Selective cytotoxicity of copper-coated magnesium composite on human hepatoma carcinoma cells : a preliminary investigation

A copper-coated magnesium (Cu@Mg) composite has been prepared by electroless plating, with the aim of generating a novel antitumor agent. The cytotoxic effects in vitro of this composite on normal hepatocyte cells (L02) and hepatoma cells (97H) were evaluated by CCK-8 assay. Extract and direct contact tests were conducted with blank groups as the control. Moreover, cell adhesion test was performed with 316L medical stainless steel as the cell carrier. It was found that Cu@Mg composite showed obvious cytotoxic effects on 97H cells but acceptable cytocompatibility whit L02 cells. As illustrated by CCK-8 assay, the cytotoxicity of Cu@Mg on 97H and L02 cells were grade I and III, respectively, and more apoptosis occurred to 97H cells than to L02 cells. During direct contact test, much more pathological reactions such as rounding, shrinking, atrophic edges and clustering were found in 97H cells than those in L02 cells. Similar evidence was shown in the adhesion tests. According to the single-factor cytotoxicity evaluation of pH, Cu2+ and Mg2+, the selective cytotoxicity of Cu@Mg on 97H cells is attributed to the fast release of Cu2+ and OH−, resulting from the degradation of Cu@Mg in the culture medium, but the Mg2+ released in the same process shows no toxicity on the both cells. Therefore, it is promising to develop novel antitumor materials on liver cancers with good biocompatibility based on Cu@Mg composite.


Introduction
Magnesium and its alloys are considered to be a new generation of biodegradable biomaterials [1,2]. Recently, the antitumor effect related to the degradation of Mg have been investigated. The biodegradation of Mg in biological environment not only inhibited proliferation and induced apoptosis in human cholangiocarcinoma cells but also exhibited an inhibitory effect on tumor growth [3]. As shown by the antitumor evaluation of Mg on colorectal tumors in vivo (rabbits) and in vitro, Mg 2+ and H 2 released from the degradation of Mg can induce apoptosis of tumor cells and block them from migrating to normal organs [4]. As potential materials for bone scaffold, Mg and Mg-Zn alloys showed strong cytotoxicity on osteosarcoma (U2-OS) cell [5][6][7], which was ascribed to the alternation of the cell cycle and the activation of mitochondrial pathway to induce apoptosis [7]. The cytotoxicity of pure Mg on another osteosarcoma (MG-63) cell was also confirmed [8]. Fortunately, Mg did not exhibit toxicity to normal cells and tissues in tumor-bearing mice model [5]. Furthermore, the antitumor effect of WE43 alloy on human prostate carcinoma cell (LNCaP) and human breast adenocarcinoma cell (MDAMB-231) was enhanced by the higher corrosion rate induced by equal-channel angular extrusion (ECAP) [9].
As a well-established antibacterial agent, Cu 2+ ion shows excellent antitumor properties. Copper nanoparticles (CuNPs) exhibited cytotoxic to human breast cancer (McF-7) and colon cancer (LoVo) cells via inducing apoptosis and increasing anticancer genes [10]. The cytotoxicity of CuNPs on MCF-7 cells was shown by rupture of cell membrane and shrinkage and oxidative stress induced by reactive oxygen species (ROS), as well as mitochondrial damage [11]. They pointed out that while exposed to CuNPs the MCF-7 cells oxidative stress produced by extracellular ROS generation in treated with might damage membrane lipids causing cell shrinkage and fragmentation leading to cell death. And a highly effective and low-toxic Cu-based Myeloid cell leukemia 1 (McL-1) inhibitor was proposed as a promising candidate for the treatment of McL-1-related tumors [12].
Some Mg-Cu alloys and copper-containing coating on Mg alloys show antibacterial effects [13,14], but their cytotoxicity on tumor cells is not reported to the authors' knowledge. In the present work, a novel composite, Cu-coated Mg (Cu@Mg), was prepared by electroless plating. And its cytotoxicity has been tested with normal hepatocytes (L02) and hepatoma cells (97H) in order to evaluate the possibility of developing a better antitumor material with good security and efficacy on normal liver cells.

