You are on page 1of 17

Available online at www.sciencedirect.

com

Cytokine & Growth Factor Reviews 20 (2009) 97–113


www.elsevier.com/locate/cytogfr

Survey

Interferon-g: A historical perspective


Alfons Billiau *, Patrick Matthys
Rega Institute, University of Leuven, Minderbroedersstraat 10, 3000 Leuven, Belgium

Available online 5 March 2009

Abstract

This article reviews the main lines of thinking and exploration that have led to our current conception of the role of IFN-g in immune defense
and autoimmunity. In 1965 the first report appeared describing production of an interferon-like virus inhibitor in cultured human leukocytes
following exposure to the mitogen phytohemagglutinin. In the early 1970s the active principle became recognized as being distinct from classical
virus-induced interferons, leading to its designation as immune interferon or Type II interferon, and eventually IFN-g. Up to that point interest in
the factor had come almost exclusively from virologists, in particular those among them who were believers in interferon. Evidence first coming
forward in the 1980s that IFN-g is indistinguishable from macrophage-activating factor (MAF), then a prototype lymphokine, was the signal for
immunologists at large to become interested. Today IFN-g ranks among the most important endogenous regulators of immune responses.
# 2009 Published by Elsevier Ltd.

Keywords: Interferon-g; History

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2. From the discovery to the cloning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3. The belated recognition of IFN-g as a lymphokine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
4. Macrophage activation by IFN-g and resulting anti-microbial activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5. IFN-g and the helper effects of T cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
6. IFN-g produced by CD8+ cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
7. IFN-g and NK cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
8. IFN-g, dendritic cells (DC) and antigen presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
9. IFN-g and the chemokines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
10. IFN-g in vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
10.1. Defence against infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
10.2. IFN-g and hypersensitivity reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
10.2.1. Delayed-type hypersensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
10.2.2. Hypersensitivity reactions elicited by endotoxin—the Shwartzman-type reactions . . . . . . . . . . . . . . . . . . . 106
10.2.3. IFN-g and autoimmune disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
11. Concluding thoughts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

1. Introduction
* Corresponding author. Tel.: +32 16 337341 (office)/622008 (home);
fax: +32 16 337340 (office)/503933 (home). If IFN-g were to be discovered today, the chances are
E-mail address: Alfons.Billiau@Rega.Kuleuven.be (A. Billiau). slim that it would be named an interferon. Its molecular

1359-6101/$ – see front matter # 2009 Published by Elsevier Ltd.


doi:10.1016/j.cytogfr.2009.02.004
98 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

structure is only superficially similar to that of other would be named interleukin-n, n representing a two-digit
interferons and its actions on cells are mediated by a receptor number. That IFN-g was identified as an antiviral factor long
different from that used by other interferons. Moreover, the before it was recognized as an interleukin-type factor is
direct antiviral effect on cells would probably be deemed mainly due to the simplicity and extreme sensitivity of the
insufficiently important, relative to the regulatory actions on antiviral assay which preceded the more complex in vitro
leukocytes, to justify creation of a second interferon assays by which other cytokines were discovered. However,
subfamily. Most probably the newly discovered cytokine the early discovery as an antiviral factor made that many

Table 1
Chronology of IFN-g research.
IFN-g—general issues
1965 Induction of an IFN activity in human PBMC by phytohemagglutinin [12]
1966 Antigen-specific induction of IFN activity during virus infections [20]
1972 IFN induced by anti-lymphocyte serum differs physicochemically from classical IFN ! named ‘immune IFN’ [16]
1973 IFN produced during DTH reactions differs physicochemically from classical IFN ! named ‘Type II IFN’ [22]
1979 Progress in purification [28–31]
1980 Nomenclature Committee: definitive name ‘IFN-g’ [27]
1982 Dimeric structure of IFN-g suggested [31]
Cloning of IFN-g from cDNA [32,33]
IFN-g vs. DTH and chemokines
1968 Migration inhibitory factor (MIF) and skin-reactive factor (SRF) described [38]
Lymphotoxin (LT) described [40,197]
1969 Lymphocyte mitogenic factor(s) and the lymphokine concept [35]
1973 MIF co-induced with Type II IFN during DTH in mice [22]
1981 Chemotactic activity of PF4 (now CXCL4) recognized [91]
1985 g-IP10 (now CXCL10) described [89]
1992 IFN-g inhibits neutrophil-targeting chemokines, and . . . [98–104]
1993 IFN-g stimulates mononuclear cell-targeting chemokines [99,102,103,105–109]
IFN-g, the macrophage-activating factor (MAF)
1969 Lymphocyte-mediated activation of macrophages – MIF responsible? [43]
1979 Macrophage-activating factor (MAF) ! several MAF assays established [44]
1981 Mr of MAF = 45.000 [50]
1983 Anti-IFN-g antibody neutralizes MAF preparations [45]
1985 Cloned IFN-g possesses MAF activity [53]
IFN-g and the evolving T helper cell concept
1973 T cell-replacing factor (TRF, T cell help for B cells) described [58]
1977 Type II IFN inhibits antibody production in vitro [198]
1984 IFN-g proposed as a TRF [59,61–63]
1988 Th1 and Th2 clones described [64]
Role of IFN-g in Th1/Th2 paradigm [65]
2005 Th17 cell lineage defined in mice [68,181]
IFN-g inhibits differentiation of Th17 cells and IL-17 production by activated Th memory cells
IFN-g and natural killer (NK) cells
1980 IFN-g upregulates NK activity [77]
1983 IL-2 may induce IFN-g in NK cells [78]
1991 IL-12 induces IFN-g in NK cells [79]
1993 Novel IFN-g-coinducing factor ! IGIF/IL-18 [80]
1995 IL-18 induces IFN-g in NK cells [81]
IFN-g vs. antigen-presenting cells (APC) and dendritic cells (DC)
1982 IFN-g enhances MHC Class II expression in mononuclear phagocytes [46]
1984 IFN-g induces enzymatic breakdown of tryptophan [199]
1990 IFN-g induces IDO in vivo [200]
1996 IFN-g enhances MHC Class II expression in DCs [83]
2000 IFN-g optimizes IL-12 production by DCs [85]
2000 IFN-g-induced IDO conditions DCs to become tolerogenic [188–190]
IFN-g vs. CD4+CD25+ Treg cells
1995 CD4+CD25+ Treg cells defined [201]
2005 Defective functioning of Treg cells in IFN-gR KO mice with collagen-induced arthritis (CIA) [187]
2006 IFN-g can convert CD4+CD25 cells into CD4+ Treg cells able to suppress experimental autoimmune [185]
encephalomyelitis (EAE)
Release of IFN-g by Treg cells [186]
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 99

aspects of the biology and biochemistry of IFN-g were the list would still grow to include monoclonal anti-CD3
explored in close connection with those of the ‘real’ antibody, additional bacterial proteins (e.g. staphylococcal
interferons, IFN-a and -b. These common aspects have been enterotoxins and staphylococcal protein A) later to be grouped
touched upon in Parts I and II of this historical review [1,2]. with others as the super-antigens, and certain enzymes (e.g.
Here I will discuss facets that deal specifically with the galactose oxidase) (for review see Refs. [17–19]). Induction
history of IFN-g research. Milestones marking this history of IFN-g by these stimulants would appear to be due to
are assembled chronologically in Table 1. binding and cross-linking of receptor-like glycoproteins on
Over the years the status of our knowledge on IFN-g has the cell membranes of lymphocytes resulting in polyclonal
often been reviewed. The reader may be interested in having activation.
the references to some of these reviews [3–11]. Polyclonal activation of lymphocytes was not yet a
current concept at the end of the 1960s, and the analogy
between stimulation by these mitogens and that by antigens
2. From the discovery to the cloning was not yet generally recognized. In fact, antigen-specific
induction of interferon was discovered quite independently.
In 1965, 8 years after the discovery of interferon induction In 1966, Glasgow [20] hypothesized that interferon
by viruses, Wheelock [12] reported on the appearance of production might take place and play an important role
interferon-like antiviral activity in supernatant fluid of not only during primary infection by viruses, but also
cultures of fresh human leukocytes following incubation following re-infection. To make his point, he immunized
with the plant lectin phytohemagglutinin (PHA). At the time, mice against Chikungunya virus and tested interferon
like today, PHA was much in use by cytogeneticists to induce production by their cultured peritoneal leukocytes following
blastogenesis and mitosis in human peripheral lymphocytes, in vitro challenge with the same virus. Leukocytes from
so as to enable chromosomes to be visualized. Like virus- immunized mice produced up to four times as much
induced interferon, the PHA-induced inhibitor was macro- interferon-like antiviral activity as cells from naı̈ve mice.
molecular (non-dialysable) but soluble (non-sedimentable) The interferon activity was stated to be acid-resistant,
and acted on human cells only. However, its activity was less though a parallel comparison with virus-induced interferon
resistant to heat and to acid, two criteria that, besides others, was not undertaken. Therefore, in retrospect, it cannot be
would remain in use up to the present day to differentiate IFN- ascertained if and to what extent IFN-g accounted for this
g from other interferons. observation. However, in discussing the data, the author
Wheelock’s observation came at a time when reports on argued that interferon production by immune-sensitized
induction of interferon-like activity by various natural leukocytes following re-exposure to the viral antigen could
products were legion. Most notable were the reports that play an important role in ‘cellular or tissue immunity’ (as
intravenous injection of endotoxin in chicks, mice or rabbits opposed to antibody-mediated immunity). Of note, at the
induced appearance of interferon-like activity in the blood time, the concept of cellular immunity against virus
stream. The possibility was considered that the PHA-induced infection, as we know it today, was poorly developed.
human interferon was the homologue of these animal Few years later (1969), these findings in mice and mouse
interferons induced in vivo by endotoxin. Endotoxin-induced cells exposed to live virus were extended so as to apply to
interferon in rabbits [13] or in cultured rabbit macrophages humans and human leukocytes following sensitization with
[14] had indeed been found to be relatively acid- and heat- non-viral antigens: Green et al. [21] demonstrated interferon
sensitive. Moreover, endotoxins, like PHA, had been shown to production by blood lymphocytes from humans sensitized to
possess mitogenic potential for certain lymphocytes. Simila- tetanus toxoid, diphtheria toxoid or tubercular PPD when such
rities in molecular weight were also invoked, though the leukocytes were exposed in vitro to the respective antigens. In
values estimated at the time were at best tentative. As we know this study the amounts of interferon and the size of the
today, circulating interferon induced by endotoxin in mice is a blastogenic response were compared with those in blood
mixture of Type I and Type II interferons, the former ones leukocytes exposed to PHA, and both were found to be only a
produced by myeloid cells [15] and the latter one by NK cells fraction of the PHA-induced response. The possibility that
and some populations of T lymphocytes. interferon induced by these agents might be different from
Another question debated at this early time was whether classical virus-induced interferon was not considered. In
interferon production following induction by PHA was linked particular, acid-lability, though explicitly mentioned by
to the blastogenic response. PHA preparations were hetero- Wheelock in 1965 as a peculiarity, was not tested for.
geneous and the possibility was considered that the two Reviewing the literature on ‘non-viral inducers of interferon’
effects were triggered by different components. However, as in 1972–1973, Merigan [17] did discuss the biological
would soon become evident, other blastogenic and/or mitotic significance of induction by mitogens and antigens, including
agents can also induce the acid-labile interferon. The early list Glasgow’s hypothesis. He also cited unpublished data
of such agents included other plant lectins (concanavalin A, indicating that ‘The molecular weight of interferon induced
pokeweed mitogen) but also the bacterial toxin streptolysin-O by PHA in white blood cell cultures is 18,000, the same as that
and antilymphocyte serum (ALS) [16]. In the 1970s and 1980s induced by Newcastle disease virus in the same cells.’
100 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

