Invariant natural killer T cells and immunotherapy of cancer 183
factors which promote tumor growth and the induction of Studies combining intravenous (i.v.) free αGalCer treat-
immunosuppressive cells (reviewed in [2–4]). An immune ment with protein vaccination demonstrated that αGalCer
response mediated by effector cells without the develop- acts as an adjuvant for the induction of antigen specific CD4+
ment of a state of chronic inflammation is vital for a and CD8+ T cell mediated immunity [26,27]. Furthermore,
successful anti-tumor immune response. Here we provide i.v. injection of dying hematologic tumor cells together with
evidence that a relatively recently discovered T cell subset αGalCer induced long lasting protective immunity, depend-
with immune controlling capacity holds promise in this ing on conventional CD4+ and CD8+ T cells. In this model
respect. αGalCer-activated iNKT cells enhanced the maturation of
Invariant CD1d restricted natural killer T (iNKT) cells can DC, that subsequently were more efficient in (cross-)
express natural killer (NK) receptors and express a canonical priming CD4+ and CD8+ T cells .
T cell receptor (TCR)-Vα-chain (Vα24.Jα18 in humans, These intriguing findings from pre-clinical studies
preferentially paired with Vβ11; Vα14.Jα18 in mice, paired prompted several groups to study iNKT cell numbers and
with Vβ2, Vβ7 or Vβ8.2) recognizing glycolipid antigens function in cancer patients and to perform clinical phase I
presented by the monomorphic CD1d molecule [5,6]. The studies in these patients to modulate the iNKT cell popula-
glycolipid α-galactosylceramide (αGalCer) was originally tion through administration of αGalCer, αGalCer-pulsed DC
isolated from the marine sponge Agelas mauritianus. or iNKT cell-enriched autologous peripheral blood mono-
Synthetically produced αGalCer, also known as KRN7000 nuclear cells (PBMC). This review will translate these studies
has been used in most pre-clinical and clinical studies to possible future strategies of iNKT cell mediated immu-
performed so-far and it has been shown to induce prolifera- notherapy. In the first part we will provide evidence for the
tion of, and cytokine production by, iNKT cells [5,7]. iNKT relevance of iNKT cells in human cancer. In the second part
cells are activated by microbial pathogen-derived glycolipids we will discuss several clinical phase I studies on iNKT cell
presented by CD1d on dendritic cells (DC) [8–11]. Activation activation and, in particular, we will illustrate the ther-
by endogenous antigens such as the (currently disputed apeutic potential of autologous adoptive transfer of purified
[12,13]) lysosomal glycosphingolipid isoglobotrihexosylcera- iNKT cell lines in patients that are severely deficient in iNKT
mide (iGb3), is amplified by DC activation through bacterial- cells.
derived toll-like receptor (TLR) ligands such as LPS [11,14].
These findings underscore the role iNKT cells play in the
Part I: Relevance of invariant Natural Killer
immune response against microorganisms. On the other hand
iNKT cells can protect against autoimmunity most probably T lymphocytes in human cancer
due to secretion of anti-inflammatory cytokines (reviewed in
). The response towards microorganisms may very well Selective decrease of iNKT cell numbers in
be linked to this protection since immune responses against peripheral blood of carcinoma patients
infections may be accompanied by anti-self responses
causing autoimmunity (reviewed in ). As listed in Table 1, Kawano et al. were the first to report a
Human and mouse iNKT cells can either be CD4+ or numeric defect in circulating iNKT cells in cancer . These
CD4−CD8− [double negative (DN)] and in humans a small findings were confirmed by some [30–32] and contradicted
proportion can express CD8. Direct ex vivo analyses by others [33–36].
suggested that CD4+ iNKT cells produce both Th1 cytokines Most of these studies however, investigated small
(e.g. GM-CSF, IFN-γ and TNF-α) and Th2 cytokines (e.g. IL-4 cohorts and the effects of age and gender, as reported by
and IL-13), whereas the DN and CD8+ iNKT cell subsets Delarosa et al.  and Sandberg et al. , were not
primarily produce Th1 cytokines [17–20]. While the always taken into account. We therefore studied circulat-
occurrence of different iNKT cell subsets may be the ing iNKT cell levels in a cohort of 120 patients with various
most likely explanation for their dichotomous regulatory epithelial cancers [melanoma, breast-, colorectal-, renal
nature, conclusive evidence for this distinction is lacking to cell-cancer and head and neck squamous cell carcinoma
date. (HNSCC)] and 69 healthy controls using multivariate
Apart from protection against microorganisms and auto- analysis and confirmed that after correction for the strong
immunity, an important physiological role for iNKT cells was influence of both age and gender, cancer patients had a
shown in the immuno-surveillance of cancers. Mice deficient selective numeric iNKT cell deficiency within the circulat-
in iNKT cells (Jα18−/− mice) were more susceptible to ing T cell pool (average 47% reduction compared to healthy
chemically [methylcholanthrene (MCA)] induced sarcomas, controls; p = 0.013, linear regression analysis) . Crough
while protection could be restored by adoptive transfer of et al. reported similar findings and observed that while
iNKT cells isolated from wild-type animals. Protection chemotherapy did not affect iNKT cell numbers, radio-
depended on CD1d, IFN-γ production by iNKT cells, and NK therapy induced a decline of circulating iNKT cells in
and CD8 Tcell function . The capacity of in vivo activated melanoma patients . We observed a mild and transient
iNKT cells to enhance protection against experimental effect of radiotherapy on iNKT circulating cell levels in
tumors has been studied extensively. Depending on the patients with HNSCC (Molling JW, unpublished results).
model, resident iNKT cells can augment innate as well as iNKT cell numbers were not influenced by tumor type or
adaptive anti-tumor immune responses. Several groups disease stage [31,33,39], nor were iNKT cells restored after
demonstrated that systemic injection of αGalCer or αGal- tumor de-bulking by surgery or radiotherapy . Taken
Cer-loaded DC activates iNKT cells, leading to the inhibition together these findings indicate that circulating iNKT cell
of metastasis formation predominantly via the downstream levels are at most only mildly affected by chemotherapy,
activation of NK cells [22–25]. radiotherapy or by the outgrowing tumor.
184 J.W. Molling et al.
Table 1 Immunologic studies on iNKT cells in cancer patients
n Peripheral Related to iNKT cells Relation to Peripheral Proliferative Cytotoxic
blood disease in tumor clinical blood iNKT response to against
iNKT levels stage infiltrate outcome IFN-γ αGalCer tumorXI
Melanoma 13 ↓ ? ? ? ? ↔ yes
(Kawano et al.) 
Prostate cancer 6 ↓ ? ? ? ↓VIII ↓X ?
(Tahir et al.) 
Lung cancer 60 ↓ no yes ? ↔IX ? ?
(Motohashi et al.) 
Lung cancer 55 ↔ no ? ? ? ↓ ?
(Konishi et al.) 
Myeloma 23 ↔ n/a yes yesV ↓ ↔ yes
(Dhodapkar et al.) 
Glioma 9 ↔ n/a ? ? ↔ ↔ yes
(Dhodapkar et al.) 
Neuroblastoma 8 (blood) ↔ n/a yes yesVI ? ? ?
(Metelitsa et al.)  98 (tumor)
Various carcinomas 109 ↔I; ↓II n/aI; ?II n/aI; ?II n/aI; ?II n/aI; ?II n/aI; ?II n/aI; ?II
(Crough et al.) 
Various carcinomas 120 ↓ noIII ? ? ↔ ? ?
(Molling et al.) 
Various carcinomas 21 ↓ n/a ? ? ? ↓X ?
(Yanagisawa et al.) 
Head and neck squamous 47 ↓III ? ? yesVII ? ? ?
(Molling et al.) 
Colorectal cancer 103 ? n/a yesIV yesVI ? ? ?
(Tachibana et al.) 
Overview of the results from various studies on the number and function of circulating invariant natural killer T (iNKT) cells in cancer
patients, their capacity to infiltrate tumors, or their relation to prognosis. I: Colorectal, renal cell and other cancers; II: Breast cancer
and melanoma; III: Not restored upon tumor de-bulking (18 weeks); IV: Vα24+ T cells, assumed to be invariant natural killer T (iNKT) cells;
V: IFN-γ secretion by iNKT in peripheral blood or tumor bed; VI: Size of tumor infiltrating iNKT cell pool; VII: Size of peripheral blood iNKT
cell pool; VIII: After in vitro expansion using α-Galactosylceramide (αGalCer) loaded autologous peripheral blood mononuclear cells (PBMC)
as antigen presenting cells (APC); IX: Detection of messenger RNA, not protein; X: Autologous αGalCer-loaded PBMC used as APC, not
monocyte derived dendritic cells; XI: Upon in vitro expansion; ↓, ↔: reduced, unaffected compared to healthy controls respectively; ?:
unknown (analysis was not performed); n/a: not applicable (no reduction in iNKT cells).
