Oncoimmunology: Some fundamental problems of cancer immunotherapy

13
ISSN 0026-8933, Molecular Biology, 2007, Vol. 41, No. 2, pp. 316–328. © Pleiades Publishing, Inc., 2007. Original Russian Text © S.A. Nedospasov, D.V. Kuprash, 2007, published in Molekulyarnaya Biologiya, 2007, Vol. 41, No. 2, pp. 355–368. 316 IMMUNE SYSTEM AND CANCER CELLS: CURRENT STATE OF THE IMMUNOLOGICAL SURVEILLANCE CONCEPT The concept of immunological surveillance of spontaneously emerging tumor cells was proposed long ago [1]. It was based on early data on the immu- nogenicity of implanted [2], carcinogen-induced [3], or virus-induced [4] tumors, including the feasibility of vaccination against viral antigens to suppress the emergence of primary tumors [5]. It was believed that a tumor cell expresses the genes encoding proteins or “factors” that the immune system can recognize to discriminate between normal and tumor cells. It should be noted that, in the 1960s, nothing was known about the molecular mechanisms of immune recogni- tion by T lymphocytes, while selection and immuno- logical tolerance were poorly understood. Curiously, Stutman [6] discarded the immunologi- cal surveillance theory in 1970. He studied tumor development in immunodeficient mice, such as athy- mic (nude) mice, lacking adaptive immunity. Contrary to what might be expected, such mice failed to show any notable acceleration in either induced carcinogen- esis or in growth of implanted tumors. Thus, because of the absence of T and B cells in such mice, the con- clusion was made that the immune system is not involved in tumor surveillance. It was found more recently that some immune surveillance components in nude mice, such as γδ T cells and natural killers (NKs) are fully functional; i.e., Stutman’s experi- ments were misinterpreted. In the 1990s, the genetic knockout technique was employed to generate mouse lines in which the adap- tive immunity system was partly or entirely disrupted. Such lines allowed the role of adaptive immunity in protection against tumors and pathogens to be prop- erly analyzed. In particular, the gold standard in immunology and, especially, cancer immunology is RAG (recombination activating gene)-deficient mice, lacking the mechanism of somatic recombination, which is required for rearrangements of both immuno- globulin genes and genes for T-cell receptors [7]. These rearrangements are mandatory for the develop- ment of functional T and B cells. In contrast to nude mice, the adaptive immunity system of RAG –/– mice is entirely out of function. Another common model with a disrupted immunity is scid mice. Their defect is due to a mutation that inactivates the gene for the catalytic subunit of DNA-dependent protein kinase, impairs DNA double-strand break repair, and forbids V(D)J recombination [8]. Schreiber et al. [9] repeated the experiments on spontaneous carcinogenesis with scid and RAG –/– mice and with mice defective in signaling pathways involving γ-interferon (γ-IFN), the most important immunoregulatory cytokine indispensable for the for- mation of functional cytolytic T cells [9]. An acceler- ated emergence of spontaneous and induced tumors MOLECULAR MEDICINE Oncoimmunology: Some Fundamental Problems of Cancer Immunotherapy S. A. Nedospasov a, b and D. V. Kuprash a a Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, 119991; e-mail: [email protected] b Belozersky Institute of Physico-Chemical Biology, Moscow State University, Moscow, 119899 Russia Received October 6, 2006 Accepted for publication November 20, 2006 Abstract—The review briefly discusses several central problems of modern oncoimmunology. The controver- sies surrounding the concept of immunological surveillance, as well as the problem of immunological tolerance to tumors, are considered. The discovery of tumor antigens is a great advance towards the identification of pos- sible therapeutic targets. However, antigen-specific vaccinations against cancer have, so far, a very limited use, mainly for prevention of virus-associated cancers, which is essentially based on the antiviral immune response. On the other hand, antibodies to cancer antigens are widely used in cancer diagnosis, and there are remarkable examples of their therapeutic applications. The future opportunities in both theoretical and applied oncoimmu- nology will directly depend on further advances in basic science. DOI: 10.1134/S0026893307020124 Key words: cancer, immunity, antibodies, T lymphocytes, vaccines UDC 612.017.1:616-006

Transcript of Oncoimmunology: Some fundamental problems of cancer immunotherapy

ISSN 0026-8933, Molecular Biology, 2007, Vol. 41, No. 2, pp. 316–328. © Pleiades Publishing, Inc., 2007.Original Russian Text © S.A. Nedospasov, D.V. Kuprash, 2007, published in Molekulyarnaya Biologiya, 2007, Vol. 41, No. 2, pp. 355–368.

316

IMMUNE SYSTEM AND CANCER CELLS: CURRENT STATE OF THE IMMUNOLOGICAL

SURVEILLANCE CONCEPT

The concept of immunological surveillance ofspontaneously emerging tumor cells was proposedlong ago [1]. It was based on early data on the immu-nogenicity of implanted [2], carcinogen-induced [3],or virus-induced [4] tumors, including the feasibilityof vaccination against viral antigens to suppress theemergence of primary tumors [5]. It was believed thata tumor cell expresses the genes encoding proteins or“factors” that the immune system can recognize todiscriminate between normal and tumor cells. Itshould be noted that, in the 1960s, nothing was knownabout the molecular mechanisms of immune recogni-tion by T lymphocytes, while selection and immuno-logical tolerance were poorly understood.

Curiously, Stutman [6] discarded the immunologi-cal surveillance theory in 1970. He studied tumordevelopment in immunodeficient mice, such as athy-mic (nude) mice, lacking adaptive immunity. Contraryto what might be expected, such mice failed to showany notable acceleration in either induced carcinogen-esis or in growth of implanted tumors. Thus, becauseof the absence of T and B cells in such mice, the con-clusion was made that the immune system is notinvolved in tumor surveillance. It was found morerecently that some immune surveillance componentsin nude mice, such as

γδ

T cells and natural killers

(NKs) are fully functional; i.e., Stutman’s experi-ments were misinterpreted.

In the 1990s, the genetic knockout technique wasemployed to generate mouse lines in which the adap-tive immunity system was partly or entirely disrupted.Such lines allowed the role of adaptive immunity inprotection against tumors and pathogens to be prop-erly analyzed. In particular, the gold standard inimmunology and, especially, cancer immunology isRAG (recombination activating gene)-deficient mice,lacking the mechanism of somatic recombination,which is required for rearrangements of both immuno-globulin genes and genes for T-cell receptors [7].These rearrangements are mandatory for the develop-ment of functional T and B cells. In contrast to nudemice, the adaptive immunity system of

RAG

–/–

mice isentirely out of function. Another common model witha disrupted immunity is

scid

mice. Their defect is dueto a mutation that inactivates the gene for the catalyticsubunit of DNA-dependent protein kinase, impairsDNA double-strand break repair, and forbids V(D)Jrecombination [8].

Schreiber et al. [9] repeated the experiments onspontaneous carcinogenesis with

scid

and

RAG

–/–

mice and with mice defective in signaling pathwaysinvolving

γ

-interferon (

γ

-IFN), the most importantimmunoregulatory cytokine indispensable for the for-mation of functional cytolytic T cells [9]. An acceler-ated emergence of spontaneous and induced tumors

MOLECULARMEDICINE

Oncoimmunology: Some Fundamental Problems of Cancer Immunotherapy

S. A. Nedospasov

a,

b

and D. V. Kuprash

a

a

Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, 119991; e-mail: [email protected]

b

Belozersky Institute of Physico-Chemical Biology, Moscow State University, Moscow, 119899 Russia

Received October 6, 2006

Accepted for publication November 20, 2006

Abstract

—The review briefly discusses several central problems of modern oncoimmunology. The controver-sies surrounding the concept of immunological surveillance, as well as the problem of immunological toleranceto tumors, are considered. The discovery of tumor antigens is a great advance towards the identification of pos-sible therapeutic targets. However, antigen-specific vaccinations against cancer have, so far, a very limited use,mainly for prevention of virus-associated cancers, which is essentially based on the antiviral immune response.On the other hand, antibodies to cancer antigens are widely used in cancer diagnosis, and there are remarkableexamples of their therapeutic applications. The future opportunities in both theoretical and applied oncoimmu-nology will directly depend on further advances in basic science.

DOI:

10.1134/S0026893307020124

Key words

: cancer, immunity, antibodies, T lymphocytes, vaccines

UDC

612.017.1:616-006

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

PROBLEMS OF CANCER IMMUNOTHERAPY 317

was observed in these experiments and the mice diedearlier [9]. These results were often interpreted as arevival of the immunological surveillance concept.Schreiber’s experiments revealed the mechanism ofthis phenomenon, indicating that surveillance requiresT cells and a functional

γ

-IFN system. Earlier, the pro-tective role of

γ

-IFN had been noted in the classicspontaneous carcinogenesis model of

p53

-deficientmice [10]. Other results indicated that an importantsource of

γ

-IFN in these model experiments was non-canonical T lymphocytes with

γδ

T-cell receptors [11].