Thus, by the end of the 1960s the concept of ‘immune evidence for the existence of homologous interferon
induction’ of interferon was rather well established. The molecular species in mice. Thus, the need developed for an
possibility that such interferon might reside in different adequate nomenclature for these proteins. In the spring of
molecular forms was entertained, but available evidence was 1980 a Committee of experts [27] agreed on proposing the
considered to weak. Then, two milestone studies appeared in terms IFN-a and IFN-b for the ‘classical interferons’ both of
1972 and 1973, confirming that immune-induced interferon which had been obtained in pure form and appeared to be
differed physicochemically and biologically from ‘classical’ homogenous. Although the immune-induced interferon had
virus-induced interferon. Rebecca Falcoff characterized not yet been obtained in pure form, the Committee recognized
interferon induced by anti-lymphocyte immunoglobulin in that its main component differed from IFN-a and IFN-b, and
cultured peripheral human lymphocytes and showed it to be gave it the name IFN-g, thus settling the debate on the choice
acid-labile and to migrate differently from virus-induced between ‘Type II interferon’ and ‘Immune interferon’.
interferon on gel filtration and on DEAE-cellulose Ironically, the terms Type I and Type II interferon have
chromatography [16]. In fact, her estimate of a molecular survived up to the present day, the only reason being that
mass of 50,000, as opposed to 25,000 for virus-induced ‘Type I interferon’ conveniently allows to collectively
interferon, was as close to reality as was possible to estimate designate the biologically related IFN-a and IFN-b.
with the available reagents of the time. On the basis of this Difficulties encountered in purifying IFN-g to homo-
evidence she concluded: ‘. . . synthesis of interferon by geneity included low yields in production systems and
immunocompetent cells under sets of circumstances which irreversible loss of biological activity during purification.
certainly represent true immune responses, suggests that this Between 1979 and 1982 these difficulties were overcome in
interferon can be called immune interferon’. In another several laboratories [28–31], such that in 1982 it had become
milestone study, Youngner and Salvin [22] showed that clear that native human IFN-g had a Mr of 45 kDa, being
BCG-infected mice inoculated intravenously with old composed of glycopeptides of 20 to 25 kDa. In the same
tuberculin (OT) release a viral inhibitor into the circulation year, two laboratories [32,33] succeeded in cloning the
(up to 8000 units/ml). This inhibitor also had all basic cDNA of human IFN-g. Cloning and analysis of the genomic
properties of interferon, but differed from virus-induced DNA revealed that, contrary to the intronless genes of IFN-a
mouse interferon by being less stable at pH 2 and at 56 8C, and IFN-b, the IFN-g gene contained three introns [34].
by a different host range specificity and by resisting
neutralization by an antiserum prepared against highly
purified virus-induce mouse interferon. On the basis of this 3. The belated recognition of IFN-g as a lymphokine
evidence, the authors concluded: ‘. . . the interferon elicited
by OT in BCG-infected or control mice is . . . a distinct In 1969 Dumonde et al. [35] proposed the term
molecular species of viral inhibitor and is designated ‘‘Type ‘lymphokines’ to designate ‘non-antibody mediators of
II interferon’’.’ They objected against the term immune cellular immunity generated by lymphocyte activation’. This
interferon arguing that ‘. . .definition of interferon based on readily accepted proposal, was based on four biological
the cell of origin or on the stimulus does not take into activities of lymphocyte/macrophage supernatants, and
account the characterization of the inhibitor molecules and marked a turning point in the years-long efforts of many
the mechanism of their induction’. This debate continued for research groups to find an in vitro correlate for cellular
several years. In their comprehensive review of 1975 Ho and immunity. At the time, the preferred experimental animal
Armstrong [18] refrained from using the terms Type I and model for studying cellular immunity was delayed-type
Type II to designate molecular entities or even families of hypersensitivity (DTH) in the guinea pig. Typically, dermal
molecules, as they rather adhered to the terms Type I and reactions were elicited in animals infected with BCG or
Type II induction. Typically, whereas Youngner and Salvin immunized with Freund’s adjuvant, and challenged intra-
[22] considered endotoxin-induced interferon as Type I, Ho dermally with PPD. It had become evident that not only
and Armstrong, in their review [18] catalogued it under lymphocytes but also macrophages play an important role in
‘Type II induction’. Also, in his exhaustive monography this local skin reaction (reviewed in Ref. [36]). Accordingly,
published in 1979, W.S. Stewart II [19] (pages 145–149) in their attempts to develop an in vitro test, researchers
contested the adequacy of the terms Type I and Type II focused on these cells. Peritoneal exsudate of guinea pigs
interferons. Nevertheless, the later 1970s saw the appear- contains ca. 75% lymphocytes and 15% macrophages. So, it
ance of more reports of interferon induction in leukocytes by was logical to culture peritoneal exsudate cells of sensitized
mitogens or antigens all confirming the observations of the animals and to observe their reaction towards exposure to the
two aforementioned studies (reviewed by Stewart [19] and antigen. An ‘avant la lettre’ observation of 1932 inspired the
by Epstein [3]). Furthermore, in these years, purification and researchers to use cell migration as read-out: Rich and Lewis
characterization of the classical mouse and human inter- had discovered that ‘migration of cells from splenic explants
ferons took giant leaps forward. In the course of 1979 the from hypersensitive guinea pigs was specifically inhibited by
amino-terminal sequences of pure human leukocyte- and antigen’ (loc. cit. [37]). This led to the development of the
fibroblast interferons were revealed [23–26], and there was classical macrophage migration inhibition assay in which
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 101

macrophages, with or without lymphocytes, are cultured in molecular structure of which has been identified’. An
open-ended capillary tubes lying on a horizontal glass important factor in settling the debate was the definition,
surface. Without exposure to antigen, the macrophages around 1980, of one additional lymphokine activity, i.e.
emigrate from the tubes and form monolayers on the glass. macrophage-activating factor (MAF).
In the presence of sensitized lymphocytes and antigen, this
migration is inhibited. Working with this assay, Bloom and
Bennett [37] established that ‘the sensitized lymphocytes, 4. Macrophage activation by IFN-g and resulting
upon interaction with the antigen in vitro, elaborate into the anti-microbial activity
medium a soluble material capable of inhibiting migration
of normal exudate cells’. In a further study, the same authors Today it is common knowledge that, during host defense
[38] called the soluble material ‘migration inhibitory factor’ against microorganisms, mononuclear phagocytes become
(MIF) and showed that, after partial purification, this activated by lymphokines and thereby acquire the capability
proteinaceous material was also able to elicit a skin reaction to efficiently kill intracellular pathogens and to attract
upon intradermal injection, later leading to the designation additional cellular immune components into the inflamma-
of skin-reactive factor (SRF). In hindsight, it seems plausible tory focus. This concept of macrophage activation owes
that the activities of these ‘factors’ probably corresponded to much to a series of studies conducted by Mackaness and
a mixture of several cytokines and chemokines. In 1967, collaborators in the course of the 1960s. One particular study
exposure of lymphocytes from sensitized guinea pigs to an from 1969 [43] deserves special attention. It addressed the
immunizing antigen (e.g. a hapten-carrier complex) had observation that BCG-infected mice mount a degree of a-
been shown to also result in blast formation [39], and in his specific resistance against Listeria. It was shown that
1969 study Dumonde et al. [35], using peritoneal exudate lymphocytes of such mice can transfer this resistance to
cells of BCG-sensitized guinea pigs, demonstrated that this naı̈ve recipients provided they are injected together with an
mitogenic activity is associated to the appearance of (a) eliciting dose of BCG. Furthermore, it was demonstrated
soluble blastogenesis-inducing factor(s) in the stimulated that ‘The peritoneal macrophages of animals so treated
cultures. Thus, lymphocyte growth factors became seen as a develop the morphology and microbicidal features of
third type of lymphokine. activated macrophages’. From these observations, the
A fourth soluble biological activity included by authors ‘. . . inferred that acquired resistance depends upon
Dumonde et al. [35] in their definition of lymphokines the activation of host macrophages through a product
was cytotoxicity. That lymphocytes could under certain resulting from specific interaction between sensitized
circumstances exert toxic effects on other cells had been lymphoid cells and the organism or its antigenic products’.
noted in the earlier 1960s. Release of a soluble cytotoxic Considering various possible mechanisms for the interaction
factor (lymphotoxin) by sensitized lymphocytes following between lymphoid cells and macrophages, the author also
exposure to antigen and the link of this phenomenon to DTH speculated that MIF, the mediator described by Bloom and
were established in 1968 [40,41]. Bennett [37], ‘could fulfill the role of the hypothetical
Thus, by 1970, the ‘lymphokine’ concept was fairly well humoral mediator which influences macrophage function’.
documented. However, neither the interferonologists nor the In the years that followed, in vitro evidence for this concept
lymphokinologists considered interferon as a lymphokine. was collected. In particular, various tests for macrophage
The first ones, having their minds set on defence against activation were developed. Also, from 1979 onwards, the term
virus infections, were still uncertain about the nature of the ‘macrophage-activating factor’ (MAF) first appears in the
peculiar interferons induced by such non-viral substances as literature, referring to the ability of supernatants of mitogen-
endotoxin and PHA. The second ones, being concerned with or antigen-challenged mononuclear cell cultures to augment
DTH as a marker of cellular immunity against mycobacteria various biological activities of macrophages. The MAF
had little interest in anti-virus defence but were more alert bioassays variably relied on intracellular killing of ingested
for the role of DTH in rejection of tumors and tissue microorganisms, increased oxidative metabolism [44,45],
transplants. Moreover, interferon workers and DTH workers enhanced expression of Class II antigens [46] or enhanced
formed separate scientific communities, with only few tumor cell killing [47,48]. In vitro studies in the early 1980s
occasions of mutual interaction. demonstrated that exposure of macrophages to lymphokine
Yet, in the debate on the nature of Type II and immune preparations results in enhancement of their potential to kill
interferon, Salvin et al. in 1973 [42] came close to the point ingested various microorganisms [49–51]. Cultured mouse
of considering immune-induced interferon as a lymphokine peritoneal cells, which supported proliferation of Leishmania
as they showed the appearance of both MIF and Type II tropica, were shown to kill the intraphagolysosomal organ-
interferon in the circulation of mice during DTH reactions. isms following exposure to crude lymphokine [50]. Fractio-
Only towards the end of the 1970s would reviewers, for nation of the lymphokine by gel filtration indicated that a
instance Epstein [3], consider IFN-g as a bona fide 45 kDa protein component accounted for this macrophage-
‘lymphokine’. In his 1984 review on IFN-g, Kirchner [4] activating effect. The fraction also produced a cytocidal effect
remarked: ‘. . .(IFN-g) was the first lymphokine the for tumor cells and for extracellular Schistosomula, whereas a
102 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

smaller fraction (23 kDa) provided only activity against the IgE and IgG1. Remarkably, this selective induction was
extracellular targets [52]. The authors’ presumption that the counteracted by addition of IFN-g [63], suggesting hat IFN-
45 kDa fraction represented IFN-g was confirmed by a later g played a role in directing isotype switch. Follow-up of
study employing cloned IFN-g [53]. these findings soon led to the classification of murine helper
Meanwhile, availability of pure IFN-g and specific T cell clones into T-helper-1 (Th1) and -2 (Th2) categories
antibodies made it possible to demonstrate that IFN-g by [64] and to the division of in vivo immune responses in Th1
itself could augment bactericidal activity of macrophages and Th2 responses. Both were characterized by particular
against Toxoplasma in mouse macrophages [45] and cytokine production profiles, whereby production of IFN-g
Chlamydia in human monocyte-derived macrophages is the Th1 hallmark by excellence. Thus, it appeared that
[54]. Furthermore, characterization with monoclonal anti- IFN-g indeed qualifies as an important component of T cell
IFN-g antibodies revealed that IFN-g is an essential helper cytokines, as it critically co-directs the isotype
component of MAF in crude lymphokine preparations switch. Evidently the question also rose as to whether IFN-g
[45]. In some assay systems for MAF, e.g. hydrogen would play a role in the proliferation and differentiation of T
peroxide production elicited by phorbol ester, IFN-g is helper cells. As far as murine T cell clones are concerned,
sufficient for activation to take place, but in others, IFN-g was found to inhibit proliferation of Th2 but not Th1
additional cytokines such as TNF-a are required. This clones [65]. Follow up of this observation would later lead to
seems to be particularly so in the case of bactericidal effects the finding that the b-chain of the IFN-g receptor is not
against Mycobacteria [53] or Listeria [55–57]. expressed in Th1 clones [66].
From the foregoing it should be evident that studies
focusing on IFN-g have been of paramount importance in
5. IFN-g and the helper effects of T cells establishing the Th1/Th2 paradigm, which holds that IFN-g
(1) favours development of Th1 over Th2 cells, (2) provides
That antibody responses require help from T cells was well help from Th1 cells to macrophages (activation) and (3)
known already in the 1960s. In the early 1970s work in several provides help from Th1 cells to B cells (isotype switch).
laboratories indicated that antigen-nonspecific soluble factors Over almost two decades the Th1/Th2 concept has inspired
secreted by the T cells could at least in part account for this numerous studies clarifying fundamental aspects of both
help. The term ‘T cell-replacing factor’ (TRF) to designate the normal and pathological immune responses. However, the
ensemble of these factors came in use around 1973 [58]. In the Th1/Th2 concept failed to accommodate several observa-
years that followed, assay systems for TRF were multiplied tions, in particular in connection with the role of IFN-g in
and refined, allowing to further underpin the distinction experimental models of autoimmune disease.
between different stages in B cell biology and to define the The discovery, in the early 2000s, of an additional T cell
corresponding factors present in TRF. At first, the source of lineage (Th17) characterized by production of IL-17 (for
TRF consisted of freshly harvested normal thymocytes or review, see Ref. [67]) finally paved the way for explaining
splenocytes stimulated with mitogens. Later on, thanks to the these long-standing unsolved questions. Most importantly,
discovery of the T cell growth factor (TCGF) that eventually IFN-g produced by Th1 cells was shown to counteract
became IL-2, cloned T cell lines capable of producing TRFs development of Th17 cells [68]. Differentiation of naı̈ve T
became available, and these facilitated purification and cells towards IL-17-producing cells was greatly enhanced
characterization work. Though interferonologists were using when the medium was supplemented with neutralizing anti-
quite similar settings to produce IFN-g, the question of its IFN-g antibodies or when the T cells were derived from IFN-
possible TRF activity could not be tackled until 1982 when g- or IFN-gR knock-out mice.
pure IFN-g and specific anti-IFN-g antibody became
available. In 1984 cloned murine IFN-g was found to induce
Ig secretion by resting B cells or by a B-cell tumour line [59]. 6. IFN-g produced by CD8+ cells
A second study [60] demonstrated identity of cloned IFN-g
with a T cell-line-derived factor required for purified B mouse In the early 1980s reports appeared demonstrating the
lymphocytes to produce an antibody response to sheep red ability of cytolytic T cells to produce IFN-g, alongside with
blood cells when stimulated by IL-1 and IL-2. Subsequently, other cytokines. These studies used alloreactive murine T
an inhibitory effect of IFN-g on B cell proliferation was cells, either freshly harvested [69] or obtained as clonal lines
shown [61,62]. Together, these observations suggested that [70,71]. It would take several years for the idea to permeate
one role of IFN-g might consist in redirecting B cells from that this might represent an accessory mechanism by which
proliferation towards differentiation. cytolytic CD8+ cells eliminate virus from infected tissues: in
As TRF activities from various sources were further 1990 Fong and Mosmann [72] noted that the cytokine
analysed, certain clonal murine T cell lines appeared to be production spectrum of murine alloreactive cytolytic clones
highly specialized in production of particular types of helper closely resembles that of Th1 cells, and postulated the
factors. One such clone (D9.1) was found to produce help existence in viral antigen-specific CD8+ clones of ‘cytokine-
that allowed LPS-stimulated B cells to selectively produce mediated mechanisms or functions that may be especially
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 103