Some controversy exists with regard to the capacity of the broad variety of cancers [43–46] and that they can suppress
residual iNKT cells of cancer patients to respond to αGalCer anti-tumor responses (reviewed in ). An additional factor
in vitro (Table 1). It is noteworthy, in this respect, that the of influence might be the limited presence and poor function
decline in ex vivo IFN-γ secretion  and expansion  of of DC in the circulation of cancer patients [48–50]. The use of
iNKT cells in response to αGalCer were also reported to be autologous PBMC as APC might thus induce in vitro iNKT cell
age related. Unfortunately, information regarding subjects' anergy. In line with this, Tahir et al. demonstrated that IFN-γ
age is not always provided. secretion by iNKT cells from prostate cancer patients was
Substantial in vitro proliferation of iNKT cells from cancer restored when recombinant IL-12 was added to iNKTcell/APC
patients was achieved in studies using αGalCer-pulsed co-cultures . Actually, in all reported studies, iNKT cells
monocytes or immature monocyte derived DC (moDC) as from cancer patients had retained their capacity to either
antigen presenting cells (APC) [29,35]. In contrast, iNKT cell produce IFN-γ ex vivo upon αGalCer stimulation, or after
expansion using αGalCer-pulsed autologous PBMC did not repeated or highly active stimulation by professional
lead to adequate expansion of iNKT cells from carcinoma antigen presenting cells [30,31,34,35,39]. Furthermore,
patients [30,32,33]. The latter might be explained by the expanded iNKT cells of cancer patients can display in
presence of other, suppressive, T cells in peripheral blood of vitro cytotoxicity against various CD1d expressing tumor
these patients, as proposed by Yanagisawa et al. . targets (Table 1) [29,32,34,35]. This is in line with our
Naturally occurring regulatory T cells (nTreg) have been observations that in vitro cultured iNKT cells of healthy
demonstrated to be capable of directly inhibiting iNKT cell donors contain granules consisting of cytotoxic effector
proliferation, cytokine secretion and cytotoxic activity via molecules such as perforin and granzyme B [51–53].
cognate interactions . Furthermore, it has been well However, cultured iNKT cells do not kill very efficiently
established that circulating nTreg numbers are enhanced in a and often require loading of CD1d+ tumor cells with
Invariant natural killer T cells and immunotherapy of cancer 185
αGalCer [54–56]. Furthermore, the ability to directly kill peripheral blood were also reduced in this cohort as
tumor cells has not been established to be a physiological compared to historical age and gender matched controls.
function of circulating iNKT cells . However, no evidence was obtained for a relation of
In summary, although some discrepancies have been peripheral blood T or NK cell levels and clinical outcome.
noted regarding iNKT cell number and function in different This was in line with previous reports on HNSCC [61–63].
cancers, evidence is accumulating that circulating iNKT cells Reduced functionality, rather than reduced numbers of
of carcinoma patients are substantially reduced compared to circulating or tumor infiltrating T and/or NK cells was related
healthy controls. Notwithstanding, residual iNKT cells in to a poor prognosis in HNSCC in additional studies [64–68].
these patients still possess the capacity to proliferate, to Interestingly, Reichert et al. demonstrated a strong correla-
secrete IFN-γ and to gain some cytotoxic activity when tion between a loss of function in peripheral T cells and a loss
properly stimulated in vitro. This suggests that the residual of function in tumor infiltrating lymphocytes (TILs) in HNSCC
iNKT cells of carcinoma patients might still be capable of patients . A similar observation was made in myeloma,
taking part in patho-physiological anti-tumor responses. where a marked defect in IFN-γ secretion by iNKT cells in the
Therapies aimed at their increase and activation in cancer circulation as well as in the tumor bed of progressive, but not
patients could thus facilitate more potent anti-tumor of non-progressive myeloma patients was found .
immunity. These two aspects will be discussed further in As mentioned above, we found evidence that peripheral
the remaining sections of this review. blood iNKT cells of HNSCC patients can still secrete IFN-γ.
Taken together these findings [34,39,60,67] suggest that the
level of circulating iNKT cells is indicative of their relative
Peripheral blood iNKT cells in relation to clinical contribution to local anti-tumor immune responses. It would
outcome of cancer thus be relevant to study the correlation between peripheral
and intra-tumor iNKT cell number (and function) in future
The reduction in circulating iNKT cells of carcinoma patients cohorts, and to relate these parameters to (local) T and NK
suggests their contribution to anti-tumor immune responses. cell function. Furthermore, prospective studies are war-
However, it has not been demonstrated in the studies re- ranted to provide more insight in how findings of such studies
viewed above, whether a reduction of circulating iNKT cells would relate to patient survival.
can be regarded as a risk factor for carcinoma development.
This also remains to be elucidated for an elevated level of
circulating nTregs, which may inhibit iNKT cell number and Tumor infiltrating iNKT cells in relation to clinical
function . We therefore studied the relation between outcome of cancer
peripheral blood iNKT cell or nTreg frequencies and the
natural course of pre-invasive cervical intraepithelial neopla- To make a significant contribution to local anti-tumor
sia (CIN), in a prospective nonintervention cohort study of 82 responses, iNKT cells would most likely have to be located
women with abnormal cervix cytology . Persistent infec- in the tumor draining lymph nodes and/or in the tumor
tion with human papillomavirus type 16 (HPV) is a major risk microenvironment. Both in humans and in germ free mice
factor for the development of high grade CIN (i.e. CIN 3) iNKT cells were found to acquire an effector (memory)
which, if left untreated, may lead to invasive cervical phenotype before birth, allowing for their distribution to
carcinoma . Circulating nTreg numbers were increased in sites of inflammation [69,70]. As such, they resemble tissue-
individuals with persistent HPV16 infection, compared to infiltrating Th1 cells and CD8+ cytotoxic T lymphocytes (CTL)
women who had cleared the infection, and were possibly and indeed have a corresponding chemokine receptor
associated with the development of a CIN 3 lesion. The expression pattern [18,71,72]. In mouse models, the initial
number of circulating iNKT cells was not related to these pre- influx of iNKT cells was required for the formation of
malignant events of cervical carcinoma, or to the level of granulomatous lesions caused by Mycobacterium tuberculo-
circulating nTregs. However, this does not rule out that iNKT sis  or Cryptococcus neoformans . In the latter study,
cells could be involved in immune responses directed at more the iNKT cells were attracted by the CCR2 ligand monocyte
advanced carcinomas. chemoattractant protein (MCP)-1 (i.e. CCL2). Metelitsa et al.
While the carcinoma patient cohort introduced earlier in demonstrated that human neuroblastomas expressing CCL2,
this review  as a whole was significantly deficient in iNKT were well infiltrated with iNKT cells. Patients with iNKT cell-
cell numbers, some individuals had iNKT cell levels resem- enriched tumors had a significantly prolonged long-term
bling those observed in age-matched healthy controls survival, compared to patients with less iNKT cells at the
whereas others had very low to undetectable numbers of tumor site . Interestingly, they recently reported that
iNKT cells in their circulation. We therefore tested the amplified MYCN oncogene, which is a hallmark of aggressive
hypothesis that a severe iNKT cell deficiency was related to a neuroblastoma , repressed CCL2 expression leading to
poor clinical outcome after radiation therapy in a prospec- impaired iNKT cell infiltration. This was strikingly evident in
tive study of 47 patients with HNSCC . Indeed patients patients with bone marrow metastases of neuroblastoma,
with a severe iNKT cell deficiency prior to radiation therapy suggesting an important contribution of iNKT cells to the
had a poor 3 year disease specific survival, compared to local immune attack against these metastases .
patients with above average iNKT cell levels (43 vs. 92% In addition, Tachibana et al. demonstrated that colorectal
3 year survival rate; p = 0.0027, Log Rank test). This effect carcinomas were well infiltrated with activated CD69+ TCR-
was independent of clinical T stage and age (hazard ratio Vα24+ T cells compared to the patients' control tissue and
[p value] = 17 [0.022], 5.3 [0.024] and 1.1 [0.030] respectively, established by Cox regression analysis that high TCR-Vα24+ T
Cox regression analysis). Pre-therapy T or NK cell levels in cell infiltration was predictive of prolonged (disease free)
186 J.W. Molling et al.
survival, independently of other prognostic variables like carcinoma patients [92,93]. Given that iNKT cells are
clinical T stage . Although the authors did provide positioned early in the immune cascade, that they express
indirect evidence that the TCR-Vα24+ T cells were indeed CCR5 and CXCR3 , can readily release IFN-γ in carcinoma
iNKT cells, they did not clarify whether these cells infiltrated patients [31,39] and apparently infiltrate colorectal carci-
the tumor more readily than other T cells. This is relevant, nomas , some challenging clues arise regarding the
since a high number of T cells infiltrating the tumor is by nature of the before mentioned “early arriving” T cells
itself related to favorable prognosis in colorectal cancer (Fig. 1Aiii). In addition, several groups have demonstrated
(reviewed in ). the capacity of iNKT cells to stimulate DC activation and
maturation [26–28,94,95]. At the tumor site, iNKTcells could
thus facilitate proper DC maturation, resulting in improved
Proposed mechanisms behind the role of iNKT cells migration of more mature DC towards the tumor draining
in anti-tumor responses lymph nodes (Fig. 1B). This would result in more effective
priming of additional T effector cells with tumor associated
A few putative mechanisms via which iNKT cells could take antigens [96–99].
part in anti-tumor responses will be discussed below.
Heat shock proteins (Hsp), which may be released by
tumor cells [79,80], have been proposed to enhance CD1d Therapies aimed at the activation of iNKT cells in
expression on epithelial cells [81–83] as well as DC . In cancer patients
vivo studies suggest that apoptotic tumor cells releasing Hsp
may induce an effective anti-tumor response . Locally Attempts have been made to target iNKT cells in vivo in
produced Hsp at sites of (metastasized) malignant lesions, by clinical phase I studies, based on the use of αGalCer as a
damaged epithelial or tumor cells, might therefore enhance stimulatory ligand (summarized in Table 2). We established in
CD1d mediated antigen presentation within the tumor 24 advanced cancer patients with solid tumors that 3 weekly
microenvironment. For example, GD3 ganglioside is highly i.v. injections of soluble αGalCer did not reach dose limiting
expressed on human tumors of neuroectodermal origin toxicity over a wide dose range (50–4800 μg/kg) .