In addition, Schreiber et al. compared the proper-ties of carcinogen-induced tumor cells in normal andimmunodeficient mice by implanting tumor cells intosecondary recipient mice. The recipient mice wereeither normal or incapable of immunological surveil-lance (Table 1).

The results allowed the immunological surveil-lance concept to be rectified. First, the adaptiveimmune system can indeed recognize tumor cells anddestroy some of them. Second, the pressure imposedby the immune system selects the tumor cell variantsthat successfully escape immune recognition. As aresult, the immunological properties of tumor cellsundergo immunoediting [12].

This conclusion can have both positive and nega-tive implications. The immune system combats tumorcells; hence, the understanding of the mechanisms ofthe combat would allow an increase in its efficiency.On the other hand, in spite of the resistance of the per-fectly healthy immune system, tumor cells can changetheir properties and escape immune surveillance. Sev-eral mechanisms of such escape have been found. Insome cases, tumors change their properties so thatthey are no more recognizable by T cells; for example,tumors decrease or cease the production of major his-

tocompatibility complex (MHC) molecules, whichpresent peptides to T cells, or lose the expression ofthe tumor antigen itself [13]. In other cases, tumorcells activate the signaling pathways that inhibit theimmune response and induce an anergic state incytolytic T lymphocytes. As it became obvious fromrecent studies, the main difficulty lies in the inherentproperties of the immune response regulation, whichinvolves various inhibiting mechanisms [14, 15].

It should be noted that the above results, obtainedmainly by Schreiber et al., are criticized by someresearchers. For example, Oin and Blankenstein [16]failed to reproduce these results and suggested quiteradically that mice deficient in the

γ

-IFN system diedof infections rather than of tumors. This statementraised hot debates [17].

In addition to T cells, NKs, including intraepithe-lial NKs, play an important role in antitumor surveil-lance [18, 19]. The mechanisms of action of thesecells, which belong to the innate immunity system,were not understood until quite recently. In the pastdecade, both activating and inhibiting receptor fami-lies were discovered on NKs [20]. These receptorssupport the role of NKs in the elimination of infectedor abnormal (including malignant) cells by the “miss-ing self” recognition mechanism [21]. In particular,reduced synthesis of MHC class I molecules is typicalof both cells infected by viruses and cells able to pro-liferate during tumor progression. A loss of MHCclass I gene expression allows cells to escape recogni-tion by cytotoxic lymphocytes but exposes them toNKs [13]. One of the necessary conditions for the sur-veillance by NK cells is recognition of some specificligands produced by abnormal cells (including tumorcells) [22, 23].

Table 1.

Early development and reduced immunogenicity of chemically induced and spontaneous tumors in immunocom-promised mice [9]

Experiment Wild-type mice

RAG

–/–

mice

Tumor induction with methylcholanthrene Mice with tumors, %

120 days 0 20

160 days 20 60

Spontaneous tumors Mice with tumors in the second year of life, %

Total, % 20 100

Mice with adenocarcinomas, % 0 50

Tumor growth in secondary recipients* Percentage of exponentially growing tumors

10

5

cells in

RAG

–/–

mic 100%** 100%***

10

6

cells in wild-type mice 100%** 50%***

Notes: * The amount of cells of the spontaneous tumor used for implantation and the recipient genotype are indicated.** Cells of tumors spontaneously arising in wild-type mice were implanted.

*** Cells of tumors spontaneously arising in

RAG

–/–

mice were implanted. Wild-type recipients are capable of rejecting approxi-mately half of tumors developing in immunocompromised donors.

318

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

NEDOSPASOV and KUPRASH

Thus, the conjectured immunological surveillancemechanisms involve both adaptive and innate immu-nity, acting in concert. Another feature of the involve-ment of innate immunity is that tumor cell invasion isaccompanied by tissue damage and repair, whichcauses local inflammation and production of stress-induced molecules [24].

It is pertinent to mention Zinkernagel’s pessimisticopinion that immunological surveillance of tumorsemerging in the postreproductive period of life is evo-lutionarily meaningless. This viewpoint is based onthe notion that evolution of the immune system ofmammals, including human, was governed by abso-lutely different factors, first of all, by combating infec-tions, which was necessary for survival [25, 26].

TUMOR ANTIGENS

The involvement of adaptive immunity in the con-trol of tumorigenesis and tumor growth implies thepresence of specific antigens recognizable by T cellson tumor cells and on antigen-presenting cells.Although the term tumor antigens, including bothantigens produced by epithelial malignant tumors andhematopoietic tumors, is more precise, the term can-cer antigens is often used for antigens expressed bynonepithelial tumors. Moreover, the recognition ofsome antigens (also referred to as cancer antigens) bythe immune system of a patient often correlates withdisease onset and course, although no abnormal pro-duction of these antigens by tumor cells is observed.

The virogenetic cancer theory, proposed by thefounder of Russian oncoimmunology Zilber in the1950s [27], could not be based on the molecularmechanisms of lymphocyte-mediated immunologicalrecognition. However, in light of present knowledge,this theory is in good agreement with the notions oftumor antigens, which may be viral proteins. Never-theless, Zilber’s student Abelev [28] identified one ofthe first tumor antigens as the autologous embryonicprotein

α

-fetoprotein, rather than as a viral protein.

Another seminal discovery in oncoimmunologywas made by Boon and colleagues [30], who foundthe MAGE family of melanoma antigens. The signifi-cance of this discovery reached far beyond the systemwhere it was made. Further study of the so-called can-cer–gamete antigens [30] revealed a class of geneswith unknown functions that are transcribed in the tes-tes and, partly, in the ovaries and placenta but are sup-pressed [31] or weakly expressed [32] in somatic cellsother than tumor cells. It should be noted that the cir-cumstances of MAGE antigen discovery imply thatthese antigens can be recognized by T cells, being pre-sumably suitable as immunotherapy targets [33].

Types of Tumor Antigens

The antigens produced by tumor cells are com-monly divided into T and B antigens. This divisionstems from the differences in the mechanisms of pro-cessing and presentation of antigens for recognitionby B and T cells, as well as from the different experi-mental methods of their identification on the basis ofthese mechanisms. B and T cells recognize differentepitopes, even when they belong to the same antigenicprotein. The correlation between the B- and T-cellresponses to tumor antigens is limited. A high titer ofantibodies against a tumor-produced protein is notalways indicative of the presence of cytotoxic lym-phocytes specific to this protein in the patient, not tomention an efficient T-cell response against the tumor,and vice versa.

Tumor antigens recognizable by the immune sys-tem can be classified according to the supposed causeof their immunogenicity (Table 2). Except for proteinsencoded by mutant genes, all antigens mentioned inTable 2 are normal endogenous cell proteins. Recog-nition of such proteins would seem to be impossibleowing to immunological tolerance, resulting fromselection of T cells in the thymus. However, it is con-ceivable for some antigens (first of all, carcinoembry-onic and cancer–gamete antigens) that T cells recog-nizing them with “desired” low affinity escape nega-tive selection because of the low production of theseantigens by thymic epithelial cells.

It should be mentioned that many cancer antigenshave evolved to perform specific functions requiredfor normal cell biology. As pointed out by R. Medzhi-tov, it is unknown whether natural selection of special-ized mechanisms of tumor detection by the immunesystem ever took place. In any case, no antigen hasbeen convincingly shown to be a product of suchselection.

Precise knowledge of the repertoire of specificantigens produced by a tumor, independent of theoriginal functions of these antigens in the cell, is ofcrucial importance for solving the main problem ofspecific cancer immunotherapy: circumvention ofimmunological tolerance. Tolerance is regulated notonly by controlling the T-cell repertoire but also byadditional mechanisms, in particular, by regulatory(suppressor) T cells [14, 34].

Antibodies and Autoantibodies against Tumor Antigens: Possibilities of Their Use

Some of the above tumor antigens can elicit ahumoral immune response and high-affinity immuno-globulins in the blood [35, 36]. Except for the cases ofmutations, the mechanisms of this response are poorlyunderstood for both patients with malignant tumorsand for the more general case of normal autoantibodyproduction, as well as for many autoimmune diseases.

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

PROBLEMS OF CANCER IMMUNOTHERAPY 319

Tab

le 2

.