Material preparation and characterization
Pure magnesium (99.99 wt%) disks sized in Φ 9.0 mm × 1.0 mm were cut from as-extruded rods. After ground and ultrasonically cleaned in anhydrous ethanol, the disks were immersed in a 10 wt% CuSO 4 •5H 2 O solution at room temperature. A substitutional reaction between Mg and Cu 2+ occurred, leading to an electroless plating. A copper layer was coated on the Mg surface 15 min later, and a copper coated magnesium composite (Cu@Mg) was obtained.
The surface morphology and crystal structure of the composite were characterized by a scanning electron microscope (SEM, JEOL JSM-7800F) and an x-ray diffractometer (XRD, D/max-1200V).

Cytotoxicity tests
In vitro cytotoxicity tests were performed according to ISO 10993-5:2009, including extract test and direct contact test. The complete medium composed of Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS, pH 7.7) in addition was used for cell culture. Normal liver cells (L02) and human hepatoma cells (97H) (American Type Culture Collection, ATCC) were used in this work. Before cytotoxicity tests, the cells were passaged every 24 h and the fourth passage were retained as the tested cells, which were then cultured at 37°C for 72 h in an incubator aerated with 5% CO 2 . Liquid extracts of Cu@Mg and pure Mg were prepared by immersing the samples in the complete medium at 37°C for 24 h, followed by filtered with a 0.22 μm filter membrane. The ratio of sample surface area to medium volume was 0.15 cm 2 ml −1 . L02 and 97H cells were incubated in 96-well cell culture plates with 5000 cells and 100 μl medium in each well. After 24 h of culture, the culture medium was substituted by the extracts. Twentyfour hours later, the cytotoxicity of the extracts was detected by the Cell Counting Kit-8 (CCK-8) assay. The spectrophotometric absorbance of each well was measured using a microplate reader (Gen5 CHS 2.00) at 450 nm. Cells cultured in the culture medium only, was used as the control (blank group).
By the same procedures used in the extract cytotoxicity test, effects of pH, Mg 2+ , and Cu 2+ concentration on the cytotoxicity of the cell culture medium was tested by CCK-8 assay, separately. The pH value was adjusted from 7.7 to 10 using NaOH solution. Concentrations of 0 to 1000 mg l −1 Mg 2+ and 0 to 50 mg l −1 Cu 2+ were regulated using MgCl 2 •6H 2 O and CuCl 2 solutions, respectively.
Before direct contact tests, the cells were seeded in 24-well plates with 50000 cells and 1 ml medium in each well. After 24 h of culture, samples that had undergone 24 h UV sterilization were added into the wells and contacted directly with cells for another 24 h. Then the cells close to the samples in each group were observed under an Olympus EP50 optical microscope (OM). A blank group was used as the control again.

Cell adhesion test
Due to the hydrogen release resulted from the corrosion of the samples, cells cannot attach to the samples. Hence, medical stainless steel (316L) disks sized in Φ 9.0 mm × 1.0 mm were used as carriers for the adhesive cells. UV-sterilized 316L disks were placed into the 24-well plate as soon as the cells were seeded in. After 24 h of co-culture, the culture medium was changed and UV-sterilized samples were added into the wells with 3 wells in each group, and the blank group with 316L disk only was used as the control. The samples did not cover the 316L disks totally. Thus, the cell morphology at different locations was observed, i. e., directly under, near, and far from the samples. After co-cultured for 60 min, the 316L disks were removed and further prepared for cell morphology observation with a Quattro S environmental scanning electron microscope (ESEM).