(which include neuroblastoma and melanoma) but not on Although no clinical responses were recorded, a transient
normal tissues. Wu et al. described that this glycolipid can be increase in serum levels of immunostimulatory cytokines (IL-
effectively cross-presented by CD1d+ DC to mouse iNKT cells 12, IFN-γ, TNF-α and GM-CSF) was observed after the first
in vivo resulting in their activation . αGalCer injection in patients with relatively high iNKT cell
iNKTcell activation leads to the subsequent augmentation levels. This was preceded by a rapid loss of detectable iNKT
of innate, e.g. NK cell mediated, anti-tumor effector cells from the circulation within 24 h after αGalCer
mechanisms. In addition, iNKT cells can enhance antigen administration and iNKT cell levels remained low for up to
specific B cell responses, leading to elevated IgG levels 21 days after the first administration. In line with the decline
[87,88]. These combined effects may lead to opsonization of in iNKT cell levels, no increase in serum cytokines was
cancer cells with immunoglobulins, leading to antibody- observed after a second and third injection with αGalCer (7
dependent cellular cytotoxicity (ADCC) mediated by NK cells and 14 days after the first injection respectively). This might
(reviewed in ). iNKT cells might thus be able to promote a be explained by the relatively high dose of the glycolipid
tumoricidal environment within the vicinity of the tumor injected and the short interval between repeated injections.
(Fig. 1Ai). In a clinical phase I/II trial in chronic Hepatitis C virus
Angiogenesis and lymphangiogenesis are of vital impor- infected patients we observed a re-appearance of circulating
tance for a malignant lesion to develop into a substantial iNKT cells within 2 weeks after i.v. αGalCer given at a lower
tumor mass . iNKT cell activation via systemic injection dose (0.1–10 μg/kg). Furthermore, we observed a cytokine
of αGalCer significantly inhibited angiogenesis of intrader- response even to a second and third αGalCer injection given
mal tumors in mice, which depended on IFN-γ release by at 4 and 8 weeks after the first injection respectively .
both iNKT cells and subsequently activated NK cells  The lack of detectable iNKT cells quickly after i.v. injection
(Fig. 1Aii). Interestingly, Dhodapkar et al. showed that the of the glycolipid might reflect their antigen-specific activa-
neo-vasculature of gliomas expressed CD1d, identifying tion by αGalCer, leading to TCR internalization, especially in
them as targets for this anti-angiogenic response . It the case of the administration of high doses of αGalCer. This
would therefore be interesting to determine neo-vasculature has been demonstrated previously in mice [102,103] and
CD1d expression in a broader panel of human tumors, in the additional studies revealed that in vivo treatment with a
light of local iNKT and NK cell activation. high dose of soluble αGalCer can lead to long-term iNKT cell
It has been proposed that once human colorectal cancers anergy, even after a single injection of the glycolipid
become clinically detectable, and thus have escaped early [104,105].
innate immune surveillance, the adaptive arm plays a The relatively low in vivo responsiveness of circulating
predominant role in preventing disease progression. Color- iNKT cells upon i.v. injection of free αGalCer parallels the
ectal cancer tumors are initially infiltrated with CCR5+ reduced in vitro responsiveness of iNKT cells when stimu-
CXCR3+ T cells . IFN-γ release by these “early arriving” T lated with autologous αGalCer-loaded PBMC [30,32,33]. Both
cells was proposed in this study to trigger events leading to phenomena may be explained by the impaired function of
the influx of effector T cells. The presence of many Th1 APC in cancer patients. Analogous to the in vitro data which
oriented memory/activated T cells in both the tumor center showed restored iNKT proliferation upon stimulation with
and in the tumor's surrounding tissue invasive margin was αGalCer-loaded, in vitro generated, fully functional mature
related to prolonged (disease specific) survival of colorectal DC , the injection of αGalCer-pulsed DC gave rise to more
Invariant natural killer T cells and immunotherapy of cancer 187
Figure 1 Proposed role of iNKT cells in enhancing immune function in tumor microenvironment. (A) iNKT cells recognize endogenous
ligands presented by CD1d on tumor cells, damaged epithelial cells or APC and subsequently release IFN-γ locally. As a result chemokine
release is enhanced, facilitating considerable influx of anti-tumor effector cells. Natural killer (NK) cells and CD8+ cytotoxic T
lymphocytes (CTL) arrive at the tumor site. i) iNKTcells in the (lymphatic) circulation recognize, and are activated by, CD1d on migrating
tumor cells. IFN-γ, released by the activated iNKT cells enhances antibody-dependent cellular cytotoxicity (ADCC) mediated by NK cells
and antibodies specific for cell surface tumor associated antigens (Ig/Fc-R). ii) iNKT cells that are activated by CD1d on endothelial cells
secrete IFN-γ to activate NK cells that subsequently release higher levels of IFN-γ, resulting in the inhibition of neo-angiogenesis initiated
by the developing tumor. iii) NK cells and CTL cross the endothelial barrier to directly kill the tumor cells. (B) Chemokine (CHEM) release
from the tumor microenvironment results in the influx of iNKTcells which subsequently co-localize with residing immature dendritic cells
(immDC) that might be kept in an immature state by tumor derived soluble factors. Binding of the iNKT TCR to the APC CD1d, presenting
endogenous glycolipids triggers IFN-γ release by the iNKTand IL-12 release by the immDC, creating a positive feedback loop that results in
elevated iNKT derived IFN-γ and DC maturation. As a result, mature DC (matDC) up-regulate CCR7 and migrate towards the tumor
draining lymph node (TDLN), where they can prime CD4+ T helper cells and CD8+ cytotoxic T lymphocytes and facilitate B cell activation.
188 J.W. Molling et al.
Table 2 Clinical phase I studies regarding activation of peripheral blood iNKT cells in cancer patients
Therapeutic n Clinical responses Serum Effects on Effects on other
approach (when observed) cytokines circulating circulating cells
after therapy iNKT cells after therapy
Solid tumors i.v. αGalCer 24 Stable disease Elevated IL-12, Decline Transient reduction
(Giaccone et al.) (in 7 pt; IFN-γ (in 1/10 pt (in all pt) in NK cell
 83–216 days) tested) TNF-α number + cytotoxicity
and GM-CSF (in all pt tested)
(in 5/21 pt
Metastatic i.v. immature 4 ? ? Decline, ?
malignancy αGalCer-pulsed followed by
(Okai et al.) moDC mild increase
 (in all pt)
Metastatic i.v. immature 12 Decreased serum Elevated IL-12 Decline, T + NK activation
carcinoma αGalCer-pulsed tumor markers (in all pt) and followed by and increased
(Nieda et al.) moDC (in 2 pt; IFN-γ (in 6/9 mild increase NK cytotoxicity
 4–12 months), patients tested) (in all pt) (in 5/11 pt tested)
(in 1 pt),
(in 2 pt with
Myeloma and i.v mature 3 and Decreased serum Elevated N 100 fold Expansion of
carcinoma αGalCer-pulsed 2 tumor markers IL-12p40 increase at CMV specific
(Chang et al.) moDC (in 3 pt; MIP-1β and peak (in all pt) CD8+ T cells
 9–10 months) IP-10 (in all pt (in 3/3 pt
and stable disease tested) tested)
(in 1 pt; 8 months)
Advanced lung i.v. autologous 12 Stable disease ? N 20 fold ?
canceri PBMC enriched (in 3/9 pt tested; increase at
(Ishikawa et al.) for moDC 23–26 weeks) peak (in 3/12 pt)
 pulsed with and increase in
αGalCer iNKT IFN-γ mRNA
(tested in 1 pt)
Non small cell i.v. autologous 6 Stable disease ? Mild increase Mild increase
lung cancerII PBMC enriched (in 4/6 pt; up to (in 3/6 pt) and in NK cells
(Motohashi et al.) for iNKT cells 12 moths) elevated IFN-γ (in 1/4 pt tested)
 in response to
αGalCer in vitro
in 5/5 pt tested)
Head and neck Intra mucosal 9 Stable disease Elevated IFN-γ Mild increase Increase in NK
cancerIII (nasal) injection (in 5/9 pt; (in 8/9 patients (in 4/9 pt) and cells (in 1/9 pt
(Uchida et al.) of autologous PBMC 8 weeks); tested) elevated IFN-γ tested; case with
 enriched for moDC partial response in response to partial response)
pulsed with (in 1/9 pt; αGalCer in vitro
αGalCer 8 weeks) (in 8/9 pt tested)
Overview of the clinical phase I studies conducted regarding activation of peripheral blood invariant natural killer T (iNKT) cells in cancer
patients. Patients were treated with αGalCer, monocyte derived dendritic cells (moDC) or ex vivo activated iNKT cells. No adverse events
were recorded in any of these studies. I: The method of DC preparation does not exclude the presence of ex vivo activated iNKTcells at time
of injection; II: Preparations contain large amounts of CD3+ Va24− T cells and NK cells; III: The method of DC preparation does not exclude
the presence of ex vivo activated (iNK)T cells at time of injection?: unknown (parameter not tested); pt: patients.
