Vol. 1, 1353-1358, November 1995 Clinical Cancer Research 1353

Involvement of Human Interleukin 6 in Experimental Cachexia

Induced by a Human Uterine Cervical Carcinoma Xenograft

Sumie Tamura, Kaori Fujimoto Ouchi,

Kazushige Mori, Mika Endo, Takehisa Matsumoto,

Hiroyuki Eda, Yutaka Tanaka,’ Hideo Lshitsuka,

Hisashi Tokita, and Ken Yamaguchi

Department of Oncology, Nippon Roche Research Center, 200

Kajiwara, Kamakura, Kanagawa 247 [5. T., K. F. 0., K. M., M. E.,

T. M., H. E., Y. T., H. I.], Japan; Division of Animal Research,Chiba Cancer Center Research Institute, 666-2 Nitona-cho, Chuo-ku,

Chiba 260 [H. T.], Japan; and Growth Factor Division, National

Cancer Center Research Institute, Tsukiji, Chuo-ku, Tokyo 104[K. Y.], Japan

ABSTRACT

A human tumor xenograft model for cancer cachexia

was established by growing a uterine cervical carcinoma,

Yumoto, in nude mice. The tumor transplanted into the mice

induced severe body weight loss (30% of body weight) when

the tumor weight was only 1 g. In addition, other indicators

for cachexia, such as adipose tissue and muscle wasting and

hypoglycemia, were also observed in the tumor-bearing

mice, suggesting that this is a proper model for experimental

cachexia induced by a human tumor. We then examined the

association of this model with various cytokines, such as

tumor necrosis factor a, interleukin (IL)-la, IL-1�1, IFN-y,

IL-6, and leukemia inhibitory factor, and identified humanIL-6, which was produced by the tumor cells, as a mediatorof cachexia. A neutralizing antibody against hIL-6 adminis-

tered to the mice after the development of cachexia symp-

toms significantly improved body weight loss, adipose tissue

wasting, hypoglycemia, acute phase reaction, and leukocy-

tosis, although it did not suppress the tumor growth. These

results demonstrate that the hIL-6 produced by the tumor

cells is an essential mediator of the cachexia induction in this

model.

INTRODUCTION

Cancer cachexia, which includes disorders of homeostasis

such as progressive wasting, weakness, anorexia, and anemia, is

commonly seen in cancer patients (1, 2). Cachexia is associated

not only with deterioration of the quality of life but also with

shorter survival (1) and poor response to chemotherapy (3).

Elucidation ofthe mechanism of cachexia induction will help in the

search for potential therapeutic interventions for advanced cancer.

In studies on the mechanism of cachexia induction, several

cytokines, such as TNF2 (4, 5), IL-i (6), IFN-’y (7, 8), IL-6

(9-12), LIF (13), and transforming growth factor �3 (14), have

been proposed as cachexia inducers in different cachexia mod-

ebs. In cachexia models of munine tumors, however, it is rather

difficult to identify whether the tumor cells or the host cells are

producing these cytokines. In contrast, in nude mouse models

bearing human tumor xenografts, tumor cell-derived cytokines

should be easily distinguishable from host cell-derived cyto-

kines. Therefore, experimental models for cachexia induced by

human tumor xenografts should be useful for the identification

of tumor cell-derived cachexia factors. Indeed, one of the au-

thors has reported a close relationship between the expression of

hLIF mRNA and the development of cachexia syndrome in

human melanoma xenograft models (13). The participation of

hIL-6 in the hypercalcemia, and cachexia induced by a human

head and neck squamous cell carcinoma was also reported (1 1).

In this report, we described a cachexia model established

by growing a human uterine cervical carcinoma, the Yumoto

line, in nude mice. We also showed that hIL-6 produced by the

tumor cells is a mediator of cachexia. This action was demon-

strated by an experiment in which a neutralizing antibody

against hIL-6 was administered to the mouse model. The roles

of IL-6 in the induction of cachexia were also discussed.

MATERIALS AND METHODS

Mice. Female BALB/c-nu/nu mice (Charles River Japan

Inc., Atsugi, Japan) were used at the age of 6-7 weeks. They

were kept at 22 ± 2#{176}Cwith a 12 h on and off lighting cycle and

given breeding diet Fl (Funabashi Farm, Funabashi, Japan)

containing 21.3% protein, 57.1% carbohydrates, 5.6% fat, 3.3%

fiber, 5.7% ash, and 7.0% moisture (metabolic calories, 3.76

kcal g ‘) and water ad libitum. All studies were conducted in

accordance with the Guidelines for the Care and Use of Labo-

ratory Animals in Nippon Roche Research Center.

