Expression of interleukin-18, interferon-γ and interleukin-10 in hepatocellular carcinoma

Expression of interleukin-18, interferon-g and interleukin-10 inhepatocellular carcinoma

Chok Seng Chia a, Kechen Ban a, Hairuszah Ithnin a, Harjit Singh b, R. Krishnan b,Suryati Mokhtar b, Nik Malihan c, Heng Fong Seow a,d,*

a Department of Clinical Laboratory Science, Faculty of Medicine and Health Science, Universiti Putra Malaysia, Serdang, Selangor 43400, Malaysiab Department of Hepato-Pancreato-Biliary Surgery, Selayang Hospital, Selangor, Malaysia

c Department of Pathology, Selayang Hospital, Selangor, Malaysiad Institute of Bioscience, Universiti Putra Malaysia, Serdang, Selangor 43400, Malaysia

Received 7 May 2002; accepted 12 July 2002

Abstract

This is the first report on the detection of IL-18, IFN-g and IL-10 proteins in hepatocelllular carcinoma. In the apparently normal

surrounding tissue, 13 out of 17 paired specimens showed positive immunoreactivity to IL-18 (76.5%) compared with six out of 17 in

the tumour portion (35.3% of specimens). Thus, a significantly higher number of IL-18 positive specimens was found in the

hepatocytes of apparently normal surrounding tissue compared with the tumour (P�/0.018). In contrast, the number of specimens

with positive immunoreactivity to the antibody against the Th1 cytokine, IFN-g expression in the hepatocytes was lower. Only one

specimen from the apparently normal surrounding tissue (one out of 17; 5.9%) and three other specimens from the tumour portion

(three out of 17; 17.6%) had positive immunoreactivity. Similarly, the expression of the Th2 cytokine, IL-10 in normal (four out of

17; 23.5%) and tumour portions (five out of 17; 29.4%) was also low. Thus, there did not appear to be predominant Th2 immune

response as denoted by IL-10 expression. Using the Spearman correlation rank test, a significant correlation between IL-18

expression in the apparently normal surrounding tissue and high a-foetoprotein (AFP) levels of �/350 IU/l. No correlation between

IL-18 expression in the tumour portion and clinicopathological factors was found. There was also no correlation found between IL-

18 and the other cytokines, namely, IFN-g and IL-10 expression These new findings provide additional information on the type of

cytokines expressed in the tumour microenvironment and give a further insight into the role of cytokines in the pathogenesis of

cancer which is critical for the development of effective immunotherapeutic approaches for cancer therapy in the future.

# 2002 Elsevier Science B.V. All rights reserved.

Keywords: Cytokines; Hepatocelllular carcinoma; Immunohistochemistry

1. Introduction

In the last decade, there has been a massive explosion

in our knowledge on the identity and functions of

cytokines. These discoveries have been rapidly applied

to the prevention and treatment of a diverse array of

diseases ranging from malignancies, infectious diseases,

autoimmune disorders and allergies [1]. The cell-

mediated response depends on cytotoxic and helper T

cell activities and function through the actions of

cytokines to regulate cells such as macrophages and

natural killer (NK) cells produced in the cells including

those in the liver are essential in defending the host

against infectious agents such as hepatitis B virus

(HBV). Various cytokines have been documented in

the liver host defence [2], liver injury [3] and liver

regeneration [4]. High levels of macrophage inhibitory

factor (MIF) have been detected in the sera of patients

with hepatocellular carcinoma (HCC) and liver cirrhosis

suggesting an association between the overexpression of

MIF and hepatocarcinogenesis [5]. Other cytokines such

as human monokine induced by interferon (IFN)-g(hMIG), interleukin (IL)-8, macrophage inflammatory

protein-1 (MIP-1)-a and b have also, been found to be

expressed in HCC. These chemokines have been sug-

* Corresponding author. Tel.: �/60-389-48-6101x8511; fax: �/60-

389-42-6957

E-mail address: [email protected] (H.F. Seow).

Immunology Letters 84 (2002) 163�/172

www.elsevier.com/locate/

0165-2478/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.

PII: S 0 1 6 5 - 2 4 7 8 ( 0 2 ) 0 0 1 7 6 - 1

mailto:[email protected]

gested to play a role in the recruitment of lymphocytes

to these tumours in vivo [6]. Thus, cytokines may be

beneficial in the host immune response against tumours.

