Repeated Cadmium Exposures Enhance the Malignant Progression of Ensuing Tumors in Rats

Michael P. Waalkes*,1, Sabine Rehm*,2 and M. George Cherian{dagger}

* Inorganic Carcinogenesis Section, Laboratory of Comparative Carcinogenesis, National Cancer Institute at the National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709; and {dagger} Department of Pathology, University of Western Ontario, London, Ontario, N6A 5C1, Canada

Received August 2, 1999; accepted November 5, 1999


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Prior studies show that a single subcutaneous (sc) exposure to cadmium (Cd) will induce injection site sarcomas (ISS) in rats. These tumors, thought clearly malignant, do not often metastasize or invade subdermal muscle layers because of their location. Recent evidence indicates that when tumorigenic cells chronically exposed to Cd in vitro are inoculated into mice, tumor progression and invasiveness in the mice are enhanced. Thus, we studied the effects of repeated Cd exposures on tumor incidence, progression, and metastatic potential in rats. Wistar (WF) and Fischer (F344) rats (30 per group) were injected sc in the dorsal thoracic midline with CdCl2 once weekly for 18 weeks with doses of 0, 10, 20, or 30 µmol Cd/kg. This resulted in total doses of 0, 180, 360, or 540 µmol/kg. One other group of each strain received a low, loading dose of Cd (3 µmol/kg) prior to 17 weekly injections of 30 µmol/kg (total dose 513 µmol/kg). Rats were observed for 2 years. Many F344 rats (57%) died within one week after the first injection of the highest dose, but WF rats were not affected. The low loading dose prevented acute lethality of the high dose in F344 rats. Surprisingly, latency (time to death by tumor) of ISS was the shortest in the groups given the low loading dose in both strains. ISS in these groups also showed the highest rate of metastasis and subdermal muscle layer invasion. Based on ISS incidence in the groups given the lowest total dose of Cd (180 µmoles/kg), F344 rats were more sensitive to tumor induction, showing an incidence of 37% compared to 3% in WF rats. On the other hand, Cd-induced ISS showed a higher overall metastatic rate in WF rats (18 metastatic ISS/68 total tumors in all treated groups; 27%) compared to F344 rats (6%). Immunohistochemically, the primary ISS showed high levels of metallothionein (MT), a cadmium-binding protein, while metastases were essentially devoid of MT. These results indicate that repeated Cd exposures more rapidly induce ISS. An initial low exposure to Cd further accelerates the appearance and enhances the metastatic potential and invasiveness of these tumors. The primary and metastatic ISS appear to have a differing phenotype, at least with regard to MT production. The association between multiple Cd exposures and enhanced metastatic potential of the ensuing tumors may have important implications in chronic exposures to Cd, or in cases of co-exposure of Cd with organic carcinogens, as in tobacco smoking.

Key Words: cadmium; carcinogenesis; rats; sarcoma; tumor progression; tumor invasion; tumor metastasis.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The toxic transition metal cadmium (Cd) is a known human carcinogen and a potent carcinogen in animals (IARC, 1993Go). In rats, a single sc exposure to soluble salts of Cd has resulted in a significant incidence (20–70%) of sarcomas at the site of the sc injection. (Gunn et al., 1964Go; IARC, 1993Go; Poirier et al., 1983Go; Waalkes et al., 1988Go, 1989Go, 1991bGo, 1999Go). Most of these injection site sarcomas (ISS) are typically fibrosarcomas. Other histological types of sarcomas, including histiocytic sarcomas, rhabdomyosarcomas, and osteosarcomas, are less common with Cd. The carcinogenic effects of Cd at the injection site have been assessed in several strains of rats including Chester Beatty (Roe et al., 1964Go), Hooded (Heath, et al. 1962Go; Heath and Webb, 1967Go), Noble (Waalkes et al., 1999Go), Fischer (Waalkes et al., 1991bGo) and several Wistar-derived strains (Gunn et al., 1963Go, 1964Go, 1967Go; Poirier et al., 1983Go; Waalkes et al., 1988Go, 1989Go, 1991aGo). The ISS first occur at about one year after a single sc Cd injection. There is some evidence of strain differences in sensitivity to Cd-induced ISS in rats, although this has not been studied directly. Though clearly malignant, these Cd-induced ISS are rarely locally invasive or metastatic (IARC, 1993Go; Waalkes et al., 1991bGo, 1999Go). For instance, in F344 rats treated with a single sc injection of 30 µmol Cd/kg, sc, only 1 metastatic ISS occurred out of the 21 total ISS that were induced (Waalkes et al., 1991bGo). ISS induced by Cd can also occur in mice and are strain-dependent (Waalkes and Rehm, 1994Go), but have not been reported in hamsters (IARC, 1993Go; Waalkes, et al., 1998). Cd exposure is also associated with tumors of the testes, prostate, adrenal, liver, pituitary, pancreas, and hematopoietic system in rodents (IARC, 1993Go; Waalkes and Misra, 1996Go; Waalkes and Rehm, 1994Go; Waalkes et al., 1999Go).

The toxic effects of metals, including Cd, can sometimes be mitigated by prior exposure to low doses of the same metal (Goering et al., 1994Go; Klaassen et al., 1999Go). The acquired self- tolerance to Cd is thought to have some basis in toxicokinetics but primarily concerns specifically modified tissue responses (Goering et al., 1994Go; Klaassen et al., 1999Go). Induction of the synthesis of metallothionein (MT), a metal-binding protein, after metal exposure is clearly one of the primary modifications in acquired tolerance to Cd (Klaassen et al., 1999Go). The carcinogenic effects of Cd can also be modified by low-dose pretreatment, but only for some tissues such as the testes (Waalkes et al., 1988Go). At the injection site, a low dose-Cd pretreatment does not block sarcoma formation induced by a subsequent carcinogenic dose of Cd (Waalkes et al., 1988Go). On the other hand, zinc treatment can reduce Cd-induction of ISS (Waalkes et al., 1989Go) while zinc deficiency states enhance the appearance ISS after Cd injection (Waalkes et al., 1991aGo). Generally speaking, the development of tolerance in Cd carcinogenesis is only poorly understood.

