Cytotoxicity of Carbon Nanotubes, Graphene, Fullerenes, and Dots

The cytotoxicity of carbon nanomaterials is a very important issue for microorganisms, animals, and humans. Here, we discuss the issues of cytotoxicity of carbon nanomaterials, carbon nanotubes, graphene, fullerene, and dots. Cytotoxicity issues, such as cell viability and drug release, are considered. The main part of the review is dedicated to important cell viability issues. They are presented for A549 human melanoma, E. coli, osteosarcoma, U2-OS, SAOS-2, MG63, U87, and U118 cell lines. Then, important drug release issues are discussed. Bioimaging results are shown here to illustrate the use of carbon derivatives as markers in any type of imaging used in vivo/in vitro. Finally, perspectives of the field are presented. The important issue is single-cell viability. It can allow a correlation of the functionality of organelles of single cells with the development of cancer. Such organelles are mitochondria, nuclei, vacuoles, and reticulum. It allows for finding biochemical evidence of cancer prevention in single cells. The development of investigation methods for single-cell level detection of viability stimulates the cytotoxicity investigative field. The development of single-cell microscopy is needed to improve the resolution and accuracy of investigations. The importance of cytotoxicity is drug release. It is important to control the amount of drug that is released. This is performed with pH, temperature, and electric stimulation. Further development of drug loading and bioimaging is important to decrease the cytotoxicity of carbon nanomaterials. We hope that this review is useful for researchers from all disciplines across the world.


Introduction
The carbon nanomaterials, such as carbon nanotubes (CNTs), graphene, fullerenes, and dots, attract the attention of researchers because of their unique physical properties. Carbon nanomaterials can cause damage and intoxication to organisms. The problem of toxicity of carbon nanomaterials for plants, microorganisms, animals, humans, and the natural environment is very important. Regarding humans, people work in laboratories worldwide in intercultural environments, and it is an important issue to save people from toxic substances.
These problems can be resolved by the chemical functionalization of carbon nanomaterials. There are five ways to chemically functionalize carbon nanomaterials. Among them is the covalent functionalization of the outer surface of carbon nanotubes [97][98][99][100], noncovalent functionalization of the outer surface of carbon nanotubes [101,102], the substitution of carbon atoms with other atoms , intercalation of carbon nanotube bundles , and filling of carbon nanotubes [153,154].
The covalent functionalization is performed via the covalent bonds between the functional groups and carbon nanomaterials. The functional groups are further used for the bioconjugation of carbon nanomaterials with therapeutic and targeting agents. This decreases the toxicity of carbon nanomaterials and increases the selectivity of drug delivery and the sensitivity and accuracy of bioimaging [155,156]. The noncovalent functionalization is performed via hydrophobic interactions, π-π interactions, and van der Waals forces between carbon nanomaterials and guest molecules. The functional groups can further be conjugated with therapeutic agents. The substitution and intercalation of carbon nanomaterials are other promising ways to functionalize them for theranostic applications. Endohedral chemical functionalization (filling of carbon nanotubes) is an important method for drug delivery and bioimaging applications. It is a very simple and viable method of chemical functionalization for biomedical applications. The filling ratios of medicines are controlled with the synthesis procedures.
In this review, we consider the functionalization methods of carbon nanomaterials and the cytotoxicity of carbon nanotubes, graphene, fullerene, and dots. In Section 3, we consider the microscopy issues. In Section 4, we highlight the theoretical methods. In Section 5, we discuss cell viability issues. In Section 6, we describe the drug release issues. In Section 7, we introduce perspectives.

Cytotoxicity Issues
There are three main issues of cytotoxicity: • Materials characterization, and theoretical considerations, • Cell viability, • Drug release.
For cytotoxicity issues, it is important to introduce the material characterization methods, such as microscopy, spectroscopy (Raman spectroscopy, optical absorption spectroscopy (OAS), Fourier transformed infrared spectroscopy (FT-IR)), and zeta potential measurements. There is a number of reports on the characterization of the material. We chose the most important reports to review here.
Theoretical works exist for carbon nanotubes, graphene, and fullerenes. Authors consider drug delivery systems with fullerenes and carbon nanotubes with defined atomic structures. The peculiarities of the drug delivery system and drug release are considered. Important structural models of drug delivery systems of such drugs as doxorubicin (DOX) are calculated. These issues are highlighted here.
The main part of the review is dedicated to the discussion of cytotoxicity studies. They are presented for A549 human melanoma cells, E. coli, osteosarcoma cell lines, the U2-OS cell line, the SAOS-2 cell line, the MG63 cell line, the U87 cell line, and the U118 cell line. The results are shown in bar diagrams, which are supported by images of cells and bacteria cultured with carbon nanomaterials.
The next important part of the review is drug release. The results are shown in drug release plots versus time under different pH for different functionalized carbon nanomaterials. Bioimaging results are shown here to illustrate the use of carbon derivatives as markers in any type of imaging used in vivo/in vitro. The review finishes with perspectives.

