ROLE OF CYTOKINES IN CANCER CACHEXIA IN A MURINE MODEL OF INTRACEREBRAL INJECTION OF HUMAN TUMOURS

doi:10.1006/cyto.2001.0899, available online at http://www.idealibrary.com on

ROLE OF CYTOKINES IN CANCER CACHEXIA IN A MURINE MODEL OF INTRACEREBRAL INJECTION OF HUMAN TUMOURS Donatella R. M. Negri,1* Delia Mezzanzanica,1 Silvano Sacco,2 Massimo Gadina,1* Fabio Benigni,2 Laura Cajola,3 Gaetano Finocchiaro,3 Pietro Ghezzi,2 Silvana Canevari1 To study the role of cytokines that are relevant in cancer cachexia syndrome due to intracerebral tumours, mice were injected with human A431 epidermoid carcinoma, OVCAR3 ovarian carcinoma and GBLF glioma cells comparing intracerebral (i.c.) and systemic (i.p. or s.c.) routes of implantation. Anorexia and weight loss developed within 7–10 days in mice injected i.c. with A431 or OVCAR3 cells well before a large tumour developed, while i.c.-injected GBLF cells did not induce cachexia until day 20, when the tumour was large. By contrast, mice injected i.p. or s.c. developed tumours without evidence of anorexia. Thus, intracerebrally-growing A431 and OVCAR3 resulted in cancer cachexia independent of tumour mass, and we investigated their cytokine pattern. Serum levels of murine and human cytokines are not predictive of cancer cachexia development. Reverse-transcriptase polymerase chain reaction (RT-PCR) analysis revealed in the brain of i.c.-injected A431 tumour-bearing mice expression of human interleukin(IL-)1
, IL-1 and LIF in all samples and IL-6 in two of four samples while in i.c.-injected OVCAR3 tumour-bearing animals IL-6, and LIF were detected in all samples and tumour necrosis factor-
(TNF
) in two of four samples. Only LIF was expressed in brains of mice injected with GBLF cells. Murine IL-6 was increased only in the brains of A431-bearing mice. Only mice injected i.c. simultaneously with a monoclonal antibody (mAb) directed against the murine IL-6 receptor and OVCAR3 cells, but not those with mAb and A431 cells, showed a significant increase in survival time with a partial and temporary attenuation of cachexia symptoms. These results suggest that IL-6 in OVCAR3 model may be important cachectogenic factor when centrally released by even a limited number of tumour cells.  2001 Academic Press

The cachexia syndrome is characterized by several homeostatic perturbations including depletion of muscle and fat tissue, hypoglycaemia and anaemia, accompanied by progressive wasting, weakness and From the 1Unit of Molecular Therapies, Department of Experimental Oncology, Istituto Nazionale per lo Studio e la Cura dei Tumori, Via Venezian 1, 20133 Milano, Italy; 2Laboratory of Neuroimmunology, Istituto di Ricerche Farmacologiche ‘‘Mario Negri’’, Via Eritrea 62, Milano, Italy; 3Unit of Neuro-Oncology and Gene Therapy, Istituto Neurologico ‘‘Besta’’, Via Celoria 11, Milano, Italy Correspondence to: Dr. Silvana Canevari, Unit of Molecular Therapies, Department of Experimental Oncology, Istituto Nazionale Tumori, Via Venezian 1, 20133 Milan, Italy. E-mail, [email protected] *Present addresses: D.R.M. Negri, Istituto Superiore di Sanita’, Roma, Italy. M. Gadina, NIAMS, NIH, Bethesda, MD, USA Received 22 September 2000; received in revised form 9 April 2001; accepted for publication 18 May 2001  2001 Academic Press 1043–4666/01/130027+12 $35.00/0 KEY WORDS: Cancer anorexia-cachexia syndrome/intracerebral tumour growth/cytokines/animal model/Interleukin 6 CYTOKINE, Vol. 15, No. 1 (7 July), 2001: pp 27–38

anorexia. The cancer cachexia syndrome, which affects a large proportion of patients with solid tumours, is associated not only with decreasing quality of life but also with a shorter survival time and poor response to therapy. Cachexia is more common in children and elderly patients and becomes more pronounced as disease progresses, however the severity of cancer cachexia syndrome may be unrelated to tumour size since profound wasting has been observed in patients with tumour burdens of less than 1% body mass.1,2 Although the pathophysiology of cancer cachexia syndrome is not well defined, several hypotheses have been explored including cytokines, circulating hormones, neuropeptides, neurotransmitters, and tumourderived factors.3,4 Studies in animal models have shown that several cytokines, including tumour necrosis factor-
(TNF-
), interleukin-(IL-)1, IL-6, interferon-(IFN-), LIF and TGF, can produce signs, symptoms and 27

28 / Negri et al.

biochemical features commonly seen in cancer cachexia syndrome.5 Clinical studies have demonstrated elevated TNF-
levels in serum samples from patients with diseases known to be associated with cachexia;6 however serum levels of TNF-
and of IL-1, IL-6, and IFN- correlate with weight loss and anorexia only in some pathologies, such as pancreatic and prostate cancers.7,8 Cytokines are likely candidates as mediators of disease-associated cachexia and anorexia. Since the discovery that TNF and the cachectogenic mediator originally termed ‘‘cachectin’’ were identical9, various studies have documented the role of TNF and IL-1 in cachexia10–12. Transplantation of LIF-producing cells induces a cancer cachexia syndrome13,14 and similar results were reported for IL-615 and IFN-16–18. IL-1 and TNF are also potent anorectic agents19,20 and intracerebral (i.c.) injection of these cytokines in mice has been shown to induce acute and profound anorexia with typical metabolic alterations.21,22 Anorexia is also induced by i.c. injection of tumour cells genetically engineered to produce TNF.23 IL-6 and other cytokines of the IL-6 family, including LIF, have also been reported to induce anorexia when injected i.c.24 Although animal models have been used to study the role of cytokines in cancer cachexia syndrome following i.c. injection of cytokines or of tumour cells genetically engineered to produce cytokines, only a few animal-transplantable tumours induce cachexia as an early effect, when tumour burden is low.25–30 Use of these models is further limited by the difficulty in determining whether cytokines are produced by the tumour or the host cells. By contrast, athymic mice bearing human tumour xenografts readily enable identification of the cytokine origin, and few cancer cachexia syndrome xenograft models have been established and used to evaluate the role of cytokines.14,31 Here we established cancer cachexia syndrome models by injecting human tumour cell lines (A431, OVCAR3 and GBLF) i.c. into a lateral cerebral ventricle of athymic mice. We measured serum levels and brain expression of host- and tumour-derived cytokines, and evaluated the effect of a monoclonal antibody against the IL-6 receptor. The results suggest that IL-6 and IL-1, respectively in OVCAR3 and A431 models, may be important cachectogenic factors when centrally released by even a limited number of tumour cells.