Several evidence support the hypothesis that a bene-ficial immunological control of tumour growth is

associated with the development of cell-mediated cyto-

toxicity either through specific T cells [7] or non-antigen

specific NK cells [8]. The generation of such cells are

under the control of cytokines such as, IL-2 and IFN-g,

produced by T helper 1 (Th1) cells [9]. IL-18, also

known as IFN-g inducing factor, has been shown to be a

strong co-factor for Th1 development [10]. IL-10, on theother hand, is associated with development of the Th2

response has been shown to inhibit the development of

cellular immune responses via a number of mechanisms

[11]. Thus, tumour cells can also evade the immune

surveillance system by various mechanisms such as the

disruption of Th1/Th2 balance, destruction of immune

effector cells by Fas/Fas ligand and production of

immunosuppressive cytokines such as transforminggrowth factor-b (TGF-b) and IL-10 [12]. In addition,

there is increasing evidence that the tumour microenvir-

onment may also contribute to cancer progression [13].

Since the cytokines expressed at the tumour site may

play important roles in determining a successful immune

response against tumours as well as cancer pathogenesis

and progression, this study has been undertaken to

determine the expression of IL-18, IFN-g and IL-10 inHCC which is a common cancer in Malaysia and other

parts of Asia such as China, Taiwan and Hong Kong. In

Malaysia, a large proportion of HCC occurs in patients

who are chronic carriers of the human HBV [14].

2. Materials and methods

2.1. Production of polyclonal antibodies to IL-18

Recombinant human IL-18 (rhIL-18) produced in our

laboratory was purified as previously described [15].

One New Zealand white rabbit was immunised with 100

mg of rhIL-18 emulsified in an equal volume of complete

Freund’s adjuvant (Sigma, USA). Two booster injec-

tions were administered at fortnightly intervals with 50

mg of rhIL-18 emulsified in incomplete Freund’s adju-vant. Blood (4 ml) was collected from the ear vein prior

to immunisation. Subsequent bleeds were collected at 2,

4 and 8 weeks after the first immunisation. The

immunoglobulin fraction IgG was purified from whole

serum using Thiophilic adsorption chromatography

(Clontech, USA). Briefly, 1 ml of serum was diluted in

9 ml of sample buffer (50 mM sodium phosphate; 0.55

M sodium sulphate pH 7.0). Thiophilic resin (1 ml) waswashed with ten column volumes of equilibration buffer

(50 mM sodium phosphate; 0.5 M sodium sulphate

pH7.0). The diluted serum was then adsorbed to the

resin and after extensive washings, the IgG was eluted in

5 ml of 50 mM sodium phosphate.

2.2. Tissues

Seventeen pairs of formalin-fixed, paraffin-embedded

liver tumour and the adjacent tissue from confirmed

HCC cases were collected from Selayang Hospital,

Selangor dated from September 2000 till August 2001.

This hospital is the National Reference Centre for Liver

Diseases in Malaysia.

2.3. Immunohistochemistry

Sections of 5 mm thickness were cut from the paraffin

blocks, dried at 37 8C, deparaffinized in xylene and

rehydrated in graded ethanols before the antigen

retrieval step by microwave treatment for 10 min at

560 W.

The tissue sections were quenched with peroxidase

block for 10 min, washed three times with phosphate-buffered saline (PBS) and then incubated for 30 min

with the 100-fold diluted polyclonal antibody to IL-18

produced as described in Section 2.1. After three washes

with PBS, the tissue sections were incubated with

streptavidin and biotin secondary antibody (LSAB kit,

DAKO Ltd., USA) for 15 min, washed in PBS and then

incubated with DAB plus substrate chromogen (DAKO

Ltd., USA) for 5 min. The sections were then rinsed indistilled water followed by counterstaining with haema-

toxylin. With the negative control sections, the primary

antibody was substituted with PBS. For the detection of

IFN-g and IL-10, the monoclonal antibodies were

purchased from Santa Cruz Technology, USA. The

tissue sections were incubated with diluted normal

blocking serum for 1 h followed by these antibodies

(1:50 dilution) for 30 min in a wet chamber. They werethen incubated with goat-anti-mouse secondary anti-

body in the DAKO EnvisionTM kit for 30 min at room

temperature followed by incubation with 3,3-diamino-

benzidine (DAB) plus substrate chromogen for 5 min.

The sections were then rinsed in distilled water followed

by counterstaining with haematoxylin.

