Metal-organic Framework [H2NMe2]3 [Tb(PDA)3] Induces Cytotoxic Effects and Changes DNA Methylation-related Gene Expression in a Human Fibroblast Cell Line

In this work, we describe the effect of six concentrations (0.05, 0.1, 0.2, 0.4, 0.8 and 1.6 mg/mL) of a metal-organic framework [H2NMe2]3 [Tb(PDA)3] on the viability of human broblast cells (hFB) after 48 h of exposure. The terbium metal-organic framework may be utilized for a variety of environmental and human health applications due to its luminescent properties and high thermal and water stability. However, the cytotoxic and epigenetic effects of lanthanide MOFs have not been fully elucidated to date. Therefore, we evaluated the cytotoxic effects, internalization and changes in DNA methylation-associated gene expression induced by exposing hFB to Tb-MOF. The images obtained by SEM, confocal microscopy and EDS analysis demonstrated the internalization of Tb-MOF at concentrations as low as 0.05 mg/mL, as well as dose-dependent toxicity in the broblasts, with a mean inhibitory concentration (IC50) of 0.260 ± 0.012 mg/mL being observed. Additionally, we detected signicant changes in the gene expression levels of DNA methyltransferases and demethylases (important regulators of development of cell) due to the presence of Tb-MOF, suggesting that Tb-MOF could contribute to cell death and destabilize the genome even at low concentrations. These results of this study may establish a foundation for future research attempting to develop and apply secure nanomaterials (e.g., MOFs) to minimize damage to the environment and human health.


Introduction
Recently, a new class of mesoporous materials, known as metal-organic frameworks (MOFs), has attracted considerable attention. These nanometric materials are composed of organic ligands interspersed with metallic ions or clusters to form networks of nanopores with large surface area and novel properties (Sajid 2016). These characteristics have enabled the materials to be applied with great success in such applications as catalysis, gas storage, energy storage, magnets, drug delivery, optoelectronics, light emitting devices and sensors (Gangu et  In general, it has been reported that the presence of the most common nanomaterials (NMs), such as TiO 2 , SiO 2 , CeO 2 , and Ag nanoparticles, causes adverse effects in vitro on lung, embryonic kidney and liver cells, as well as on skin cells ( broblasts and keratinocytes) and blood cells, such as macrophages (Franchi et al. 2015; Gong et al. 2010; Sharma et al. 2012; Wang et al. 2009). In these studies, cytotoxic and genotoxic effects were observed, and in the case of SiO 2 , epigenetic changes on keratinocytes were demonstrated (Gong et al. 2010). In this regard, these cells and others have been employed to test the toxicity generated by NMs according to the primary routes of exposure, such as ingestion, inhalation, blood circulation, and dermal penetration (De Matteis 2017). However, because the skin is a major route of exposure, there is interest in further characterizing the dermal absorption and toxicity produced by NMs, which is usually undertaken using skin cell models, such as keratinocytes and broblasts (Wang et al. 2018).
Regarding MOFs, contradictory results have been obtained in several different cell lines regarding the toxicity that these materials induce. Lin et al. (2016) suggest that zinc-based MOFs have potential applications for human healthcare due to their characteristics and because they do not cause signi cant cytotoxic effects on human cells. In contrast, other studies have shown that MOFs are more toxic than noble metal nanoparticles and cause microbial cell death and damage human cells (Bashir et al. 2015a). Moreover, it has also been observed that the presence of iron-based MOFs in retinal pigmented epithelium human cells damages the plasmatic membrane and depolarizes the mitochondrial membrane due to oxidative stress (Bashir et al. 2015b).
Due to the versatility and composition of MOFs, their exposure to humans is increasing rapidly, which means that studies should be conducted to evaluate the safety of this type of NM if their applications in biomedical, environmental or chemical elds are to be expanded. In particular, it has been suggested that it is important to complement the studies of traditional toxicity pathways with epigenetic studies, since epigenetic changes can be triggered by exposure to NMs (Dusinska et al. 2017 DNA methylation, the best-characterized epigenetic modi cation associated with transcriptional repression (Rinaldi and Benitah 2015), is tightly regulated by sophisticated enzymes that act as writers and erasers on DNA. The rst group consists of members of the DNA methyltransferase family, that is, DNMT1, DNMT3a, and DNMT3b, which add a methyl group at the carbon-5 position of cytosine to form 5-methylcytosine (5mC) (Plongthongkum et al. 2014). In contrast, demethylation is catalyzed by teneleven translocation enzymes (TET1, TET2, and TET3) through 5mC oxidation to hydroxymethylcytosine, and derivates from this reaction to enable proper gene regulation, cell differentiation, genome integrity, and several biological processes (Rasmussen and Helin 2016 3 ] (Ln = Sm, Eu, Tb), using a solvothermal method in the presence of a base and from inexpensive and readily available reactants, which might be used for diverse environmental and human health applications due to their luminescent properties and high thermal and water stability