RESULTS Induction of cancer cachexia syndrome following tumour cell injection in the Central Nervous System (CNS) To establish a mouse model of human cancer cachexia syndrome, nude mice were injected with each

CYTOKINE, Vol. 15, No. 1 (7 July, 2001: 27–38)

of the tumour cell lines peripherally (s.c. or i.p.) or in the CNS (i.c.). The s.c. injection of A431 (2.5106) and GBLF (10106) cells resulted in formation of evident tumour masses and mice were sacrificed after 25 days from tumour injection due to the large size and necrosis of the tumour mass. OVCAR3 cells, following i.p. injection (10106), were able to grow as solid masses and ascitic tumours; mice injected with these cells survived between 12 and 20 days (median: 15 days). In either the s.c. or i.p. peripheral models of tumour growth, no sign or symptom of cancer cachexia syndrome was evident, even when the animal became moribund (data not shown). By contrast, mice transplanted i.c. developed a cachexia-like syndrome, with a marked body weight loss that started, in the case of A431 or OVCAR3 cells, in the early stage of tumour growth (7–9 days after injection) and only later (18–20 days after injection) in the case of GBLF cells (Fig. 1). The early cachectogenic cell lines A431 and OVCAR3, even at the moment of death, induced only a slight increase in brain weight compared to that of control mice, by contrast GBLF tumour-bearing animals showed a significant increase in brain weight (P<0.05) compatible with a larger tumour mass Fig. 2). At necropsy, in all the three model systems, no brain oedema was evident. Eight brains (2 OVCAR3, 3 A431 and 3 GBLF) were studied histologically and neoplastic cells were clearly identified in 7/8. In six cases (1/2 OVCAR3, 3/3 A431 and 2/3 GBLF) the tumour invasion reached the hypothalamus. Figure 3 shows a coronal section of the brain from an A431 bearing mouse. Similar tumour infiltration was observed in the other tumour models (not shown). Food intake of CNS tumour-bearing animals also indicated that i.c transplantation of A431 or OVCAR3 cells is accompanied by an early anorexia particularly evident in A431-treated animals in which food intake continually declined (Fig. 4). Paralleling the body weight data, mice injected i.c. with GBLF cells evidenced reduced food intake only later (18–20 days after injection) (Fig. 4). A431 and OVCAR3 i.c. tumour-bearing mice had a dramatically short MST of 10 (range 7–14) and 11 (range 6–22) days, respectively, while GBLF-injected mice survived significantly longer (MST=26 days, range 11–34) (see Fig. 2). The short MST in these mice was not attributable to the development of large tumour masses in the brain.

Cytokine pattern in vitro and in vivo When maintained in vitro all the cell lines revealed by RT-PCR comparable levels of IL-6 and LIF; IL1
and IL1 were detected at different levels in A431 and OVCAR3 cells; TNF-
was detected in GBLF cells,

Cytokines and cancer cachexia / 29

120

120

110

110

100

100 90

90

80

% of initial body weight

80

70

70

60

Saline

60 0

5

10

15

20

25

30

50

35

120

120

110

110

100

100

90

A431 0

5

10

15

20

25

30

35

90

80

80

70 70 60 50

OVCAR3 0

5

10

15

20

25

30

GBLF

60 35

0

5

10

15

20

25

30

35

Days Figure 1.

Weight loss following i.c. injection of tumour cell lines or saline (control).

Mice were injected i.c. with saline (20 l/mouse, n=4), A431 (2.5105 cells/mouse, n=5), OVCAR3 (2.5105 cells/mouse, n=4), or GBLF (1106 cells/mouse, n=4), and individual body weight was recorded daily. Data are expressed as percent of the initial mouse weight, measured before the treatment. Arbitrary symbols are reported to identify each mouse in the same treatment group. A reduction of 20–25% in body weight as compared to saline-injected controls was significant (P<0.05) by analysis of variance. Data are from 1 of 4 separate experiments with similar results.

and it was almost undetectable in OVCAR3 cells (Fig. 5). All 3 cell lines released detectable amounts of TGF (data not shown). Analysis of serum samples from tumour-bearing mice for murine and human cytokines revealed no detectable hu-LIF, hu-IL1-
or hu-TNF-
, and no mu-IFN- and mu-TNF-
, by enzyme linked immunoabsorbent assay (ELISA) or bioassays (data not shown), whereas various levels of human and/or murine IL-6 were detected in serum samples from most of the mice, regardless of injection route (peripherally and i.c), with the exception of GBLF tumour-bearing mice, which showed only hu-IL-6 after i.c. injection (Table 1). Most of the i.c. growing tumours released detectable levels of hu-IL6 into the serum, reaching 30 pg/ml in the case of the GBLF-derived large tumour masses. Levels of mu-IL6 were highest in sera of mice bearing A431 i.c. tumours (median 400 pg/ml). mu-IL6 levels were also high in sera of s.c. A431- or i.p. OVCAR3 tumour-bearing mice (Table 1), which did not evidence

any cancer cachexia syndrome signs or symptoms. Altogether these data indicate that cytokine serum levels are not good predictors of cancer cachexia syndrome development. We then analyzed the brain tissue of i.c. tumourinjected mice (4 A431- and OVCAR3- and 3 GBLFbearing brains) by reverse transcriptase-polymerase chain reaction (RT-PCR) for the presence of human cytokine mRNA. A PCR analysis of one representative brain for each group of tumour-bearing mice and controls is depicted in Figure 5. No clear correlation between in vitro and in vivo cytokine release was observed. Specific bands can be revealed only in brains from A431- and OVCAR3-, but not from GBLFbearing mice. In particular, IL-1
, IL-1 and LIF mRNA were detected in all brain samples from A431bearing mice, and IL-6 was detected in 2 of 4 samples. In OVCAR3 tumour-bearing animals, IL-6 and LIF mRNA were expressed in all brains tested and TNF-
in two of four brains. In two of three brains from GBLF tumour-bearing mice the only detectable

30 / Negri et al.