2.4. Scoring of immunoreactivity

The immunoreactive score was determined by the sum

of extension and intensity as previously reported [16,17].

The intensity of staining was scored as 0 for negative, 1

for weakly postive, 2 for moderately positive and 3 for

strongly positive. The extent of positive immunoreactiv-

ity was estimated as 0 for negative, 1 for positive

immunoreactivity for 1�/25% of the cells, 2 for 26�/

50%, 3 for 51�/75% and 4 for 76�/100%. The combined

staining score (score from intensity plus score from the

extent of positive immunoreactivity) ]/3 was considered

C.S. Chia et al. / Immunology Letters 84 (2002) 163�/172164

as positive immunoreactivity. The slides were viewed

using the digitalised Olympus BX50 microscope and

images were captured at selected magnification using the

Olympus pm10-sp35 camera attached.

2.5. Statistical analysis

Data analysis was performed with the Statistical

Package of Social Science (SPSS) v10.0.1. The signifi-

cance differences between categorical variables werecompared using the Fischer’s exact probability test.

Values of P B/0.05 were considered to be statistically

significant. Correlation between factors was evaluated

using the Spearman rank correlation coefficient test.

3. Results

3.1. Higher IL-18 expression in apparently normal

adjacent tissues

IL-18 was detected in the hepatocytes of the appar-ently normal tissue adjacent to the tumour (13 of the 17

specimens; Table 1) whereas in the tumour portion, only

six of the 17 specimens showed positive immunoreactiv-

ity (Table 1). Fig. 1 shows the representative slides for

positive immunoreactivity to the antibody against IL-

18. The specimens considered to have intense (three

specimens), moderate (six specimens), weak (four speci-

mens) and negative immunoreactivity in the apparentlynormal tissue surrounding the tumour are as shown in

Fig. 1a, c, d and f, respectively. Representative slides

showing IL-18 immunoreactivity in the tumour portion

for moderate (one specimen) and weak (five specimens)

immunostaining are as shown in Fig. 1b and e,

respectively.

There was a 2-fold higher number of cases with IL-18

positive immunoreactivity in the apparently normal

tissue compared with the tumour portion. Statistical

analysis showed that this was significant (P�/0.018).Analysis of the case distribution of IL-18 immunoreac-

tivity in the adjacent tissue with the extent of the area

versus intense and moderate immunostaining showed

that not only was there a higher rate of immunoposi-

tivity for IL-18 (nine cases), the extent of the intense

immunostaining was large (76�/100%) compared with

the tumour portion (one case with moderate staining

over 51�/75% of area; five with weak staining) (Fig. 2).The correlation between IL-18 expression with sex,

age and race (Table 1) as well as with clinicopathological

factors such as tumour size, infection with HBV,

cirrhosis and AFP levels was made (Table 2). Using

the Spearman correlation rank test, it was found that

there was a significant correlation between IL-18

immunoreactivity in the apparently normal tissue and

high AFP levels�/350 IU/l (P�/0.015) (Table 2). How-ever, no correlation between IL-18 expression in the

tumour and the clinicopathological factors was found.

There was also no correlation between IL-18 and IFN-gor IL-10 expression (Table 3).

3.2. Low incidence of expression of IFN-g in apparently

normal and tumour tissue

In three out of the 17 paired specimens, positive

immunoreactivity to IFN-g monoclonal antibody was

detected in the hepatocytes of the adjacent normal tissue(Fig. 3a) and only in one out of the 17 tumour tissue

(Fig. 3b). The negative controls for the normal and

tumour tissues are shown in Fig. 3c and d, respectively.

The expression of IFN-g in the two specimens from

the adjacent normal tissue was intense, in contrast to the

weak and granular-like staining for one specimen from

Table 1

Relationship between IL-18, IFN-g and IL-10 expression with sex, race and age

Variables IL-18 IFN-g IL-10

Surrounding tissue Tumour Surrounding tissue Tumour Surrounding tissue Tumour

� � P � � P � � P � � P � � P � � P

Sex

Male 11 3 0.579 6 8 0.243 3 11 0.824 1 13 0.824 4 10 0.421 4 10 0.676

Female 2 1 0 3 0 3 0 3 0 3 1 2

Race

Malay 3 3 0.261 1 5 0.099 2 4 0.272 0 6 0.647 0 6 0.139 2 4 0.605

Chinese 10 1 5 6 1 10 4 7 3 8

Age

B60 7 2 0.665 3 6 0.627 2 7 0.547 0 9 0.471 1 8 0.247 3 6 0.563

]60 6 2 3 5 1 7 1 7 3 5 2 6

Total 13 4 6 11 3 14 1 16 4 13 5 12

�, Positive immunoreactivity; �, negative immunoreactivity.

