Parathyroid hormone–related protein (PTHrP) as a causative factor of cancer-associated wasting: Possible involvement of PTHrP in the repression of locomotor activity in rats bearing human tumor xenografts
Pharmaceutical Research Department IV, Kamakura Research Laboratories, Chugai Pharmaceutical Co., Kanagawa, Japan
Cancer-associated wasting (cancer wasting) is a complex disease often associated with anorexia and cachexia and characterized by several metabolic and behavioral abnormalities, such as early satiety, weakness, fatigue, depression and weight loss. Cancer wasting occurs in about 10% of advanced cases of malignancy. It correlates with poor prognosis irrespective of tumor mass or tumor metastases1 and affects the susceptibility to and tolerability of chemotherapy.2 Furthermore, cancer wasting is not easily resolved by forced caloric intake.3 Several factors are thought to be involved in cancer wasting. TNF-α, IL-1β, IL-6, LIF and IFN-γ have been demonstrated to cause cachectic symptoms in animal models.4 Activin, a member of the TGF-β gene family, was found to induce cachectic symptoms in mice.5 PIF, a sulfated glycoprotein that induces protein catabolism in isolated muscle cells, has been detected in serum samples from cachectic mice and from cancer patients.1 LMF, a homologue of the plasma Zn-α2 glycoprotein which induces lipid mobilization and catabolism in rats, has been isolated from the urine of cancer patients.6 Nevertheless, the etiology of cancer cachexia is still not well understood.
Previously, we demonstrated that nude rats bearing the LC-6-JCK human lung cancer xenograft contained high levels of blood PTHrP and exhibited HHM. Administration of the humanized anti-PTHrP MAb raised against the N-terminal 34 amino acids of the human PTHrP (PTHrP1–34) restored blood calcium levels.7 In addition, anti-PTHrP antibody was effective against HHM even after animals received repeated administration of an osteoclast inhibitor, such as a bisphosphonate drug or a calcitonin analogue, and thereby became refractory to the drug (Onuma E, Saito H, Tsunenari T, Sato K, Yamada-Okabe H, Ogata E, data not shown). However, animals carrying tumors that secreted PTHrP often showed not only hypercalcemia but also wasting.8 These facts suggest that PTHrP plays important roles in wasting in addition to its systemic hypercalcemic effects. Because PTHrP has been shown to induce production of several proinflammatory cytokines,9 it may activate cytokine networks, which may result in cachexia. Indeed, Takahashi et al.10 reported that cancer patients with elevated plasma PTHrP levels tended to exhibit higher blood levels of IL-6, TNF-α and IL-8.
In this study, we examined the involvement of PTHrP in cancer wasting. Nude rats carrying the LC-6-JCK human lung cancer xenograft contained high blood levels of PTHrP and several human proinflammatory cytokines, such as IL-6, LIF and IL-8. Administration of anti-PTHrP antibody significantly improved both hypercalcemia and body weight loss but did not diminish either production of proinflammatory cytokines or acute-phase reactions. The increase in body weight from anti-PTHrP antibody was remarkable under an unlimited feeding condition but not under limited pair feeding. Additionally, nude rats bearing LC-6-JCK exhibited reduced locomotor activity and a low blood phosphorus level, which were restored by anti-PTHrP antibody. Although alendronate, a bisphosphonate drug, transiently but significantly decreased blood calcium, it affected neither locomotor activity nor the serum phosphorus level. These results indicate that PTHrP causes cancer-associated wasting syndrome at least in part by repressing physical activities and energy metabolism.
Human cancer cell lines LC-6-JCK (lung clear cell carcinoma), PAN-3-JCK (pancreatic adenocarcinoma), PAN-7-JCK (pancreatic adenocarcinoma), OCC-1-JCK (larynx squamous carcinoma), RCC-11-JCK (renal adenocarcinoma) and PHA-1-JCK (pharynx squamous carcinoma) were obtained from the Central Laboratory Animal Research Center (Kawasaki, Japan). FA-6 (pancreatic adenocarcinoma) was kindly provided by Dr. K. Sato (Tokyo Women's Medical University, Tokyo, Japan). All of the above cells were maintained in vivo in nude mice (BALB/cAJcl-nu).