potent in vivo expansion of endogenous iNKT cells. Three respectively with metastatic carcinomas with autologous
independent groups have now shown that this approach αGalCer-pulsed immature moDC and found that this was well
was well tolerated in humans. In two consecutive studies, tolerated and resulted in a mild increase in circulating iNKT
the group of Nicol treated 4  and 12  patients cell numbers. In the latter study, transient but potent pro-
Invariant natural killer T cells and immunotherapy of cancer 189
inflammatory effects were observed in peripheral blood towards a type 1 cytokine profile by stimulation of isolated
samples after DC treatment, consisting of elevated levels of TCR-Vα24+ T cells with mature αGalCer-pulsed moDC in the
IL-12 and IFN-γ, reduced levels of IL-4, activation of T and NK presence of IL-15 . These type 1 polarized iNKT cultures
cells and an increase in NK cell number and cytotoxicity. secreted large amounts of the pro-inflammatory cytokines
These effects were reproduced upon a second injection of IFN-γ, TNF-α and GM-CSF. Although some patient derived
immature moDC pulsed with αGalCer. Next, Chang et al. were iNKT cell cultures showed an initial delay in proliferation, as
able to induce dramatic expansion of circulating iNKT cells in reported previously [30,32,33], this could be overcome by
5 out of 5 advanced cancer patients who received i.v. repeated stimulation with αGalCer-pulsed moDC, resulting
injections with high purity, properly matured, αGalCer- in functionally competent iNKTcells. This suggests that, even
pulsed autologous moDC (N100 fold expansion at peak level in patients with a putative state of iNKT cell anergy, in vitro
in all cases) . Strikingly, despite having undetectable expansion of autologous iNKT cells using mature DC may
peripheral blood iNKT cell counts at the time of study allow for subsequent adoptive transfer of these cells.
enrollment, the iNKT cell level remained above baseline Another advantage of autologous adoptive transfer of ex
after DC treatment for more than 85 days in all patients and vivo expanded iNKT cells would be that their in vitro
for up to 6 months in two patients with longer follow up. expansion and activation does not depend on autologous
Ishikawa et al reported activation and transient expansion of DC, since CD1d is monomorphic. This allows for the use of DC
resident iNKT cells in 3 out of 12 advanced lung cancer cell lines as more standardized in vitro APC leading to more
patients by injecting αGalCer-pulsed APC . However, standardized iNKT cells for adoptive transfer. We have
given the culturing protocol of their APC preparations (low previously demonstrated that human iNKT cells can be
purity of APC in autologous PBMC cultured with GM-CSF, IL-2 expanded using αGalCer-pulsed DC derived from the CD34+
and αGalCer for 7 to 14 days) this could also have resulted human acute myeloid leukemia derived cell line MUTZ-3 (M3-
from a direct iNKT cell infusion. In a follow up study the DC) . In successive experiments we stimulated iNKT
authors indeed demonstrated that this method provided cells with M3-DC over-expressing both CD1d and IL-12
preparations that were enriched for iNKTcells. In 6 non-small (M312CD1d-DC). The obtained iNKT cells had an increased
cell lung cancer patients they re-confirmed that infusion was activation phenotype and were capable of producing high
well tolerated . Furthermore, a transient increase in levels of IFN-γ. Addition of these iNKTcells to autologous CD8
circulating iNKT cells and direct ex vivo IFN-γ production in and tumor associated antigen (TAA) peptide-loaded moDC
response to αGalCer in an ELISPOTassay were observed. More leads to an enhanced TAA specific CTL response . In
recently, this group demonstrated that the administration of similar experiments we found enhanced NK cell functionality
these αGalCer-pulsed APC into the nasal sub mucosa was also upon adding expanded iNKT cells to co-cultures of NK cells
well tolerated . and moDC . The effect of the iNKT cells depended in
It can be concluded from these clinical phase I studies that both cases on the presence of αGalCer on the moDC. Thus,
injection of preparations containing αGalCer-pulsed DC and/ M312CD1d-DC provide us with an attractive “off the shelf”
or iNKT cells is feasible, since it can be performed safely in DC source for the large scale expansion of functional iNKT
advanced cancer patients and results in distinct activation of cells from cancer patients.
iNKT and downstream effector cells.
Retained capacity of long-term mouse iNKT cell
Part II: Towards autologous adoptive transfer of cultures to enhance the in vivo immune response
highly purified and well defined against experimental tumor metastases
pro-inflammatory iNKT cells
Although the strategy of treating carcinoma patients with
autologous adoptive transfer of ex vivo expanded iNKT cells
Ex vivo establishment of iNKT cell lines from healthy is very appealing, long-term in vitro culture would be
controls and carcinoma patients necessary in order to obtain sufficient numbers to repopulate
patients with iNKT cells up to healthy control levels. This
Although the phase I clinical trials performed thus far led to would especially be the case in those individuals who would
promising results there is ample room for improvement. probably benefit most from this therapy, namely those with a
When mature DC are used as “in vivo iNKT cell activators“, a severe deficiency in circulating iNKT cell number and/or a
high iNKT cell expansion can be achieved, but the ability to poor in vitro proliferative response towards αGalCer.
control the functional aspects of the expanded iNKT cells in However, the effect of repetitive in vitro stimulation with
vivo (e.g. cytokine profile or the capacity to home towards DC pulsed with the strong agonist αGalCer on in vivo iNKTcell
tumor sites) is limited. For instance, in the clinical phase I functionality had not previously been investigated. We
study of Chang et al., iNKT cells that were isolated after the therefore developed a method to generate long-term high
administration of αGalCer-pulsed mature moDC lacked the purity oligoclonal mouse iNKT cell cultures . Although
capacity to secrete IFN-γ in an αGalCer ELISPOT assay . other studies previously demonstrated that it is possible to
In addition, the use of highly purified and well defined iNKT generate from mice either short lived cultures containing
cells for adoptive transfer enables to ascribe any immuno- bona fide iNKT cells [57,116,117], or long lived clone derived
logical or clinical effects observed to the injected iNKT cells. iNKT cell hybridomas [118–120], our study was the first to
As already mentioned, we have developed a method to make available large scale, highly pure oligoclonal mouse
expand peripheral blood iNKTcells of healthy controls as well iNKT cell lines representative of in vivo iNKT cells. The iNKT
as advanced cancer patients in vitro, and to polarize them cell cultures retained their most important functional
190 J.W. Molling et al.
aspects since they could bind to αGalCer-loaded mouse certain HLA type or expressing a particular tumor antigen is
CD1d:IgG1 dimers and readily released substantial amounts necessary. The risk of selection of TAA and/or HLA negative
of IFN-γ, GM-CSF, IL-4, IL-5, IL-6, IL-10 and IL-13 upon tumor cells is expected to be low, compared to TAA peptide
antigen specific TCR triggering. In a subsequent study we based immunotherapy. Furthermore, safety issues always
demonstrated that these expanded iNKT cell lines could need to be addressed with the application of TCR gene
induce a partial NK cell-dependent protection against B16. transfer because of the use of retroviral vectors .
F10 lung metastases, upon their adoptive transfer into wild- iNKT cells can be expanded in vitro using the “off the
type mice shortly after tumor injection . Crowe et al. shelf” MUTZ-3 cell line derived DC pulsed with αGalCer. Using
investigated the effect of transfer of freshly isolated iNKT this ex vivo approach, high purity iNKT cells can be obtained
cells on B16.F10 lung metastases in iNKT cell deficient mice. and their potential to enhance anti-tumor responses can be
In their study, the activation of NK cells and thus the strengthened by defining the desired culture conditions (e.g.
inhibition of metastasis formation required additional treat- the use of IL-12 over expressing MUTZ-3 DC). More impor-
ment of mice with αGalCer. We demonstrated that additional tantly, their capacity to secrete high amounts of type 1
αGalCer injection was not required when iNKT cells had been cytokines (e.g. IFN-γ and GM-CSF), to enhance the function-
pre-activated in vitro with αGalCer-pulsed DC . This is ality of other immune cells (e.g. DC, antigen specific CTL or
in line with the findings by Shin et al., who observed NK cells), or their expression of “tumor homing receptors”
inhibition of experimental B16.F10 liver metastases upon (e.g. CXCR3 and CCR5 for carcinomas or CCR2 for neuro-
adoptive transfer of in vitro IL-12 pre-activated iNKT cells blastomas) can be checked prior to adoptive transfer.
into iNKT cell deficient mice without additional αGalCer There are many questions remaining that could lead to a
treatment . In contrast to these results we found, as more efficient therapeutic outcome when answered. For
described in the previous paragraph, that additional trigger- instance: How do peripheral and intra-tumor iNKT cell
ing of expanded human iNKT cells by αGalCer was necessary numbers and function relate? What is the impact of systemic
to achieve enhanced CTL and NK cell responses in vitro. or local (tumor site) enhancement of iNKT cell numbers on
Whether this discrepancy is due to a difference in species (tumor infiltrating) DC, T and NK cell function, also with
(mouse vs. human) or experimental setting (in vitro vs. in regard to disease outcome? Does additional treatment with
vivo) remains to be established. αGalCer enhance these responses and should it then be
Interestingly, Shin et al. also demonstrated that iNKT cell administered systemically (e.g. loaded on DC) or locally? Can
transfer mediated protection was superior to the injection of responses be enhanced when the negative influence of nTreg
high dose IFN-γ or a combination of high dose IFN-γ, IL-2 and on anti-tumor responses is abrogated (e.g. by blocking CTLA4
IL-4. This protection is possibly due to a selective recruit- function [127–130], by TLR2 triggering [131,132]) or by nTreg
ment of iNKT cells to the tumor site followed by their local depletion using e.g. denileukin diftitox )?
release of chemokines and cytokines, allowing for the Nonetheless, we already observed that the amount of pre-
recruitment and trans-activation of downstream effector therapy circulating iNKT cells has a dramatic impact on
cells. Although direct evidence for this is still lacking, we did disease specific survival upon curative radiotherapy of
observe that long-term iNKTcell lines can express the “tumor HNSCC patients. Immuno-adjuvant treatment of iNKT cell
homing receptors” CXCR3 and CCR5 (Molling JW, unpublished deficient HNSCC patients, by radiotherapy accompanied by
results) and that they secrete a broad variety of cytokines in re-constitution of their peripheral blood iNKT cell compart-
vitro (as described above ). These observations, in ment using autologous adoptive transfer of iNKT cells, could
combination with the finding that NK cells were selectively provide a solid base from which questions as those above
recruited to the lungs after i.v. iNKT injection, are in favor of could be further addressed.
the sequential influx of iNKT cells and NK cells into the lungs
of B16.F10 tumor challenged mice. These pre-clinical studies Acknowledgments
illustrate the potential of adoptive transfer of autologous ex
vivo expanded and activated iNKT cells as an immunother- Funded by Dutch Cancer Society: Grant # VU2002-2607
apeutic strategy for the treatment of cancer patients. (AJMvdE, BMEvB and RJS); Netherlands Organization for
Scientific Research: NWO-TALENT grant and Grant # 920-03-
Based on the (pre-)clinical data reviewed here, re-constitu-
tion of the circulating iNKT cell pool as an immunotherapeutic  I. Penn, Tumors of the immunocompromised patient, Annu.
adjuvant therapy of cancer appears to be feasible. In Rev. Med. 39 (1988) 63–73.
addition, screening for those patients who are severely  T.L. Whiteside, Immune suppression in cancer: effects on
deficient in circulating iNKT cells would identify them as immune cells, mechanisms and future therapeutic interven-
the individuals who could benefit the most from this tion, Semin. Cancer Biol. 16 (1) (2006) 3–15.