Cla

ssif

icat

ion

of tu

mor

ant

igen

s ac

cord

ing

to th

e pr

esum

ptiv

e ca

use

of th

eir

imm

unog

enic

ity

Tum

or a

ntig

en ty

peE

xam

ples

of

antig

ens

and

thei

r ge

nes

Ref

er-

ence

Com

men

ts

1. N

ew e

pito

pes

in n

orm

al c

ell

prot

eins

Dom

inan

t ant

igen

s fo

und

inch

emic

ally

indu

ced

tum

ors

Unk

now

n in

mos

t cas

es[9

0, 9

1]A

che

mic

ally

indu

ced

som

atic

mut

atio

n ca

n ge

nera

te a

new

epi

tope

. With

eff

icie

nt b

ind-

ing

of th

e ne

w p

eptid

e to

cla

ss I

MH

C m

olec

ules

, an

effi

cien

t T-c

ell-

med

iate

d re

spon

se

can

be e

licite

d. M

olec

ular

iden

tific

atio

n of

the

antig

ens

with

exp

erim

enta

l mod

els

is h

am-

pere

d by

the

som

atic

cha

ract

er o

f th

e m

utat

ions

Tum

or s

uppr

esso

r p5

3

p53

[35]

Man

y

p53

mut

atio

ns b

oth

stab

ilize

the

prot

ein

and

chan

ge it

s co

nfor

mat

ion.

As

a re

sult,

man

y ep

itope

s hi

dden

with

in th

e p5

3 te

tram

er b

ecom

e ac

cess

ible

for T

and

B ly

mph

ocyt

es in

volv

ed

in th

e re

cogn

ition

of a

ntig

ens

rele

ased

as

a re

sult

of a

popt

osis

/nec

rosi

s of

tum

or c

ells

Alte

rnat

ive

splic

ing

prod

ucts

Res

tin, a

ssoc

iate

d w

ith

Hod

gkin

's d

isea

se[3

7]M

any

splic

ing

vari

ants

are

pro

duce

d on

ly in

the

brai

n or

test

es, w

hich

are

imm

unop

rivi

-le

ged

orga

ns, a

nd in

tum

ors.

The

reg

ions

enc

oded

by

alte

rnat

ive

exon

s ar

e pe

rcei

ved

as

new

ant

igen

s by

the

imm

une

syst

emSp

licin

g va

rian

t of

OA

NN

1

[92]

2. O

verp

rodu

ced

antig

ens

eIF-

4

γ

in lu

ng tu

mor

s[9

3]

The

imm

une

syst

em is

gen

eral

ly to

lera

nt to

ant

igen

s pr

oduc

ed in

nor

mal

tiss

ues.

How

ev-

er, i

n so

me

case

s, o

verp

rodu

ctio

n of

an

antig

en in

the

tum

or is

suf

fici

ent f

or o

verc

omin

g im

mun

olog

ical

tole

ranc

eC

orta

ctin

[94]

Tel

omer

ase

[95]

3. V

iral

ant

igen

sH

PV, E

BV

, HE

RV

-K, H

SV,

hepa

titis

vir

uses

[83,

84]

Vir

al p

rote

ins

are

alie

n fo

r th

e bo

dy a

nd a

re e

ffic

ient

ly r

ecog

nize

d by

the

imm

une

syst

em

4. A

ntig

ens

with

lim

ited

prod

uc-

tion

in n

orm

al ti

ssue

s

Dif

fere

ntia

tion

antig

ens

Mel

an-A

, tyr

osin

ase

[96,

97]

For

man

y tis

sues

, ind

uctio

n of

the

imm

une

resp

onse

aga

inst

dif

fere

ntia

tion

antig

ens

is

an a

ppro

pria

te tr

eatm

ent

Can

cer–

gam

ete

antig

ens

MA

GE

-1

,

NY

-ESO

-1

[31]

Tes

tis is

an

imm

unop

rivi

lege

d or

gan,

who

se p

rote

in p

rodu

cts

are

invi

sibl

e fo

r the

imm

une

syst

em. C

urio

usly

, the

bio

logi

cal r

ole

of m

ost c

ance

r–ga

met

e an

tigen

s is

unk

now

n

5. C

hrom

osom

al tr

ansl

ocat

ion

prod

ucts

p210

BC

R-A

BL

[98]

Prod

ucts

of c

hrom

osom

al tr

ansl

ocat

ions

invo

lved

in tr

ansf

orm

atio

n ca

n co

ntai

n a

pept

ide

at th

e bo

unda

ry o

f th

e tw

o pr

otei

n se

quen

ces

that

elic

its a

n ef

fici

ent i

mm

une

resp

onse

6. T

umor

-ass

ocia

ted

auto

antig

ens

Num

erou

s ex

ampl

es[9

9, 1

00]

It is

sug

gest

ed th

at im

mun

olog

ical

tole

ranc

e ca

n be

ove

rcom

e if

the

imm

une

syst

em r

ecog

-ni

zes

auto

antig

ens

in th

e co

ntex

t of

necr

osis

or

apop

tosi

s oc

curr

ing

in th

e gr

owin

g tu

mor

320

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

NEDOSPASOV and KUPRASH

It is thought that autoantibodies against intracellularantigens form in response to cell lysis, which accom-panies repair of injured tissues, tumor cell invasion,etc. This assumption can partly explain the formationof tumor-specific autoantibodies and indicates thatmost antigens recognized by such autoantibodies can-not serve as therapeutic targets for T lymphocytes.

It should be emphasized that antibodies againsttumor antigens, including those obtained by bioengi-neering, are widely used in diagnosis and, in somecases, in treatment of malignant tumors (see below).

At present, there are two practical applications ofautoantibodies against tumor antigens. First, the useof cancer patient serum antibodies for immunoscreen-ing of expression gene libraries allows the identifica-tion of rare tumor antigens appropriate as immuno-therapy targets. Serological identification of recombi-nantly expressed clones (SEREX) is an example ofthis approach [37]. SEREX was used for cloning anumber of new autoantigens, including the cancer–gamete antigen NY-ESO-1 [30], more recently stud-ied, in detail, both experimentally and clinically [38].Second, quantitation of tumor-specific autoantibodies(frequently detected in cancer patients but rarely, if atall, found in healthy persons) can be used for noninva-sive cancer diagnosis and for tracing the effect of anti-tumor therapy (in particular, immunotherapy) in can-cer patients [39–41].

DOMINANT IMMUNOLOGICAL TOLERANCE AS THE MAIN OBSTACLE

FOR THE ANTITUMOR IMMUNE RESPONSE

Recent studies performed with both experimentalmodels and patients have demonstrated that peripheralcytolytic

CD8

+

T cells with high affinity for tumorantigens can appear in an immunocompetent organ-ism. Moreover, such T cells can infiltrate the tumortissue [42]. Seemingly, modern advances in the exvivo production of such T cells and the possibility oftheir adoptive transfer allow mobilization of a widerepertoire of highly specific effector killer cells [43,44]. Then why do they fail to destroy the tumor inmost cases?

It has been found that progressive tumors can uti-lize intrinsic regulatory mechanisms of the immunesystem for survival. In addition to radical protectivemeans such as a loss of a tumor antigen or arrest of theproduction of class I MHC [45], mechanisms usingso-called regulatory T cells are involved. Regulatory Tcells constitute a subpopulation of

CD4

+

lymphocytesthat produce the CD25, the

α

subunit of the interleu-kin 2 (IL-2) receptor. These cells can infiltrate thetumor and suppress the activity of proximal effector Tlymphocytes. The development of these special-pur-pose cells and their suppressor effect are governed bythe intracellular transcription factors Foxp3 and NFAT

[46] and by TGF-

β

, produced by cells of the microen-vironment of the tumor [15]. The molecular mecha-nisms governing the suppression of the autoreactive T-cell response are now a subject of extensive studies,and new important information is accumulating. Inparticular, a new population of T cells, Th17, has beenrecently recognized and described. Th17 cells play akey role in autoimmune conditions [47]. These cells,producing IL-17, differ from both Th1 and Th2 T-helper types. Like regulatory T cells, they depend onTGF-

β

[48].

At present, clinical trials of the reduction of thelevel of regulatory T cells in cancer patients with theuse of IL-2-based immunotoxins or antibodies againstCD25 are in progress. Such preparations also killbeneficent cytotoxic T cells, but it is believed that thisapproach can be efficient if followed by adoptivetransfer of activated cytotoxic lymphocytes [49].Undoubtedly, a precise control of the population ofregulatory T cells requires the identification of thepopulation-specific surface markers and design ofnew reagents specific for these markers. At the sametime, the specificity of tumor cell recognition can beimproved using new tumor-associated antigens. Thefinal goal of manipulations with regulatory T cells isto reach a therapeutic balance between antitumorcytotoxicity and autoimmunity; that is, to reliablyeliminate tumor cells while maintaining an acceptablelevel of autoimmune damage to healthy tissues.

TUMOR IMMUNOTHERAPY

Nonspecific Immunotherapy

This approach includes cytokine-based therapy,various innate immunity activators, etc. (Table 3).

Initially, cytokine therapy was considered verypromising because of its simplicity and the possibilityof its combination with other well-developed treat-ment methods [50]. Unfortunately, systemic applica-tion of the cytokines that seemed most promising inexperiments in vitro or in model animals proved to beunacceptable because of their toxicity, related todiversity of biological effects. Trials of the tumornecrosis factor (TNF),

γ

-IFN, GM-CSF, IL-12, andsome others were terminated, although TNF and

γ

-IFN continue to be used in combination with Mel-phalan for isolated limb perfusion [51, 52].