Although the carcinogenic mechanism of Cd is not well defined, recent in vitro evidence indicates that this metal may also enhance progression of tumor cells (Abshire et al., 1996Go; Haga et al., 1996Go, 1997Go). For instance, when a malignantly transformed rat myoblast cell line is chronically exposed to certain levels of Cd in vitro, and is then inoculated into immunodeficient mice, tumors progress much more rapidly and are highly invasive in comparison to tumors arising from control cells (Abshire et al., 1996Go). The enhanced malignant progression seen with in vitro Cd exposure causes increased host lethality after inoculation (Abshire et al., 1996Go). This enhancement of progression, based on accumulated in vitro dose of Cd (Abshire et al., 1996Go), indicates that repeated or continuous exposures to Cd may modify malignant progression. Similarly, fibrosarcoma cells made resistant to Cd by chronic in vitro exposure showed a much higher invasiveness into recombinant basement membranes than the parent cell line (Haga et al., 1997Go). Furthermore, Cd exposure of host-cell monolayer fibroblasts or endothelial cells can enhance invasiveness of a fibrosarcomatous tumor cell line (Haga et al., 1996Go). These results indicate that both tumor-cell and host tissue can be altered by Cd exposure in a way that facilitates tumor invasion and presumably metastasis. Such a "progressor" function for Cd could be very important in multiple exposures to this carcinogenic metal or in combination exposures of Cd and other carcinogens, as would occur, for example, in cigarette smoking. However, the effects of Cd on tumor progression or metastasis in vivo are unknown.

Thus, this study was designed to determine the effects of repeated Cd exposures on tumor incidence, progression, and metastasis in vivo. In this regard, two strains of rat were compared: one which is sensitive to acute Cd toxicity (F344) and one which is resistant (Wistar). We selected ISS as the tumor model, which can be induced in F344 and Wistar rats by a single sc injection of Cd (Waalkes et al., 1988Go, 1991bGo). In addition, it can be monitored externally to test the effects of multiple Cd exposures on tumor progression. Since MT is clearly associated with repeated Cd exposure (Klaassen, et al., 1999Go) and may be important in enhancing tumor cell invasiveness and metastasis (Haga et al., 1996Go, 1997Go, 1998Go), we carried out additional studies to investigate MT expression in primary and metastatic sarcomas by immunohistochemistry.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Animals and treatment.
A total of 150 F344/NCr (F344) and 150 WF/NCr (WF) male rats were obtained at 4 weeks of age from the Animal Production Area, NCI-FCRDC, Frederick, MD. Animals were housed 3 per hanging polycarbonate cage (cage size: 19 x 10.5 x 8 inches) with hardwood-chip bedding and were provided food (NIH-31 Open Formula 6% Modified Teklad Standard Diets, Madison, WI) and water (acidified tap) ad libitum. Lighting (flourescent) schedule was 12 h on (6 A.M. to 6 P.M.) and 12 h off (6 P.M. to 6 A.M.). Environmental temperatures were held between 68 and 72°F, with a relative humidity of 50 ± 5%. Animals were cared for and used humanely according to the U.S. Public Health Policy on the Care and Use of Animals and the Guide for the Care and Use of Laboratory Animals. The NCI-FCRDC animal facility, where the bioassay portion of this study was conducted, and its animal program are accredited by the Association for the Assessment of Laboratory Animal Care International.

The experimental design of the present study is shown in Table 1Go. For injection cadmium (CdCl2•µ21/2 H2O; J. T. Baker Co.) solutions were prepared in sterile normal saline. At 8 weeks-of-age, animals were divided into groups as designated in Table 1Go and treated subcutaneously (10.0 ml/kg) with 10, 20, or 30 µmol Cd/kg, as CdCl2, once weekly for 18 consecutive weeks. One other group of each strain was given a low loading dose (3 µmol/kg, sc) one week prior to 17 weekly injections of 30 µmol/kg. Controls received vehicle (saline). Injections were given in the dorsal thoracic midline (henceforth referred to as the injection site). Body weights, survival, and clinical signs were recorded throughout the experiment. Body weights were recorded weekly for the first 26 weeks and biweekly thereafter. Clinical signs were checked daily. Animals were killed when significant clinical signs developed or at 104 experimental weeks.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Experimental Design: Carcinogenicity of Multiple Injections of Cadmium Chloride to Male WF and F344 Rats
 
Pathology.
An extensive necropsy was performed on all animals whether found dead, killed during the experiment when appropriate clinical signs developed, or killed at the conclusion of the experiment. Injection site, testes, liver, kidney, prostate, pancreas, lung, and all abnormal tissues from each animal were fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned at 5 µm and stained with hematoxylin and eosin for histological examination. Development of injection site sarcomas (ISS) was a primary goal of this study. The externally visible nature of this tumor provided for accurate assessment of its development, which in turn, allowed for the sacrifice of animals under uniform clinical conditions based primarily on the absence of impairment of free access to food and water.

Metallothionein in primary and metastatic sarcomas.
Localization of MT was examined immunohistochemically in primary ISS and pulmonary metastases. Multiple sections of several sets (WF, 6 primaries and 6 pulmonary metastases; F344, 3 each) were stained for MT. The immunohistochemical localization of MT was performed by the peroxidase-antiperoxidase method using a polyclonal rabbit antibody against rat liver MT (Banerjee et al., 1982Go). This antibody readily cross-reacts with MT from various species as demonstrated in several previous studies (Banerjee et al., 1982Go; Nartey et al., 1987Go). In the final reaction, the color was developed with diaminobenzidine in the presence of 0.3% hydrogen peroxide. Hematoxylin was used as a counterstain. The specificity of the antibody for MT was tested with different control experiments as described earlier (Nartey et al., 1987Go). Normal rabbit serum was substituted for the primary antibody and used as the negative control, while sections of kidney from rats acutely treated with cadmium, which induces high levels of MT, were used as the positive control.

