The Th17 plasticity demonstrated in
these two reports comes from cells
generated in canonical culture conditions
in vitro during which they were constantly
exposed to instructive cytokine signals.
In vivo, cells likely receive additional
signals both favoring and opposing Th17
differentiation, and the nature and relative
abundance ofthesesignals, isinconstant.
Thus, an important question is whether
Th17 cells generated in vivo exhibit the
same instability. The evidence at present
isboth limited and conflicting. Bycontrast
to Lee et al. (2009) who purified Th17 cells
based on their expression of an IL-17F-
fied cells from unmanipulated mice based
on their ability to secrete IL-17A directly
ex vivo and found that the fraction of
cells producing IL-17A alone or along
with IFN-g was relatively stable in the
absence of added cytokines for 6 days.
Similarly, when human Th17 cells were
purified from blood based on their ability
to secrete IL-17A directly ex vivo (Streeck
declined and IFN-g increased by 6 weeks
but were considerably more stable than
observed by Lee et al. (2009) whereas An-
nunziato et al. (2007) found that human
Th17 cells cloned directly ex vivo in non-
induced to produce substantial amounts
of IFN-g when activated in the presence
of IL-12. Thus, further study will be
required to reconcile these differences,
to determine whether cells that can
produce IL-17A directly ex vivo represent
a relatively more committed subset of
Th17 cells, and to explore the underlying
In any case, the two reports in this issue
of Immunity indicate that the ability of
CD4+T cell subsets to adopt alternate or
simply by the epigenetic states of cyto-
kine loci but by the epigenetic states of
the entire set of genes associated with
these lineages. Although the data of Wei
et al. (2009) characterized only two
histone modifications, their multilineage
comparison provides hints that global
mapping of histone modifications may
be predictive of lineages that are most
susceptible to deviation and toward
which lineages they are most likely to be
reprogrammed. In the near future, we
can hope to see ever more complete
epigenomic and gene expression profiles
in T helper cell lineages. These profiles
as carried out by Lee et al. (2009) should
help to unravel the many unanswered
question regarding T helper cell lineage
commitment, plasticity, and overlapping
Annunziato, F., Cosmi, F., Santarlasci, V., Maggi,
L., Liotta, F., Mazzinghi, B., Parente, E., Fili, L.,
Ferri, S., Frosali, F., et al. (2007). J. Exp. Med.
Lee, Y., Turner, H., Maynard, C.L., Oliver, J.R.,
Chen, D., Elson, C.O., and Weaver, C.T. (2009).
Immunity 30, this issue, 92–107.
Richter, A., Kamradt, T., Radbruch, A., and Chang,
H.-D. (2008). Eur. J. Immunol. 38, 2654–2664.
McGeachy, M.J., and Cua, D.J. (2008). Immunity
Streeck, H., Cohen, K.W., Jolin, J.S., Brockman,
M.A., Meier, A., Power, K.A., Waring, M.T., Alter,
G., and Altfeld, M.J. (2008). J. Immunol. Methods
Wan, Y.Y., and Flavell, R.A. (2007). Nature 445,
Wei, G., Wei, L., Zhu, J., Zhang, C., Hu-Li, J., Yao,
Z., Cui, K., Kanno, Y., Roh, T.-Y., Watford, W.T.,
et al. (2009). Immunity 30, this issue, 155–167.
Xu, L., Kitani, A., Fuss, I., and Strober, W. (2007).
J. Immunol. 178, 6725–6729.
Yang, X.O., Nurieva, R., Martinez, G.J., Kang, H.S.,
Chung, Y., Pappu, B.P., Shah, B., Chang, S.H.,
Schluns, K.S., and Watowich, S.S. (2008). Immu-
nity 29, 44–56.
Interleukin-17A and Interleukin-17F:
A Tale of Two Cytokines
Patricia J. Dubin1and Jay K. Kolls2,*
1Children’s Hospital of Pittsburgh, Rangos Research Center, 530 45th Street, Pittsburgh, PA 15201, USA
2Department of Genetics, LSU Health Sciences Center, CSRB Room 657, New Orleans, LA 70112, USA
In this issue of Immunity, Ishigame et al. (2009) show that interleukin-17A (IL-17A) mediates autoimmunity
whereas both IL-17A and IL-17F are required for mucosal immunity. IL-17A may be more pathologic by
inducing proinflammatory cytokines.
