Serendipities ofacquired immunity

Tasuku Honjo

December 7, 2018Nobel Lecture

Kyoto University Institute for Advanced Studyand Graduate School of Medicine

My family (1955)

Saturn.Actual photo taken on June 5, 2016https://www.nasa.gov/mission_pages/cassini/multimedia/pia17171.html

Cassini: Earth and SaturnThe Day the Earth Smiled

Through the brilliance of Saturn’s rings, Cassini caught a glimpse of a far-away planet and its moon. At a distance of just under 900 million miles, Earth shines bright among the many stars in the sky, distinguished by its bluish tint.

EarthThe telescopic view of Saturn fascinated me. I dreamed of becoming an astronomer.” “

Inspired by biography of Hideyo Noguchi (1876~1928)

Rockefeller Univ.

・Identified Syphilis spirochete as the cause of progressive paralysis

・Died in Ghana during pursuit of yellow fever pathogen

With Osamu Hayaishi

With Jacques Lucien Monod 1966

Diphtheria toxin inactivates protein synthesis factor by ADP-ribosylation

T. Honjo et al., J.Biol.Chem. (1968)

active inactive

NAD

+EF-II

nicotinamideEF-II

Ribo

ADP

EF-II

diphtheria toxin acts as a catalyst

Ribo

ADPRibo

ADP

DT

+

8

Donald Brown at Carnegie Institution in Baltimore 1971

Mystery of immune responsein 1950~1970

How can animals generate antibodies specific to an almost infinite number of antigens, including artificial chemicals?

Why can animals generate specific antibodies to almost all

unexperienced compounds?

Anti-N-benzene Ab

Anti-N-phenol Ab

Anti-anthracene Ab

Anti-toluene-DIC Ab

Anti-oxazolone Ab

nitrobenzene-protein

nitrophenol-protein

anthracene-protein

toluene diisocyanate-protein

oxazolone-protein

Modified from K. Landsteiner 1919-22

11

Structure of antibody identified by 1970

H chain (heavy chain)L chain (light chain)

Constant region (C)(Antibody class determination)

Variable region (V)(Antigen-recognition site)

CCCC

Philip Leder at NIH 1973

C. Brack et al., Cell (1978)

VDJ recombination generates V region repertoire during differentiation

VDJ recombination

JH CµS

CδDHVH

S

Constant region

B cellIgDIgM

S. Tonegawa

University of Tokyo, Dept. of Nutrition(Professor Yoshinaga Mano) 1974

10000

1000

100

10

0

1

0 7 14 21 28 35 42

Primary response Secondary response

Tite

r

Days

Primary immunization Secondary immunization

Antibody memory generation byvaccine (antigen) administration

Increase in antigen binding capability (somatic hypermutation of variable region)

Somatic hypermutation (SHM) mutates V region and only

good antibodies are selected

YY

Y Y

Ypathogen

Y strongest

Y

Y

Y

YY

Y YY

Darwinian principle

B cell

Increase in antigen processing ability(class switch of constant region)

10000

1000

100

10

0

1

0 7 14 21 28 35 42

Primary response Secondary response

Tite

r

Days

IgMIgG

Primary immunization Secondary immunization

Antibody memory generation byvaccine (antigen) administration

Increase in antigen binding capability (somatic hypermutation of variable region)

?

Class switching changes the H chain constant region and antibody function

L chain

H chain

CCCC

Constant region (C)

Variable region (V)

IgM

CCCC

CSRCCCC

IgG

CCCC CCCC

IgA

CCCC CC

CC

IgE

Y

Y

Y

Gut bacteria

Parasite

Virus

CC

class switch recombination(CSR)

looped-outcircular DNA

Class switch recombination takes place by deletion of a large DNA segment

S S S S S S S

V µ δ γ ε αγ γ γH C C C 3 C CC 1 C 2b C 2a

µCδC

γC 3γC 1

γC 2b

VH γC 2a εC αCS S

S

T. Honjo & T. Kataoka, PNAS (1978)T. Kataoka et al., PNAS (1980)A. Shimizu et al., Cell (1982)

T. Kataoka A. Shimizu

Matthias Wabl, Göran Möller (coorganizer) Leroy Hood

YIY

Discovery of AID by comparison of gene expression before and after CSR

Stimulation with CD40L, IL-4, TGF-β1

M. Muramatsu et al., J.Biol. Chem. (1999)

IgM IgA

Activation Induced cytidine Deaminase

AID

AIDB cells

Expressed in germinal center

Comparison of expressed gene transcripts

Defective IgG response to antigens (Sheep Red Blood Cell) in AID deficient mice

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 215 10 14

Days

IgM

0 215 10 14

IgG1

Ant

ibod

y ti

ter

AID +/-AID -/-

AID+/-

AID-/-

Days

Ant

ibod

y ti

ter

M. Muramatsu et al., Cell (2000)

VH186.2 sequence (residue)

