TGF-B Receptor

Question 17

How is Wnt signalling important in cancer?

Answer 17

Wnt can signals through the Frizzled receptor, but coreceptors such as LRP 5/6 can lead to different responses.

Frizzled signalling is involved in motility, invasiveness and self-renewal (see 3014 zebrafish 2).

LRP signalling has a critical role in cancer pathogenesis, especially in the epithelial-mesenchymal transition (EMT).
Unlike the Frizzled receptor, this is not a GPCR, but allows for interactions with other proteins enabling other, non-canonical Wnt pathways to be activated via Disheveled.

Question 18

What are integrins?

Answer 18

Integrins are transmembrane receptors that are the bridges for cell-cell and cell-extracellular matrix (ECM) interactions (both in terms of attachment and signalling)

These are heterodimers with alpha and beta subunits both of which penetrate the membrane to produce small cytoplasmic domains that often couple to the cytoskeleton via RTKs FAC and SRC and adaptor protein p130Cas

Question 19

How are GPCRs activated?

Answer 19

GPCRs binding structure consists of two large extracellular ligand binding domains. Because the ligands are generally present only at very low concentrations, the flexible and jutting N domain acts as an affinity trap, very strongly binding the ligand by its C-terminus.

The ligand, usually a disordered peptide in solution, is often found to have a turn structure and adopt a helical structure once bound to the N domain.

This then allows the N-terminus part of the ligand to bind to the higher KD J domain, which causes activation.
The J domain connects via several bundles of transmembrane helices to a G-protein binding domain, activating it through a conformation change.

Question 20

What is the general structure of a kinase domain?

Answer 20

A large C-lobe and a small N-lobe which binds the ATP initially. Between the lobes is a substrate-peptide groove which forms the active site upon closure of the lobes as this properly positions the ATP and substrate.

A central ‘activation loop’ controls the activation state, and the alpha-C helix must also adopt a specific conformation for activity.

Question 21

Symmetrical RTK activation. How does that work?

Answer 21

Ligand binding leads to homodimerisation of the RTKs to bring two intracellular kinase domains together to trans-phosphorylate each other.

The first round generally phosphorylates the activation loop within the kinase domain to relieve inhibition, allowing them to phosphorylate more residues in the C-terminal tails.

Question 22

What is asymmetrical RTK activation?

Answer 22

Some RTKs, such as HER3 (AKA ErbB3, of the EGFR family) lacks kinase ability. It can however form head to tail dimers upon ligand stimulation with active RTKs (HER2 in this case, which has no known ligands so may exist only to colocalise in this way).

The HER3 inactive kinase domain is called the activator kinase as - without phosphorylating it - it activates the reciever kinase domain.

Question 23

How are RTKs mutated in cancer?

Answer 23

Generally they are made to be constitutively or over-active, and less ligand dependent.

Question 24

What are the two classes of oncogenic RTK mutation?

Answer 24

Ligand dependent and ligand independent.

Question 25

What are the ligand dependent RTK oncogenic mutations?

Answer 25

These can be paracrine or autocrine.

Paracrine relies on exogenous production of the ligand, but increases specificity/affinity of the binding domain to increase activation.

Autocrine mutations lead to RTK ligand binding loops recognising a ligand secreted by the tumour cell itself, leading to constitutive activation.

Question 26

What are the ligand independent RTK oncogenic mutations?

Answer 26

Overexpression of the RTK due to gene amplification or abberant transcriptional regulation.

Activating mutations - some point mutations can lead to increased dimerisation or kinase domain activation.

Fusion proteins - Chromosomal translocation can result in fusion of the kinase domain to a dimerization domain from another protein leading to constitutive kinase activation.

Question 27

What is vIII EGFR?

Answer 27

This is an oncogenic variant of the EGFR receptor, which is without its N-terminal extracellular domains 1 and (most of) 2 due to deletion of exon 1. It is common in glioblastomas.

It cannot dimerise through the wild-type mechanism, but has constitutive low level kinase activity. This is thought to be due to misfolding which leads to accumulation in the ER, and subsequent EGF-independent activation.