Charles Dickens’s epic tale of Paris and
London begins with ‘‘It was the best of
times, it was the worst of times,’’ empha-
sizing the recent advances in wisdom that
coincided with a time of great ignorance.
In the past 10 years, our understanding of
interleukin-17 (IL-17) in mucosal immunity
and autoimmunity has greatly expanded.
IL-17A, the founding member of the
IL-17 family of cytokines, is produced by
a subset of CD4+T cells termed Th17
cells. IL-17A is potent inducer of antimi-
crobial peptides as well as neutrophil
colony-stimulating factor (G-CSF) and it
Immunity 30, January 16, 2009 ª2009 Elsevier Inc.
also mediates host resistance to extracel-
lular bacterial and fungal infections. An
unintended consequence of this arm of
the immune system is autoimmunity. To
this end, Th17 cells have been shown to
be critical mediators of many autoim-
mune disorders. However, the individual
contributions of the two most closely
related IL-17 family members, IL-17A
and IL-17F (which share 55% homology
at the amino acid level), and the role of
their receptors in these processes has
been an area of needed enlightenment.
The paper by Ishigame et al. (2009) in
this issue ofImmunity
expands our wisdom of the individual
contributions of the ligands IL-17A and
IL-17F (and their receptors IL-17RA and
IL-17RC) in mucosal immunity and auto-
immune tissue inflammation.
The authors show that in adjuvant
models of autoimmunity (including experi-
arthritis), nearly all of the CD4+T cells
that produce IL-17A also produce IL-17F.
Despite the large number of double-
positive cells in these models, IL-17F is
mice show reduced disease (Ishigame
et al., 2009). To further study the role of
IL-17F, the investigators also generated
mice deficient in both IL-17A and IL-17F,
and these mice showed no additional
protection in experimental autoimmune
encephalomyelitis (EAE) or arthritis (the
latter model was in the absence of IL-1R
signaling), compared to mice deficient in
IL-17A only. These data show that the
absence of IL-17F is neither protective
nor harmful in these autoimmune models.
The predominant role
has been observed in another Il17f
line (Yang et al., 2008). The latter issue is
important because it has been shown
that IL-17A and IL-17F can form hetero-
dimers that have intermediate bioactivity
in vitro (Wright et al., 2007), and it could
have been possible that the absence of
IL-17F could lead to increased IL-17A
homodimer formation (which could have
exacerbated disease), but this does not
appear to be the case. Furthermore, Ishi-
game et al. (2009) found that IL-17A was
more critical than IL-17F for ovalubumin-
induced neutrophilic airway inflammation.
However, the prior study by Yang et al.
(2008) showed that IL-17F was more crit-
ical than IL-17A in inducing airway neutro-
philic inflammation to Aspergillus oryzae.
The differences in these responses are
not clear but may depend on signaling
pathways driven by the different adjuvant
properties of these models.
Incontrast to these
inflammation models, IL-17F is required
for mucosal immunity to two extracellular
pathogens. Ishigame et al. (2009) show
that single-gene deletions of Il17a or
Il17f alone do not result in spontaneous
susceptibility to infection. However, dele-
tion of the locus encoding both Il17a and
Il17f (which are separated by approxi-
mately 50 kB on mouse chromosome 1)
leads to the spontaneous development
of perinasal and perioral abscesses colo-
nized by Staphylococcus aureus. This is
a phenotype that is nearly identical to
one reported for Il17ra?/?mice, which
also develop cutaneous S. aureus infec-
tion (Schwarzenberger and Kolls, 2002).
This phenotype was restricted to mucosal
infection, as shown by the fact that
Il17a?/?Il17f?/?mice were not more
susceptible to a systemic infection with
S. aureus. The spontaneous development
of cutaneous infections with S. aureus is
similar to that observed in humans with
Hyper IgE syndrome (HIES) who are defi-
cient in IL-17-producing T cells because
of mutations in Stat3 (Milner et al., 2008).
Although Stat3 controls many aspects of
cell development and cytokine signaling
(by both the IL-6 family of cytokines
as well as IL-22), the data published
in this issue of Immunity strongly impli-
cates IL-17A and IL-17F in mucosal
immunity against S. aureus in mice and
Th17 cytokines have also been impli-
cated in mucosal immunity in the gut
against the Gram-negative pathogen Cit-
Il17a?/?, Il17f?/?, and Il17a?/?Il17f?/?
double-deficient mice showed similar
susceptibility to infection compared to
wild-type mice (Ishigame et al., 2009).