Type I interferons and some clinical protocols withlow concentrations of IL-2, too toxic at higher con-centrations, are still in use for cancer treatment [53].

It should be noted that the most important applica-tion of cytokines in cancer therapy is supportive treat-ment with erythropoietin and G-CSF, aimed at elevat-ing the levels of red cells and neutrophils to preventtheir dangerous decrease in many chemotherapeuticalprotocols [54].

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

PROBLEMS OF CANCER IMMUNOTHERAPY 321

Tab

le 3

.

App

roac

hes

to im

mun

othe

rapy

of

mal

igna

nt tu

mor

s

No.

Des

crip

tion

of a

ther

apeu

tic a

ppro

ach

Imm

unol

ogic

al m

echa

nism

and

fact

ors

favo

ring

suc

cess

ful t

reat

men

tO

bsta

cles

Ref

eren

ce

1A

ctiv

atio

n of

non

spec

ific

imm

unity

, bac

teri

al

adju

vant

s, in

clud

ing

trea

tmen

t with

non

spec

ific

cy

toki

nes

(IL

-2 a

nd G

M-C

SF)

Bas

ed o

n th

e no

tion

that

spe

cifi

c T

cel

ls e

ither

do

not r

ecei

ve a

ny c

ostim

ulat

ory

sign

al o

r th

e co

sti-

mul

ator

y si

gnal

s ar

e no

t suf

fici

ent.

The

met

hod

is p

rom

isin

g in

com

bina

tion

with

oth

er a

ntig

en-

spec

ific

trea

tmen

ts

The

app

roac

h is

lim

ited

by to

xici

ty a

nd is

inef

-fe

ctiv

e in

man

y ca

ses.

Pro

per

eval

uatio

n of

the

deve

lopi

ng im

mun

e re

spon

se is

dif

ficu

lt or

im-

poss

ible

[101

]

2Im

mun

izat

ion

with

a s

peci

fic

antig

en (

reco

mbi

-na

nt p

rote

in, p

eptid

es, r

ecom

bina

nt n

onpa

thog

e-ni

c vi

rus

expr

essi

ng a

tum

or a

ntig

en, p

eptid

es in

co

mpl

ex w

ith h

eat s

hock

pro

tein

s)

Act

ivat

ion

of th

e cl

assi

c an

tigen

-spe

cifi

c im

mu-

ne r

espo

nse

resu

lting

in g

ener

atio

n of

hig

h-af

fi-

nity

cyt

olyt

ic T

lym

phoc

ytes

. Not

toxi

c. M

any

imm

uniz

atio

n pr

otoc

ols

are

wel

l tol

erat

ed

Tum

or-s

peci

fic

T c

ells

for

m a

nd in

filtr

ate

the

tu-

mor

but

app

ear

to b

e su

ppre

ssed

by

tum

or c

ells

or

reg

ulat

ory

T c

ells

[81,

102

,10

3, 1

11,

112]

3V

acci

nes

base

d on

mod

ifie

d tu

mor

cel

ls (p

atie

nt's

own)

Bas

ed o

n th

e su

gges

tion

that

the

patie

nt h

as“c

orre

ct”

T c

ells

, but

they

eith

er fa

il to

reco

gniz

e th

e tu

mor

ant

igen

or

are

pres

ent i

n in

suff

icie

nt

quan

titie

s

Indi

vidu

al tr

eatm

ent.

Unb

iase

d m

onito

ring

of

the

imm

une

resp

onse

in th

e pa

tient

is d

iffi

cult,

if

at a

ll po

ssib

le

[66,

104

]

4V

acci

nes

base

d on

den

driti

c ce

llsA

ctiv

atio

n of

spe

cifi

c T

cel

ls w

ith a

ntig

en-p

re-

sent

ing

cells

, whi

ch c

an a

lso

give

a s

timul

atin

g si

gnal

to T

cel

ls

The

fun

ctio

n of

end

ogen

ous

dend

ritic

cel

ls in

a

canc

er p

atie

nt is

oft

en d

isru

pted

. Thi

s ca

n m

ake

vacc

inat

ion

inef

fici

ent.

Mor

eove

r, th

ere

can

be

an e

xces

sive

aut

oim

mun

e re

spon

se

[105

, 106

]

5V

acci

nes

base

d on

ado

ptiv

e tr

ansf

er o

f T

cel

ls,

incl

udin

g in

filtr

atin

g ly

mph

ocyt

es. C

an b

e ap

-pl

ied

in c

ombi

natio

n w

ith d

eple

tion

of p

atie

nt's

ly

mph

ocyt

es

Bas

ed o

n th

e fa

ct th

at ly

mph

ocyt

es in

filtr

atin

g tu

mor

s in

clud

e hi

gh-a

ffin

ity c

ytol

ytic

T c

ells

, w

hich

can

be

puri

fied

and

pro

paga

ted

Indi

vidu

al th

erap

y. D

iffi

cult

to p

erfo

rm, e

spe-

cial

ly in

com

bina

tion

with

med

icin

al in

activ

atio

n of

the

patie

nt's

lym

phoc

ytes

[42,

107

]

6V

acci

nes

base

d on

ado

ptiv

e tr

ansf

er o

f re

com

bi-

nant

T c

ells

tran

sfec

ted

with

T-c

ell r

ecep

tors

re-

cogn

izin

g th

e “c

orre

ct”

tum

or a

ntig

en in

the

con-

text

of

the

corr

ect M

HC

The

oret

ical

ly, a

per

fect

gen

e-im

mot

hera

peut

ical

ap

proa

ch. B

ypas

ses

man

y di

ffic

ultie

s re

late

d to

in

suff

icie

nt a

ntig

en d

ensi

ty o

r in

suff

icie

nt r

ecep

-to

r af

fini

ty

Indi

vidu

al th

erap

y. V

ery

diff

icul

t to

perf

orm

. D

eman

ds k

now

ledg

e of

at l

east

one

ant

igen

(o

r, b

ette

r, s

ever

al a

ntig

ens)

[108

, 109

]

7E

limin

atio

n or

blo

ckad

e of

reg

ulat

ory

(sup

pres

-so

r) T

lym

phoc

ytes

Bas

ed o

n th

e “t

urni

ng d

own”

mec

hani

sm o

f im

-m

une

supp

ress

ion

in a

con

trol

labl

e w

ayIn

duct

ion

of a

utoi

mm

une

cond

ition

s, w

hich

can

be

a li

miti

ng f

acto

r[1

10]

8A

pplic

atio

n of

ant

ibod

ies

as e

ffec

tor

imm

uno-

ther

apeu

tic to

ols

Ant

ibod

ies

can

be u

sed

for c

ontr

ollin

g th

e si

gnal

-in

g pa

thw

ays

rela

ted

to th

e ac

tivity

of a

tum

or a

n-tig

en (

e.g.

, rec

epto

r), e

limin

atio

n of

cel

l pop

ula-

tions

exp

ress

ing

the

antig

en, o

r tar

gete

d de

liver

y of

a to

xic

agen

t to

the

tum

or

Fund

amen

tally

impo

ssib

le to

com

plet

ely

elim

i-na

te a

ll tu

mor

cel

ls[7

7, 7

8]

322

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

NEDOSPASOV and KUPRASH

Specific Immunotherapy

The goal of specific immunotherapy is to elicit orenhance the immune response to tumor antigens, evenif the antigens have not been identified.

Specific immunotherapy includes various types ofvaccination: prophylactic (a brilliant example is pro-vided by recently successful clinical trials of vaccinespreventing virus-associated cervical cancer [55]) andtherapeutic ones. Therapeutic vaccination involvesadministration of autologous cancer cells obtainedfrom the patient (generally, after passages and variousin vitro manipulations), various kinds of immuniza-tion against specific tumor antigens (with recombi-nant proteins, peptides, or viruses harboring such anti-gens) together with various adjuvants (in particular,innate immunity activators) [56].

Adoptive transfer of specific T cells, includinggenetically modified T cells [57] or dendritic cellsengineered to present certain tumor antigens [58], alsobelongs to specific immunotherapy.

Finally, specific immunotherapy includes clinicalapplications of specific antibodies against tumor anti-gens as well as immunoconjugates and immunotoxinsdesigned on their basis [59].

OUTLOOK FOR DESIGN OF ANTITUMOR VACCINES

Preventive Vaccines

It is not a new idea to develop protective vaccinesthat would elicit stable immunity against tumor cellsin healthy people provided the cells bear certain tumorantigen(s) on their surfaces [60]. Similar vaccinationsallowed immunologists to overcome many infectiousdiseases, including some virus-associated ones [61].

The origin of certain types of cancer is related toviruses. It is known that cervical cancer is associatedwith the papillomavirus [62], while hepatocellularcarcinoma is associated with the hepatitis B and Cviruses [63]. No wonder that virus-associated tumorsare the most promising targets for preventive vaccina-tion. However, it should be emphasized that, strictlyspeaking, the vaccines are antiviral rather than antitu-mor. Other important virus targets include EBV,HTLV-1, and HHV-8, related to various lymphomas,especially in immunocompromised patients [64].