Data analysis.
In all cases, a one-sided probability level of p <= 0.05 was considered to indicate a significant difference. In pairwise comparison of lesion incidence, Fisher's exact test was used. Body weight data were examined by Dunnett's t-test after ANOVA. Survival was examined with the Cox test and the generalized Kruskal/Wallis test and was considered significantly different only if so indicated by both tests. In all cases, the number of rats at risk is defined as the number of rats surviving at the time of the appearance of the first tumor of any type (21 weeks). Latency of ISS was defined as the average time in weeks to death or sacrifice with ISS.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Effects of the Cd treatments on survival are shown in Table 2Go. In the animals given the same dose for 18 weeks, the initial 30 µmol/kg dose of Cd was quite toxic to the F344 rats, killing 57% of this group within the first week after exposure. The subsequent Cd doses were well tolerated in this group. The 20 µmol/kg dose of Cd also caused 3/30 acute deaths in F344 rats. The low loading dose (3 µmol/kg) abolished the acute lethality of the 30 µmol/kg dose in F344 rats. None of the Cd doses caused any acute lethality in WF rats. All other doses were well tolerated by both strains over the 18 weeks of exposure. Cd treatment significantly reduced average survival (in experimental weeks) compared to control in all treatment groups in both strains with the exceptions of the lowest accumulated Cd dosage group in both strains (180 µmol/kg). The lowest average survival in WF rats was in the group that received the low loading dose (3 µmol/kg) prior to repeated injections of 30 µmol/kg. The lowest survival in F344 was in the group receiving 30 µmol/kg repeatedly. Decreased survival was largely due to ISS. Early deaths prior to the appearance of the first tumor of any type in any group (21 experimental weeks) were eliminated from further consideration.


View this table:
[in this window]
[in a new window]
 
TABLE 2 Survival of WF and F344 Rats Treated with Multiple Injections of Cadmium Chloride
 
Body weights over the course of the experiment for the various groups are shown in Figure 1Go (WF) and Figure 2Go (F344). Body weights of the rats were suppressed in all Cd-treated groups. The WF rats showed a clear dose-related weight suppression (control > 180 > 360 > 513 >= 540 µmol/kg). On the other hand, F344 rats showed an unexpected pattern of body weight suppression relative to dose. This pattern in F344 rats at risk was control > 540 > 180 > 360 > 513 µmol/kg. The basis of this pattern in F344 rats is not immediately clear, but perhaps represents the response of subpopulations of varying sensitivity within the F344 rat strain, although this is speculative.



View larger version (20K):
[in this window]
[in a new window]
 
FIG. 1. Effect of Cd treatment on body weights of WF (Wistar) rats. Rats were treated according to protocol detailed in Table 1Go. Significant depression of body weights occurred at all doses of Cd. The WF rats showed a clear dose-related weight suppression as follows: control body weights > 180 > 360 > 513 >= 540 µmol/kg.

 


View larger version (21K):
[in this window]
[in a new window]
 
FIG. 2. Effect of Cd treatment on body weights of F344 (Fischer) rats. Rats were treated according to protocol detailed in Table 1Go. Significant depression of body weights occurred at all doses of Cd. The F344 rats showed an unexpected pattern of body weight suppression relative to dose as follows: control body weights > 540 > 180 > 360 > 513 µmol/kg.

 
The incidence, latency, and metastatic and invasive rates of ISS induced by multiple exposures to Cd are shown in Table 3Go. Metastasis was primarily to the lung while invasion occurred primarily into the subdermal muscle. In WF rats, the incidence of ISS showed a marked increase between 180 µmol/kg (3% ISS incidence) and 360 µmol/kg (>70% ISS incidence). The maximal rate of 79–82% ISS was seen at the highest 2 accumulated doses. The metastatic rate of the ISS in WF rats showed a complex relationship with total dose and was maximal in the group receiving the low loading dose, which was the second highest cumulative dose (513 µmol/kg). Invasiveness and latency of ISS showed a similarly complex dose-relationship in the WF rats and was maximal in the group receiving the low loading dose (513 µmol/kg). These data indicate that an initial low dose of Cd clearly enhances the progression of ISS in WF rats. In F344 rats, the incidence of ISS showed dose-related increases between 0 and 360 µmol/kg to a maximum of 68% and then held at a fairly constant rate (60 to 67%) at the highest 2 accumulated doses. The metastatic and invasive rates of the ISS in F344 rats was much less than in WF rats. Metastatic sarcomas occurred only in the group given a low loading dose (513 µmol/kg). Invasiveness and latency of ISS showed a similar pattern in the F344 rat and was maximal in the group receiving the low loading dose (513 µmol/kg). These data also indicate that an initial low dose of Cd appears to enhance the progression of ISS in F344 rats, although the relationship here is weaker than in WF rats.


View this table:
[in this window]
[in a new window]
 
TABLE 3 Incidence of Injection Site Sarcomas (ISS) Following Multiple Injections of Cadmium Chloride
 
Table 4Go compares the overall incidence of metastatic and invasive tumors and sites of metastasis. When directly compared, WF rats had 18 metastatic ISS out of 67 ISS in all Cd-treated groups (27%) compared to only a 6% rate of metastasis in F344 rats. The primary site for ISS metastasis was the lung and nearly 80% of metastatic ISS in WF rats included a pulmonary component. The ISS frequently showed multiple sites of metastasis, mostly to 2 but occasionally to 3 sites. Multiple metastases generally included the lung and some other tissue, most commonly lymph nodes or liver. The overall rate of invasiveness of the ISS in WF rats (21%) was similar to that in F344 rats (17%). Invasion was most commonly into the subdermal muscle layer but occasionally into the local bone at the injection site. The primary histological types of ISS that were induced by Cd treatment were fibrosarcomas, but histiocytic sarcomas were not uncommon. Rhabdomyosarcomas, mixed sarcomas, and osteosarcomas occurred much less frequently as ISS.