CYTOKINE, Vol. 15, No. 1 (7 July, 2001: 27–38)

700

6

650

5 g/mouse

600

mg

550 500

1

400 350

MST: Figure 2.

3 2

450

300

4

0

5

10

15

20

25

30

35

Days A431

OVCAR3

GBL-F

10 (7–14)

11 (6–22)

26 (17–34)

Brain weight of i.c. tumour-bearing mice.

Mice were injected i.c with saline (20 l/mouse, n=4), A431 (2.5105 cells/mouse, n=6), OVCAR3 (2.5105 cells/mouse, n=6), or GBLF (1106 cells/mouse, n=5) and sacrificed when body weight was reduced at least 25% or when moribund. Data are expressed as meanSE (errors bars) of 4–6 brains per group. The solid and the broken lines indicate meanSE, respectively, of brain weight of mice injected with saline. MST=median survival time (range). Increase of brain weight due to tumour growth was assessed based on comparison of mean brain weight of tumour-injected versus saline-injected mice by Student’s t-test. Only GBLF tumourbearing brains showed a significant (P<0.05) weight increase. Data are from 1 of 4 separate experiments with similar results.

Figure 4. Food consumption following i.c. injection of tumour cell lines or saline (control). Mice were injected i.c. with saline (20 l/mouse, n=10) ( ), A431 (2.5105 cells/mouse, n=6) ( ), OVCAR3 (2.5105 cells/mouse, n=6) ( ), or GBLF (1106 cells/mouse, n=6) (), and food intake was measured daily. Data are expressed as g/mouse (mean value for n mice per cage). The decline in food intake (from days 1 to 14 and from days 14 to 32) in mice bearing A431 and GBLF tumours, respectively, and the consistently low food intake in OVCAR3treated mice, as compared to saline-injected controls, were statistically significant (P<0.05) by analysis of variance. Data are from 1 of 4 separate experiments with similar results.

mice. Brains of animals injected with GBL-F were not analyzed for murine cytokines, because this cell line did not give a pronounced cachexia syndrome (see Fig. 1).

Improvement of cancer cachexia syndrome by treatment with antibody specific for murine IL-6 receptor (muIL-6R)

Figure 3. Hystologic section of brain from an A431 i.c. tumourbrearing mouse. Mice, injected i.c. with tumours were sacrified when mouribund. Brains were removed and fixed in formalin. The 2 m coronal section, stained with haematoxilin-eosin shows tumour infiltrating the brain from the injection site (white asterisk) down to the hypothalamus (arrows). (4 magnification).

cytokine expression was a faint band corresponding to LIF mRNA. Brains of i.c. A431 or OVCAR3 tumour-injected mice were analyzed for the presence of murine cytokine mRNA. Representative data are shown in Figure 6. Compared to saline-injected mice there was an increase of IL-6 mRNA and IFN- mRNA in A431-bearing

To investigate the role of brain-released IL-6 in cancer cachexia syndrome, mice were injected i.c. with a neutralizing mAb, 15A7, directed against the muIL-6R (two different doses), or with control isotypematched mAb simultaneously with OVCAR3 or A431 cells given i.c. IL-6 neutralizing mAb 15A7 led to a dose dependent delay in body weight loss and to an increase in food intake that were significant only in the OVCAR3 tumour-bearing mice treated with the higher dose of anti-IL-6R mAb (Fig. 7A and B). This abatement of cancer cachexia syndrome symptoms was reflected in the dose dependent longer survival time of IL-6R antibody-treated mice, which was significant at the higher mAb dose only in the OVCAR3 tumourbearing mice (Fig. 8, Table 2). The treatment did not suppress tumour growth, as demonstrated by in vitro proliferation assays in presence of mAb 15A7 and by the slight (not significant) increase in brain weight of the tumour-bearing mice treated with this mAb as compared to that of control mAb-treated mice (Fig. 7 C and F).

Cytokines and cancer cachexia / 31

Figure 5.

Expression of human cytokines by PCR analysis.

PCR was performed on mRNA from human cell lines maintained in vitro and from tumour-bearing brains using specific primers for human cytokines and -actin. Results are from analysis of one brain from each group of tumour-bearing mice and saline-injected controls. M: Molecular weight marker (x 174 DNA Hae III digest).

TABLE 1. Serum cytokines in mice injected with human tumour cell lines. Serum interleukin-(IL-)6 levels were measured by enzyme linked immunoabsorbent assay (ELISA). Data are expressed as median of positive sera for each group and number of positive samples/number of samples tested (in parentheses). In control mice injected with saline, no IL-6 was detectable (level of sensitivity: 2 pg/ml). All the positive and negative results were confirmed by bioassay Cell line injection site

hu-IL-6 (pg/ml)

mu-IL-6 (pg/ml)

A431 s.c. i.c.

5.0 (4/5) 2.8 (8/9)

190 (3/5) 400 (7/9)

OVCAR3 i.p. i.c.

c2 (0/5) 3.9 (5/5)

43 (5/5) 40 (5/5)

GBLF s.c. i.c.

c2 (0/5) 30 (7/7)

c2 (0/5) c2 (0/7)

DISCUSSION Studies of the metabolic changes observed in cancer have implicated soluble molecules as mediators of cancer cachexia syndrome. In particular, several investigators have suggested the existence of

Figure 6. Expression of murine cytokines by PCR analysis. PCR was performed on mRNA from tumour-bearing brains using specific primers for murine cytokines and -actin. Results are from analysis of one brain from each group of tumour-bearing mice and saline-injected controls. M: Molecular weight marker (MassRuler DNA Ladder Low Range).

32 / Negri et al.