0

40

Q1 R98W33

0

40

0

80 CDR1 CDR2RSγ1

µ

µMut

atio

n fr

eque

ncy(

%)

AID+/-

AID+/-

AID-/-

No mutation

AID deficient mice fail to accumulate mutations

K. Kinoshita M. Muramatsu et al., Cell (2000)

・AID deficiency in human is the cause of Hyper IgM Syndrome Type II: exactly the same phenotypes as mouse.

P. Revy et al., Cell (2000)

・Thus, AID is the enzyme that engraves antigen memory in the antibody gene, the mechanistic basis of vaccination.

VDJ recombination

V D J µ δ γ ε αγ γ γH H C C C 3 C CC 1 C 2b C 2aS S S S S S S

RAGsNatural AbNatural Ab

Pathogensomatic hypermutation(SHM)

mutated V gene

class switch recombination(CSR)

chromosomal product

＋

Antibody memoryformation

looped-outcircular DNA

MemoryAb

VDJ recombination

V D J µ δ γ ε αγ γ γH H C C C 3 C CC 1 C 2b C 2a

Repertoire formation

S S S S S S S

RAGsNatural Ab

tumors

AID engraves Ab memory in the genome for effective vaccination

AID

Immune surveillance against cancer

Proposed by Sir Frank Macfarlane Burnet (1970)

However, numerous attempts to develop immunotherapy were unsuccessful.

Cancer immunotherapy by boosting accelerators has not given convincing

clinical outcomes

1. Cancer vaccine

2. In vitro activation of T lymphocytes

3. Cytokine treatment (IFNγ, IL-2, IL-12 etc)

This was because no immune brake molecules were known before 1995

driving［Attack］

accelerator［ICOS］

brake［PD-1］

〜100k/h

parking ［Activation］

ignition［CD28］

parking brake［CTLA4］

ON/OFF

drive stop action mode

Brakes and accelerators control immune reactions like those in a car

［Drastic］

［Mild］

action phase

Y. Ishida et al., EMBO J. (1992)

Discovery of PD-1 (programmed death-1) cDNA

Structure of cytoplasmic tail suggests PD-1 is a surface signaling molecule

Y. Ishida Y. Agata

A long journey to understanding the function of PD-1

・1994 PD-1 knock out (KO) on mixed backgroundmice, no phenotype change

・1996 PD-1 KO on C57BL/6, no phenotype change for 6M

Nephritis and arthritis after 5M in PD-1KO x lpr/lpr background

・1998

・1997

But over-response to antigen stimulation

Clear autoimmunity in PD-1 KO by 14M

PD-1 is a negative regulator

PD-1 KO (Knock Out)

Nephritis ArthritisY. Nishimura et al., Immunity (1999)

Dilated cardiomyopathy

PD-1 KOWTBALB/c

NOD

Diabetes

NODxPD-1 KO MRL

MyocarditisT. Okazaki et al., Nat. Medicine (2003)J. Wang et al., Int. Immunol. (2010)

MRLxPD-1 KO

C57BL/6

Molecular mechanism of immune inhibition by PD-1 signaling

N N

PD-1

N

NNα β

γ/ε δ/ε

ζ ζ

Antigen receptor

negative signal

PPP

P

Coreceptor

T. Okazaki et al., PNAS（2001）

ITSM

Activationsignal

Kinase

ZAP70

S-S

S-S

S-S

S-S S-S

S-S

S-SS-S

S-S

P YY

Antigen

PP

P

P

PD-L1

Balance between immune surveillance and immune tolerance

Immune toleranceImmune suppression

Treatment of autoimmunityHigher risk of infectious diseases and cancer

Treatment of infectious diseases and cancerRisk of autoimmunity

Immune surveillanceHyper immunity

PD-1blockade

Days after inoculation5 10 15 20

Inhibition of tumorigenesis of myeloma (J558L) in PD-1-/- mice

Y. Iwai et al., PNAS (2002)

J558L

BALB/cWT

Tum

or v

olum

e (m

m3 )

5 10 15 20

N=4

J558L

BALB/cPD-1-/-

N=410000

7500

5000

2500

0

Tum

or v

olum

e (m

m3 )