Microscopy
In order to observe the morphology of nanomaterials, different microscopies have been employed. Authors of Ref. [56] studied the effect of graphene oxide (GO) nanoparticles Nanomaterials 2023, 13, 1458 3 of 25 (GONPs), GONPs-polyethylene glycol (PEG) nanocomposite, and GONPs-PEG-N. sativa on the structure of organisms under treatment using the scanning electron microscopy (SEM) technique. Figure 1(a1) shows the scanning electron microscopy (SEM) images of non-treated bacterial cells [56]. Figure 1(a2) shows bacterial cells treated with GONPs. Figure 1(a3) shows bacteria cells treated with GONPs-PEG. Figure 1(a4) shows bacterial cell exposure with GONPs-PEG-N. sativa. The left panels correspond to E. coli, and the right panels belong to S. aureus.  SEM allows visualizing the effects. E. coli represent elongated-form bacteria, and all material influences its morphology. The modifications are denoted by white arrows in the left panels of the SEM images. S. aureus represent spherical-like material, whose modifications are marked by red arrows on the right panels. The modifications are observed for all material, and it is clearly visible on the SEM images.

Raman Spectroscopy
Raman spectroscopy investigations of carbon nanomaterials for cytotoxic studies were performed in Refs. [43,75]. Figure 1e shows the Raman spectra of the NL001 sample that belongs to the sample pristine sample, the NL002 samples that are the pristine material functionalized by -COOH groups, and the NL004 sample that is functionalized by polyethyleneimine@Naproxen.

Fourier Transformed Infrared Spectroscopy
The FT-IR spectra of carbon nanomaterials for cytotoxic investigations were performed in Ref. [44]. Here, we review the most important examples. Figure 1f shows partial FT-IR spectra of the purified reduced carbon nanotube nanocomposite Nano 6 ( Figure 1(f1)) and the same sample at the end of hydrocortisone release (Figure 1(f2)) [44].

Theoretical Methods
Authors of Ref. [42] developed the drug delivery system for the drug doxorubicin. This drug delivery system was capable of drug release under lower pH. Figure 2a shows the schematics of the drug delivery system [42]. Figure 2b shows the structure of SWCNT-CR-DOX systems at neutral pH values: SWCNT (10,0), 20 CR (red) and 10 DOX (blue) (Figure 2 [42]. Part of DOX goes inside CNTs, as we see in Figure 2c. Authors of Ref. [63] theoretically studied the dendro [60] fullerene/molnupiravir drug delivery system (Figure 2d    Authors of Ref. [17] studied the influence of different materials on the gills and digestive glands of Mytilus galloprovincialis. For gills, Figure 3b (upper part) shows the histopathological alterations. The hemocyte infiltration (arrows), enlargement of the central vessel (long arrows), and abundance of lipofuscin aggregates (*) were observed, especially for Hg-treated samples [17]. For digestive glands, Figure 3b (lower part) shows the histopathological alterations. The hemocyte infiltration (arrows), atrophy, and necrosis were observed, especially in mussel tissue exposed to Hg (Figure 3b) [17].

