| Objective: To discuss the relationship between concentrations of arsenic in drinkingwater and risks of liver cancer and evaluate the risks of liver cancer in concentrations ofarsenic in drinking water and the rationality of health standards for arsenic in drinkingwater, by establishing a mathematical model of dose-response relationship; To discussthe arsenic benchmark dose of liver injury in animals, to establish mathematical model ofdose-response relationship for arsenic-induced liver injury exposure in animals, and toselect sensitive indicators for liver injury induced by arsenic, providing scientific basisfor early diagnosis and prevention of arsenic-induced liver injury; To discuss effects ofinorganic arsenic on survival and apoptosis rates of Chang liver cells in normal humansand observe effects of arsenic on antioxidant enzymes, as well as on RNA expression ofHGF and Bax genes, and to establish mathematical model of time-dose-response (effect)relationship of injury induced by inorganic arsenic on normal liver cells cultured in vitro.Methods: Meta-regression was adopted to merge the slope coefficients of dose-responserelationship between concentrations of arsenic in water and risks of liver cancer derivedfrom different studies, establishing mathematical model of concentrations of arsenic inwater and dose-response relationship based on multiple independent studies; Integralanimal experiments were adopted to test liver toxicity in rabbits with various doses (0,0.075,0.110,0.150,0.250,0.300,0.375,0.500,0.750, and1.500mg/kg·w) of inorganicarsenic, and arsenic-induced hepatotoxicity in normal human Chang liver (24h,48h,72h,and96h) was tested with various concentrations of (0.2,2,5,10,15,20,25,50,75, and100μmol/L) sodium arsenite (NaAsO2) and arsenate. Enzyme-linked immunosorbentassay was applied to determine contents of ALT, AST and HA, TBA, γ-GT, TNF-α, IL-6,TIMP-1, and MMP-8in animal serum and liver tissues. MTT reduction was used to detect survival rates of Chang liver cells; apoptosis was detected with flow cytometry;contents of amounts of reduced glutathione (GSH) were detected with micro-ELISA;activity of catalase (CAT) was detected with visible light; activity of superoxidedismutase (SOD) was detected with WST-1; activity of glutathione peroxidase (GSH-Px)was detected with colorimetry; mRNA expression levels of TIMP-1, TIMP-2, andMMP-8in liver and HGF and Bax in Chang liver cells in rabbits were detected withReal-time PCR. Curve Estimate was adopted to fit a mathematical model ofdose-response relationship of arsenic-induced injury to rabbit liver. Time-dose-effectmodel of arsenic taking effects on Chang liver cells was established withthree-dimensional modeling method. Results:1) Dose-response relationship betweenconcentrations of arsenic exposure in drinking water and risks of liver cancer could beexpressed as: lnRR=0.00137X. The best slope factor was5×10-6, and the maximumslope factor calculated on the basis of the upper limit was8×10-6. When risk of livercancer reached1×10-5, arsenic exposure was1.25μg/kg/day; according to subgroupanalysis, the best slope coefficient of dose-response relationship in males was1×10-5, andthe maximum slope factor calculated on the basis of the upper limit was5×10-6; the bestslope coefficient of dose-response relationship in females was1×10-5, and the maximumslope factor calculated on the basis of the upper limit was5×10-6. When male AR reached10-6, arsenic concentration was approximately2μg/L; doses inducing liver cancer inmales were lower than those in females. When risks of liver cancer was1/100,000,arsenic concentration in drinking water was45μg/L, with risks of liver cancer in males as20μg/L and females as103μg/L.2) The total arsenic levels in the urine of rabbits wereincreased along with the increasing doses of arsenic exposure. When the dose reached0.5mg/(kg.w) and above, arsenic levels in urine showed statistically significantdifference compared with0-dose group (P<0.05), all with distribution patterns asDMA>MMA> iAsⅢ>iAsⅤexcept the0.5mg/(kg.w)-dose group which showed adistribution pattern of urinary arsenic as DMA>MMA> iAsⅤ>iAsⅢ; The total arseniclevels in liver were increased along with