 M.R. Young, Protective mechanisms of head and neck
approach. In our view, the autologous adoptive transfer of
squamous cell carcinomas from immune assault, Head Neck
ex vivo expanded iNKT cells provides a promising strategy.
28 (5) (2006) 462–470.
An advantage above adoptive transfer of e.g. TAA-pulsed DC  T.L. Whiteside, The role of immune cells in the tumor
vaccines , or transfer of TAA specific T cells generated in microenvironment, Cancer Treat. Res. 130 (2006) 103–124.
vitro via expansion using TAA-pulsed DC or via TCR gene  D.I. Godfrey, K.J. Hammond, L.D. Poulton, M.J. Smyth, A.G.
transfer [125,126] is their universal immunostimulating Baxter, NKT cells: facts, functions and fallacies, Immunol.
potential. Therefore no pre-selection for patients carrying a Today 21 (11) (2000) 573–583.
Invariant natural killer T cells and immunotherapy of cancer 191
 S. Porcelli, C.E. Yockey, M.B. Brenner, S.P. Balk, Analysis of T  M. Nakui, A. Ohta, M. Sekimoto, M. Sato, K. Iwakabe, T.
cell antigen receptor (TCR) expression by human peripheral Yahata, et al., Potentiation of antitumor effect of NKT cell
blood CD4-8- alpha/beta T cells demonstrates preferential use ligand, alpha-galactosylceramide by combination with IL-12
of several V beta genes and an invariant TCR alpha chain, on lung metastasis of malignant melanoma cells, Clin. Exp.
J. Exp. Med. 178 (1) (1993) 1–16. Metastasis 18 (2) (2000) 147–153.
 F.M. Spada, Y. Koezuka, S.A. Porcelli, CD1d-restricted recogni-  M.J. Smyth, M.E. Wallace, S.L. Nutt, H. Yagita, D.I. Godfrey, Y.
tion of synthetic glycolipid antigens by human natural killer T Hayakawa, Sequential activation of NKT cells and NK cells
cells, J. Exp. Med. 188 (8) (1998) 1529–1534. provides effective innate immunotherapy of cancer, J. Exp.
 T. Kawano, J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, K. Motoki, Med. 201 (12) (2005) 1973–1985.
et al., CD1d-restricted and TCR-mediated activation of  S. Fujii, K. Shimizu, C. Smith, L. Bonifaz, R.M. Steinman,
valpha14 NKT cells by glycosylceramides, Science 278 (5343) Activation of natural killer Tcells by alpha-galactosylceramide
(1997) 1626–1629. rapidly induces the full maturation of dendritic cells in vivo
 D. Zhou, J. Mattner, C. Cantu III, N. Schrantz, N. Yin, Y. Gao, and thereby acts as an adjuvant for combined CD4 and CD8 T
et al., Lysosomal glycosphingolipid recognition by NKT cells, cell immunity to a coadministered protein, J. Exp. Med. 198
Science 306 (5702) (2004) 1786–1789. (2) (2003) 267–279.
 Y. Kinjo, D. Wu, G. Kim, G.W. Xing, M.A. Poles, D.D. Ho, et al.,  I.F. Hermans, J.D. Silk, U. Gileadi, M. Salio, B. Mathew, G.
Recognition of bacterial glycosphingolipids by natural killer T Ritter, et al., NKT cells enhance CD4+ and CD8+ T cell
cells, Nature 434 (7032) (2005) 520–525. responses to soluble antigen in vivo through direct interaction
 J. Mattner, K.L. Debord, N. Ismail, R.D. Goff, C. Cantu III, D. with dendritic cells, J. Immunol. 171 (10) (2003) 5140–5147.
Zhou, et al., Exogenous and endogenous glycolipid antigens  K. Liu, J. Idoyaga, A. Charalambous, S. Fujii, A. Bonito, J.
activate NKT cells during microbial infections, Nature 434 Mordoh, et al., Innate NKT lymphocytes confer superior
(7032) (2005) 525–529. adaptive immunity via tumor-capturing dendritic cells,
 A.O. Speak, M. Salio, D.C. Neville, J. Fontaine, D.A. Priestman, J. Exp. Med. 202 (11) (2005) 1507–1516.
N. Platt, et al., Implications for invariant natural killer T cell  T. Kawano, T. Nakayama, N. Kamada, Y. Kaneko, M. Harada, N.
ligands due to the restricted presence of isoglobotrihexosylcer- Ogura, et al., Antitumor cytotoxicity mediated by ligand-
amide in mammals, Proc. Natl. Acad. Sci. U. S. A. 104 (14) activated human V alpha24 NKT cells, Cancer Res. 59 (20)
(2007) 5971–5976. (1999) 5102–5105.
 S. Porubsky, A.O. Speak, B. Luckow, V. Cerundolo, F.M. Platt,  S.M. Tahir, O. Cheng, A. Shaulov, Y. Koezuka, G.J. Bubley, S.B.
H.J. Grone, Normal development and function of invariant Wilson, et al., Loss of IFN-gamma production by invariant NK T
natural killer T cells in mice with isoglobotrihexosylceramide cells in advanced cancer, J. Immunol. 167 (7) (2001) 4046–4050.
(iGb3) deficiency, Proc. Natl. Acad. Sci. U. S. A. 104 (14)  S. Motohashi, S. Kobayashi, T. Ito, K.K. Magara, O. Mikuni, N.
(2007) 5977–5982. Kamada, et al., Preserved IFN-alpha production of circulating
 M. Brigl, L. Bry, S.C. Kent, J.E. Gumperz, M.B. Brenner, Valpha24 NKT cells in primary lung cancer patients, Int. J.
Mechanism of CD1d-restricted natural killer T cell activation Cancer 102 (2) (2002) 159–165.
during microbial infection, Nat. Immunol. 4 (12) (2003)  K. Yanagisawa, K. Seino, Y. Ishikawa, M. Nozue, T. Todoroki, K.
1230–1237. Fukao, Impaired proliferative response of V alpha 24 NKT cells
 H.J. van der Vliet, J.W. Molling, B.M. von Blomberg, N. Nishi, from cancer patients against alpha-galactosylceramide,
W. Kolgen, A.J. van den Eertwegh, et al., The immunoregu- J. Immunol. 168 (12) (2002) 6494–6499.
latory role of CD1d-restricted natural killer T cells in disease,  J. Konishi, K. Yamazaki, H. Yokouchi, N. Shinagawa, K.
Clin. Immunol. 112 (1) (2004) 8–23. Iwabuchi, M. Nishimura, The characteristics of human NKT
 M.G. von Herrath, R.S. Fujinami, J.L. Whitton, Microorgan- cells in lung cancer—CD1d independent cytotoxicity against
isms and autoimmunity: making the barren field fertile? Nat. lung cancer cells by NKT cells and decreased human NKT cell
Rev. Microbiol. 1 (2) (2003) 151–157. response in lung cancer patients, Hum. Immunol. 65 (11)
 J.E. Gumperz, S. Miyake, T. Yamamura, M.B. Brenner, (2004) 1377–1388.
Functionally distinct subsets of CD1d-restricted natural killer  M.V. Dhodapkar, M.D. Geller, D.H. Chang, K. Shimizu, S. Fujii,
T cells revealed by CD1d tetramer staining, J. Exp. Med. 195 K.M. Dhodapkar, et al., A reversible defect in natural killer T
(5) (2002) 625–636. cell function characterizes the progression of premalignant to
 C.H. Kim, E.C. Butcher, B. Johnston, Distinct subsets of human malignant multiple myeloma, J. Exp. Med. 197 (12) (2003)
Valpha24-invariant NKT cells: cytokine responses and chemo- 1667–1676.
kine receptor expression, Trends Immunol. 23 (11) (2002)  K.M. Dhodapkar, B. Cirignano, F. Chamian, D. Zagzag, D.C.
516–519. Miller, J.L. Finlay, et al., Invariant natural killer T cells are
 P.T. Lee, K. Benlagha, L. Teyton, A. Bendelac, Distinct preserved in patients with glioma and exhibit antitumor lytic
functional lineages of human V(alpha)24 natural killer T activity following dendritic cell-mediated expansion, Int. J.
cells, J. Exp. Med. 195 (5) (2002) 637–641. Cancer 109 (6) (2004) 893–899.
 K. Seino, M. Taniguchi, Functionally distinct NKT cell subsets  L.S. Metelitsa, H.W. Wu, H. Wang, Y. Yang, Z. Warsi, S.
and subtypes, J. Exp. Med. 202 (12) (2005) 1623–1626. Asgharzadeh, et al., Natural killer T cells infiltrate neuro-
 N.Y. Crowe, M.J. Smyth, D.I. Godfrey, A critical role for natural blastomas expressing the chemokine CCL2, J. Exp. Med. 199
killer T cells in immunosurveillance of methylcholanthrene- (9) (2004) 1213–1221.
induced sarcomas, J. Exp. Med. 196 (1) (2002) 119–127.  O. DelaRosa, R. Tarazona, J.G. Casado, C. Alonso, B. Ostos, J.