CYTOKINE, Vol. 15, No. 1 (7 July, 2001: 27–38)

OVCAR-3

A431

% initial body weight (g/mouse)

130

130 A

D 110

110

90

90 **

70

70

50

0

10

30

20

50

6

0

20

30

10

20

30

6 B

Food intake (g/mouse)

10

E

5

5

4

4 *

3

3

2

2

1

1

0

10

20

30

0 Days

600

600

Brain weight (mg)

C 550

500

500

450

450

400

400

350

350

300 Ab amount Figure 7.

F

550

–

α-IL6R 5 µg

α-IL6R 25 µg

Control 25 µg

300 Ab amount

–

α-IL6R 5 µg

α-IL6R 25 µg

Control 25 µg

Weight loss, food consumption and brain weight of i.c. tumour-bearing mice following i.c. administration of anti-IL-6R mAb.

Mice were injected i.c. with saline (20 l/mouse, n=5) (——), 2.5105 tumour cells/mouse with 5 g of anti-IL-6R mAb (n=6) (), 2.5105 tumour cells/mouse with 25 g of anti-IL-6R mAb (n=6) ( ), and 2.5105 tumour cells/mouse with 25 g of control mAb (n=5) (- - - -). Panels A and D: body weight loss. Data are expressed as percent of the initial mouse weight recorded daily. The decrease in body weight loss in OVCAR3-bearing mice treated with 25 g of anti-IL-6R mAb, as compared to mice treated with control mAb, was statistically significant (P=0.002) by analysis of variance. Panels B and E: food intake. Data are expressed as g/mouse (mean value for n mice per cage). The increase in food intake in OVCAR3-bearing mice treated with 25 g of anti-Il-6R mAb, as compared to mice treated with control mAb, was statistically significant (P=0.014) by analysis of variance. Panels C and F: brain weight. For evaluation of brain weight see Figure 2 legend. The increase in brain weight observed in animals treated with 25 g of anti-Il-6R mAb was not significant as compared to mice treated with saline.

‘‘toxohormones’’ responsible for effects such as impairment of liver drug metabolism and catalase activity.32 This suggestion was supported by the cachectogenic effect of implanted tumour cells genetically engineered to produce TNF23 and by the recently observed

association between high serum TNF levels and cachexia in prostate cancer patients.8 However several animal studies13,16,33 and some clinical reports34 have challenged this view and instead point to the key role of other cytokines such as LIF, IL-1, IL-6 and IFN-

Cytokines and cancer cachexia / 33

A431

100

100

80

80 % Survival

% Survival

OVCAR3

60

40

60

40

20

0

20

5

10

15 Days

20

25

30

0

5

10

15 Days

20

25

30

Figure 8. Kaplan-Meier survival curves for i.c. tumour-bearing mice treated i.c. with anti-mu-IL-6R or isotype-matched mAb of unrelated specificity. Mice were injected i.c with A431 or OVCAR3 (2.5105 cells/mouse) mixed with saline ( ), 5 g ( ) or 25 g ( ) of anti-murine IL-6R mAb, or 25 g ( ) of control mAb in 20 l/mouse. Results were pooled from two experiments in which identical treatments were used. See also Table 2.

TABLE 2. Effect of intracerebral treatment with anti-mu-interleukin-6-receptor (IL6R) on survival of i.c. tumour bearing mice. Results were pooled from two experiments in which identical treatments were used Cell line Treatment (dose)

Log-rank P value

No. of mice

MST (days)

Range (days)

A

OVCAR3 saline anti-mu-IL-6R mAb (5 g) anti-mu-IL-6R mAb (25 g) control mAb (25 g)

9 10 6 5

10 18 22 15

6–22 6–22 15–28 13–15

0.0797 0.0043 0.3171

A431 saline anti-mu-IL-6R mAb (5 g) anti-mu-IL-6R mAb (25 g) control mAb (25 g)

14 11 7 6

10 10 13 12

7–14 7–14 11–14 8–13

0.5315 0.0863 0.7280

B

0.0078

0.2536

A Each mAb treatment group vs saline; B The anti-mu-IL-6R mAb (25 g) treatment groups vs control mAb (25 g). Bold values are statistically significant.

in cancer cachexia syndrome. Here, we observed the cachetogenic effect of two human tumour cell lines, A431 and OVCAR3, when implanted i.c. in athymic mice, and the predominant release in the CNS of human IL-1 and IL-6 respectively. Treatment with a rat anti-mouse IL-6R mAb gives rise to a significant attenuation of cancer cachexia syndrome, accompanied by a significant longer survival in OVCAR3 bearing mice, strongly suggesting a key role for CNS released IL-6 in this model system. In both model systems, cancer cachexia syndrome was evident only when tumour cells were injected i.c. Tumour infiltration of the hypothalamus, visible in the majority of the brains considered histologically, might

have contributed to the development of cachexia but is unlikely to be its major determinant. In OVCAR3 and A431 tumours cachexia appears early, when the tumour mass is relatively small. In GBLF, in contrast, cachexia appears later and hypothalamic infiltration was present in 2/3 analyzed brains. We therefore tested whether cytokines, released either directly by the cancer cells or as a host response, might also underline the cancer cachexia syndrome. The human tumour cells, upon implantation in mice display a different pattern of cytokine production. In particular, GBLF cells produced in vitro TNF-
, while human TNF-
was detected only in the brains of mice i.c. bearing OVCAR3 tumours. These data point

34 / Negri et al.