Tumors. Tumor tissue of the human uterine cervical

carcinoma Yumoto, a well-differentiated keratotic epidermoid

type, was transplanted s.c. into a nude mouse and established as

a transplantable tumor line, as described previously (15). In the

present study, this in vivo tumor line was adapted for in vitro

tissue culture using the following procedure. The tumor was

excised, minced into small fragments, and then plated onto a

25-cm2 Pnimania flask (Falcon 3813; Becton Dickinson, Lincoln

Park, NJ) with MEM supplemented with 10% FCS to suppress

the growth of contaminating fibroblasts. After three passages in

the Primaria flasks, the growing adherent tumor cells were

maintained in ordinary tissue culture flasks (Falcon 3028; Bec-

ton Dickinson) at 37#{176}C,5% C02, in RPMI 1640 medium

supplemented with 10% FCS. Compared with observations in

Received 6/2/95; accepted 7/12/95.

I To whom requests for reprints should be addressed.2 The abbreviations used are: TNF, tumor necrosis factor; IL, interleu-

kin; G-CSF, Granubocyte-cobony stimulating factor; LIF, leukemia in-

hibitory factor; hLIF, human LIF; hIL, human IL; lAP, immunosup-

pressive acidic protein; hG-CSF, human G-CSF; hIFN, human IFN;

mIL, murine IL; hTNF, human TNF.

Research. on March 15, 2020. © 1995 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

1354 hIL-6-induced Experimental Cancer Cachexia

the mice bearing the in vivo tumor line (15), cells from this

established tissue culture line transplanted (s.c.) into nude mice

showed similar histological characteristics and caused a similar

extent of weight loss in the tumor-bearing mice. This tissue

culture line is, however, still a mixture of various clones. The

ability of such a line to induce cachexia might be unstable, since

individual clones may have different abilities to induce cachexia

and the proportions of various clones could be changeable in a

long-term tissue culture. Therefore, we have established a sin-

gle-cell derived clone which retains the stable ability to induce

cachexia using a limiting dilution method. Clone 17 was isolated

for such purposes and utilized for further studies on the involve-

ment of cytokines. All of these cell lines were confirmed to be

free of Mycoplasma by testing with 6-methylpunine deoxynibose

(Boehninger Mannheim, Mannheim, Germany).

Measurement of Body Wasting and Other CachexiaParameters. A single-cell suspension of tumor cells (106 cells/

mouse) was inoculated s.c. into the right flank of the nude mice.

The body weight and the length (a) and width (b) of the tumors

were measured two or three times a week. For time course exper-

iments, the tumor weight was estimated by calculating the tumor

volume (ab2/2) and multiplying this by a correction factor. The

correction factor was determined by comparing actual tumor vol-

umes and tumor weights in separate experiments. When the mice

were sacrificed, the actual tumor weight was measured. Carcass

weight was then calculated as the difference in weight between the

whole body and the tumor tissue. To determine the extent of tissue

wasting, we measured the weights of the left peniovarian white

adipose tissues and the gastrocnemius muscles of the left hind legs.

The concentrations of substances in serum or plasma were mea-

sured in blood samples collected from the heart by using the

following methods and reagents: an enzyme reaction method with

mutarotase and glucose oxidase for measuring glucose (Glucose

CII test; Wako Pure Chemical md. Ltd., Osaka. Japan); an immu-

nodiffusion assay with antimouse lAP antibody for measuring

munine lAP (Ref. 16; mouse lAP plate; Sanko Junyaku, Tokyo,

Japan); a clotting time test with thrombin using COBAS FIBRO

(Nippon Roche, Tokyo, Japan) for measuring fibninogen in citrate

plasma; a cobor-chelate reaction method with o-cresol phthalein

complexone for measuring calcium (Calcium C-Test; Wako); and

a radioimmunoassay for thyroxin (T-3 RIA Beads; Dynabot, To-

kyo, Japan).

Cytokine Assays and Neutralizing Antibodies. The

levels of cytokines secreted by the tumor cells in vitro were

determined by assaying the supernatants from 18-h cultures of

tumor cells (i0� cells/ml). Circulating levels of cytokines in

tumor-bearing mice were determined for serum samples ob-

tamed from the heart at autopsy. For the determination of tumor

tissue levels of cytokines, tumor tissues were minced in PBS,

disrupted by three 10-s sonication bursts with an immersion

probe, and centrifuged at 10,000 X g for 10 mm. Concentrations

of cytokines in these samples were determined by the following

ELISA systems: Factor-Test system for munine TNF-a (Gen-

zyme, Cambridge, MA); Quantikine systems for human TNF-a,

hIL-la, hIL-1�3, hIL-6, hLIF, hG-CSF (R & D Systems, Mm-

neapolis, MN); Cytoscreen system for hIFN-� (BioSource In-

ternational, Inc., Camanillo, CA); Inter-Test systems for mIL-lot

and munine IFN-y (Genzyme); and munine IL-6 ELISA kit for

mIL-6 (Endogen, Inc., Boston, MA). None of the ELISA sys-

Table 1 Changes in parameters associated with cancer cachexia in

mice bearing a human cervical carcinoma, Yumoto line

Tumor-bearing Nontumor-bearing

Parameters mice” mice

Tumor wt (g) 0.72 ± 0.46

Carcass wt (g) 18.4 ± 2.2” 28.2 ± 2.2

Adipose tissue wt (mg) 8 ± 11” 160 ± 41

Muscle wt (mg) 74 ± 19” 145 ± 10

Serum glucose (mg/dl) 121 ± 24b 173 ± 24

Plasma fibninogen (mg/dl) 21.9 ± 7.6’ 9.98 ± 3.0

Serum lAP (jig/mI) 491 ± 119 274 ± 180

Serum calcium (mg/dI) 12.7 ± 4.0’ 6.28 ± 0.2

Serum thyroxin (p.g/dl) 0.17 ± 036b 2.68 ± 1.04

WBC count (X106 cells/ 192 ± l3l� 7.8 ± 1.4

ml)