 T. Kawano, J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, H. Sato, et Pena, et al., Valpha24+ NKT cells are decreased in elderly
al., Natural killer-like nonspecific tumor cell lysis mediated by humans, Exp. Gerontol. 37 (2–3) (2002) 213–217.
specific ligand-activated Valpha14 NKT cells, Proc. Natl. Acad.  J.K. Sandberg, N. Bhardwaj, D.F. Nixon, Dominant effector
Sci. U. S. A. 95 (10) (1998) 5690–5693. memory characteristics, capacity for dynamic adaptive
 R. Nakagawa, I. Nagafune, Y. Tazunoki, H. Ehara, H. Tomura, expansion, and sex bias in the innate Valpha24 NKT cell
R. Iijima, et al., Mechanisms of the antimetastatic effect in compartment, Eur. J. Immunol. 33 (3) (2003) 588–596.
the liver and of the hepatocyte injury induced by alpha-  J.W. Molling, W. Kolgen, H.J. van der Vliet, M.F. Boomsma, H.
galactosylceramide in mice, J. Immunol. 166 (11) (2001) Kruizenga, C.H. Smorenburg, et al., Peripheral blood IFN-
6578–6584. gamma-secreting Valpha24+Vbeta11+ NKT cell numbers are
192 J.W. Molling et al.
decreased in cancer patients independent of tumor type or  L.S. Metelitsa, O.V. Naidenko, A. Kant, H.W. Wu, M.J. Loza, B.
tumor load, Int. J. Cancer 116 (1) (2005) 87–93. Perussia, et al., Human NKT cells mediate antitumor cyto-
 T. Crough, D.M. Purdie, M. Okai, A. Maksoud, M. Nieda, A.J. toxicity directly by recognizing target cell CD1d with bound
Nicol, Modulation of human Valpha24(+)Vbeta11(+) NKT cells ligand or indirectly by producing IL-2 to activate NK cells, J.
by age, malignancy and conventional anticancer therapies, Br. Immunol. 167 (6) (2001) 3114–3122.
J. Cancer 91 (11) (2004) 1880–1886.  T. Takahashi, M. Nieda, Y. Koezuka, A. Nicol, S.A. Porcelli, Y.
 E. Peralbo, O. DelaRosa, I. Gayoso, M.L. Pita, R. Tarazona, R. Ishikawa, et al., Analysis of human V alpha 24+ CD4+ NKT cells
Solana, Decreased frequency and proliferative response of activated by alpha-glycosylceramide-pulsed monocyte-
invariant Valpha24Vbeta11 natural killer T (iNKT) cells in derived dendritic cells, J. Immunol. 164 (9) (2000) 4458–4464.
healthy elderly, Biogerontology 7 (5–6) (2006) 483–492.  K. Chamoto, T. Takeshima, A. Kosaka, T. Tsuji, J. Matsuzaki, Y.
 T. Azuma, T. Takahashi, A. Kunisato, T. Kitamura, H. Hirai, Togashi, et al., NKT cells act as regulatory cells rather than
Human CD4+ CD25+ regulatory T cells suppress NKT cell killer cells during activation of NK cell-mediated cytotoxicity
functions, Cancer Res. 63 (15) (2003) 4516–4520. by alpha-galactosylceramide in vivo, Immunol. Lett. 95 (1)
 A.M. Wolf, D. Wolf, M. Steurer, G. Gastl, E. Gunsilius, B. (2004) 5–11.
Grubeck-Loebenstein, Increase of regulatory T cells in the  J.W. Molling, T.D. de Gruijl, J. Glim, M. Moreno, L. Rozendaal,
peripheral blood of cancer patients, Clin. Cancer Res. 9 (2) C.J. Meijer, et al., CD4(+)CD25hi regulatory T-cell frequency
(2003) 606–612. correlates with persistence of human papillomavirus type 16
 R. Okita, T. Saeki, S. Takashima, Y. Yamaguchi, T. Toge, CD4+ and T helper cell responses in patients with cervical
CD25+ regulatory T cells in the peripheral blood of patients intraepithelial neoplasia, Int. J. Cancer 121 (8) (2007)
with breast cancer and non-small cell lung cancer, Oncol. Rep. 1749–1755.
14 (5) (2005) 1269–1273.  J.M. Walboomers, C.J. Meijer, R.D. Steenbergen, M. van Duin,
 F. Ichihara, K. Kono, A. Takahashi, H. Kawaida, H. Sugai, H. T.J. Helmerhorst, P.J. Snijders, Human papillomavirus and the
Fujii, Increased populations of regulatory T cells in peripheral development of cervical cancer: concept of carcinogenesis,
blood and tumor-infiltrating lymphocytes in patients with Ned. Tijdschr. Geneeskd. 144 (35) (2000) 1671–1674.
gastric and esophageal cancers, Clin. Cancer Res. 9 (12) (2003)  J.W. Molling, J.A. Langius, J.A. Langendijk, C.R. Leemans, H.
4404–4408. J. Bontkes, H.J. van der Vliet, et al., Low levels of circulating
 U.K. Liyanage, T.T. Moore, H.G. Joo, Y. Tanaka, V. Herrmann, invariant natural killer T cells predict poor clinical outcome in
G. Doherty, et al., Prevalence of regulatory T cells is increased patients with head and neck squamous cell carcinoma, J. Clin.
in peripheral blood and tumor microenvironment of patients Oncol. 25 (7) (2007) 862–868.
with pancreas or breast adenocarcinoma, J. Immunol. 169 (5)  V. Soysal, O.G. Yigitbasi, M. Alper, T. Patiroglu, E. Guney, Total
(2002) 2756–2761. lymphocyte and T lymphocyte subpopulation levels in head
 E.M. Shevach, CD4+ CD25+ suppressor T cells: more questions and neck squamous cell carcinomas, J. Exp. Clin. Cancer Res.
than answers, Nat. Rev. Immunol. 2 (6) (2002) 389–400. 17 (2) (1998) 207–212.
 T.K. Hoffmann, J. Muller-Berghaus, R.L. Ferris, J.T. Johnson,  G.T. Wolf, C.R. Bradford, S. Urba, A. Smith, A. Eisbruch, D.B.
W.J. Storkus, T.L. Whiteside, Alterations in the frequency of Chepeha, et al., Immune reactivity does not predict che-
dendritic cell subsets in the peripheral circulation of patients motherapy response, organ preservation, or survival in
with squamous cell carcinomas of the head and neck, Clin. advanced laryngeal cancer, Laryngoscope 112 (8 Pt 1) (2002)
Cancer Res. 8 (6) (2002) 1787–1793. 1351–1356.
 H.J. van der Vliet, R. Wang, S.C. Yue, H.B. Koon, S.P Balk, M.A.
.  I. Kuss, B. Hathaway, R.L. Ferris, W. Gooding, T.L. Whiteside,
Exley, Circulating myeloid dendritic cells of advanced cancer Decreased absolute counts of T lymphocyte subsets and their
patients result in reduced activation and a biased cytokine relation to disease in squamous cell carcinoma of the head and
profile in invariant NKT cells, J. Immunol. 180 (11) (2008) neck, Clin. Cancer Res. 10 (11) (2004) 3755–3762.
7287–7293.  S.P. Schantz, E.J. Shillitoe, B. Brown, B. Campbell, Natural
 O. Imataki, Y. Heike, H. Makiyama, A. Iizuka, Y. Ikarashi, T. killer cell activity and head and neck cancer: a clinical
Ishida, et al., Insufficient ex vivo expansion of Valpha24(+) assessment, J. Natl. Cancer Inst. 77 (4) (1986) 869–875.
natural killer T cells in malignant lymphoma patients related  S.P. Schantz, N.G. Ordonez, Quantitation of natural killer cell
to the suppressed expression of CD1d molecules on CD14(+) function and risk of metastatic poorly differentiated head and
cells, Cytotherapy (2008) 1–10. neck cancer, Nat. Immun. Cell Growth Regul. 10 (5) (1991)
 H.J. van der Vliet, H.M. Pinedo, B.M. von Blomberg, A.J. van 278–288.
den Eertwegh, R.J. Scheper, G. Giaccone, Natural killer T  F.M. Gonzalez, J.A. Vargas, C. Lopez-Cortijo, R. Castejon, C.
cells, Lancet Oncol. 3 (9) (2002) 574. Gorriz, R. Ramirez-Camacho, et al., Prognostic significance of
 H.J. van der Vliet, N. Nishi, Y. Koezuka, B.M. von Blomberg, natural killer cell activity in patients with laryngeal carci-
A.J. van den Eertwegh, S.A. Porcelli, et al., Potent expansion noma, Arch. Otolaryngol Head Neck Surg. 124 (8) (1998)
of human natural killer T cells using alpha-galactosylceramide 852–856.
(KRN7000)-loaded monocyte-derived dendritic cells, cultured  T.E. Reichert, R. Day, E.M. Wagner, T.L. Whiteside, Absent or
in the presence of IL-7 and IL-15, J. Immunol. Methods 247 low expression of the zeta chain in T cells at the tumor site
(1–2) (2001) 61–72. correlates with poor survival in patients with oral carcinoma,
 N. Nishi, H.J. van der Vliet, Y. Koezuka, B.M. von Blomberg, R. Cancer Res. 58 (23) (1998) 5344–5347.
J. Scheper, H.M. Pinedo, et al., Synergistic effect of KRN7000  J.H. Heimdal, H.J. Aarstad, J. Olofsson, Peripheral blood T-
with interleukin-15, -7, and -2 on the expansion of human V lymphocyte and monocyte function and survival in patients
alpha 24+V beta 11+ T cells in vitro, Hum. Immunol. 61 (4) with head and neck carcinoma, Laryngoscope 110 (3 Pt 1)
(2000) 357–365. (2000) 402–407.