to the role of environment factors and in particular of the host inflammatory response in determining the tumour cytokine production. It should be noted that all the cytokines analyzed, with the only exception of IFN, exhibit a cross-species activity and accordingly the human cytokines are fully competent in paracrine activation of the murine environment. Several cytokines were detected both locally and peripherally, with IL-6 production in particular above threshold levels in most of the tumour-bearing mice. The presence of human IL-6 in the serum of mice with i.c. tumours indicates that IL-6 leaks into the periphery, in agreement with reports demonstrating that CNS-administered cytokines are rapidly transported from the brain to the blood.35–37 While in all three model systems, including the GBLF glioblastoma, the i.c.-produced IL-6 appears to be transported in the blood at levels directly dependent on tumour volume, no clear correlation between tumour mass and human IL-6 serum levels was observed when the tumours were grown peripherally. The latter discrepancy might rest in a difference in tumour blood supply (for instance, s.c. A431 tumours are highly vascularized as compared to s.c. GBLF tumours) or to the local presence of catabolic factors. Mice injected i.c. with A431 and OVCAR3 cells also exhibited very high serum levels of host-derived (murine) IL-6. This could reflect in part the efflux of IL-6 produced in the brain in response to the implanted tumour (as in the case of A431 tumours, see Fig. 6), or a peripheral response to IL-6-inducing, tumour-derived cytokines. We previously showed that CNS-injected IL-1 is a potent inducer of peripheral IL-6,38 raising the possibility that the high murine IL-6 levels observed in A431-tumour bearing mice are due to the human IL-1 production by this tumour. It should be noted that because our studies were performed in athymic mice, the host-mediated responses are not due to the immunogenicity of the tumours but probably reflect responses by other cell types. In particular, IL-6 is produced mostly by cells of the monocyte/macrophage lineage and by microglial and endothelial cells. It should be noted that increased murine IL6 serum levels are necessary but not sufficient for the induction of cancer cachexia syndrome, since IL-6 serum levels were also high even in the absence of syndrome. Moreover, the acute phase protein SAA, previously shown to accompany IL-6 serum increases,25 was present at high levels in A431 tumourbearing mice regardless of injection route or cachectic effects (data not shown). Together, these observations indicate that cytokine serum levels could not be used as predictors of cancer cachexia syndrome development and that CNS activity of the cytokine is important in the induction of syndrome. These data suggest that the

CYTOKINE, Vol. 15, No. 1 (7 July, 2001: 27–38)

net metabolic effects of cachectic factors may depend on paracrine effects of their local concentration in the brain and that hypothalamic actions of these mediators may outweigh the effects of circulating levels. Information about the contribution of cytokines to cachexia induction has come from studies using mAb specifically directed against cytokines or their receptors in animal models29,39,40 and even in clinical settings.34 In light of the increased circulating levels of both human and murine IL-6 in our study, we used an anti-IL-6R mAb as a tool to evaluate the role of IL-6, irrespective of its origin, in OVCAR3- and A431induced cancer cachexia syndrome. Syndrome’s signs and symptoms were significantly reduced and survival time significantly increased only in the case of mAbtreated OVCAR3-tumour bearing mice, even if a trend toward survival prolongation was observed in A431tumour bearing mice. These data, consistent with other reports of organ-specific local action of cytokines23, and in keeping with the different pattern of released cytokines, strongly suggest that centrally released IL-6 is the principal mediator of OVCAR3-induced cancer cachexia syndrome. In view of the heterogeneity of tumours, it seems unlikely that IL-6 is the common mediator of all cancer cachexia syndrome. In fact, several cytokines might play a role or at least their administration can produce both clinical and biochemical features commonly seen in syndrome. For instance, A431 has the highest expression of IL-1 (alpha and beta) among the tumours tested, and CNS-administered IL-1 is a much more potent anorectic cytokine than IL-6.22,24 The particularly evident anorectic effect in the A431-treated animals in which food intake steadily declined, further suggest the primary involvement of IL-1 rather than IL-6 in mediating cancer cachexia syndrome in the i.c. A431 model. It is likely that various cytokines and other cachectic factors act in synergy to produce cancer cachexia syndrome, analogous to the events in other diseases associated with cytokine production, such as septic shock and systemic inflammatory response syndrome. For instance, LIF, a member of the IL-6 family of cytokines whose receptors use the gp130 signal transducer, and which induces cachexia and anorexia,13,24 and that was detected in both A431 and OVCAR3 cells, might also contribute to cancer cachexia syndrome. Recently, Inui described the possible alterations in the hypothalamic neuropeptide circuitry in cancer cachexia syndrome and pointed out that in tumourbearing states, cytokines could elicit effects on energy homeostasis that mimic leptin, a hormone secreted by adipose tissue, that controls food intake and energy expenditure via neuropeptidergic effector molecules within the hypothalamus.4

Cytokines and cancer cachexia / 35

TABLE 3. Polymerase chain reaction (PCR) primers used to amplify human and murine cytokines. The primers were synthesized based on published sequences or purchased from CLONTECH

Cytokine huIL-6 Forward Reverse muIL-6 Forward Reverse huIL-1
Forward Reverse muIL-1
Forward Reverse huIL-1 Forward Reverse muIL-1 Forward Reverse huIFN- Forward Reverse muIFN- Forward Reverse huTNF-
Forward Reverse huTNF-
Forward Reverse huLIF Forward Reverse

Length amplified fragment (bp)

PCR primers

Reference

ATGAACTCCTTCTCCACAAGCGC GAAGAGCCCTCAGGCTGGACTG

627

46

ATGAAGTTCCTCTCTGCAAGAGACT CACTAGGTTTGCCGAGTAGATCTC

638

Clontech

CAAGGAGAGCATGGTGGTAGTAGCAACCAACG TAGTGCCGTGAGTTTCCCAGAAGAAGAGGAGG

491

47

ATGGCCAAAGTTCCTGACTTGTTT CCTTCAGCAACACGGGCTGGTC

625

Clontech

ATGGCAGAAGTACCTAAGCTCGC ACACAAATTGCATGGTGAAGTCAGTT

801

48

GACGTTCCCATTAGACAG TTCACCATGGAATCTG

451

Clontech

ATGAAATATACAAGTTATATCTTGGCTTT GATGCTTTCGACCTCGAAACAGCAT

493

49

TGAACGCTACACACTGCATCTTGG CGACTCCTTTTCCGCTTCCTGAG

460

Clontech

ATGAGCACTGAAAGCATGATCCGG GCAATGATCCCAAAGTAGACCTGCCC

694

ATGAGCACAGAAAGCATGATCCGC CCAAAGTAGACCTGCCCGGACTC

692

Clontech

CGCCACCCATGTCACAACAACC CCAACACGGCGATGATCTGCTTA

487

51

The availability of two new human tumour models of cancer cachexia syndrome, as well as the possibility to test in these models the disregulation in hypothalamic neuropeptide circuitry, will hopefully help in the elucidation of the mechanism of cancer cachexia syndrome induction and in the search for potential therapeutic interventions for this syndrome.

