Invariant NKT cells and CD1d+ cells amass in human omentum and are depleted in patients with cancer...

Invariant NKT cells and CD1d+ cells amassin human omentum and are depleted in patientswith cancer and obesity

Lydia Lynch1, Donal O’ Shea2, Desmond C. Winter3, Justin Geoghegan3,

Derek G. Doherty�4,5 and Cliona O’Farrelly�6

1 Education and Research Centre, St.Vincent’s University Hospital, Dublin, Ireland2 Department of Endocrinology, St.Vincent’s University Hospital, Dublin, Ireland3 Department of Surgery, St.Vincent’s University Hospital, Dublin, Ireland4 Department of Immunology, Trinity College Dublin, Ireland5 St. James’s Hospital, Institute of Immunology and Department of Biology, National University

of Ireland, Maynooth, Dublin, Ireland6 School of Biochemistry and Immunology, Trinity College Dublin, Ireland

Invariant NKT (iNKT) cells recognize lipid antigens presented by CD1d and respond rapidly by

killing tumor cells and release cytokines that activate and regulate adaptive immune

responses. They are essential for tumor rejection in various mouse models, but clinical trials

in humans involving iNKT cells have been less successful, partly due to their rarity in

humans compared with mice. Here we describe an accumulation of functional iNKT cells in

human omentum, a migratory organ with healing properties. Analysis of 39 omental samples

revealed that T cells are the predominant lymphoid cell type and of these, 10% expressed the

invariant Va24Ja18 TCR chain, found on iNKT cells, higher than in any other human organ

tested to date. About 15% of omental hematopoietic cells expressed CD1d, compared with 1%

in blood (po0.001). Enriched omental iNKT cells killed CD1d+ targets and released IFN-c and

IL-4 upon activation. Omental iNKT-cell frequencies were lower in patients with severe

obesity (p 5 0.005), and with colorectal carcinoma (p 5 0.004) compared with lean healthy

subjects. These data suggest a novel role for the omentum in immune regulation and tumor

immunity and identify it as a potential source of iNKT cells for therapeutic use.

Key words: CD1d . Human . NKT cells . Obesity . Tumor immunity

Introduction

While most T lymphocytes recognize peptides bound to MHC

molecules, a minority recognize lipids and glycolipids bound to the

MHC-like glycoprotein, CD1d [1–3]. CD1d-restricted T cells

frequently express cell-surface markers that are typically found on

NK cells and a highly conserved antigen receptor (TCR) a-chain

(Va14Ja18 in mice and Va24Ja18 in humans), which pairs with a

limited number of b-chains [1, 2, 4] and are hence termed invariant

NKT (iNKT) cells. The identity of natural antigenic ligands recognized

by iNKT cells remains controversial, but the lysosomal glyco-

sphingolipid, isoglobotrihexosylceramide [5], and glycosphingolipids

found in some bacteria [6] have been proposed to be endogenous

and exogenous antigens, respectively. Additionally, the xenogeneic

glycolipid, a-galactosylceramide (a-GC), isolated from the marine

sponge, Agelas mauritianus, is a potent agonist for murine and human

iNKT cells [7]. Upon activation with a-GC, iNKT cells display

powerful anti-tumor cytotoxic activity [8], rapidly release cytokines

(such as GM-CSF, TNF-a, IFN-g, IL-4, IL-10 and IL-13) [2, 4, 9] and

promote maturation of dendritic cells into APC that are capable of

�These authors contributed equally to this work.Correspondence: Dr. Lydia Lynche-mail: [email protected]

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2009. 39: 1893–1901 DOI 10.1002/eji.200939349 Innate immunity 1893

activating conventional ab-T cells [10]. Therapeutic activation of

iNKT cells using a-GC can prevent and reverse tumor growth

[11–13], prevent autoimmune disease and mediate protection

against multiple infectious agents [1, 2] in murine models. However,

clinical trials involving a-GC in humans have shown limited success

[14, 15]. One reason for this is that, while iNKT cells account for up

to 5% of peripheral and up to 30% of hepatic T cells in mice,

depending on strain [2], they are found at 100-fold lower frequencies

at these locations in humans [16, 17]. Here, we describe large

numbers of functional iNKT cells in human omentum.

The omentum is a large apron-like peritoneal fold that connects

the spleen, pancreas, stomach and transverse colon, terminating in an

apron-like structure, which in obese people accumulates in consid-

erable quantities [18]. It is composed of two mesothelial sheets which

enclose adipocytes embedded in loose connective tissue with islands

of compact tissue, known as ‘‘milky spots’’, which contain macro-

phages, B cells, T cells, mast cells and dendritic cells [19, 20]. As well

as storing fat within adipocytes, the omentum has a long-established

reputation as the ‘‘abdominal policeman’’ due to its unique ability to

adhere to foreign bodies and travel to sites of inflammation, injury

and infection. Here it restores order by surrounding the compromised

site where it seals microperforations, localizes inflammation and

limits the spread of infection, and provokes revascularization and

tissue regeneration [18, 21]. Surgical transposition of omental tissue

to other body sites has been used for over 100 years for tissue

regeneration [22], which is mediated, in part, by the release of

growth and angiogenic factors, chemokines and cytokines [21, 23].

Since the omentum is a site of lipid accumulation and one

which is capable of recognizing foreign bodies and sites of injury,

mediating innate immune responses and promoting healing, we

hypothesized that lipid-reactive iNKT cells may play a role in

these activities. We show that the omentum contains the highest

concentrations of both CD1d+ cells and functional iNKT cells

known to date at any location in the human body. The immune

system appears to be compromised in obese individuals [24], and

obesity is associated with a significantly increased risk of devel-

oping several malignancies [25, 26]. We found that numbers of

iNKT cells and CD1d+ cells were reduced in obese and cancer

patients compared with lean control subjects. Our data suggest a

possible role for omental iNKT cells in malignancy and identifies

the omentum as a novel source of putative anti-tumor effectors

with potential for use in adoptive immunotherapy.