Basic immunology points to a fundamental differ-ence between viral (replicating) antigens and autoan-tigens. The former can become the target of an effi-cient immune response when either T lymphocytesdestroy virtually all infected cells or efficient neutral-izing antibodies are produced. In both cases, activa-tion of specific T lymphocytes is considered to be themain stage of the immune response, and the repertoireof these potentially useful T cells does not undergo

selection in the thymus, because the antigen to be rec-ognized is genuinely foreign.

A different situation arises with autoantigens,which, for some reasons, are produced in tumor butnot in normal cells. Attempts to induce a prolongedefficient immune response against such target anti-gens encounter certain obstacles, such as partial toler-ance of autoantigens, their low density on the surface(or even the absence of exposed epitopes in the case ofmany intracellular molecules), inability to generateefficient memory cells and produce neutralizing anti-bodies, etc. Current intense studies are dedicated tosolving the fundamental problem of dominant toler-ance, which is eventually responsible for suppressionof the effector function of T lymphocytes even whenthey have been formed in sufficient quantities andinfiltrated the tumor environment [65].

Therapeutic Vaccination

Various protocols of active immunotherapy aimedat specific tumor antigens (even when the antigens areunidentified, e.g., when vaccination is performed withautologous cells) belong to the most complex thera-peutic approaches. The advance in this field is rathermodest.

The most recent and promising protocols of“ideal” vaccination are aimed at (1) eliciting the spe-cific adaptive immune response upon administrationof a tumor antigen within a nonpathogenic virus (sig-nificant advantage is taken of replication of the viralDNA-encoded antigen), (2) concurrently activatinginnate immunity with various adjuvants (to providecostimulating signals necessary for T cells), and (3)completely or partly blocking the main inhibitorypathways controlling peripheral tolerance (Table 3).

OTHER APPROACHES TO ANTIGEN-SPECIFIC IMMUNOTHERAPY

The three approaches described in this section(Table 3) are based on using tumor cells or theimmune system of the patient. The patient’s cells aretreated by various methods under laboratory condi-tions and transferred back into the body. Apparently,these immunotherapeutic approaches are individualand, therefore, too complicated and expensive for rou-tine use.

Manipulations involving gene transfer are collec-tively called gene therapy. It should be clearly under-stood that gene therapy is actually only part of com-plex immunotherapy.

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

PROBLEMS OF CANCER IMMUNOTHERAPY 323

Manipulations with Autologous Tumor and Antigen-Presenting Cells

The modern version of this approach was proposedabout 15 years ago. It consists in enhancing tumor cellimmunogenicity by transfection with genes encodingcytokines and other factors able to activate adaptive orinnate immunity branches. For example, introductionof the GM-CSF gene enhances the antitumor response[66], probably, by acting on antigen-presenting cells.Another original approach involves

Tag-7

, encodingone of the proteins involved in regulation of innateimmunity [67].

As a rule, approaches of this sort leave the natureof tumor antigens unidentified. It is assumed that theirproduction is maintained in tumor cells, which can beverified in vitro by treatment with activated cytolyticT cells isolated from the patient’s blood cells, and thatthey are recognized in vivo.

The weak side of this approach is that the specificimmune response is extremely difficul to monitor, asit is necessary to count

CD8

+

T lymphocytes anddetermine their functionality. Typical parameters(affinity and concentration) of the T-cell response aresuch that a monitoring of specific T cells of a patientrequires tetramer technology [68], which, in turn,requires molecular identification of the antigens that Tcells recognize.

Another approach under development relies on theimprovement of the properties of antigen-presentingcells in vitro followed by their transfer into thepatient’s body. For example, dendritic cells are iso-lated from peripheral blood cells of a patient andtransfected with tumor antigens or factors enhancingtheir ability to activate T cells [58].

Improvement of the Efficacy of Antitumor T Lymphocytes

One of the early approaches to obtaining lympho-cytes suitable for therapeutic transfer to cancerpatients involved a culturing of so-called lymphokine-activated killer cells (i.e., T lymphocytes from theperipheral blood of a patient), which was followed bytreatment with IL-2 or a cytokine cocktail [69]. Later,a more efficient approach was proposed, whichincluded isolation of small amounts of so-calledtumor-associated T lymphocytes from tumor samples,their propagation in vitro, activation, and transferback to the patient’s body [43]. The idea of the methodwas that T cells infiltrating the tumor contain “cor-rect” cytolytic T cells recognizing tumor antigens.Indeed, as mentioned above, such lymphocytesinclude high-affinity effector-competent

CD8

+

T cells.If their amount can be augmented and they can be acti-vated, one may hope that, after transfer, they can infil-trate the tumor again and kill tumor cells.

At present, the most remarkable achievement inindividual immunotherapy is the approach in whichone or more genes for a rearranged T-cell receptor(s)able to recognize a tumor antigen(s) within MHC mol-ecules of the patient are transferred to his T lympho-cytes [57, 70]. Such lymphocytes, transformed andspecific to a certain antigen, are activated and trans-ferred to the patient. Obviously, cloning manipula-tions in this approach are rather complicated. Theyrequire not only knowledge of the nature of the anti-gen but also availability of constructs allowing theartificial T-cell receptor to recognize the antigen in thecomplex with MHC molecules from various patients.Another difficulty is that this method requires individ-ual test reagents for monitoring the antitumor immuneresponse (e.g., so-called tetramers).

Note that none of the protocols presented in Table 3demonstrates unambiguous efficacy, although objec-tive therapeutic success has been reported in somecases [71].

Application of Modified Antibodies to Cancer Diagnosis and Treatment

Antibodies have already been of limited use as aneffector tool in cancer immunotherapy, and undeni-able clinical success has been achieved.

First, several formulations based on monoclonalantibodies are used in adjuvant therapy protocols. Thebest known of them are “therapeutic” antibodiesagainst the Her2 tumor cell marker and CD20 B-celldifferentiation marker [72, 73]. Second, antibodiesagainst a wider range of tumor antigens are applied fordisease diagnosis and monitoring and for tumor local-ization. For the latter purpose, antibodies are labeledwith either radioactive isotopes [74] or dyes (e.g., flu-orescent) [75].

A specific line of research is an attempt to develop“effector” antibodies conjugated with an agent able todestroy tumor cells [76]. These tools include antibod-ies bound with isotopes, which cause radioactive dam-age to tumor cells [77], or with various toxins (immu-notoxins) [78].

These approaches cannot eliminate all cells bear-ing a certain tumor antigen; therefore, a single courseof such immunotherapy can only have a limited suc-cess.

Present State of Clinical Trials of Various Immunotherapy Protocols

After the discovery of tumor antigens, includingcancer–gamete antigens, many research teamslaunched clinical studies of various vaccination meth-ods (with peptides or recombinant proteins) using var-ious adjuvants, etc. Currently, the most promising for-mulations include vaccines based on the smallpox

324

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

NEDOSPASOV and KUPRASH

virus with an adjuvant cocktail. Such recombinantvirus can encode more than one tumor antigen [79].

Numerous studies performed using sophisticatedmethods have convincingly demonstrated the possi-bility of induction of “correct” antigen-specificcytolytic T lymphocytes and their infiltration to thetumor [80, 81]. As in viral infections, a significant, ifnot major, portion of

CD8

+

T lymphocytes of patientshad the required specificity, being able to kill tumorcells in vitro. Unfortunately, no clear reproducibleclinical effects were observed in these studies [82]. Tomaintain objectivity, note that early phases of antitu-mor immunization or adoptive transfer trials includepatients with severe malignancies resistant to all avail-able treatment methods.

It should also be mentioned that, although adoptivetransfer of T cells after ex vivo manipulations provedto be applicable for tumor treatment, it will hardly beaccessible for the majority of patients, even in devel-oped countries. Therefore, one cannot predict, now,whether this immunotherapeutic method will be ableto supersede other modern cancer treatment methods.

PRACTICAL IMMUNOLOGY AND BASIC RESEARCH

The revival of the immunological surveillance con-cept coincided with a series of new ideas in cancerimmunotherapy, which were based on importantknowledge obtained by basic studies. An example isthe notion that surveillance could be performed notonly by T cells. This knowledge considerably broad-ened the possibilities of therapy. Analysis of the scien-tific support of clinical trials of various cancer immu-notherapy protocols in countries with advanced clini-cal immunology indicates that this field ofbiomedicine is science intensive. Knowledge ofachievements in basic immunology and biology of thetumor cell will allow clinicians to put forward moresound, practical ideas and provide additional chancesfor the final success.

ONCOIMMUNOLOGY PROGRAM IN RUSSIA

The educational program Oncoimmunology(www.oncoimmunology.ru), which we organized, hasspread modern knowledge of cancer immunology inRussia for more than six years. During this time,nearly all distinguished scholars in modern cancerimmunology, whose works are mentioned in all sec-tions of the review, visited Moscow and gave lectures.

