Biomedical Microdevices

, Volume 14, Issue 4, pp 709–720

Concentration-dependent cytotoxicity of copper ions on mouse fibroblasts in vitro: effects of copper ion release from TCu380A vs TCu220C intra-uterine devices


  • Bianmei Cao
    • School of Materials Science and EngineeringUniversity of Science and Technology Beijing
    • Beijing Institute of Medical Device Testing
    • School of Materials Science and EngineeringUniversity of Science and Technology Beijing
    • School of Materials Science and EngineeringUniversity of Science and Technology Beijing
    • National Institute for the Control of Pharmaceutical & Biological Products
  • Chuanchuan Zhang
    • National Institute for the Control of Pharmaceutical & Biological Products
  • Wenhui Song
    • Wolfson Center for Materials ProcessingSchool of Engineering and Design, Brunel University
  • Krishna Burugapalli
    • Brunel Institute for BioengineeringBrunel University
  • Huai Yang
    • School of Materials Science and EngineeringUniversity of Science and Technology Beijing
  • Yanxuan Ma
    • School of Materials Science and EngineeringUniversity of Science and Technology Beijing

DOI: 10.1007/s10544-012-9651-x

Cite this article as:
Cao, B., Zheng, Y., Xi, T. et al. Biomed Microdevices (2012) 14: 709. doi:10.1007/s10544-012-9651-x


Sustained release of copper (Cu) ions from Cu-containing intrauterine devices (CuIUD) is quite efficient for contraception. However, the tissue surrounding the CuIUD is exposed to toxic Cu ion levels. The objective for this study was to quantify the concentration dependent cytotoxic effects of Cu ions and correlate the toxicity due to Cu ion burst release for two popular T-shaped IUDs - TCu380A and TCu220C on L929 mouse fibroblasts. Fibroblasts were cultured in 98 well tissue culture plates and 3-(4,5-dimethylthiazol- 2-yl)-2,5-diphehyltetrazolium bromide (MTT) assay was used to determine their viability and proliferation as a function of time. For cell seeding numbers ranging from 10,000 to 100,000, a maximum culture time of 48 h was identified for fibroblasts without significant reduction in cell proliferation due to contact inhibition. Thus, for Cu cytotoxicity assays, a cell seeding density of 50,000 and a maximum culture time of 48 h in 96 well plates were used. 24 h after cell seeding, culture media were replaced with Cu ion containing media solutions of different concentrations, including 24 and 72 h extracts from TCuIUDs and incubated for a further 24 h. Cell viability decreased with increasing Cu ion concentration, with 30 % and 100 % reduction for 40 μg/ml and 100 μg/ml respectively at 24 h. The cytotoxic effects were further evaluated using light microscopy, apoptosis and cell cycle analysis assays. Fibroblasts became rounded and eventually detached from TCP surface due to Cu ion toxicity. A linear increase in apoptotic cell population with increasing Cu ion concentration was observed in the tested range of 0 to 50 μg/ml. Cell cycle analysis indicated the arrest of cell division for the tested 25 to 50 μg/ml Cu ion treatments. Among the TCuIUDs, TCu220C having 265 mm2 Cu surface area released 9.08 ± 0.16 and 26.02 ± 0.25 μg/ml, while TCu380A having 400 mm2 released 96.7 ± 0.11 and 159.3 ± 0.15 μg/ml respectively following 24 and 72 h extractions. The effects of TCuIUD extracts on viability, morphology, apoptosis and cell cycle assay on L929 mouse fibroblasts cells, were appropriate for their respective Cu ion concentrations. Thus, a concentration of about 46 μg/ml (~29 μM) was identified as the LD50 dose for L929 mouse fibroblasts when exposed for 24 h based on our MTT cell viability assay. The burst release of lethal concentration of Cu ions from TCu380A, especially at the implant site, is a cause of concern, and it is advisable to use TCuIUD designs that release Cu ions within cytotoxic limits yet therapeutic, similar to TCu220C.


Intrauterine devicesCopperCell behaviorMaterial-cell interactions

1 Introduction

Since the first reporting of the contraceptive efficacy of Cu in IUDs by Zipper et al., Cu metal has been widely used in the form of wires, tubes and beads for contraception (Zipper et al. 1969). CuIUDs gained popularity because, they are non-hormonal, highly effective, inexpensive, long-lasting (use of single device reported for 20 years), rapidly reversible and safe contraceptives (Hov et al. 2007; Hubacher et al. 2001; Sivin 2007; Thiery et al. 1985). Currently, there are over 150 million IUD users (about 15 % of the world’s women of reproductive age) around the world (Hubacher et al. 2006, WHO 2002).

CuIUDs are usually made of Cu wires or tubes wound around T or U shaped flexible plastic material and their names typically contain a number indicating the total available Cu surface area in mm2. Depending on the available Cu surface area, currently manufactured TCuIUDs, are indicated for short-term usage of 3 years (200, 220, 250 mm2) or for long-term of 5, 8 or 10 years (300, 375, & 380 mm2) (Kulier et al. 2007). However, in practice, effective contraception was reported for TCu220C and TCu380A for at least 10 and 20 years respectively (Sivin 2007; Thiery et al. 1985).