Question 28

How can RTKs be involved in therapy?

Answer 28

RTKs are prominent oncogenes to which tumours can become addicted. Their activity can be decreased by ligand binding inhibition using antibodies, or by kinase domain inhibition using small molecule inhibitors.

Question 29

What is a large challenge in RTK-targeting therapies?

Answer 29

RTKs are liable to develop resistance to small-molecule kinase domain inhibition, due to mutation in the ‘gatekeeper residue’ of the active site altering the binding properties of the inhibitor’s binding site.

Question 30

Give an example of a gatekeeper mutation.

Answer 30

This is how resistance to erlotinib developed in EGFR receptors (Daub et al 2004). Mutation of Thr-766 (the gatekeeper residue) to a larger residue prevented erlotinib binding to the hydrophobic pocket through steric clash without affecting ATP, allowing resumption of EGFR signalling.

Question 31

What are second messengers? Give examples.

Answer 31

Molecules that interact with the signal transducer (receptor) and are modified to carry the signal intracellularly.

• cAMP, cGMP
• Lipid-derived (DAG, IP3, ceramide)
• Lipids (PIP3)
• Calcium
• NO

Question 32

What is interesting about cAMP as a second messenger?

Answer 32

cAMP can act as a first messenger by activating GPCRs (in Dyctiostelium) and as a second messenger by activating

• cAMP-dependent protein kinase (PKA)
• cNucleotide-regulated ion channels
• cAMP- regulated GEF (Epac2)

Question 33

What is the structural nature of protein signalling components?

Answer 33

They are modular, comprised of enzymatic/catalytic effector domains and interaction domains that can be regulatory and provide interaction sites with proteins, lipids and nucleic acids.

Question 34

What are the responsibilities of interaction domains?

Answer 34

These can mediate their interaction constitutively or in a signal-induced manner (eg PTMs, conformational change).

They can interact with regulatory molecules, mediate sub-cellular localisation or contribute to the inhibition or activation of the catalytic domain.

A single interaction domain may bind multiple ligands and vice versa.

Question 35

What can interaction domains interact with?

Answer 35

The huge variety of common interaction domains are still being mapped.

Some such as SH2 and PTB bind specifically to modified peptides (in this case p-Tyr).

Others such as PDZ bind to specific domains/surfaces.

Others such as PH domains bind to phospholipids.

Yet more bind to peptide regions, such as FHA which recognises proline rich sequences.

Others bind RNA (PUM) or DNA (Tubby).

Question 36

How is the signalling potential of PIP mediated?

Answer 36

Phosphatidyl Inositol Phosphate has three phosphorylable sites. Each can be phosphorylated or dephosphorylates to produce a large number of different conformations, each with the ability to bind different interaction domains.

For example, PH and FERM domains have a small binding pocket so to be sensitive to the phosphorylation state. The result of this is often membrane localisation due to the membrane embedded lipid of PIP.

Question 37

What are examples of a domain/surface binding interaction domain?

Answer 37

The Ras Association/Ras Binding (RA/RBD) fold. This is present in Raf-kinase, PI3-kinase, Ral GEF (RalGDS) and several other Ras-effectors.

RA/RBD folds (amongst many others) bind to another interaction domain, the G-box, though their primary role is enabling interaction with Ras.

Question 38

How does the SH2 domain/surface binding interaction domain mediate autoinhibition?

Answer 38

The human RTK called SRC is composed of four SH domains, in descending order from the N-terminus where SH4 links it into the lipid bilayer.

The SH2 domain inhibits the kinase activity of the SH1 domain when bound to it, which only occurs when the SH1 Tyr530 is phosphorylated by CSK.

The phospho-tyrosine must be removed by phosphatases before the SH1 kinase activity can be activated, and full activity also required another, this time activating phosphorylation to the SH1 domain at Tyr419.

Question 39

Describe Adapter domain/surface binding interaction domains.

Answer 39

Adapters domains may link an enzyme to its substrate, providing colocalisation.