Results

The human omentum is a site of accumulation ofT cells

Flow cytometric analysis of single-cell suspensions of omental tissue

samples (after removal of adipocytes) from 39 patients undergoing

elective abdominal surgery indicated that lymphocytes (detected by

the presence of CD45 and by size and granularity) are abundant in

the omentum (Fig. 1A). Approximately 107 lymphocytes (10% of the

cells in the stomovascular fractions (SVF)) were recovered from 50g

of omental tissue, and of these, 80% (range 60–95%) were CD3+

T cells (Fig. 1B) and 10% (5–16%) were NK cells (CD56+CD3�) and

o2% (0–1.5%) were CD19+ B cells. Immunohistochemical analysis

of CD45 and CD3 expression in paraffin-embedded sections of

omental tissue indicated that T cells are predominantly located

between neighboring adipocytes, where macrophages are normally

found [19, 21] and in the perivascular areas (Fig. 1C). Phenotypic

analysis of omental T cells indicated that the majority (mean 72%)

expressed the NK-cell-associated molecule CD56. This compares with

CD3+CD56+ cell frequencies of o5% of matched peripheral blood

lymphocytes (p 5 0.0001; Fig. 2A and E) and up to 50% of T cells in

human liver and intestine [27, 28].

The omentum contains large numbers of iNKT cells

Analysis of TCR expression by omental T cells indicated that up to

50% (mean 10%; range 0.6–54%) expressed the invariant TCR

present on iNKT cells, as detected by the 6B11 mAb, which

detects the CDR3 loop of the Va24Ja18 TCR a-chain [29]. This is

a striking accumulation of iNKT cells when compared with 0.05%S

ide

scat

ter

A B

81%

C

CD45

10 10 10 10 10 10 10 10 10 10

Nu

mb

er o

f ce

llsCD3

Figure 1. Human omentum contains multiple leukocytes. (A) Flowcytometry dot plot showing the expression of CD45 by total omentalstromovascular cells. (B) Flow cytometry histogram showing expres-sion of CD3 after electronically gating on CD45+ cells with low sidescatter (lymphocytes). The number indicates the % of CD45+ lympho-cytes that are CD3+. (A) and (B) are representative of 39 samples.(C) Immunohistochemical localization of CD45+ cells (top) and CD3+

cells (bottom) in omental tissue showing their association withadipocytes (top and bottom left) and in perivascular areas (bottomright). Arrows indicate CD3+ cells (brown staining). Magnification� 1000 (left) and �2000 (right).

Eur. J. Immunol. 2009. 39: 1893–1901Lydia Lynch et al.1894


(0.02–0.8%) of T cells in matched blood (p 5 0.0001; Fig. 2B and

F), o0.1% of intestinal [28] and bone marrow T cells and o1%

of liver T cells [16, 17]. The high frequencies of iNKT cells

observed were confirmed in some samples by flow cytometric

detection of cells expressing the Va24 and Vb11 TCR chains

(Fig. 2C) and using an a-GC-loaded CD1d tetramer (Fig. 2D).

Therefore, the omentum contains the largest density of iNKT cells

known at any location in the human body [1, 2, 16, 17, 28].

CD1d is constitutively expressed by the majority ofomental leukocytes

Since iNKT cells were found to be present in high numbers in

human omentum, we investigated whether CD1d, the antigen-

presenting ligand for the iNKT cells, is expressed in omental

tissue. RT-PCR analysis of mRNA isolated from whole omental

tissue indicated the presence of CD1d mRNA in omentum from

healthy subjects as well as in samples of liver and colon, but not

in C1R cells expressing transfected CD1a (Fig. 3A). Using flow

cytometry, CD1d protein was found on the surface of 15% (range

1–55%) of omental hematopoietic (CD45+) cells compared with

1% (range 0.4–3.5%) of CD45+ cell in matched blood (po0.001;

Fig. 3B). Analysis of leukocyte subsets indicated that CD1d was

also expressed by about 50% of CD14+ cells and�10% of CD3+

cells from human omentum (data not shown). These data

confirm that, as well the omentum being a site of accumulation

of iNKT cells, a large proportion of omental leukocytes

constitutively express CD1d at the cell surface.

Omental iNKT cells kill CD1d-expressing and NK-sensitive target cells

Murine and human iNKT-cell clones kill CD1d-expressing target

cells and allogeneic and autologous tumor cells in vitro [8]. We

0.05%

V24

IgG2a

IgG

1

V 11

23.8%

3.7%

CD

3

77%

CD56IgG1

IgG

1

A

E F

B

p=0.0001

Blood Omentum0

20

40

60

80

100

% o

f ly

mp

ho

cyte

s

CD56+ T cells

% o

f C

D3+

cel

ls

Omentum0

10

20

30

40

50

60

Blood

iNKT cellsp=0.0001

Tetr

amer

CD3

1%

-GC

-lo

aded

tetr

amer

8.1%

CD3

C

DNKT cell OmentumUnloaded

tetramer

Blood OmentumIg controls

IgG

1

IgG1

0.09% 19%

6B11

CD3

Figure 2. CD3+CD56+ T cells and Va24Vb11+ iNKT cells accumulate inhuman omentum. (A) Flow cytometric analysis of Ig isotype controlstaining (left) and CD3 and CD56 expression by lymphocytes fromblood (center) and omentum (right). (B) Ig control staining (left) andCD3 and 6B11 staining of lymphocytes from blood (center) andomentum (right). (C) Ig control staining (left) and TCR Va24 and Vb11chain expression by CD3+ T cells from blood (center) and omentum(right). (D) Flow cytometric analysis of CD3 expression and unloadedCD1d tetramer staining by omental lymphocytes (left) and CD3 and a-GC-loaded tetramer staining by a human peripheral blood iNKT-cellline (center) and omental lymphocytes (right). (E) and (F) Scatterplotshowing the frequencies of CD56+ T cells (E) and iNKT cells defined bypositivity for CD3 and 6B11 (F) in blood and omentum of 39 individuals.All Ig isotype control and unloaded CD1d tetramer plots used omentalcells. Numbers in quadrants indicate the percentages of lymphocytes(A) or T cells (B–D) with the indicated phenotypes. Horizontal lines in(E) and (F) indicate means.