MATERIALS AND METHODS Tumour cell lines The human A431 vulva epidermoid carcinoma and OVCAR3 ovarian carcinoma cells (provided by ATCC, Rockville, MD) and GBLF glioblastoma cell line (derived from a primary tumour and established in vitro at Istituto Nazionale Tumouri)41 were selected on the basis of their ability to grow in vivo in athymic mice. Cells were maintained in RPMI-1640 medium supplemented with 10% FCS under standard culture conditions. The lines have different

50

in vitro doubling times (A431, 22.7 h; OVCAR3, 54.6 h; GBLF, 32.4 h).

Antibodies Rat mAb 15A7, specific for the mouse IL-6 receptor (IL-6R), was kindly provided by Dr. Jacques Van Snick (Bruxelles, Belgium).42 The mAb was purified from hybridoma supernatant by affinity chromatography on a Sepharose-protein G column (Pierce, Rockford, IL). Purity was confirmed by 4–15% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) in a precast slab gel using the automated microprocessor-driven Phastsystem (Pharmacia). Rat mAb 187.1, directed against the light chain of murine immunoglobulin and purified from the supernatant of the hybridoma (ATCC), served as an isotype-matched control.

Mice Pathogen-free female athymic (nu/nu, CD1 background) mice, 6 to 8 weeks old, were obtained from Charles River (Calco, Como, Italy). Mice were held for 1 to 2 weeks before

36 / Negri et al.

the start of experiments. Animals were housed under sterile conditions and received autoclaved food and water. Treatments were approved by an ethical committee and performed according to institutional and European guidelines.

Intracerebral injection and measurement of cancer cachexia syndrome parameters Cell of the three lines (2.5105 A431 and OVCAR3 cells; 1.0106 of GBLF cells) were injected i.c. through a 28-gauge needle into a lateral cerebral ventricle of etheranesthetized mice. Briefly, a 3/8-inch, 28-gauge needle attached to a 0.5-ml allergy syringe was inserted perpendicularly through the skull into the brain. All i.c. injections were in a minimal volume of 20 l. Control mice received an equal volume of sterile saline. Tumour take was assessed by histology and immunohistochemistry with anti-human cytokeratin 8/18 (Dako), and by an increase in brain weight. In parallel, groups of 4–5 mice were injected s.c. with A431 (2.5106/mouse) or GBLF (10106/mouse) cells and i.p. with OVCAR3 (10106/mouse) cells resuspended in 200 l of saline. In two independent experiments, anti-IL-6R or the control mAb was injected i.c. at the doses indicated in the text simultaneously with cells. The mAb neutralization effect was evaluated by comparison of body weight loss, food consumption and survival curves. Blood was obtained from the retro-orbital plexus under ether anesthesia and serum was prepared. Individual body weight and food intake per cage were recorded daily at 1000 hours. Weight change was calculated based on the weight measured at the beginning of treatment. At the end of the experiments (25 days for s.c. tumour-bearing mice) or when moribund (for i.c. or i.p.-injected tumour-bearing mice), animals were sacrificed and brains removed, weighed and frozen in liquid nitrogen or fixed in formalin or Bouin reagent and embedded in paraffin. Slices of 2 m were stained by haematoxilin-eosin and examined microscopically.

Statistical analysis Body weight changes (individual animals), average food intake (mean value from 5–6 mice per cage) and brain weight (meanSE) were calculated. ANOVA, Student t test and analysis of variance were used to determine significant differences (P<0.05) between tumour-bearing and control mice (saline-treated). The effect of anti-IL-6R mAb treatment on survival was evaluated by Wilcoxon and log-rank nonparametric tests.

Cytokine quantitation Released cytokines were quantified in culture supernatants and biological fluids (serum, ascitic fluid) of mice by biological assays or ELISA. TNF-
was assessed as cytotoxicity on L929 cells in the presence of 1 g/ml of actinomycin D as described,43 using recombinant TNF-
as a standard. Sensitivity of the assay was around 20 pg/ml. IL-6 was measured as hybridoma growth factor using 7TD1 cells (a kind gift of Dr van Snick ) as described.44 IL-6 activity is expressed as costimulatory units/ml using rIL-6 as a standard. Sensitivity of the assay was 50 U/ml.

CYTOKINE, Vol. 15, No. 1 (7 July, 2001: 27–38)

hIL-6, muIL-6, hIL-1
, hIL-1, hLIF, and muIFN- were measured using commercially available ELISA kits (range of sensitivity:1–5 pg/ml) (Genzyme, Cambridge, MA, USA; Amersham, UK; Benfer-Scheller, Milan, Italy). According to the manufacturer’s specifications all ELISA kits were specific for the indicated cytokine and did not cross-react with other cytokines. SAA was measured by ELISA as described.45

RT-PCR analysis of cytokine expression Total RNA was prepared from 2-5106 cells or brains homogenized in 4M guanidine isothiocyanate and passed over a cesium chloride gradient: 1 g of total RNA was reverse-transcribed for 2 h at 42C using the Gene Amp kit (Perkin-Elmer). -actin cDNA was amplified to confirm the efficiency of reverse transcription. cDNA (1 l) was amplified by cytokine-specific PCR primers synthesized based on published sequences or commercially available (Table 3). All specific primers were designed to span intron-exon junctions in order to identify PCR bands amplified from RNA or from contaminating DNA. PCR was carried out in a 25-l reaction volume containing 10 pM primers, 0.2 mM each dNTP, 1 mM MgCl2 and 1.25 U Taq polymerase (Promega) for 40 cycles at 94C for 1 min, 60C for 2 min and 72C for 3 min. PCR products were analyzed on a 2% ethidium bromide-stained agarose gel. Analyses were repeated 2–3 times with superimposable results.

Acknowledgements This work was partially supported by AIRC/ FIRC and CNR Target Project in Biotechnology. We thank A. Cipollina for secretarial assistance and M. Azzini for photographic reproductions.