Materials
All reagents used for the synthesis of Tb-MOF were purchased from Sigma-Aldrich Chemical Co., and reagent-grade solvents were dried and distilled according to procedures previously established in the literature (Viveros-Andrade et al. 2017). Single crystals were generated using spectroscopy-grade solvents. Melting points were measured in sealed capillaries on an IA-9100 model 10034137/01 Thermo Scienti c electrothermal apparatus. Elemental analyses were performed on a PerkinElmer Series II CHNS/O model 2400 analyzer using cystine as a standard. IR spectra in the range of 4000 to 400 cm − 1 were recorded in a PerkinElmer FT-IR Spectrum 400 spectrophotometer with a universal ATR sampling accessory at 298 K. The crystal was mounted on glass ber and studied with an Oxford Diffraction Gemini "A" diffractometer with a CCD area detector, with radiation source of λ MoKα = 0.71073 Å using graphite-monochromatized radiation.

General MOF [H 2 NMe 2 ] 3 [Tb(PDA) 3 ] synthesis "Tb-MOF"
The Tb-MOF was synthesized as previously reported (Viveros-Andrade et al. 2017). Brie y, a solution of terbium(III) nitrate pentahydrate (69.6 mg, 0.16 mmol) in 5 mL of deionized water was added dropwise to 5 mL of a methanol solution of 4,4′-bipyridine (53.5 mg, 0.32 mmol). The reaction mixture was re uxed for 90 min with stirring, after which a solution of pyridine-2,6-dicarboxylate acid (0.32 mmol) in 5 mL of N,N'-dimethylformamide (DMF) was slowly added at room temperature (Scheme 1). The reaction mixture was transferred into a Te on-lined stainless-steel vessel. The solution was heated at 120°C for 4 days under autogenous pressure; the reaction was slowly cooled to room temperature and subsequently heated at 45°C for 2 days; colorless prism-shaped crystals formed from the reaction were ltered off,

Cell line and culture
hFB from healthy breast tissue was provided by the Instituto Tecnológico y de Estudios Superiores de Monterrey, México. These cells were routinely grown in DMEM/F12 without phenol red medium (D2906, Sigma-Aldrich) supplemented with 10% fetal bovine serum (BIO-S1650, Biowest) and 100 µg/mL-100 U/mL-0.25 µg/mL streptomycin, penicillin and amphotericin B (ABL02, Caisson Laboratories), respectively, at 37°C in a humidi ed atmosphere with 5% CO 2 . To be analyzed, cells were detached from cell culture asks with 0.25% trypsin-EDTA (15090046, Invitrogen). Trypsin was neutralized with the same amount of complete culture medium. Concentration and viable cells were measured by the trypan blue exclusion method with the use of a hemocytometer.

Nanomaterial stock solution and dilutions
A stock solution of Tb-MOF at 3.2 mg/mL was prepared in Dulbecco´s phosphate buffered saline (DPBS 1X, D5652, Sigma-Aldrich). Brie y, the material was pre-sterilized with UV irradiation for 30 min and weighed in a 50-mL sterile tube. Next, the compound was dissolved in sterile DPBS, and the solution was homogenized with a sonicator microtip probe at 130-Watt, 70% amplitude, for 15 min with a Cole-Parmer 130-Watt Ultrasonic Processor 4347 (UK).
Six concentrations (0.05, 0.1, 0.2, 0.4, 0.8 and 1.6 mg/mL) were prepared through serial twofold dilution. For this step, the 1.6 mg/mL dilution was made by adding stock solution into complete 2X DMEM/F12 medium in a 1:1 proportion and then ltered with a 0.2 µm syringe lter. The subsequent dilutions were made in 1X culture medium prepared by dissolving the 2X medium in DPBS. Stock solution and dilutions were prepared immediately before use.