The humanized anti-PTHrP MAb raised against the N-terminal 34 amino acids of human PTHrP (PTHrP1–34)7 and alendronate (Teijin, Osaka, Japan) were dissolved and/or diluted in saline, and 3 mg/kg of the anti-PTHrP antibody or 5 mg/kg of alendronate were administrated i.v. Control rats received the same volume of saline. For each injection, rats received 1 ml/kg of saline containing or not containing anti-PTHrP antibody or alendronate.
Tumor transplantation and determination of blood levels of calcium and PTHrP.
Five-week-old male nude rats (F344/N Jcl-rnu) and 5-week-old male nude mice (BALB/cAJcl-nu) were purchased from Clea (Tokyo, Japan). Unless otherwise specified, rats and mice were freely fed rodent chow containing 1.25% calcium and 1.06% phosphate (CE-2, Clea) and tap water. Tumors grown in nude mice were excised, and small pieces (approx. 3 mm3 blocks) were inoculated s.c. into nude rats and nude mice. Unless otherwise specified, tumor volume, body weight, blood iCa and plasma PTHrP were measured once a week after tumor inoculation. To determine blood iCa and plasma PTHrP, whole blood was obtained from the abdominal aorta under anesthesia with sodium pentobarbital (Nembutal; Abbott, North Chicago, IL). Concentrations of blood iCa and plasma PTHrP were determined with an autoanalyzer (Hitachi 7170, Tokyo, Japan) and by PTHrP IRMA (Mitsubishi, Tokyo, Japan), respectively.7 Animals were kept under illumination for 14 hr a day (5 A.M. to 7 P.M.) and without illumination for 10 hr a day (7 P.M. to 5 A.M.). Animals used in this experiment were treated in accordance with Chugai Pharmaceutical's ethical guidelines of animal care, handling and termination.
Determination of fats, muscle, albumin, α1-acid glycoprotein and proinflammatory cytokines with or without tumor resection and administration of anti-PTHrP antibody.
Forty-nine days after tumor transplantation, nude rats bearing the LC-6-JCK xenograft were divided into 3 groups: (i) vehicle, in which rats were administrated saline; (ii) oncotectomy, in which rats underwent tumor resection; and (iii) anti-PTHrP antibody, in which rats were administrated anti-PTHrP MAb. Non-tumor-bearing rats were also used as a control group. Each group consisted of 8 animals. Tumor resection was performed under anesthesia with sodium pentobarbital.11 Tumor resection and administration of anti-PTHrP MAb (3 mg/kg) or saline were carried out 49 days after tumor transplantation; all rats were killed 59 days after tumor transplantation, to measure body weight, indicated tissue weights and indicated parameters. Fat and muscle contents were determined by measuring the weights of the epididymal adipose tissue and gastrocnemius muscle, as described.8
Concentrations in the serum of calcium and albumin were determined with whole blood obtained from the abdominal aorta and an autoanalyzer (Hitachi 7170), and that of α1-acid glycoprotein was determined by a rat α1-acid glycoprotein EIA kit (Sceti, Tokyo, Japan). Serum levels of human cytokines (hIL-6, hIL-1β, hTNF-α, hLIF, hIL-8, hIL-5 and hIL-11) and rat cytokines [rIL-6, rIL-8 (GRO/CINC1), rTNF-α, and rIL-1β] were also determined with whole blood obtained from the abdominal aorta using ELISA kits [Quantikine human IL-6; Quantikine TNF-α; Quantikine human LIF; Quantikine human IL-5; Quantikine human IL-11; Quantikine TNF-α (R&D Systems, Minneapolis, MN); BIOTRAK human IL-β; BIOTRAK rat IL-6; BIOTRAK GRO/CINK1; and BIOTRAK rat IL-1β (Amersham, Aylesbury, UK)] and human IL-8 (TFB Inc., Tokyo, Japan) according to the manufacturers' instructions.
Determination of body weight change with or without limited feeding.
Fifty-six days after tumor transplantation, nude rats bearing the LC-6-JCK xenograft were divided into 3 groups: (i) vehicle, consisting of 5 animals; (ii) anti-PTHrP antibody, consisting of 5 animals fed the same amounts as the vehicle group (pair feeding); and (iii) anti-PTHrP antibody, consisting of 2 animals fed freely. Anti-PTHrP MAb (3 mg/kg) and saline were administrated 56 days after tumor transplantation; body weights were measured every day after administration of the antibody or vehicle. For pair feeding experiments, pairs of rats bearing LC-6-JCK were kept in separate cages equipped with a pellet feeder and pellet detector (PairMex; Osakamicro, Osaka Japan) but provided the same amount of CE-2 chow.