The high effectiveness of Cu for contraception, unlike metals such as Pt and Ag, is due to Cu metal’s high solubility and contraceptive ability even in the presence of chloride ions (Chang et al. 1970). The corrosion products, primarily Cu2+ ions, released from the CuIUD cause the release of uterine inflammatory reaction products - leukocytes and prostaglandins - by the endometrium in response to the inserted CuIUD; together, they form a hostile environment in the uterus that reduces not only the viability of sperms and eggs, but also the receptivity of endometrium to implantation of embryos (Arancibia et al. 2003; Araya et al. 2003; Beltran-Garcia et al. 2000; Hagenfeldt 1972; Kaplan et al. 1998; Ortiz et al. 1996; Roblero et al. 1996; Shimizu et al. 1991; Toder et al. 1988; Yin et al. 1993). However, like any therapeutic drug, CuIUDs also induce side effects(pain, bleeding and pelvic inflammatory disease), the causes for which are not yet fully understood, but are reported to decrease to some extent over time (Hubacher et al. 2009; Stanback and Grimes 1998).

Two aspects of Cu ion toxicity need particular attention: one, the exposure of tissues in intimate contact with the device to (often lethal) Cu ion concentrations, and two, the chronic systemic exposure of rest of the body (Arancibia et al. 2003; Cai et al. 2005; Cao et al. 2008; Grillo et al. 2009; Hefnawi et al. 1974). In either case, cells take up Cu ions, and when the uptake crosses a threshold limit, the cells become apoptotic and die (Araya et al. 2003; Aston et al. 2000; Grillo et al. 2009; Hayashi et al. 2006; Obata et al. 1996; Prasad et al. 1996; Singh et al. 2006). The entire genital tract is reported to be exposed to 25–80 μg/day of Cu2+ ions released from an inserted CuIUD (Arancibia et al. 2003; Kjaer et al. 1993). The resulting damaged uterine tissue might largely be removed by menstruation, often increasing the severity and time of menses (Beltran-Garcia et al. 2000). Furthermore, the sustained daily release of Cu2+ ions also causes chronic systemic exposure (De la Cruz et al. 2005; Okereke et al. 1972; Shubber et al. 1998). For instance, irrespective of the age of the user or the length of time of using TCu380A, De la Cruz et al., reported the sustenance of nearly double the normal values for Cu concentration in blood (De la Cruz et al. 2005).

The elevated plasma Cu levels due to TCu380A are reported to cause time-dependant increases in clinical oxidative stress biomarkers such as thiobarbituric acid reactive substances, protein carbonyls, glutathione, nitrates, nitrites, metallothioneins, ceruloplasmin, and the hepatic enzymes - lactate dehydrogenase and transaminases. As a result, to prevent oxidative damage, Arnal et al. recommended avoiding the continuous use of TCu380A beyond 2 years (Arnal et al. 2010).

In addition, the sustained high concentration of Cu, if not excreted fast enough, results in accumulation of Cu in tissues (Cu overloading). The Cu uptake by cells is reported to be mediated primarily by ceruloplasmin, histidine and other Cu containing proteins (DiDonato and Sarkar 1997). Within the cell, Cu is distributed in all components, including nucleus, mitochondria, lysosomes, endoplasmic reticulum and cytosol (Linder 1991; Prasad et al. 1996). Excess Cu in cells is thought to interact non-specifically with various macromolecules, modify their conformation or cause site-specific damage (Burkitt 1994; Grillo et al. 2010; Kang et al. 2004; Letelier et al. 2005). The resulting disruption of fundamental cellular processes triggers apoptosis (Araya et al. 2003; Aston et al. 2000; Grillo et al. 2009; Hayashi et al. 2006; Obata et al. 1996; Prasad et al. 1996; Singh et al. 2006). However, threshold limits of Cu accumulation beyond which cellular damage is triggered, as demonstrated in Wilson and Menkes disease models having abnormal Cu accumulation, varies between different cell types (hepatocytes, neurons, and kidney cells) (Aston et al. 2000; Hayashi et al. 2006). Thus, a major objective for this study was to identify the threshold for Cu induced cytotoxicity to fibroblasts, one of the important (connective tissue formation and remodeling) cell type, affected in the uterus.

The overall objectives for this study were to evaluate Cu ion cytotoxicity on L929 fibroblasts as a function of concentration and correlate with that due to Cu ions in the burst release extracts of TCu220C and TCu380A. Cells were cultured on tissue culture plastic and MTT cell viability, Annexin V—propidium iodide (PI) apoptosis, and cell cycle analysis assays were used to assess the effects of Cu ions on the viability and proliferation of L929 fibroblasts. To this goal cells were treated with Cu ion solutions of concentration 0.1, 0.5, 1, 5, 10, 25, 40, 50 and 100 μg/ml, as well as 24 and 72 h extracts of TCu220C and TCu380A IUDs. The cell culture experiments were performed similar to that recommended in ISO-10993-5 standard and the results demonstrated a Cu ion concentration-dependent (including that in TCuIUD extracts) cytotoxicity (ISO10993-5 2009).