important . . .in combating viruses, . . ..’ In fact, over the IFN-g also augmented expression of MHC Class II
ensuing years, studies devoted to this question did confirm molecules, but inhibited expression of other membrane
production of IFN-g by virus-specific CD8+ cells, but also receptor molecules that are also important in antigen
indicated that the role of IFN-g is particularly evident in presentation [84]. Production of IL-12 is an important
models of virus infections where the cytolytic capacity of marker of DC activation and there is evidence that, under
CD8+ cells is not expressed. Non-cytolytic viral clearance certain circumstances, exposure to IFN-g is a prerequisite
that depends both on CD8+ cells and on IFN-g has been for optimal production of this cytokine [85]. That IFN-g may
reported in models of persistent nervous system infection crucially affect DC function appears from experiments
with viruses such herpes virus [73], measles virus [74] or showing that injections of DC that have been exposed to
borna virus [75]. In fact, working with LCM-infected mice, IFN-g suppress autoimmune diabetes in NOD mice [86] and
Oldstone et al. [76] had already noted that the central reduce auto-antibody production and disease severity in
nervous system possesses a mechanism to avoid cytolysis mice with autoimmune myasthenia gravis [87].
and inflammation which in other tissues such as liver or
lungs accompanies viral clearance by CD8+ cells. In tissues
with a restricted potential for tissue renewal the ‘usual’ form 9. IFN-g and the chemokines
of viral clearance by elimination of virus-infected cells may
not be an optimal strategy. Chemokines came of age in the late 1980s with the
discovery of IL-8 (now also termed CXC chemokine ligand
8 or CXCL8). At that time, immunologists still considered
7. IFN-g and NK cells IFN-g mainly as a stimulant of immune defense against
viruses and ‘intracellular’ bacteria. The concept of it being
NK cells are both targets of IFN-g action and producers of an inflammation-regulating cytokine was poorly developed
IFN-g. Following the discovery that Type I interferons can up- (for a contemporary review, see [88]). Nevertheless, as a
regulate NK cell activity in vitro, a similar effect for IFN-g prologue to the chemokine era, in 1985 a report appeared
was reported [77]. The ability of NK cells to produce IFN-g demonstrating induction by IFN-g in human cells of a
was suggested in 1983 by Handa et al. [78] who showed IFN-g transcript encoding a protein named gIP-10 (g-IFN induced
production by mouse splenocytes enriched in NK cells by peptide-10, now also named CXCL10) with sequence
growth in IL-2. Later studies confirmed that NK cell-enriched homology to platelet factor 4 (PF4) [89]. Human PF4 (now
lymphocyte populations produce IFN-g in response to various also designated as CXCL4), had already been extensively
exogenous and endogenous stimuli, including the NK- studied by haemostasis researchers, who described it in 1964
stimulating factors IL-12 [79] and IL-18. In fact, IL-18 as an antiheparin factor [90]. Evidence for its chemotactic
was discovered by virtue of its ability to act as a co-stimulator potential would come forward as late as 1981 [91]. This 17
for IFN-g induction in resting T cells or NK cells by IL-2 or years-long delay testifies to the fact that haemostasis
anti-CD3 antibody. The factor was first observed in serum of researchers had until then manifested little interest in the
P. acnes-primed and LPS-challenged mice [80]. Subse- relation between blood clotting and inflammation.
quently, it was isolated from liver tissue of the mice [81] and In 1987 cDNAs of PF4 and gIP-10 were cloned and found
found to qualify as an IFN-g-inducing factor (IGIF). to share extensive sequence homology [92]. Also around that
time, interest for secretion of chemotactic cytokines had
been growing in several laboratories of leukocyte biology
8. IFN-g, dendritic cells (DC) and antigen such that in 1987 three teams independently reported the
presentation release by such cells of granulocyte chemotactic factor(s)
[93–96]. Originally the factors were named differently in
DCs are the most potent antigen-presenting cells (APC). each laboratory. Because in each instance the purified
Exploration of their diversity, maturation and function has protein appeared indistinguishable from the others, and
made fast progress in recent years. Insight into the role of because the general impression was that it represented the
IFN-g in their biology is of recent date and still evolving. next interleukin to be discovered, it was rather hastily named
Human peripheral blood monocytes differentiate into interleukin-8 (Larsen et al., 1989; International Meeting of
antigen-presenting DCs by cultivation in GM-CSF + IL-4; Novel Neutrophil Stimulating Peptides: Source, Structure
the alternative differentiation pathway leading to a macro- and Role in Inflammation, London, 1988, loc. cit. Ref. [97]).
phage phenotype is favoured by M-CSF, IL-6 and IL-10. In hindsight, this was not a wise decision, as both this new
When added during differentiation of DCs, IFN-g switches factor and PF4 would soon appear to belong to a large family
the differentiation pathway towards macrophages [82], of proteins with a molecular structure quite different from
partly due to endogenous production of M-CSF and IL-6. In that of all interleukins isolated up to that point. This episode
mice with a subnormal expression of MHC Class II signified the start of an avalanche of revelations on novel
molecules on DC, treatment with IFN-g was found to chemotactic factors, produced by various cell types and also
correct the defect [83]. In cultured human blood-derived DC affecting movement of various leukocyte categories. Today,
104 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

the numerous chemokines and the cellular receptors targeted most important in vivo role would consist in mounting an
by them are organized as molecular families and sub- effective response towards pathogens (bacteria, molds,
families. Synthesis and release of chemokines by leukocytes protozoans) whose elimination from the body depends on
as well as other cells can be constitutive or inducible by phagocytosis and intracellular killing. Today, our conception
exogenous or endogenous stimuli. Chemokines not only of the immune response to such agents holds that, early after
regulate leukocyte migration and homing; some of them also entry of these agents, NK cells produce IFN-g which primes
induce degranulation of leukocytes and regulate angiogen- mononuclear phagocytes for production of monokines,
esis. Production of chemokines and expression of their including TNF-a and IL-12. IFN-g and TNF-a then act in
receptors is under regulatory control by cytokines. As to IFN- concert to augment the bacteriostatic potential of the
g, studies have documented either positive or negative effects phagocytes. Later on, guided by IL-12, a Th1 response is
depending on the particular chemokine and cell system mounted resulting in additional IFN-g production by
considered. The general trend is for induction of neutrophil- activated CD4+ and CD8+ T cells.
chemotactic chemokines to be inhibited by IFN-g [98–104], Experimental work leading to these insights started
and for induction of mononuclear cell-chemotactic chemo- when, also around 1980, pure mouse IFN-g and specific
kines to be stimulated [99,102,103,105–109]. But according neutralizing (monoclonal) antibodies were made avail-
to some reports, exceptions to this trend may exist [110]. able. More refined analysis followed when gene knock-out
Moreover, in the whole animal, or in in vitro systems mice with defective expression of IFN-g or its receptor
consisting of more than a single cell type, IFN-g is more likely became available from 1993 onwards. Application of
to act in an indirect fashion by affecting production of other these tools in studies with animal models revealed a
cytokines (e.g. TNF-a) that may counteract the primary effect crucial role of IFN-g in defence against infections with
of IFN-g [111]. the most diverse bacteria. Beyond confirming the
These data are indicative of an important role for IFN-g importance of IFN-g-mediated macrophage activation,
in chemokine-controlled processes, such as traffic and these experiments also revealed additional mechanisms by
degranulation of leukocytes and angiogenesis. Whereas an which IFN-g influences the pathogenesis of infections. In
interpretation for stimulatory effects is usually easy to Listeria infection, administration of IFN-g was found to
conceive in terms of well-established pro-inflammatory augment normal resistance [113] or restore compromised
effects of IFN-g, an interpretation of inhibitory effects is less resistance [114,115]. Also, treatment with neutralizing
evident. In vague terms, inhibition of chemokine production antibodies against IFN-g abrogated resistance [116,117].
by IFN-g has been invoked to provide an explanation for the Importantly, however, IFN-g production during the first 2
sometimes puzzling anti-inflammatory roles of IFN-g in days of infection was shown to be critical for development
certain murine models of autoimmune diseases. However, of protective antigen-specific T cells [118]. Thus, it
detailed mechanistic schemes are so far mostly lacking. appeared that IFN-g not only optimizes effector cell
Clearly, more in depth analysis of the IFN-g/chemokine function, but also acts as a regulator of the adaptive
relation in defined in vivo situations is needed. One example immune response to the bacterial agent.
is the demonstration in collagen-induced arthritis (CIA) in As to the role of IFN-g in Mycobacterial infections, an
mice that ablation of IFN-g results in an excessive important early observation was that repeated percutaneous
proportion of neutrophils in the inflamed joint synovia, as application of IFN-g on skin lesions of lepromatous leprosy
opposed to predominance of mononuclears in IFN-g- patients caused increased infiltration with lymphocytes and
competent mice, a difference that was associated with reduction in the local bacterial load [119]. In the early 1990s,
higher levels of GCP-2/CXCL-6 [104]. Another example is as gene knock-out mice became available, mice deficient in
the demonstration of transient IFN-g-dependent inhibition IFN-g production proved unable to control otherwise
of the expression of two homeostatic chemokines (SLC and sublethal doses M. tuberculosis [120,121] or M. bovis
BCA-1/BLC) in the lymphoid tissues of mice after infection [122]. Bacteria multiplied more extensively and caused
with LCM (lymphocytic choreomeningitis) virus. This more widespread damage in affected tissues. Likewise, mice
modulation was associated with altered distribution of with a disrupted gene for the IFN-g receptor were found to
lymphocytes and dendritic cells in these tissues and with a fail controlling infection with M. bovis [123]. In these IFN-
transiently impaired immune response to a second primary g-ablated mice not only innate resistance in the early phase
challenge with a different pathogen [112]. of infection, but also later development of protective
immunity was compromised. Although some DTH did
develop, the mice failed to recover from infection suggesting
10. IFN-g in vivo that IFN-g, while dispensable for DTH, is essential for
protective immunity [120]. Along this line of thinking, IL-12
10.1. Defence against infection production was found to be absent in IFN-gR KO mice
infected with M. bovis [124], indicating that early
The specific activating effect of IFN-g on macrophages, production of IFN-g (by NK cells) is an important factor
established around 1980, was seen as an indication that its in directing the immune response towards the Th1 pathway.
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 105