View this table:
[in this window]
[in a new window]
 
TABLE 4 Metastatic and/or Invasive Injection Site Sarcomas in WF and F344 Rats Given Multiple Injections of Cadmium Chloride
 
Because MT is clearly associated with chronic Cd exposure and is frequently linked to cadmium tolerance (Klaassen et al., 1999Go), immunohistochemical analysis of MT was performed in primary ISS and their corresponding pulmonary metastases in Cd-treated rats. A representative example of a primary fibrosarcoma and its pulmonary metastasis in a WF rat are shown (Fig. 3Go). In WF rats, the primary ISS stained strongly for MT and was frequently localized within the tumor-cell nucleus. On the other hand, pulmonary metastases showed little or no evidence of MT within the tumor cells. This staining pattern, with the primary ISS showing intense MT staining and the metastatic ISS essentially devoid of MT, was similar in primary and metastatic lesions from F344 rats. Despite showing minimal MT content, the metastatic ISS maintained their aggressive nature as evidenced by invasion and compression of normal adjacent pulmonary tissue, as shown in Figure 4Go.



View larger version (136K):
[in this window]
[in a new window]
 
FIG. 3. Localization of metallothionein (MT) in an injection site sarcoma (top) and its pulmonary metastasis (bottom) from a WF rat. These lesions are from a rat that had been treated with a total Cd dose of 513 µmol/kg, sc, as detailed in Table 2Go. Peroxidase-antiperoxidase immunohistochemistry with hematoxylin counter staining. Brown staining represents the presence of MT. Magnification x164. Primary fibrosarcoma showing strong immunoreactivity for MT, particularly within the nucleus, in many tumor cells (top). MT immunoreactivity is negative in the corresponding pulmonary metastasis (bottom). This staining pattern was typical for sets of primary and metastatic injection site sarcomas from WF and F344 rats.

 


View larger version (158K):
[in this window]
[in a new window]
 
FIG. 4. Lower magnification view of a pulmonary metastasis of an injection site sarcoma depicting aggressive behavior. This lesion is from a WF rat that had been treated with a total Cd dose of 513 µmol/kg, sc, as detailed in Table 2Go. Peroxidase-antiperoxidase immunohistochemistry with hematoxylin counter staining. Magnification x82. Metastatic tumor (right) shown compressing and invading normal adjacent lung tissue (left). Metallothionein immunoreactivity is essentially negative in this lesion.

 
Cd exposure frequently induces testicular interstitial cell (Leydig) tumors in rats (IARC, 1993Go), but these are also common spontaneous tumors in many strains. The effects of repeated Cd exposures on testicular tumors are shown in Table 5Go. In this study, WF rats showed induction of testicular tumors in the group receiving a total of 180 µmol/kg (93% incidence rate of rats at risk) and 360 µmol/kg (66%) over the spontaneous control rates (37%). At higher accumulated doses (540 and 513 µmol/kg), testicular tumors decreased to control rates. In F344 rats, the spontaneous control rate was very high (97%) as is typical for this strain in 2-year studies. The 3 highest accumulated doses in F344 rats (360, 513, 540 µmol/kg) significantly suppressed interstitial-cell tumor incidence to levels well below control (40–67%).


View this table:
[in this window]
[in a new window]
 
TABLE 5 Incidence of Testicular Tumors in WF and F344 Rats Treated with Multiple Cadmium Chloride Injections
 
Other tumors were also suppressed in these rats. In WF rats, lipomatous renal tumors and pituitary adenomas were suppressed in many cases at the higher doses (360–540 µmol/kg) of Cd (Table 6Go). Likewise, in F344 rats, the occurrence of some tumors were suppressed at the 3 highest doses of Cd. This included suppression of leukemias, pituitary adenomas, and adrenal pheochromocytomas. The suppression of these fairly common, spontaneous tumors may well be due to a combination of reduced body weight, reduced survival and/or enhanced cytotoxicity.


View this table:
[in this window]
[in a new window]
 
TABLE 6 Incidence of Tumors in Either WF or F344 Treated Rats That Was Suppressed by Multiple Injections of Cadmium Chloride
 
There were several tumors in both strains that were unrelated to any treatment. In the 144 WF rats at risk, these included 5 prostatic adenomas, 3 papillomas (2 in forestomach, 1 in urinary bladder), 3 schwannomas (colon, eye, thoracic cavity), 2 pancreatic islet cell adenomas, and one each of the following: a mammary gland fibroadenoma, a pulmonary adenoma, a renal carcinoma, a renal adenoma, a basal cell carcinoma of the skin, a keratoacanthoma, a leiomyosarcoma (intestine), a fibrosarcoma of the spleen, a non-injection site subcutaneous fibrosarcoma, a non-injection site subcutaneous fibroma, and a thyroid carcinoma. In the 127 F344 rats at risk, tumors not associated with Cd treatment included: 9 pancreatic islet cell adenomas, 7 preputial gland tumors, 3 pancreatic acinar cell adenomas, 3 pulmonary adenomas, 2 bone osteosarcomas, 2 hemangiomas, 2 mammary gland fibroadenomas, 2 papillomas (skin), 2 prostatic adenomas, 2 squamous cell carcinomas (gingiva and nasal cavity), 2 testicular mesotheliomas, 2 thyroid c-cell adenomas, and one each of the following: a parathyroid adenoma, a renal carcinoma, a keratoacanthoma, a non-injection site subcutaneous fibroma, and a non-injection site subcutaneous lipoma.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In the present study, repeated exposures to Cd resulted in the more rapid onset and increased malignancy of ISS in rats in comparison to prior observations with ISS resulting from single exposures (IARC, 1993Go; Waalkes et al., 1991bGo). This was observed in both the WF and F344 rat strains, although this effect was more pronounced in the WF rat, and is likely related to both the repeated nature of the exposure and higher total dose of Cd. The aggressive nature of the tumors formed by repeated exposure to Cd is reflected in a higher rate of invasion into the subdermal muscle layers and bone as well as a greater tendency of the tumors to metastasize to distant tissues, particularly in the lungs of WF rat. This is the first observation that Cd can act in vivo as a "progressor" in the sense that the malignant progression of tumors formed by Cd is enhanced by repeated exposures to the metal. Prior studies have shown that Cd can enhance tumor progression or invasion in a variety of model systems, including in vitro systems, using model basement membranes (Haga et al., 1997Go), ex vivo systems using invasion of explanted tissues (Haga et al., 1996Go), and in systems where tumor cells are exposed to Cd in vitro and then inoculated into host mice (Abshire et al., 1996Go). However, this is the first completely in vivo demonstration that repeated exposures to Cd can have a dramatic impact on tumor progression. Although Cd is certainly an effective single-dose carcinogen in rodents (IARC, 1993Go), it is obvious that human populations, during an average lifetime, would experience repeated exposures to Cd. Additionally, it should be pointed out that, in general, tumor metastases are more often the cause of cancer deaths than the primary tumor.