 H.J. van der Vliet, J.W. Molling, N. Nishi, A.J. Masterson, W.  H.J. Der Vliet, N. Nishi, T.D. de Gruijl, B.M. von Blomberg, A.J.
Kolgen, S.A. Porcelli, et al., Polarization of Valpha24+ van den Eertwegh, H.M. Pinedo, et al., Human natural killer T
Vbeta11+ natural killer T cells of healthy volunteers and cells acquire a memory-activated phenotype before birth,
cancer patients using alpha-galactosylceramide-loaded and Blood 95 (7) (2000) 2440–2442.
environmentally instructed dendritic cells, Cancer Res. 63  S.H. Park, K. Benlagha, D. Lee, E. Balish, A. Bendelac,
(14) (2003) 4101–4106. Unaltered phenotype, tissue distribution and function of
Invariant natural killer T cells and immunotherapy of cancer 193
Valpha14(+) NKT cells in germ-free mice, Eur. J. Immunol. 30  G. Galli, P. Pittoni, E. Tonti, C. Malzone, Y. Uematsu, M.
(2) (2000) 620–625. Tortoli, et al., Invariant NKT cells sustain specific B cell
 S.Y. Thomas, R. Hou, J.E. Boyson, T.K. Means, C. Hess, D.P. responses and memory, Proc. Natl. Acad. Sci. U. S. A. 104 (10)
Olson, et al., CD1d-restricted NKT cells express a chemokine (2007) 3984–3989.
receptor profile indicative of Th1-type inflammatory homing  N. Nishida, H. Yano, T. Nishida, T. Kamura, M. Kojiro,
cells, J. Immunol. 171 (5) (2003) 2571–2580. Angiogenesis in cancer, Vasc. Health Risk Manag. 2 (3) (2006)
 C.H. Kim, B. Johnston, E.C. Butcher, Trafficking machinery of 213–219.
NKT cells: shared and differential chemokine receptor  Y. Hayakawa, K. Takeda, H. Yagita, M.J. Smyth, L. Van Kaer, K.
expression among V alpha 24(+)V beta 11(+) NKT cell subsets Okumura, et al., IFN-gamma-mediated inhibition of tumor
with distinct cytokine-producing capacity, Blood 100 (1) angiogenesis by natural killer T-cell ligand, alpha-galactosyl-
(2002) 11–16. ceramide, Blood 100 (5) (2002) 1728–1733.
 I. Apostolou, Y. Takahama, C. Belmant, T. Kawano, M. Huerre,  H. Musha, H. Ohtani, T. Mizoi, M. Kinouchi, T. Nakayama, K.
G. Marchal, et al., Murine natural killer T(NKT) cells Shiiba, et al., Selective infiltration of CCR5(+)CXCR3(+) T
[correction of natural killer cells] contribute to the granulo- lymphocytes in human colorectal carcinoma, Int. J. Cancer
matous reaction caused by mycobacterial cell walls, Proc. 116 (6) (2005) 949–956.
Natl. Acad. Sci. U. S. A. 96 (9) (1999) 5141–5146.  F. Pages, A. Berger, M. Camus, F. Sanchez-Cabo, A. Costes, R.
 K. Kawakami, Y. Kinjo, K. Uezu, S. Yara, K. Miyagi, Y. Koguchi, Molidor, et al., Effector memory T cells, early metastasis, and
et al., Monocyte chemoattractant protein-1-dependent survival in colorectal cancer, N. Engl. J. Med. 353 (25) (2005)
increase of V alpha 14 NKT cells in lungs and their roles in 2654–2666.
Th1 response and host defense in cryptococcal infection,  J. Galon, A. Costes, F. Sanchez-Cabo, A. Kirilovsky, B. Mlecnik,
J. Immunol. 167 (11) (2001) 6525–6532. C. Lagorce-Pages, et al., Type, density, and location of
 R.C. Seeger, G.M. Brodeur, H. Sather, A. Dalton, S.E. Siegel, immune cells within human colorectal tumors predict clinical
K.Y. Wong, et al., Association of multiple copies of the N-myc outcome, Science 313 (5795) (2006) 1960–1964.
oncogene with rapid progression of neuroblastomas, N. Engl.  S.C. Yue, A. Shaulov, R. Wang, S.P. Balk, M.A. Exley, CD1d
J. Med. 313 (18) (1985) 1111–1116. ligation on human monocytes directly signals rapid NF-kappaB
 L. Song, T. Ara, H.W. Wu, C.W. Woo, C.P. Reynolds, R.C. activation and production of bioactive IL-12, Proc. Natl. Acad.
Seeger, et al., Oncogene MYCN regulates localization of NKT Sci. U. S. A. 102 (33) (2005) 11811–11816.
cells to the site of disease in neuroblastoma, J. Clin. Invest.  G. Eberl, P. Brawand, H.R. MacDonald, Selective bystander
117 (9) (2007) 2702–2712. proliferation of memory CD4+ and CD8+ T cells upon NK T or T
 T. Tachibana, H. Onodera, T. Tsuruyama, A. Mori, S. cell activation, J. Immunol. 165 (8) (2000) 4305–4311.
Nagayama, H. Hiai, et al., Increased intratumor Valpha24-  A.V. Gorbachev, R.L. Fairchild, Activated NKT cells increase
positive natural killer T cells: a prognostic factor for primary dendritic cell migration and enhance CD8+ T cell responses in
colorectal carcinomas, Clin. Cancer Res. 11 (20) (2005) the skin, Eur. J. Immunol. 36 (9) (2006) 2494–2503.
7322–7327.  C.J. Montoya, H.B. Jie, L. Al Harthi, C. Mulder, P.J. Patino, M.
 H. Ohtani, Focus on TILs: prognostic significance of tumor T. Rugeles, et al., Activation of plasmacytoid dendritic cells
infiltrating lymphocytes in human colorectal cancer, Cancer with TLR9 agonists initiates invariant NKT cell-mediated cross-
Immun. 7 (2007) 4. talk with myeloid dendritic cells, J. Immunol. 177 (2) (2006)
 H. Isomoto, M. Oka, Y. Yano, Y. Kanazawa, H. Soda, R. Terada, 1028–1039.
et al., Expression of heat shock protein (Hsp) 70 and Hsp 40 in  M.S. Vincent, D.S. Leslie, J.E. Gumperz, X. Xiong, E.P. Grant,
gastric cancer, Cancer Lett. 198 (2) (2003) 219–228. M.B. Brenner, CD1-dependent dendritic cell instruction, Nat.
 J. Sasaki, M. Dejehansart, J. De Bruyn, The expression of Immunol. 3 (12) (2002) 1163–1168.
mycobacterial heat shock protein (HSP64) on Meth A tumour  R.J. Vuylsteke, B.G. Molenkamp, P.A. van Leeuwen, S. Meijer,
cells, Immunol. Cell Biol. 72 (5) (1994) 415–418. P.G. Wijnands, J.B. Haanen, et al., Tumor-specific CD8+ T cell
 S.P. Colgan, R.S. Pitman, T. Nagaishi, A. Mizoguchi, E. reactivity in the sentinel lymph node of GM-CSF-treated stage
Mizoguchi, L.F. Mayer, et al., Intestinal heat shock protein I melanoma patients is associated with high myeloid dendritic
110 regulates expression of CD1d on intestinal epithelial cells, cell content, Clin. Cancer Res. 12 (9) (2006) 2826–2833.
J. Clin. Invest. 112 (5) (2003) 745–754.  G. Giaccone, C.J. Punt, Y. Ando, R. Ruijter, N. Nishi, M. Peters,
 C.V. Nicchitta, Come forth CD1d: Hsp110 in the regulation of et al., A phase I study of the natural killer T-cell ligand alpha-
intestinal epithelial CD1d expression, J. Clin. Invest. 112 (5) galactosylceramide (KRN7000) in patients with solid tumors,
(2003) 646–648. Clin. Cancer Res. 8 (12) (2002) 3702–3709.
 B. Bonish, D. Jullien, Y. Dutronc, B.B. Huang, R. Modlin, F.M.  B.J. Veldt, H.J. van der Vliet, B.M. von Blomberg, H. van
Spada, et al., Overexpression of CD1d by keratinocytes in Vlierberghe, G. Gerken, N. Nishi, et al., Randomized placebo
psoriasis and CD1d-dependent IFN-gamma production by NK-T controlled phase I/II trial of alpha-galactosylceramide for the
cells, J. Immunol. 165 (7) (2000) 4076–4085. treatment of chronic hepatitis C, J. Hepatol. 47 (3) (2007)
 Y.V. Bobryshev, R.S. Lord, Expression of heat shock protein-70 356–365.
by dendritic cells in the arterial intima and its potential  N.Y. Crowe, A.P. Uldrich, K. Kyparissoudis, K.J. Hammond, Y.
significance in atherogenesis, J. Vasc. Surg. 35 (2) (2002) Hayakawa, S. Sidobre, et al., Glycolipid antigen drives rapid
368–375. expansion and sustained cytokine production by NK T cells,
 H. Feng, Y. Zeng, L. Whitesell, E. Katsanis, Stressed apoptotic J. Immunol. 171 (8) (2003) 4020–4027.
tumor cells express heat shock proteins and elicit tumor-  M.T. Wilson, C. Johansson, D. Olivares-Villagomez, A.K. Singh,
specific immunity, Blood 97 (11) (2001) 3505–3512. A.K. Stanic, C.R. Wang, et al., The response of natural killer T
 D.Y. Wu, N.H. Segal, S. Sidobre, M. Kronenberg, P.B. Chapman, cells to glycolipid antigens is characterized by surface
Cross-presentation of disialoganglioside GD3 to natural killer T receptor down-modulation and expansion, Proc. Natl. Acad.
cells, J. Exp. Med. 198 (1) (2003) 173–181. Sci. U. S. A. 100 (19) (2003) 10913–10918.