These observations were also seen as an indication that IFN- antibodies [130], supporting the concept that production of
g might prove to be useful as vaccine adjuvant. However, IFN-g by TH1 cells is essential for the reaction. Using a
attempts to obtain protective immunity by adding IFN-g to local adoptive transfer assay with Ag-specific mouse TH1
killed vaccine preparations for Listeria or Mycobacteria clones, administration of anti-IFN-g antibodies was found to
have remained unsuccessful. inhibit responses in some but not all instances [131],
In further work with IFN-g-ablated mice, a role for IFN-g indicating that the positive contribution of IFN-g to the DTH
in immunopathological changes during mycobacterial infec- effector phase is real but can vary perhaps with specificity of
tions became evident. Splenomegaly and altered splenic the clone or with mouse strain. In the rat, administration of
architecture that accompanies such infections was shown to IFN-g-neutralizing monoclonal antibodies inhibited lym-
be enhanced in IFN-gR KO mice infected with M. bovis, phocyte recruitment in DTH sites induced with a
indicating that IFN-g mitigates excessive extramedullary haemocyanin [132]. Transgenic mice with ectopic expres-
myelopoiesis [125]. IFN-g was also found responsible for sion of IFN-g in the photoreceptors of the retina, were found
apoptosis of CD4+ T cells taking place in the later phases of to develop a DTH state, suggesting that intraocular IFN-g
the immune response to mycobacteria [126], and for the production can overcome the so-called anterior chamber-
appearance of ‘immunosuppressive’ macrophages [127]. associated immune deviation [133].
In a passive transfer model in mice allowing to study a
10.2. IFN-g and hypersensitivity reactions population of T cells that suppresses haptene (dinitrofluor-
obenzene – DNFB) contact sensitivity responsiveness,
10.2.1. Delayed-type hypersensitivity exogenous IFN-g was found to antagonize suppressive
The delayed-type hypersensitivity (DTH) cutaneous cells, suggesting the existence of a pathway by which IFN-g
reaction, first defined in the context of the immune response potentiates contact sensitivity [134]. However, in a DNFB
to mycobacteria, is universally considered as the prototype of contact sensitivity model in rats [135] IFN-g was found to
T-cell-mediated, antigen-specific inflammation. DTH inflam- act as a counter-regulator of the inflammatory changes:
matory reactions are crucial for defense against intracellular systemic administration of anti-IFN-g augmented ear
pathogens, but also represent the cellular mechanism under- swelling towards haptene challenge in sensitized animals.
lying pathologic responses to allergens, e.g. in contact This effect was accompanied by reduced MHC Class II
dermatitis, and autoantigens, e.g. in autoimmune diseases. antigen expression on keratinocytes. Administration of IFN-
Despite recognition, around 1980, of the close association of g in this same model inhibited ear swelling if given prior to
IFN-g with the host response to intracellular pathogens, and challenge, but not if given later [136]. The authors
although 7 years earlier Youngner and Salvin [22] had considered that MHC Class II antigen-expressing keratino-
described production of acid-labile interferon following PPD cytes are instrumental in the counterregulatory effect of IFN-
challenge in BCG-infected mice, the possibility that IFN-g g. Such cells had indeed been shown to be able to inhibit T
might somehow control DTH-associated inflammation was cell responses [137]. In disagreement with these studies in
not considered until the mid 1980s. This hesitation is rats, IFN-g receptor-deficient mice were found to display
remarkable if one considers that in 1975 De Maeyer had reduced hypersensitivity to haptens, as evident from reduced
described inhibitory effects of Type I interferon on DTH cellular infiltrates [138].
reactions in mice [128,129]. In his review on ‘Gamma More recent studies have indicated that IFN-g present in
Interferon’ of 1984, Kirchner wrote: ‘Based upon the well- DTH inflammatory sites originates mainly in CD8+ Tc1
known effects of [Type I] interferon on immune reactivities, a cells [139,140], while another CD8+ subpopulation releases
major role was postulated for IFN-g as an immunoregulatory IL-17 [141], which acts in a pro-inflammatory fashion.
molecule in reactions in which it is produced. However, such a Furthermore, some of the CD4+ cells were identified as
role is far from being proven’. A simple explanation may be CD25+ Treg cells [142,143]. Involvement of IL-15 was
that the tools to examine the in vivo role of IFN-g, pure IFN-g evident from a study using IL-15 receptor-deficient mice, as
and monospecific neutralizing antibodies, were still lacking. these mice displayed reduced hapten contact hypersensi-
When these tools became available in the mid-1980s, their tivity together with reduced infiltration by CD8+ cells,
application to several animal models of DTH revealed the diminished levels of IFN-g and chemokines CCL5/
existence of both stimulatory and inhibitory IFN-g-dependent RANTES and CXCL10/IP-10 [144].
circuits of which one or the other appeared to predominate Most likely, the ambiguity in reports analyzing the role of
depending on the model system under study. This finding went IFN-g in DTH reactions reflects the pathogenic complexity
against the expectation that IFN-g would play a positive role of the systems under study. The effects may differ depending
in DTH, mainly because, at the time, DTH hypersensitivity on the antigen used (protein or hapten), the route of exposure
had become seen as a typical Th1-type immune response, (injection or contact with the skin) and the time point during
involving mainly IFN-g-producing CD4+ cells. the reaction (during the sensitization or the elicitation
In a DTH induction model using sheep erythrocytes as an phase). Furthermore, DTH reactions rely on both natural and
antigen in mice, exogenous IFN-g was found to reverse acquired immune response mechanisms and, again, IFN-g
inhibition of the response by anti-CD4 or anti-IL-2R may act differently on these components.
106 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

10.2.2. Hypersensitivity reactions elicited by and possibly also TNF’s target cells (endothelial cells and
endotoxin—the Shwartzman-type reactions various leukocytes). Several studies were undertaken to
First described in the late 1920s in rabbits, the Shwartz- identify the cellular source of IFN-g during Shwartzman-
man reaction exists in a localized and a generalized variant type reactions [156–158]. Together these studies pointed
(for review, see Ref. [145]). Both are two-stage phenomena, mainly at NK and NKT cells.
requiring a preparative (sensitizing) injection of endotoxin
followed after about 24 h by an eliciting (provoking) 10.2.3. IFN-g and autoimmune disease
injection. Classically, to obtain a local Shwartzman reaction In the mid-1980s recombinant IFN-g and neutralizing
in the rabbit, the preparative dose is given in the skin, and the monoclonal antibodies became available for in vivo experi-
eliciting dose is given intravenously. Within a few hours a ments on the role of endogenous IFN-g in experimental
thrombohemorrhagic reaction appears at the prepared skin diseases. At the time, the major lead regarding pathogenesis of
site. When both doses are given intravenously, a generalized autoimmune diseases was the association with MHC genes,
reaction can occur, which often leads to death. This reaction and IFN-g had just become known as a prominent stimulator
resembles the thrombohemorrhagic shock that sometimes of expression of MHC Class II genes. Most notably, IFN-g
occurs in humans affected by meningococcal sepsis. The was found able to induce MHC Class II gene expression by
hypothesis of involvement of endogenous ‘mediators’ (e.g. cells that are normally Class II-negative [159,160], and it was
complement, serotonin and prostglandins) in the Swartzman proposed that such ‘ectopic’ expression MHC Class II
reaction already prevailed when in the early 1980s antigens might result in cross-presentation of auto-antigens,
interleukin-1 and TNF/cachectin were characterized bio- leading to breakdown of peripheral tolerance and develop-
chemically. Both these cytokines were conceptually linked ment of autoimmune disease. In support of this view,
to the biological effects of endotoxin, IL-1 as the proposed transgenic mouse strains harboring the MHC Class II or IFN-
‘endogenous pyrogen’ responsible for fever induction by g genes linked to the insulin promoter were shown to develop
endotoxin, and TNF/cachectin as the mediator of the tumor insulin-dependent diabetes mellitus [161]. Furthermore, a
necrosis effect of endotoxin. No wonder that, in the early disease-promoting role of IFN-g in autoimmune diseases was
1980s, investigators of the Shwartzman phenomenon turned confirmed in other animal models of autoimmune diabetes
to these two cytokines as the possible endogenous mediators [162,163] as well as in lupus-like nephritis [164] and
of their phenomenon [146]. autoimmune neuritis [165].
Contrastingly, the question as to a possible involvement The concept of a possible role of IFN-g in autoimmune
of IFN-g did not immediately come up. The reason for this pathogenesis became more specific with the then emerging
may have been that, at the time, evidence for specific Th1/Th2 paradigm [166] which heralded IFN-g as the
induction of IFN-g by endotoxin, or for any role of IFN-g- hallmark of the Th1-type immune response. Auto-immune
producing cells (T cells and NK cells) in endotoxin encephalomyelitis (EAE) and similar autoimmune diseases
pathology was minimal. In fact, the first evidence that that had previously been classified as ‘cell-mediated’, now
IFN-g does indeed play a crucial role in the Shwartzman came to be seen as typically Th1-mediated. Accordingly,
phenomenon came from a separate line of investigation in ablation of endogenous IFN-g was presumed to inhibit such
which mouse footpad reactions elicited by endotoxin were diseases. Surprisingly, however, experiments with neutraliz-
used as an assay to detect anti-inflammatory mediators ing anti-IFN-g antibodies had already been yielding
present in various biological fluids [147]. Pre-treatment with evidence to the contrary. Most notably, administration of
neutralizing anti-IFN-g completely protected the mice anti-IFN-g antibody to mice was found to cause aggravation
against this reaction [148,149]. The essence of this of EAE [167], an observation many times confirmed
observation was confirmed by studies using several other subsequently [168–170] and complemented by evidence
variants of endotoxin-induced lethal reactions in mice [150– for a similar protective effect of endogenous IFN-g in other
155]. One of these studies compared the protective models of T cell-mediated autoimmune disease, notably
effectiveness of anti-IFN-g antibody with that of anti- experimental autoimmune uveitis (EAU) [171] and CIA
TNF-a antibody and found the latter one to be ineffective [172–175].
[154], suggesting that in some forms of endotoxin-mediated Over the 20 years since the first report on these
lethality, endogenous IFN-g may be more important than counterintuitive effects of IFN-g, several possible underlying
endogenous TNF-a. mechanisms have been proposed (for reviews, see Ref. [176–
Injection of the anti-IFN-g antibody at different time 180]). Though each of these may play a certain role at some
points allowed to situate the action of IFN-g in the interval stage, their integration into a comprehensive explanatory
between the preparative and the eliciting dose. Furthermore, model has only recently become possible, thanks to
mice treated with anti-IFN-g antibody showed reduced recognition of important novel cell lineages, in particular
production of circulating TNF following the eliciting dose Th17 cells, Treg cells and tolerogenic DCs. The identification
[149]. Together, these data supported a model according to of the Th17 cell lineage in 2005 [68,181] has necessitated
which the preparative dose relies on IFN-g in order to revision of the Th1/Th2 paradigm (for review, see Ref. [67]).
sensitize TNF-producing cells (mononuclear phagocytes) The experimental autoimmune diseases, EAE, EAU and CIA,
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 107