the increasing doses of arsenic exposure. Whenthe dose reached0.11mg/(kg.w) and above, arsenic levels in liver showed statisticallysignificant difference compared with0mg/(kg.w) group (P<0.05), with a wide variety ofarsenic distribution patterns in animal liver exposed to arsenic, and no specific rule hadbeen found; arsenic exposure dose of0.11mg/(kg.w) and above could lead toarsenic-induced liver injury; levels of ALT, AST, TBA, HA, TIMP-1, and MMP-8variedin each exposure dose group, with statistically significant differences (P<0.05); activity levels of γ-GT in each exposure dose group showed no statistically significant difference(P>0.05). AST, ALT, TBA, and γ-GT showed positive correlations (P<0.01) incorrelation analysis; there was a positive correlation between AST, ALT, TBA, γ-GT andliver pathological score (P<0.05); levels of TNF-α and activities of IL-6were differentamong all exposure dose groups (P<0.05), and IL-6, TNF-α, HA, and TIMP-1showed apositive correlation with exposure dose, as well as with liver pathological score (P<0.01).There was a negative correlation between MMP-8and exposure dose (P<0.05). Throughstepwise regression analysis of multivariate linear regression, it was found that AST,TBA, and MMP-8IL-6, TNF-α could be used as markers of arsenic-induced liver injury.Expressions of TIMP-1, TIMP-2, and TGF-β1varied in different dose groups (P<0.05)and gene expression levels of TIMP-1, TIMP-2, and TGF-β1showed an upward trend;the optimal model of dose-response relationship of external and internal exposure doseswas the Quadratic model; the Logarithmic model was the optimal model ofdose-response relationships between external exposure dose and biological effective dose,internal exposure dose and biological effective dose, and arsenic levels in liver andpathological score; BMD of arsenic-induced liver injury was0.0607mg/(kg.w) daily, andBMDL was0.0441mg/(kg.w) daily.3)72h after exposure, arsenic levels in cells weredifferent among different concentration groups (P<0.05); compared with control group,with increasing doses of arsenic exposure, survival rates were lower in exposure groupand showed a certain time-dose dependency (P<0.05); apoptosis rates were increasedwith the dose of arsenic exposure in exposure group (P<0.05); GSH and GSH-Px inChang liver cells exposed to iAsⅢand iAsⅤwere both found to be increased with lowdose and decreased with extending time of exposure and increasing doses (P<0.05),while activities of SOD and CAT were found to be decreased with extending time andincreasing dose; expression levels of Bax mRNA in cells exposed to iAsⅢand iAsⅤincreased (P<0.05), while mRNA expression of HGF was decreased (P<0.05); throughfitting results from a three-dimensional time-dose-response mathematical model, it wasfound that the optimal mathematical model for time-dose-response of survival rate,apoptosis rate, GSH, GSH-Px, and SOD was the three-dimensional Taylor polynomialmodel. Conclusions: There was a positive dose-response relationship between thearsenic concentrations in drinking water and risks of liver cancer. Based on the currenthealth standard for arsenic(10μg/L),We can infered that arsenic could induce liver injuryin rabbits and showed a dose-effect relationship. The models, mainly polynomial, of thedose-response relationship could be different with different indicators. Based on the reference dose extrapolated from experiments on liver injury in rabbits, current healthstandards can be considered reasonable and feasible. Arsenic can enter the liver cellsthrough the cell membrane, increasing the arsenic concentrations in cells. Althoughcytotoxicity of iAsⅢis higher than that of iAsⅤ, both of them can induce apoptosis ofChang liver cells to decrease survival rates of cells, block cell damage repair byinhibiting the expression of HGF and inhibit cell proliferation. They show stimulatingeffect at low-dose and toxicity at high-dose by adjusting the levels of antioxidantenzymes. Time-dose-response model of iAs to cells is in compliance with generalizedlinear models, and the three-dimensional time-dose-response mathematical model hasbroad application prospects in toxicological researches. |
PDF Full Text Request