Aluminum (Al) is a nonessential biological element that is toxic to fish, interfering with ion regulation (Verbost et al., 1989; Booth et al., 1988) and respiratory function (Karlsson-Norrgren et al., 1985; Wood et al., 1988) in the gills. Fish re-production is also impaired by this metal (Haux et al., 1988; Rask et al., 1990). For example, the spawning of perch is de-layed in acidic water with high Al levels (Rask et al., 1990). Such environments have been shown to impair oogenesis due to the poor accumulation of egg yolk in the perch Perca flu-viatilis (Runn et al., 1977) and by reducing egg number in the pupfish Cyprinodon nevadensis (Lee and Gerking, 1980).

After absorption by fish, Al accumulates in the liver (Buer-gel and Soltero, 1983; Norrgren et al., 1991; Exley, 1996), where vitellogenin (VTG, a calcium-binding phosphlipogly-co-protein), an egg yolk precursor protein, is synthesized in mature females under estrogen (estradiol-17β, E₂) stimulation. VTG is also used as a nutrient source for embryonic and larval development until the beginning of feeding.

In recent years, attention has been paid to the potential significant impacts of ocean acidification on many calcifying organisms, especially corals and other invertebrates that pre-cipitate aragonite skeletons (Hoegh-Guldberg, 2009). In con-trast, the range of impacts that ocean acidification may have on marine fish remains poorly understood. Acidic water, in-cluding high Al concentrations may damage the reproductive physiology of stationary fish in coastal zones near estuarine outflow. However, few studies have investigated the effects of Al on the reproductive physiological processes of marine fish species. We used immature rockfish, Sebastes schlegelii, as an experimental fish. This species is an excellent indica-tor of coastal environment pollution, as it is fairly stationary in coastal rocky areas. The present study examined the toxic effects of Al on the liver, which is closely related to repro-ductive physiology in marine fish. Changes in plasma VTG induced by Al administration were analyzed by electrophore-sis. The concentrations of plasma alkaline-labile phosphorus (ALPP) and calcium (Ca) were also analyzed using a quan-titative analysis of inorganic phosphorus (Drummond L and Maher, 1984) and the o-cresolphthalein-complexone method (Bjorsson and Haux, 1985), respectively. An eventual toxic effect of Al administration was also analyzed by measuring the concentrations of plasma glutamate pyruvate transaminase (GPT), which has previously been used as an indicator of toxic effects on the liver (Hwang and Kang, 2002; Hwang et al., 2007). The hepatosomatic index (HSI) was expressed as liver weight × 100/body weight (BW). The concentration of Al in the liver was analyzed using an inductively coupled plasma-mass spectrometer (ICP-MS).

Immature Sebastes schlegelii, weighing about 50 g, were obtained from the Institute of Fisheries Sciences, Pukyoung National University (Busan, Korea) and kept in indoor tanks with continuously running water with 33±0.5 psu salinity and a constant temperature of 18℃. The fish were not fed during experimental periods.

Fish were intraperitoneally injected with E₂ (5 mg/kg BW) in 70% ethanol and Al (AlCl₃·6H₂0, Wako) (0.1, 1, 5, and 10 mg/kg BW) in distilled water twice at 3-day intervals, and blood samples were obtained from the fish (n = 7 per experimental tank), as described below, 7 days after the last admin-istration. Control fish received the E₂ only.

Fish were anesthetized with 2-phenoxyethanol and their blood was collected by injecting the tail of the fish and by draining the blood into heparinized capillary tubes. Plasma was separated by centrifugation (350 g, 8 min) and frozen at -20℃ until the analysis of VTG, ALPP, Ca, and GPT concen-trations.

Protein concentrations of the plasma were determined ac-cording to the method of Bradford (1976). Plasma proteins were also analyzed by 5-20% sodium dodecyl sulfate poly-acrylamide gel electrophoresis (SDS-PAGE) according to the method of Laemmli (1970). After SDS-PAGE, the integrated optical density (IOD) of the VTG band was measured using a Bio Image System (Millipore, Bedford, MA, USA) and was expressed as a percentage of the IOD of the total proteins in-cluding VTG. Minor subunits of VTG were not considered VTG, because the subunits constituted only a fairly small part of VTG and overlapped with other proteins (Kwon et al., 1993).