Schreiber was one of the first lecturers of the pro-gram. He presented the revitalized immunologicalsurveillance theory in Moscow in the year of its pub-lication [9]. He was accompanied by Greenberg, apioneer of clinical approaches based on adoptivetransfer of T cells [42]. Lectures in Moscow were also

given by researchers who had discovered the best-known human tumor antigens: Abelev [28], Boon[33], Knuth [29], and Pfreundschuh [37]. The lectur-ers of the program Boon, Knuth, Jager, and Pfreunds-chuh conducted or are conducting trials of clinicalprotocols involving therapeutic antitumor vaccination[82]. Lectures in Moscow covered their results, prob-lems to be solved, and unsolved contradictions. Sev-eral lectures and small courses dedicated to the rela-tionships between viruses and cancer were deliveredby the outstanding scientists H. zur Hausen, G. Klein,and E. Klein, who had made the major contribution tothe study of the papillomavirus and Epstein–Barrvirus, and revealed a correlation between the virusesand tumors [83, 84]. Lectures by Blankenstein [85],Hayday [86], and Zinkernagel [87] were dedicated tothe immunological surveillance theory. Therapeuticapproaches based on the use of antibodies (includingmodified antibodies) were covered in lectures byJ.P. Mach and Pfreundschuh.

A significant contribution to the discussion ofmodern problems and challenges in oncoimmunologywas made in brilliant lectures by outstanding Russianinvestigators working abroad: K. Rajewsky, R. Med-zhitov, A. Rudensky, and A. Chervonsky. Most ofthese lectures were videotaped. The records are com-monly available from the archive of the program.

We sincerely hope that our educational activity aspart of the Oncoimmunology program (which will becontinued) as well as this review will draw young,active, and purposeful researchers to this importantfield of theoretical and practical biomedicine.

We note, again, that this review was not aimed atcomprehensively covering all problems of tumorimmunology, particularly, its history. These issues areconsidered in much more detail in Russian and foreignreviews (e.g., [88, 89]). We confined ourselves to dis-cussing problems raised in lectures of the Oncoimmu-nology program.

ACKNOWLEDGMENTS

We are grateful to A.Yu. Rudensky, Yu.V. Sheb-zukhov, and, especially, P.V. Belousov for criticalreview of the manuscript and valuable comments.

This work was supported by the Federal programHigh-Priority Research and Development in Scienceand Technology, the program Molecular and CellBiology of the Presidium of the Russian Academy ofSciences, and the Russian Foundation for BasicResearch (project no. 05-04-49075).

REFERENCES

1. Burnet F.M. 1970. The concept of immunological sur-veillance.

Prog. Exp. Tumor Res

.

13

, 1–27.

MOLECULAR BIOLOGY

Vol. 41

No. 2

2007

PROBLEMS OF CANCER IMMUNOTHERAPY 325

2. Foley E.J. 1952. Immunity of C3H mice to lymphosar-coma 6-C3H-Ed following regression of the implantedtumor.

Proc. Soc. Exp. Biol. Med

.

80

, 675–677.3. Foley E.J. 1953. Antigenic properties of methylcholan-

threne-induced tumors in mice of the strain of origin.

Cancer Res

.

13

, 835–837.4. Sjogren H.O. 1961. Further studies on the induced

resistance against isotransplantation of polyomatumors.

Virology

.

15

, 214–219.5. Deichman G.I. 1969. Immunological aspects of car-

cinogenesis by deoxyribonucleic acid tumor viruses.

Adv. Cancer Res

.

12

, 101–136.6. Stutman O. 1974. Tumor development after 3-methyl-

cholanthrene in immunologically deficient athymic-nude mice.

Science

. 183, 534–536.7. Mombaerts P., Iacomini J., Johnson R.S., Herrup K.,

Tonegawa S., Papaioannou V.E. 1992. RAG-1-deficientmice have no mature B and T lymphocytes. Cell. 68,869–877.

8. Kirchgessner C.U., Patil C.K., Evans J.W., Cuomo C.A.,Fried L.M., Carter T., Oettinger M.A., Brown J.M.1995. DNA-dependent kinase (p350) as a candidategene for the murine SCID defect. Science. 267, 1178–1183.

9. Shankaran V., Ikeda H., Bruce A.T., White J.M., Swan-son P.E., Old L.J., Schreiber R.D. 2001. IFNγ and lym-phocytes prevent primary tumor development and shapetumor immunogenicity. Nature. 410, 1107–1111.

10. Kaplan D.H., Shankaran V., Dighe A.S., Stockert E.,Aguet M., Old L.J., Schreiber R.D. 1998. Demonstra-tion of an interferon γ-dependent tumor surveillancesystem in immunocompetent mice. Proc. Natl. Acad.Sci. USA. 95, 7556–7561.

11. Gao Y., Yang W., Pan M., Scully E., Girardi M., Augen-licht L.H., Craft J., Yin Z. 2003.γδ T cells provide anearly source of interferon γ in tumor immunity. J. Exp.Med. 198, 433–442.

12. Dunn G.P., Old L.J., Schreiber R.D. 2004. The immu-nobiology of cancer immunosurveillance and immuno-editing. Immunity. 21, 137–148.

13. Seliger B., Ritz U., Ferrone S. 2006. Molecular mecha-nisms of HLA class I antigen abnormalities followingviral infection and transformation. Int. J. Cancer. 118,129–138.

14. O’Garra A., Vieira P. 2004. Regulatory T cells andmechanisms of immune system control. Nat. Med. 10,801–805.

15. Bierie B., Moses H.L. 2006. Tumor microenvironment:TGFβ: The molecular Jekyll and Hyde of cancer. Nat.Rev. Cancer. 6, 506–520.

16. Qin Z., Blankenstein T. 2004. A cancer immunosurveil-lance controversy. Nat. Immunol. 5, 3–4.

17. Schreiber R.D., Old L.J., Hayday A.C., Smyth M.J.2004. Response to “a cancer immunosurveillance con-troversy.” Nat. Immunol. 5, 4–5.

18. Tagliabue A., Luini W., Soldateschi D., Boraschi D.1981. Natural killer activity of gut mucosal lymphoidcells in mice. Eur. J. Immunol. 11, 919–922.

19. Beagley K.W., Husband A.J. 1998. Intraepithelial lym-phocytes: Origins, distribution, and function. Crit. Rev.Immunol. 18, 237–254.

20. Held W., Coudert J.D., Zimmer J. 2003. The NK cellreceptor repertoire: Formation, adaptation and exploita-tion. Curr. Opin. Immunol. 15, 233–237.

21. Raulet D.H. 2006. Missing self recognition and self tol-erance of natural killer (NK) cells. Semin. Immunol. 18,145–150.

22. Raulet D.H. 2003. Roles of the NKG2D immunorecep-tor and its ligands. Nat. Rev. Immunol. 3, 781–790.

23. Bahram S., Inoko H., Shiina T., Radosavljevic M. 2005.MIC and other NKG2D ligands: From none to toomany. Curr. Opin. Immunol. 17, 505–509.

24. Cairns R.A., Khokha R., Hill R.P. 2003. Molecularmechanisms of tumor invasion and metastasis: An inte-grated view. Curr. Mol. Med. 3, 659–671.

25. Zinkernagel R.M. 2001. Immunity against solidtumors? Int. J. Cancer. 93, 1–5.

26. Zinkernagel R.M., Hengartner H. 2004. On immunityagainst infections and vaccines: Credo 2004. Scand. J.Immunol. 60, 9–13.

27. Zilber L.A. 1961. On the interaction between tumorviruses and cells: A virogenetic concept of tumorigene-sis. J. Natl. Cancer Inst. 26, 1311–1319.

28. Abelev G.I. 1968. Production of embryonal serumα-globulin by hepatomas: Review of experimental andclinical data. Cancer Res. 28, 1344–1350.

29. Van der Bruggen P., Traversari C., Chomez P., Lur-quin C., De Plaen E., Van den E.B., Knuth A., Boon T.1991. A gene encoding an antigen recognized bycytolytic T lymphocytes on a human melanoma. Sci-ence. 254, 1643–1647.

30. Chen Y.T., Scanlan M.J., Sahin U., Tureci O., Gure A.O.,Tsang S., Williamson B., Stockert E., Pfreundschuh M.,Old L.J. 1997. A testicular antigen aberrantly expressedin human cancers detected by autologous antibodyscreening. Proc. Natl. Acad. Sci. USA. 94, 1914–1918.

31. Scanlan M.J., Simpson A.J., Old L.J. 2004. The can-cer/testis genes: Review, standardization, and commen-tary. Cancer Immun. 4, 1.

32. Simpson A.J., Caballero O.L., Jungbluth A., Chen Y.T.,Old L.J. 2005. Cancer/testis antigens, gametogenesisand cancer. Nat. Rev. Cancer. 5, 615–625.