Similar to the results in vivo in the present study, there are several model systems of tumor invasion or progression in which Cd has a striking effect (Haga et al., 1996Go, 1997Go; Abshire et al., 1996Go). In this regard, one study has shown that when myoblastic tumor cells are exposed to Cd in vitro, it can markedly enhance tumor growth of the cells upon inoculation into nude mice (Abshire et al., 1996Go). Furthermore, the tumors resulting from inoculation of cells exposed to Cd in vitro appeared histologically more malignant, were more invasive, and more readily precipitated host death as a reflection of aggressive behavior than the tumors resulting from inoculation of untreated cells (Abshire et al., 1996Go). Another report shows that, when human fibrosarcoma (HT-1080) cells are chronically exposed to Cd in vitro, the exposure promotes tumor cell invasion of reconstituted membranes, indicative of enhanced tumor invasiveness (Haga et al., 1997Go). The authors suspect that these characteristics induced by Cd treatment would promote malignancy and tumor metastasis in vivo (Haga et al., 1997Go), which is in accord with the present observations. In fact, it is of interest that the malignant potential of both rodent (present study) and human (Haga et al., 1997Go) fibrosarcoma cells appear to be enhanced by Cd exposure. Evidence also indicates that Cd, besides having a direct effect on tumors cells, can modify host tissue response resulting in increased tumor invasiveness (Haga et al., 1996Go). In this regard, B16 melanoma cells will invade into samples of explanted organs from Cd-treated mice much more readily than similar tissue samples from control mice (Haga et al., 1996Go). This includes explanted lung and liver (Haga et al., 1996Go), both of which were target tissues for metastasis of ISS in the present study. Additionally, in cell monolayer assays of invasiveness using either human fibroblast (WI-38) or bovine carotid artery endothelial (HH) cells as host monolayers, human HT-1080 fibrosarcoma cells invaded much more readily into host monolayers pretreated with Cd (Haga et al., 1996Go). So it appears that, although the precise mechanism cannot be defined in the present study, both tumor cell and host tissue factors may play a role in the enhanced tumor progression and aggressiveness seen with repeated Cd exposures. The long biological half-life of Cd indicates that relatively late, host tissue-mediated effects would be a reasonable possibility.

The strain-related differences seen in the present study with Cd-induced, enhanced tumor progression indicate that genetic factors may well play an important role in this effect of Cd. In fact, it appears that the strain that was most tolerant to the acute toxic effects of Cd (WF) was also the one that was the most sensitive to Cd-induced, enhanced tumor progression. The F344 rat is one of the more sensitive strains of rat with regard to the acute toxicity of Cd, as seen in this and other studies (Kuester et al., 1999Go; Waalkes et al., 1991bGo). This indicates that initial tolerance may actually predispose the animal to Cd induction of the metastatic and/or invasive phenotype. In this regard, it is well recognized that a low loading dose of Cd will make animals or cells tolerant to the effects of subsequent, normally toxic doses of Cd (see Goering et al., 1995 and Klaassen et al., 1999 for review). In the present study, the low loading dose (3 µmol/kg) effectively abolished the high acute lethality (57%) of the highest dose of Cd (30 µmol/kg) in F344 rats, clearly demonstrating this acquired-tolerance phenomenon. Interestingly, it was the groups that received this low loading dose of Cd, and were, therefore, made tolerant to acute effects of Cd, that eventually showed the most aggressive tumors in both strains. This also supports an association between acute tolerance and predisposition to development of aggressive tumors with repeated Cd exposure. This could be due to altered cell-population dynamics in that, with a resistant cell population, more cells, including cells more heavily damaged, would survive the initial Cd exposure. In this regard, the 2 highest weekly doses used in the present study (i.e., 20 and 30 µmol/kg) are in fact often effective in inducing ISS if given as a single dose, although they rarely metastasize (IARC, 1993Go). Enhanced survival of cells with more DNA damage would presumably allow tumor progression to occur more rapidly. Recent evidence indicates that at least under some circumstances, Cd can block apoptosis produced by certain DNA damaging agents (Shimada et al., 1998Go; Yuan et al., 1999Go). Thus, if the first several doses of Cd were effective in tumor initiation, subsequent doses, by blocking apoptosis, may allow survival of genetically damaged cells. In any event, it appears cellular survival of the initial Cd exposure, for whatever reasons, is a key in the development of these Cd-induced aggressive tumors.

Initial tolerance also appears to affect ISS incidence as well as latency, at least in Wistar rats. In Wistar rats given a single injection of 30 µmol/kg, the average latency (time to death with tumor) of ISS was over 20 weeks longer (87.0 ± 4.3 weeks) (Waalkes et al., 1991bGo, 1989Go) than latency of ISS in rats given repeated injections of the same dose in the present study (65.7 ± 5.3). Similarly, the incidence of ISS (40%) in Wistar rats given a single dose of 30 µmol/kg (Waalkes et al., 1991aGo, 1989Go) appeared higher in rats given the same dosage, only repeatedly, in the present study (67%). These comparisons may not be totally valid as they compare outbred Wistar [Crl:(WI)BR] rats used in the prior study (Waalkes et al., 1989Go) and inbred WF rats in the present study. However, both these Wistar substrains are certainly resistant to the acute toxic effects of Cd when compared to the F344 rat. Furthermore, with F344 rats a prior study showed that a single sc dose of 30 µmol/kg produced a 66% incidence and a 65.3 ± 2.9 week latency of ISS (Waalkes et al., 1991bGo), which is remarkably similar to the incidence (67%) and latency (67.0 ± 7.6 weeks) in the present study. Based on these results, it appears that the strain generally less sensitive to acute Cd toxicity (Wistar) is more sensitive to the enhancing effects of repeated Cd exposure on tumor progression. Thus, again it appears that initial tolerance allows the more rapid formation of more highly aggressive tumors with repeated Cd exposure. The observation that acute tolerance to Cd toxicity does not translate into tolerance to the carcinogenic effects of Cd, at least with regard to Cd-induced enhanced tumor progression and metastasis, is both counter-intuitive and surprising.