Fish were administered with Al twice at 3-day intervals, and were sampled 7 days after the last administration. The dosage was the same as described above. Plasma was separated from the blood of the fish and analyzed for the main VTG band by SDS-PAGE.

ALPP was analyzed from a 0.01 ml sample according to the method of Wallace and Jared (1968) and determined by colo-rimetric assay of the acidified phosphomolybdate complex using a commercially available kit (670-A; Sigma, St. Louis, MO, USA). This kit consists of the necessary reagents includ-ing a phosphorus reagent and calcium/phosphorus standard. The absorbance of the standards and samples were measured by an atomic absorption spectrophotometer (10-AU Fluoro meter; Turner Designs, Sunnyvale, CA, USA) at 340 nm.

Ca concentration was examined using a Ca assay kit (Asan Pharm. Co., Ltd, Seoul, Korea) by the o-cresolphthalein-com-plexone method (Bjorsson and Haux, 1985). The absorbance of the standard solution and samples were measured by an atomic absorption spectrophotometer (10-AU Fluoro meter; Turner Designs) at 575 nm.

GPT was examined using an assay kit (Asan Pharm. Co., Ltd) flowing the method of Reitman-Frankel (Racicot et al., 1975). This kit consists of the necessary reagents including standard solution, reaction solution, assay solution, and 4 N NaOH solution.

The assay was conducted according to the kit instruction. Briefly, 1 mL reaction solution was added to 0.2 mL plasma in Al-administrated rockfish and then 0.4 N NaOH was added and mixed at room temperature for 10 min. Distilled water was used as the blank. The standard solution of GPT and sam-ples were measured by an atomic absorption spectrophotom-eter (10-AU Fluoro meter; Turner Designs) at 505 nm.

The body and liver weights of Al-administrated fish were examined, and HSI was expressed as liver weight × 100/ BW.

Fish were administered with Al twice at 3-day intervals and were dissected in the laboratory 7 days after the last adminis-tration. Liver samples were stored in a deep-freezer at -20℃

[Fig 1.] Concentration-dependent inhibition of vitellogenin (VTG) production in the plasma of aluminum (Al)-administered rockfish. The activity of VTG production was estimated as a percentage of VTG to total proteins after sodium dodecyl sulfate polyacrylamide gel electrophoresis on Day 7 after the Al last administration. Vertical bars represent the SE of the mean for the five individuals. *P < 0.05 and **P < 0.01 for control (E2 only).

until analysis.

Liver samples were dried for 12 h at 80℃ and 0.1 g dried tissue was digested with 1.5 ml concentrated analytical HNO₃ in a Teflon polytetrafluoroethylene tube in a microwave oven (Yasunaga et al., 2000; Anan et al., 2001). Al content was de-termined using an ICP-MS (HP-4500; Hewlett-Packard, Illi-nois, USA).

Data were analyzed by one-way analysis of variance (ANOVA) (Fisher protected least significant difference test). Fisher’s tests were also used to examine the significance of correlation coefficients. Significance was accepted at P <0.05.

Plasma samples were extracted and then analyzed by SDS-PAGE in 7 days after E₂ and/or Al administration. VTG ac-counted for 23.6% of the total proteins in the control group, but this value decreased with increasing Al dosage (Fig. 1). Significant inhibition was confirmed at Al dosages of 1, 5, and 10 mg/kg, under which VTG productions were reduced to 54.2% (P < 0.05), 20.8% (P < 0.01), and 7.1% (P < 0.01) of the control, respectively (Fig. 1).

Immature rockfishes were intraperitoneally administrated with Al to compare the eventual dosage-dependent response and to avoid large amounts of Al-polluted waste water dur-ing experiment periods. Enzyme linked immunosorbent assay is a sensitive method for VTG determination and has been widely used for this purpose. In the present study, however, SDS-PAGE was used to separate VTG before optical quan-tification. VTG production in the presence and absence of Al was expressed as a percentage of total proteins. This type of expression has the benefit of excluding the effects of varia-tion in the amount of protein applied to the gel. It also has the advantage of checking the effects of Al on the production of proteins other than VTG. A significant decrease in percentage indicates that the synthesis of VTG is more susceptible to Al than are other hepatocyte-derived proteins.