Determination of locomotor activity, food consumption, water intake and serum phosphorus with or without administration of anti-PTHrP antibody and alendronate.
Forty-two days after tumor transplantation, nude rats bearing LC-6-JCK were divided into 3 groups: (i) vehicle, in which rats were administrated saline; (ii) alendronate, in which rats were administrated alendronate; and (iii) anti-PTHrP antibody, in which rats were administrated anti-PTHrP MAb. Each group consisted of 8 animals. Anti-PTHrP MAb (3 mg/kg) and alendronate (5 mg/kg) were administrated 42 days after tumor transplantation. To determine locomotor activities, animals were kept in individual cages equipped with a Supermex infrared sensor system (Muromachi Kikai, Tokyo, Japan).12 Body weight and locomotor activity during the 10 hr dark conditions (7 P.M. to 5 A.M.) were monitored every day. To measure food and water intake, animals were kept in individual cages and the reduction in amounts of food and water within 24 hr was considered as the daily food and water intake. Concentrations of serum phosphorus were determined with whole blood obtained from the abdominal aorta and an autoanalyzer (Hitachi 7170).
Statistical analyses of the experimental results were carried out with the SAS (Cary, NC) statistical package (version 6.12). Differences in mean values between the 2 groups were examined using the unpaired t-test; p < 0.05 was considered significant.
Involvement of PTHrP in cancer wasting
Nude rats carrying the LC-6-JCK human lung cancer xenograft, designated LC-6 rats in this study, showed HHM and significant body weight loss starting about 3 weeks after tumor transplantation. Elevation of blood calcium and body weight loss correlated with increasing plasma levels of human PTHrP (Fig. 1). Although tumors reached 15 × 103 mm3 by 10 weeks after tumor transplantation, body weight rapidly decreased by about 50 g after 4 weeks of tumor transplantation (Fig. 1), suggesting the involvement of PTHrP in both HHM and cancer wasting.
Because PTHrP can induce several proinflammatory cytokines, we measured the serum concentrations of those cytokines that might cause wasting in LC-6 rats. As shown in Table I, LC-6 rats contained various human proinflammatory cytokines, such as hIL-6, hLIF, hIL-8, hIL-5 and hIL-11. The level of GRO/CINC1, the rat counterpart to hIL-8, also increased in LC-6 rats. However, despite increased levels in the blood of human PTHrP and several other human proinflammatory cytokines, the levels of rat IL-6 (rIL-6) and rTNF-α were not elevated in LC-6 rats (Table I). In addition, neither human nor rat IL-1β levels were elevated in the blood of LC-6 rats.
Table I. Blood Levels of Proinflammatory Cytokines in Nude Rats Bearing the LC-6-JCK Tumor Xenograft1
Whole blood was obtained from LC-6 rats, and concentrations of the cytokines were determined 42 days after tumor inoculation. Results are means with SEs.
Next, we examined whether blockage of PTHrP action restored body weight. Tumor resection and administration of anti-PTHrP antibody attenuated body weight loss to a similar extent, and the resulting body weight gain was accompanied by increased amounts of fat and muscle (Fig. 2). Serum glucose and triglyceride were also increased in LC-6 rats that received anti-PTHrP antibody (94 mg/dl in vehicle group vs. 115 mg/dl in antibody-treated group for glucose and 29 mg/dl in vehicle group vs. 36 mg/dl in antibody-treated group for triglyceride). Interestingly, tumor resection, but not administration of anti-PTHrP antibody, restored serum albumin and α1-acid glycoprotein, the murine counterpart to human CRP, both of which typically reflect acute-phase reactions caused by proinflammatory cytokines.13 We also examined whether anti-PTHrP antibody affected blood levels of proinflammatory cytokines. As shown in Figure 3, none of the hIL-6, hIL-11, hLIF, hIL-8 or GRO/CINC1 levels was decreased by administration of anti-PTHrP antibody in LC-6 rats, which is consistent with the finding that anti-PTHrP antibody does not improve acute-phase reactions. These results demonstrate that PTHrP caused wasting independently of the proinflammatory cytokine actions.
