Tumour immunology and immune escape

Question 1

How does the immune system know what to kill and leave alone?

Answer 1

The model we use is discrimination of self and non-self through recognition of sugars, proteins, glycosylation patterns, PAMPS

Question 2

If tumour cells are self, does the immune system recognise cancer?

Answer 2

Tumour cells are self

And yet…

Transplantable tumours in mice showed that a tumour specific immune response can be generated

Mice can be vaccinated against a tumour, therefore showing that the immune system can recognise cancer

Question 3

Explain how the experiment by Shankaren et al (2001) showed the implication of the immune system and CD8+ T cells in eliminating tumours

Answer 3

RAG2-/- = mice have no way of generating antigen receptors = no lymphocytes therefore no adaptive response

Gave mice methylcolanthine, 80 days after drug given, KO mice start to have tumour, this means that the adaprive immune reponse is dealing with tumours and killing them normally

They also found that almost all tumour cells were killed by CD8+ T cells, with a small contribution from NK cells

Question 4

What do the TCRs of T cells recognise infected and maligant cells through?

Answer 4

major histocompatibility molecules (MHC) on antigen presenting cells (APCs)

Question 5

Surface and secreted molecules on CD4 vs CD8 cells

Answer 5

Surface: same (ab TCR, CD3), different, CD4 and CD8

Secreted:
CD4: IFNy, IL-2, IL-4, IL-5, IL-13, IL-10
CD8: Perorin, granzyme, IFNy

Question 6

Features of MHC i

Answer 6

Present endogenous antigens (intracellular, internal)

Display self proteins, virus proteins, intracellular pathogens

Present antigen to cytotoxic T cells (CD8)

Question 7

Features of MHC II

Answer 7

Found primarily on Antigen presenting cells (APCs) – DENDRITIC CELLS, B cells, macrophages

Present exogenous antigens (extracellular, external)

Phagocytosis, receptor mediate endocytosis

Present antigen to helper T cells (CD4)

Question 8

What 2 signals does a T cell need

Answer 8

1 = antigen

2= co-stimulation from an APC

Question 9

Features of NK cells

Answer 9

Can kill infected or tumour cells without priming

Release large amounts of IFN-γ

Have an activatibg receptor and an inhibatory ones

Activating ligands could be:
Stress molecules
Caspases from dying cells

Inhibitory ligans are:
MHC (attach to KIR receptor or NKG2A (CD94)

Question 10

Explain how NK cells make decisions based on the balance of activating and inhibitory signals

Answer 10

NO ATTACK:

1. no MHC I moleculs, no activating ligand
2. MHC I molecules, no activating ligand

ATTACK:
No MHCI molecule, Activating ligands

BALANCE:
If there is a balance of both present

Question 11

What molecule forms a pore to allow granzymes to kill?

Answer 11

Perforin

Smyth 2000

pfp-/- mice lack perforin

p53+/- mice more susceptible to tumours

Mice that lack porferin only have about 40% with no lymphoma

next:

wild type B6 mice

B6.pfp-/- = lack perforin

B6 + aThy1 = lack T cells

B6 + aNK1.1 = lack NK T cells

B6 + aCD8 = lack CD8 T cells

Low cell numbers do not cause tumours

Mice lacking perforin, T cells and CD8 T cells rapidly develop tumours

KO T cells (white ovals)

RW

Question 12

Experiment highlighting the importance of T cells and NK cells

Answer 12

Shankaren 2001
WT 20% sarcomas

RAG2-/- = mice have no way of generating antigen receptors = no lymphocytes (60%)

IFNGR-/- = mice cannot respond to IFNy release (60%)

STAT1-/- = STAT1 mice cannot respond to inflammation (55%)

RkSk – both RAG2 and STAT1 knocked out (15% incease in mice with sarcomas) likly due to NK cells, showing these also contibute

Question 13

What are tumour antigens also known as

Answer 13

Altered self

Question 14

How can tumour antigens/altered self occur?

Answer 14

if tumour is triggered via a virus (Oncogenic virus ) e.g. HPV

Differences in epigenetic silencing within a tumour cell

Mutations generally that make a different peptide

These are all very specific so are easier to treat with drugs

Gene over-expression, can be seen by T cells, but more difficult to treat

Tumour specific gene expression – T cells may be able to sense this but again difficult

Question 15

For a tumour antigen to develop

Answer 15

The mutation needs to be generated, through structural or epigenetic mutations

The mutation needs to be translated to protein

The peptide needs to be generated by the proteosome and pass into the ER

The peptide needs to be bound by an MHC class I molecule

The peptide/MHC complex needs to initiate a robust response in a T cell

Question 16

Mutated antigens

Answer 16

The first identified mouse tumour antigen was cloned in 1988.