33. Boon T., van der Bruggen P. 1996. Human tumor anti-gens recognized by T lymphocytes. J. Exp. Med. 183,725–729.

34. Nishikawa H., Kato T., Tawara I., Saito K., Ikeda H.,Kuribayashi K., Allen P.M., Schreiber R.D., Sakaguchi S.,Old L.J., Shiku H. 2005. Definition of target antigensfor naturally occurring CD4+ CD25+ regulatory T cells.J. Exp. Med. 201, 681–686.

35. Schlichtholz B., Legros Y., Gillet D., Gaillard C.,Marty M., Lane D., Calvo F., Soussi T. 1992. Theimmune response to p53 in breast cancer patients isdirected against immunodominant epitopes unrelated tothe mutational hot spot. Cancer Res. 52, 6380–6384.

36. Disis M.L., Calenoff E., McLaughlin G., Murphy A.E.,Chen W., Groner B., Jeschke M., Lydon N., McGlynn E.,Livingston R.B. 1994. Existent T-cell and antibodyimmunity to HER-2/neu protein in patients with breastcancer. Cancer Res. 54, 16–20.

37. Sahin U., Tureci O., Schmitt H., Cochlovius B.,Johannes T., Schmits R., Stenner F., Luo G., Schobert I.,

326

MOLECULAR BIOLOGY Vol. 41 No. 2 2007

NEDOSPASOV and KUPRASH

Pfreundschuh M. 1995. Human neoplasms elicit multi-ple specific immune responses in the autologous host.Proc. Natl. Acad. Sci. USA. 92, 11810–11813.

38. Gnjatic S., Nishikawa H., Jungbluth A.A., Gure A.O.,Ritter G., Jager E., Knuth A., Chen Y.T., Old L.J. 2006.NY-ESO-1: Review of an immunogenic tumor antigen.Adv. Cancer Res. 95, 1–30.

39. Goldstein M.J., Mitchell E.P. 2005. Carcinoembryonicantigen in the staging and follow-up of patients withcolorectal cancer. Cancer Invest. 23, 338–351.

40. Shebzukhov Y.V., Koroleva E.P., Khlgatian S.V., Lagar-kova M.A., Meshcheryakov A.A., Lichinitser M.R.,Karbach J., Jager E., Kuprash D.V., Nedospasov S.A.2005. Humoral immune response to thymidylate syn-thase in colon cancer patients after 5-FU chemotherapy.Immunol. Lett. 100, 88–93.

41. Shebzukhov Y.V., Koroleva E.P., Khlgatian S.V.,Belousov P.V., Kuz’mina K.E., Radko B.V., Longpre F.,Lagarkova M.A., Kadachigova T.S., Gurova O.V.,Meshcheryakov A.A., Lichinitser M.R., Knuth A.,Jager E., Kuprash D.V., Nedospasov S.A. 2005. Anti-body response to a non-conserved C-terminal part ofhuman histone deacetylase 3 in colon cancer patients.Int. J. Cancer. 117, 800–806.

42. Blattman J.N., Greenberg P.D. 2004. Cancer immuno-therapy: A treatment for the masses. Science. 305, 200–205.

43. Rosenberg S.A., Spiess P., Lafreniere R. 1986. A newapproach to the adoptive immunotherapy of cancer withtumor-infiltrating lymphocytes. Science. 233, 1318–1321.

44. Riddell S.R. 2004. Finding a place for tumor-specific Tcells in targeted cancer therapy. J. Exp. Med. 200,1533–1537.

45. Khong H.T., Restifo N.P. 2002. Natural selection oftumor variants in the generation of “tumor escape” phe-notypes. Nat. Immunol. 3, 999–1005.

46. Rudensky A.Y., Gavin M., Zheng Y. 2006. FOXP3 andNFAT: Partners in tolerance. Cell. 126, 253–256.

47. Harrington L.E., Mangan P.R., Weaver C.T. 2006.Expanding the effector CD4 T-cell repertoire: The Th17lineage. Curr. Opin. Immunol. 18, 349–356.

48. Weaver C.T., Harrington L.E., Mangan P.R., Gavrieli M.,Murphy K.M. 2006. Th17: An effector CD4 T cell lin-eage with regulatory T cell ties. Immunity. 24, 677–688.

49. Dudley M.E., Wunderlich J.R., Yang J.C., Sherry R.M.,Topalian S.L., Restifo N.P., Royal R.E., Kammula U.,White D.E., Mavroukakis S.A., Rogers L.J., Gracia G.J.,Jones S.A., Mangiameli D.P., Pelletier M.M., Gea-Bana-cloche J., Robinson M.R., Berman D.M., Filie A.C.,Abati A., Rosenberg S.A. 2005. Adoptive cell transfertherapy following non-myeloablative but lymphode-pleting chemotherapy for the treatment of patients withrefractory metastatic melanoma. J. Clin. Oncol. 23,2346–2357.

50. Dranoff G. 2004. Cytokines in cancer pathogenesis andcancer therapy. Nat. Rev. Cancer. 4, 11–22.

51. Van Horssen R., Ten Hagen T.L., Eggermont A.M.2006. TNF-α in cancer treatment: Molecular insights,antitumor effects, and clinical utility. Oncologist. 11,397–408.

52. Grunhagen D.J., de Wilt J.H., Ten Hagen T.L., Egger-mont A.M. 2006. Technology insight: Utility of TNF-α-based isolated limb perfusion to avoid amputation ofirresectable tumors of the extremities. Nat. Clin. Pract.Oncol. 3, 94–103.

53. Waldmann T.A. 2006. The biology of interleukin-2 andinterleukin-15: implications for cancer therapy and vac-cine design. Nat. Rev. Immunol. 6, 595–601.

54. Harper P., Littlewood T. 2005. Anaemia of cancer:Impact on patient fatigue and long-term outcome.Oncology. 69 Suppl. 2, 2–7.

55. Roden R., Wu T.C. 2006. How will HPV vaccines affectcervical cancer? Nat. Rev. Cancer. 6, 753–763.

56. Kirkwood J.M., Moschos S., Wang W. 2006. Strategiesfor the development of more effective adjuvant therapyof melanoma: Current and future explorations of anti-bodies, cytokines, vaccines, and combinations. Clin.Cancer Res. 12, 2331s–2336s.

57. Kershaw M.H., Teng M.W., Smyth M.J., Darcy P.K.2005. Supernatural T cells: genetic modification of Tcells for cancer therapy. Nat. Rev. Immunol. 5, 928–940.

58. Nestle F.O., Farkas A., Conrad C. 2005. Dendritic-cell-based therapeutic vaccination against cancer. Curr.Opin. Immunol. 17, 163–169.

59. Pastan I., Hassan R., FitzGerald D.J., Kreitman R.J.2006. Immunotoxin therapy of cancer. Nat. Rev. Can-cer. 6, 559–565.

60. Richman S.P., Gutterman J.U., Hersh E.M. 1979. Can-cer immunotherapy. Can. Med. Assoc. J. 120, 322–329.

61. Plotkin S.A. 2005. Vaccines: past, present and future.Nat. Med. 11, S5–S11.

62. Cain J.M., Howett M.K. 2000. Preventing cervical can-cer. Science. 288, 1753–1755.

63. Heathcote E.J. 2004. Prevention of hepatitis C virus-related hepatocellular carcinoma. Gastroenterology.127, S294–S302.

64. Lyons S.F., Liebowitz D.N. 1998. The roles of humanviruses in the pathogenesis of lymphoma. Semin.Oncol. 25, 461–475.

65. Zou W. 2005. Immunosuppressive networks in thetumour environment and their therapeutic relevance.Nat. Rev. Cancer. 5, 263–274.

66. Dranoff G. 2002. GM-CSF-based cancer vaccines.Immunol Rev. 188, 147–154.

67. Larin S.S., Korobko E.V., Kustikova O.S., Borodu-lina O.R., Raikhlin N.T., Brisgalov I.P., Georgiev G.P.,Kiselev S.L. 2004. Immunotherapy with autologoustumor cells engineered to secrete Tag7/PGRP, an innateimmunity recognition molecule. J. Gene Med. 6, 798–808.

68. Pittet M.J., Speiser D.E., Valmori D., Rimoldi D.,Lienard D., Lejeune F., Cerottini J.C., Romero P. 2001.Ex vivo analysis of tumor antigen specific CD8+ T cellresponses using MHC/peptide tetramers in cancerpatients. Int. Immunopharmacol. 1, 1235–1247.

69. Rosenberg S. 1985. Lymphokine-activated killer cells:A new approach to immunotherapy of cancer. J. Natl.Cancer Inst. 75, 595–603.

70. Schumacher T.N. 2002. T-cell-receptor gene therapy.Nat. Rev. Immunol. 2, 512–519.

MOLECULAR BIOLOGY Vol. 41 No. 2 2007

PROBLEMS OF CANCER IMMUNOTHERAPY 327

71. Gattinoni L., Powell D.J., Jr., Rosenberg S.A., Restifo N.P.2006. Adoptive immunotherapy for cancer: Building onsuccess. Nat. Rev. Immunol. 6, 383–393.

