Micro-spectroscopic investigation of selenium-bearing minerals from the Western US Phosphate Resource Area

Mining activities in the US Western Phosphate Resource Area (WPRA) have released Se into the environment. Selenium has several different oxidation states and species, each having varying degrees of solubility, reactivity, and bioavailability. In this study we are investigating the speciation of Se in mine-waste rocks. Selenium speciation was determined using bulk and micro-x-ray absorption spectroscopy (XAS), as well as micro-x-ray fluorescence mapping. Rocks used for bulk-XAS were ground into fine powders. Shale used for micro-XAS was broken along depositional planes to expose unweathered surfaces. The near edge region of the XAS spectra (XANES) for the bulk rock samples revealed multiple oxidation states, with peaks indicative of Se(-II), Se(IV), and Se(+VI) species. Micro-XANES analysis of the shale indicated that three unique Se-bearing species were present. Using the XANES data together with ab initio fitting of the extended x-ray absorption fine structure region of the micro-XAS data (micro-EXAFS) the three Se-bearing species were identified as dzharkenite, a di-selenide carbon compound, and Se-substituted pyrite. Results from this research will allow for a better understanding of the biogeochemical cycling of Se in the WPRA.

Summary Proliferative activities in 181 primary Borrmann type 4 gastric carcinomas were investigated using percentage labelling of proliferating cell nuclear antigen (PCNA) and an argyrophilic nucleolar organiser region (AgNOR) count. Tumours with a high proliferative activity often metastasised to lymph nodes (P<0.01), and these patients had a lower survival rate (P<0.05). A significant correlation was recognised between the PCNA labelling percentage and AgNOR count (r = 0.452, P< 0.001). Cox's regression analysis showed that PCNA labelling percentage is an independent prognostic factor. These results indicate that estimating proliferative activity may be useful in predicting lymph node metastasis and patients' prognosis in cases of Borrmann type 4 gastric carcinoma.
Advanced carcinoma of the stomach can be classified based on Borrmann's criteria into one of four types (1-4) (Borrmann, 1926). A knowledge of this classification is important for endoscopists, radiologists and surgeons (Borchard, 1990). Of the Borrmann types of carcinomas, type 4 is a diffuse malignant lesion with indistinct borders, and is usually identified only at a very advanced stage (Maehara et al., 1992a). The lack of sharp borders can lead to underestimation of the size. As these cancers grow in the plane of the submucosa beneath an otherwise normal mucosa, establishing the histological diagnosis is difficult (Borchad, 1990). The clinical course is usually unfavourable and the 5 year survival rates are only 0-20% (Furukawa et al., 1988;Maehara et al., 1992a). Though many investigators have studied this entity from various aspects, the biological characteristics of Borrmann type 4 gastric carcinoma remain an open question.
As it is now feasible to measure proliferative activities of cells in formalin-fixed paraffin-embedded sections of surgical samples, two parameters of proliferative activity, proliferating cell nuclear antigen (PCNA) and argyrophilic nucleolar organiser regions (AgNOR), were measured. PCNA, a 36 kDa non-histone nuclear polypeptide, is an auxiliary protein of DNA polymerase delta , and plays a critical role in the initiation of cell proliferation (Jaskulski et al., 1988). The levels of PCNA increase in the nucleus during the late GI phase immediately prior to the onset of DNA synthesis, become maximal during the S phase, decline again during G2 and are low in M phase and quiescent cells (Kurki et al., 1987). Though PCNA staining has limitations in that the molecule has a long half-life which can also lead to staining of cells which have exited from the cycle , it is a useful marker for proliferating cells (McCormick & Hall, 1992). Nucleolar organiser regions (NORs) are loops of DNA (rDNA) encoded for ribosomal RNA (rRNA) production (Watson et al., 1987). The proteins associated with the NORs, the so-called AgNOR proteins, are argyrophilic, acidic and non-histone (Fakan & Hernandez-Verdan, 1986) and may serve as a marker for rDNA transcription activity or of rDNA transcriptional potential (Dimova et al., 1982;Busch, 1984;Walker, 1988). Thus, AgNOR staining can also serve as a parameter of proliferation (Egan & Crocker, 1992).
We examined the relationship between these two parameters and clinicopathological factors of gastric carcinomas. The objective of this study was to clarify the proliferative Correspondence: Y. Kakeji, Department of Gastroenterologic Surgery, National Kyushu Cancer Center, Notame 3-1-1, Minamiku, Fukuoka 815, Japan. Received 10 August 1993; and in revised form 12 November 1993. activity of Borrmann type 4 gastric carcinoma, with regard to clinical prognosis.