Statistical analysis
Data are shown as the 'mean ± standard deviation' of results from at least three independent repetitions of each experiment. Statistical analysis was performed using the SPSS 20.0 software. One-way ANOVA was used to determine the difference among the groups, where p < 0.05 was considered to be a statistically significant difference.

Material characterization
The surface morphology and XRD spectrum of the Cu@Mg composite are shown in figure 1. As revealed by XRD analysis, only Mg and Cu phases are found in the composite, indicating that a Cu metal layer has been coated on the Mg matrix. The coating is composed of irregular Cu particles sized less than 2 μm, but the particles do not connect seamlessly with each other. Very tiny gaps and pores exist among the Cu particles, which are resulted from the side reaction of H 2 evolution during the electroless plating. More details of the microstructure coating will be reported next step. Figure 2(a) shows the CCK-8 cytotoxicity of different extracts on L02 and 97H cells. The proliferation of L02 cells is not inhibited obviously by the extracts of pure Mg (RGR102.8%) and Cu@Mg (RGR92.1%), indicating no and slight cytotoxicity expression, respectively. On the other hand, the relative proliferation rate of 97H cells is not affected by the pure Mg extract (RGR100.4%), but it is inhibited obviously by the Cu@Mg extract (RGR61.3%).

Extract tests
Before CCK-8 assay, the morphology of cells was observed under an Olympus EP50 microscope (figures 2(b)-(g)). The morphology of cells cultured in pure Mg extract are almost the same as those in the control, while the L02 and 97H cells in the Cu@Mg group changed in morphology. There are more shrinkage and death phenomena in 97H cells than that in L02 cells, implying stronger cytotoxicity of Cu@Mg on the former.

Direct contact tests
Cell morphology after 24 h of co-culture are shown in figure 3. Compared with the cells of the control, pure Mg inhibits the proliferation of L02 cells and 97H cells slightly, but shows no obvious toxic effects on the both. The proliferations of both cells are restrained obviously in Cu@Mg group, but the reduction of 97H cells is much more than that of L02 cells. And the pathological reactions such as rounding and shrinking (shown by the blue arrow in figure 3(f)) are more serious in 97H cells than in L02 cells. In addition, after 24 h of contact with Cu@Mg, almost all the 97H cells had atrophic edges (shown by the red arrow in figure 3(f)) and clustered together, indicating that the cells underwent apoptosis during the co-culture process.

Cell morphology adhering on 316L disks
An obvious cytotoxic effect on 97H cells was found within 30 min and became even worse 60 min later during the adhesion experiment. Figure 4 shows the ESEM morphology of adherent cells on 316L disks after 60 min of contact with the samples. The L02 cell density in pure Mg and Cu@Mg groups are both slightly more than that in the control. On the other hand, the density of 97H cell in pure Mg group is slightly decreased while that in Cu@Mg group is much decreased.
More detailed morphology is shown in the inserts in figures 4(a)-(f). In pure Mg group, the morphology of both L02 and 97H cells are similar to that in the control group, without visible abnormal features. Difference is found in Cu@Mg group. The morphology of L02 cells is not changed after co-cultured with Cu@Mg. Nevertheless, rounding and shrinking are found in 97H cells after the co-culture process. After 60 min of contact, the connection between 97H cells is broken and adhesion weakened significantly, meaning the apoptosis of cells.  In addition, the location-dependence of cell adhesion is found in 97H cells co-cultured with Cu@Mg. As shown in figures 3(f)-(h), the quantity of adhesion cells decreases gradually with the decreasing distance from the sample, and a large proportion of cell-free area appears directly under the sample. On the other hand, uniform cell adhesion is found in all the other three groups, i. e., L02 and 97H cells co-cultured with pure Mg and L02 cells co-cultured with Cu@Mg.
3.5. Influence of pH, Mg 2+ , and Cu 2+ on activity of cells As can be seen from figure 5(a), the proliferation of L02 cells is not affected by the increase of pH. But the RGR value of 97H cells decreased from 100% to 40% or lower once the pH goes up to 9.5 or more, meaning obvious cytotoxicity of grade III or higher. The proliferation rates of both the L02 and 97H cells are not inhibited by the increasing Mg 2+ concentration ( figure 5(b)), and even promoted slightly in some ranges. The effect of Cu 2+ concentration is more complicated ( figure 5(c)). The proliferation rate of L02 cells keeps at around 100% and then decreases with the increasing Cu 2+ concentration gradually from 20 mg l −1 on, indicating increasing cytotoxicity. However, the RGR value is higher than 75% as long as the Cu 2+ concentration is higher than 30 mg l −1 . Interestingly, the proliferation of 97H cells exhibits a fluctuation with the variation of Cu 2+ concentration. At low Cu 2+ concentrations, the RGR value is around 100%. But the RGR decreases abruptly to lower than 20% when the Cu 2+ concentration reaches 10 mg l −1 , and keeps at this level until the Cu 2+ concentration lower than 30 mg l −1 , implying strong cytotoxicity of grade IV. When the Cu 2+ concentration is increased to 35 mg l −1 or more, the RGR value goes back to around 100%.