Cytotoxicity assay
The cell density of hFB exposed and not exposed to Tb-MOF was determined by staining with crystal violet (CV) according to Feoktistova et al. (2016) with some modi cations, in which the CV stains proteins and DNA of cells and the intensity of the color is proportional to the number of hypothetically viable adhered cells. Brie y, cells in the exponential phase of growth were inoculated in a at-bottomed 96-well polystyrene-coated plate at 2 × 10 4 cells/well in 100 µL of complete DMEM/F12 medium and allowed to attach for 24 h. Then, the medium was removed and replaced with 200 µL of diluted nanomaterial or only medium in triplicate. As a positive control known to be toxic to the cells, pure DMSO was used. The cells were incubated again for 48 h. After incubation, the cells were washed twice in a gentle stream of tap water, and the plate was inverted and left to dry on lter paper. In each well, 50 µL of 0.5% crystal violet (C6158, Sigma-Aldrich) in methanol (34860, Sigma-Aldrich) was added, and the plate was incubated for 20 min at room temperature. After being washed four times in a stream of tap water, the plate was air-dried, and the remaining stain was dissolved with 200 µL of methanol. Absorbance was measured with a microplate reader at 540 nm (Multiskan FC Thermo Scienti c). The percentage of cell density was calculated based on a comparison with control cells as (absorbance of exposed cells/absorbance of control cells) x 100. The cytotoxic concentration 50 (IC 50 ) was calculated by tting the data to a sigmoidal dose-response equation using GraphPad Prism 7.0. Three independent trials were performed.
In addition, photographic records were acquired before staining to observe the morphological changes of the cells and again after staining to demonstrate the effects of Tb-MOF on adhesion, as re ected in cell density.
2.6. Characterization of effects on human broblasts exposed to Tb-MOF 2.6.1. Cell preparation for microscopy The cells exposed to Tb-MOF were observed by eld emission scanning electron microscopy (FESEM JEOL JSM-7600F model) and confocal microscopy to observe the effect and the presence of nanoparticles on or within them. To this end, the cells were grown on Thermanox™ coverslips (13 mm diameter). Previously, the slides were placed on a sterile Petri dish and exposed to UV irradiation for 30 min. Next, the slides were washed two times with DPBS and decontaminated by being incubated for 24 h in DPBS with 1% antibiotic-antimycotic solution. The slides were placed on the wells of 24-well plates and washed twice with DPBS. Cells were inoculated at 5 x 10 4 cell/well in 1 mL of culture medium and incubated for 24 h. Next, the culture medium was removed, and dilutions of NMs were added to the wells (1 mL). The plates were incubated for 48 h, and at the end of that time period, the cells were washed twice with DPBS and examined under a microscope.

Cell analysis by FESEM
For the FESEM images, 3% glutaraldehyde in 0.1 M sodium cacodylate buffer was added to the slides with cells, which were incubated at 4°C overnight. After incubation, the slides were washed two times with DPBS, and the samples were dehydrated in increasing concentrations of ethanol (50, 70, 90 and 100%) for 15 min each time. At the end of dehydration, the samples were placed in a sample holder, which was soaked in ethanol. Next, the sample holder was submitted to a process of critical point drying with liquid CO 2 (equipment Quorum K850 model) for 1 h. Later, the samples were xed on aluminum stubs and were subsequently coated with a layer of gold-palladium (60:40% each) through a sputtering coater process (equipment Quorum Q150R ES model). Finally, the cells were observed using FESEM. The images were taken by a LABE detector with back-scattered electrons under high vacuum at 10 or 15 kV of voltage with a working distance of approximately 12 mm and magni cations of 150X and 2000X. Linear energy dispersive spectroscopy (EDS) analysis was performed to identify the presence of nanoparticles in a speci c area with an OXFORD-INCA detector.

