Radionuclides in the lichen-caribou-human food chain near uranium mining operations in northern Saskatchewan, Canada.

The richest uranium ore bodies ever discovered (Cigar Lake and McArthur River) are presently under development in northeastern Saskatchewan. This subarctic region is also home to several operating uranium mines and aboriginal communities, partly dependent upon caribou for subsistence. Because of concerns over mining impacts and the efficient transfer of airborne radionuclides through the lichen-caribou-human food chain, radionuclides were analyzed in tissues from 18 barren-ground caribou (Rangifer tarandus groenlandicus). Radionuclides included uranium (U), radium (226Ra), lead (210Pb), and polonium (210Po) from the uranium decay series; the fission product (137Cs) from fallout; and naturally occurring potassium (40K). Natural background radiation doses average 2-4 mSv/year from cosmic rays, external gamma rays, radon inhalation, and ingestion of food items. The ingestion of 210Po and 137Cs when caribou are consumed adds to these background doses. The dose increment was 0.85 mSv/year for adults who consumed 100 g of caribou meat per day and up to 1.7 mSv/year if one liver and 10 kidneys per year were also consumed. We discuss the cancer risk from these doses. Concentration ratios (CRs), relating caribou tissues to lichens or rumen (stomach) contents, were calculated to estimate food chain transfer. The CRs for caribou muscle ranged from 1 to 16% for U, 6 to 25% for 226Ra, 1 to 2% for 210Pb, 6 to 26% for 210Po, 260 to 370% for 137Cs, and 76 to 130% for 40K, with 137Cs biomagnifying by a factor of 3-4. These CRs are useful in predicting caribou meat concentrations from the lichens, measured in monitoring programs, for the future evaluation of uranium mining impacts on this critical food chain.

Summary.-The immunogenicity of a soluble fraction containing Gross-virusassociated cell-surface antigen (GCSAa) obtained from (C58NT)D lymphoma cells either by detergent (NP40) solubilization or by 3M KCI extraction, was studied in syngeneic W/Fu rats. Rats immunized by 2 s.c. injections of soluble antigen or soluble antigen mixed with empty liposomes and emulsified in complete Freund's adjuvant (CFA) failed to produce significant levels of cytotoxic antibodies to GCSAa. On the other hand, rats similarly immunized by negatively charged liposomes containing NP40-solubilized GCSAa, and emulsified in CFA, developed high and persistent levels of cytotoxic antibodies, and their response could even mimic that induced by viable (C58NT)D cells. A similar response could also be obtained in rats immunized with liposome-associated NP40-solubilized GCSAa, but without CFA. Rats immunized by comparable amounts of liposome-associated 3M KC1 -extracted GCSAa developed only low levels of cytotoxic antibodies, and their response was of shorter duration. These results strongly suggest that inclusion into liposomes of a solubilized proteic tumour-associated cell-surface antigen can provide an immunogen as potent as viable tumour cells in inducing an antibody response, and that the solubilization method may be critical.
WE HAVE DESCRIBED (Sakal et al., 1980) immunochemical characters of the association with liposomes of Gross cellsurface antigen (GCSAa), a major cell-surface antigen, of proteic nature (Ledbetter & Nowinski, 1977;Snyder et al., 1977) associated with Gross virus-induced lymphomas in the mouse (Old et al., 1965) and rat (Geering et al., 1966;Herberman, 1972) which appears to play an important role in hosttumour relationship and can induce high antibody response in syngeneic rats (Gerlier et al., 1977a;Herberman & Oren, 1971). The present work describes the antibody response elicited in syngeneic W/Fu rats by immunization with liposome-associated partly purified GCSAa, and the results suggest that liposomal presentation of this antigen can induce cytotoxic antibodies to GCSAa, reaching in some instances the level obtained after immunization with viable syngeneic tumour cells.