The antigen was a complex between a mutated peptide (one amino acid altered ) and the MHC class I molecule. The wt peptide did not bind well to MHC.

Allowed CD8 T cells to recognize and kill cells expressing the mutated but not the wildtype peptide

Question 17

Overexpressed antigens in tumour cells lead to…

Answer 17

Some tumour antigens are not unique to the tumour but are hugely overexpressed compared to healthy tissue. This is a danger signal for the T cells and for antibody recognition. BUT these antigens are not unique to the tumour and so that makes them very difficult to target

Question 18

Germline antigens

Answer 18

Sometimes antigens are expressed in tumours but not in normal adult tissues – so there is no immune tolerance to that antigen established. E.g. gene switched on in the tumour cell which is normally only expressed in sperm, which do not have MHC and cannot therefore present it to T cells – so T cells do not normally see that antigen

These antigens are therefore tumour-specific but are not mutated

Question 19

Viral oncoproteins

Answer 19

About 15% of tumours are triggered by viruses (e.g. cervical cancer / papilloma viruses). Viral peptides will be presented by MHC on the surface of the cells they invade. These are obviously non-self and trigger an immune response from the CD8 T cells.

Question 20

Differences in testing in mice vs humans

Answer 20

Mice:

We use genetically identical mice

We only use a few, very in-bred mouse strains

Cell lines:

We use a few tumour cell lines

They are often extremely mutated after years in culture

We inject at millions / mouse

As we move forward in cancer biology it may become better to test in humans first, this is not the case currently

Question 21

What would you refer to lymphocytes in a tumour as

Answer 21

TILs (tumour infiltrating lymphocytes)

Question 22

Are TILs important in humans?

Answer 22

One way of telling if T cells have important anti-cancer roles in patients is seeing if patients on immunosuppressive therapies have increased cancer risk

Or if patients with inherited immunodeficiencies have increased cancer risk

Large population databases are required for this sort of work

Van Leeuwen 2010:
Do patients with kidney transplants have more cases of cancer? And due to being on immunosuppressive drugs

Also had group where transplant failed and went back to dialysis - this is a good control group that once had immunosuppressive drugs but now don’t

Also split cancers they looked at into cancers triggered by infections and ones that weren’t

On immunosuppressive drugs you are 100% more likely to develop lip cancer, drops back to normal ones off, but there are many cancers that this wasn’t the case, showing that T cells are important in some cases

Question 23

Another example showing if T cells are important

Answer 23

Mayor 1018

tracked people with inherited immunodifficuancies, in US pooled people with them, split into ages when cancer emerged and then compared to normal population

Compared to sample control population, patients with inherited immunodificiencies has a much higher cancer incidence, and this was general immundificiencies, it could be hypothesised that if yo just looked at T cell difficuencies it would be even greater

Question 24

Another example of showing that TILs are helpful

Answer 24

Azimi 2012

Using a TIL grading system:

TIL grade 0 – no TIL

Grade 1 – few

Grade 2 – moderate

Grade 3 - marked

removed tumour and investigated:
Split melanoma patients into different grades and tracked their survival

Grade 0 had lowest survival rate, then 1,then 2, 3 had 100% survival

Question 25

Example of looking at position of TILs in CRC (and granzyme B)

Answer 25

Galon 2006
looked at if its important where the T cells are in the tumour

Looked to see if in centre or invasive margin

Answer was yes, if T cells in centre of tumour, better survival, same with invasive margin

Looked at T cells expressing granzyme B

if these are in centre of tumour, average survival is 50 months, when these are in the mergin, only 20 months, only 10 when none present

Question 26

Different T cells and what they are involved in

Answer 26

Th1 - intracellular pathogens - chronic inflammation

Th2 - extracellular parasites - allergic disorders

Treg - self-tolerance

Th17 - extracellular pathogens - autoimmunity

Th22 - wound healing - inflammatory diseases

Question 27

Example of how the immune contexture in human tumous has impact on clinical outco e

Answer 27

Fridman et al. (2012)

Looked at articles published that investigated TIL types present in tumour

Many types of Th cells (classed by the cytokines they produce) but only ones important in tumours was Th1

CD8+ even more important

Th2, 17 and Reg all had poor effects on prognosis

Also looked at all immune cells in fifferent cabncer types: found that cancer and immune cell involvment vary between cancers e.g. CD8 good in most, M2 bad in most, Macrophages vary between types

Question 28

Different cancers having hot VS cold tumours

Answer 28

HOT tumour types: melanoma, kidney, bladder, head and neck, non-small cell lung cancer

COLD tumour types: most breast cancers, prostate, pancreatic, glioblastoma, ovarian

Question 29

Types of tumour associated antigen

Answer 29

Germline antigens - derived from genes expressed in the foetus. These genes are epigenetically silenced in most normal adult tissues except germline and placenta cells, which do not express MHC. They can be expressed in tumours as a result of altered DNA methylation.