Days after inoculation

0 5 10 15 200

500

1000

1500

2000

2500

0 5 10 15 20

n=10n=10

Rat IgG

P815/PD-L1 DBA/2

a-PD-L1

Inhibition of tumorigenesis of P815/PD-L1 by anti-PD-L1

Tum

or v

olum

e (m

m3 )

2500

2000

1500

1000

500

00 5 10 15 200 5 10 15 20

N=10 N=10

Days after inoculation

Tumor growth Survival rate

(%)

Rat IgG

a-PD-L1

100

0

20

40

60

80

020 40 60 80

Days after inoculation

N. Minato

Y. Iwai et al., Int. Immunol. （2005）

PD-1 blockade inhibits metastasis of B16 melanoma (mouse model)

weig

htof

live

r (g

)

0

2

4

6

8

WT

Spleen to liver

Anti PD-1 AbWT Anti PD-1 Ab

PD-1 blockade by antibody against either PD-1 or PD-L1 can cure cancer

Killer T cell(CD8+)

Tumor/infected cell

antiPD-L1

Other immune cells

PD-L1 is expressed on various immune cells

SHP2PD-L1 PD-1

MHC TCR

PD-L1PD-1

MHCTCR

antiPD-L1

antiPD-1

antiPD-1

+-

+

-

Human anti-PD-1 antibody

Approved as Investigation New Drug by FDA(USA; Aug 1, 2006)

Synthesized in mice containing human immunoglobulin gene by Medarex

Subclass: IgG4S228Pmutant IgG4 (S228P) stabilizes the protein and reduces ADCC(antibody-dependent cell-mediated cytotoxicity)KD = 2.6 nmol/L

Named Nivolumab

296 terminal stage patients recruited Nivolumab treatment for two years

Complete or partial response rates

18% ( 76 patients) of non small cell lung cancer28% ( 94 patients) of melanoma27% ( 33 patients) of renal cell carcinoma

S. Topalian et al., NEJM (2012)

Clinical trials began in US (2006) and Japan (2008)

Summary of Phase I clinical trial

Durable response to PD-1 blockade“Responses were durable; 20 of 31 responses lasted

1 year or more and some even after stopping therapy”

baseline

tumorenlargement

tumorregression

Chan

ge in

tar

get

lesi

ons

from

bas

elin

e(%

)

Patients with melanomaFirst occurrence of new lesionPatient off study

Weeks since treatment initiation

S. Topalian et al., NEJM (2012)

stop treatment

Phase II trial of anti-PD-1 antibody in patients with platinum-

resistant ovarian cancer

Oct 21, 2011-Dec 7, 2014

Dose total(n) CR PR SD PD NE RR DCR

1 mg/kg 10 0 1 4 4 1 1/10(10%)

5/10(50%)

3 mg/kg 10 2 0 2 6 0 2/10(20%)

4/10(40%)

Total 20 2 1 6 10 1 3/20(15%)

9/20(45%)

Tumor growth stopped in 40-50% of terminal stage patients

J. Hamanishi et al., J. Clin. Oncol. (2015)

A responder with ovarian cancer(clear cell): Nivolumab 3mg/kg

Baseline 4 months

Peritoneal dissemination ⇒ disappeared

History: 60 yr. Stage Ic with progressive disease afterRSO, MMC/CPT11*3, SCH+BSO, CPT/CDDP*5, TC*2

Cancer marker CA125 (U/ml)

１ 2

(day)

316

16 100

200

400

0 50 100 150

43J. Hamanishi et al., J. Clin. Oncol. (2015)

Durable complete responses of ovarian cancer patients to Nivolumab

J. Hamanishi, The International Federation of Gynecology and Obstetrics FIGO (2018)

0

50

100

150

0 1 2 3 4 5

CA12

5(U

/ml)

NivoCase 1

CA12

5(U

/ml)

0

100

200

300

400

0 1 2 3 4 5years

NivoCase 2

No medication

> 4 years

No recurrence > 5 years

No medication

> 3.5 years

No recurrence > 4.5 years

Randomized Study on Untreated Melanoma Patients with Nivolumab and Dacarbazine (chemotherapy)

Overall survival

Dacarbazine

Nivolumab

Patients who died median survival

Hazard ratio for death, 0.42(99,7% Cl, 0.25-0.73)P<0.001

DacarbazineNivolumab

No,/total no. mo (95% Cl)50/210 Not reached96/208 10.8 (9.3-12.1)