Spleen wt (mg) 416 ± 293” 108 ± 30

12 The mice were inoculated s.c. with tissue culture cells of the

Yumoto line (106 cells/mouse). Tumor-bearing mice and age- and

sex-matched normal mice were sacrificed on day 66. Data are expressed

as the mean ± SD from six mice per group.

b Statistically different from nontumor-bearing mice using the

Mann-Whitney U test (P < 0.01).C Statistically different from nontumor-bearing mice using the

Mann-Whitney U test (P < 0.05).

tems for human cytokines cross-reacted with corresponding

mouse cytokines and vice versa. The biological activities of

human and mouse IL-6 were determined by an IL-6-dependent

B9 cell proliferation assay (17). Mouse monoclonal antihuman

IL-6 antibody (B-E8) (18) was obtained from BioSource Inter-

national. Twenty jig of this antibody are able to neutralize 1 unit

of natural and recombinant hIL-6, whereas it does not neutralize

mIL-6. Rabbit polyclonal antihuman IL-la antibody (LP-710)

was obtained from Genzyme. One �ig of this antibody is able to

neutralize 1 unit of natural and recombinant hIL-ict.

Statistical Analysis. Statistical analysis of data was per-

formed using SAS software. Differences in tumor size, carcass

weight, tissue weight, and concentrations of substances were

analyzed using the Mann-Whitney U test. Differences in the

time courses of carcass weight changes were analyzed by

ANOVA. Differences in IL-6 levels in the culture supernatant of

tumor cells were analyzed using Student’s t test. Differences

were considered to be significant when the P was < 0.05.

RESULTS

Experimental Cancer Cachexia Induced in Nude Mice

by Human Cervical Cancer Yumoto. A xenograft model

established by growing a human cervical carcinoma, Yumoto

line, in nude mice was capable of inducing marked body weight

loss in the early stage of tumor growth. The tumor-bearing mice

started losing weight by day 35, when the tumor weight was

only 1 g (4% of total body weight), and continued to lose weight

until day 45, when their calculated carcass weight was 10 g less

than that of age- and sex-matched controls. As Table 1 shows,

these mice also showed significant loss of adipose tissue and

muscle weight, as well as hypoglycemia, indicating the presence

of disorders in energy and nitrogen homeostases. Other changes

observed in these mice included elevated levels of the acute

phase proteins (fibninogen and lAP), reduced level of thyroxin,

splenomegaly, hypercalcemia, and leukocytosis. These abnor-

Research. on March 15, 2020. © 1995 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Clinical Cancer Research 1355

Table 2 In vitro production of cytokines by clone 17 of the

Yumoto line

Cytokines

Concentrations in

conditioned medium”

(ng/ml)

hTNF-a <0.01hIL-lu 0.09 ± 0.03”

hIL-113 <0.01hIL-6 3.57 ± 0.56

hLIF <0.01hIFN--y <0.02

hG-CSF 0.14 ± 0.07

‘4 Concentrations of various cytokines in conditioned medium of

the Yumoto line, clone 17 (10� cells/mb for 18 h). were determined using

ELISA systems for individual cytokines.h Data are expressed as the mean ± SD from three determinations.

malities of the host are also characteristics of paraneoplastic

syndrome (19), although they are not necessarily associated with

weight loss.

Cytokines Detected in the Yumoto Model. For further

experiments on cachexia mediators in the Yumoto model, we

used clone 17, a single-cell-derived clone of the tissue culture

Yumoto line. Probably because of the homogeneous population

of tumor cells, the mice bearing this clone showed more syn-

chronized weight loss than those bearing the original tissue

culture line.

Production of possible cachexia mediators, TNF-a, IL-la,

IL-1�3, IFN--y, IL-6, LIF, and G-CSF by Yumoto clone 17, was

examined in vitro and in vivo using the ELISA. In the condi-

tioned medium from tissue cultures of Yumoto clone 17, large

amounts of hIL-6 as well as small amounts of hIL-la and

hG-CSF were detected, but hTNF-a, hIL-1�3, hLIF, or hIFN--y

was not detected (Table 2). In the sera of tumor-bearing mice,

increased levels of hIL-6, mIL-6, and hG-CSF were detected,

but hTNF-a, hIL-la, �3, hLIF, hIFN-’y, munine TNF-a, mIL-ia,

or munine IFN--�, was not. In the homogenates of the tumor

tissue, however, high levels of hIL-1a,�3 and mIL-la were

detected using the ELISA, in addition to hIL-6, mIL-6, and

hG-CSF (Table 3). The human-type cytokines detected in vivo

must be produced by the tumor cells themselves, whereas mu-

nine-type cytokines in the tumor tissues must be produced by

infiltrating host cells or stroma cells. Similar patterns of cyto-

kine production in vitro and in vivo were observed in the

original tissue culture line (data not shown). When the correba-

tion between the serum levels of hIL-6 and the body weight was

analyzed for 28 mice bearing tumors of various sizes, a negative

correlation between these two parameters was confirmed, with a

correlation coefficient of r = 0.531; probability of the correla-

tion P < 0.005.