C.S. Chia et al. / Immunology Letters 84 (2002) 163�/172 165

Fig. 1. Immunostaining with anti-IL-18. Intense immunostaining was detected in the hepatocytes of the apparently normal tissue surrounding the

tumour in three specimens (a). The representative slides show moderate staining in a tumour (b) and apparently normal tissue surrounding the

tumour (c) and weak immunostaining in the normal vs. tumour in (d) and (e), respectively. Negative immunoreactivity is shown in (f) in the normal

tissue (200�/ magnification).


normal and the other from the tumour tissue. In

addition, all the staining covered only 26�/50% of the

area for both the tumour and surrounding tissue (Fig.

2).Statistical analysis showed no significant difference in

the incidence of IFN-g expression in the tumour versus

the normal tissue. No correlation was found between the

IFN-g expression, sex, race and age (Table 1) as well as

clinicopathological factors (Table 2). There was also no

correlation in the expression of IFN-g and the other

cytokines (Table 3).

3.3. Expression of IL-10 in apparently normal and

tumour tissues

IL-10 immunoreactivity was detected in the hepato-cytes of apparently normal tissue adjacent to the tumour

in four out of the 17 (23.5%) compared with five out of

the 17 tumour (29.4%) specimens (Table 1). The

representative slides for intense and weak immunoreac-

tivity for the apparently normal tissues are as shown in

Fig. 4a and c, respectively. The corresponding immu-

nostaining for the tumour portion are is shown in Fig.

4b and d, respectively.Fig. 2 shows the case distribution of IL-10 expression

with the intensity and extent of the area. The immunos-

taining was either intense or weak and the extent of

moderate staining is at 51�/75% of the area (one in

normal vs. none in tumour) whereas for weak staining, it

could either be 26�/50% (two cases for both apparently

normal and tumour tissue) or 76�/100% (only one for

both normal and tumour portion) of the area. Overall, it

appeared that for both apparently normal and tumour

tissue, the number of positive IL-10 cases were low, the

staining was only moderate or weak and extent of the

area with staining is similar.

There was no correlation between tumour IL-10

expression and clinicopathological factors (Table 2) as

Fig. 2. Case distribution of IL-18, IFN-g and IL-10 expression in the surrounding and tumour tissue as shown by the extent of the area and intensity

of immunostaining.


well as with other cytokines (Table 3) but it should be

kept in mind that this could be due to the small number

of samples.

4. Discussion

IL-18 has been initially known as a factor that is

primarily involved in the inflammatory immune re-

sponse and is a potent IFN-g inducing factor [18].

However, more recently, it has been reported that IL-18

has the capacity to stimulate innate immunity and both

Th1 and Th2 mediated responses [19]. IL-18 exerts anti-

tumour action via a number of mechanisms such as

enhancement of NK cell activity [20], induction of

apoptosis via Fas/Fas ligand interaction and inhibition

of angiogenesis [21]. In the present study, the number of

cases with positive IL-18 immunoreactivity is signifi-

cantly higher (2-fold) in the apparently normal tissue

compared with the tumour portion of the same indivi-

dual. This appears to be consistent with a previous study

showing that the presence of IL-18 transcript is much

Table 2

Relationship between IL-18, IFN-g and IL-10 expression and clinicopathological data



� � P � � P � � P � � P � � P � � P

Tumour size (cm)

B5 6 1 0.441 4 3 0.145 2 5 0.360 1 6 0.421 3 4 0.162 3 4 0.314

�5 7 3 2 8 1 9 0 10 1 9 2 8

HbsAg

Positive 9 1 0.162 3 7 0.484 2 8 0.641 0 10 0.412 4 6 0.088 3 7 0.686

Negative 4 3 3 4 1 6 1 6 0 7 2 5

HCV

Positive 1 1 0.426 1 1 0.426 1 1 0.331 0 2 0.882 0 2 0.574 2 0 0.074

Negative 12 3 12 3 2 13 1 14 4 11 3 12

Cirrhosis

Present 11 4 0.750 6 9 0.750 3 11 0.800 1 14 0.938 3 12 0.250 4 11 0.750

Absent 1 0 0 1 0 1 0 1 1 0 0 1

Unknown 1 0 0 1 0 1 0 0 0 1 1 0

AFP(IU/l)