HERVs - make up 8% of genomic DNA but are normally epigenetically silenced. Can be expressed in tumours owing to DNA demethylation alterations (as above), and some of the peptides can be recognised by T cells. They are immunogenic. Virtually all cancers express these peptides but there has been very limited clinical experience of directly targeting them (may be off target effects). They can be expressed in other tissues so very careful screening is necessary – off-target toxicity possible.

Overexpressed antigens. These are the products of normal genes which are expressed in other tissues and overexpressed in cancer - e.g. HER2. These have of course the highest risk of on-target toxicity. e.g. in a trial of three patients with colorectal cancer, targeting an overexpressed antigen led to cancer shrinkage but also irreversible destruction of normal colonic mucosa.

Question 30

What are tumour-specific antigens and give examples

Answer 30

These result from genomic perturbations that occur specifically in tumour cells and have entirely novel amino acid sequences. These are rarely shared between patients (often caused by environmental effects). Least chance of off-target toxicity.

The predominant form is SNV - single nucleotide variables. These proteins may become immunogenic if the peptide that results either gains the ability to bind to surface MHC or to trigger a T cell response. In a systematic screening, 83% of patients had TILs that could recognise at least one SNV neoantigen. BUT only about 1.5% of mutated SNV peptides in tumours are immunogenic and 99% were unique to each patient – difficult to manage in the clinic

INDELs. (insertion/deletions) These can be immunogenic - eg a frameshift mutation in TGFBRII is immunogenic and recognised by CD4s and CD8s in colorectal cancer. INDELs are less prevalent than SNVs – however, a higher proportion of INDELs generate high-affinity MHC binding peptides.

Gene fusions - arise through large scale genomic rearrangements such as chromosome translocations or inversions. These can result in fusion proteins which are unique to the tumour. Far fewer gene fusions that SNVs in tumours but they seem to be more therapeutically-targetable. Also, some gene fusions are pan-cancer - e.g. an FGFR-TACC3 fusion which is seen in 2% of patients with multiple different cancers (so may be good to target)

Viral oncoproteins. A subset of tumours arise from viral infection, where the integration of viral genes into the genome leads to expression of viral proteins with oncogenic properties. E.g. the E6 and E7 oncoproteins result from HPV infections. These are ‘foreign’ and can be recognised by T cells. These do not have much on or off target toxicity.

Question 31

Types of unconventional antigens

Answer 31

Conventional tumour antigens = generated from the coding regions via conventional transcription, translation, proteasomal digestion. In contrast, unconventional antigens arise from either non-coding regions of the genome, or from coding regions by means of aberrant transcription, translation, or PTM. Some of these processes may not be tumour specific.

Splice variants. Aberrant mRNA splicing in tumours can result in intron retention. Then, sequences of the retained introns can be translated and can generate peptides that bind to MHC and are recognised by T cells. Aberrant splicing can also result in novel exon-exon junctions which can produce immunogenic peptides. Not really clear whether this only happens in tumours or the clinical significance.

Aberrant translation can produce cryptic antigens. These result from either translation of protein-coding genes in alternative ORFs or from translation of non-coding sequences. Therefore, the intronic antigens above can also be cryptic antigens. Mostly this results from a phenomenon called leaky scanning, where the ribosomes continue to scan the mRNA beyond where they should have stopped. There can be stop-codon read-through events. In some tumours cryptic peptides are up to 13% of all MHC-bound peptides (but not many are immunogenic)

Post-translational changes. This can be fusion peptides resulting from proteasomal splicing of two peptide fragments - this may be up to 25% of all MHC-bound peptides. The peptides could also go through phosphorylation / acetylation / citrullination / glycosylation / deamidation.

Question 32

Which antigens can we target in cancer?

Answer 32

Viral oncoproteins and HERVs are good targets as they have low structural similarity to normal proteins and are commonly shared among patients

Question 33

What makes a tumour antigen?