Improvement of Cachexia by Antihuman IL-6 mAb.

To investigate the role of IL-6 in vivo, we administered anti-

hIL-6-neutralizing antibody B-E8 to the cachectic mice. The

mice bearing the Yumoto clone 17 showed significant body

weight loss by day 40. Anti-hIL-6 antibody administered to

these mice i.p. at a dose of 50 jig/day for 7 days from day 44

caused a rapid and significant recovery of their body weight. In

contrast, the mice receiving the control antibody showed no

improvement of body weight loss (Fig. 1). The antibody-treated

mice and control mice were sacrificed on day 50, 3 h after the

final antibody injection, and their carcass weights, tumor

weights, and various cachexia parameters were compared (Table

4). Anti-hIL-6 antibody treatment caused a significant increase

in the carcass weight, although it did not suppress tumor growth.

Loss of adipose tissue weight and hypoglycemia were also

significantly improved. The serum level of lAP and WBC

counts were reduced. The biological activity of hIL-6 in the

serum and tumor tissue, which was determined by hybnidoma

B9 cell proliferation assay, was indeed reduced by the antibody.

These results strongly suggest that hIL-6 produced by tumor

cells is one of the essential inducers of cachexia in the Yumoto

model. It is worth noting that tumor tissue levels of mIL-6 and

serum levels of hG-CSF were also reduced by anti-hIL-6 anti-

body.

Stimulation of IL-6 Production by Cytokines in Vitro.

We detected other cytokines produced by the tumor cells, such

as hIL-la, in the tumor tissues. Since IL-i is known to induce

IL-6 production in various types of cells (20), we examined the

possibility that hIL-la is an autocnine stimulator for hIL-6

production in clone 17 of the Yumoto line. Fig. 2 shows in vitro

hIL-6 production by Yumoto clone 17 cells treated with anti-

hIL-la antibody. The neutralization of IL-i activity signifi-

cantly reduced hIL-6 production by the tumor cells. Thus, it is

highly possible that the hIL-1 detected in the tumor tissues could

be acting as an autocnine stimulator for hIL-6 production by the

tumor cells in the tumor tissues.

DISCUSSION

Appropriate animal models are indispensable to examine

the induction mechanism of cancer cachexia. Although several

animal models have been reported as models for cancer ca-

chexia (21-24), not many of them were established from human

tumor xenografts. It was reported in the previous article that a

human uterine cervical carcinoma, the Yumoto line, induced

remarkable weight loss in tumor-bearing mice (15). In the

present study, we established a tissue culture line of the Yumoto

line and demonstrated that the in vitro line induced the typical

cachexia syndrome after transplantation (s.c.) into nude mice

even when the tumor burden was small. Severe adipose tissue

and muscle wasting and hypoglycemia as well as decreased

carcass weight were observed in the mice bearing less than 1 g

of tumor. These findings suggest that there are tumor-induced

disorders in the host that involve energy and nitrogen homeosta-

sis. These mice also exhibited elevated levels of acute phase

proteins, hypercalcemia, and leukocytosis, abnormalities that

are often observed in advanced cancer patients (19). Thus, the

mice bearing the human cervical carcinoma Yumoto line can be

considered an appropriate model to investigate mechanisms of

human tumor-induced cachexia and disorders in homeostasis of

the host.

Host cell-derived and tumor cell-derived cytokines have

been suggested as mediators of the metabolic changes associ-

ated with cachexia (25, 26). In the munine carcinoma colon 26

model, mIL-6 has been proposed as a cachexia factor, because

a mAb to mIL-6 (but not anti-TNF antibody) was able to

suppress the development of key parameters of cachexia in

tumor-bearing mice (10, 12). In the colon 26 model, however, it

Research. on March 15, 2020. © 1995 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Table 3 In vivo cytokine productions in mice bearing clone 17 of the

Yumoto line

I

22

20

18

16

0 10 20 30 40 50

Days after Inoculation

1356 hIL-6-induced Experimental Cancer Cachexia

Cytokines

Levels of various cytokines”(ng/ml serum or ng/g tumor tissue)