�350 10 0 0.015 4 6 0.516 3 7 0.17 0 10 0.412 3 7 0.441 4 6 0.278

B350 3 4 2 5 0 7 1 6 1 6 1 6

6

Total: 13 4 6 11 3 14 1 16 4 13 5 12


Table 3

Relationship between IL-18, IFN-g and IL-10 expression



� � P � � P � � P � � P � � P � � P

IL-18

Positive 3 10 0.421 1 5 0.353 4 9 0.300 2 4 0.605

Negative 0 4 0 11 0 4 3 8

IFN-gPositive 3 0 0.421 1 0 0.353 1 2 0.465 0 1 0.765

Negative 10 4 5 11 2 12 5 11

IL-10

Positive 4 0 0.300 2 3 0.605 1 2 0.465 0 5 0.765

Negative 9 4 4 8 2 12 1 11

Total 13 4 6 11 3 14 1 16 4 13 5 12



higher in the normal compared with the tumour colon

cells of the same individual [22]. The latter study also

indicated that the IL-18 expressed in the tumour will

most likely not produce bioactive protein because of the

lack of the interleukin-1b converting enzyme (ICE)

transcript. ICE activity is required to cleave the pre-

cursor form of IL-18 in order to produce a biologically

active IL-18. In contrast, IL-18 and ICE transcripts

have been detected in the corresponding normal colon

cells suggesting the bioactive IL-18 is most likely

produced by the adjacent normal cells [22]. So far, the

role of IL-18 produced in the surrounding normal tissue

in the anti-tumour immune response or cancer progres-

sion is still unknown. IL-18 alone or in combination

with IL-12 can be a potent adjuvant for reducing

tumour growth and metastasis in animal tumour models

[21]. Since our initial study has indicated the presence of

IL-18 in HCC, it is worthwhile performing further

studies to investigate whether the IL-18 detected is

biologically active and to elucidate the role of IL-18 in

liver and other cancers.

A previous study showed that IFN-g transcript was

expressed in HCC but not the protein [23]. Since IFN-gpromotes a Th1 response, which will be favourable for

the induction of the anti-tumour response, the absence

of IFN-g protein in the majority of the specimens in this

study suggests a favourable environment for evasion of

the immune surveillance system. The very low number

of cases with IFN-g immunoreactivity is consistent with

the hypothesis that tumour cells are allowed to survive

due the lack of a predominant Th1 response required for

the anti-tumour response. Further studies on factors

Fig. 3. Representative slides showing the immunohistochemical staining with anti-IFN-g. Moderate expression of IFN-g was detected in the

hepatocytes of the apparently normal tissue surrounding the tumour (a) and weak expression in the tumour (b). The negative control for the normal

and tumour tissue are as shown in (c) and (d) (200�/ magnification).


that can contribute to enhance tumour cell survival

rather than death via apoptosis need to be investigated.

The presence of IFN-g in the tumour has been

associated with the rejection of the tumour. Non-

immunogenic tumour cells of high metastasing potential

and with the IFN-g gene insertion when injected into

syngeneic mice have a lower ability than the parental

cells to develop into tumours [24]. However, other

studies have shown that the presence of IFN-g may

not be beneficial, as the administration of IFN-g did not

inhibit the growth of the tumour in animal experimental

models. Thus, the role of IFN-g in the anti-tumour

response requires further investigation.

It is generally thought that IL-10 has the capacity to

suppress cellular responses via a number of ways such as

the inhibition of cytotoxicity and cytokine production

by tumour-specific T cells [25], suppression of the

tumoricidal activity of macrophages [26] and inhibition

of dendritic cell functions against tumour cells [27,28].