Al is a nonessential biological element and therefore fish may have no regulatory mechanisms for it. The chemical form of Al depends on pH. At neutral pH, it occurs as Al(OH)₃ precipitate and/or in a soluble Al(OH)₄^- form (Verbost et al., 1992). As no precipitates were found, Al exerted its effect on vitellogenesis as an ion of Al(OH)₄^-. Al is toxic to fish, inter-fering with ion regulation (Booth et al., 1988) and respiratory function (Wood et al., 1988) in the gills. Fish reproduction may also be impaired by this metal (Rask et al., 1990). Hwang et al. (2000) reported that Al inhibits VTG and its mRNA in-duction by E₂ in the primary culture of hepatocytes in the rain-bow trout Oncorhynchus mykiss. Nakao et al. (2002) reported that Al inhibits E₂-induced VTG production in immature and male salmon Oncorhynchus masou. Similarly, in this study, Al inhibited E₂-induced VTG production. These results suggest that the inhibitory action of Al on E₂-induced VTG production in the rockfish Sebastes schlegeli is similar to that of previ-ously studied freshwater fish species.

Plasma samples were extracted and analyzed by colori-metric assay 7 days after E₂ and/or Al administration. Plasma ALPP decreased in a concentration-dependent manner in all groups of Al-administered fish compared to the control group that had been treated with E₂ only. Significant inhibition was confirmed at Al dosages of 5 and 10 mg/kg, at which ALPP concentrations were reduced to 70.7% (P <0.05) and 66.2% (P < 0.05) of the control, respectively (Fig. 2).

Emmersen et al. (1979) reported that plasma ALPP concen-tration increased in male flounder Platichthys flesus with E₂ treatment in a way similar to that of VTG. Also, the increase in VTG synthesis by 4-nonylphenol exposure is similar to that of ALPP concentration, which has been shown to be a very reli-able indicator of circulating VTG in fish (Nagler et al., 1987; Kramer et al., 1998). Similarly, the reduction in plasma ALPP concentration induced by Al administration was similar to that of VTG in this experiment using immature rockfish Sebastes schlegelii.

[Fig. 2.] Concentration-dependent inhibition of alkaline-labile phosphorus (ALPP) in the plasma of aluminum (Al)-administered rockfish. Vertical bars represent the SE of the mean for the five individuals. *P< 0.05 for control (E2 only).

[Fig. 3.] Concentration-dependent inhibition of Ca in the plasma of aluminum (Al)-administered rockfish. Vertical bars represent the SE of the mean for the five individuals. *P< 0.05 for control (E2 only).

Plasma samples were extracted and analyzed by the o-cresolphthalein-complexone method 7 days after E₂ and/or Al administration. Plasma Ca also decreased in a concentration-dependent manner in all groups of Al-administered fish com-pared to the control group treated with E₂ only. Significant inhibition was confirmed at Al dosages of 5 and 10 mg/kg, at which Ca concentrations were reduced to 74.7% (P < 0.05) and 70.0% (P < 0.05) of the control, respectively (Fig. 3).

Nagler et al. (1987) and Tinsley (1985) reported that the concentration of plasma Ca can also indicate circulating VTG.

[Fig. 4.] Concentration-dependent increase of glutamate pyruvate transaminase (GPT) in the plasma of aluminum (Al)-administered rockfish. Vertical bars represent the SE of the mean for the five individuals. *P< 0.05 for control (E2 only).

Ca binds to VTG, so that increased Ca may be mainly due to an increase in the protein-bound calcium fraction of the blood (Bjorsson and Haux, 1985). In the present study, the reduc-tions in Ca concentration induced by Al administration were similar to those of VTG. In the flounder Platichthys flesus, total protein concentrations decrease with increasing dosages of 4-nonylphenol, because of the effect of 4-nonylphenol on hepatic tissue (Christensen et al., 1999). In this study, the con-centrations of plasma Ca and total protein (data not shown) were reduced by Al administration.