Tolerance to acute Cd toxicity can be induced by a variety of substances, including other metals, and is generally associated with induction of MT synthesis (Goering et al., 1995; Klaassen et al., 1999Go). However, tolerance to Cd does not, in all cases, require MT and can be induced in cells through means that involve reduced uptake of the toxic metal ion (Takiguchi and Waalkes, 1999Go). In the present study, the primary sarcomas showed high levels of MT. The staining for MT in the primary ISS was frequently most intense in the nucleus, a characteristic thought to be associated with rapid proliferation (Cherian, 1993Go). Therefore, the expression of MT in the ISS of the present study could be a reflection of their rapid proliferative growth. In fact, many tumors show intense staining for MT, but others are heterogenous or show minimal staining (Cherian, 1993Go ;Waalkes and Perez-Ollie, 1999Go). Given the long residence time of Cd at the injection site (Kasprzak and Poirier, 1985Go), the MT in the ISS could also represent a response to residual Cd, as Cd is one of the better MT-inducing agents (Klaassen et al., 1999Go). Alternatively, this heightened MT expression in ISS could be a genotypic response affording tolerance to the repeated Cd exposure, as seen in vitro or in vivo with acquired self tolerance to Cd (Klaassen et al., 1999Go). In any event, there is some evidence that MT may be important to invasion and metastasis (Haga et al., 1998Go). When HT-1080 fibrosarcoma cells are treated with propargylglycine (PPG) to deplete cellular MT, the ability of these cells to invade a reconstituted basement membrane is significantly reduced (Haga et al., 1998Go). The authors hypothesize that increased MT levels correlate with invasiveness and metastatic potential (Haga et al., 1998Go), which would be consistent with the metastatic potential and high MT levels in the primary ISS in both F344 and WF rats from the present study. PPG can have many effects beyond a simple depletion of MT, so this cannot be considered conclusive evidence of a role of MT in progression. In a study which used antisense MT and found suppression of leukemia P388 cell growth, the authors concluded that the growth of neoplastic cells depends on expression of the MT gene (Takeda et al., 1999Go). However, tumors and tumor cell lines show a great diversity of MT expression from minimal to high (Waalkes and Perez-Ollie, 1999GoWoo et al., 1997Go) and the role of MT in tumor cell pathobiology and growth is, at best, complex and poorly defined.

Whatever the basis for the high levels of MT in the primary ISS, our results show that the ISS cells, once having metastasized, are essentially devoid of MT in both the WF and F344 rats. This argues against a major role for MT in metastasis, although very early metastases were not observed and the stimulus for MT may be lost because of the new tissue environment of the metastatic tumor. Alternatively, the metastatic cells may originate from the least well-differentiated subpopulation in the primary tumor and hence the associated loss of MT expression. In this scenario, the cells most prone to metastasis may be the cells expressing the least MT out of the primary ISS cell population. In this regard, the work of Rossman et al. (1997) indicates that suppression of MT expression results in increased spontaneous mutations, indicating that MT may act as an antimutagen, and the loss of MT expression could create a less differentiated state. In fact, it is not infrequent that tumors develop distinctive subpopulations (Cherian, 1993Go). Our previous studies have shown that MT cannot be detected in the metastatic adenocarcinoma of human liver but the metastases have a greatly increased number of apoptotic bodies (Deng, et al., 1998Go). These results suggest that MT may also have an anti-apoptotic effect in human liver (Deng, et al., 1998Go). Additionally, in patients with metastatic colorectal cancers, the metastatic lesion contains very little MT (Mulder et al., 1992Go). Likewise in lymph node metastases of breast cancers in humans, there is typically minimal evidence of MT expression (Haerslev et al., 1994Go) although some breast carcinomas prominently express MT (Cherian, 1993Go; Douglas-Jones et al., 1997Go). Based on limited available data, it appears that MT is often expressed to a much lesser extent in metastatic tumors when compared to the primary tumor. Further study will be required to define the role of MT in tumor-cell progression and metastasis. Perhaps the use of MT-null mice in assays of tumor cell metastasis would be appropriate.

The observation that repeated exposures to Cd has an impact on tumor progression could have important implications in humans exposed to Cd. Industrial Cd exposures are undoubtedly multiple events over many years. Furthermore, it is estimated that cigarette smoking will double the total lifetime body burden of Cd, because of the Cd content in tobacco (Elinder, 1985Go). So, in this case there would be repeated exposures to Cd along with all the other organic carcinogens contained in tobacco smoke and a role of progressor for Cd could be quite important. In this regard, smoking cessation can reduce cancer deaths even after diagnosis of a primary lung cancer, although it is very premature to attribute any such effect to Cd.

Repeated Cd exposures, with the consequent high accumulated doses, and associated body-weight suppression and reduced survival, suppressed a variety of spontaneously occurring tumors in the present study. This included suppression of leukemias, pituitary adenomas, testicular interstitial-cell tumors, and adrenal pheochromocytomas. The suppression of these common spontaneous tumors is not unusual under such circumstances and is likely due to a combination of reduced body weight, reduced survival, and/or enhanced cytotoxicity. With the testicular tumors in WF rats, Cd at lower doses first increased tumor incidence; then, at higher doses, caused tumor suppression. Cd induction of testicular interstitial cell tumors has been widely reported after single systemic exposures (IARC, 1993Go; Waalkes et al., 1997Go). Cd causes acute testicular necrosis and subsequent atrophy and loss of testicular function, including androgen production (Waalkes et al., 1997Go). After this acute phase reaction, it is likely that the remnant interstitial cell population, which has largely lost its androgen production capacity, is chronically overstimulated by gonadotropin, resulting in a proliferative response and hence, tumor formation (Waalkes et al., 1997Go). The loss of tumor response in the testes, at the highest accumulated doses of Cd in the present study, may well be due to a more complete destruction of the interstitial cell population, allowing the survival of remnant cells in sub-minimal numbers to eventually form tumors.