Human-type cytokines Murine-type cytokines

Serum Tumor Serum Tumor

TNF-a <0.05 0.12 ± 0.03” <0.05 0.09 ± 0.02

IL-la <0.01 34.9 ± 5.6 <0.03 1.18 ± 0.45

IL-113 <0.01 21.4 ± 5.6 Not done Not done

IL-6 0.38 ± 0.19 15.2 ± 7.6 0.14 ± 0.05 3.13 ± 2.23

LIF <0.05 0.11 ± 0.04 Not done Not done

IFN--y <0.02 0.17 ± 0.18 <0.125 Not done

G-CSF 4.82 ± 2.36 188 ± 71 Not done Not done

a Cytokine levels in the sera of tumor-bearing mice and in the

tumor tissues were determined by corresponding ELISA systems. Sam-pies were obtained from mice bearing clone 17 of the Yumoto line on

day 50 for hIL-6, mIL-6, and hG-CSF, and on day 48 for the othercytokines. None of these cytokines was detected in the sera of non

tumor-bearing mice. The bower detection limits for hIL-6 and mIL-6 inthe serum were 0.01 and 0.05 ng/ml, respectively.

b Data are expressed as the mean ± SD for samples from six

animals.

0 non-tumor-bearing mice

U tumor-bearing mice

0 treated with anti-hIL-6 antibody

. treated with control IgG

Fig. 1 Improvement of cachexia by anti-hIL-6 antibody. Cells (l0�cells) of the Yumoto line clone 17 were inoculated into nude mice onday 0 (#{149}).Anti-hIL-6 antibody, B-E8 (0), or control IgG (#{149})wasadministered i.p. to the tumor-bearing mice (six mice per group) at adose of 50 jig/mouse/day for 7 days from day 44. Calculated carcass

weight and tumor volume were determined periodically. The control

nontumor-bearing nude mice (0) were matched for age and sex. The

difference in carcass weight changes between the anti-IL-6 antibodygroup and the control IgG group was statistically significant (ANOVA,P < 0.05). Points and bars, mean ± SD from six mice per group.

Table 4 Improvement of cachexia by anti-hIL-6 anbearing clone 17 of the Yumoto line

tibody in mice

Administration” Nontumor-

bearing

miceControl IgG Anti-hIL-6 Ab

Carcass wt (g) 17.8 ± 2.1 20.7 ± 1.3” 21.4 ± 1.3

Tumor wt (g) 1.83 ± 0.82 2.22 ± 1.09

WBC count 181 ± 113 30.3 ± 13.8’ 9.91 ± 1.90

(X l06/ml)

Adipose tissue 15 ± 8 68 ± 19’ 77 ± 76

wt (mg)Muscle wt (mg) 89 ± 12 101 ± 19

Serum glucose 126.0 ± 78.8 216.4 ± 54.4”

1 13 ± 7.4

246.4 ± 85.6

(mg/dl)

Serum calcium 11.9 ± 1.7 10.5 ± 1.7 8.6 ± 0.7

(mg/dl)Serum lAP 1142 ± 294 388 ± 105’ 183 ± 120

(jig/ml)

Serum hIL-6 0.38 ± 0.19 0.026 ± 0.017’

(ng/ml)”

Tumor hIL-6 15.2 ± 7.63 0.89 ± 0.57’

(ng/g)”Tumor mIL-6 3.13 ± 2.23 0.57 ± 0.41”

(ng/g)e

Serum hG-CSF 4.82 ± 2.36 1.01 ± 0.57’

(ng/ml)e

a Anti-hIL-6 mAb (B-E8) or control IgG was administered i.p. tothe mice bearing clone 17 of the Yumoto line at a dose of 50 jig/mouse/

day for 7 days from day 44. Various parameters were assayed on day 50.Data are expressed as the mean ± SD from six mice per group.

b Statistically different from the control IgG group using the Mann-

Whitney U test (P < 0.05).

C Statistically different from the control IgG group using the Mann-

Whitney U test (P < 0.01).d Determined by bioassay (B9 cell proliferation) in the presence of

anti-mIL-6 antibody.1� Determined using the ELISA.

was difficult to identify whether the cells producing mIL-6 were

tumor cells or host cells. In the present study with the Yumoto

model, we could detect elevated serum levels of both hIL-6 and

mIL-6, indicating that both the tumor cells and the host cells

were producing IL-6. Furthermore, administration of the anti-

hIL-6 antibody to cachectic mice significantly improved various

symptoms of cachexia, including weight loss. Therefore, the

hIL-6 produced by the tumor cells is considered to be an inducer

of cachexia in this model. The circulating level of mIL-6 de-

tected in the mice bearing the Yumoto line was reduced by the

treatment with anti-hIL-6 antibody, suggesting that hIL-6 causes

the host cells to produce mIL-6 in this model. Thus, even if

mIL-6 plays a role in the cachexia induction, hIL-6 should be a

causative factor for cachexia in this model.

The association of munine or human IL-6 with cachexia

induction was recently reported in several animal models. Chi-

nese hamster ovarian cells transfected with the mlL-6 gene

induced body weight loss (27). Munine Lewis lung carcinoma

cells transfected with the hIL-6 gene induced cachexia (28).