Biological Effect of Nanomaterials upon Cell (In Vitro)
Hematoxylin-eosin (HE) staining of lung sections exposed to MWCNTs is presented in Figure 3c. The images of lung tissues after intratracheal instillation after 3 days, 1 week, 1 month, and 6 months for control samples of dosages of 0.2 mg and 0.6 mg are shown. Authors of Ref. [2] observed that animals exposed to CNTs and euthanized at 0 weeks postexposure as well as 2 weeks post-exposure did not demonstrate major pathological changes ( Figure 3d). Authors of Ref. [5] investigated the toxicity of SWCNTs for fish. SWCNTs can not be seen using light microscopy. Near-infrared fluorescence molecular imaging (NIRF) was used to detect the nanotubes in cross-sections. Bright fluorescence of SWCNTs was observed in treated tissues (Figure 3e) [5].    license [21]. * p-value < 0.05, ** p-value < 0.01,*** p-value < 0.001, **** p-value < 0.0001, α, β, Φ: the mean ± standard error of the mean, n.s: not significant.
Authors of Ref. [60] studied the cell viability in osteosarcoma cell lines U87 (Figure 4(b1)) and U118 (Figure 4(b2)) in graphene oxide@PEG. MTT assay was made after 72 h of glioblastoma cell line (Figure 4(b3)) [60]. The protein expression level index (PELI) values in stress conditions for different concentrations of graphene oxide samples are shown in Figure 4c [19]. It is visible that in specific conditions, there is an enhancement of toxicity (asterisk, Figure 4c). In Ref. [21], the cytotoxicity of graphene nanoparticles in human keratinocytes cells (HaCaT) for an exposure period of 24 h and concentrations from 0 (control) to 10 µg/mL were analyzed. It is visible that there is no cytotoxicity for concentrations below ≤0.05 µg/mL (Figure 4d).
Authors of Ref. [22] investigated the developmental toxicity of reduced graphene oxide. It was shown that 2 µm × 2 µm reduced graphene oxide caused higher embryonic mortality than the sample of 400 nm × 400 nm reduced graphene oxide. For the 2 µm × 2 µm reduced graphene oxide, there was significant mortality for concentrations starting at 10.7 µg/mL, whereas 400 nm × 400 nm reduced graphene oxide did not lead to significant mortality at tested concentrations (Figure 5b). Authors of Ref. [26] compared the wastewater quality parameters, biological oxygen demand (BOD), chemical oxygen demand (COD), and total organic carbon (TOC) before and after photocatalytic treatment with aerogel photocatalytic membrane based on graphene oxide with Cr-Mn-doped TiO 2 (Figure 5c) [26]. Figure 5d shows the dependence of the efficiency of membranes on the number of cycles [26]. It is visible that the efficiency is retained and then decreases with increasing the number of cycles. Even after eight cycles, the membrane kept its efficiency of more than 50% in the wastewater.
Graphene oxide cytotoxicity is dependent on flake size and oxygen group composition. In Ref. [155], the authors studied the dependence of cytotoxicity on graphene flake size. Two samples with an average size of 20 nm (GO-20) and 100 nm (GO-100) were studied. It was shown that both samples inhibit the viability of TM3 and TM4 cells. However, with 100 nm-size graphene flakes, the cell viability is higher ( Figure 5(e1,e2)). Figure 5(e3,e4) shows TM3, and TM4 cell morphology, respectively, under a light microscope, compared with the control sample (Con) and silver nanoparticles (AgNPs) (scale bar 200 µm) [155].
In Ref. [156], the authors studied the dependence of cytotoxicity on oxygen group composition. Figure 5f shows TEM images of cellular internalization of reduced graphene oxide (rGO) and graphene oxide (GO) in Caco-2 cells. Figure 5(f1,f4) shows an untreated control sample. Figure 5(f2,f3) shows Caco-2 treated with rGO after 24 h and 48 h, respectively. Figure 5(f5,f6) shows Caco-2 treated with GO after 24 h and 48 h, respectively (scale bar: 2 µm). They found that cells treated with rGO showed significant changes, whereas cells treated with GO did not show such modifications. Cells treated with rGO for 24 h (Figure 5(f2)) demonstrated improved heterochromatin in the nucleus. Dense bodies in the cytoplasm, as well as cell fragmentation, were observed. There is rGO inside cells. Cells treated with GO for 24 h (Figure 5(f5)) heterochromatic nuclei were detected. Mitochondria are better conserved than rGO. There is GO inside Caco-2 cells, too. Cells treated with rGO for 48 h (Figure 5(f3)) exhibited the segregation of the nucleolus. The endoplasmatic reticulum appears as a dense body. Cells treated with GO for 48 h (Figure 5(f6)) showed more modified mitochondria than after 24 h [156]. Figure 5(g1) shows in vitro cytotoxicity of different samples of carbon dots, carbon dots-gel, diclofenac sodium (DS)-carbon dots, DS-carbon dots-Gel, and DS-Gel. Figure 5(g2) shows histopathological microscopy of treated ocular tissues [64].