formerly considered as Th1-driven conditions, are now seen some other cells [192]. B7-H1 and B7-H2 are two recently
as Th17-driven. Indeed, deficient production or action of IL- discovered members of the B7 family that negatively regulate
17 is associated with loss of susceptibility to these T cell activation by interaction with the ‘programmed death 1’
experimental autoimmune diseases [182,183]. Naı̈ve T cells (PD1) molecule on T cells.
differentiate into Th17 cells when their cytokine environment, Whereas these models of IFN-g-mediated control of
created by antigen-presenting cells, is dominated by IL-6 and autoimmune disease are focused on antigen-specific T cells
IL-23. However, conditions for optimal in vitro induction of as taking the central position in gain or loss of peripheral
IL-17-producing T cells by IL-23 were found to include tolerance, it should be recognized that the protective effect
neutralization of endogenous IFN-g [68]. Also, during in vivo of IFN-g in experimental autoimmune disease may derive
infection with mycobacteria, ablation of IFN-g resulted in from effects on non-antigen-specific stages. Intriguing is the
increased numbers of IL-17-producing T cells [184]. This led observation that the protective effect of IFN-g seems to be
to the conclusion that endogenous IFN-g inhibits differentia- limited to models that use CFA as an adjuvant to immunize
tion of Th17 cells. Accordingly it is not more than normal that animals with the (auto)antigen. In the case of collagen-
functional ablation of endogenous IFN-g should boost disease induced arthritis, omission of CFA from the experimental
in EAE and CIA. protocol was found to reverse the effect of endogenous IFN-
Another significant development of recent date has been g from being protective to disease-promoting. Further
the re-emergence of the concept that T cell autoreactivity is analysis of this system revealed the possibility that IFN-g
kept in check by ‘‘suppressor’’ T cells. Best studied so far are might act by inhibiting extramedullary myelopoiesis
the thymus-derived CD4+CD25+ natural regulatory T cells brought about by the immunoadjuvant [193–195]. Even
(Treg cells). Their importance in controlling autoimmunity more striking was the finding that injection of IFN-gR KO
is most clearly illustrated by the development of auto- mice with only CFA (without a joint-specific antigen)
immune disease in athymic nude mice upon transfer of Treg- resulted in the development of a full erosive arthritis [196].
depleted CD4+ T cells. In addition, many studies have These observations are highly suggestive of an important
documented blocking of autoimmune diseases by injection contribution of innate immune mechanisms in these model
of syngeneic Treg cells. Two important aspects concern the autoimmune diseases. Since CFA contains killed myco-
role of IFN-g in Treg functioning. In vitro treatment of bacteria, toll-like receptors (TLR) come to mind as possible
CD4+CD25 cells with IFN-g was shown to cause their players in autoimmune pathogenesis. Recognition of
conversion into CD4+ Treg cells, with increased expression microbial stimuli by TLRs on myeloid cells is essential
of FoxP3, a specific marker for Treg cells, and acquisition of for the regulation of inflammatory reactions and adaptive
full regulatory properties as evident by their ability to immunity but imbalances in the feedback control of TLR-
suppress EAE by adoptive transfer [185]. Furthermore, activated innate immune cells, by non-functioning of one
evidence exists for induced Treg cells to rapidly release IFN- single cytokine such as IFN-g, may lead to autoimmunity,
g, that may be important for their suppressive activity [186]. and this will be an interesting field for further research.
Increased severity of CIA in IFN-gR KO mice was found to
be associated with reduced Treg activity. In vitro analysis by
co-culturing Treg with Teff cells from both IFN-gR KO and 11. Concluding thoughts
wild-type mice suggested that a third population of
accessory cells participates in the stimulatory effect of Almost 5 decades separate us from the first report on the
IFN-g on Treg cells in wild-type mice [187]. release by leukocyte cultures of an interferon-like activity
Thirdly, studies on the role of IFN-g in regulating diversity that would later appear to represent IFN-g. Since then, the
and function of DCs have revealed pathways by which IFN-g number and variety of biological activities assigned to IFN-g
can safeguard peripheral tolerance. One such pathway have steadily increased. In immune networks IFN-g operates
concerns release by certain DCs of the enzyme indoleamino at multiple successive levels. Moreover, actions at one level
2,3-dioxygenase (IDO) [188,189] (for review, see Ref. may counteract effects at others. Though such pleotypic
[190,191]). Among cytokines IFN-g is the most potent effects and negative feedback loops are usual features of
inducer of IDO. The enzyme degrades the indole moiety of cytokines in general, in the case of IFN-g, the long history of
tryptophan, serotonin and melatonin, and produces kynur- research has made that we have a fairly detailed conception
enines, some of which elicit apoptosis of T cells. By depleting of the resulting network and its impact on in vivo situations.
tryptophan, IDO also contributes to innate host resistance Today IFN-g, perhaps more than any other cytokine, figures
against those pathogens that depend metabolically on high in many of our pathogenesis schemes of anti-infectious
exogenous supply of the amino acid. However, local depletion host defense, of inflammatory conditions and of auto-
of tryptophan also inhibits proliferation of T cells. Similar immune disease. A remarkable feature in the history of IFN-
effects on T cell proliferation and induction of apoptosis have g research has been the evolution of its conceived place in
been described for the IFN-g-induced production of nitric the immune system from a merely antiviral factor to a broad-
oxide. IFN-g may also control peripheral tolerance by up- spectrum antimicrobial agent, and finally to an important
regulating expression of B7-H1 and B7-H2 on APCs and player in overall inflammatory responses to exogenous as
108 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

well as endogenous agents. As with other cytokines, [14] Smith TJ, Wagner RR. Rabbit macrophage interferons. II. Some
neutralizing antibodies and gene knock-out mice have been physicochemical properties and estimations of molecular weights. J
Exp Med 1967;125:579–93.
foremost important tools to reveal these roles. However, [15] Havell EA, Spitalny GL. Endotoxin-induced interferon synthesis in
more than with other cytokines, the experimental results macrophage cultures. J Reticuloendothel Soc 1983;33:369–80.
have confronted researchers with difficult-to-resolve para- [16] Falcoff R. Some properties of virus- and immune-induced human
doxes. Whereas multiple immunoregulatory pathways have lymphocyte interferons. J Gen Virol 1972;16:251–3.
[17] Merigan TC. Non-viral substances which induce interferons. In:
been identified by which these paradoxes can possibly be
Finter NB, editor. Interferons and interferon inducers. Amsterdam,
resolved, the relative importance of each pathway in each London, New York: North-Holland Publishing Cy. & American
particular experimental situation remains to be determined. Elsevier Co., Inc.; 1973. p. 45–72.
A topic not covered in this historical review is the clinical [18] Ho M, Armstrong JA. Interferon. Annu Rev Microbiol 1975;29:131–
application of IFN-g. In contrast to Type I interferons, IFN-g 61.
has not reached the stage of a clinically useful drug. While [19] Stewart WEI. The interferon system. Wien, New York: Springer-
Verlag; 1979.
pilot trials with exogenous administration of IFN-g have been [20] Glasgow LA. Leukocytes and interferon in the host response to viral
indicative of beneficial effects in some clinical conditions, in infections. II. Enhanced interferon response of leukocytes from
particular visceral leishmaniasis and chronic granulomatous immune animals. J Bacteriol 1966;91:2185–91.
disease, trials establishing applicability of IFN-g in major [21] Green JA, Cooperband SR, Kibrick S. Immune-specific induction of
human pathologies are not available. As evident from this interferon production in cultures of human blood lymphocytes.
Science 1969;164:1415–7.
review, IFN-g has been demonstrated to play a critical role in [22] Youngner JS, Salvin SB. Production and properties of migration
models of acute and chronic inflammatory diseases. Whereas inhibitory factor and interferon in the circulation of mice with
this role is comparable to that of TNF-a, antibodies against delayed type hypersensitivity. J Immunol 1973;111:1914–22.
IFN-g, unlike those against TNF-a, have so far not found their [23] Knight Jr E, Hunkapillar MW, Korant BD, Hardy RWF, Hood LE.
Human fibroblast interferon: amino acid analysis and amino terminal
way into the clinic.
amino acid sequence. Science 1980;207:525–6.
In conclusion, two important further paths of discovery [24] Knight Jr E, Hunkapillar MW. Purification, amino acid composition,
still lie ahead: (1) annotating the network of interactions of and N-terminal amino acid sequence of human fibroblast interferon.
IFN-g such that it becomes increasingly suitable to interpret Ann NY Acad Sci 1980;350:385–9.
its role in natural diseases, and (2) expanding the [25] Zoon KC, Smith ME, Bridgen PJ, Anfinsen CB, Hunkapiller MW,
applicability of IFN-g (or IFN-g antagonists) in the clinic. Hood LE. Amino terminal sequence of the major component of
human lymphoblastoid interferon. Science 1980;207:527–8.
[26] Okamura H, Berthold W, Hood L, Hunkapiller M, Inoue M, Smith-
Johannsen H, et al. Human fibroblastoid interferon: immunosorbent
References column chromatography and N-terminal amino acid sequence. Bio-
chemistry 1980;19:3831–5.
[1] Billiau A. Interferon: the pathways of discovery. I. Molecular and [27] Anonymous. Interferon nomenclature. Nature 1980;286:110.
cellular aspects. Cytokine Growth Factor Rev 2006;17:381–409. [28] Langford MP, Georgiades JA, Stanton GJ, Dianzani F, Johnson HM.
[2] Billiau A. Interferons: the pathways of discovery. II. Immunomodu- Large-scale production and physicochemical characterization of
latory, in vivo and applied aspects. J Clin Virol 2007;39:241–65. human immune interferon. Infect Immun 1979;26:36–41.
[3] Epstein L. The comparative biology of immune and classical [29] De Ley M, Van Damme J, Claeys H, Weening H, Heine JW, Billiau
interferons. In: Cohen S, Pick E, Oppenheim JJ, editors. Biology A, et al. Interferon induced in human leukocytes by mitogens:
of lymphokines. Academic Press, Inc.; 1979. p. 443–514. production, partial purification and characterization. Eur J Immunol
[4] Kirchner H. Interferon gamma. Progr Clin Biochem Med 1984;1: 1980;10:877–83.
171–203. [30] Yip YK, Pang RH, Urban C, Vilcek J. Partial purification and
[5] Ijzermans J, Marquet RL. Interferon-g: a review. Immunobiology characterization of human gamma (immune) interferon. Proc Natl
1989;179:456–73. Acad Sci USA 1981;78:1601–5.
[6] Young HA, Hardy KJ. Interferon-g: producer cells, activation stimuli, [31] Yip YK, Barrowclough BS, Urban C, Vilcek J. Molecular weight of
and molecular genetic regulation. Pharmacol Ther 1990;45:137–51. human gamma interferon is similar to that of other human inter-
[7] Vilcek J, Oliveira IC. Recent progress in the elucidation of interferon- ferons. Science 1982;215:411–3.
gamma actions: molecular biology and biological functions. Int Arch [32] Gray PW, Goeddel DV. Structure of the human immune interferon
Allergy Immunol 1994;104:311–6. gene. Nature 1982;298:859–63.
[8] Young HA, Hardy KJ. Role of interferon-g in immune cell regulation. [33] Devos R, Cheroutre H, Taya Y, Degrave W, Van Heuverswyn H, Fiers
J Leukoc Biol 1995;58:373–81. W. Molecular cloning of human immune interferon cDNA and its
[9] Billiau A. Interferon-g: biology and role in pathogenesis. Adv expression in eukaryotic cells. Nucleic Acids Res 1982;10:2487–501.
Immunol 1996;62:61–130. [34] Taya Y, Devos R, Tavernier J, Cheroutre H, Engler G, Fiers W.
[10] Billiau A, Vandenbroeck K. IFN-g. In: Oppenheim JJ, Feldman M, Cloning and structure of the human immune interferon-gamma
Durum SK, Hirano T, Vilcek J, Nicola NA, editors. Cytokine chromosomal gene. EMBO J 1982;1:953–8.
reference. Academic Press; 2000. p. 642–88. [35] Dumonde DC, Wolstencroft RA, Panyi GS, Matthew M, Morley J,
[11] Schroder K, Hertzog PJ, Ravasi T, Hume DA. Interferon-g: an Howson WT. ‘‘Lymphokines’’: non-antibody mediators of cellular
overview of signals, mechanisms and functions. J Leukoc Biol immunity generated by lymphocyte activation. Nature 1969;224:38–
2004;75:163–89. 42.
[12] Wheelock EF. Interferon-like virus inhibitor induced in human [36] Dumonde DC. The role of the macrophage in delayed hypersensi-
leukocytes by phytohemagglutinin. Science 1965;149:310–1. tivity. Br Med Bull 1967;23:9–14.
[13] Ho M. Interferon-like viral inhibitor in rabbits after intravenous [37] Bloom BR, Bennett B. Mechanism of a reaction in vitro associated
administration of endotoxin. Science 1964;146:1472–4. with delayed-type hypersensitivity. Science 1966;153:80–2.
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 109