Answer 33

Not normally expressed or highly expressed in healthy tissues

A peptide that binds to MHC I or MHC II

AND induces T cell activation

Question 34

T cells can see human tumour antigens

T cells can infiltrate some ‘hot’ tumours with high TMB

why do we still get these cancers?

Answer 34

Think about the theory of darwinian cancer development (Greaves and Maley 2012)

Most cancers do get killed by T cells, but sometimes they sneak through, the immune system can also be used against the tumour, by becoming immunoedited

Question 35

What are cancer stem cells?

Answer 35

Tumours are composed of a proportion of Cancer Stem Cells (CSCs) – cells with stem-like properties that are able to give rise to more differentiated daughter cells

Question 36

Explain how tumours consist of different tumour cell subtypes and an example of what this can mean for prognosis/outcome

Answer 36

Greaves and Maley 2012

Brown = normal tumour, has funsion of ERG tumour protease meaning it can get in to tissue more easily, selection pressure lead to a further mutation (green) whicj is the Eden tu,our (the ERG gene but with part cut off)

Within this there were then more mutations

Esplit tumour (red) and, 2 Edel (pink cells) (deletion of PTEN gene – poor outcome)

If tumour removed and a few pink cells left, lik;ey poor prognosis,

Question 37

Three stages/states of immunoediting

Answer 37

Elimination – Tumour cells are targeted by both the innate and adaptive immune system, many tumour cells are destroyed

Equilibrium – Tumour cells are controlled by the immune system, but not eliminated

Escape – Tumour cells acquire new mutations that allow them to escape the immune response

Question 38

Explain paper Shanlaren et all (2001) on immunoediting

Answer 38

MCA tumours grown in WT(129/SvEv) and Rag-/- mice

Transplanted to WT mice

Only tumours grown in Rag-/- mice are rejected

Tumours grown in immunofidicient mice grpwnin WT mice, the Tumour will die, dur to no immunoediting taking place

Tumours grown in WT mice put inti a different mouse will survive due to these tyumours already having undergne immunoediting, meaninbg its more resistant to the immune syetem

Question 39

Examples of immunosupressive mechanisms of tumours

Answer 39

1 loss of MHC

2 VEGF

3 FasL

4 suppressive cytokines

5 PDL1 – PD1

Question 40

immunosupressive mechanisms of tumours: MHC downregulation

Answer 40

• to hide antigens
• NK cells woukd be triggered when this happens as NK cells respond to loss of MHC, but this is a gamble as more impirtant to getvrid of T cell response

Diffrent types:
1. Total loss
2. Selective loss
3. Haplotype loss
4. Total downregulation
5. Selective downregulation
(Campoli & Ferrone 2008)

Question 41

Explain immunosupressive mechanisms in tumour: Prevention of migration into the tumour

Answer 41

– by inturupting chemokine gradients

The presence of TILs correlates with the expression of CCL2, CCL3, CCL5 and other chemokines

CXCR3 signalling is important for T cells binding to endothelial cells and moving into the tumour

Tumours disrupt chemokine production to prevent TIL migration or to attract suppressive cells like macrophages

Question 42

Explain the immunosupressive tumour mechanism of FasL

Answer 42

Kills T cells that do make it in:

FasL is upregulated in tumours, when Fas is trimarised it leads to cleavage of procaspase 8 to actve caspase 8, caspase cascade and therefore apoptosis/cell death

Tumours cause upregulation of FasL and therefore incerease chaces of trimerisation

Question 43

Explain immunosupressive tumour role - dampening tumour response

Answer 43

In normal inflammation:

Adhesion molecule expression (ICAM-1, VCAM-1) in response to inflammation, interaction with receptors on T cells (LFA-1, VLA-1), T cell extravasion

In the TME:

Tumour cell expression of VEGF, bFGR, intercat with VEGFR, FGFR and these block adhesion molecule expression, therefore inhibiting T cell activation

RW

Question 44

Explain the technique of infusing TILs into the patient

Answer 44

Adoptive T cell therapy involves removing a patient’s TILs, selecting tumour-specific ones, expanding them ex vivo and infusing them back in

Before the infusion patients are treated with lymphodepleting chemotherapy

Along with the TIL, IL-2 injections are given to support the growth of the T cells

Very expensive, personalised therapy

Does it work?

In ‘hot’ tumours this has had a lot of success (complete remission in 15% of patients with treatment-refractory metastatic melanoma).