Figure 3. CD1d is constitutively expressed in human omentum.(A) RT-PCR analysis of GAPDH (top) and CD1d (bottom) in snap-frozenand homogenized samples of healthy liver, colonic tissue taken from acolonic carcinoma patient, and omentum. CD1a- and CD1d-trans-fected C1R cells were included as controls (representative of threesamples). (B) Flow cytometric histogram showing Ig isotype controlstaining (unfilled histograms) and cell-surface CD1d expression (grayhistograms) by CD45+ cells in blood (left) and two omental samples(representative of 21 samples). Numbers show % positivity for CD1d.

Eur. J. Immunol. 2009. 39: 1893–1901 Innate immunity 1895


investigated whether human omental iNKT cells display similar

functional activities. Enriched preparations of iNKT cells were

obtained by indirect magnetic bead selection of omental cells that

bind to the 6B11 mAb [29]. Enriched iNKT cells from omental

samples (n 5 4) were used as effectors in cytotoxicity assays

against C1R cells expressing transfected CD1d [4]. Non-iNKT-cell

fractions were used as controls. Addition of iNKT cells at effector/

target ratios of 15:1, in the absence of glycolipid, resulted in the

death of 12% of CD1d+ targets compared with 6% when non-

iNKT cells were added (p 5 0.02). Addition of a-GC caused a

dramatic increase in cytotoxic activity (mean of 48% of target

cells killed, p 5 0.01, Fig. 4A). Low or no cytotoxicity was

seen when mock-transfected C1R cells were used as targets (data

not shown), indicating a dependence on both CD1d and lipid

antigen. Thus, omental iNKT cells display CD1d-dependent

cytotoxic activity similar to murine and human iNKT-cell clones.

Omental iNKT cells (n 5 4) also lysed the NK-sensitive target

cell line K562 in a dose-dependent manner, in the absence of

CD1d or added glycolipid (Fig. 4B) indicating that these cells

display antigen-non-specific cytotoxic activity against certain

tumor cells.

Omental iNKT cells exhibit dual Th1 and Th2 cytokinesecretion

Murine and human iNKT-cell clones secrete large amounts of the

Th1 cytokine IFN-g and/or the Th2 cytokine IL-4 [4, 9]. We

investigated whether enriched omental iNKT cells could release

IFN-g or IL-4 upon stimulation with CD1d+ C1R cells in the

absence and presence of a-GC. Upon co-culture with CD1d+ C1R

cells, but not mock-transfected C1R cells, omental iNKT cells

released significant amounts of IFN-g and IL-4 in the absence of

a-GC (po0.05, n 5 4), although the amounts of both cytokines

released were significantly enhanced when a-GC was added

(po0.05, Fig. 4C). Thus, human omental iNKT cells display

similar activation requirements and dual Th1/Th2 cytokine

secretion profiles to previously described murine and human

iNKT-cell lines and clones [2, 4, 30].

Omental iNKT-cell numbers are lower in patients withobesity and cancer

Hepatic expression of CD1d and hepatic iNKT-cell numbers are

lower in fatty liver compared with control liver [31]. Further-

more, peripheral and hepatic iNKT-cell numbers are reduced in

Figure 4. Omental iNKT cells display CD1d-dependent and naturalcytotoxicity and exhibit dual Th1 and Th2 cytokine secretion. (A) and(B) Cytotoxic activities of magnetic bead-enriched omental iNKT andnon-iNKT lymphocytes against C1R cells expressing transfected CD1d(A) and K562 cells (B) in the absence or presence of a-GC. Results aremeans of four omental samples. Error bars show standard deviations.(C) IFN-g (left) and IL-4 (right) release by magnetic bead-enrichedomental iNKT and non-iNKT omental T cells in response to C1R cellsexpressing transfected CD1d in the absence and presence of a-GC.Results are expressed as means and standard deviations of fourindependent experiments. �po0.05 using Student’s t-test.

Figure 5. Omental iNKT-cell numbers and CD1d expression levels are lower in patients with obesity and colorectal cancer. (A) Scatter plotsshowing the percentages of omental T cells prepared from lean healthy individuals and patients with obesity and colorectal carcinoma (cancer)that express Va24Ja18 TCR chains, which define iNKT cells (p 5 0.005, normal versus obese and p 5 0.004, normal versus cancer, using ANOVA).(B) Scatter plots showing the percentages of peripheral and omental hematopoietic (CD45+) cells prepared from lean healthy individuals andpatients with obesity and colorectal carcinoma (cancer) that express CD1d (p 5 0.04, normal versus obese, using ANOVA). Horizontal lines indicatemeans. (C) Correlation of omental iNKT cell and CD1d+ cell frequencies (p 5 0.001 using Pearsons correlation).



several cancers [17, 32], for which obesity is a risk factor [25,

26]. We investigated whether omental iNKT-cell numbers or

CD1d expression by omental CD45+ cells is altered in patients

with obesity (n 5 15) or colonic cancer (n 5 16), compared with

lean healthy individuals (n 5 8). Figure 5A shows that omental

iNKT-cell frequencies are significantly lower in patients with

obesity (mean 5.5%, range 0.6–22%) compared with controls

(19.4%, 8.8–56%; p 5 0.005). Similar decreases in iNKT-cell

frequencies were found in the cancer patients (4.5%, 2.0–8.7%;

p 5 0.004). The frequencies of CD1d expression by hematopoietic

cells in omentum were slightly lower in obese individuals (7.5%,

1.1–26%) compared with lean healthy individuals (11.4%,

4.5–15.8%) and significantly lower in the cancer patients

(4.3%, 0.9–12%; p 5 0.04) (Fig. 5B). Indeed, omental CD1d

expression correlated positively with the frequencies of omental

iNKT cells (p 5 0.001, Fig. 5C).