In summary, the present results indicate that repeated exposures to the carcinogenic, inorganic Cd can result in the more rapid onset of more highly aggressive tumors. The mechanism of this effect is yet undefined, but this could have an important impact on hazards posed by multiple Cd exposures alone or in combination with exposure to other carcinogens. Finally, acute tolerance to Cd toxicity does not appear to translate into tolerance to the carcinogenic effects of Cd, at least with regard to Cd-induced enhanced tumor progression and metastasis.


    ACKNOWLEDGMENTS
 
The authors thank Charles Riggs, Shirley Hale, Deborah Devor, and Robert Bare for excellent technical assistance and Drs. Jerry Ward, Larry Keefer, Jie Liu, and Hua Chen for critical comments.


    NOTES
 
1 To whom correspondence should be addressed at Laboratory of Comparative Carcinogenesis, NCI at NIEHS, 111 Alexander Drive, PO Box 12233, MD F0-09, Research Triangle Park, NC 27709. Back

2 Present address: SmithKline Beecham, 709 Swedeland Road, P.O. Box 1539, King of Prussia, PA 19406. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Abshire, M. K., Devor, D. E., Diwan, B. A., Shaughnessy, J. D., Jr., and Waalkes, M. P. (1996). In vitro exposure to cadmium in rat L6 myoblasts can result in both enhancement and suppression of malignant progression in vivo. Carcinogenesis 17, 1349–1356.[Abstract]

Banerjee, D., Onosaka, S., and Cherian, M. G. (1982). Immunohistochemical localization of metallothionein in cell nucleus and cytoplasm of rat liver and kidney. Toxicology 24, 95–105.[ISI][Medline]

Cherian, M. G. (1993). Nuclear and cytoplasmic localization of metallothionein in human liver during development and in tumor cells. In Metallothionein III, (K. T. Suzuki, N. Imura, and M. Kimura, Eds.), pp. 175–187. Birkhäuser Verlag, Basel.

Deng, D. X., Chakrabarti, S., Waalkes, M. P., and Cherian, M. G. (1998). Metallothionein and apoptosis in primary human hepatocellular carcinoma and metastatic adenocarcinoma. Histopathology 32, 340–347.[ISI][Medline]

Douglas-Jones, A. G., Navabi, H., Morgan, J. M., and Jasani, B. (1997). Immunoreactive p53 and metallothionein expression in duct carcinoma in situ of the breast: No correlation. Virchows Archives 430, 373–379.[ISI][Medline]

Elinder, C.-G. (1985). Normal values for cadmium in human tissue, blood, and urine in different countries. In Cadmium and Health: A Toxicological and Epidemiological Appraisal (L. Friberg, C.-G. Elinder, T. Kjellström, and G. F. Nordberg, Eds.), Vol. 1, pp. 81–102. CRC Press, Boca Raton, FL.

Goering, P. L., Waalkes, M. P., and Klaassen, C. D. (1994). Toxicology of cadmium. In Handbook of Experimental Pharmacology; Toxicology of Metals, Biochemical Effects, (R. A. Goyer, and M. G. Cherian, Eds.), Vol. 115, pp. 189–214. Springer-Verlag, New York.

Gunn, S. A., Gould, T. C., and Anderson, W. A. D. (1963). Cadmium-induced interstitial cell tumors in rats and mice and their prevention by zinc. J. Natl. Cancer Inst. 31, 745–753.[ISI][Medline]

Gunn, S. A., Gould, T. C., and Anderson, W. A. D. (1964). Effect of zinc on cancerogenesis by cadmium. Proc. Soc. Exp. Biol. Med. 115, 653–657.

Gunn, S. A., Gould, T. C., and Anderson, W. A. D. (1967). Specific response of mesenchymal tissue to cancerogenesis by cadmium. Arch. Pathol. 83, 493–499.[ISI][Medline]

Haerslev, T., Jacobsen, K., Nedergaard, L., and Zedeler, K. (1994). Immunohistochemical detection of metallothionein in primary breast carcinomas and their auxiliary lymph node metastises. Pathology Res. Practice 190, 675–681.[ISI][Medline]

Haga, A., Nagase, H., Kito, H., and Sato, T. (1996). Enhanced invasiveness of tumour cells after host exposure to heavy metals. Europ. J. Cancer 32A, 2342–2347.

Haga, A., Nagase, H., Kito, H., and Sato, T. (1997). Invasive properties of cadmium-resistant human fibrosarcoma HT-1080 cells. Cancer Biochem. Biophys. 15, 275–284.[ISI][Medline]

Haga, A., Nagase, H., Kito, H., and Sato, T. (1998). Inhibition effect of propargylglycine on human HT-1080 cell invasivevess. Cancer Biochem. Biophys. 16, 31–43.[Medline]

Heath, J. C., and Webb, M. (1967). Content and intracellular distribution of the inducing metal in the primary rhabdomyosarcomata induced in the rat by cobalt, nickel and cadmium. Br. J. Cancer 21, 768–779.[ISI][Medline]

Heath, J. C., Daniel, I. R., Dingle, J. T., and Webb, M. (1962). Cadmium as a carcinogen. Nature 193, 592–593.[ISI][Medline]

IARC (1993). International Agency for Research on Cancer Monographs, Vol 58. Beryllium, cadmium, mercury, and exposures in the glass manufacturing industry, pp. 119–238. IARC Press, Lyon.

Kasprzak, K. S., and Poirier, L. A. (1985). Effects of calcium and magnesium actetates on tissue distribution of carcinogenic doses of cadmium chloride in Wistar rats. Toxicology 34, 221–230.[ISI][Medline]

Klaassen, C. D., Liu, J., and Choudhuri, S. (1999). Metallothionein: An intracellular protein to protect against cadmium toxicity. Annu. Rev. Pharmacol. Toxicol. 39, 267–294.[ISI][Medline]

Kuester, R. K., Waalkes, M., Goering, P. L., Li, G., Fisher, B., Younis, H., and Sipes, I. G. (1999). Differential induction of metallothionein and heat shock protein 72 by cadmium in the livers of the male Fischer-344 and the Sprague-Dawley rat. Toxicologist 48, 355.