2 Materials and methods

2.1 Materials

The IUDs, TCu380A and TCu220C were supplied by Tianjin Medical Instrument Factory, China. Both IUDs are made of bare Cu wires or tubes wound on the two cross-arms and the stem of T-shaped high-density polyethylene (HDPE). The TCu380A had a Cu wire of 0.24 ± 0.01 mm diameter wound on the stem and Cu tubes on each side of the two cross-arms, while TCu220C had five Cu tubes on the stem and two Cu tubes on each side of the two cross-arms. The Cu tubes on both devices had outer and inner diameters of 2.18 ± 0.03 and 1.48 ± 0.01 mm respectively. The purity of Cu in both IUDs was 99.99 %. The devices were produced about 6 months before being used in the experiments.

The L929 mouse fibroblast cell line was provided by Shanghai Institute for Biological Sciences, Chinese Academy of Sciences. MTT was purchased from AMRESCO-Inc; dimethyl sulfoxide (DMSO) and 0.25 % Trypsin (with 0.02 % ethylenediaminetetraacetic acid (EDTA)) from Sigma–Aldrich; Pen-Strep, L-Glutamine, Dulbecco’s modified Eagle’s medium (DMEM) without phenol red and Fetal bovine serum (FBS) from GIBCO®, Invitrogen; and Dulbecco’s phosphate-buffered saline (DPBS) (Ca/Mg free) and DMEM with phenol red from Hyclone, Thermo Scientific Inc.

2.2 Cu ion extracts from the IUDs

Sterile TCu380A, TCu220C, and T-shaped HDPE stem without Cu wires or tubes were incubated in 10 % FBS supplemented DMEM (FBS-DMEM) culture medium (1 ml per 0.2 g of Cu) at 37 °C for 24 and 72 h. HDPE extract was used as a negative control and 5 % DMSO solution as a positive control. The total weight of Cu on TCu220C and TCu380A, devices used in this study, was 0.3250 and 0.1845 g respectively, and total Cu volume was 70.57 and 39.87 mm3 respectively. As a result, the weight to volume ratios for Cu TCu220C and TCu380A were similar (~0.00462 g/mm3) (Table 1). However, their total Cu surface areas were different: 265 and 400 mm2 respectively. Furthermore, the surface of the fine Cu wire wound on TCu380A appeared to have micro-cracks, compared to a smoother surface on Cu tubes, on the tested TCuIUDs (Cao et al. 2008).
Table 1

Total geometric surface area and weight to volume ratio of Cu on TCuIUDs available for Cu extraction (in 10 % FBS supplemented DMEM culture medium at 37 °C) and the corresponding Cu ion concentration in TCuIUD extracts


Cu Surface area (mm2)

Cu Weight to Volume ratio (g/mm3)

24 h Extract (μg/ml)

72 h Extract (μg/ml)




9.08 ± 0.16

26.02 ± 0.25




96.7 ± 0.11

159.3 ± 0.15

The Cu ion concentrations in Cu solution standards and extracts from the TCuIUDs were analyzed by atomic absorption spectroscopy (FAAS Thermo Electron Corporation M6AA System) using a Cu hollow-cathode lamp with an air–acetylene flame (acetylene, 1.0 L/min). The measurement was performed at a wavelength of 324.8 nm. Cu ion concentrations in extracts were determined against a standard curve of absorbance at a wavelength 324.8 nm obtained using Cu ion standard solutions of concentrations 0.0, 0.1, 0.2, 0.4, 0.8, 1.6, 2.0 μg mL−1. The standard curve (A = 1.9902 C(mg/ml) + 0.0009, where A is the absorbance and C the concentration of Cu ions) had R2 (linearity) of 0.9999.

2.3 Cell culture

The L929 mouse fibroblasts were cultured to about 80 % confluence in a T75 flask containing DMEM culture medium supplemented with 10 % FBS, 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mmol/L L-Glutamine, at 37 °C in humidified atmosphere of 5 % CO2 and 95 % air. The cells were washed with fresh DMEM, trypsinized, centrifuged, re-suspended and counted under an inverted microscope (ZEISS, Axiovent) using haemocytometer, and cell viability was determined using tryphan blue exclusion method. Cell seeding concentrations, for the in vitro biocompatibility assays, were varied to suit the different test requirements described below.

2.4 MTT cell viability assay

The viability of L929 fibroblasts on tissue culture plastic was measured by monitoring their metabolic activity using MTT assay. At the time of assay, cells cultured in 96 well plates were washed with DPBS, and then incubated in 100 μl of DMEM (without phenol red) containing 5 mg/ml MTT for 4 h at 37 °C in a 5 % CO2 humid atmosphere incubator. After incubation, the medium was removed and cells washed with DPBS. To develop the color 100μL of DMSO was added and the plate shaken gently for 10 min. The absorbance was measured on a microplate reader (SPECTRA MAX plus 384, Molecular Devices) at a wavelength of 570 nm, with a reference wavelength of 650 nm, against medium only blank.

2.5 Cell viability as function of number of seeded cells and time

To assess the MTT cell viability as a function of cell numbers and time, cells were seeded at 1 × 104, 2 × 104, 4 × 104, 5 × 104, 6 × 104, 8 × 104 and 10 × 104 cells/well (100 μl/well) in 96 well plates. Following cell seeding, MTT cell viability assay was done every day for 7 days to study the growth and proliferation of L929 mouse fibroblast cells.