The decreased iNKT and CD1d+ cell percentages in omentum

from obese and cancer patients could result from an expansion or

influx of another cell population, rather than a decrease in iNKT

or CD1d+ cell numbers. To investigate this possibility, we

compared the frequencies of several other cell populations in the

three subject groups. We found that the frequencies of iNKT cells

and CD1d+ cells, only, but not NK cells or CD4+, CD8+ or CD56+

T cells, were decreased in obese and cancer patients (Table 1).

We also calculated the absolute numbers of iNKT cells present per

gram of omental tissue from six healthy, 13 obese and seven

cancer patients and found that iNKT-cell numbers were slightly

lower in obesity (1.3� 105 iNKT cells/g compared with

2.3�105/g in healthy individuals; not significant) and signifi-

cantly lower in cancer patients (0.7�105/g; po0.05). These

data provide strong evidence that omental iNKT-cell numbers and

CD1d expression are specifically decreased in patients with

obesity and cancer.

Discussion

Until recently, the omentum has not received much attention as

an organ of the immune system, being predominantly composed

of adipose and connective tissue and thought of as a reserve of

lipids for energy storage. However, analyses of omental milky

spots have revealed abundant macrophages, B cells, T cells, mast

cells and dendritic cells [19, 20], many of which are thought to

develop locally [20]. Furthermore, adipocytes themselves can

express toll-like receptors and secrete mediators of inflammation

and immunity [33]. Indeed the omentum displays the phenom-

enon of ‘‘creeping fat’’, being able to attach to sites of

inflammation, infection and injury, where it contains infection,

promotes vascularization and contributes to tissue regeneration

[18, 21]. The unusual immunological properties of the omentum

and its role as a fat depot prompted us to investigate whether

immune recognition of lipids is a feature of the omentum.

Analysis of the SVF of 39 human omental samples indicated that

lymphocytes are abundant in the omentum and of these �80%

are T cells. Phenotypic analysis of omental T cells indicated that

the majority express CD56, an adhesion molecule that is present

on a subset of memory T cells that are capable of MHC-

unrestricted cytotoxicity and rapid cytokine secretion [34].

CD56+ T cells generally account for o5% of peripheral T cells,

but are enriched in human liver [27], and intestine [28]. Analysis

of TCR expression by omental T cells indicated that up to

30%, and 450% in one individual, express the Va24Ja18 TCR

chain found on iNKT cells. This high iNKT frequency was

confirmed in some samples by detecting co-expression of the

Va24 and Vb11 chains and using a a-GC-loaded CD1d tetramer.

Although found in significant numbers in mice [2], iNKT cells

account for o0.1% of peripheral, intestinal and bone marrow

T cells and o1% of liver T cells in humans [16, 17, 28].

Therefore, the omentum contains the highest density of iNKT

cells known at any location in the human body. A previous study

in mice [35] has reported elevated NK1.1+ T-cell numbers in

epididymal fat tissue, but this study did not determine if these

cells are iNKT cells.

The ligand recognized by iNKT cells is composed of self or

foreign lipid-based antigens bound to the CD1d glycoprotein

[1, 2]. CD1d is expressed by APC, such as dendritic cells,

macrophages and B cells, as well as various epithelial, parench-

ymal and vascular smooth muscle cells [1, 36, 37] and its

expression can be altered in cancer, autoimmunity and infectious

disease [36]. We found that CD1d is constitutively expressed by

omental hematopoietic cells, including monocytes and T cells.

Furthermore, the frequencies of omental CD1d+ cells correlated

positively with the frequencies of iNKT cells in different indivi-

duals, suggesting that omental iNKT-cell numbers may be directly

governed by local cell-surface expression of CD1d. If this is true,

the omentum may prove to be a source of physiologically relevant

autoantigen recognized by iNKT cells, the identity of which is

controversial [38].

Table 1. Frequencies of omental cell subsets (% of lymphocyte) and CD1d+ cells (% 0f CD45+ cells) in lean healthy individuals (n 5 6) and in patientswith obesity (n 5 15) and colorectal cancer (n 5 16)

Patient group NK cells (CD3�CD56+) T cells (CD3+) CD4 T cells CD8+ T cells CD56 T cells iNKT cells CD1d+ cells

Healthy 13.2 85.5 57.0 26.0 49.9 19.4 11.4

Obese 16.1 81.2 59.7 31.1 61.9 5.46� 7.5

Cancer 20.3 76.7 51.5 31.5 55.4 4.5�� 4.3��

�p 5 0.005; ��p 5 0.004; ��p 5 0.04.



The iNKT cells are thought to be innate lymphocytes that are

capable of rapid cytotoxicity [8] and rapid cytokine secretion [2,

4, 9]. Most notably, they have a unique capability to secrete

cytokines with opposing effects on adaptive immune responses,

including those released by Th1, Th2, Treg and Th17 cells [4, 9,

39, 40]. These functional properties of iNKT cells place them as

innate immune effector cells as well as regulators of adaptive

immune responses. We found that human omental iNKT cells

have similar functional activities. In cytotoxicity assays, enriched

preparations of omental iNKT cells, but not non-iNKT cells, killed

CD1d-transfected C1R target cells pulsed with a-GC. They did not

kill CD1d� C1R cells nor CD1d+ cells in the absence of a-GC,

indicating a dependence on both CD1d and lipid antigen.

