http://orcid.org/0000-0001-9392-5210
Abstract
In recent years, nanocellulose (NC) obtained by defibrating cellulose to the nanometer level has been developed, and its development for various applications, e.g. as an additive for cosmetics and as a component of structural elements, is progressing. However, because NC has unique physico-chemical properties that are not found in conventional nanomaterials, particularly when inhaled, there are concerns about unexpected effects on organisms. This review summarizes the progress of in vivo experiments on the effects of NC on the respiratory system by inhalation. In addition, this review will provide new insights into NC toxicity studies by comparing the effects of fibrous nanomaterials.
Introduction
Cellulose is the most abundant polymer in the world, being abundantly present in plant cell walls, and it is used as a raw material for paper. The carbohydrate comprises two repeating units of glucose linked with β-1,4 glycosidic bonds. Cellulose chains assemble into a crystal structure through intermolecular and intramolecular hydrogen bonds.
The most basic unit of cellulose is the cellulose microfibril, having a width of approximately 4 nm, and cellulose microfibrils can bundle into cellulose microfibril bundles with a width of 10–20 nm and serve as a basic unit in the cell wall. Furthermore, in microcrystalline cellulose, microfibril bundles form large bundles of up to several hundred nanometers and comprise a nest-like network (Eichhorn et al. Citation2010).
Microcrystalline cellulose has long been used in medical products such as wound tissue healing and dialysis membranes, in addition to food additives. The development of nanocellulose (NC) in which its fibers are dispersed into nanomaterials, which are 100 nm or smaller in at least one dimension, have progressed in recent years. Nanocellulose is classified according to the aspect ratio and crystallinity. Nanocellulose with a relatively large diameter and a low aspect ratio is called cellulose nanocrystal (CNC), and fibrous nanofiber with a high aspect ratio is called cellulose nanofiber (CNF). However, the categories are not strictly defined (Tayeb et al. Citation2018). CNC is a cellulose fiber cut by sulfuric acid treatment, and some of the surface hydroxyl groups are replaced by sulfate groups (Rajinipriya et al. Citation2018). Conversely, the hydroxyl groups on the surface of CNF are substituted without cutting the cellulose fiber.
In recent years, NC has been developed for various uses such as an additive for cosmetics and component of structural elements. In many cases, public institutions and companies cooperate regarding manufacturing methods, uses, and mass production under projects launched by the government. For example, ‘VTT’ in Finland, ‘USDA Forest Service’ and ‘Forest Products Laboratory’ in the US, ‘Arbora Nano’ and ‘FPInnovations’ in Canada, and ‘New Energy and Industrial Technology Development Organization (NEDO)’ in Japan are leading relevant projects.
Nanocellulose with characteristic physico-chemical properties is being developed by replacing hydroxyl groups on its surface with hydrophilic groups such as carboxyl groups and phosphate groups, chemical and mechanical defibration, and further substitution of modifying groups (Rol et al. Citation2019). Therefore, it is expected that producers and consumers will be exposed to NC in various forms in the future. Furthermore, because NC has characteristic properties not found in conventional nanomaterials, such as high viscosity and a variety of impurities derived from natural products, it may have unexpected biological effects.
Reports on the safety of NC are increasing, but long-term in vivo tests have only been initiated in recent years. Even among the aforementioned projects, only Japan has started to evaluate the effects of respiratory exposure to NC. In addition, reviews summarizing studies of NC have started to be reported, but few of these reviews provided safety evaluations.
Furthermore, the target materials in most previous studies of NC, including reports by the Food and Drug Administration (FDA) (FDA Citation1973) and FAO/WHO Joint Expert Committee on Food Additives (JECFA) (Elder Citation1986), have been limited to cellulose derivatives with hydrophilic groups on the fiber surface. However, it is known that research in the direction of adding various functions using hydrophobic substituents in NC is progressing (Rol et al. Citation2019). Therefore, it is considered that risk assessment based on the toxicity of NC will become necessary in the future.
In this report, we have compiled previous inhalation studies on the toxicity of cellulose-containing NC. Specifically, the results of studies mainly using animals were summarized for the purpose of evaluating changes observed in vivo. In vitro and immune response studies were referenced as needed. Furthermore, we compared the findings with those for multiwalled carbon nanotubes (MWCNTs), which represent a fibrous nanomaterial that has been extensively studied, to further clarify the toxicity of NC.