MATERIAL AND METHODS
Animals and tumour.-W/Fu/Rholco rats and C57BL/6/Rholco mice were bred in our colony. Five-weeks-old male W/Fu rats were used for immunization. Gross-virus-induced (C58NT) D lymphoma (Geering et al., 1966) was maintained in ascitic form by weekly passage into syngeneic weanling W/Fu rats.
Gross-virus-induced EdG2 lymphoma (Old et al., 1965) was also weekly transplanted into syngeneic C57BL/mice. Antigen preparation.-Gross cell-surface antigen (GCSAa) was extracted either by Nonidet P40 (NP40) or 3M KCI from (C58NT)D lymphoma cells and partially purified after 60% ammonium sulphate precipitation and Sephadex G200 filtration. Details are given in Sakal et al. (1980). Liposome preparation.-Negatively charged liposomes were prepared as described by Gregoriadis et al. (1971). Details of liposome sensitization with GCSAa have been reported elsewhere (Sakal et al., 1980). Briefly, a film of dipalmitoylphosphatidylcholine, cholesterol and dicetylphosphate in 7:2:1 molar ratio, was dispersed in antigenic extract obtained either by NP40 or by 3M KCI solubilization. Liposomes used in these experiments had a protein/phospholipid ratio of 0.15-0-20, most GCSAa activity being firmly associated with lipids (Sakal et al., 1980) and were injected immediately without previous storage. As control, empty liposomes were similarly prepared by dispersion of lipids in the buffer.
Immunizations.-Groups of W/Fu rats were immunized by 2 s.c. injections given 5 weeks apart with GCSAa preparations. In one set of experiments NP40-solubilized GCSAa was used as immunogen, presented either as soluble antigen, soluble antigen mixed with empty liposomes, or GCSAasensitized liposomes, and injected with or without complete Freund adjuvant (CFA). In a second set of experiments, the kinetics of the antibody response was studied using groups of 4 rats receiving either soluble or liposomeassociated NP40-solubilized GCSAa with or without CFA, and, in a third set of experiments, the kinetics of the antibody response to 3M KCl-solubilized GCSAa was studied under similar conditions. Doses of injected antigen are detailed in the Table. As control, a group of W/Fu rats was immunized by a single s.c. injection of 2 x 108 syngeneic (C58NT)D viable lymphoma cells, since calculation based on the specific activity of antigenic extract indicated that rats in the other groups were immunized with a quantity of GCSAa grossly amounting to the cellsurface expression by 2 x 108 (C58NT)D cells (Gerlier et al., 1977b). In all groups, a blood sample was weekly collected from the animals' tails.
Antibody production assay.-Sera from animals under immunization were tested for antibody to GCSAa, using a complementdependent cytotoxicity test as previously described (Gerlier et al., 1977a). Briefly, 50 ,ul of E,G2 cell suspension (4 x 106 cells/ ml) was incubated for 45 min at 3700 with 50 ,ul of serial dilutions of serum and 50 ,ul of an appropriate dilution of rabbit complement selected for absence of natural antimouse activity. Percentage of dead cells was determined by trypan-blue dye uptake. Results are expressed as cytotoxic index (CI) calculated as follows:

CI = % dead cells in test-%/ dead cells in control
100-% dead cells in control and the endpoint titre was expressed as the last serum dilution giving a C > 0-5. Controls of the specificity of GCSAa detection in this cytotoxicity test were performed by absorbing sera on mouse normal lymphoid cells, ESG2 lymphoma cells, or GCSAalymphoma cells as previously described (Gerlier et al., 1977b).

RESULTS
Antibody response to NP40-solubilized GCSAa Injection of liposome containing GCSAa (0.44 mg and 0-25 mg protein) emulsified with CFA induced an antibody response 3 weeks after the booster injection in 6/10 rats, the antibody response being 1:64 or more in 4 of these (Fig. 1). The high cytotoxic-antibody titres in these 4 rats (1:64, 1:128, 1:256, 1:512) were comparable to that in rats immunized with viable cells, as previously described, although immunization with tumour cells usually elicits an antibody response in all animals (Gerlier et at., 1977a). When rats were immunized with the same GCSAasensitized liposomes, but without CFA, or with a higher amount of NP40-extracted soluble GCSAa (0.69 mg and 0 52 mg protein) emulsified in CFA, or with soluble GCSAa (0-61 mg and 0 54 mg protein) mixed with empty liposomes and emulsified in CFA (Fig. 1) all animals failed to develop a significant antibody response. Primary antibody response was also determined in every group of animals3-4 weeks after the first injection and was always of a low level in this set of experiments.
Antibody response to NP40-solubilized GCSAa was further studied with the same immunization schedule, to determine the kinetics of this response, in comparison with that of rats receiving viable tumour cells. As observed in the preceding experiment, injection of liposomes containing GCSAa (0.43 mg and 0 37 mg protein) emulsified with CFA induced a good antibody response in 3/4 rats at the 8th week (Fig. 2b) which reached in one rat the same intensity as that produced by immunization with viable tumour cells. Low-level antibody responses were obtained in rats immunized with soluble antigen (0.9 mg and 0-56 mg protein) emulsified with CFA ( Fig. 2a) or with some liposomes containing GCSAa but without CFA (Fig. 2c) as previously observed, with the exception of one rat immunized with liposome containing GCSAa without CFA (Fig. 2c)  injection of viable tumour cells (Gerlier et al., 1977a). Moreover, the secondary peak of these antibody responses was somewhat higher than the primary one and persisted at a high level up to 13 weeks after the booster injection, similarly to the viable tumour-cell-elicited antibody response.
Antibody response to 3M KCI-solubilized GCSAa Similar immunization experiments were performed with liposomes sensitized with   (Fig. 3b, c).