Discussion
During immersion in DMEM culture medium, the corrosion of Mg is significantly accelerated by Cu coating because a typic bimetal corrosion process is resulted from the large difference in standard electrode potential of Mg (−2.37 V) and Cu (+0.34 V). As a result, the pH and Mg 2+ in the medium are much increased [3,5]. Moreover, the Cu coating is peeled off from the Mg substrate gradually as the corrosion proceeds, and dissolves in the medium, leading to the increase of Cu 2+ concentration in the medium [12]. As a result, the pH, Mg 2+ , and Cu 2+ concentrations in the extract of Cu@Mg are much higher than those in the extract of pure Mg (table 1), indicating the much faster dissolution of the former. And the difference in cytotoxicity of the samples can be attributed primarily to the variation in composition of the medium.

Cytotoxicity related to the release of ions
Magnesium and magnesium alloy have been proven to have certain antibacterial [15,16] and anticancer effects [3][4][5][6][7][8]. However, the toxic effect of Mg on bacteria and tumor cells is resulted from the increase of pH but not Mg 2+ concentration, although both of which are related to the corrosion of Mg [3,12]. The good compatibility of Mg 2+ with L02 and 97H cells is also confirmed in this work ( figure 5(b)). Considering the highest Mg 2+ concentration is around 400 mg l −1 in the extract of Cu@Mg, which is within the Mg 2+ concentration range shown in figure 5(b), the cytotoxicity of Cu@Mg has nothing to do with the rapid release of Mg 2+ ion.
On the other hand, the toxic effect of Cu 2+ ion itself has been reported on tumor cells [10,12,13,17,18]. Cu 2+ ion can not only induce apoptosis of cancer cells [10] but also inhibit the formation of tumor blood vessels [17]. And thereby it is an effective antitumor element. It is noteworthy that the cytotoxicity of Cu 2+ ion on L02 cells is quite different than that on 97H cells. Moreover, this difference is concentration-dependent. At the Cu 2+ concentration ranged from 10 mg l −1 to 25 mg l −1 , very strong cytotoxicity on 97H cells (grade VI) is obtained while the cytotoxicity on L02 cells is acceptable (lower than grade I as shown in figure 5(c)). Hence, a selective antitumor effect accompanied by good cytocompatibility with normal cells is desirable via controlling the concentration of Cu 2+ ion.
Cytotoxic effects of high pH on osteosarcoma cells (U2-OS) [5][6][7] and human erythroleukemia tumor cells (K562) [19] were reported and related to the corrosion of Mg alloys [6]. In this work, it has been found that the RGR of 97H cells decreases successively with increasing pH, leading to strong cytotoxicity at the pH over 9.5. This result is close to that of Y Zhang et al [6], where strong cytotoxicity on osteosarcoma U2-OS cells were found at pH values over 9.4. Nevertheless, the proliferation rate and morphology of L02 cells are not affected by pH up to 10. Thus, a selective toxic effect on 97H cells can be obtained by adjusting the pH to a proper range.  Additionally, H 2 is produced during the corrosion of Mg and Cu@Mg. But its effect on cell proliferation is not explored at this stage due to the difficulty of controlling the H 2 concentration in the culture medium. However, H 2 is an effective antitumor agent too [4,20,21]. And the antitumor effect of H 2 released by the biodegradation of Mg was confirmed [22]. The cytotoxicity on 97H cells is expected to be enhanced by the release of H 2 .