Perhaps, this will allow for progressive tumour growth

Fig. 4. Immunostaining with anti-IL-10. Intense staining was detected in the apparently normal tissue surrounding the tumour (a) and moderate

staining in the tumour (b). A representative slide showing weak staining in the normal surrounding tissue and in the tumour is as shown in (c) and (d),

respectively (200�/ magnification).


by prevention of the cellular immune responses. This is

consistent with the elevation of IL-10 expression in a

variety of human cancers including melanoma, non-

small cell lung carcinoma, gastric and colon adenocar-

cinoma [29,30]. Thus, according to this hypothesis, a

higher percentage of the tumour is expected to have

positive immunoreactivity for IL-10 as the presence of

IL-10 in the tumour microenvironment will allow for

suppression of the productive cellular immune response

that contributes to the anti-tumour effect. Our results

have shown that, a higher percentage of cells in the

tumour portion (five out of 17; 29.4%) produce IL-10,

and only one out of 17 (5.9%) tumour was positive for

IFN-g but it should be kept in mind that, the number of

samples available for this study was small. In the

apparently normal tissue adjacent to the tumour, only

one specimen was found to be have positive immunor-

eactivity for both IL-10 and IFN-g and two specimens

expressed IFN-g only. However, the immunoreactivity

for IFN-g in these cases is very weak. Due to the small

number of samples, it is difficult to be definitive that the

hypothesis that a predominant Th2 response exists in

the tumour microenvironment is upheld but the lack of

IL-10 in a large proportion of the tumours suggests a

lack of a predominant Th2 response.

The correlation between IL-18, IL-10 and IFN-gexpression was investigated due to a number of reasons.

Amongst the many biological activities of IL-18 and IL-

10 such as in anti-angiogenesis and immunosuppression,

respectively, IL-18 has been shown to inhibit production

of IL-10 in vitro [31], the correlation between IL-18 and

IL-10 production will give further insight into the

possible roles of these two cytokines in the anti-tumour

response. In this study, we found no relationship

between the expression of IL-10, IL-18 and IFN-g in

both tumour and the surrounding tissues. As IL-18 can

stimulate Th1 or Th2 responses depending on its

cytokine milieu [19], it will be interesting to further

investigate the profile of other cytokines.

In addition, this study needs to be further extended to

find out the phenotype and activation status of other

cells such as the T-lymphocytes in the tumour micro-

environment. One of the limitations of immunohisto-

chemical staining is that the amount of protein cannot

be quantitated and the intensity of staining can some-

times be fairly subjective. With the advances in laser

capture microdissection techniques and real-time poly-

merase chain reaction, further studies can be done to

analyse the level of the transcript for various cytokines.

The latter approach combined with immunohistochem-

ical staining of the specific cell types will provide

additional information on the source of the cytokines

produced at the tumour microenvironment. Thus, these

studies will contribute to the development of rationale

new therapeutic approaches for the treatment of cancer.

References

[1] M. Feldmann, F.M. Brennan, in: J.J. Oppenheim, M. Feldmann,

S.K. Durum, T. Hirano, J. Vilcek, N.A. Nicola (Eds.), Cytokine

Reference, vol. 1, Academic Press, 2000, pp. 35�/52.

[2] S. Seki, Y. Habu, T. Kawamura, K. Takeda, H. Dobashi, T.

Ohkawa, H. Hiraide, Immunol. Rev. 174 (2000) 35�/46.

[3] G. Ramadori, T. Armbrust, Eur. J. Gastroenterol. Hepatol. 13

(2001) 777�/784.

[4] N. Fausto, J. Hepatol. 32 (2000) 19�/31.

[5] S.M. Akbar, M. Abe, H. Murakami, K. Tanimoto, T. Kumagi, Y.

Yamashita, K. Michitaka, N. Horiike, M. Onji, Cancer Lett. 171

(2001) 125�/132.

[6] A.F. Yoong, S.C. Afford, R. Jones, P. Aujla, S. Qin, K. Price,

S.G. Hubscher, D.H. Adams, Hepatology 30 (1999) 100�/111.

[7] T. Boon, P.G. Coulie, V. den Eynde, Immunol. Today 18 (1997)

267�/268.

[8] R.B. Herberman, M.E. Nunn, H.T. Holden, D.H. Lavrin, Int. J.

Cancer 16 (1975) 23�/239.

[9] T.R. Mosmann, R.L. Coffman, Adv. Immunol. 46 (1989) 111�/

147.

[10] D. Robinson, K. Shibuya, A. Mui, F. Zonin, E. Murphy, T. Sana,

S.B. Hartley, S. Menon, R. Kastelein, F. Bazan, A. O’Garra,

Immunity 7 (1997) 571�/581.

[11] M. Giovarelli, P. Musiani, A. Modesi, P. Dellabona, G. Cavallo,

F. di Pierro, C. DeGiovanni, J. Immunol. 155 (1995) 3112�/3123.

[12] B.K. Halak, H.C. Maguire, E.C. Lattime, Cancer Res. 59 (1999)

911�/917.