Anti-mIL-6-neutralizing antibody partially prevented the induc-

tion of cachexia in a colon 26 model (10, 12). Anti-hIL-6-

neutralizing antibody improved some cachexia parameters in a

human head and neck squamous cell carcinoma model, although

the reversal of body weight loss was not statistically significant

(1 1). Anti-mIL-6 antibody partially prevented body weight loss

Research. on March 15, 2020. © 1995 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

hIL-6 production (ng/ml)

shown). On the other hand, the anti-hIL-6 antibody treatment

did not significantly reduce serum levels of calcium, indicating

that hIL-6 is not the sole inducer of hypercalcemia in this model.

One possibility is suggested by the finding that the human

parathyroid hormone-relating protein was detected in tissueI cultures of the Yumoto line.3 Involvement of hG-CSF or human

0 2 4 6 8 10 parathyroid hormone-relating protein in the body weight loss or

tissue wasting, however, is still unclear. These factors might

play auxiliary roles in cachexia induction, although this has to

be confirmed.

Thus, our data indicate that the hIL-6 produced by the

tumor cells is an essential mediator for the cachexia induction in

this human cervical carcinoma xenograft model. Whether or not

IL-6 is a mediator common to clinical cancer cachexia or just to

certain types of cachexia has yet to be established.

Clinical Cancer Research 1357

3 K. Nagasaki (Growth Factor Division, National Cancer Center Re-

search Institute, Tokyo), personal communication.

Control

antilLlaAb � *

Fig. 2 Regulation of hIL-6 production by an antibody against hIL-la

in vitro. Cells (l0� cells/mb) of the Yumoto line clone 17 were incubated

in the presence or absence of anti-hIL-ics antibody (100 p.g/ml) for 72

h. The concentration of hIL-6 produced in the conditioned medium oftriplicate assays was determined using the ELISA. �, P < 0.05, com-

pared with the control cultures (Student’s t test).

induced by acute inflammation (29). It is also reported that IL-6

has an ability to suppress the expression in adipocytes of li-

poprotein lipase, one of the key enzymes regulating lipid me-

tabolism in cachexia (30). It is noteworthy that IL-6 and LIF, the

latter of which was proposed as a cachexia factor in human

melanoma xenograft models by one of the authors (13), share a

common subunit for their receptor, glycoprotein 130 (31). Thus,

IL-6 and LIF might play similar biological roles in the induction

of cachexia.

In contrast, body weight loss was not reported in mice

given injections of recombinant hIL-6 (32, 33), or in hIL-6

transgenic mice (34). Thus, the role of IL-6 in cachexia induc-

tion is still controversial. There are several possible explanations

for these apparent discrepancies. First, although IL-6 is an

essential factor for the induction of cachexia in certain models,

it might not be the sole inducer of cachexia. Systemic symptoms

of cachexia such as body weight loss may only occur in the

presence of some additional factors. Indeed, we detected hIL-

la, hIL-1�3, and mIL-la in the tumor tissues of the Yumoto

model. These cytokines might also be involved in the induction

of cachexia, although we have not yet examined their roles in

combination with IL-6. The other possibility is that IL-6 is not

a direct inducer of cachexia, but is able to make tumor cells

produce some other unidentified cachexia factor(s). If the pro-

duction of such a factor is dependent on the cell type, then the

IL-6 production would not always be associated with the ca-

chexia induction in some cell types. Thus, further studies are

still required to clarify the role of IL-6 in cachexia induction.

However, the result of the anti-hIL-6 antibody treatment, we

reported here, gives clear-cut evidence for the participation of

tumor cell-derived hIL-6 in experimental cachexia models.

Besides typical disorders involving a host metabolism in

cancer cachexia, we also observed other disorders in this model,

such as leukocytosis and hypercalcemia. The leukocytosis was

improved by the anti-hIL-6 antibody treatment. We have de-

tected a high level of hG-CSF by ELISA in the serum of

tumor-bearing mice, which is decreased by the anti-hIL-6 anti-

body treatment. Therefore, the suppression of leukocytosis

could be mediated by the reduction of the serum levels of

hG-CSF. The mechanism for in vivo suppression of hG-CSF

production by anti-hIL-6 antibody is still unclear, since in vitro

production of hG-CSF by the tumor cells was not affected by a

treatment with either hIL-6 or anti-hIL-6 antibody (data not

REFERENCES1. DeWys, W. D., Begg, C., Lavin, P. T., Band, P. R., Bennett, J. M.,

Bertino, J. R., Cohen, H. H., Douglass, H. 0., Jr., Engstrom, P. F.,

Ezdinli, E. Z., Horton, J., Johnson, G. J., Moerteb, C. G., Oken, M. M.,

Perbia, C., Rosenbaum, C., Silverstein, M. N., Skeel, R. T., Sponzo,

R. W., and Tormey, D. C. Prognostic effect of weight loss prior tochemotherapy in cancer patients. Am. J. Med., 69: 491-497, 1980.

2. Norton, J. A., Peacock, J. L., and Morrison, S. D. Cancer cachexia.

CRC Crit. Rev. Oncol. Hematol., 7: 289-327, 1987.