Plasma samples were extracted and analyzed following the methods of Reitman-Frankel 7 days after E₂ and/or Al administration. GPT increased in a concentration-dependent manner in all groups of Al-administered fish compared to the control group treated with E₂ only. A significant increase was confirmed at an Al dosage of 10 mg/kg, at which GPT was 25.3% higher than that of the control (Fig. 4).

The concentration of the plasma enzyme GPT has fre-quently been used to detect eventual damage to the liver cells. Hwang and Kang (2002) reported VTG induction by exog-enous E₂ damage to hepatocytes, and the concentration of plasma GPT was temporarily increased in immature Sebastes schlegelii. Also, a significant increase in GPT concentration was reported in immature rockfish administrated 4-nonylphe-nol (Hwang et al., 2007). In the present study, a significant increase was observed in the concentration of plasma GPT at Al dosages over 10 mg/kg. This result indicates that Al dam-ages the hepatic tissue.

[Fig. 5.] The changes of hepatosomatic index (HSI) in Al-administered rockfish. Vertical bars represent the SE of the mean for the five individuals. *P< 0.05 for control (E2 only).

HSI was measured 7 days after E₂ and/or Al administration (Fig. 5). HSI decreased in a concentration-dependent manner in all groups of Al-administered fish compared to the control group administered with E₂ only. A significant reduction was confirmed at an Al dosage of 10 mg/kg, at which HSI was reduced to 85.3% of the control (Fig. 5).

Emmersen et al. (1979) reported that HSI increases in fish with E₂, because the fish liver is primarily hypertrophic for VTG synthesis. Hwang et al. (2007) also reported increases in HSI by 4-nonylphenol administration similar to those induced by E₂. In the present study, Al inhibited E₂-induced primary hypertrophy in the liver which reduced HSI.

The concentrations of Al in the liver 7 days after E₂ and/or Al administration are shown in Fig. 6. Al concentration increased in a concentration-dependent manner in all groups of Al-administered fish compared to the control group treated with E₂ only (Fig. 6). Significant increases were confirmed at Al dosages of 1, 5, and 10 mg/kg, at which Al concentrations were 1 ± 0.29, 3.46 ± 0.91, and 5.38 ± 0.92 μg/g dry wt, re-spectively (Fig. 6).

The livers of vertebrates such as fishes are an important organs for storing metals and for detoxification by biotrans-formation (Brown et al., 1984; Olsson et al., 1989). Due to its higher exposure to metal concentration compared to other tis-sues, the liver can also be used for the bio-monitoring of coast-al pollution (Von Landwust et al., 1996; Broeg et al., 1999).

In the present study, as Al levels increased, so did GPT levels, whereas a decreasing HSI indicated that Al affected

[Fig. 6.] Concentration-dependent increase of aluminum (Al) in the liver of Al-administered rockfish. Vertical bars represent the SE of the mean for the five individuals. *P< 0.05 and **P< 0.01 for control (E2 only).

the physiological activity of the liver. The internal accumula-tion of metals in aquatic organisms such as fishes is attributed to the introduction of dissolved metals through their diet or through the gills as they breathe.

Although the results of this study are inconsistent with the natural absorption of metals, they are very helpful in that the eventual dose-dependent responses could be induced and large amounts of Al-polluted wastewater could be avoided during the experiments, as described above. The accumulation of heavy metal in the liver tissue of fishes depends on temper-ature, age, aquatic chemicals, exposure time, and the amount of exposure (Pagenkopf, 1983; Heath, 1987; Goyer, 1991). De la Torre et al. (2000) reported that as Cyprinus carpio was exposed to heavy metals, its plasmid GPT level showed a concentration-related response. In the present study, GPT and Al levels in the liver were also high in the experimental group, which was exposed to high concentrations.

In conclusion, our results suggest that the concentrations of plasma ALPP and Ca could be utilized as an index of re-productive physiology in fish, because the changes in these concentrations after Al administration were similar to those of VTG. GPT concentration and HSI changed temporarily as in-creasing Al accmulation damaged hepatocytes. These results indicate that Al affects E₂-induced vitellogenesis in immature rockfish Sebastes schlegelii, which may result in the dysfunc-tion of reproductive physiological processes by having toxic effects on hepatocytes.