Literature review and discussion
Previously, various inhalation toxicity studies were conducted on 1,4-linked hexose-derived polysaccharides such as cellulose and cellulose derivatives, as well as substances obtained by nanosizing them. This section introduces the main contents of each study and lists the findings in tables. In addition, the results and discussion of previous studies were summarized and compared with those for MWCNTs.
Respiratory exposure studies are summarized in tables ( for intratracheal [i.t.] administration or pharyngeal aspiration, for inhalation exposure). Respiratory exposure methods include i.t. administration, pharyngeal aspiration, and inhalation exposure. In i.t. administration and pharyngeal aspiration, the test substance is dispersed in a liquid and injected into the lungs as a bolus in large quantities. In inhalation exposure, a powder or mist-like test substance is suspended in the air at a low concentration for chronic exposure. Because no inhalation exposure study of NC has been reported, the discussed studies involve only i.t. administration or pharyngeal aspiration. All test substances introduced in are non-NC.
Table 1. Toxicity of nanocellulose and related materials by intratracheal administration or pharyngeal aspiration.
Table 2. Inhalation toxicity of cellulose.
Severe dyspnea was observed immediately after the i.t. administration of unmodified CNF and carboxymethylated CNF at a dose of 6.4 mg/kg in female C57BL/6 mice (Hadrup et al. Citation2019). At a dose of 0.3 or 0.9 mg/kg, increased leukocyte counts in bronchoalveolar lavage fluid (BALF) and acute inflammatory marker (SAA3) levels in blood were immediately and dose-dependently observed after administration. The unmodified CNF administration group displayed a stronger response, but almost complete recovery was noted 28 days after administration. The authors noted that cellulose may cause a stronger acute response than MWCNTs based on the SAA3 marker results. The difference in the reaction between unmodified CNF and carboxymethylated CNF is presumed to be attributable to the number of −OH groups, and authors concluded that reduction of numbers of −OH groups by carboxymethylation reduces the toxicity of CNF.
Meanwhile, increases in interleukins (ILs) and tumor necrosis factor-α (TNF-α) expressions were observed in the lungs of female C57BL/6 mice after the pharyngeal inhalation of TEMPO-CNF at doses of 4 mg/kg or higher (Catalán et al. Citation2017). The reason why dose-dependent results were not obtained for some endpoints was presumed to be that the high-concentration samples were agglomerated and quickly removed from the respiratory tract.
Pharyngeal aspiration of BALB/c mice with 80 µg/mouse (∼4 mg/kg) CNC or CNF, resulted in changes in cytokine expression levels in BALF after 14 days (Park et al. Citation2018). Changes in clinical symptoms were not investigated. In the BALF test, there were differences in the expression of CD antigens on the leukocyte surface and ILs between the CNC and CNF groups. A characteristic change in gene expression was also observed in mice administered asbestos and single-walled carbon nanotubes, suggesting that even with the same NC, there may be differences in the mechanism of inflammation induction according to its physico-chemical properties.
In a separate study, male and female C57BL mice were administered CNC via pharyngeal aspiration at a dose of 40 µg/mouse (2 mg/kg) twice a week for 3 weeks and evaluated 3 months after the last dose (Shvedova et al. Citation2016). The expression of cytokines and oxidative stress response factors in BALF and the amount of collagen in the lungs was increased. Therefore, the authors speculated that the effect of CNC was stronger in female mice than in male mice. However, Shatkin and Oberdörster (Citation2016) noted that extremely high exposure levels may have affected the rate of CNC adsorption and the effects of a bolus, which is one of the characteristics of i.t. administration, onto the lungs. In addition, the same amount of CNC was administered to both sexes in the experiment, but in C57BL/6 mice, body weight differs by approximately 20% between the sexes in mice 8–11 weeks of age, indicating an effect of the dose difference in relation to body weight (Shatkin and Oberdörster Citation2016).
When powdered or gelled CNC was administered to female C57BL/6 mice via pharyngeal aspiration at 50–200 µg/mice (2.5–10 mg/kg), the number of cells and levels of lactate dehydrogenase (LDH), cytokines, and oxidative stress markers in BALF were increased on the next day. Regarding CNC gels in particular, marked increases in oxidative stress marker levels were observed (Yanamala et al. Citation2014). The changes in oxidative stress marker and ILs expression differed between mice administered powdered and gelled CNC, which was presumed to be due to size differences (powdered CNC is approximately 3-fold longer than gelled CNC).