Cell imaging by confocal microscopy
To evaluate the presence of Tb-MOF in the cells, based on its auto uorescence, the hFB cells were grown on Thermanox™ coverslips and exposed to 0.05, 0.2 and 16 mg/mL of nanomaterial for 48 h, as previously described. Subsequently, the cells were xed for 20 min at room temperature with 4% paraformaldehyde in DPBS followed by several washes with DPBS, and the uorescence properties of the cells were observed upon excitation at 330 and 490 nm accordingly to excitation spectra reported in Viveros-Andrade et al. (2017) by using a confocal laser scanning microscope (Olympus, FV100 SW). The images were acquired and analyzed by using FV10 ASW 3.1 viewer software (Olympus). As a reference, cell samples on slides were counterstained with DAPI mounting medium (H-1200, Vectashield ® ) according to the manufacturer's instructions, and the staining signal was detected using an excitation wavelength of 405 nm.

RNA isolation and reverse transcriptase-polymerase chain reaction (RT-PCR)
To analyze the changes in the expression of DNA methyltransferases and demethylases that are encoded by DNMTs (DNMT1, DNMT3a, DNMT3b) and TETs (TET1, TET2 and TET3), respectively, in cells exposed to Tb-MOF, cells were inoculated in a 24-well polystyrene-coated plate at 5 × 10 4 cells/well in 1 mL of medium and incubated for 24 h. After incubation, cells were exposed or not (as a control) to three concentrations of Tb-MOF (0.05, 0.2 and 1.6 mg/mL) and were incubated for 48 h. Thereafter, the cells were washed (twice) with 1 mL of DPBS 1X and trypsinized with 0.30 mL of 0.25% trypsin-EDTA for 5 min. Subsequently, trypsin was neutralized by the addition of 0.30 mL of growth medium (DMEM/F12), and the cells were collected in the same conical tube. The tube was centrifuged at 1000 rpm for 5 min, and the pellet was resuspended in 1 mL of DPBS. The cells were again centrifuged at 12,000 g for 5 min, and the pellet was employed to extract and purify RNA. Brie y, total RNA was isolated using the Direct-zol RNA miniprep kit (R2051, Zymo Research) according to the manufacturer's protocol. Reverse transcription reactions were performed using one microgram of RNA using the SuperScript III rst-strand synthesis system (18080051, Invitrogen) following the manufacturer's instructions. To evaluate the messenger RNA expression of DNA demethylases and methyltransferases, regular PCR was performed with MyTaq™ DNA Polymerase (21106, Bioline) using the constitutive gene β-actin as a control. The PCR products were electrophoresed in 1.5% agarose, stained with ethidium bromide (E1510, Sigma-Aldrich) and photographed using the Gel DocTM XR System (BIO-RAD). The primer sequences employed for the PCR analysis are listed in Supplementary Table S1.

Statistical analyses
CV tests were performed in three independent trials with at least three replicates per sample, and data were statistically analyzed through parametric one-way ANOVA and Tukey´s HSD post hoc test for each mean comparison. Statistical analysis and graphics were analyzed using GraphPad Prism 7.0 software. The data were plotted as the mean ± standard deviation (SD).

Effects of Tb-MOF on cell density
To determine the cytotoxic effects of Tb-MOF, hFB cells were exposed or not (for the control) to different concentrations of Tb-MOF for 48 h. After exposure, the effects on cell morphology were visible starting at a concentration of 0.4 mg/mL, from which a percentage of the cells shrank and exhibited a rounded morphology, which was more evident with increased concentrations of Tb-MOF (Fig. 1). In addition, an important reduction in cell density was also observed when the cells were evaluated by CV assay. These results indicate that Tb-MOF induces the loss of adhesion capacity to the substrate and, nally, cell viability (Fig. 1).
Although the effects of Tb-MOF on cell morphology could only be visualized with optical microscopy starting at the concentration of 0.4 mg/mL, the statistical analysis of the data obtained through the CV assay showed a dose-dependent response, with signi cant differences (p < 0.05) being observed in percentages of cell density with respect to the control at all concentrations (Fig. 2). Thus, the results obtained with these assays suggest that the density of the cells exposed to 0.05 mg/mL Tb-MOF is different from that of the control and that from the concentration of 0.8 mg/mL Tb-MOF, the cell density drops to less than 70% (Fig. 2). Similarly, it was possible to determine the inhibitory concentration 50% (IC50) by tting a sigmoidal dose-response curve (Fig. 3). The resulting average was IC50 = 0.260 ± 0.012 mg/mL. Therefore, these results indicated that broblast cells are damaged by exposure to Tb-MOF.