2.6 Effect of Cu ions on cell viability

A series of Cu ion solutions, 0.1, 0.5, 1.0, 5.0, 10.0, 25.0, 40.0, 50.0 and 100.0 μg/ml were prepared by serial dilution of a commercial standard cupric ion solution using FBS-DMEM cell culture medium (the National Institute of Metrology, China) and their effects on cell viability determined using MTT assay. L929 fibroblast cells were seeded at 5 × 104 cells/well in 96 well plates (100 μl/well). After 24 h cell attachment, cells were washed with DPBS and incubated in culture media containing the different Cu ion concentrations for 24 h at 37 °C in a CO2 incubator. Culture medium without Cu ions was used as a negative control, while 5 % DMSO solution in distilled water as a positive control. After 24 h treatment, the medium was removed and cells washed in DPBS followed by the MTT assay. The results were expressed as percentage relative growth rate (%RGR) calculated using the equation:
$$ \% {\hbox{RGR}} = \left( {{\hbox{OD}}\;{\hbox{for}}\;{\hbox{Test}}\;{\hbox{Material}}/{\hbox{OD}}\;{\hbox{for}}\;{\hbox{Negative}}\;{\hbox{Control}}} \right) \times {1}00. $$

Similarly, the effect of Cu ion concentration in extracts from TCu380A and TCu220C on cell viability was also evaluated by the MTT assay and the results expressed as %RGR.

2.7 Morphology of cells on tissue culture plastic under inverted light microscope

The effect of TCuIUD extracts on fibroblast morphology was also studied using inverted light microscope equipped with phase contrast accessories.

2.8 Annexin V apoptosis assay

For the Annexin V apoptosis assay, L929 fibroblasts were seeded in 96-well plates at a density of 1×106 cells/well. Cu ions solutions in FBS-DMEM culture medium making up to 25, 40 and 50 μg/well; and 24 and 72 h extracts for both TCu380A and TCu220C were added to the wells seeded with fibroblasts. Culture medium without Cu ions was the negative control. After 24 h of treatment, Annexin V assay was performed using Annexin V-Fluorescein isothiocyanate (FITC)/propidium iodide (PI) apoptosis kit (BD Pharmingen, San Jose, CA) as described previously [25]. Briefly, the test cells were trypsinized, washed in DPBS and re-suspended in binding buffer at a concentration of 1×106 cells/ml. To 100 μl of cell suspension, 5 μl each of Annexin V-FITC and PI solutions were added and incubated in the dark for 15 min at room temperature. An additional 400 μl of binding buffer was added and the cells were analyzed by flow cytometry using BD FACSCalibur™ (BD, Franklin Lake, NJ) equipped with CellQuest software. Cells were categorized as viable (Annexin V-/PI-), early apoptotic (Annexin V+/PI-) or late apoptotic (Annexin V+/PI+), and expressed as a percent of total gated cells.

2.9 Cell cycle analysis

CycleTEST™ PLUS DNA Reagent Kit (BD Biosciences, San Jose, CA) in combination with PI was used for cell cycle analysis. Cell seeding and treatments were similar to that used for apoptosis assay. The cells were incubated in the dark with DNA staining solution at 4 °C for 10 min and then analyzed using BD FACSCalibur™ equipped with Modifit3.0 software. The number of gated cells in G0/G1, S, or G2/M phases was presented as % of total cells.

2.10 Statistical analysis

All cell cultures were performed in triplicate for each measurement at each time point. All statistical computations were performed using SPSS Version 16.0. Data is expressed as mean ± standard deviation. Differences of P < 0.05 based on Student’s t-test were considered significant.

3 Results

3.1 Cu extraction from TCuIUDs

Cu content in TCu380A extracts was substantially higher than in TCu220C extracts at both 24 and 72 h of extraction in FBS-DMEM culture medium (Table 1). TCu380A having ~1.5 times higher apparent geometric surface area (400 mm2) released ~10 and 6 times more Cu ions at 24 and 72 h respectively compared to TCu220C (265 mm2). This significantly higher Cu burst release from TCu380A is attributed to the physical form of Cu wound on the T-shaped HDPE. On the stem of TCu380A Cu wire is wound, which apparently releases Cu at much faster rate than Cu in the form of tubes. The two cross-arms on TCu380A and whole of TCu220C have Cu tubes wound around (Cao et al. 2008; Lu et al. 2008).

3.2 Cell viability as a function of number of seeded cells and time

Initially, the L929 fibroblasts’ proliferation in 96 well plates was tested. Cells were seeded at numbers of 1 × 104, 2 × 104, 4 × 104, 5 × 104, 6 × 104, 8 × 104 and 10 × 104, each in triplicate. After 1, 2, 3, 4, 5, 6, and 7 day intervals, cells were washed and assayed for MTT cell viability. Linear regression curve fitting on the optical density results were linear at 1 day (y = 0.000004x-0.0016; R2 = 0.9970) for the cell seeding range of 1 × 104 to 10 × 104, and thereafter the linearity decreased with time (R2 values of 0.9970, 0.9787, 0.9567, 0.8875, 0.8339, 0.6955 and 0.5829 respectively for the time points 1, 2, 3, 4, 5, 6 and 7 days).