[38] Bennett B, Bloom BR. Reactions in vivo and in vitro produced by a [58] Schimpl A, Wecker E. Stimulation of IgG antibody response in vitro
soluble substance associated with delayed-type hypersensitivity. Proc by T-cell-replacing factor. J Exp Med 1973;137:547–52.
Natl Acad Sci USA 1968;59:756–62. [59] Sidman CL, Marshall JD, Schulz LD, Gray PW, Johnson HM. g-
[39] Oppenheim JJ, Wolstencroft RA, Gell PG. Delayed hypersensitivity Interferon is one of several direct B cell-maturing lymphokines.
in the guinea-pig to a protein-hapten conjugate and its relationship to Nature 1984;309:801–3.
in vitro transformation of lymph node, spleen, thymus and peripheral [60] Leibson HJ, Gefter M, Zlotnik A, Marrack P, Kappler JW. Role of
blood lymphocytes. Immunology 1967;12:89–102. gamma-interferon in antibody-producing responses. Nature
[40] Granger GA, Williams TW. Lymphocyte cytotoxicity in vitro: acti- 1984;309:799–801.
vation and release of a cytotoxic factor. Nature 1968;218:1253–4. [61] Mond JJ, Finkelman FD, Sarma C, Ohara J, Serrate S. Recombinant
[41] Ruddle NH, Waksman BH. Cytotoxicity mediated by soluble antigen interferon-gamma inhibits the B cell proliferative response stimu-
and lymphocytes in delayed hypersensitivity. II. Correlation of the in lated by soluble but not by Sepharose-bound anti-immunoglobulin
vitro response with skin reactivity. J Exp Med 1968;128:1255–65. antibody. J Immunol 1985;135:2513–7.
[42] Salvin SB, Youngner JS, Lederer WH. Migration inhibitory factor [62] Rabin EM, Mond JJ, Ohara J, Paul WE. Interferon-gamma inhibits
and interferon in the circulation of mice with delayed hypersensi- the action of B cell stimulatory factor (BSF)-1 on resting B cells. J
tivity. Infect Immun 1973;7:68–75. Immunol 1986;137:1573–6.
[43] Mackaness GB. The influence of immunologically committed lym- [63] Coffman RL, Carty J. A T cell activity that enhances polyclonal IgE
phoid cells on macrophage activity in vivo. J Exp Med 1969;129: production and its inhibition by interferon-g. J Immunol
973–92. 1986;136:949–54.
[44] Buchmüller Y, Mauel J. Studies on the mechanisms of macrophage [64] Stevens TL, Bossie A, Sanders VM, Fernandez-Botran R, Coffman
activation. II. Parasite destruction in macrophages activated by RL, Mosmann TR, et al. Regulation of antibody isotype secretion by
supernates from concanavalin A-stimulated lymphocytes. J Exp subsets of antigen-specific helper T cells. Nature 1988;334:255–8.
Med 1979;150:359–70. [65] Gajewski TF, Fitch FW. Anti-proliferative effect of IFN-g in
[45] Nathan C, Murray HW, Wiebe ME, Rubin BY. Identification of immune regulation. I. IFN-g inhibits the proliferation of Th2 but
interferon-g as the lymphokine that activates human macrophage not Th1 murine helper T lymphocyte clones. J Immunol 1988;140:
oxidative metabolism and antimicrobial activity. J Exp Med 4252.
1983;158:670–81. [66] Pernis A, Gupta, Gollob KJ, Garfein E, Coffman RL, Schindler C,
[46] Steeg P, Moore RN, Johnson HM, Oppenheim JJ. Regulation of et al. Lack of interferon g receptor b chain and the prevention of
murine macrophage Ia antigen expression by a lymphokine with interferon g signalling in TH1 cells. Science 1995;269:245–7.
immune interferon activity. J Exp Med 1982;156:1780–93. [67] Weaver CT, Hatton RD, Mangan PR, Harrington LE. IL-17 family
[47] Lohmann-Matthes ML, Ziegler FG, Fischer H. Macrophage cyto- cytokines and the expanding diversity of effector T cell lineages.
toxicity factor. A product of in vitro sensitized thymus-dependent Annu Rev Immunol 2007;25:821–52.
cells. Eur J Immunol 1973;3:56–8. [68] Harrington LE, Hatton RD, Mangan PR, Turner H, Murphy TL,
[48] Weinberg JB, Chapman HA, Hibbs JB. Characterization of the effects Murphy KM, et al. Interleukin 17-producing CD4+ effector T cells
of endotoxin on macrophage tumor cell killing. J Immunol develop via a lineage distinct from the T helper type 1 and 2 lineages.
1978;121:72–80. Nat Immunol 2005;6:1123–32.
[49] Crowle AJ, May M. Preliminary demonstration of human tubercu- [69] Prystowsky MB, Ely JM, Beller DI, Eisenberg L, Goldman J,
loimmunity in vitro. Infect Immun 1981;31:453–64. Goldman M, et al. Alloreactive cloned T cell lines. VI. Multiple
[50] Nacy CA, Meltzer MS, Leonard EJ, Wyler DJ. Intracellular replica- lymphokine activities secreted by helper and cytolytic cloned T
tion and lymphokine-induced destruction of Leishmania tropica in lymphocytes. J Immunol 1982;129:2337–44.
C3H/HeN mouse macrophages. J Immunol 1981;127:2381–6. [70] Kelso A, Glasebrook AL. Secretion of interleukin 2, macrophage-
[51] Zlotnik A, Crowle AJ. Lymphokine-induced mycobacteriostatic activ- activating factor, interferon, and colony-stimulating factor by allor-
ity in mouse pleural macrophages. Infect Immun 1982;37:786–93. eactive T lymphocyte clones. J Immunol 1984;132:2924–31.
[52] Nacy CA, James SL, Benjamin WR, Farrar JJ, Hockmeyer WT, [71] Guerne PA, Piguet PF, Vassalli P. Production of interleukin 2,
Meltzer MS. Activation of macrophages for microbicidal and tumor- interleukin 3, and interferon by mouse T lymphocyte clones of
icidal effector functions by soluble factors from EL-4, a continuous T Lyt-2+ and -2 phenotype. J Immunol 1984;132:1869–71.
cell line. Infect Immun 1983;40:820–4. [72] Fong TA, Mosmann TR. Alloreactive murine CD8+ T cell clones
[53] Douvas GS, Looker DL, Vatter AE, Crowle AJ. Gamma interferon secrete the Th1 pattern of cytokines. J Immunol 1990;144:1744–52.
activates human macrophages to become tumoricidal and leishma- [73] Liu T, Khanna KM, Carriere BN, Hendricks RL. Gamma interferon
nicidal but enhances replication of macrophage-associated myco- can prevent herpes simplex virus type 1 reactivation from latency in
bacteria. Infect Immun 1985;50:1–8. sensory neurons. J Virol 2001;75:11178–84.
[54] Rothermel CD, Rubin BY, Murray HW. Gamma-interferon is the [74] Patterson CE, Lawrence DM, Echols LA, Rall GF. Immune-mediated
factor in lymphokine that activates human macrophages to inhibit protection from measles virus-induced central nervous system dis-
intracellular Chlamydia psittaci replication. J Immunol ease is noncytolytic and gamma interferon dependent. J Virol
1983;131:2542–4. 2002;76:4497–506.
[55] van Dissel JT, Stikkelbroeck JJM, Michel BC, Van den Barselaar MT, [75] Hausmann J, Pagenstecher A, Baur K, Richter K, Rziha HJ, Staeheli
Leijh PC, Van Furth R. Inability of recombinant interferon-g to P. CD8 T cells require gamma interferon to clear borna disease virus
activate the antibacterial activity of mouse peritoneal macrophages from the brain and prevent immune system-mediated neuronal
against Listeria monocytogenes and Salmonella typhimurium. J damage. J Virol 2005;79:13509–18.
Immunol 1987;2-139:1673–8. [76] Oldstone MB, Blount P, Southern PJ, Lampert PW. Cytoimmunother-
[56] Campbell PA, Canono BP, Cook JL. Mouse macrophages stimulated apy for persistent virus infection reveals a unique clearance pattern
by recombinant gamma interferon to kill tumor cells are not bacter- from the central nervous system. Nature 1986;321:239–43.
icidal for the facultative intracellular bacterium Listeria monocyto- [77] Senik A, Stefanos S, Kolb JP, Lucero M, Falcoff E. Enhancement of
genes. Infect Immun 1988;56:1371–5. mouse natural killer cell activity by Type II interferon. Ann Immunol
[57] Leenen PJ, Canono BP, Drevets DA, Voerman JS, Campbell PA. (Inst Pasteur) 1980;131c:349–61.
TNF-a and IFN-g stimulate a macrophage precursor cell line to kill [78] Handa K, Suzuki R, Matsui H, Shimizu Y, Kumagai K. Natural killer
Listeria monocytogenes in a nitric oxide-independent manner. J (NK) cells as a responder to interleukin 2 (IL 2). II. IL 2-induced
Immunol 1994;153:5141–7. interferon gamma production. J Immunol 1983;130:988–92.
110 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

[79] Chan SH, Perussia B, Gupta JW, Kobayashi M, Pospisil M, Young response to interleukin-1 or tumour necrosis factor. Immunology
HA, et al. Induction of interferon gamma production by natural killer 1989;68:31–6.
cell stimulatory factor: characterization of the responder cells and [98] Oliveira IC, Sciavolino PJ, Lee TH, Vilcek J. Downregulation of
synergy with other inducers. J Exp Med 1991;173:869–79. interleukin 8 gene expression in human fibroblasts: unique mechan-
[80] Nakamura K, Okamura H, Nagata K, Komatsu T, Tamura T. Pur- ism of transcriptional inhibition by interferon. Proc Natl Acad Sci
ification of a factor which provides a costimulatory signal for gamma USA 1992;89:9049–53.
interferon production. Infect Immun 1993;61:64–70. [99] Rathanaswami P, Hachicha M, Sadick M, Schall TJ, McColl SR.
[81] Okamura H, Nagata K, Komatsu T, Tanimoto T, Nukata Y, Tanabe F, Expression of the cytokine RANTES in human rheumatoid synovial
et al. A novel costimulatory factor for gamma interferon induction fibroblasts. Differential regulation of RANTES and interleukin-8
found in the livers of mice causes endotoxic shock. Infect Immun genes by inflammatory cytokines. J Biol Chem 1993;268:5834–9.
1995;63:3966–72. [100] Gusella GL, Musso T, Bosco MC, Espinoza-Delgado I, Matsushima
[82] Delneste Y, Charbonnier P, Herbault N, Magistrelli G, Caron G, K, Varesio L. IL-2 up-regulates but IFN-g suppresses IL-8 expression
Bonnefoy JY, et al. Interferon-gamma switches monocyte differen- in human monocytes. J Immunol 1993;151:2725–32.
tiation from dendritic cells to macrophages. Blood 2003;101:143–50. [101] Oliveira IC, Mukaida N, Matsushima K, Vilcek J. Transcription
[83] Akbar SM, Inaba K, Onji M. Upregulation of MHC class II antigen inhibition of the interleukin-8 gene by interferon is mediated by
on dendritic cells from hepatitis B virus transgenic mice by inter- the NF-kB site. Mol Cell Biol 1994;14:5300–8.
feron-gamma: abrogation of immune response defect to a T-cell- [102] Tran EH, Prince EN, Owens T. IFN-g shapes immune invasion of the
dependent antigen. Immunology 1996;87:519–27. central nervous system via regulation of chemokines. J Immunol
[84] Rongcun Y, Maes H, Corsi M, Dellner F, Wen T, Kiessling R. Interferon 2000;164:2759–68.
gamma impairs the ability of monocyte-derived dendritic cells to [103] Robson RL, McLoughlin RM, Witowski J, Loetscher P, Wilkinson
present tumour-specific and allo-specific antigens and reduces their TS, Jones SA, et al. Differential regulation of chemokine production
expression of CD1A, CD80 AND CD4. Cytokine 1998;10:747–55. in human peritoneal mesothelial cells: IFN-gamma controls neutro-
[85] Mosca PJ, Hobeika AC, Clay TM, Nair SK, Thomas EK, Morse MA, phil migration across the mesothelium in vitro and in vivo. J Immunol
et al. A subset of human monocyte-derived dendritic cells expresses 2001;167:1028–38.
high levels of interleukin-12 in response to combined CD40 ligand [104] Kelchtermans H, Struyf S, De Klerck B, Mitera T, Alen M, Geboes L,
and interferon-gamma treatment. Blood 2000;96:3499–504. et al. Protective role of IFN-gamma in collagen-induced arthritis
[86] Shinomiya M, Fazle Akbar SM, Shinomiya H, Onji M. Transfer of conferred by inhibition of mycobacteria-induced granulocyte che-
dendritic cells (DC) ex vivo stimulated with interferon-gamma (IFN- motactic protein-2 production. J Leukoc Biol 2007;81:1044–53.
gamma) down-modulates autoimmune diabetes in non-obese dia- [105] Hachicha M, Rathanaswami P, Schall TJ, McColl SR. Production of
betic (NOD) mice. Clin Exp Immunol 1999;117:38–43. monocyte chemotactic protein-1 in human type B synoviocytes:
[87] Adikari SB, Lian H, Link H, Huang YM, Xiao BG. Interferon- synergistic effect of tumor necrosis factor a and interferon-G.
gamma-modified dendritic cells suppress B cell function and ame- Arthritis Rheum 1993;36:26–34.
liorate the development of experimental autoimmune myasthenia [106] Marfaing-Koka A, Devergne O, Gorgone G, Portier A, Schall TJ,
gravis. Clin Exp Immunol 2004;138:230–6. Galanaud P, et al. Regulation of the production of the RANTES
[88] Billiau A. Interferons and inflammation. J Interferon Res 1987;7: chemokine by endothelial cells: synergistic induction by IFN-g plus
559–67. TNF-a and inhibition by IL-4 and IL-13. J Immunol 1995;154:
[89] Luster AD, Unkeless JC, Ravetch JV. Gamma-interferon transcrip- 1870–8.
tionally regulates an early-response gene containing homology to [107] Meda L, Bernasconi S, Bonaiuto C, Sozzani S, Zhou D, Otvos L,
platelet proteins. Nature 1985;315:672. et al. b-Amyloid (25–35) peptide and IFN-g synergistically induce
[90] Poplawsky A, Niewiarowaki S. Dissociation of platelet antiheparin the production of the chemotactic cytokine MCP-1/JE in monocytes
factor (platelet factor 4) from lipoprotein lipase inhibitor. Biochim and microglial cells. J Immunol 1996;157:1213–8.
Biophys Acta 1964;90:403–5. [108] Zhou Z-HL, Chaturvedi P, Han YL, Aras S, Li YS, Kollattukudy PE,
[91] Deuel TF, Senior RM, Chang D, Griffin GL, Heinrikson RL, Kaiser et al. IFN-g induction of the human monocyrte chemoattractant
ET. Platelet factor 4 is chemotactic for neutrophils and monocytes. protein (hMCP)-1 gene in astrocytoma cells: functional interaction
Proc Natl Acad Sci USA 1981;78:4584–7. between an IFN-g-activated site and a GC-rich element. J Immunol
[92] Poncz M, Surrey S, LaRocco P, Weiss MJ, Rappaport EF, Conway TM, 1998;160:3908–16.
et al. Cloning and characterization of platelet factor 4 cDNA derived [109] Mohammed KA, Nasreen N, Ward MJ, Antony VB. Helper T cell
from a human erythroleukemic cell line. Blood 1987;69:219–23. type 1 and 2 cytokines regulate C–C chemokine expression in mouse
[93] Yoshimura T, Matsushima K, Tanaka S, Robinson EA, Appella E, pleural mesothelial cells. Am J Crit Care Med 1999;159:1653–9.
Oppenheim JJ, et al. Purification of a human monocyte-derived [110] Loos T, Dekeyzer L, Struyf S, Schutyser E, Gijsbers K, Gouwy M,
neutrophil chemotactic factor that has peptide sequence similarity et al. TLR ligands and cytokines induce CXCR3 ligands in endothe-
to other host defense cytokines. Proc Natl Acad Sci USA 1987;84: lial cells: enhanced CXCL9 in autoimmune arthritis. Lab Invest
9233–7. 2006;86:902–16.
[94] Walz A, Peveri P, Aschauer H, Baggiolini M. Purification and amino [111] Kasama T, Strieter RM, Lukacs NW, Lincoln PM, Burdick MD,
acid sequencing of NAF, a novel neutrophil-activating factor pro- Kunkel SL. Interferon g modulates the expression of neutrophil-
duced by monocytes. Biochem Biophys Res Commun 1987;149: derived chemokines. J Invest Med 1995;43:58–67.
755–61. [112] Mueller SN, Hosiawa-Meagher KA, Konieczny BT, Sullivan BM,
[95] Schroder JM, Mrowietz U, Morita E, Christophers E. Purification and Bachmann MF, Locksley RM, et al. Regulation of homeostatic
partial biochemical characterization of a human monocyte-derived, chemokine expression and cell trafficking during immune responses.
neutrophil-activating peptide that lacks interleukin 1 activity. J Science 2007;317:670–4.
Immunol 1987;139:3474–83. [113] Kiderlen AF, Kaufmann SHE, Lohmann-Matthes ML. Protection
[96] Van Damme J, Van Beeumen J, Opdenakker G, Billiau A. A novel, of mice against the intracellular bacterium Listeria monocyto-
NH2-terminal sequence-characterized human monokine possessing genes by recombinant immune interferon. Eur J Immunol
neutrophil chemotactic, skin-reactive, and granulocytosis-promoting 1984;14:964–7.
activity. J Exp Med 1988;167:1364–76. [114] Nakane A, Minagawa T, Yasuda I, Chen Y, Kato K. Prevention by
[97] Larsen CG, Anderson AO, Oppenheim JJ, Matsushima K. Production gamma interferon of fatal infection with Listeria monocytogenes in
of interleukin-8 by human dermal fibroblasts and keratinocytes in mice treated with cyclosporin A. Infect Immun 1988;56:2011–5.
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 111