Cell interaction with Tb-MOF
To evaluate the cell interaction of Tb-MOF, hFB cells were exposed to the lowest (0.05 mg/mL) and highest (1.6 mg/mL) concentrations of Tb-MOF for 48 h (Fig. 4). In both cases, the presence of Tb-MOF was observed, and the clusters of nanomaterial could be clearly distinguished over cells exposed to 1.6 mg/mL, where MOF also induced a contraction of the cellular membrane (Fig. 4C). To determine whether the negative effects on cell morphology were due to Tb-MOF internalization within the cells, we carried out FESEM-EDS analysis. This analysis demonstrated the presence of terbium within the cells in the presence of 0.05 mg/mL Tb-MOF (Fig. 4D). This result shows that cell internalization of Tb-MOF occurs at concentrations as low as 0.05 mg/mL, which could be related to the observed cytotoxicity.
Moreover, considering that Tb-MOF exhibits photoluminescent properties (Viveros-Andrade et al. 2017), cell internalization within hFB of MOF was also evaluated by confocal microscopy. Consistent with the EDS analysis, we noted that Tb-MOF photoluminescent signals were detected in the cell clusters exposed to both 0.05 and 1.6 mg/mL MOF (Fig. 5) by using a wavelength of 330 and 490 nm. The results showed full nanoparticle internalization within the cells, particularly in the cytoplasmic region and probably in the nucleus, when the cells were exposed to a higher concentration (1.6 mg/mL) of Tb-MOF (Fig. 5). This result suggests that decreased cell density is correlated with increased internalization of Tb-MOF into broblasts.
3.3. Gene expression patterns of DNA methylation-related genes in hFB cells exposed to Tb-MOF.
DNA methylation, a well-established epigenetic mark, is widely involved in the control of gene transcription and cellular phenotypic changes, as well as genome stability. Therefore, in this study, we evaluated the gene expression pro le of DNMTs (DNMT1, DNMT3a, DNMT3b) and TETs (TET1, TET2 and TET3) on DNA in hFB cells in response to treatment with three different concentrations of Tb-MOF (0.05, 0.2 and 1.6 mg/mL) after 48 h of exposure (Fig. 6). We observed that Tb-MOF affected the expression of DNMTs. For instance, the expression of DNMT1 was only slightly affected by all Tb-MOF concentrations, whereas the expression of DNMT3a was downregulated by 0.2 mg/mL. In contrast, the levels of DNMT3B expression were affected by 1.6 mg/mL but not by 0.05 or 0.2 mg/mL Tb-MOF. Additionally, because 5-mC can be oxidized to 5-hydroxymethylcytosine (5hmC) as part of a pathway to reverse DNA methylation, we further determined the expression pro le of the TET genes. Treating hFB with Tb-MOF for 48 h induced downregulation of TET1 in all treatments, while both TET2 and TET3 expression were affected by 0.2 and 1.6 mg/mL but not by 0.05 mg/L of Tb-MOF. Taken together, these results suggest that the decreasing proliferation of hFB could be related to the imbalance of DNA methylation patterns, since the expression levels of both DNMT and TET are affected by Tb-MOF exposure.