REFERENCES 1. Nelson KA, Walsh D, Sheehan FA (1994) The cancer anorexia-cachexia syndrome. J Clin Oncol 12:213–225. 2. Tisdale MJ (1997) Biology of cachexia. J Natl Cancer Inst 89:1763–1773. 3. Plata-Salaman CR (1996) Anorexia during acute and chronic disease. Nutrition 12:69–78. 4. Inui A (1999) Cancer anorexia-cachexia syndrome: are neuropeptides the key? Cancer Res 59:4493–4501. 5. Matthys P, Billiau A (1997) Cytokines and cachexia. Nutrition 13:763–770. 6. Argiles JM, Lopez-Soriano J, Busquets S, Lopez-Soriano FJ (1997) Journey from cachexia to obesity by TNF. FASEB J 11:743–751. 7. Okada S, Okusaka T, Ishii H, Kyogoku A, Yoshimori M, Kajimura N, Yamaguchi K, Kakizoe T (1998) Elevated serum interleukin-6 levels in patients with pancreatic cancer. Jap J Clin Oncol 28:12–15. 8. Nakashima J, Tachibana M, Ueno M, Miyajima A, Baba S, Murai M (1998) Association between tumor necrosis factor in serum and cachexia in patients with prostate cancer. Clin Cancer Res 4:1743–1748. 9. Beutler B, Greenwald D, Hulmes JD, Chang M, Pan YC, Mathison J, Ulevitch R, Cerami A (1985) Identity of tumour

Cytokines and cancer cachexia / 37 necrosis factor and the macrophage-secreted factor cachectin. Nature 316:552–554. 10. Strassmann G, Masui Y, Chizzonite R, Fong M (1993) Mechanisms of experimental cancer cachexia. Local involvement of IL-1 in colon-26 tumor. J Immunol 150:2341–2345. 11. Moldawer LL, Georgieff M, Lundholm K (1987) Interleukin 1, tumour necrosis factor-alpha (cachectin) and the pathogenesis of cancer cachexia. Clin Physiol 7:263–274. 12. Fong Y, Moldawer LL, Marano M, Wei H, Barber A, Manogue K, Tracey KJ, Kuo G, Fischman DA, Cerami A (1989) Cachectin/TNF or IL-1 alpha induces cachexia with redistribution of body proteins. Am J Physiol 256:R659–R665. 13. Metcalf D, Gearing DP (1989) Fatal syndrome in mice engrafted with cells producing high levels of the leukemia inhibitory factor. Proc Natl Acad Sci USA, 86:5948–5952. 14. Mori M, Yamaguchi K, Honda S, Nagasaki K, Ueda M, Abe O, Abe K (1991) Cancer cachexia syndrome developed in nude mice bearing melanoma cells producing leukemia-inhibitory factor. Cancer Res 51:6656–6659. 15. Hack V, Gross A, Kinscherf R, Bockstette M, Fiers W, Berke G, Droge W (1996) Abnormal glutathione and sulfate levels after interleukin 6 treatment and in tumor-induced cachexia. FASEB J 10:1219–1226. 16. Matthys P, Dijkmans R, Proost P, Van Damme J, Heremans H, Sobis H, Billiau A (1991) Severe cachexia in mice inoculated with interferon-gamma-producing tumor cells. Int J Cancer 49:77–82. 17. Black K, Garrett IR, Mundy GR (1991) Chinese hamster ovarian cells transfected with the murine interleukin-6 gene cause hypercalcemia as well as cachexia, leukocytosis and thrombocytosis in tumor-bearing nude mice. Endocrinology 128:2657–2659. 18. Fujimoto-Ouchi K, Tamura S, Mori K, Tanaka Y, Ishitsuka H (1995) Establishment and characterization of cachexiainducing and non-inducing clones of murine colon 26 carcinoma. Int J Cancer 61:522–528. 19. Faggioni R, Benigni F, Ghezzi P (1995) Proinflammatory cytokines as pathogenetic mediators in the central nervous system: brain-periphery connections. Neuroimmunomodulation 2:2–15. 20. Plata-Salaman CR (1998) Cytokines and anorexia: a brief overview. Sem Oncol 25:64–72. 21. Masotto C, Caspani G, De Simoni MG, Mengozzi M, Scatturin M, Sironi, Carenzi A, Ghezzi P (1992) Evidence for a different sensitivity to various central effects of interleukin-1 beta in mice. Brain Res Bull 28:161–165. 22. Plata-Salaman CR, Sonti G, Borkoski JP, Wilson CD, French-Mullen JM, brahms (1996) Anorexia induced by chronic central administration of cytokines at estimated pathophysiological concentrations. Physiol Behavior 60:867–875. 23. Tracey KJ, Morgello S, Koplin B, Fahey TJ, Fox J, Aledo A, Manogue KR, Cerami A (1990) Metabolic effects of cachectin/ tumor necrosis factor are modified by site of production. Cachectin/ tumor necrosis factor-secreting tumor in skeletal muscle induces chronic cachexia, while implantation in brain induces predominantly acute anorexia. J Clin Invest 86:2014–2024. 24. Plata-Salaman CR (1996) Anorexia induced by activators of the signal transducer gp 130. Neuroreport 7:841–844. 25. Langstein HN, Doherty GM, Fraker DL, Buresh CM, Norton A (1991) The role of gamma-interferon and tumor necrosis factor
in an experimental rat model of cancer cachexia. Cancer Res 51:2302–2306. 26. Plata-Salaman CR, Ilyin SE, Gayle D (1998) Brain cytokine mRNAs in anorectic rats bearing prostate adenocarcinoma tumor cells. Am J Physiol 275:R566–R573. 27. Ilyin SE, Gayle D, Gonzalez-Gomez I, Miele ME, Plata-Salaman CR (1999) Brain tumor development in rats is associated with changes in central nervous system cytokine and neuropeptide systems. Brain Res Bull 48:363–373. 28. Bibby MC, Double JA, Ali SA, Fearon KCH, Brennan RA, Tisdale MJ (1987) Characterization of a transplantable adenocarcinoma of the mouse colon producing cachexia in recipient animals. J Natl Cancer Inst 78:539–546. 29. Strassmann G, Fong M, Kenney JS, Jacob CO (1992) Evidence for the involvement of interleukin 6 in experimental cancer cachexia. J Clin Invest 89:1681–1684.