Involvement of PTHrP in locomotor activity
Next, we examined whether PTHrP induces cancer wasting by affecting food intake. LC-6 rats administrated anti-PTHrP antibody showed significant body weight gain when they were freely fed but not when limited feeding was imposed (Fig. 4). When LC-6 rats that were administrated anti-PTHrP antibody were fed the same (pair feeding) as those that received only vehicle, there was no significant increase in body weight, at least by 7 days after administration of anti-PTHrP antibody (Fig. 4). This raises the possibility that the anti-PTHrP antibody restored body weight by modulating eating behavior and/or physical activity. To address this possibility, we examined locomotor activity in LC-6 rats. LC-6 rats displayed reduced locomotor activity compared to non-tumor-bearing rats: locomotor activity (from 7 P.M. to 5 A.M. of the next day) was 1,780 counts in non-tumor-bearing rats and 704 counts in LC-6 rats. Administration of anti-PTHrP antibody significantly augmented locomotor activity (Fig. 5c), indicative of suppression of physical activity by PTHrP.
Next, we examined if wasting in HHM is solely dependent on increased blood calcium level. To address this question, we used alendronate, a bisphosphonate drug that suppresses osteoclasts, thereby reducing blood calcium in HHM patients.14 Although the augmentation of locomotor activity by anti-PTHrP antibody coincided with a decrease in blood calcium levels and body weight gain, administration of alendronate did not enhance locomotor activity despite significantly but transiently decreasing the blood calcium level and increasing body weight (Fig. 5a,b). In LC-6 rats that received alendronate, blood calcium was still at a high level on day 1 and dropped to the level attained by anti-PTHrP antibody on day 3, and change of body weight negatively correlated with that of blood calcium levels. However, locomotor activity in these rats remained at a low level, at least by day 7 (Fig. 5c). Furthermore, body weight gain of LC-6 rats after administration of anti-PTHrP antibody was associated with increased food and water intake (Fig. 5d,e). Because LC-6 rats showed a low serum phosphorus level, we also examined the effects of anti-PTHrP antibody on serum phosphorus and found that the antibody, but not alendronate, elevated serum phosphorus (Fig. 5f). These results demonstrate that PTHrP caused wasting at least in part by affecting physical activity and energy metabolism and that repression of physical activity and energy metabolism caused by PTHrP occurs independently of blood calcium levels.
PTHrP expression and cancer wasting in other xenografts
If PTHrP is a causative factor of wasting, wasting should occur in other xenograft models that secrete PTHrP. To address this question, we explored PAN-7-JCK human pancreatic cancer xenografts. Rats bearing PAN-7-JCK (PAN-7 rats) showed significant body weight loss (177 g in PAN-7 rats vs. 205 g in non-tumor-bearing rats) and increased blood calcium (1.8 mM in PAN-7 rats vs. 1.4 mM in non-tumor-bearing rats). Administration of anti-PTHrP antibody restored body weight and blood calcium (Table II). Wasting of animals bearing a xenograft that secreted PTHrP was not specific to rats; body weight loss and restoration by anti-PTHrP antibody were also observed in nude mice transplanted with LC-6-JCK or PAN-7-JCK (Table II). Moreover, wasting and hypercalcemia occurred in mice bearing other xenografts, such as FA-6, PAN-3-JCK, OCC-1-JCK, RCC-11-JCK or PHA-1-JCK, all of which expressed PTHrP and in which administration of anti-PTHrP antibody restored body weight and blood calcium levels, which further supports the hypothesis of direct involvement of PTHrP in wasting.
Table II. Plasma PTHrP Levels, Body Weights and Blood iCa Levels in Various Human Tumor Xenograft Models
Date of measurements (days after tumor inoculation): PAN-7-JCK (in nude rats), day 19; LC-6-JCK (in nude rats), day 45; FA-6 (in nude mice), day 34; PAN-3-JCK (in nude mice), day 28; PAN-7-JCK (in nude mice), day 19; OCC-1-JCK (in nude mice), day 14; LC-6-JCK (in nude mice), day 45; RCC-11-JCK (in nude mice), day 35; PHA-1-JCK (in nude mice), day 29.
Results are means with SEs.