Biological Effect of Nano Materials upon In Vivo
Authors of Ref. [30] studied the toxicity of graphene and GO on plant Lemna minor. It was observed that there was a decrease in the size of the leaves (white arrows, Figure 6a), and the roots had an appearance of white highlights. This was observed for graphene and GO. Nanomaterials 2023, 13, x FOR PEER REVIEW 16 of 28   Figure 6b shows brightfield imaging of 6 hpf embryos exposed to 400 nm × 400 nm reduced graphene oxide for different post-exposure periods in comparison with exposures to ultrapure water controls [22]. In Ref. [76], to study pharmokinetics, and biodistribution of 153 Sm@SWCNTs and 153 Sm@MWCNTs, bioimaging, a whole-body single-photon emission computed tomography/computed tomography (SPECT/CT) imaging, and quantitative γ-counting were applied The accumulation in the spleen, lung, and liver within 30 min was observed, as it is shown in Figure 6c. Positron emission tomography (PET), CT, with the functionalization of radioactive elements, was applied [94]. In Ref. [94], 64 Cu was linked with GO-PEG via 1,4,7-triazacyclo nonane-1,4,7 triacetic acid (NOTA, a chelating agent of 64 Cu). 64 Cu-NOTA-GO conjugated with TRC105 ( 64 Cu-NOTA-GO-TRC105) for targeting a CD105 (endogline) was synthesized. Figure 6d shows its targeting ability toward 4T1 tumor-bearing mice [94]. Authors of Ref. [27] performed the studies of cytotoxicity of graphene oxide and GO-Ginsenoside (Rg3)-Doxorubicin against Huh7 cells. Figure 7(b1) shows the transmission electron microscopy images of treated cells. Figure 7(b2) shows AlamarBlue cell viability assay 24 h after treatment with GO-Rg3-DOX. Figure 7(b3) shows the scanning electron microscopy of treated cells. Figure 7(b4) shows reactive oxygen species (ROS). Figure 7(b5) shows schematics of GO-Rg3-DOX internalization [27]. Figure 7c shows Rg3, DOX, and graphene oxide cytotoxicity in human breast cancer MDA-MB-231 cells [27]. AlamarBlue assay 24 h after GO, GO-Rg3, GO-Rg3-DOX treatment is shown together with ROS production. The Rg3 component reduced ROS generation. Therefore, it reduced the side effects on non-cancerous tissues. Nanomaterials 2023, 13, x FOR PEER REVIEW 18 of 28 Authors of Ref. [27] performed the studies of cytotoxicity of graphene oxide and GO-Ginsenoside (Rg3)-Doxorubicin against Huh7 cells. Figure 7(b1) shows the transmission electron microscopy images of treated cells. Figure 7(b2) shows AlamarBlue cell viability assay 24 h after treatment with GO-Rg3-DOX. Figure 7(b3) shows the scanning electron microscopy of treated cells. Figure 7(b4) shows reactive oxygen species (ROS). Figure  7(b5) shows schematics of GO-Rg3-DOX internalization [27]. Figure 7c shows Rg3, DOX, and graphene oxide cytotoxicity in human breast cancer MDA-MB-231 cells [27]. AlamarBlue assay 24 h after GO, GO-Rg3, GO-Rg3-DOX treatment is shown together with ROS production. The Rg3 component reduced ROS generation. Therefore, it reduced the side effects on non-cancerous tissues.  Figure 8a shows the drug release from the PEI-functionalized carbon nanotubes at different pH [43]. It is visible that the release of drugs is maximal at pH 4.1. Other pH leads to much lower releases (19% and 16%, accordingly).  Figure 8a shows the drug release from the PEI-functionalized carbon nanotubes at different pH [43]. It is visible that the release of drugs is maximal at pH 4.1. Other pH leads to much lower releases (19% and 16%, accordingly).  Authors of Ref. [40] studied sequential versus simultaneous addition of Congo Red and anticancer drug doxorubicin to single-walled carbon nanotubes. In Figure 8b, it is visible that the amount of DOX bound with triple complex SWCNT-CR-DOX is larger than the amount of DOX bound with CR-DOX complex. Figure 8c shows DOX release from the SWCNT-CR-DOX complex at pH 5 and pH 7.4 at room temperature [40]. It is visible that there is a gradual release of DOX. The values are obtained by the spectrofluorimetrical method. DOX release is better at lower pH because it is an acidic media. Figure 8d shows the dependence of DOX release on time PK 5 E 7 (PEI-rGO/DOX) at different pH [50]. There is an easy release of DOX in acidic conditions.