[115] Chen Y, Nakane A, Minagawa T. Recombinant murine gamma [134] Taborski U, Freitag W, Heremans H, Knop J. Inhibitory effects of
interferon induces enhanced resistance to Listeria monocytogenes interferon-g on the suppressor cell circuit in contact sensitivity.
infection in neonatal mice. Infect Immun 1989;57:2345–9. Immunobiol 1986;171:329–38.
[116] Buchmeier NA, Schreiber RD. Requirement of endogenous inter- [135] Skoglund C, Scheynius A, Holmdahl R, Van der Meide PH. Enhance-
feron-gamma production for resolution of Listeria monocytogenes ment of DTH reactions and inhibition of the expression of class II
infection. Proc Natl Acad Sci USA 1985;82:7404–8. transplantation antigens by in vivo treatment with antibodies against
[117] Nakane A, Minagawa T, Kohanawa M, Chen Y, Sato H, Moriyama M, g-interferon. Clin Exp Immunol 1988;71:428–30.
et al. Interactions between endogenous gamma interferon and tumor [136] Skoglund C, Scheynius A. Effects of interferon-g treatment on the
necrosis factor in host resistance against primary and secondary cutaneous DTH reaction in rats. Arch Dermatol Res 1990;282:
Listeria monocytogenes infections. Infect Immun 1989;57:3331–7. 318–24.
[118] Yang J, Kawamura I, Mitsuyama M. Requirement of the initial [137] Gaspari AA, Jenkins MK, Katz SI. Class II MHC-bearing kerati-
production of gamma interferon in the generation of protective noytes induce antigen-specific unresponsiveness in hapten-specific
immunity of mice against Listeria monocytogenes. Infect Immun Th1 clones. J Immunol 1988;141:2216–20.
1997;65:72–7. [138] Saulnier M, Huang S, Aguet M, Ryffel B. Role of interferon-gamma
[119] Nathan CF, Kaplan G, Levis WR, Nusrat A, Witmer MD, Sherwin in contact hypersensitivity assessed in interferon-gamma receptor-
SA, et al. Local and systemic effects of intradermal recombinant deficient mice. Toxicology 1995;102:301–12.
interferon-g in patients with lepromatous leprosy. N Engl J Med [139] Tokuriki A, Seo N, Ito T, Kumakiri M, Takigawa M, Tokura Y.
1986;315:6–15. Dominant expression of CXCR3 is associated with induced expres-
[120] Cooper AM, Dalton DK, Stewart TA, Griffin JP, Russell DG, Orme sion of IP-10 at hapten-challenged sites of murine contact hyper-
IM. Disseminated Tuberculosis in interferon g gene-disrupted mice. J sensitivity: a possible role for interferon-gamma-producing CD8(+)
Exp Med 1993;178:2243–7. T cells in IP-10 expression. J Dermatol Sci 2002;28:234–41.
[121] Flynn JL, Chan J, Triebold KJ, Dalton DK, Stewart TA, Bloom BR. [140] Vocanson M, Hennino A, Cluzel-Tailhardat M, Saint-Mezard P,
An essential role for interferon g in resistance to Mycobacterium Benetiere J, Chavagnac C, et al. CD8+ T cells are effector cells
tuberculosis infection. J Exp Med 1993;178:2249–54. of contact dermatitis to common skin allergens in mice. J Invest
[122] Dalton DK, Pitts-Meek S, Keshav S, Figari IS, Bradley A, Stewart Dermatol 2006;126:815–20.
TA. Multiple defects of immune cell function in mice with disrupted [141] He D, Wu L, Kim HK, Li H, Elmets CA, Xu H. CD8+ IL-17-
interferon-gamma genes. Science 1993;259:1739–42. producing T cells are important in effector functions for the elicita-
[123] Kamijo R, Le J, Shapiro D, Havell EA, Huang S, Aguet M, et al. Mice tion of contact hypersensitivity responses. J Immunol 2006;177:
that lack the interferon-g receptor have profoundly altered responses 6852–8.
to infection with Bacillus Calmette-Guérin and subsequent challenge [142] Dubois B, Chapat L, Goubier A, Papiernik M, Nicolas JF, Kaiserlian
with lipopolysaccharide. J Exp Med 1993;178:1435–40. D. Innate CD4+CD25+ regulatory T cells are required for oral
[124] Flesch IE, Hess JH, Huang S, Aguet M, Rothe J, Bluethmann H, et al. tolerance and inhibition of CD8+ T cells mediating skin inflamma-
Early interleukin 12 production by macrophages in response to tion. Blood 2003;102:3295–301.
mycobacterial infection depends on interferon g and tumor necrosis [143] Kish DD, Gorbachev AV, Fairchild RL. CD8+ T cells produce IL-2,
factor a. J Exp Med 1995;181:1615–21. which is required for CD(4+)CD25+ T cell regulation of effector
[125] Murray PJ, Young RA, Daley GQ. Hematopoietic remodelling in CD8+ T cell development for contact hypersensitivity responses. J
interferon-g-deficient mice infected with mycobacteria. Blood Leukoc Biol 2005;78:725–35.
1998;91:2914–24. [144] Chen JP, Liao NS, Lai SL, Hsu L, Mao WY, Ku MC, et al. Reduced
[126] Dalton DK, Haynes L, Chu C-Q, Swain SL, Wittmer S. Interferon g 2,4-dinitro-1-fluorobenzene-induced contact hypersensitivity
eliminates responding CD4 T cells during mycobacterial infection by response in IL-15 receptor alpha-deficient mice correlates with
inducing apoptosis of activated CD4 T cells. J Exp Med 2000;192: diminished CCL5/RANTES and CXCL10/IP-10 expression. Eur J
117–22. Immunol 2005;35:690–8.
[127] Tomioka H, Sato K, Maw WW, Saito H. The role of tumor necrosis [145] Brozna JP. Shwartzman reaction. Semin Thromb Haemostasis
factor, interferon-G, transforming growth factor b, and nitric oxide in 1990;16:326–32.
the expression of immunosuppressive functions of splenic macro- [146] Movat HZ, Burrowes CE, Cybulsky MI, Dinarello CA. Role of
phages induced by Mycobacterium avium complex infection. J complement, interleukin-1 and tumor necrosis factor in a local
Leukoc Biol 1995;58:704–12. Shwartzman-like reaction. In: Movat HZ, editor. Leukocyte emigra-
[128] De Maeyer EJ, De Maeyer-Guignard J, Vandeputte M. Inhibition by tion and its sequelae. Satellite Symp. 6th Int. Cong. Immunology.
interferon of delayed-type hypersensitivity in the mouse. Proc Natl Basel: Karger; 1987. p. 69–78.
Acad Sci USA 1975;72:1753–7. [147] Heremans H, Billiau A, Coutelier JP, De Somer P. The inhibition of
[129] De Maeyer-Guignard J, Cachard A, De Maeyer E. Delayed-type endotoxin-induced local inflammation by LDH virus or LDH virus-
hypersensitivity to sheep red blood cells: inhibition of sensitization infected tumors is mediated by interferon. Proc Soc Exp Biol Med
by interferon. Science 1975;190:574–6. 1987;185:6–15.
[130] Diamantstein T, Eckert R, Volk H-D, Kupier-Weglinski J-W. Rever- [148] Billiau A, Heremans H, Vandekerckhove F, Dillen C. Anti-inter-
sal by interferon-g of inhibition of delayed-type hypersensitivity feron-g antibody protects mice against the generalized Shwartzman
induction by anti-CD4 or anti-interleukin 2 receptor (CD25) mono- reaction. Eur J Immunol 1987;17:1851–4.
clonal antibodies. Evidence for the physiological role of the CD4+ [149] Heremans H, Van Damme J, Dillen C, Dijkmans R, Billiau A. Inter-
TH1+ subset in mice. Eur J Immunol 1988;18:2101–3. feron-g, a mediator of lethal lipopolysaccharide-induced Shwartzman-
[131] Fong TA, Mosmann TR. The role of IFN-g in delayed-type like shock reactions in mice. J Exp Med 1990;171:1853–69.
hypersensitivity mediated by Th1 clones. J Immunol 1989;143: [150] Heinzel FP. The role of IFN-g in the pathology of experimental
2887–93. endotoxemia. J Immunol 1990;145:2920–4.
[132] Issekutz TB, Stoltz JM, Van der Meide P. Lymphocyte recruitment in [151] Doherty GM, Lange JR, Langstein HN, Alexander RH, Buresh CM,
delayed-type hypersensitivity. The role of IFN-g. J Immunol Norton JA. Evidence for IFN-g as a mediator of the lethality of
1988;140:2889–93. endotoxin and tumor-necrosis factor-a. J Immunol 1992;149:
[133] Geiger K, Sarvetnick N. Local production of IFN-g abrogates the 1666–70.
intraocular immune privilege in transgenic mice and prevents the [152] Silva AT, Cohen J. Role of interferon-gamma in experimental gram-
induction of ACAID. J Immunol 1994;153:5239–46. negative sepsis. J Infect Dis 1992;166:331–5.
112 A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113