DacarbazineNivolumabNo. at risk

210208

185177

150123

10582

4522

83

00

Months

Patien

ts s

urviving

(%)

C. Robert et al., NEJM (2015)

Cancers approved for PD-1 blockade therapy

• 2014 melanoma• 2015 lung cancer• 2016 renal cancer

Hodgkin’s lymphomahead and neck cancersurothelial cancer

• 2017 colorectal cancergastric cancerhepatocellular carcinomaMerkel cancerall highly mutated cancers

• 2018 cervical cancerprimary mediastinal large B-Cell lymphoma

Paradigm shift of cancer therapy by anti-PD-1 treatment

1. Less adverse effects because normal cells are unaffected

2. Effective for a wide range of tumors(more than 1000 clinical trials)

3. Durable effects to responders after stopping treatment

47

Cancer cells accumulate mutations

I. Martincorena et al., Science (2015)

10000

1000

100

10

1

100

10

1

0.1

Mut

atio

nspe

r m

egab

ase

Codi

ng m

utat

ions

per

tum

our

What we learned from hugecancer genome projects

1. Cancer cells accumulate a large number of mutations to express neo-antigens that can berecognized by the immune system as non-self.This is why cancer immunotherapy is effective.

2. Too many mutations to pinpoint the dominant mutations for targeted chemotherapy.

Continuous mutations generate resistant tumor cells

cancercells

drug A

Selectionmutagenesis

Resistant cells grow

Lymphocytes can recognize many more mutants& attack them

drug B

Selectionmutagenesis

Resistant cells grow

Current issues in PD-1 blockade therapy

Biomarkers for responders・ High mutagenesis in tumors・ Potency of individual’s immunity

Improvement of immunotherapy・ Accessibility of killer T cells to tumor sites・ Potentiation of killer T cell function

PD-1 blockade initiates killer T cell expansion in lymph nodes

Chemokines attract killer T cell

K. Chamoto et al., PNAS (2017)

1. PD-1 blockade enhances killing within tumor which secrets chemokines

2. PD-1 blockade enhances priming and induces chemokine receptor to helpmigration of new killer T cell towards tumor

DLN(draining lymph node)

Tumor

53

Numbers of PD-1 blockade trials using combinations with :

1. Anti-CTLA-4 agents: 2512. Chemotherapies: 1703. Radiotherapies: 644. Anti-VEGFA agents: 435. Chemoradiotherapy combos: 42

J. Tang et al., Ann. Oncol. (2018)

Cancer immunotherapy by PD-1-based combination studies underway in 2017

Requirement of mitochondrial activation for killer T cell activation and proliferation

TCR stimulation

Proliferation/activation of killer T cells

Mitochondria Mitochondrial biogenesis provides cell with energy

-

Anti-PD-1PD-1

- + +

? boost

tumor

AMPK

PGC-1α

mTOR

PGC-1αPPARα/γ

Activation of PGC-1⟨ /PPAR complex improves the efficacy of PD-1 blockade

-PD-1

K. Chamoto et al., PNAS (2017)

Bezafibrate

2500

2000

1500

1000

500

Days after MC38 inoculation

Tum

or v

olum

e(m

m3 )

0 5 10 15 20 25

ControlAnti-PD-L1 mAbAnti-PD-L1 mAb + BezafibrateBezafibrate

Mitochondria mass + Energy boost

TCR stimulation

-+ +

Bezafibrate increases killer T cell proliferation and blocks cell death

PD-1 - +-

-

Survival, proliferation &memory generation

Effector killer T cells

Exhaustion & cell death

P. Chowdhury et al., Cancer Immunol. Res., (2018)

Bezafibrate+

PPARα/γ

M. Miyajima, B. Zhang et al., Nat. Immunol. (2017)

Hyperimmune activity can be read inblood biochemistry of PD-1-/- mice

Tryptophan metabolism

Malate-aspartate shuttleGluconeogenesis

Urea cycleGalactose metabolism

Starch and sucrose metabolismFructose and mannose degradation

Aspartate metabolismAmmonia recycling

Alanine metabolismGlucose-alanine cycleHistidine metabolism

Glutamate metabolism

P value

1e-01

6e-01

1e+00

Fold enrichment0 2 4

Citric acid cycleMitochondrial electron transport chain

Mitochondrial activation &Tryptophan consumption

PC1 (24.8%)

PC2

(16.

6%)

PD-1-/-WT

Sidonia Fagarasan

PD-1-/- mice biology is very complex

?