In other tumours there are many problems getting the TIL to move to the correct site
Side effects are severe owing to cytokine release by the T cells (cytokine storm). Particularly capillary leak syndrome – can lead to multiple organ failure

Question 45

Explain the method of infusing activating cytokines for cancer treatment (to turn on T cells)

Answer 45

T cells need a lot of signals from cytokines

For proliferation and survival of all T cells IL-2 is absolutely critical. Taken up through the IL-2R (CD25 and other chains)

For activation, IL-12 and IL-15 promote Th1 cells and activate NK cells

Suppressive cytokines (TGF-beta, IL-10) or Th2-driving cytokines (IL-4, IL-13) are less useful in this context

IL-2 injections have been approved for melonoma and renal cell cancer, other cytokine therapies (IL-15, IL-21, IL-12 are curremtly in trial for melanoma and renal cell)

Question 46

Explain increasing T cell priming by dendritic cells (cancer treatment)

Answer 46

RW

Does this work?

Adoptive transfer of DC + antigen is much harder than transfer of the T cells – they are fewer in number, must get to the right area of the lymph node to trigger T cell proliferation, must have all the correct co-receptors

This has been tried for decades but positive results are scarce – it works much better in mice than in humans, for reasons we don’t understand well. Doesn’t really work when injected in patients

Question 47

Explain the cancer treatment of infecting tumours with viruses

Answer 47

Oncolytic virus therapy is the deliberate infection of tumour cells with oncolytic viruses. Follows on from decades of observations that infection with viruses occasionally leads to regression

These viruses alone induce death of the cancer cells (not by activating T cells) AND release of tumour antigens, sparking immune responses to tumours which may have otherwise been ‘cold’

They can also be engineered to deliver other therapeutic moieties to the tumour such as cytokines

Often they are engineered to also be less infectious to healthy cells

Next step is to work out why certain people respond while others don’t

Question 48

Explain inducing T cell exhaustion

Answer 48

After continued antigen stimulation CD8+ T cells become ‘exhausted’. They up-regulate PD1 and Tim3, no longer proliferate, no longer kill infected cells or produce cytokines

Reasons are unclear but this is perhaps a method of limiting immunopathology from having chronically-activated T cells accumulating

This happens with chronic viral infections but also in tumours

Question 49

Explain the basic stimulation of a T cell

Answer 49

Basic (very simplified) series of events – signal one (MHC); signal two (co-stimulation); signal three (other signals) e.g. cytokines

Question 50

Explain negative co-stimulation

Answer 50

Some co-stimulatory molecules are negative rather than positive signals

CTLA-4 – initiall intracellular, after about 48 hours it moves to the surface to help switch off T cells

PD-1 turns off T cell effector function over a longer period of time, in cancer we don t want these to be active, so that T cells can keep killing tumour cells without exhastioun

Question 51

Explain the use of checkpoint inhibatirs to turn on T cells in cancer treatment

Answer 51

CTLA4 major negative regulator. Initially intracellular – moves to cell surface after TCR signalling

Higher affinity for CD80 / CD86 than CD28 – binds all the CD80/86 so prevents CD28 signalling

PD1 is up-regulated on exhausted T cells and switches off T cell effector function

In cancer therapies we want lots of very activated T cells

therefore PD-1, PD-L1 and CTLA-4 inhibitors were made

These drugs are known as checkpoint inhibitors

How successful is checkpoint inhibition?

20% of all patients have durable positive results following checkpoint inhibition.

For example, just four doses of anti-CTLA4 have led to 10-year regression in some melanoma patients. Anti-PD1 therapies have a 40% response rate in NSCLC

BUT this is very disease-specific (e.g. hot vs cold tumours) and patient-specific. E.g. breast cancer response is only 15%

Question 52

What are some side effects of cancer treatments targeting T cells

Answer 52

Immune related adverse effects (IRAEs) and autoimmunity

Question 53

Explain IRAEs

Answer 53

These therapies do lead to side effects – as every T cell in the body is turned on. Development of side effects positively correlates with regression of the tumour

These IRAEs (immune related adverse events) are often localised to the intestine, where millions of T cells reside in the intestinal wall

Switching off anergy also means self-reactive T cells, which are normally suppressed, will wake up. This has led to patients developing autoimmune diseases.

Careful management of co-therapies is required

Question 54

Where next with immunotherapies to fight cancer

Answer 54

Careful consideration of co-therapies is now underway. E.g. adoptive transfer of ex vivo-activated T cells WITH anti-PD1 therapy AND recombinant IL-2

Reduction in side effects through careful dosing

Understanding immune-related adverse events through in-depth immunology

Biomarkers to predict patient response