However, omental iNKT cells also killed the NK-sensitive cell line,

K562 in a manner that is independent of CD1d or added glyco-

lipid. Although not tested in the present study, this cytotoxicity is

likely to be triggered by the ligation of any of a number of

stimulatory receptors that activate NK cells and some T cells,

including iNKT cells [9, 41].

The omental iNKT-cell population was also found to be capable

of dual Th1/Th2 cytokine production. Upon co-culture with

CD1d+ C1R cells, they released IFN-g and IL-4 in the absence of a-

GC, but addition of a-GC resulted in increased production of both

cytokines. The iNKT-cell recognition of CD1d in the absence of

added antigen has previously been reported [4, 30] and it has been

proposed that self-antigen reactivity may prime iNKT cells to

respond rapidly to pathogens and tumors and to subsequently

activate and regulate other cells of the immune system [30].

Indeed, this immune ‘‘kick-start’’ mechanism appears to mainly be

mediated via the secretion of cytokines [2, 4, 9, 39, 40]. Cytokine

release by iNKT cells may also play a role in the anti-inflammatory

and healing properties of the omentum. Murine iNKT cells can

produce immunosuppressive cytokines, such as IL-4, IL-10 and IL-

13, within 1–2 h of activation [9, 39]. They can induce the

generation of CD8+ T regulatory cells and can prevent auto-

immune disease and allograft rejection [39]. As well as being anti-

tumor effectors [11–13], iNKT cells can suppress anti-tumor

immunity in some mouse models [42]. Further studies on the

mechanisms by which omental tissue can promote healing and

suppress inflammation are required before iNKT cells can be

implicated in this phenomenon.

CD1d-restricted iNKT cells play key roles in anti-tumor

defense in murine models. Injection of mice with a-GC can

prevent and reverse tumor growth [12, 13] and mice deficient in

CD1d or iNKT cells fail to mediate IL-12-induced tumor rejection

[11]. Numerical and functional deficiencies of iNKT cells have

been reported in humans with tumors [17, 32, 43] and therapies

involving iNKT cells are being tested in clinical trials for cancer

[14, 15]. A potential role for omental iNKT cells in tumor

immunity is suggested by our observation that patients with

colorectal carcinoma had 3–4 times less iNKT cells and 2–3 times

less CD1d-expressing cells than control subjects. It is possible that

depletions of these IFN-g-producing, cytotoxic cells may predis-

pose individuals to malignancy, however, it remains unknown

whether this reduction in omental iNKT cells in cancer patients

contributes to or is a result of malignancy. The omentum may in

the future prove to be a valuable source of iNKT cells for adoptive

immunotherapy for cancer.

We also report that severely obese individuals had lower

frequencies of omental iNKT cells compared with healthy

subjects. Obesity is associated with a compromised immune

system and an increased susceptibility to infection [24] and risk

of developing several malignancies [25, 26]. Obesity is also

frequently accompanied by inflammation within adipose tissue,

including macrophage accumulation, the release of pro-inflam-

matory cytokines and expression of receptors that mediate innate

immune responses by adipocytes [44]. Whether iNKT cells are

involved in the development of obesity is unknown, but liver and

splenic iNKT numbers and functions are reported to be altered in

mice on high fat diets [45] and CD1d expression and iNKT-cell

numbers are lower in mice and humans with fatty liver disease

compared with controls [31].

In conclusion, we have shown that human omentum harbors

large populations of CD1d-restricted iNKT cells with potent cytotoxic

activity and rapid cytokine secretion. This concentration of a lipid-

reactive immune system with important anti-tumor potential in a

migrating organ that can control inflammation and tissue damage

suggests a unique immuneoregulatory and/or anti-metastatic role for

the omentum. Consistent with this idea, omental iNKT cells were

significantly depleted in patients with malignancy and obesity, the

latter condition being associated with an increased risk of cancer.

Omental tissue is accessible and can be transposed for surgical

reconstruction, and with the highest density of functional iNKT cells

known in the human body, it may provide a novel source of anti-

tumor effectors for use in adoptive immunotherapy.

Materials and methods

Subjects

Omental biopsies and matched peripheral blood were

collected from 39 consenting patients, including 15 severely

obese patients (body mass index 440) undergoing gastric

bypass surgery (mean age 48, mean body mass index 56, six

males, nine females), 16 patients with colorectal cancer under-

going resection (mean age 65, nine males, seven females)

and eight lean non-malignant patients undergoing elective

surgery for hernia repair (n 5 2), rectal prolapse (n 5 1), colonic

blockage (n 5 1), or benign polyp removal (n 5 4). Ethical

approval for this study was obtained from St. Vincent’s University

Hospital.

Isolation of leukocytes from omental and bloodsamples

Omental biopsies were collected in DMEM/Nutrient Mixture

F-12 (Gibco, Paisley, UK) supplemented with antibiotics



(100 U/mL penicillin, 100 mg/mL streptomycin) and 5% FCS

(Invitrogen Life Technologies, Paisley, UK). Approximately 2 g of

tissue was snap-frozen and stored in liquid nitrogen for PCR

analysis. The remaining tissue was processed immediately to

obtain single-cell suspensions of the omental SVF. Omental

samples were minced and incubated with 2 mg/mL type II

collagenase I (Sigma Aldrich, Poole, UK) in DMEM, shaking at

371C for 60 min. Samples were washed and passed twice through

a 350mm nylon mesh (Sefar, Lancashire, UK) to remove

adipocytes. The cell suspensions were centrifuged at 500� g for

10 min and the pellet was lysed with FACS lysing solution (BD

Biosciences, Oxford, UK) to remove erythrocytes. Cell suspension

was passed through a 30 mm nylon mesh and centrifuged at

500� g for 10 min. The pellet was resuspended and cell yields

and viability were assessed by ethidium bromide/acridine orange

staining.