Metabolite shift

Behavioralchanges

Expansion T cells

H. Nishimura et al. Immunity 1999

Consumption ofmetabolites

M. Miyajima, B. Zhang et al., Nat. Immunol. (2017)

Gut bacterial changes

PD-1 selects IgA critical to microbiota regulation

IgA-coated bacteria in the gut ***

WT PD-1-/-

IgA

coa

ted

bact

eria

(%)

IgA DAPI

Less IgA-coating of bacteria in PD-1-/- mice

Bacterial dysbiosis

TotalAnaerobic

AerobicBacteroidiaceae

103 104 105 106 107 108 109

No. bacteria/g small intestine content

***

NDBifidobacteriumLactobacillus

StreptococcusEnterobact ***

***

WT

PD-1-/-

S. Kawamoto et al., Science (2012)

GC B cells GC T cellsPD1hiAIDhi

AID and PD-1 cooperate in germinal centersfor high affinity IgA selection

to maintain microbiome

YY Y

AID

IgAPD1

IgM

S. Kawamoto et al., Science (2012)

AID-/-WT

Critical role of AID for controlling microbiota & whole body immune homeostasis

S. Fagarasan et al., Science (2002)K. Suzuki et al., PNAS (2004)

WT AID-/-

94%Lactobacillus

72%SFB

28%Clostridium

Mucosal immune activation

WT AID-/-Systemic immune activation, spleen

SFB

From Meyerholz et al,2002

Enhanced anti-tumor immunity in AID-/- mice depends on microbiota

M. Akrami, R. Menzies, M. Miyajima, Y. Nakajima. unpublished data

WT SPFAID-/- SPF

WT GFAID-/- GF

Specific-pathogen free Germ free

Microbiome-immune system regulation

YYAID

PD1

metabolites

Y

IgA

System homeostasis Immune tolerance

Changes Microbiota & Metabolites

YAID

PD1 YY

IgA

blockade

Anxiety, autoimmunity

Microbiome-immune system regulation

Enhanced Anti-tumoractivity

CLOSING IN ON CANCERAndy Coghlan New Scientist, 5 March 2016

“We’re at the point where we’ve discovered the cancer equivalent of penicillin” says Chen. Although penicillin itself couldn’t cure all infections, it gave rise to a whole generation of antibiotics that changed medicine forever, consigning most previously fatal infections to history.

Future prospects in cancer therapy

1. Efficacy of PD-1 blockade therapy improved.

2. Many more cancers may be treated by immunotherapy.

3. Cancer may not completely disappear, but be controlled by immunotherapy. Cancer maybecome one of chronic diseases.

2016

2020

2030?

Enormous benefit by acquired immunity

20th century

Penicillin

Pneumonia

Streptomycin

Tuberculosis

Eradication of infectious diseasesby vaccination and antibiotics

21st century Cancer may be controlled by immunotherapy and its improvement including microbiome manipulation

Acquired immunity evolved in vertebrates

Molluscs

Arthropods

Nemertea

Annelides

Urochordata

Platyhelminthes

Echinoderms

Cnidaria

ProtozoaPorifera

Mammals

BirdsAmphibians

Reptiles

Jawed fishJawless fish

Acquired immunity

Naturalimmunity

Fortunate outcomes from evolution of acquired immunityMolluscs

Arthropods

Nemertea

Annelides

Urochordata

Platyhelminthes

Echinoderms

Cnidaria

ProtozoaPorifera

Mammals

BirdsAmphibians

Reptiles

Jawed fishJawless fish

Acquired immunity

Naturalimmunity

Collaborators

Dept. of Immunology and Genomic Medicine,Graduate School of Medicine, Kyoto University

Sidonia FagarasanIMS, RIKEN

Fumihiko MatsudaKyoto University

Major outside collaboratorsAntibody diversityE. Severinson (Stockholm Univ.)F. Alt (Harvard Univ.)M. Nussenzweig (Rockefeller Univ.)A. Fischer (Necker Hospital)A. Durandy (Necker Hospital)T. Chiba (Kyoto Univ. Hospital)

Cancer immunotherapy by PD-1 blockadeG. Freeman (Dana Farber Cancer Center)N. Minato (Kyoto Univ.)S. Fujii (Kyoto Univ.)I. Konishi (Kyoto Univ.)

Acknowledgement for research funding support

ONO Pharmaceutical CO. LTD

MEXT MHLW

JSPS AMED NIBIOHN

NIHN

Bristol-Myers Squibb

Jane Coffin Child Memorial FundTang Prize Foundation

Thank you for your attention