Advantages and Disadvantages of Using Nanomaterials
Authors of Ref. [52] studied the cumulative release of Quercetin from the microfiber scaffolds without electric stimulus (Figure 8(e1)), and under 10 Hz (Figure 8(e2)), 50 Hz (Figure 8(e3)). Figure 8f shows in vitro release profiles. Figure 8g shows the comparison of in vitro release profile of the LNZ, LNZ-BCDs nanocomposite [67]. It is visible that there is the release of almost 98 ± 0.5% after 6 h of LNZ from the free drug LNZ solution. The release of LNZ-BCDs nanocomposite has a similar manner. It is visible that the release of LNZ does not exceed 52 ± 1.3% after 6 h from LNZ-BCDs nanocomposite. This system has the following advantages and disadvantages:

•
BCDs have a small size, suitable optical and photoluminescence properties, and good photostability, and that is why they are promising nanocarriers of LNZ, • LNZ-BCDs nanocomposites show biphase release, which is important for improving tissue healing, • LNZ-BCDs nanocomposites were shown to have good biocompatibility and low cytotoxicity for human cells, • LNZ-BCDs nanocomposites have good antibacterial properties, • LNZ-BCDs nanocomposites have increased cell proliferation, which improves tissue regeneration and healing effect, • LNZ-BCDs nanocomposites can be considered as a replacement for toxic nanoparticles in biomedical applications and for drug delivery to mend humans [67].

Perspectives
We discussed in Section 2 that there are three cytotoxicity issues: material characterization, theoretical considerations, cell viability, and drug release.
Materials characterization methods are developed, and the progress of this topic is dependent on the achievements of microscopy, spectroscopy, and other methods, such as zeta potential measurements. The increase in accuracy and resolution of characterization techniques are significant, and it is connected with progress in synthesis methods of carbon nanomaterials, and theoretical considerations help.
Cell viability is important to be improved here. Cell viability can be used to investigate chemicals for harmful influence on different cells. It can be used to choose anticancer drugs. It is important to improve the detection sensitivity and experiments. This leads to improved knowledge about the cytotoxicity of chemicals and carbon nanomaterials on tissues, organs, and the whole body.
The important issue is single-cell viability. It can allow a correlation of the functionality of organelles of single cells with the development of cancer. Such organelles are mitochondria, nuclei, vacuoles, and reticulum. It allows for finding biochemical evidence of cancer prevention in single cells. This means improving the time of development of the disease and the time of treatment of the disease.
The development of investigation methods for single-cell level detection of viability stimulates the development of the cytotoxicity investigative field. The development of methods of single-cell microscopy is needed to improve the resolution and accuracy of investigations, and the influence of chemicals on single cells is important to study here.
The important for cytotoxicity is drug release. It is important to control the amount of drug that is released. This is performed with pH, temperature, and electric stimulation. The further development of methods of drug loading and bioimaging is important to decrease the cytotoxicity of carbon nanomaterials. The delivery of both biomedical contrast agents and therapeutic drugs using carbon nanomaterials is required.

Conclusions
In this review, we considered the issues of cytotoxicity of carbon nanomaterials. We performed detailed studies on the topics and revealed that such issues of cytotoxicity as solubility and cellular uptake could be overcome with the chemical functionalization of carbon nanomaterials. These can be performed by covalence, noncovalence, intercalation, substitution, and filling. The future work is dedicated to the chemical functionalization of carbon nanomaterials toward applications, in particular, in the biomedical field, nanoelectronics, bioelectronics, biosensors, light emission, biolight emission, bioelectronics light emission, thermoelectric power generation, solar cells, catalysts, spintronics, drug delivery, and bioimaging for plants. Our next review is dedicated to drug delivery, bioimaging of carbon nanomaterials, and biosensors for plants, microorganisms, animals, and humans. Acknowledgments: Authors would like to acknowledge the authors of all reviews, papers. Open Access Funding by the University of Vienna.

Conflicts of Interest:
Author may have the conflict of interest with Andrei Eliseev (Lomonosov Moscow State University) on the data from the papers of authors. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.