Patients
The 181 Japanese patients with primary Borrmann type 4 gastric cancer studied herein had undergone gastrectomy in the National Kyushu Cancer Center, Fukuoka, Japan, from 1972 to 1990. Partial gastrectomy was done in 43 patients and total gastrectomy with lymph node dissection in 138. A thorough histological examination was made on haematoxylin and eosin-stained preparations, and the histological classification was according to the tumour-node-metastasis classification system of the International Union Against Cancer (UICC, 1987). Macroscopic subtype, giant fold type, stenotic type and eroded type were classified according to Iwanaga et al. (1983). Adjuvant chemotherapy was given to 171 patients.
Immunohistochemical study for PCNA Sections from paraffin blocks were dewaxed and stained using the avidin-biotin-peroxidase complex method. The primary antibody, PC10, a monoclonal mouse antibody for human PCNA, was purchased from Dako (Carpinteria, CA, USA). The sections were incubated for 2 h with PC1O (dilution 1:20) at room temperature, with biotinylated goat antimouse IgG (1:200 for 30 min; Vector Laboratories), and with the avidin-biotin-peroxidase complex (for 30min; Vector Laboratories). Peroxidase labelling was developed with 3,3'diaminobenzidine and hydrogen peroxide, and the sections were counterstained with Mayer's haematoxylin.
To ensure consistency of PCNA staining between batches, a known positive control gastric carcinoma was included in each round. Negative controls were included by performing duplicate assays, in one of which the primary antibody was replaced by phosphate-buffered saline.
All of the nuclei stained were regarded as positive for PCNA ( Figure la). The percentage PCNA labelling was determined by observing 1,000 nuclei in areas of the section with the highest labelling, and the percentage of PCNAlabelfed nuclei was used for analysis. The principal method for determination of heterogeneity was as follows: (1) the entire area of each section was observed with low-power magnification (x 20) to determine the area where the cells positive for PCNA had gathered most densely, and (2) the counting of PCNA-positive cells was done in this area, under conditions of high-power magnification (x 400).

AgNOR staining
From the complete group of 181 patients, 174 tissues were also examined using AgNOR staining. The one-step silver colloid method was used. The NOR staining solution was prepared according to the description of Ploton et al. (1982). A mixture of one volume of 2% gelatin in 1% formic acid and two volumes of a 50% silver nitrate solution was poured over the sections and the preparations were left for 1 h at room temperature in the dark. On the AgNOR-stained slides, careful focusing made visible the AgNORs in the nucleus, in the form of black dots (Figure lb). At a magnification of x 1,000 (oil immersion) all dots, both satellite and those within clusters, were counted. One hundred cells from each lesion were analysed and a mean score of AgNOR count was recorded.

Statistical analysis
Clinicopathological data were stored in an IBM 4381 mainframe computer. The Biomedical Computer Program (BMDP) was used for all statistical analyses (Dixon, 1988). The BMDP P4F and P3S programs were used for the chi-square test and the Mann-Whitney test to compare characteristics between high and low groups with individual proliferative activities. Linear regression analyses were used to determine the correlation between the percentage PCNA labelling and the AgNOR count. Quantitative data on PCNA and AgNOR were compared using Student's t-test. The BMDP PIL program was used to analyse survival by the Kaplan-Meier method, and to compare survival curves, by the method of Mantel and Cox. The BMDP P2L program was used for a multivariate adjustment of all covariates, simultaneously, using the Cox regression analysis (Cox, 1972).

Results
Proliferative activity and clinicopathological characteristics PCIO immunostaining was almost entirely confined to the nucleus, and was diffuse, granular or a mixture of both. The distribution of PC10-positive cells was not homogeneous in many cases, and varied in different areas of even the same tumour. PC10-positive cells were frequently present in the advancing margin of the tumour, therefore counting was done in this area. The PCNA labelling index varied from 9.8% to 85.4%. The mean was 36.5%. The cases were divided into two groups: a high labelling group (> 36.5) and a low labelling group (<36.5). Table I summarises the clinicopathological characteristics of the high and low PCNA labelling groups. Tumours with a high PCNA percentage of labelling were associated with a higher incidence of lymphatic permeation, venous invasion and metastasis to lymph nodes than were those with low PCNA labelling (P<0.01). The percentage PCNA labelling was not related to the sex, age, tumour size, macroscopic subtype, depth of invasion, histological type, peritoneal dissemination, liver metastasis or operative curability.
As for AgNOR staining, the result was much the same as PCNA staining. AgNOR counts varied from 1.89 to 5.88, and the mean was 3.58. Tumours with high proliferative Eactivity (>3.58) were more likely to invade lymphatics, veins and lymph nodes than were those with low proliferative activity (<3.58). Figure 2 shows the results of linear regression analysis of percentage PCNA labelling and AgNOR count in primary gastric tumours. There was a significant correlation between the percentage PCNA labelling and the AgNOR count (r = 0.452, P<0.001).