[13] N. Barker, H. Clevers, Trends Mol. Med. 7 (2001) 535�/537.

[14] I. Merican, R. Guan, D. Amarapuka, M.J. Alexander, A.

Chutaputti, R.N. Chien, S. Hasnian, N. Leung, L. Lesmana,

P.H. Piet, N.H.M. Sjalfoellah, J. Sollano, H.S. Sun, D.Z. Xu, J.

Gastroenterol. Hepatol. 15 (2000) 1356�/1361.

[15] G.W. Wong, A.M. Ali, H.F. Seow, Asia Pacific J. Mol. Biol. 7

(1999) 109�/118.

[16] R. Koomagi, M. Volm, Int. J. Cancer 84 (1999) 239�/243.

[17] M.A. Rahman, D.K. Dhar, E. Yamaguchi, S. Maruyama, T.

Sato, H. Hayashi, T. Ono, A. Yamanoi, N. Nagasue, Clin. Cancer

Res. 7 (2001) 1325�/1332.

[18] H. Okamura, H. Tsutsui, T. Komatsu, M. Yutsudo, A. Hakura,

T. Tanimoto, J. Torigoe, T. Okura, Y. Nukada, K. Hattori,

Nature 378 (1995) 88�/91.

[19] K. Nakanishi, T. Yoshimoto, H. Tsutsui, H. Okamura, Annu.

Rev. Immunol. 19 (2001) 423�/474.

[20] F. Tanaka, W. Hashimoto, H. Okamura, P.D. Robbins, M.T.

Lotze, H. Tahara, Cancer Res. 60 (2000) 4838�/4844.

[21] C.M. Coughlin, K.E. Salhany, M. Wysocka, E. Aruga, H.

Kurzawa, A.E. Chang, C.A. Hunter, J.C. Fox, G. Trinchieri,

W.J. Lee, J. Clin. Invest. 101 (1998) 1441�/1452.

[22] F. Pages, V. Vives, C.S. Fridman, F. Fossiez, A. Berger, P.H.

Cugnenc, E. Tartour, W.H. Fridman, Immunol. Lett. 68 (1999)

135�/139.

[23] M. Nagao, Y. Nakajima, H. Kanehiro, M. Hisanaga, Y.

Aomatsu, S. Ko, Y. Tatekawa, N. Ikeda, H. Kanokogi, Y.

Urizono, T. Kobayashi, T. Shibaji, T. Kanamura, S. Ogawa, H.

Nakano, Hepatology 32 (2000) 491�/500.

[24] A. Billiau, K. Vandenbroeck, in: J.J. Oppenheim, M. Feldmann,

S.K. Durum, T. Hirano, J. Vilcek, N.A. Nicola (Eds.), Cytokine

Reference, vol. 1, Academic Press, 2000, pp. 689�/702.

[25] J.W. Rohrer, J.H. Coffin, J. Immunol. 155 (1995) 5719�/5722.

[26] R. Nabioullin, S. Sone, K. Mizuno, S. Yano, Y. Nishiko, T.

Haku, T. Ogura, J. Leukocyte Biol. 55 (1994) 437�/442.

[27] J.X. Gao, J. Madrenas, W. Zeng, M.J. Cameron, Z. Zhang, J.J.

Wang, R. Zhong, D. Grant, Immunology 98 (1999) 159�/170.

[28] J. Steinbrink, H. Jonuleit, G. Muller, G. Schuler, J. Knop, A.H.

Enk, Blood 93 (1999) 1634�/1642.


[29] T. Sato, P. McCue, K. Masuoka, S. Salwen, E.C. Lattime, M.J.

Mastrangelo, D. Berd, Clin. Cancer Res. 2 (1996) 1383�/1390.

[30] C. Fortis, M. Foppoli, L. Gianotti, L. Galli, G. Citterio, G.

Consogno, O. Gentilini, M. Braga, Cancer Lett. 104 (1996) 1�/5.

[31] S.J. Ushio, M. Namba, T. Okura, K. Hattori, Y. Nukada, K.

Akita, F. Tanabe, K. Konishi, M. Micallef, M. Fujii, K. Torigoe,

T. Tanimoto, S. Fukuda, M. Ikeda, H. Okamura, M. Kurimoto,

J. Immunol. 156 (1996) 4274�/4279.


Documents

Expression of interleukin-18, interferon-γ and interleukin-10 in hepatocellular carcinoma