[153] Smith SR, Calzetta A, Bankowski J, Kenworthy-Bott L, Terminelli C. [172] Vermeire K, Heremans H, Vandeputte M, Huang J, Billiau A,
Lipopolysaccharide-induced cytokine production and mortality in Matthys P. Accelerated collagen-induced arthritis in interferon-g
mice treated with Corynebacterium parvum. J Leukoc Biol receptor-deficient mice. J Immunol 1997;158:5507–13.
1993;54:23–9. [173] Manoury-Schwartz B, Chiocchia G, Bessis N, Abehsira-Amar O,
[154] Kohler J, Heumann D, Garotta G, LeRoy D, Bailat S, Barras C, et al. Batteux F, Muller S, et al. High susceptibility to collagen-induced
IFN-gamma involvement in the severity of gram-negative infections arthritis in mice lacking IFN-g receptors. J Immunol 1997;158:
in mice. J Immunol 1993;151:916–21. 5501–6.
[155] Car BD, Eng VM, Schnyder B, Ozmen L, Huang S, Gallay P, et al. [174] Guedez YB, Whittington KB, Clayton JL, Joosten LAB, van de Loo
Interferon g receptor deficient mice are resistant to endotoxic shock. J FAJ, van den Berg WB, et al. Genetic ablation of interferon-g up-
Exp Med 1994;179:1437–44. regulates interleukin-1b expression and enables the elicitation of
[156] Heremans H, Dillen C, Van Damme J, Billiau A. Essential role for collagen-induced arthritis in a nonsusceptible mouse strain. Arthritis
natural killer cells in the lethal lipopolysaccharide-induced Shwartz- Rheum 2001;44:2413–24.
man-like reaction in mice. Eur J Immunol 1994;24:1155–60. [175] Chu CQ, Song Z, Mayton L, Wu B, Wooley PH. IFNgamma deficient
[157] Ogasawara K, Takeda K, Hashimoto W, Satoh M, Okuyama R, Yanai C57BL/6 (H-2b) mice develop collagen induced arthritis with pre-
N, et al. Involvement of NK1+ T cells and their IFN-gamma production dominant usage of T cell receptor Vbeta6 and Vbeta8 in arthritic
in the generalized Shwartzman reaction. J Immunol 1998;160:3522–7. joints. Ann Rheum Dis 2003;62:983–90.
[158] Dieli F, Sireci G, Russo D, Taniguchi M, Ivanyi J, Fernandez C, et al. [176] Billiau A. Interferon-g in autoimmunity. Cytokine Growth Factor
Resistance of natural killer T cell-deficient mice to systemic Rev 1996;7:25–34.
Shwartzman reaction. J Exp Med 2000;192:1645–52. [177] Heremans H, Billiau A. The effects of interferons and other cytokines
[159] Skoskiewicz MJ, Colvin RB, Schneeberger EE, Russell PS. Wide- on experimental autoimmune encephalomyelitis. In: Reder AT,
spread and selective induction of major histocompatibility complex- editor. Interferon therapy of multiple sclerosis. New York: Marcel
determined antigens in vivo by gamma interferon. J Exp Med Dekker, Inc.; 1997. p. 215–44.
1985;162:1645–64. [178] Billiau A, Heremans H, Matthys P. The role of interferon-g in
[160] Momburg F, Koch N, Moller P, Moldenhauer G, Butcher GW, experimental autoimmune disease. In: Bona CA, Revillard J-P,
Hammerling GJ. Differential expression of Ia and Ia-associated editors. Cytokines and cytokine receptors. Physiology and patho-
invariant chain in mouse tissues after in vivo treatment with IFN- logical disorders. Harwood Academic Publishers; 2000 . p. 300–
gamma. J Immunol 1986;136:940–8. 15.
[161] Sarvetnick N, Liggitt D, Pitts SL, Hansen SE, Stewart TA. Insulin- [179] Heremans H, Billiau A. The role of interferons in experimental
dependent-diabetes mellitus induced in transgenic mice by ectopic autoimmune encephalomyelitis. In: Lavi E, Constantinescu CC,
expression of Class II MHC and interferon-gamma. Cell editors. Experimental models of multiple sclerosis. New York:
1988;52:773–82. Springer; 2005. p. 314–42.
[162] Campbell IL, Kay TWH, Oxbrow L, Harrison LC. Essential role for [180] Kelchtermans H, Billiau A, Matthys P. How interferon-g keeps
interferon-g and interleukin-6 in autoimmune insulin-dependent autoinflammatory diseases in check. Trends Immunol 2008;29:
diabetes in NOD/Wehi mice. J Clin Invest 1991;87:739–42. 479–86.
[163] Debraye-Sachs M, Carnaud C, Boitard C, Cohen H, Gresser I, [181] Park H, Li Z, Yang XO, Chang SH, Nurieva R, Wang YH, et al. A
Bedossa P, et al. Prevention of diabetes in NOD mice treated with distinct lineage of CD4 T cells regulates tissue inflammation by
antibody to murine IFNg. J Autoimmunity 1991;4:237–48. producing interleukin 17. Nat Immunol 2005;6:1133–41.
[164] Jacob CO, Van der Meide PH, McDevitt HO. In vivo treatment of [182] Cua DJ, Sherlock J, Chen Y, Murphy CA, Joyce B, Seymour B, et al.
(NZBxNZW)F1 mice with monoclonal antibody to g interferon. J Interleukin-23 rather than interleukin-12 is the critical cytokine
Exp Med 1987;166:798–803. for autoimmune inflammation of the brain. Nature 2003;421:
[165] Hartung H-P, Schäfer B, Van der Meide PH, Fierz W, Heininger K, 744–8.
Toyka KV. The role of interferon-g in the pathogenesis of experi- [183] Murphy CA, Langrish CL, Chen Y, Blumenschein W, McClanahan T,
mental autoimmune disease of the peripheral nervous system. Ann Kastelein RA, et al. Divergent pro- and antiinflammatory roles for IL-
Neurol 1990;27:247–57. 23 and IL-12 in joint autoimmune inflammation. J Exp Med
[166] Mosmann TR, Coffman RL. TH1 and TH2 cells: different patterns of 2003;198:1951–7.
lymphokine secretion lead to different functional properties. Annu [184] Cruz A, Khader SA, Torrado E, Fraga A, Pearl JE, Pedrosa J, et al.
Rev Immunol 1989;7:145–73. Cutting edge: IFN-gamma regulates the induction and expansion of
[167] Billiau A, Heremans H, Vandekerckhove F, Dijkmans R, Sobis H, IL-17-producing CD4 T cells during mycobacterial infection. J
Meulepas E, et al. Enhancement of experimental allergic encepha- Immunol 2006;177:1416–20.
lomyelitis in mice by antibodies against IFN-g. J Immunol [185] Wang Z, Hong J, Sun W, Xu G, Li N, Chen X, et al. Role of IFN-
1988;140:1506–10. gamma in induction of Foxp3 and conversion of CD4+ CD25 T
[168] Duong TT, St.Louis J, Gilbert JJ, Finkelman FD, Strejan GH. Effect cells to CD4+ Tregs. J Clin Invest 2006;116:2434–41.
of anti-interferon-g and anti-interleukin-2 monoclonal antibody [186] Wood KJ, Sawitzki B. Interferon gamma: a crucial role in the
treatment on the development of actively and passively induced function of induced regulatory T cells in vivo. Trends Immunol
experimental allergic encephalomyelitis in the SJL/J mouse. J Neu- 2006;27:183–7.
roimmunol 1992;36:105–15. [187] Kelchtermans H, De Klerck B, Mitera T, Van Balen M, Bullens D,
[169] Duong TT, Finkelman FD, Singh B, Strejan GH. Effect of anti- Billiau A, et al. Defective CD4+ CD25+ regulatory T cell functioning
interferon-g monoclonal antibody treatment on the development of in collagen-induced arthritis: an important factor in pathogenesis,
experimental allergic encephalomyelitis in resistant mouse strains. J counter-regulated by endogenous IFN-g. Arthritis Res Ther
Neuroimmunol 1994;53:101–7. 2005;7:402–15.
[170] Willenborg DO, Fordham S, Bernard CCA, Cowden WB, Ramshaw [188] Hwu P, Du MX, Lapointe R, Do M, Taylor MW, Young HA.
IA. IFN-g plays a critical down-regulatory role in the induction and Indoleamine 2,3-dioxygenase production by human dendritic cells
effector phase of myelin oligodendrocyte glycoprotein-induced auto- results in the inhibition of T cell proliferation. J Immunol
immune encephalomyelitis. J Immunol 1996;157:3223–7. 2000;164:3596–9.
[171] Caspi RR, Chan C-C, Grubbs BG, Silver PB, Wiggert B, Parsa CF, [189] Munn DH, Sharma MD, Lee JR, Jhaver KG, Johnson TS, Keskin DB,
et al. Endogenous systemic IFN-g has a protective role against ocular et al. Potential regulatory function of human dendritic cells expres-
autoimmunity in mice. J Immunol 1994;152:890–9. sing indoleamine 2,3-dioxygenase. Science 2002;297:1867–70.
A. Billiau, P. Matthys / Cytokine & Growth Factor Reviews 20 (2009) 97–113 113

[190] Grohman U, Fallarino F, Puccetti P. Tolerance, DCs and tryptophan: tolerance causes various autoimmune diseases. J Immunol 1995;155:
much ado about IDO. Trends Immunol 2003;24:242–8. 1151–64.
[191] Puccetti P. On watching the watchers: IDO and type I/II IFN. Eur J
Immunol 2007;37:876–9. Alfons Billiau (January 20th 1937) is emeritus
[192] Yamazaki T, Akiba H, Iwai H, Matsuda H, Aoki M, Tanno Y, et al. professor of microbiology and immunology at
Expression of programmed death 1 ligands by murine T cells and the University of Leuven (Katholieke Universi-
APC. J Immunol 2002;169:5538–45. teit Leuven, KU Leuven). He formerly headed
[193] Matthys P, Vermeire K, Mitera T, Heremans H, Huang J, Schols D, the Laboratory of Immunobiology at the Rega
et al. Enhanced autoimmune arthritis in IFN-g receptor-deficient Institute for Medical Research, affiliated to this
mice is conditioned by mycobacteria in Freund’s adjuvant and by University. This laboratory has a long-standing
increased expansion of Mac-1+ myeloid cells. J Immunol 1999;163: tradition of pioneering work in the field of inter-
3503–10. ferons, interleukins and other cytokines, as well
[194] Matthys P, Vermeire K, Heremans H, Billiau A. The protective effect as chemokines and matrix metalloproteinases. In
of IFN-g in experimental autoimmune diseases: a central role of the 1970s Billiau was particularly involved in pre-industrial production of
mycobacterial adjuvant-induced myelopoiesis. J Leukoc Biol natural human fibroblast interferon (Interferon-b). Expertise of his labora-
2000;68:447–54. tory in this domain seamlessly led to the isolation and characterization, in
[195] Matthys P, Vermeire K, Billiau A. Mac-1+ myelopoiesis induced by the 1980s, of interleukin-1 as an interferon-b-inducing cytokine and of
complete Freund’s adjuvant (CFA): a clue to the paradoxical effects interleukin-6 as a hybridoma growth factor co-produced with interferon-b.
of IFN-g in autoimmune disease models. Trends Immunol Billiau’s name is also connected with studies on the peculiar mode of action
2001;22:367–71. of interferon against retroviruses. More recent work under his leadership
[196] Geboes L, De Klerck B, Van Balen M, Kelchtermans H, Mitera T, bears on the role of interferon-g in animal models of inflammation and
Boon L, et al. Freund’s complete adjuvant induces arthritis in mice autoimmune disease, in particular experimental autoimmune encephalo-
lacking a functional interferon-gamma receptor by triggering tumor myelitis and collagen-induced arthritis.
necrosis factor alpha-driven osteoclastogenesis. Arthritis Rheum
2007;56:2595–607. Patrick Matthys (September 19th 1963) is Pro-
[197] Ruddle NH, Waksman BH. Cytotoxicity mediated by soluble antigen fessor in immunology at the University of Leu-
and lymphocytes in delayed hypersensitivity. I. Characterization of ven and a research director at the Rega Institute.
the phenomenon. J Exp Med 1968;128:1237–54. He earned his PhD in 1993 with experimental
[198] Sonnenfeld G, Mandel AD, Merigan TC. The immunosuppessive work demonstrating the key role of IFN-g in
effect of Type II mouse interferon on antibody production. Cell eliciting tumor-associated cachexia, a wasting
Immunol 1977;34:193–206. syndrome that had until then been attributed
[199] Pfefferkorn ER. Interferon g blocks the growth of Toxoplasma gondii mainly to the production of tumor necrosis fac-
in human fibroblasts by inducing the host cells to degrade tryptophan. tor/cachectin. As a postdoctoral fellow of the
Proc Natl Acad Sci USA 1984;81:908–12. Fund for Scientific Research Flanders, Matthys
[200] Takikawa O, Habara-Ohkubo A, Yoshida R. IFN-g is the inducer of continues conducting research at the Rega Institute, focusing on the
indoleamine 2,3-dioxygenase in allografted tumor cells undergoing pathogenesis of autoimmune diseases, in particular autoimmune arthritis
rejection. J Immunol 1990;145:1246–50. and related pathology. The complex role of IFN-g in pathogenesis remains a
[201] Sakaguchi S, Sakaguchi N, Asano M, Itoh M, Toda M. Immunologic preferred subject of his investigations, however also with in-depth analysis
self-tolerance maintained by activated T cells expressing IL-2 recep- of interactions with other cytokines, chemokines and other inflammatory
tor a-chains (CD25). Breakdown of a single mechanism of self- mediators.

You might also like