Antibody and tetramer staining and flow cytometry

Fluorochrome-labeled mAb specific for human CD1d (clone

CD1d42), CD3 (SK7), CD4 (RPA-T4), CD8 (HIT8a), CD11c

(B-ly6), CD14 (M5E2), CD19 (HIB19), CD34 (581), CD56

(B159), the iNKT TCR (6B11) and isotype-matched controls

were obtained from BD Biosciences. mAb specific for the Va24

(C15) and Vb11 (C21) TCR chains were obtained from Coulter

Immunotech (Marseilles, France). Following staining [27], cells

were analyzed using a FACSCalibur flow cytometer using

CellQuest software (BD Biosciences). CD1d tetramers were

loaded with a-GC (Alexis Biochemicals) and used to stain cells

as per manufacturers’ instructions (Proimmune, Oxford, UK).

Immunohistochemistry

Paraffin-embedded human omental sections were obtained from

ten patients. Deparaffinizing, antigen retrieval, washing and

staining were carried out according to standard procedures [28].

Tissues were blocked with endogenous peroxidase (Vector

Laboratories, Peterborough, UK) and stained with mAb specific

for human CD45 (UCHL1) or CD3 (UCHT1) (Southern Biotech-

nology Associates, Birmingham, UK, Al; 1/50 dilution in PBS).

Analysis of CD1d mRNA expression

RT-PCR for CD1d and GAPDH was carried out on snap-frozen,

homogenized samples of omentum, liver, colon and (as controls)

CD1a- and CD1d-transfected C1R cells. Total RNA was extracted

from homogenized tissue using the RNeasy lipid tissue kit

(Qiagen, Crawley, UK) according to the manufacturer’s instruc-

tions. RNA was reverse transcribed to cDNA using Omniscript

reverse transcriptase (Qiagen). cDNA was used as a template

for the amplification of full-length CD1d using the primers

50-CTGTTTCTGCTGCTCTGGG-30 and 50-AGAGACACAGATGTGG-

CAAGG-30. PCR was performed in 50 mL volumes containing

100 ng cDNA, 0.1 mg each primer, 350mM of each dNTP in a

reaction buffer containing 1.5 mM MgCl2 and 1 U AmpliTaq Gold

DNA polymerase (Applied Biosystems, Warrington, UK). Thermo-

cycling conditions consisted of an initial incubation at 951C to

activate the polymerase enzyme, followed by 35 cycles consisting

of 45 s at 951C, 45 s at 601C and 45 s at 721C. PCR products

were separated in 2% agarose gels and visualized by ethidium

bromide staining. As a control, the housekeeping gene GAPDH

(50-GCCTCAAGATCATCAGCAA and 50-CCAGCGTCAAAGGTG-

GAG) was used. PCR products were separated in 2% agarose

gels.

Cytotoxicity assays

The iNKT cells were isolated from omental SVF of five patients by

staining with a phycoerythrin-labeled anti-iNKT mAb (6B11) and

positive selection using anti-PE mAb-coated magnetic beads

(Miltenyi Biotec, Gladbach Bergische, Germany). Cytotoxicity

was measured using the Total Cytotoxicity Detection Kit

(Immunochemistry, Bloomington, MN, USA). CFSE-labeled

target cells (mock-transfected or CD1d-transfected C1R cells

or K562 cells) were added to the iNKT-cell populations at

various effector/target ratios and incubated at 371C for 4 h, in

the absence or presence of 100 ng/mL a-GC. To detect killing,

7-aminoactinomycin D was added immediately and analyzed by

flow cytometry.

Analysis of cytokine production

The iNKT cells were positively selected from omental samples as

described above and stimulated for 72 h with equal numbers of

mock-transfected or CD1d-transfected C1R cells in the absence or

presence of 100 ng/mL a-GC. IFN-g and IL-4 levels in the culture

supernatants were determined by ELISA using antibody pairs

purchased from R&D Systems (Abingdon, UK).

Statistical analysis

Differences between groups were assessed using the Mann–

Whitney U-test or one-way ANOVA when three groups were

compared. The p-values ofo0.05 were considered significant.

Relationships between iNKT-cell and CD1d+ cell numbers were

investigated using Pearson’s correlation coefficient.

Acknowledgements: We thank Emma McGrath and Anna

Kwasnik for the technical support and Andrew Hogan for

providing the iNKT-cell line. The mock- and CD1d-transfectant

cells were a kind gift from Dr. Steven Porcelli, Albert Einstein



College of Medicine, New York, USA. This work was supported by

the Health Research Board, Ireland.

Conflict of interest: The authors declare no financial or

commercial conflict of interest.

References

1 Brigl, M. and Brenner, M. B., CD1: antigen presentation and T cell

function. Annu. Rev. Immunol. 2004. 22: 817–890.

2 Bendelac, A., Savage, P. B. and Teyton, L., The biology of NKT cells. Annu.

Rev. Immunol. 2007. 25: 297–336.

3 Gumperz, J. E., The ins and outs of CD1 molecules: bringing lipids under

immunological surveillance. Traffic 2006. 7: 2–13.

4 Exley, M., Garcia, J., Balk, S. P. and Porcelli, S., Requirements for CD1d

recognition by human invariant Va241 CD4–CD8-T cells. J. Exp. Med. 1997.

186: 109–120.

5 Zhou, D., Mattner, J., Cantu, C., III, Schrantz, N., Yin, N., Gao, Y., Sagiv, Y.

et al., Lysosomal glycosphingolipid recognition by NKT cells. Science 2004.

306: 1786–1789.

6 Kinjo, Y., Wu, D., Kim, G., Xing, G. W., Poles, M. A., Ho, D. D., Tsuji, M.

et al., Recognition of bacterial glycosphingolipids by natural killer T cells.