Thus in this infection, IL-17F has an inde-
pendent role apart from IL-17A. However,
the deletion of both Il17s did not exacer-
bate this phenotype in terms of bacterial
load in the tissues. However, the loss of
Il17f singly or in combination with Il17a
resulted in greater hypertrophy of the
colon and tissue inflammation in the
C. rondentium infection model. Both Il17s
were required for the optimal expression
of the antimicrobial genes b-defensin 1,
3, 4 and this likely contributes to the
increased bacterial growth in this model.
Interestingly, the authors did not observe
differences in genes previously shown to
beregulated byIL-22 orIL-17in combina-
tion with IL-22 such as lipocalin-2 (Aujla
et al., 2008) or Reg3b and Reg3g (Zheng
et al., 2008). These data show that IL-
17F regulates a subset of antimicrobial
proteins that are distinct from IL-22.
The protection of Il17a?/?mice in EAE
as opposed to Il17f?/?mice has now
been observed in two independent lines
of Il17a?/?and Il17f?/?mice (Ishigame
et al., 2009; Yang et al., 2008). This raises
the following question: why is IL-17A
more pathologic than Il-17F even though
it has been shown in fibroblasts and
epithelial cells (Toy et al., 2006) that IL-
17A and IL-17F induce similar chemo-
kines and growth factors (McAllister
et al., 2005)? One model that has been
proposed is that IL-17A and IL-17F signal
through a heterodimeric receptor com-
posed of IL-17RA and IL-17RC (Toy et al.,
2006). In fibroblasts, both these receptor
chains are required for optimal CXC che-
mokine production to IL-17A and IL-17F.
However, IL-17A is approximately 10-fold
more potent than IL-17F in these chemo-
kine responses (Figure 1A). The biochem-
ical mechanisms of these differences in
potency are still poorly understood but
could explain in part why IL-17A is more
critical in autoimmune inflammation than
Ishigame et al. (2009) propose an alter-
native to this model that is related to the
differences in distribution of IL-17RA and
IL-17RC on myeloid versus nonmyeloid
cells, thus leading to different responses
to IL-17A and IL-17F in these cell
populations (Figure 1, right). Unlike IL-
17F, IL-17A was able to induce the cyto-
kines IL-1b, IL-9, IL-12p70, and GM-CSF
as well as more potently induce the che-
mokines CCL3 and CCL2 by peritoneal
macrophages (Figure 1, right). IL-17A
but not IL-17F also induced similar proin-
flammatory cytokines and chemokines by
CD4+T cells (Figure 1, right). These
altered responses to these ligands were
associated with restricted expression of
IL-17RA and IL-17RC.
observed that Thy1.2+, B220+, CD11c+,
and CD11b+cells expressed IL-17RA
and Act1 but not IL-17RC. IL-17RC
Immunity 30, January 16, 2009 ª2009 Elsevier Inc.
nonhematopoietic cells such as colonic
epithelium and (to a lesser extent) perito-
neal macrophages. This may explain in
part the difference in downstream targets
expressed by these cells to IL-17A and
IL-17F. This hypothesis would suggest
that perhaps there are different receptor
conformations within these subgroups of
cells. Alternatively, signal transduction
and adaptor proteins may also be differ-
entially expressed within these cell popu-
lations. Lubberts et al. (2005) has used
radiation chimeras to demonstrate that
the myeloid expression
appears to be dispensable for the devel-
definitive and thus conditional deletions
of IL-17RA and IL-17RC in specific cell
approach to address the specific roles of
these receptors on IL-17A and IL-17F
signaling in myeloid and epithelial cells.
In conclusion, Ishigame et al. (2009)
have demonstrated distinct functions for
IL-17A and IL-17F in mediating inflamma-
tion resulting from mucosal infection and
autoimmune processes. In doing so,
they have helped bring us closer to
answering a conundrum that has plagued
investigators since the initial description
of IL-17F: if IL-17A and IL-17F have such
similar structures and properties and
signal through the same receptors (IL-
17RA and IL-17RC), how do we explain
IL-17A and IL-17F’s distinct functions?