72. Cragg M.S., French R.R., Glennie M.J. 1999. Signalingantibodies in cancer therapy. Curr. Opin. Immunol. 11,541–547.

73. Weiner L.M. 1999. An overview of monoclonal anti-body therapy of cancer. Semin. Oncol. 26, 41–50.

74. Mach J.P., Pelegrin A., Buchegger F. 1991. Imaging andtherapy with monoclonal antibodies in non-hematopoi-etic tumors. Curr. Opin. Immunol. 3, 685–693.

75. Ballou B., Ernst L.A., Waggoner A.S. 2005. Fluores-cence imaging of tumors in vivo. Curr. Med. Chem. 12,795–805.

76. Glinka E.M., Edelweiss E.F., Deyev S.M. 2006.Eukaryotic expression vectors and immunoconjugatesfor cancer therapy. Biokhimiya. 71, 597–606.

77. Buchegger F., Antonescu C., Delaloye A.B., Helg C.,Kovacsovics T., Kosinski M., Mach J.P., Ketterer N.2006. Long-term complete responses after 131I-tositu-momab therapy for relapsed or refractory indolent non-Hodgkin’s lymphoma. Br. J. Cancer. 94, 1770–1776.

78. FitzGerald D.J., Kreitman R., Wilson W., Squires D.,Pastan I. 2004. Recombinant immunotoxins for treatingcancer. Int. J. Med. Microbiol 293, 577–582.

79. Moingeon P. 2004. Recombinant cancer vaccines basedon viral vectors. Dev. Biol. (Basel). 116, 117–122.

80. Jager D., Jager E., Knuth A. 2001. Immune responses totumor antigens: Implications for antigen specific immu-notherapy of cancer. J. Clin. Pathol. 54, 669–674.

81. Jager E., Jager D., Knuth A. 2003. Antigen-specificimmunotherapy and cancer vaccines. Int. J. Cancer.106, 817–820.

82. Jager E., Jager D., Knuth A. 2002. Clinical cancer vac-cine trials. Curr. Opin. Immunol. 14, 178–182.

83. zur Hausen H. 1999. Viruses in human cancers. Eur. J.Cancer. 35, 1878–1885.

84. Klein G. 1994. Epstein–Barr virus strategy in normaland neoplastic B cells. Cell. 77, 791–793.

85. Willimsky G., Blankenstein T. 2005. Sporadic immuno-genic tumors avoid destruction by inducing T-cell toler-ance. Nature. 437, 141–146.

86. Hayday A., Tigelaar R. 2003. Immunoregulation in thetissues by γδ T cells. Nat. Rev. Immunol. 3, 233–242.

87. Ochsenbein A.F., Sierro S., Odermatt B., Pericin M.,Karrer U., Hermans J., Hemmi S., Hengartner H., Zin-kernagel R.M. 2001. Roles of tumor localization, sec-ond signals and cross priming in cytotoxic T-cell induc-tion. Nature. 411, 1058–1064.

88. Abelev G.I. 1999. Immunology of cancer. Vestn. Ross.Akad. Med. Nauk. 21–25.

89. Rosenberg S.A. 1999. A new era for cancer immuno-therapy based on the genes that encode cancer antigens.Immunity. 10, 281–287.

90. Ikeda H., Ohta N., Furukawa K., Miyazaki H., Wang L.,Kuribayashi K., Old L.J., Shiku H. 1997. Mutated mito-gen-activated protein kinase: A tumor rejection antigenof mouse sarcoma. Proc. Natl. Acad. Sci. USA. 94,6375–6379.

91. Klein G. 1997. Rejection antigens in chemicallyinduced tumors. Proc. Natl. Acad. Sci. USA. 94, 5991–5992.

92. Line A., Stengrevics A., Slucka Z., Li G., Jankevics E.,Rees R.C. 2002. Serological identification and expres-sion analysis of gastric cancer-associated genes. Br. J.Cancer. 86, 1824–1830.

93. Brass N., Heckel D., Sahin U., Pfreundschuh M.,Sybrecht G.W., Meese E. 1997. Translation initiationfactor eIF-4gamma is encoded by an amplified gene andinduces an immune response in squamous cell lung car-cinoma. Hum. Mol. Genet. 6, 33–39.

94. Lagarkova M.A., Boitchenko V.E., Mescheryakov A.A.,Kashkarova U.A., Nedospasov S.A. 2000. Human cort-actin as putative cancer antigen. Oncogene. 19, 5204–5207.

95. Vonderheide R.H., Hahn W.C., Schultze J.L., Nadler L.M.1999. The telomerase catalytic subunit is a widelyexpressed tumor-associated antigen recognized by cyto-toxic T lymphocytes. Immunity. 10, 673–679.

96. Boon T., Cerottini J.C., van den Eynde B.J., van derBruggen P., van Pel A. 1994. Tumor antigens recog-nized by T lymphocytes. Annu. Rev. Immunol. 12, 337–365.

97. Kawakami Y., Nishimura M.I., Restifo N.P., Topalian S.L.,O’Neil B.H., Shilyansky J., Yannelli J.R., Rosenberg S.A.1993. T-cell recognition of human melanoma antigens.J. Immunother. 14, 88–93.

98. Chen W., Peace D.J., Rovira D.K., You S.G., Cheever M.A.1992. T-cell immunity to the joining region ofp210BCR-ABL protein. Proc. Natl. Acad. Sci. USA. 89,1468–1472.

99. Zhang J.Y., Casiano C.A., Peng X.X., Koziol J.A.,Chan E.K., Tan E.M. 2003. Enhancement of antibodydetection in cancer using panel of recombinant tumor-associated antigens. Cancer Epidemiol. BiomarkersPrev. 12, 136–143.

100. Casiano C.A., Mediavilla-Varela M., Tan E.M. 2006.Tumor-associated antigen arrays for the serologicaldiagnosis of cancer. Mol. Cell. Proteomics. 5, 1745–1759.

101. Hodge J.W., Greiner J.W., Tsang K.Y., Sabzevari H.,Kudo-Saito C., Grosenbach D.W., Gulley J.L., Arlen P.M.,Marshall J.L., Panicali D., Schlom J. 2006. Costimula-tory molecules as adjuvants for immunotherapy. Front.Biosci. 11, 788–803.

102. Maclean J., Rybicki E.P., Williamson A.L. 2005. Vacci-nation strategies for the prevention of cervical cancer.Expert. Rev. Anticancer Ther. 5, 97–107.

103. Duenas-Carrera S. 2004. DNA vaccination against hep-atitis C. Curr. Opin. Mol. Ther. 6, 146–150.

104. Wysocki P.J., Karczewska-Dzionk A., Mackiewicz-Wysocka M., Mackiewicz A. 2004. Human cancer genetherapy with cytokine gene-modified cells. Expert.Opin. Biol. Ther. 4, 1595–1607.

105. Turtle C.J., Hart D.N. 2004. Dendritic cells in tumorimmunology and immunotherapy. Curr. Drug Targets.5, 17–39.

106. Soruri A., Zwirner J. 2005. Dendritic cells: Limitedpotential in immunotherapy. Int. J. Biochem. Cell Biol.37, 241–245.

328

MOLECULAR BIOLOGY Vol. 41 No. 2 2007

NEDOSPASOV and KUPRASH

107. Dudley M.E., Rosenberg S.A. 2003. Adoptive-cell-transfer therapy for the treatment of patients with can-cer. Nat. Rev. Cancer. 3, 666–675.

108. Ho W.Y., Blattman J.N., Dossett M.L., Yee C., Green-berg P.D. 2003. Adoptive immunotherapy: EngineeringT cell responses as biologic weapons for tumor massdestruction. Cancer Cell. 3, 431–437.

109. Morgan R.A., Dudley M.E., Wunderlich J.R.,Hughes M.S., Yang J.C., Sherry R.M., Royal R.E.,Topalian S.L., Kammula U.S., Restifo N.P., Zheng Z.,Nahvi A., de Vries C.R., Rogers-Freezer L.J.,Mavroukakis S.A., Rosenberg S.A. 2006. Cancer

Regression in patients after transfer of genetically engi-neered lymphocytes. Science. 314, 126–129.

110. Baecher-Allan C., Anderson D.E. 2006. Immune regu-lation in tumor-bearing hosts. Curr. Opin. Immunol. 18,214–219.

111. Binder R.J. 2006. Heat shock protein vaccines: Frombench to bedside. Int. Rev. Immunol. 18, 214–219.

112. Maki R.G., Old L.J., Srivastava P.K. 1990. Humanhomologue of murine tumor rejection antigen gp96:5'-regulatory and coding regions and relationship tostress-induced proteins. Proc. Natl. Acad. Sci. USA. 87,5658–5662.