30. Mori K, Fujimoto-Ouchi K, Ishikawa T, Sekiguchi F, Ishitsuka H, Tanaka Y (1996) Murine interleukin-12 prevents the development of cancer cachexia in a murine model. Int J Cancer 67:849–855. 31. Tamura S, Ouchi KF, Mori K, Endo M, Matsumoto T, Eda H, Tanaka Y, Ishitsuka H, Tokita H, Yamaguchi K (1995) Involvement of human interleukin 6 in experimental cachexia induced by a human uterine cervical carcinoma xenograft. Clin Cancer Res 1:1353–1358. 32. Garattini S, Ghezzi P, D’Incalci M (1988) Effects of cancer disease on the metabolism of anticancer agents. Pharmacology & Therapeutics 37:57–65. 33. Cahlin C, Korner A, Axelsson H, Wang W, Lundholm K, Svanberg E (2000) Experimental cancer cachexia: the role of hostderived cytokines interleukin (IL)-6, IL-12, interferon-gamma, and tumor necrosis factor alpha evaluated in gene knockout, tumorbearing mice on C57 Bl background and eicosanoid-dependent cachexia. Cancer Res 60:5488–5493. 34. Emilie D, Wijdeness J, Gisselbrecht C, Jarrousse B, Billaud E, Blay J-Y, Gabarre J, Gaillard J-P, Brochier J, Raphael M, Boue F, Galanaud P (1994) Administration of an anti-interleukin-6 monoclonal antibody to patients with acquired immunodeficiency syndrome and lymphoma: effect on lymphoma growth and on B clinical symptoms. Blood 84:2472–2479. 35. Banks WA, Kastin AJ, Durham DA (1989) Bidirectional transport of interleukin-1 alpha across the blood-brain barrier. Brain Research Bulletin 23:433–437. 36. Banks WA, Kastin AJ, Ehrensing CA (1994) Blood-borne interleukin-1 alpha is transported across the endothelial blood-spinal cord barrier of mice. Journal of Physiology 479:257–264. 37. Kumar S, Pandey P, Bharti A, Jin SF, Weichselbaum R, Weavers D, Kufe D, Kharbanda S (1998) Regulation of DNAdependent protein kinase by the Lyn tyrosine kinase. J Biol Chem 273:25654–25658. 38. De Simoni MG, Sironi M, De Luigi A, Manfrini A, Mantovani A, Ghezzi P (1990) Intracerebroventricular injection of interleukin 1 induces high circulating levels of interleukin 6. J Exp Med 171:1773–1778. 39. Tsujinaka T, Fujita J, Ebisui C, Yano M, Kominami E, Suzuki K, Tanaka K, Katsume A, Ohsugi Y, Shiozaki H, Monden M (1996) Interleukin 6 receptor antibody inhibits muscle atrophy and modulates proteolytic systems in interleukin 6 transgenic mice. J Clin Invest 97:244–249. 40. Matthys P, Heremans H, Opdenakker G, Billiau A (1991) Anti-interferon-gamma antibody treatment, growth of Lewis lung tumours in mice and tumour-associated cachexia. Eur J Cancer 27:182–187. 41. Ruffini PA, Rivoltini L, Silvani A, Boiardi A, Parmiani G (1993) Factors, including transforming growth factor , released in the glioblastoma residual cavity, impair activity of adherent lymphokine-activated killer cells. Cancer Immunol Immunother 36:409–416. 42. Coulie PG, Vink A, Van Snick J (1990) A monoclonal antibody specific for the murine IL-6-receptor inhibits the growth of a mouse plasmacytoma in vivo. Curr Top Microbiol Immunol 166:43–46. 43. Aggarwal BB, Kohr WJ, Hass PE, Moffat B, Spencer SA, Henzel WJ, Bringman TS, Nedwin GE, Goeddel DV, Harkins RN (1985) Human tumor necrosis factor. Production, purification, and characterization. J Biol Chem 260:2345–2354. 44. Sironi M, Breviario F, Proserpio P, Biondi A, Vecchi A, Van Damme J, Dejana E, Mantovani A (1989) IL-1 stimulates IL-6 production in endothelial cells. J Immunol 142:549–553. 45. de Oliveira RM, Sipe JD, de Beer FC, Loose LD, Bartle LM, Cecil D, Franzblau C (1994) Rapid, sensitive enzyme-linked immunosorbent assay (ELISA) for serum amyloid A (apoSAA) in human plasma and tissue culture fluids. Int J Exp Clin Invest 1:23–29. 46. Hirano T, Yasukawa K, Harada H, Taga T, Watanabe Y, Matsuda T, Kashiwamura S, Nakajima K, Koyama K, Iwamatsu A (1986) Complementary DNA for a novel human interleukin (BSF-2) that induces B lymphocytes to produce immunoglobulin. Nature 324:73–76.

38 / Negri et al. 47. Furutani Y, Notake M, Yamayoshi M, Yamagishi J, Nomura H, Ohue M, Furuta R, Fukui T, Yamada M, Nakamura S (1985) Cloning and characterization of the cDNAs for human and rabbit interleukin-1 precursor. Nucleic Acids Res 13:5869–5882. 48. Auron PE, Webb AC, Rosenwasser LJ, Mucci SF, Rich A, Wolff SM, Dinarello CA (1984) Nucleotide sequence of human monocyte interleukin 1 precursor cDNA. Proc Natl Acad Sci USA , 81:7907–7911. 49. Devos R, Cheroutre H, Taya Y, Degrave W, Van Heuverswyn H, Fiers W (1982) Molecular cloning of human immune

CYTOKINE, Vol. 15, No. 1 (7 July, 2001: 27–38) interferon cDNA and its expression in eukaryotic cells. Nucleic Acids Res 10:2487–2501. 50. Pennica D, Nedwin GE, Hayflick JS, Seeburg PH, Derynck R, Palladino MA, Kohr WJ, Aggarwal BB, Goeddel DV (1984) Human tumour necrosis factor: precursor structure, expression and homology to lymphotoxin. Nature 312:724–729. 51. Gough NM, Gearing DP, King JA, Willson TA, Hilton DJ, Nicola NA, Metcalf D (1988) Molecular cloning and expression of the human homologue of the murine gene encoding myeloid leukemia-inhibitory factor. Proc Natl Acad Sci USA 85:2623–2627.