 H. Lin, M. Nieda, V. Rozenkov, A.J. Nicol, Analysis of the effect  V.V. Parekh, M.T. Wilson, D. Olivares-Villagomez, A.K. Singh,
of different NKT cell subpopulations on the activation of CD4 L. Wu, C.R. Wang, et al., Glycolipid antigen induces long-term
and CD8 T cells, NK cells, and B cells, Exp. Hematol. 34 (3) natural killer T cell anergy in mice, J. Clin. Invest. 115 (9)
(2006) 289–295. (2005) 2572–2583.
194 J.W. Molling et al.
 A.P. Uldrich, N.Y. Crowe, K. Kyparissoudis, D.G. Pellicci, Y.  J.E. Gumperz, C. Roy, A. Makowska, D. Lum, M. Sugita, T.
Zhan, A.M. Lew, et al., NKT cell stimulation with glycolipid Podrebarac, et al., Murine CD1d-restricted T cell recognition
antigen in vivo: costimulation-dependent expansion, Bim- of cellular lipids, Immunity 12 (2) (2000) 211–221.
dependent contraction, and hyporesponsiveness to further  O. Lantz, A. Bendelac, An invariant T cell receptor alpha chain
antigenic challenge, J. Immunol. 175 (5) (2005) 3092–3101. is used by a unique subset of major histocompatibility complex
 M. Okai, M. Nieda, A. Tazbirkova, D. Horley, A. Kikuchi, S. class I-specific CD4+ and CD4-8- T cells in mice and humans,
Durrant, et al., Human peripheral blood Valpha24+ Vbeta11+ J. Exp. Med. 180 (3) (1994) 1097–1106.
NKT cells expand following administration of alpha-galacto-  S.M. Behar, T.A. Podrebarac, C.J. Roy, C.R. Wang, M.B.
sylceramide-pulsed dendritic cells, Vox Sang. 83 (3) (2002) Brenner, Diverse TCRs recognize murine CD1, J. Immunol.
250–253. 162 (1) (1999) 161–167.
 M. Nieda, M. Okai, A. Tazbirkova, H. Lin, A. Yamaura, K. Ide, et  J.W. Molling, M. Moreno, J. de Groot, H.J. van der Vliet, B.M.
al., Therapeutic activation of Valpha24+Vbeta11+ NKT cells in von Blomberg, A.J. van den Eertwegh, et al., Chronically
human subjects results in highly coordinated secondary stimulated mouse invariant NKT cell lines have a preserved
activation of acquired and innate immunity, Blood 103 (2) capacity to enhance protection against experimental tumor
(2004) 383–389. metastases, Immunol. Lett. 118 (1) (2008) 36–43.
 D.H. Chang, K. Osman, J. Connolly, A. Kukreja, J. Krasovsky,  N.Y. Crowe, J.M. Coquet, S.P. Berzins, K. Kyparissoudis, R.
M. Pack, et al., Sustained expansion of NKT cells and antigen- Keating, D.G. Pellicci, et al., Differential antitumor immunity
specific T cells after injection of alpha-galactosyl-ceramide mediated by NKT cell subsets in vivo, J. Exp. Med. 202 (9)
loaded mature dendritic cells in cancer patients, J. Exp. Med. (2005) 1279–1288.
201 (9) (2005) 1503–1517.  T. Shin, T. Nakayama, Y. Akutsu, S. Motohashi, Y. Shibata, M.
 A. Ishikawa, S. Motohashi, E. Ishikawa, H. Fuchida, K. Harada, et al., Inhibition of tumor metastasis by adoptive
Higashino, M. Otsuji, et al., A phase I study of alpha- transfer of IL-12-activated Valpha14 NKT cells, Int. J. Cancer
galactosylceramide (KRN7000)-pulsed dendritic cells in 91 (4) (2001) 523–528.
patients with advanced and recurrent non-small cell lung  E. Gilboa, DC-based cancer vaccines, J. Clin. Invest. 117 (5)
cancer, Clin. Cancer Res. 11 (5) (2005) 1910–1917. (2007) 1195–1203.
 S. Motohashi, A. Ishikawa, E. Ishikawa, M. Otsuji, T. Iizasa, H.  C.H. June, Principles of adoptive T cell cancer therapy, J. Clin.
Hanaoka, et al., A phase I study of in vitro expanded natural Invest. 117 (5) (2007) 1204–1212.
killer T cells in patients with advanced and recurrent non-  S.A. Xue, H.J. Stauss, Enhancing immune responses for cancer
small cell lung cancer, Clin. Cancer Res. 12 (20 Pt 1) (2006) therapy, Cell Mol. Immunol. 4 (3) (2007) 173–184.
6079–6086.  F.S. Hodi, M.C. Mihm, R.J. Soiffer, F.G. Haluska, M. Butler, M.V.
 T. Uchida, S. Horiguchi, Y. Tanaka, H. Yamamoto, N. Kunii, S. Seiden, et al., Biologic activity of cytotoxic T lymphocyte-
Motohashi, et al., Phase I study of alpha-galactosylceramide- associated antigen 4 antibody blockade in previously vacci-
pulsed antigen presenting cells administration to the nasal nated metastatic melanoma and ovarian carcinoma patients,
submucosa in unresectable or recurrent head and neck cancer, Proc. Natl. Acad. Sci. U. S. A. 100 (8) (2003) 4712–4717.
Cancer Immunol. Immunother. 57 (3) (2008) 337–345.  A. Ribas, L.H. Camacho, G. Lopez-Berestein, D. Pavlov, C.A.
 A.J. Masterson, C.C. Sombroek, T.D. de Gruijl, Y.M. Graus, H.J. Bulanhagui, R. Millham, et al., Antitumor activity in
van der Vliet, S.M. Lougheed, et al., MUTZ-3, a human cell line melanoma and anti-self responses in a phase I trial with
model for the cytokine-induced differentiation of dendritic the anti-cytotoxic T lymphocyte-associated antigen 4 mono-
cells from CD34+ precursors, Blood 100 (2) (2002) 701–703. clonal antibody CP-675,206, J. Clin. Oncol. 23 (35) (2005)
 M. Moreno, J.W. Molling, S. von Mensdorff-Pouilly, E. 8968–8977.
Hooiberg, D. Kramer, A.W. Reurs, et al., Interferon-y  K. Sanderson, R. Scotland, P. Lee, D. Liu, S. Groshen, J.
producing human invariant natural killer T-cells promote Snively, et al., Autoimmunity in a phase I trial of a fully human
tumor associated antigen-specific cytotoxic T cell responses, anti-cytotoxic T-lymphocyte antigen-4 monoclonal antibody
J. Immunol. (2008) In press. with multiple melanoma peptides and Montanide ISA 51 for
 M. Moreno, J.W. Molling, S. von Mensdorff-Pouilly, R.H. patients with resected stages III and IV melanoma, J. Clin.
verheijen, B.M. von Blomberg, A.J. van den Eertwegh, et al., Oncol. 23 (4) (2005) 741–750.
In vitro expanded human invariant natural killer T-cells  G.Q. Phan, J.C. Yang, R.M. Sherry, P. Hwu, S.L. Topalian, D.J.
enhance functional activity of NK cells, Clin. Immunol. Schwartzentruber, et al., Cancer regression and autoimmunity
(2008) In press. induced by cytotoxic T lymphocyte-associated antigen 4
 J.W. Molling, M. Moreno, H.J. van der Vliet, B.M. von blockade in patients with metastatic melanoma, Proc. Natl.
Blomberg, A.J. van den Eertwegh, R.J. Scheper, et al., Acad. Sci. U. S. A. 100 (14) (2003) 8372–8377.
Generation and sustained expansion of mouse spleen invariant  R.P. Sutmuller, M.H. den Brok, M. Kramer, E.J. Bennink, L.W.
NKT cell lines with preserved cytokine releasing capacity, Toonen, B.J. Kullberg, et al., Toll-like receptor 2 controls
J. Immunol. Methods 322 (1–2) (2007) 70–81. expansion and function of regulatory T cells, J. Clin. Invest.
 M. Maeda, S. Lohwasser, T. Yamamura, F. Takei, Regulation of 116 (2) (2006) 485–494.
NKT cells by Ly49: analysis of primary NKT cells and generation  H. Liu, M. Komai-Koma, D. Xu, F.Y. Liew, Toll-like receptor
of NKT cell line, J. Immunol. 167 (8) (2001) 4180–4186. 2 signaling modulates the functions of CD4+ CD25+ regu-
 Y. Ikarashi, A. Iizuka, Y. Heike, M. Yoshida, Y. Takaue, H. latory T cells, Proc. Natl. Acad. Sci. U. S. A. 103 (18) (2006)
Wakasugi, Cytokine production and migration of in vitro- 7048–7053.
expanded NK1.1(-) invariant Valpha14 natural killer T  B. Barnett, I. Kryczek, P. Cheng, W. Zou, T.J. Curiel,
(Valpha14i NKT) cells using alpha-galactosylceramide and IL-2, Regulatory T cells in ovarian cancer: biology and therapeutic
Immunol. Lett. 101 (2) (2005) 160–167. potential, Am. J. Reprod. Immunol. 54 (6) (2005) 369–377.