Administration of anti-PTHrP antibody: 100 μg/mouse 2 times/week i.v. in nude mice, 3 mg/kg single i.v. in nude rats. Tumor origin:
Animals bearing xenografts that secreted PTHrP and thereby contained elevated blood PTHrP levels displayed both hypercalcemia and wasting. Administration of anti-PTHrP antibody restored blood calcium and body weight. However, hypercalcemia and wasting may occur by different mechanisms because a vitamin D3 analogue, cholecalciferol, caused hypercalcemia without significantly affecting body weight.15 Wasting that occurred independently of hypercalcemia was also reported by Capparelli et al.16 They demonstrated that osteoprotegerin improved hypercalcemia but not body weight loss in mice bearing the Colon26 tumor, which secretes PTHrP. Thus, it is likely that PTHrP-induced wasting is not simply caused by increased blood calcium but that other mechanisms are also involved.
In this study, we demonstrated, using LC-6 rats, that (i) body weight gain after administration of anti-PTHrP antibody occurred at a rate comparable to that of tumor resection under an unlimited feeding condition, (ii) body weight gain by the anti-PTHrP antibody was accompanied by restoration of locomotor activity and food and water intake and (iii) alendronate effectively but transiently decreased the blood calcium level and had no effect on either locomotor activity or serum phosphorus level. These results support the idea that the mechanism by which wasting occurs at least in part involves repression of physical activity and energy metabolism by PTHrP, which is evidenced by decreased locomotor activity, less eating and drinking and low serum phosphorus. We observed that the anti-PTHrP antibody did not significantly affect the food intake or body weights of non-tumor-bearing rats, demonstrating that neutralization of PTHrP essentially has no effect on normal animals (not shown). This is reasonable because PTHrP is a local factor and is not released to the general circulation in normal physiologic conditions. There was a transient but significant decrease of blood calcium level by a single administration of alendronate, but even repeated administrations of alendronate did not cause the prolonged and sustained suppression of blood calcium observed after administration of anti-PTHrP antibody (Onuma E, Saito H, Tsunenari T, Sato K, Yamada-Okabe H, Ogata E, data not shown).
In preliminary experiments, we found that osmoregulatory responses such as blood concentration of arginine vasopressin and free water clearance in the kidney, expression of neuropeptides that are involved in eating and drinking behavior and periodic behavioral patterns were deregulated in mice bearing the LC-6-JCK xenograft and that administration of anti-PTHrP antibody promptly restored them (details to be reported in a separate paper). Furthermore, both PTHrP and PTHR1 mRNAs were also expressed in the brain.17, 18 Thus, it is possible that PTHrP affects behavior through some of the neurotrophic factors and thereby through the CNS.
LC-6 rats contained certain blood levels of human IL-6, LIF, IL-8, IL-5 and IL-11 in addition to PTHrP. IL-6 and LIF were considered to cause cachexia, and IL-6 has been thought to be a key factor of cachexia.19, 20 In this study, anti-PTHrP antibody restored body weight but did not reduce serum IL-6 levels in LC-6 rats. The serum level of IL-6 after administration of anti-PTHrP antibody was 339 pg/ml, a level even higher than that of LC-6 rats which received only vehicle (199 pg/ml). Restoration of body weight independently of blood IL-6 level is consistent with the finding that anti-PTHrP antibody caused body weight gain without affecting acute-phase reactions caused by proinflammatory cytokines such as IL-6. Nevertheless, we cannot rule out the possibility that PTHrP cooperates with proinflammatory cytokines, especially with those produced by host cells, to cause wasting. Indeed, although PTHrP has been shown to induce production of proinflammatory cytokines in host animals,9 neither rat IL-6 nor rat TNF-α was induced in LC-6 rats. This is presumably due to the fact that nude mice and nude rats lack T cells that can produce these cytokines in response to PTHrP. However, blood levels of rat IL-8 were elevated in LC-6 rats, which can be explained by the fact that IL-8 is produced by osteoclasts and osteoclast-like cells21 and that PTHrP activates osteoclasts.22
In conclusion, PTHrP has the ability to repress locomotor activity, eating and drinking and energy metabolism; and deregulation of these physiologic activities and functions by PTHrP may be at least in part involved in cancer-associated wasting in patients with PTHrP-producing tumors.
We thank Y. Azuma, N. Shimizu, T. Watanabe and M. Hirabayashi for assistance with the experiments and F. Ford for proof reading the manuscript.