3. DeWys, W. D., Begg, C., Band, P., and Tormey, D. The impact of

malnutrition on treatment results in breast cancer. Cancer Treat. Rep.,

65: 87-91, 1981.

4. Oliff, A., Defeo-Jones, D., Boyer, M., Martinez, D., Kiefer, D.

Vuocobo, G., Wolfe, A., and Socher, S. H. Tumor secreting human

TNF/cachectin induce cachexia in mice. Cell, 50: 555-563, 1987.

5. Stovroff, M. C., Fraker, D. L., Swedenborg, J. A., and Norton, J. A.

Cachectin/tumor necrosis factor-a possible mediator of cancer cachexia

in the rat. Cancer Res., 48: 4567-4572, 1988.

6. Moldawer, L. L., Georgieff, M., and Lundholm, K. Interbeukin 1,

tumor necrosis factor-a (cachectin) and the pathogenesis of cancer

cachexia. Clin. Physiob., 7: 263-274, 1987.

7. Matthys, P., Dijkmans, R., Proost, P., Van Damme, J., Heremans, H.,

Sobis, H., and Billiau, A. Severe cachexia in mice inoculated with

interferon--y-producing tumor cells. mt. J. Cancer, 49: 77-82, 1991.

8. Matthys, P., Heremans, H., Opednakker, G., and Billiau, A. Anti-

interferon-y antibody treatment, growth of Lewis lung tumors in mice

and tumour-associated cachexia. Eur. J. Cancer, 27: 182-187, 1991.

9. Gelin, J., Mobdawer, L. L., Lonnroth, C., Sherry, B., Chizzonite, R.,

and Lundholm, K. Robe of endogenous tumor necrosis factor a and

interleukin 1 for experimental tumor growth and the development of

cancer cachexia. Cancer Res., 51: 415-421, 1991.

10. Strassmann, G., Fong, M., Kenney, J. S., and Jacob, C. 0. Evidence

for the involvement of interleukin 6 in experimental cancer cachexia. J.

Clin. Invest., 89: 1681-1684, 1992.

I 1 . Yoneda, T., Nakai, M., Moriyama, K., Scott, L., Ida, N., Kunitomo,

T., and Mundy, G. R. Neutralizing antibodies to human interleukin 6reverse hypercalcemia associated with a human squamous carcinoma.

Cancer Res., 53: 737-740, 1993.

12. Ouchi, K. F., Tamura, S., Moni, K., Tanaka, Y., and Ishitsuka, H.

Establishment and characterization of cachexia inducing and noninduc-

ing clones of murine colon 26 carcinoma. mt. J. Cancer, 61: 522-528, 1995.

13. Moni, M., Yamaguchi, K., Honda, S., Nagasaki, K., Ueda, M., Abe,

0., and Abe, K. Cancer cachexia syndrome developed in nude mice

Research. on March 15, 2020. © 1995 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

1358 hIL-6-induced Experimental Cancer Cachexia

bearing melanoma cells producing leukemia-inhibitory factor. Cancer

Res., 51: 6656-6659, 1991.

14. Zugmaier, G., Paik, S., Wilding, G., Knabbe, C., Bano, M., Lupu,R., Deschauer, B., Simpson, S., Dickson, R. B., and Lippman, M.Transforming growth factor 131 induces cachexia and systemic fibrosis

without an antitumor effect in nude mice. Cancer Res., 51: 3590-3594,

1991.

15. Tokita, H., Tanaka, N., Sekimoto, K., Ueno, T., Okamoto, K., and

Fujimura, S. Experimental model for combination chemotherapy withmetonidazobe using human uterine cervical carcinoma transplanted into

nude mice. Cancer Res., 40: 4287-4294, 1980.

16. Shibata, Y., Tamura, K., and Ishida, N. In vivo analysis of the

suppressive effects of immunosuppressive acidic protein, a type of

al-acid glycoprotein, in connection with its high level in tumor-bearing

mice. Cancer Res., 43: 2889-2896, 1983.

17. Aarden, L. A., Dc Groot, E. R., Schaarp, 0. L., and Lansdorp, P. M.

Production of hybnidoma growth factor by human monocytes. Eur. J.Immunol., 17: 1411-1416, 1987.

18. Wijdenes, J., Clement, C., Klein, B., Moreb-Fournier, B., Vita, N.,

Ferrara, P., and Peters, A. Human recombinant dimeric IL-6 binds to itsreceptor as detected by anti-IL-6 monoclonal antibodies. Mob. Immu-

nob., 28: 1183-1192, 1991.

19. Bunn, Jr. P. A. and Ridgway, E. C. Paraneoplastic syndromes. ln:

V. T. DeVita, Jr., S. Hellman, and S. A. Rosenberg (eds.), Cancer,Principles & Practice of Oncology. Vol. 2, pp. 1896-1940. Philadeb-

phia: J. B. Lippincott Co., 1989.

20. Heinrich, P. C., Castell, J. V., and Andus, T. Interleukin-6 and the

acute phase response. Biochem. J., 265: 621-636, 1990.