Selective cytotoxicity of Cu@Mg
The most important finding of this work is the obvious cytotoxicity of Cu@Mg on 97H cells accompanied with good compatibility with L02 cells, that is, selective cytotoxicity on human hepatoma carcinoma cells. As shown in figure 1, the RGR values of L02 and 97H cells in Cu@Mg extract are 92.1% and 61.3%, respectively. And much more cytopathic morphology such as shrinkage, rounding and apoptosis, occurred to 97H cells than to L02 cells. More evidence was revealed by contact experiments. The selective cytotoxicity of Cu@Mg on 97H cells is exhibited by the different pathological morphology in addition to the difference in the density of cells. An obvious cytotoxic effect on 97H cells was found within 30 min and became even worse 60 min later during the adhesion experiment. Furthermore, the density of 97H cells adhering to the 316L carrier is significantly decreased by the decreasing distance from the Cu@Mg sample. But the proliferation and adhesion of L02 cells were not affected visibly.
As shown in table 1, pH and Cu 2+ and Mg 2+ concentrations in the extract of Cu@Mg are 9.59, 19.3 mg l −1 , and 408.8 mg l −1 , respectively, which are much high than those of the blank complete medium. At the Cu 2+ concentration of 19.34 mg l −1 , cytotoxicity of grade I and grade VI on L02 and 97H cells can be resulted, respectively (figure 5(c)). At the pH of 9.59, cytotoxicity of grade I and grade III can arise on L02 and 97H cells, respectively ( figure 5(a)). That is, the cytotoxicity of the extract is attributed to the high pH and Cu 2+ concentration, but the Mg 2+ ion shows good cytocompatibility on both cells at this concentration [6]. Furthermore, both the pH and Cu 2+ concentration show acceptable cytocompatibility on L02 cell (RGR 99.5% and 88.9%) but strong cytotoxicity on 97H cell (RGR 44.6% and 16.2%). The similar selective cytotoxicity on tumor cells with acceptable cytocompatibility on healthy cells has been found in pure Mg [5] and curcumin [23]. It is interesting that the cytotoxicity on 97H cells of Cu@Mg extract is moderated if the cytotoxicity of pH and Cu 2+ concentration is considered separately. Therefore, more investigation will be carried out on the comprehensive influence of the pH, Cu 2+ , Mg 2+ ions, and H 2 gas as a whole on the activity of cells.

Summary
In summary, the most important finding of this work is the obvious cytotoxicity of Cu@Mg on 97H cells accompanied with good compatibility with L02 cells, that is, selective cytotoxicity on human hepatoma cells. Selective cytotoxicity on 97H cells can be acquired by controlling pH and Cu 2+ concentration in proper ranges. However, the pH value and Cu 2+ concentration in the microenvironment are difficult to control in vivo. On the other hand, the corrosion rate of Cu@Mg composite can be controlled effectively by adjusting the coverage rate of Cu-coating on Mg-substrate, i. e., the number of galvanic cells. And thereby the release of Cu 2+ , Mg 2+ , and H 2 as well as the pH of the microenvironment with Cu@Mg implant is ready to be administrated, so as to regulate the composition in a proper range and acquire an antitumor effect with acceptable compatibility with the normal tissues.