Nature 2005. 434: 520–525.

7 Kawano, T., Cui, J., Koezuka, Y., Toura, I., Kaneko, Y., Motoki, K., Ueno, H.

et al., CD1d-restricted and TCR-mediated activation of Va14 NKT cells by

glycosylceramides. Science 1997. 278: 1626–1629.

8 Metelitsa, L. S., Naidenko, O. V., Kant, A., Wu, H. W., Loza, M. J., Perussia,

B., Kronenberg, M. et al., Human NKT cells mediate antitumor cytotoxi-

city directly by recognizing target cell CD1d with bound ligand or

indirectly by producing IL-2 to activate NK cells. J. Immunol. 2001. 167:

3114–3122.

9 Gumperz, J. E., Miyake, S., Yamamura, T. and Brenner, M. B., Functionally

distinct subsets of CD1d-restricted natural killer T cells revealed by CD1d

tetramer staining. J. Exp. Med. 2002. 195: 625–636.

10 Vincent, M. S., Leslie, D. S., Gumperz, J. E., Xiong, X., Grant, E. P. and

Brenner, M. B., CD1-dependent dendritic cell instruction. Nat. Immunol.

2002. 3: 1163–1168.

11 Cui, J., Shin, T., Kawano, T., Sato, H., Kondo, E., Toura, I., Kaneko, Y. et al.,

Requirement for Va14 NKT cells in IL-12-mediated rejection of tumors.

Science 1997. 278: 1623–1626.

12 Kawano, T., Cui, J., Koezuka, Y., Toura, I., Kaneko, Y., Sato, H., Kondo, E.

et al., Natural killer-like nonspecific tumor cell lysis mediated by specific

ligand-activated Va14 NKT cells. Proc. Natl. Acad. Sci. USA 1998. 95:

5690–5693.

13 Nakagawa, R., Nagafune, I., Tazunoki, Y., Ehara, H., Tomura, H., Iijima,

R., Motoki, K. et al., Mechanisms of the antimetastatic effect in the liver

and of the hepatocyte injury induced by a-galactosylceramide in mice.

J. Immunol. 2001. 166: 6578–6584.

14 Chang, D. H., Osman, K., Connolly, J., Kukreja, A., Krasovsky, J., Pack, M.,

Hutchinson, A. et al., Sustained expansion of NKT cells and antigen-

specific T cells after injection of a-galactosyl-ceramide loaded mature

dendritic cells in cancer patients. J. Exp. Med. 2005. 201: 1503–1517.

15 Motohashi, S., Nagoto, K., Kunii, N., Yamamoto, H., Yamasaki, K., Okita,

K., Nakayama, T. et al., A phase I-II study of alpha-galactosylceramide-

pulsed IL-2/GM-CSF-cultured peripheral blood mononuclear cells in

patients with advanced and recurrent non-small cell lung cancer.

J. Immunol. 2009. 182: 2492–2501.

16 Karadimitris, A., Gadola, S., Altamirano, M., Brown, D., Woolfson, A.,

Klenerman, P., Chen, J. L. et al., Human CD1d-glycolipid tetramers

generated by in vitro oxidative refolding chromatography. Proc. Natl.

Acad. Sci. USA 2001. 98: 3294–3298.

17 Kenna, T., Golden-Mason, L., Porcelli, S. A., Koezuka, Y., Hegarty, J. E.,

O’Farrelly, C., and Doherty, D. E., NKT cells from normal and tumor-

bearing human livers are phenotypically and functionally distinct from

murine NKT cells. J. Immunol. 2003. 171: 1775–1779.

18 Platell, C., Cooper, D., Papadimitriou, J. M. and Hall, J. C., The omentum.

World J. Gastroenterol. 2000. 6: 169–176.

19 Shimotsuma, M., Takahashi, T., Kawata, M. and Dux, K., Cellular subsets

of the milky spots in the human greater omentum. Cell Tissue Res. 1991.

264: 599–601.

20 Bedford, P. A., Todorovic, V., Westcott, E. D., Windsor, A. C., English, N. R.,

Al-Hassi, H. O., Raju, K. S. et al., Adipose tissue of human

omentum is a major source of dendritic cells, which lose MHC Class II

and stimulatory function in Crohn’s disease. J. Leukoc. Biol. 2006.

80:546–554.

21 Litbarg, N. O., Gudehithlu, K. P., Sethupathi, P., Arruda, J. A., Dunea,

G. and Singh, A. K., Activated omentum becomes rich in factors that

promote healing and tissue regeneration. Cell Tissue Res. 2007. 328:

487–497.

22 Goldsmith, H. S., The evolution of omentum transposition: from

lymphedema to spinal cord, stroke and Alzheimer’s disease. Neurol. Res.

2004. 26: 586–593.

23 Schaffler, A., Furst, A., Buchler, C., Paul, G., Rogler, G., Scholmerich, J. and

Herfarth, H., Secretion of RANTES (CCL5) and interleukin-10 from

mesenteric adipose tissue and from creeping fat in Crohn’s disease:

regulation by steroid treatment. J. Gastroenterol. Hepatol. 2006. 21:

1412–1418.

24 Falagas, M. E. and Kompoti, M., Obesity and infection. Lancet Infect. Dis.

2006. 6: 438–446.

25 Bianchini, F., Kaaks, R. and Vainio, H., Overweight, obesity, and cancer

risk. Lancet Oncol. 2002. 3: 565–574.

26 Renehan, A. G., Roberts, D. L. and Dive, C., Obesity and cancer:

pathophysiological and biological mechanisms. Arch. Physiol. Biochem.

2008. 114: 71–83.