Mulder, T. P., Janssens, A. R., Verspaget, H. W., van Hattum, J., and Lamers, C. B. (1992). Metallothionein concentration in the liver of patients with Wilson's disease, primary biliary cirrhosis, and liver metastasis of colorectal cancer. J. Hepatol 16, 346–350.[ISI][Medline]

Nartey, N., Cherian, M. G., and Banerjee, D. (1987). Immunohistochemical localization of metallothionein in human thyroid tumors. Am. J. Pathol 129, 177–182.[Abstract]

Poirier, L. A., Kasprzak, K. S., Hoover, K., and Wenk, M. L. (1983). Effects of calcium and magnesium acetates on the carcinogenicity of cadmium chloride in Wistar rats. Cancer Res. 43, 4575–4581.[Abstract]

Roe, F. J. C., Dukes, C. E., Cameron, K. I., Pugh, R. C. B., and Mitchley, B. C. V. (1964). Cadmium neoplasia: Testicular atrophy and Leydig cell hyperplasia and neoplasia in rats and mice following subcutaneous injection of cadmium salts. Br. J. Cancer 18, 674–681.[ISI][Medline]

Rossman, T. G., Goncharova, E. I., Nadas, A., and Dolzhanskaya, N. (1997). Chinese hamster cells expressing antisense to metallothionein become spontaneous mutators. Mutat. Res. 373, 75–85.[ISI][Medline]

Shimada, H., Shiao, Y.-H., Shibata, M.-A., and Waalkes, M. P. (1998). Cadmium suppresses apoptosis induced by chromium. J. Toxicol. Environ. Health, 54, 159–168.[ISI]

Takeda, A., Hisada, H., Okada, S., Mata, J. E., Iversen, P. L., and Ebadi, M. (1999). The suppression of metallothionein inhibits the growth of leukemia P388 cells. In Metallothionein IV, (C. D. Klaassen, Ed.), pp. 613–617. Birkhäuser Verlag, Basel.

Takiguchi, M., and Waalkes, M. P. (1999). Cyproterone acetate, the antiandrogen, induces a novel form of cellular tolerance to cadmium involving reduced cellular tolerance to cadmium involving reduced cellular accumulation. Toxicologist 48, 354.

Waalkes, M. P., Anver, M., and Diwan, B. (1999). Carcinogenic effects of cadmium in the Noble (NBL/Cr) rat: Induction of pituitary, testicular, and injection site tumors and intraepithelial proliferative lesions of the dorsolateral prostate. Toxicol. Sci. 52, 154–161.[Abstract]

Waalkes, M. P., Kovatch, R., and Rehm, S. (1991a). Effect of chronic dietary zinc deficiency on cadmium toxicity and carcinogenesis in the male Wistar [Hsd:(WI)BR] rat. Toxicol. Appl. Pharmacol. 108, 448–456.[ISI][Medline]

Waalkes, M. P., and Misra, R. R. (1996). Cadmium carcinogenicity and genotoxicity. In Toxicology of Metals (L. W. Chang, Ed.), pp. 231–244. CRC Press, Boca Raton, FL.

Waalkes, M. P., and Perez-Olle, R. (1999). Role of metallothionein in metabolism, transport, and toxicity of metals. Molecular Biology and Toxicology of Metals, (D. J. Koropatnick and R. Zalups, Eds.), Taylor & Francis, London (in press).

Waalkes, M. P., and Rehm, S. (1994). Chronic toxic and carcinogenic effects of cadmium chloride in male DBA/2NCr and NFS/NCr mice: Strain-dependent association with tumors of the hematopoietic system, injection site, liver, and lung. Fundam. Appl. Toxicol. 23, 21–31.[ISI][Medline]

Waalkes, M. P., and Rehm, S. (1998). Lack of carcinogenicity of cadmium chloride in female Syrian hamsters. Toxicology 126, 173–178.[ISI][Medline]

Waalkes, M. P., Rehm, S., and Devor, D. E. (1997). The effects of continuous testosterone exposure on spontaneous and cadmium-induced tumors in the male Fischer (F344/NCr) rat: Loss of testicular response. Toxicol. Appl. Pharmacol. 142, 40–46.[ISI][Medline]

Waalkes, M. P., Rehm, S., Riggs, C. W., Bare, R. M., Devor, D. E., Poirier, L. A., Wenk, M. L., and Henneman, J. R. (1989). Cadmium carcinogenesis in the male Wistar [Crl:(WI)BR] rat: Dose-response analysis of effects of zinc on tumor induction in the prostate and in the testes and at the injection site. Cancer Res. 49, 4282–4288.[Abstract]

Waalkes, M. P., Rehm, S., Riggs, C. W., Bare, R. M., Devor, D. E., Poirier, L. A., Wenk, M. L., Henneman, J. R., and Balaschak, M. S. (1988). Cadmium carcinogenesis in male Wistar [Crl:(WI)BR] rats: Dose-response analysis of tumor induction in the prostate and testes and at the injection site. Cancer Res. 48, 4656–4663.[Abstract]

Waalkes, M. P., Rehm, S., Sass, B., Konishi, N., and Ward, J. M. (1991b). Chronic carcinogenic and toxic effects of a single subcutaneous dose of cadmium in the male Fischer rat. Environ. Res. 55, 40–50.[ISI][Medline]

Woo, E. S., Monks, A., Watkins, S. C., Wang, A. S., and Lazo, J. S. (1997). Diversity of metallothionein content and subcellular localization in the National Cancer Institute tumor panel. Cancer Chemother. Pharmacol. 41, 61–68.[ISI][Medline]

Yuan, C., Kadiiska, M. B., Mason, R., and Waalkes, M. P. (1999). Cadmium blocks apoptosis induced by chromium, hydromycin B, and actinomycin D: The role of caspase-3 inhibition. Toxicologist 48, 355.