3.3 Effect of Cu ion concentrations on viability of mouse L929 fibroblasts

We seeded cells at a density of 5 × 104 cells per well of a 96 well plate, intending to maximize the number of cells exposed to the different Cu ion solutions (concentrations ranging from 0.1 to 100 μg/ml or 24 and 72 h extracts from TCuIUDs containing up to 160 μg/ml). The time points relevant for our Cu cytotoxicity assays were 24 and 48 h. Cells were allowed to attach to the tissue culture plastic for 24 h, following which the non-attached cells washed off, before adding fresh medium with or without Cu ions for cytotoxicity testing. The Cu ion exposure was typically for up to 24 h, i.e., 2 days since initial cell seeding, unless otherwise mentioned. At 24 h after cell seeding, the cells appear to be in the logarithmic phase (R2 = 0.9970), and the linearity between viability and cell seeding numbers for the range of 1 × 104 to 10 × 104 cell/well at 2 days (R2 = 0.9787), did not decrease significantly. Hence, we assumed the effects of contact inhibition by the study end point of 2 days, interfering with Cu cytotoxicity assays, to be minimal for the cell seeding density of 5 × 104 cells/well.

Dose-dependent toxic effects of Cu ions were observed on the mouse L929 fibroblasts (Fig. 1). The data is expressed as %RGR with reference to negative controls. The viability of cells decreased with increasing Cu ion concentration, reaching zero viability when exposed to 100 μg/ml for 24 h.
Fig. 1

Concentration-dependent Cu ion toxicity on mouse L929 fibroblasts represented as % RGR based on MTT cell viability assay. Negative control cells were not exposed to Cu ion solution. n = 3

3.4 Cytotoxic effect of Cu ions in TCuIUD extracts

The cytotoxicity results for Cu ions in TCuIUDs extracts are presented in Fig. 2. The cells were seeded in 96 well plates, allowed to attach and spread on tissue culture plastic for 24 h followed by exposure to the TCuIUD extracts in media for 24 h. The positive control, DMSO, also significantly decreased the cell viability (%RGR of 5.58). Except for 24 h extracts of TCu220C, all the other groups significantly lowered %RGR compared to the negative control group. Effectively, the toxicity of the TCuIUDs is dependent on the amount of Cu ions released into the surrounding medium. The data expressed as % RGR (Fig. 2) showed that the extracts decreased L929 fibroblast viability in line with their Cu ion content, similar to that observed in Fig. 1. A difference in the mean % RGR values for 72 h extracts (77.61 %) containing 26.02 μg/ml Cu compared to that (85 %) observed for standard 25 μg/ml Cu ion standard solution. However, this difference was not statistically significant, due to the relatively high standard deviations observed for cell viability data for TCuIUD extracts.
Fig. 2

Cell viability assay results for mouse L929 fibroblasts exposed to extracts from TCu380A and TCu220C, expressed as % RGR versus negative control. * indicates p < 0.05 vs negative control

3.5 Morphology of L929 fibroblasts treated with TCuIUD extracts

Figure 3 shows the light microscopic images of cells on tissue culture plastic after 24 h of exposure to test (TCuIUD extracts) and control media solutions. The cell exposed to negative control (HDPE extract) show a confluent layer of fibroblasts, with well spread cells (Fig. 3a). In contrast, the cells were rounded and mostly detached from the surface of tissue culture plastic for the positive control (Fig. 3b). The cells exposed to extracts from TCu220C at both 24 and 48 h showed near normal fibroblast morphology (Fig. 3c&d). However, there is a decrease in the density of cells with increasing time of incubation with TCu220C extracts. In contrast, the cells exposed to TCu380A extracts were detached from surface of tissue culture plastic and rounded (Fig. 3e & f). Thus, both 24 and 72 h extracts from TCu380A proved fatal (due to the high Cu ion concentrations ~96 and 159 μg/ml) to mouse L929 fibroblasts.
Fig. 3

Inverted light microscope images showing the morphology of mouse L-929 fibroblasts treated with, (a) negative control (0 % Cu); (b) positive control (DMSA, 0 % Cu); (c) TCu220C 24 h extract (9.08 ± 0.16 μg/ml Cu); (d) TCu220C 72 h extract (26.02 ± 0.25 μg/ml Cu); (e) TCu380A 24 h extract (96.70 ± 0.11 μg/ml Cu); and f) TCu380A 72 h extract (159.30 ± 0.15 μg/ml Cu). Magnification a–d 100x; and e–f 200x

3.6 Apoptosis analysis

Flow cytometry plots showing the cell count events of L929 fibroblasts labeled with Annexin V and PI are presented in Fig. 4. The events in bottom left quadrant indicate Annexin V-/PI- viable cells, bottom right—Annexin V+/PI- early apoptotic cells, and top right quadrant—Annexin V+/PI + late apoptotic cells. Dose dependent Cu ion-induced apoptosis of L929 mouse fibroblasts cells is evident from the gradual shift of the fluorescent events from viable to early apoptotic and late apoptotic cells for Cu ion concentrations 0, 25, 40 through to 50 μg/ml (Fig. 4). The dose-dependent toxic effects of Cu ions on L929 cells are also quantitatively illustrated by a proportional and a significant (p < 0.05) increase in late apoptotic cells vs Cu ion concentration in Fig. 5. 25, 40 and 50 μg/ml Cu ion solutions caused 14.23, 30.5 and 47.12 % apoptotic cells respectively. Thus nearly 50 % cells were dead when exposed to 50 μg/ml.
Fig. 4