Discussion
In this study, cellular toxicity in hFB was observed after treatment with 0.05 to 1.6 mg/mL Tb-MOF, exhibiting a dose-dependent response. Thus, the lowest ( On the other hand, in our work, we determined a Tb-MOF IC50 of 0.260 ± 0.012 mg/mL (Tb concentration of 0.052 mg/mL) on hFB. In this regard, Iram et al. (2016) showed that the presence of Tb 2 O 3 particles induced an IC50 of 0.001 mg/mL on the viability of osteosarcoma cells. This value was lower than that obtained in our work, probably due to the Tb concentration in Tb 2 O 3 particles (0.0008 mg/mL), which was lower than in our work. However, contrasting results were obtained by Setyawati et al. (2013) regarding the viability of hFB in the presence of Tb-Gd 2 O 3 . These researchers observed that 0.75 mg/mL of Tb-Gd 2 O 3 (0.228 mg Tb/mL) induced a 40% reduction in the viability of hFB. Thus, the toxic effect obtained at a lower concentration by the Tb may indicate that the speci c effects on cell lines could be due to the characteristics of the complexes of NMs and not due to the Tb concentration. Thus, more studies must be conducted to determine the mechanisms by which materials with Tb as Tb-MOF generate cytotoxicity.
On the other hand, through inverted microscopy and FESEM-EDS analysis, it was determined that Tb-MOF (1.6 mg/mL) induced important changes in broblast morphology with respect to the control. For instance, irregular prolongations and oval nuclei, as well as a remarkable contraction of the cells, making them round, decreased their adhesion capacity for the culture surface, which was con rmed when they were lifted when washing the culture plates. Similar results have also been obtained in osteosarcoma cells, dermal broblasts and L929 cells when exposed to We showed that remarkable changes in cell morphology, such as small protuberances and oval nuclei, in hFB cells were mediated by the internalization of Tb-MOF and by its deposition onto the cell surface On the other hand, we showed that Tb-MOF induced a change in DNMT expression, mainly affecting DNMT3A and DMT3B, by exposing hFB cells at 0.2 and 1.6 mg/mL, respectively, but without clear evidence of being dose-dependent (Fig. 6). Although there are gaps in the body of knowledge regarding the epigenetic events that happen after cells are exposed to nanomaterials, the induced changes in DNMT expression by Tb-MOF might be related to a reduction in cell proliferation (Fig. 2) and viability (Fig. 1) observed in this work, probably affecting cell cycle-related genes; however, further studies are warranted to elucidate the mechanism by which cytotoxicity occurs on hFB cells. For instance, it may be interesting to determine whether Tb-MOF impairs DNA methyltransferase activity or only leads to repression of DNMT expression. In this regard, it has been reported that exposing keratinocyte cells (HaCaT) to SiO2 nanoparticles leads slight reduction in gene expression of all DNMTs in human embryonic kidney cells (HEK-293) exposed to ZnO (0.05 mg/mL) (Choudhury et al. 2017). Moreover, it has been reported that Ag nanoparticles are able to induce the overexpression of DNMTs in neuronal cells of the mouse (HT22) (Mytych et al. 2017). These results suggest that DNMT expression is dependent on the nanoparticle's speci c features and the type of cell to be evaluated.
On the other hand, TET family members (TET1, TET2 and TET3) catalyze the process of demethylation to maintain the correct balance of DNA methylation in the genome (Liu et al. 2013). In this regard, we found that while all Tb-MOF concentrations decreased TET1 and TET3 expression, Tb-MOF induced an increase in TET2 expression (Fig. 6).
Although the expression levels of TETs can vary among cells/organs, their speci c activities in many biological processes have not been elucidated ( Therefore, the dysregulation of TET proteins could compromise human health. Our results suggest that cells exposed to Tb-MOF may exhibit cytotoxic effects by changing the epigenetic landscape (in vitro and in vivo) with a signi cant impact on cell behavior, which could lead to the development of human diseases; however, there is little information regarding the mechanism that leads to the downregulation of TET expression by such compounds, indicating that further studies are warranted to clarify in greater depth the possible mechanisms governing the effects that nanomaterials, such as Tb-MOF, have on human health.

Conclusion
Taken together, our results clearly show that Tb-MOF can be absorbed and internalized in hFB cells, disrupting the morphology, adhesion properties, viability and cytotoxicity induction of the cells, probably by affecting DNA methylation-based processes by interfering with the catalytic domains of DNMT and TET family proteins. In addition, the responses obtained in the hFB suggest that Tb-MOF may activate ROS generation. Clearly, more studies are needed to determine the functional consequences and potential effects on human health caused by exposure to Tb-MOF and/or nanomaterial complexes. This study presents evidence relevant to the possible damage caused by some nanomaterials, such as MOFs, when they are used without knowledge of their effects on the environment and human health.

Declarations
Ethics approval and consent to participate Not applicable Figure 2 Cell density of human broblasts cells exposed to Tb-MOF. The cell density was calculated with respect to the density of the control (cells incubated without nanomaterials). The error bars indicate the standard deviation (n=3). * Indicates signi cant differences (p<0.05) calculated by Tukey´s HSD post-hoc test