It may also be seperate from the enzyme and bind them both individually to colocalise them.

The use of this is to provide a regulatory layer - regulation of the adapter allows for regulation of the reaction.

Question 40

Describe Scaffold domain/surface binding interaction domains.

Answer 40

This can bind several transducer proteins in one location, and in a specific organisation. This can greatly increase catalytic efficiency of a pathway.

Question 41

What feature must all signalling pathways have?

Answer 41

A mechanism by which their transiency is ensured - all the signals must be undone to ensure that one stimulation does not permanently change all of the signalling components into one position.

This is often structured as a common writer-reader-eraser system, with the writer inducing the change (eg PTM) which is recognised by the reader and removed by the eraser.

Question 42

Give three examples of writer-reader-eraser systems,

Answer 42

Phosphotyrosine signalling - A tyrosine kinase writer (RTK) phosphorylates a tyrosine. This is recognised by the SH2 domain of the reader (PLC-gamma), which relays the signal. The phosphate is removed by a phosphotyrosine phosphatase.

Small GTPase Signalling - RasGDP is phosphorylated to Ras GTP by the writer GEF, which is read by RA/RBD domain containing proteins and removed by GAP erasers.

PIP Signalling - PtdIns is phosphorylated by PI3K writer, read by PH domain containing proteins (PKB and PDK1) and erased by PTEN, a PI-3 Phosphatase (PI-3P). Here this provides colocalisation of PDK1 and PKB so the former can phosphorylate the latter.

Question 43

How are SRC RTKs mutated in cancer?

Answer 43

First identified in a chicken sarcoma-produced virus called Rous Sarcoma Virus, v-SRC is an RTK with a deletion in the CTD of the SH2 domain, removing a pY site involved in intra-molecular interaction with the SH2 domain and lifting the SH2 inhibition, enabling constitutive stimulation.

Question 44

How is Ras often mutated in Cancer?

Answer 44

G12 and G13 mutations prevent Ras-GAP interaction, GTP hydrolysis cannot take place so the eraser cannot remove the signal and the pathway is constitutively active.

Q61 mutations impair GTP hydrolysis by interfering with the nucleophilic attack on the γ‑phosphate of GTP, again leading to constitutive activity.

Question 45

What are the different types of Ras?

Answer 45

There are three classified variants, N-Ras, H-Ras and K-Ras.

Mutations in specific variants are more of less common in certain cancers, eg K-ras mutation occurs in 60% of pancreatic cancers while the rates for the other two are negligible. Ras being a major cause of pancreatic cancer is one of the reasons it is so difficult to treat as there are very few viable Ras inhibitors. The downstream inhibitors are less effective.

Question 46

How is B-Raf often mutated in cancer?

Answer 46

Two parts of the kinase domain (the glycine-rich loop and activation segment) interact to produce an autoinhibitory effect.

Mutation in either of these can prevent this interaction leading to constitutive activation.

The frequency of V600E mutation in the activation segment occurs in BRAF mutant melanomas ~80–90% of the time.

Question 47

What cancers have B-Raf mutations been identified in?

Answer 47

BRAF mutations have been identified in a wide range of cancers including 50% of malignant melanomas, 45% of papillary thyroid cancer, 10% of colorectal cancers, and also in ovarian, breast, and lung cancers.

Question 48

What are common cancerous mutations of PI-3K?

Answer 48

Most mutations are in the p110a catalytic subunit.

E542 and E545 mutations remove the inhibitory effect of the p85 regulatory subunit they require binding to Ras.

H1047 mutations result in an activating effect on the activation loop through a conformational change; this requires binding to the p85 regulatory subunit.

Question 49

How is PTEN often mutated in cancer?

Answer 49

Mutations in the phosphatase domain (R130, R173) reduce the catalytic activity, preventing it from erasing the PIP3 signal written by PI3K.

Mutations in the C2 regulatory domain (R233, K267 and N323) prevent interaction with anionic lipids in the membrane, preventing localisation to its site of action and thus decreasing activity.

The PTEN gene is also often subject to deletions, truncations and epigenetic repression.