DISCUSSION
In order to determine whether viable tumour cells could be substituted by soluble cell-surface antigen linked to artificial membrane, in inducing an antitumour response, we have previously included GCSAa, a tumour-associated virus-directed cell-surface antigen, into negatively charged liposomes (Sakai et al., 1980). The purpose of the present work was to compare in syngeneic animals the immunogenicity of soluble GCSAa extracted from W/Fu (C58NT)D lymphoma by two currently used methods to that of the liposome-associated antigen and to that of viable lymphoma cells.
While immunizations with viable lymphoma cells usually lead to a high and persistent antibody response (Gerlier et al., 1977a;Herberman & Oren, 1971) immunizations with similar amounts of solubilized GCSAa emulsified in CFA induced only a weak antibody response (out of 14 rats, 13 had an antibody titre (AT) <1:8, 1 had AT=1:16). On the other hand, immunization with liposomes containing solubilized GCSAa and emulsified in CFA induces a significant antibody response, which in some instances may be as high and persistent as that induced by immunization with live tumour cells (AT) 1:64 in 7/14 rats) and GCSAa must be presented as part of liposome structure to obtain this good antibody response, since mixing soluble GCSAa with empty liposomes elicits no antibody response (AT,< 1:4 in 6/6 rats). However, the antibody response of animals receiving liposome-associated GCSAa is less homogeneous than that of the animals injected with live tumour cells. This could be due to a non-optimal immunization schedule, since it has been previously demonstrated that the achievement of a high and homogeneous antibody response to (C58NT)D lymphoma cells depends upon the immunization schedule (G(erlier et al., 1977a).
The antibody response of rats immunized with liposomes sensitized with 3M KCIsolubilized GCSAa was much lower in magnitude and shorter in duration than that of rats similarly immunized with liposomes sensitized with NP40-solubilized GCSAa. This could be attributed neither to a difference in antigen dose, since the in vitro specific GCSAa activities of both types of cellular extract used in these experiments were comparable, nor to a difference in GCSAa association with liposomes, since it has been shown in a previous work that the liposome composition and the distribution of GCSAa among liposomal structure are almost identical whatever the sensitizing cellular extract used (Sakai et al., 1980). Nevertheless, it may be questioned whether the 2 different antigen-solubilization procedures lead to GCSAa-bearing molecules of identical immunogenicity, since 3M KCI extraction may induce proteolytic cleavage (Mann, 1972) and since detergent solubilization produces micellar association of the solubilized molecules (Helenius & Simons, 1975).
It can be questioned whether emulsifying sensitized liposomes in CFA is a prerequisite for the induction of a high and persistent antibody response to GCSAa, since it has been reported (Nicolotti et al., 1976) that antibody to liposome-associated synthetic antigen can be raised only in the presence of CFA. Microscopic examination of the sensitized liposomes used in the experiments reported here showed that, as previously observed (Kinsky & Nicolotti, 1977) they remained intact when emulsified in CFA. From the present results it appears that the use of CFA is not an absolute prerequisite, since the antibody response induced by liposome-associated GCSAa without CFA could in some cases (in 1/14 rats, AT > 1:64) parallel the results using CFA. CFA emulsification greatly increases the number of responding animals (7/14 rats, AT,> 1]:64).
Thus it appears that liposome association of GCSAa may produce an adjuvant effect, which accords with previously reported effects of liposome presentation of various antigens (Allison & Gregoriadis, 1974;Heath et al., 1976) provided a phospholipid of high transition temperature is used to form the liposome (Dancey et al., 1978;Yasuda et al., 1977) and this is actually the case with dipalmitoylphosphatidycholine used in these experiments (transition temperature: 41.5°C). It cannot be excluded that the adjuvant effect exerted by liposome association of the antigen may be due to a membrane presentation effect since it has been shown that solubilized membrane antigen can stimulate lymphocytes in vitro when exposed on liposomes (Curman et al., 1978;Engelhard et al., 1978). However it is worth noting that the sensitized liposome used here exposed only a small proportion of the associated GCSAa at their surface (Sakai et al., 1980).
In some of the responding animals the antibody response persisted at a high level for up to 18 weeks. This may be due to a depot effect of the antigen associated with liposomes made of high-transitiontemperature phospholipid, and which are likely to be of poor fluidity at body temperature. Further studies are necessary to gain further insight into the mechanisms involved in the adjuvant effect exerted by liposomes in inducing cytotoxic antibodies against cell-surface antigens. It is likely that, for instance, efficient immunization might require the presentation of tumour cell-surface antigen in association with the major histocompatibility complex (MHC) antigens on the membrane. Either of 2 mechanisms could fulfil this requirement: (1) liposome might be sensitized with MHC antigens containing GCSAa; (2) in the absence of MHC antigens in a GCSAa preparation, this association might be obtained as a result of an in vivo fusion between liposomes and host cells. So, it would be of the utmost interest to study the interaction of host macrophages (Yasuda et al., 1977) and lymphocytes (Blumenthal et al., 1977;Ozato et al., 1978) with liposome-associated solubilized cell-surface antigen.
Results from the present studies strongly suggest that, as far as antibody production to cell-surface tumour-associated antigen is concerned, liposome-associated solubilized membrane proteins can substitute viable tumour cells as immunogen, and that the solubilization method used is critical.