21. Bibby, M. C., Double, J. A., Au, S. A., Fearon, K. C. H., Brennan,R. A., and Tisdale, M. J. Characterization of a transplantable adenocar-cinoma of the mouse colon producing cachexia in recipient animals. J.NatI. Cancer Inst., 78: 539-546, 1987.

22. Lundholm, K., Edstrom, S., Kanbberg, I., Ekman, L., and Schrsten,T. Relationship of food intake, body composition, and tumor growth tohost metabolism in nongrowing mice with sarcoma. Cancer Res., 40:

2516-2522, 1980.

23. Guaitani, A., Recchia, M., Carli, M., Racchetti, M., Bartoseki, I.,

and Garratini, S. Walker carcinoma 256: a model for studies on tumor-

induced anorexia and cachexia. Oncology, 39: 173-178, 1982.

24. Moley, J. F., Morrison, S. D., and Norton, J. A. Insulin reversal of

cancer cachexia in rats. Cancer Res., 45: 4925-4931, 1985.

25. Lowry, S. F., and Moldawer, L. L. Tumor necrosis factor and other

cytokines in the pathogenesis of cancer cachexia. In: V. T. DeVita, Jr.,S. Hellman, and S. A. Rosenberg (eds.), Cancer, Principles & Practiceof Oncology. Vol. 4, pp. 1-12. Philadelphia: J. B. Lippincott Co., 1990.

26. Tisdale, M. J. Cancer cachexia. Anticancer Drugs, 4: 1 15-1 25,

1993.

27. Black, K., Garrett, I. R., and Mundy, G. R. Chinese hamster ovariancells transfected with the murine interbeukin-6 gene cause hypercalce-

mia as well as cachexia, leukocytosis and thrombocytosis in tumor-

bearing nude mice. Endocrinology, 128: 2657-2659, 1991.

28. Ohe, Y., Podack, E. R., Olsen, K. J., Miyahara, Y., Miura, K., Saito,H., Koishihara, Y., Ohsugi, Y., Ohira, T., Nishio, K., and Saijo, N.Interbeukin-6 eDNA transfected Lewis lung carcinoma cells show un-

altered net tumour growth rate but cause weight boss and shorten

survival in syngenic mice. Br. J. Cancer, 67: 939-944, 1993.

29. Oldenburg, H. S. A., Rogy, M. A., Lazarus, D. D., Zee, K. J. V.,

Keeber, B. P., Chizzonite, R. A., Lowry, S. F., and Moldawer, L. L.

Cachexia and the acute-phase protein response in inflammation are

regulated by interleukin-6. Eur. J. Immunob., 23: 1899-1894, 1993.

30. Greenberg, A. S., Nordan, R. P., McIntosh, J., Calvo, J. C., Scow,

R. 0., and Jabbons, D. Interleukin 6 reduces lipoprotein lipase activity in

adipose tissue of mice in vivo and in 3T3-L1 adipocytes: a possible robe

for interleukin 6 in cancer cachexia. Cancer Res., 52: 41 13-4116, 1992.

31. Gearing, D. P., Comeau, M. R., Friend, D. J., Gimpeb, S. D., Thut,

C. J., McGourty, J., Brasher, K. K., King, J. A., Gibbis, S., Mosley, B.,Ziegler, S. F., and Cosman, D. The IL-6 signal transducer, gpl3O: an

Oncostatin M receptor and affinity converter for the LIF receptor.

Science (Washington DC), 255: 1434-1437, 1992.

32. Neta, R., Vogel, S. N., Sipe, J. D., Wong, G. G., and Nordan, R. P.

Comparison of in vito effects of human recombinant IL-l and human

recombinant IL-6 in mice. Lymphokine Res., 7: 403-412, 1988.

33. Asano, S., Okano, A., Ozawa, K., Nakahata, T., Ishibashi, T.,

Koike, K., Kimura, H., Tanioka, Y., Shibuya, A., Hirano, T., Kishimoto,

T., Takaku, F., and Akiyama, Y. In vivo effects of recombinant human

interleukin-6 in primates: stimulated production of platelets. Blood, 75:

1602-1605, 1990.

34. Suematsu, S., Matsuda, T., Aozasa, K., Akira, S., Nakano, N.,

Ohno, S., Miyazaki, J., Yamamura, K., Hirano, T., and Kishimoto, T.

IgG plasmacytosis in interleukin 6 transgenic mice. Natl. Acad. Sci.USA, 86: 7547-7551, 1989.

Research. on March 15, 2020. © 1995 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

1995;1:1353-1358. Clin Cancer Res   S Tamura, K F Ouchi, K Mori, et al.   induced by a human uterine cervical carcinoma xenograft.Involvement of human interleukin 6 in experimental cachexia

  Updated version

  http://clincancerres.aacrjournals.org/content/1/11/1353

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/1/11/1353To request permission to re-use all or part of this article, use this link

Research. on March 15, 2020. © 1995 American Association for Cancerclincancerres.aacrjournals.org Downloaded from