IL-17A and IL-17F clearly have distinct
bioactivities on myeloid versus nonmye-
loid cells and these divergent tissue
responses may explain the conservation
of these two closely related genes in
mammals. Moreover, this study along
with the prior study on Il17f?/?mice
(Yang et al., 2008) shows that IL-17A is
more critical than IL-17F in mediating
Aujla, S.J., Chan, Y.R., Zheng, M., Fei, M., Askew,
D.J., Pociask, D.A., Reinhart, T.A., McAllister, F.,
Edeal, J., Gaus, K., et al. (2008). Nat. Med. 14,
Ishigame, H., Kakuta, S., Nagai, T., Kadoki, M.,
Nambu, A., Komiyama, Y., Fujikado, N., Tanahashi,
Y., Akitsu, A., Kotaki, H., et al. (2009). Immunity 30,
this issue, 108–119.
Lubberts, E., Schwarzenberger, P., Huang, W.,
Schurr, J.R., Peschon, J.J., Van den Berg, W.B.,
McAllister, F., Henry, A., Kreindler, J.L., Dubin,
P.J., Ulrich, L., Steele, C., Finder, J.D., Pilewski,
J.M., Carreno, B.M., Goldman, S.J., et al. (2005).
J. Immunol. 175, 404–412.
Freeman, A.F., Hill, B.J., Elias, K.M., Kanno, Y.,
Spalding, C., Elloumi, H.Z., Paulson, M.L., et al.
(2008). Nature 452, 773–776.
Schwarzenberger, P., and Kolls, J.K. (2002).
J. Cell. Biochem. 38, 88–95.
Toy, D., Kugler, D., Wolfson, M., Vanden Bos, T.,
Gurgel, J., Derry, J., Tocker, J., and Peschon, J.
(2006). J. Immunol. 177, 36–39.
Wright, J.F., Guo, Y., Quazi, A., Luxenberg, D.P.,
Bennett, F., Ross, J.F., Qiu, Y., Whitters, M.J.,
et al. (2007). J. Biol. Chem. 282, 13447–13455.
Yang, X.O., Chang, S.H., Park, H., Nurieva, R.,
Shah, B., Acero, L., Wang, Y.H., Schluns, K.S.,
Broaddus, R.R., Zhu, Z., and Dong, C. (2008).
J. Exp. Med. 205, 1063–1075.
Zheng, Y., Valdez, P.A., Danilenko, D.M., Hu, Y.,
Sa, S.M., Gong, Q., Abbas, A.R., Modrusan, Z.,
Ghilardi, N., De Sauvage, F.J., and Ouyang, W.
(2008). Nat. Med. 14, 282–289.
CD4+ T cell
CCL2, CCL3, GM-CSF IL-1β, IL-9
G-CSF, IL-6, CXC chemokines
IL-6, CCL3, G-CSF,
CCL2, IL-1β, IL-12p70,
IL-9, GM-CSF, CXCL1
Figure 1. Differential Responses to IL-17A and IL-17F
Model 1: It has been previously shown that epithelial cells express both IL-17RA and IL-17RC and these cells produce G-CSF, IL-6, and CXC chemokines in
response to IL-17A and IL-17F, although IL-17A is approximately one log more potent than IL-17F in these assays. Thus, whether IL-17A and IL-17F have similar
roles in inducing similar genes and contributing to autoimmunity was unclear.
Model 2: In this issue of Immunity, Ishigame et al. (2009) show that IL-17A and IL-17F are both required for mucosal immunity against S. aureus or C. rodentium;
however, IL-17A is much morecriticalfor bothEAE and adjuvant-induced arthritis. Thismay beexplained by altered receptordistribution of IL-17RA and IL-17RC
as well as divergent gene expression programs within the myeloid cell compartment. IL-17A and IL-17F induce IL-6, CCL3, and G-CSF by peritoneal macro-
phages, but IL-17A also induced proinflammatory cytokines such as IL-1b, IL-12p70, and GM-CSF (shown in red). Furthermore, IL-17A and not IL-17F induces
CCL2, CCL3, GM-CSF, IL-1b, and IL-9 by CD4+T cells. The authors suggest that these divergent myeloid responses may explain why IL-17A is more critical in
autoimmunity than is IL-17F.
Immunity 30, January 16, 2009 ª2009 Elsevier Inc.