Cu ion-induced apoptosis for 1 × 106 L929 mouse fibroblasts exposed to 100 μl of cell culture medium containing Cu ion concentrations of (a) 0, b) 25 (c) 40 and (d) 50 μg/ml for 24 h. For the plots of PI versus Annexin V-FITC in (a) to (d), early apoptotic cells are in the bottom right; Late apoptotic cells in the right top; and live cells in the left bottom quadrants. *p < 0.05 for Cu ion solutions in (b), (c) and (d) vs negative control (a). n = 3
Fig. 5

Percentage of late apoptotic L929 fibroblasts as a function of exposure to different Cu ion concentrations. 1 × 106 L929 fibroblasts were exposed to 100 μl of cell culture medium with and without Cu ions. n = 3

The effects of 24 and 72 h Cu ion extracts from TCuIUDs 220 C and 380A on L929 fibroblast apoptosis was also evaluated (Figs. 6 & 7). The results further confirmed the dose-dependent effects of Cu ions on cells. TCu220C’s 24 and 72 h extracts contained ~9 and 26 μg/ml Cu ions respectively. Accordingly the percentage of late apoptotic cells (7.65 and 11.96) was low (Figs. 6 & 7). In contrast, all cells were dead by 24 h of treatment with both extracts from TCu380A. This can be attributed to the lethal Cu ion concentrations of ~96 and 159 μg/ml. To obtain meaningful intermediate stage cell apoptosis results, the treatment time of cells with TCu380A extracts was reduced to 12 h. The results showed that majority of the cells lost their viability and are in early or late apoptotic stages (Fig. 6). About 33 and 47 % of the total cells respectively for 24 and 72 h extract the TCu380A were in the late apoptotic cell stage within 12 h of exposure (Fig. 7).
Fig. 6

Cu ion-induced apoptosis for 1 × 106 L929 mouse fibroblasts exposed to 100 μl of (a) 24 h TCu220C, (b) 72 h TCu220C, (c) 24 h TCu380A, and (d) 72 h TCu380A extracts in cell culture medium. The Cu ion exposure time for (a) and (b) was 24 h and that for (c) and (d) was 12 h. For the plots of PI versus Annexin V-FITC in (a) to (d), early apoptotic cells are in the bottom right; Late apoptotic cells in the right top; and live cells in the left bottom quadrants. *p < 0.05 for a) to d) vs negative control (0 μg/ml Cu). n = 3
Fig. 7

Percentage of late apoptotic L929 fibroblasts exposed to the different TCuIUD extracts. The Cu ion exposure time for the TCu220C extracts was 24 h and that for TCu380A extracts was 12 h. n = 3

3.7 Cell cycle analysis

Cell cycle analysis was used to further characterize the Cu ion toxicity on L929 fibroblasts. Nuclear DNA was stained with propidium iodide and the fluorescence intensity of PI stained nuclei was measured using flow cytometry to classify the cells into G1, S or G2/M phases of cell cycle. L929 mouse fibroblasts cells were exposed to a series of Cu ion solutions to evaluate Cu ion induced cell cycle perturbations. As shown in Fig. 8 and Table 2, after 24 h of Cu ion solution exposure, the percentage of cells in S phase decreased significantly for all Cu ion concentrations compared with negative control cells (p < 0.05), while the percentage of the cells in G2/M phase increased (p < 0.05). The decrease in S-phase cells is indicative of inhibition of DNA replication, while the increase in G2/M phase cells of perturbation or arrest of cell division, the reduction in cell proliferation. Similar effects were observed when cells were treated with TCu220C extracts (results not shown), while all cells were dead by 24 h when treated with TCu380A extracts due to which we couldn’t capture any meaningful cell cycle perturbation effects.
Fig. 8

Cell cycle assay results showing the cell population distributions based on their stage in the cell cycle (G0/G1; S and G2/M), as a function of exposure to (a) 0, (b) 25, (c) 40 and (d) 50 μg/ml Cu ion concentrations. 1 × 106 L929 fibroblasts were treated with 100 μl of different Cu ion solutions for 24 h. n = 3

Table 2

Cell cycle perturbation induced by Cu ions—percentage of cells in the different phases of the cell cycle (G0/G1; S and G2/M) after treatment of 1 × 106 L929 fibroblasts with 100 μl of different Cu ion solutions for 24 h. n = 3

Cu ion Treatment (μg/ml)

G0/G1 (%)

S (%)

G2/M (%)


54.44 ± 0.4

30.49 ± 1.0

15.08 ± 1.1


55.93 ± 0.7

21.72 ± 0.7

22.35 ± 0.8


50.58 ± 1.3

29.45 ± 0.5

19.96 ± 0.6


53.81 ± 1.5

24.55 ± 0.8

21.64 ± 0.4

4 Discussion

Cu-containing IUDs are popular contraceptives because they are inexpensive, rapidly reversible, highly efficient and reliable. Moreover, a CuIUD once implanted can function for a long-time (3 to 20 years) (Sivin 2007). The contraceptive effects are due to the Cu ions released from the Cu wires/tubes wound around a T shaped polymer. Effectively the CuIUDs are biomaterials capable of sustained Cu ion release. In our earlier study we reported the initial burst release of Cu in the first three days from two commercial TCuIUDs followed by a sustained release in simulated uterine and body fluids. Further, we also showed CuIUDs get corroded faster when implanted in rat uterus (in vivo) than in simulated fluids (in vitro) (Cao et al. 2008). In this study, we analyzed the Cu ion concentration dependent cytotoxicity to L929 mouse fibroblasts cells and correlated with that due to the burst release extracts from two commonly used CuIUDs - TCu220C and TCu380A.