Proliferating activity and prognosis
Survival curves for patients with carcinomas in the low and high PCNA labelling groups are shown in Figure 3. Surgical mortality was excluded in the analysis of survival. In patients with tumours with a high percentage of PCNA labelling survival rates were less favourable than in those with tumours with low labelling (P<0.001).
Of the 181 patients, 28 who underwent curative operation died within 18 months, and 22 patients lived for over 3 years. Table II shows the mean proliferative activities of these two groups. Tumours in patients who died within 18 months had a significantly higher percentage of PCNA labelling and higher AgNOR count than did those from patients who lived for over 3 years (P<0.05).
To search for an independent prognostic factor of Borrmann type 4 carcinoma, we carried out a multivariate Cox regression analysis. Factors examined included the sex, age, tumour size, macroscopic subtype, peritoneal dissemination, liver metastasis, lymph node metastasis, histological type, depth of invasion, surgical method, operative curability, adjuvant chemotherapy, percentage PCNA labelling, AgNOR count and the period of diagnosis (time trends).
Multivariate analysis revealed that tumour size, gross appearance, operative curability and percentage PCNA labelling were independent prognostic factors of Borrmann type 4 gastric carcinoma (Table III).

Discussion
The results of clinical treatment of patients with Borrmann type 4 gastric carcinoma remain poor. The associated lymph node metastasis, invasion into neighbouring structures and peritoneal dissemination present a great challenge for medical  care (Furukawa et al., 1988). Our previous data (Mori et al., 1993) showed that the mean percentage of PCNA labelling of * Borrmann types 3 and 4 gastric carcinoma was 37.6% and that of Borrmann types 1 and 2 was 30.2%. The proliferative activity of invasive type carcinoma (types 3 and 4) was significantly higher than that of localised lesions (types 1 and * 2) (P <0.01). Between types 3 and 4, there is only a slight differences; thus, the proliferative activity of Borrmann type 4 was somewhat higher than that of other types of gastric  the mean peiod from the earliest recognisable lesions of gastric carcinomas to advanced scirrhous carcinoma. The rapid intramural invasiveness and the late detection of Borrmann type 4 carcinoma in the advanced stage may account for the bad prognosis. Even among patients with the same Borrmann type 4 carcinoma, there are variations in lifespan. Our investigation had revealed that patients with Borrmann type 4 gastric carcinoma of high proliferative activity had a poorer prognosis than did those with carcinoma of low proliferative activity. We previously reported that gastric carcinoma with high proliferative activity often metastasised to lymph nodes (Kakeji et al., 1991). The same trend was recognised even when the study was restricted to Borrmann type 4 carcinoma, and for patients with tumours of high proliferative activity the prognosis was poor. There was a significant relationship between PCNA labelling and AgNOR count; hence these two parameters are probably interdependent. As both factors stain easily and paraffin-embedded tissue sections can be used, either is likely to lead to a better understanding of the proliferative activity of cancer cells.
Lymph node involvement, serosal invasion, peritoneal metastasis and macroscopic subtype have been considered useful prognostic indicators of Borrmann type 4 gastric carcinomas (Nagayo et al., 1974;Furukawa et al., 1988). In the current multivariate analysis, tumour size, macroscopic subtype, operative curability and percentage PCNA labelling were independent factors associated with the prognosis. Pro- liferative activity is one of the independent prognostic factors of Borrmann type 4 carcinoma. As for macroscopic subtype, Sowa et al. (1989) found that extensive lymphatic spread was more often recognised in those tumours with giant folds than those without such folds. Iwanaga et al. (1983) found that giant fold type or stenotic type gradually extended to adjacent organs or to the peritoneum, and that the eroded type invaded via lymphatic vessels in a rather short time. In our study, though patients with giant fold-type carcinoma died earlier than those with the eroded type, there was no significant difference in proliferative activity among these macroscopic subtypes. We consider that proliferative activity is an objective factor to predict survival of a patient. By estimating the proliferative activity, the physician can estimate the extent of lymph node metastasis and the prognosis, and can tailor post-operative adjuvant chemotherapy for individual patients. For patients with carcinoma of a high proliferative activity, aggressive adjuvant chemotherapy is the policy in our clinics.