27 Doherty, D. G., Norris, S., Madrigal-Estebas, L., McEntee, G., Traynor, O.,

Hegarty, J. E. and O’Farrelly, C., The human liver contains multiple

populations of NK cells, T cells, and CD31CD561 natural T cells with

distinct cytotoxic activities and Th1, Th2, and Th0 cytokine secretion

patterns. J. Immunol. 1999. 163: 2314–2321.

28 O’Keeffe, J., Doherty, D. G., Kenna, T., Sheahan, K., O’Donoghue, D. P.,

Hyland, J. M. and O’Farrelly, C., Diverse populations of T cells with NK cell

receptors accumulate in the human intestine in health and in colorectal

cancer. Eur. J. Immunol. 2004. 34: 2110–2119.

29 Exley, M. A., Hou, R., Shaulov, A., Tonti, E., Dellabona, P., Casorati, G.,

Akbari, O. et al., Selective activation, expansion, and monitoring of

human iNKT cells with a monoclonal antibody specific for the TCR

a-chain CDR3 loop. Eur. J. Immunol. 2008. 38: 1756–1766.

30 Vincent, M. S., Gumperz, J. E. and Brenner, M. B., Understanding the

function of CD1-restricted T cells. Nat. Immunol. 2003. 4: 517–523.

31 Yang, L., Jhaveri, R., Huang, J., Qi, Y. and Diehl, A. M., Endoplasmic

reticulum stress, hepatocyte CD1d and NKT cell abnormalities in murine

fatty livers. Lab. Invest. 2007. 87: 927–937.

32 Molling, J. W., Kolge, W., van der Vliet, H. J., Boomsma, M. F., Kruizenga,

H., Smorenburg, C. H., Molenkamp, B. G. et al., Peripheral blood IFN-g-

secreting Va241Vb111 NKT cell numbers are decreased in cancer



patients independent of tumor type or tumor load. Int. J. Cancer 2005. 116:

87–93.

33 Tilg, H. and Moschen, A. R., Adipocytokines: mediators linking

adipose tissue, inflammation and immunity. Nat. Rev. Immunol. 2006. 6:

772–783.

34 Kelly-Rogers, J., Madrigal-Estebas, L., O’Connor, T. and Doherty, D. G.,

Activation-induced expression of CD56 by T cells is associated with a

reprogramming of cytolytic activity and cytokine secretion profile in vitro.

Hum. Immunol. 2006. 67: 863–873.

35 Caspar-Bauguil, S., Cousin, B., Galinier, A., Segafredo, C., Nibbelink, M.,

Andre, M., Casteilla, L. and Penicaud, L., Adipose tissues as an ancestral

immune organ: site-specific change in obesity. FEBS Lett. 2005. 579:

3487–3492.

36 Dougan, S. K., Kaser, A. and Blumberg, R. S., CD1 expression on antigen-

presenting cells. Curr. Top. Microbiol. Immunol. 2007. 314: 113–141.

37 Brossay, L., Jullien, D., Cardell, S., Sydora, B. C., Burdin, N., Modlin, R. L.

and Kronenberg, M., Mouse CD1 is mainly expressed on hemopoietic-

derived cells. J. Immunol. 1997. 159: 1216–1224.

38 Speak, A. O., Salio, M., Neville, D. C., Fontaine, J., Priestman, D. A., Platt,

N., Heare, T. et al., Implications for invariant natural killer T cell ligands

due to the restricted presence of isoglobotrihexosylceramide in

mammals. Proc. Natl. Acad. Sci. USA 2007. 104: 5971–5976.

39 Godfrey, D. I. and Kronenberg, M., Going both ways: immune regulation

via CD1d-dependent NKT cells. J. Clin. Invest. 2004. 114: 1379–1388.

40 Rachitskaya, A. V., Hansen, A. M., Horai, R., Li, Z., Villasmil, R., Luger, D.,

Nussenblatt, R. B. and Caspi, R. R., Cutting edge: NKT cells constitutively

express IL-23 receptor and RORat and rapidly produce IL-17 upon

receptor ligation in an IL-6-independent fashion. J. Immunol. 2008. 180:

5167–5171.

41 Lanier, L. L., Up on the tightrope: natural killer cell activation and

inhibition. Nat. Immunol. 2008. 9: 495–502.

42 Berzofsky, J. A. and Terabe, M., NKT cells in tumor immunity: opposing

subsets define a new immunoregulatory axis. J. Immunol. 2008. 180:

3627–3635.

43 Dhodapkar, M. V., Geller, M. D., Chang, D. H., Shimizu, K., Fujii, S.,

Dhodapkar, K. M. and Krasovsky, J., A reversible defect in natural killer T

cell function characterizes the progression of premalignant to malignant

multiple myeloma. J. Exp. Med. 2003. 197: 1667–1676.

44 Karagiannides, I. and Pothoulakis, C., Obesity, innate immunity and gut

inflammation. Curr. Opin. Gastroenterol. 2007. 23: 661–666.

45 Miyazaki, Y., Iwabuchi, K., Iwata, D., Miyazaki, A., Kon, Y., Niino, M.,

Kikuchi, S. et al., Effect of high fat diet on NKT cell function and NKT cell-

mediated regulation of Th1 responses. Scand. J. Immunol. 2008. 67:

230–237.

Abbreviations: a-GC: a-galactosylceramide � iNKT-cell: invariant

natural killer T cell � SVF: stomovascular fraction

Full correspondence and current address: Dr. Lydia Lynch, Dept. of

Medicine, Beth Israel Deaconess Medical Centre, Harvard Medical

School, 330 Brookline Avenue, Boston, MA 02215, USA

Fax: 11-617-975-5235

e-mail: [email protected]

Received: 18/2/2009

Revised: 19/3/2009

Accepted: 20/4/2009



Invariant NKT cells and CD1d+ cells amass in human omentum and are depleted in patients with cancer...

Documents

Transcript of Invariant NKT cells and CD1d+ cells amass in human omentum and are depleted in patients with cancer...