The weight to volume ratio of total Cu on each of the TCu220C and the TCu380A IUDs was similar (~0.00462 g/mm3), but the total surface area available differed (265 and 400 mm2 respectively). Yet, an abnormally high (6 to 10 times higher) burst release of Cu ions from Cu wires on TCu380A (0.24 ± 0.01 mm diameter) was observed compared to TCu220C (Table 1). This could be due to a size dependent non-linear increase in corrosion with decreasing Cu wire diameter, as reported by Lu et al. They observed the formation of a diffusion layer having an optimum thickness of 0.56 mm on Cu wire surface when immersed in a salt solution. If the radius of the immersed Cu wires was smaller than the diffusion layer thickness, a size dependent non-linear acceleration in Cu corrosion was observed with decreasing Cu wire radius (Lu et al. 2008).

The mode of action for TCuIUDs is through the inhibition of sperms’ motility and viability, thus preventing sperms from fertilizing oocytes. However, like any therapeutic drug, Cu ions released from TCuIUDs also cause cytotoxicity to other cells both in the vicinity of the implant site (in the uterus) and in remote tissues (including liver, kidney, spleen and lungs). Fibroblasts are one such cells in the close vicinity of implant that are affected by the Cu ion toxicity. They are crucial in the repair of the tissue damaged by inflammatory responses due to the implants and new tissue formation in uterus. Furthermore, L929 fibroblasts are the cells recommended by the international standard ISO-10993-5 for in vitro biocompatibility evaluations. Hence we chose L929 fibroblasts for our evaluation of Cu ion concentration dependent cytotoxicity.

In culture, fibroblasts move (locomotion), multiply (proliferation) and grow (growth) until they cover all the surface area available on the tissue culture plastic forming a confluent mono-layer. However, when a fibroblast comes in contact with its neighboring cell, the cell walls fuse at the contact points triggering a decrease in growth and stop proliferating, which phenomenon is called contact inhibition. To prevent the interference of contact inhibition with the Cu ion toxicity assays, cell seeding density and culture time were varied and MTT cell viability assessed. For a well of a 96 well plate, a linear increase in cell viability was observed when cells were cultured for 24 h for the cell seeding numbers of 1 × 104 to 10 × 104. A gradual decrease in linearity of MTT cell viability data was observed on a daily basis up to 7 days of testing, which was indicative of the high cell numbers in the small space of a well in 96 well plate causing contact inhibition (Meisler 1973a; Meisler 1973b). For the MTT assay to test Cu cytotoxicity, a cell seeding density of 5 × 104 cells/well and a maximum culture time of 48 h was chosen to ensure that the number of cells exposed to the different Cu ion concentrations was maximized without interference from contact inhibition.

The MTT cell viability assay revealed a concentration dependent decrease in viability of L929 fibroblasts (Fig. 1). Exposure of about 50,000 fibroblasts to 100 μl of 40 μg/ml (25 μM) Cu ion solution caused about 30 % decrease in the viability of L929 mouse fibroblasts. Considering a 30 % decrease in cell viability to be cytotoxic, as recommended by ISO 10993–5 standard, 40 μg/ml was identified as cytotoxic for L929 fibroblasts (Fig. 2) (ISO10993-5 2009). Furthermore, the MTT cell viability results correlated well with apoptosis assays. We observed a decrease in cell viability and a matching raise in apoptotic cell population with increasing Cu ion concentration. Cell cycle assay indicated the arrest of DNA replication and as a result, the arrest of cell proliferation. Overall, our cell viability data suggests that the LD50 for L929 fibroblasts is about 46 μg/ml (~29 μM) and almost all cells were dead when exposed to 100 μg/ml (~62 μM) of Cu ions (Fig. 1).

The cytotoxic effects due to Cu ions in burst release extracts from TCu200C and TCu380A were consistent with their respective Cu ion content. Our results, further reiterate that, the burst release in the first few days, especially for the TCu380A, exposes cells in close proximity to the device to lethal concentration of Cu ions (Pereda et al. 2008). TCu380A had a burst release extract concentration of >95 μg/ml which was lethal for the L929 mouse fibroblasts, as demonstrated by our cell viability, apoptosis and cell cycle analysis assay. Clinical reports document that TCu380A required removal in 15 % of users due to pain and bleeding (side effects) within 1 year of insertion (Hubacher et al. 2009). The TCu220C burst release extracts, on the other hand, had concentrations that are not as lethal. However, in clinical practice, pregnancy and expulsion rates of 0.8 to 2.2 % and 0 to 6.4 % respectively for TCu220C, and 0 to 1 % and 2.4 to 8.2 % for TCu380A were reported following use for one year (Kulier et al. 2007). Thus, the difference in performance between the two devices is only marginal.

The sustained high concentrations of Cu ions in the body due to TCuIUD cause an accumulation of Cu ions inside cells both in the vicinity of the implant and systemically (Arnal et al. 2010; Okereke et al. 1972). However, the definitive correlation of specific disease conditions, e.g., cancer, oxidative stress and liver dysfunction, to the long-term use of TCuIUDs in humans has been difficult to prove because the elevations in toxicity biomarkers are often subclinical (Arnal et al. 2010). The resulting clinical symptoms are often disregarded as harmless side effects. But there is growing evidence from a large number of in vitro and in vivo studies, in literature, suggesting the ill-effects of Cu ion toxicity.

Apparently, elevation in Cu ion levels in healthy individuals cause an increase in proteins involved in Cu ion metabolism, which in turn increase the active accumulation of Cu ions in cells systemically in the body (Arnal et al. 2010). The threshold for accumulation of Cu ions in cells beyond which Cu ion toxicity manifests varies between different cell types. Hayashi et al. demonstrated that 8.5 to 16 ppm of Cu produced single strand breaks in brain cells of Long Evans Cinnamon rats, an animal model for human Wilson Disease. However, higher concentrations (200 to 400 ppm) were needed to damage hepatic and renal cells (Hayashi et al. 2006). Furthermore, Aston et al. observed a time dependant linear accumulation of Cu in human hepatoma cells. At 72 h of culture in the presence of 4 and 64 μM (0.64 and 10.22 μg/ml) of Cu, they estimated a Cu content of 0.11 and 1.22 pg/cell and observed ~2 to 4 and 18 % necrotic cells respectively (Aston et al. 2000). The LD50 of Cu for 72 h culture of peripheral blood mononuclear cells (PBMCs) was reported by Singh et al. to be 115 μM (18.4 μg/ml), which increased to 710 μM (113.3 μg/ml) when the cells were pretreated with 200 μM of Zinc (Singh et al. 2006). For Chinese hamster ovary cells, Grillo et al. reported a significant decrease in viability of cells treated with ≥7.42 μg/ml (4.65 μM) Cu and ~90 % decrease when treated with 10.85 μg/ml (6.8 μM) for 72 h. Cortizo et al. tested Cu toxicity to URM106 rat osteosarcoma and MC3T3E1 osteoblast cell lines. They incubated the cells with Cu wires having 0.1 cm diameter and 0.314 cm2 total Cu surface area immersed in the culture medium. At 24 h of incubation, they observed a release of 104 μg/ml Cu ions into the culture medium, with 40 to 50 % decrease in surviving URM106 or MC3T3E1 cells and all cells were dead by 48 h (225 μg/ml Cu) (Cortizo et al. 2004). Cell necrosis was evident from as early as 4 h of incubation with Cu wires. Wahata et al. demonstrated the cytotoxic effects of Cu ions on human monocytes that play an important role in the biological response to implanted biomaterials. They tested the effects of Cu and other metals on the mitochondrial function and total cellular protein in THP-1 monocytes for 4 weeks. A Cu ion concentration ≥20 μM/l (3 μg/ml) produced a 35 % increase in total protein content, while 60 μM/l (9.6 μg/ml) produced a 75 to 125 % and 75 to 150 % rise in total protein and succinic dehydrogenase activity. The concentrations tested were said to be near lethal for the THP-1 monocytes by 24 h of cell culture (Wahata et al. 2002). Cu ion toxicity to human vascular endothelial (HVE) and fibroblasts (HAIN-55) was reported by Kishimoto et al. It is interesting to note that HVE cells were more susceptible to concentration-dependent Cu cytotoxicity than HAIN-55 cells (Kishimoto et al. 1992). Overall, our results are also indicative of similar cytotoxic effects on L929 mouse fibroblasts as that reported in literature for a variety of cells lines. Differences in the nature of cells, culture time, and units for the presented data make it difficult to tabulate comparable data. However, from the above discussion, it is evident that the thresholds for Cu accumulation vary for the different cell types in the body.

In human use, elevated plasma concentrations of Cu in blood were shown to cause chromosomal aberrations in blood lymphocytes (Shubber et al. 1998), when CuIUDs were used for over one year. Such positive correlations for long-term use of CuIUDs and DNA damage must be taken as a safety warning. At the same time, exposure to a minimum of 8 × 10−6 mol/L (0.5 μg/ml) of Cu ion concentration for 20 min is required to significantly reduce the motility of spermatozoa (Araya et al. 2003). Thus a balance of Cu ion concentration needs to be achieved in the uterus fluid, wherein the therapeutic effect is 100 %, yet the side effects are minimal.

To conclude, our findings suggest a LD50 dose of about 46 μg/ml (~29 μM) Cu ions for L929 mouse fibroblasts and >99 % cell death with 24 h exposure to 100 μg/ml (~62 μM). Cu ion concentration in burst release extracts (up to 160 μg/ml) for TCu380A is lethal to fibroblasts, and exposure of uterine tissues to such lethal concentrations (>100 μg/ml) should be viewed as a cause of concern. Furthermore, there is growing clinical evidence to suggest the deleterious cytotoxic and overloading effects due to TCu380A (Arnal et al. 2010; Aston et al. 2000; Beltran-Garcia et al. 2000; De la Cruz et al. 2005; Grillo et al. 2010). As a result, it is advisable to encourage the use of alternative TCuIUDs that release Cu ions well within the cytotoxic limit, yet highly effective for contraception.


The authors acknowledge the financial support from National Science and Technology Support Project of China (Grant No.2006BAI15B08), National Natural Science Foundation of China Project (Grant No. 51073024) and the Royal Society-NSFC international joint project grant (No. 5111130207).

Copyright information

© Springer Science+Business Media, LLC 2012