Pharyngeal inhalation of four types of NC with different physico-chemical properties (such as zeta potential and contained metal impurities) at a dose of 2 mg/kg (40 µg/mice) in female C57BL mice led to increases in total cell numbers and inflammatory markers (ILs and TNF-α) expression in BALF on the next day. After 28 days, aggregate deposition, infiltration of inflammatory cells, and phagocytosis of foreign bodies by multinucleated giant cells were observed histopathologically (Ilves et al. Citation2018). This paper is one of the few reports in which ‘i.t. administration of NC and MWCNTs was performed in the same study using animals.’ The increase of inflammatory marker levels in BALF was weaker in the NC group than in the MWCNTs group. However, because follow-up after administration was only performed for up to 1 day (and not quantitatively evaluated histopathologically 1 month after administration), the effects of the bolus are included.
Bacteria-derived NC and carboxylated NC were intratracheally administered to female C57BL mice at 100 µg/animal (4 mg/kg) once a week for 4 weeks (16 mg/kg). Mice administered carboxylated NC were observed for 3 months, and mice administered other celluloses were observed for 6 months. Histopathological examination revealed mild diffuse degeneration of terminal bronchioles and infiltration of inflammatory cells. The findings in the bacterial NC group were comparable to those in the microcrystalline cellulose group, whereas the effects were slightly stronger in the carboxylated NC group. When the oxidative stress response and cell profile of BALF were evaluated 3 months after the final administration, no significant changes in marker levels compared with the control group levels were observed (Silva-Carvalho et al. Citation2019). The difference in the shape of the test substance was cited as the cause of the differences in the increase of B and T cell counts in BALF between the carboxylated NC and microcrystalline cellulose groups.
Chronic histopathological changes were not evaluated in studies involving respiratory exposure to NC because most reports had an observation period of less than 1 month (Ilves et al. Citation2018; Park et al. Citation2018; Hadrup et al. Citation2019). However, in a studies featuring 3 months of observation after a total dose of 12 mg/kg (Shvedova et al. Citation2016), 3–6 months of observation after a total dose of 16 mg/kg (Silva-Carvalho et al. Citation2019), and 1 month of observation after a dose of 2 mg/kg (Ilves et al. Citation2018), chronic inflammatory reactions were observed in the lungs.
Regarding reports of the i.t. administration of non-NC, long-term results (generally 2 months after administration) have been compared. In rodents, chronic histopathological changes were observed at doses of approximately 4 mg/kg and higher, and no results were significantly different from those of NC. From these findings, it was speculated that the dose required for NC to induce chronic inflammation in the lung was approximately 2–4 mg/kg or higher. In a previous report, there was no difference in BALF between the 60 mg/kg microcrystalline cellulose and vehicle groups 1 week after administration (Adamis et al. Citation1997). However, histopathological findings were not presented. The results of respiratory exposure studies with observation periods exceeding 6 months are limited, including one report on NC and three reports on non-NC.
As shown in previous reports, even low doses (less than 2 mg/kg) of cellulose, regardless of the presence of nanomaterials, immediately induced various biochemical inflammatory changes, such as changes in cytokine and gene expression and inflammatory cell exudation in BALF, after i.t. administration or pharyngeal aspiration (approximately 1 day to 4 weeks) (Ilves et al. Citation2018; Hadrup et al. Citation2019). At higher doses, the initial biochemical inflammatory response tends to disappear over time (approximately 2–3 months), whereas the histopathological changes worsen over time (Adamis et al. Citation1997), eventually leading to chronic changes including inflammation and fibrosis in the lungs (Shvedova et al. Citation2016; Silva-Carvalho et al. Citation2019).
In i.t. administration, the test substance is administered as a bolus, and thus, bolus-specific effects can occur (Shatkin and Oberdörster Citation2016). Therefore, the response immediately after administration is unlikely to reflect the effects of actual exposure, and biochemical inflammatory marker levels decreased over time after administration in all of the introduced studies. Hadrup et al. (Citation2019) calculated the lowest observed adverse effect level (LOAEL) in several i.t. administration studies of NC, but their study included a report that evaluated the effects for only 1 day after administration (Yanamala et al. Citation2014). The biochemical responses immediately after i.t. administration and pharyngeal aspiration may be qualitatively different from changes induced by long-term low-level inhalation exposure, which is the same as actual exposure, or chronic changes over time after i.t. administration and pharyngeal aspiration.
The mechanism of inflammation induced by NC has been evaluated. The findings illustrated that inflammation and cell death are basically induced by oxidative stress due to the increased production of reactive oxygen species (ROS) in mitochondria (Farcas et al. Citation2016; Shvedova et al. Citation2016). This is the same mechanism by which inflammation is induced after non-NC exposure in alveoli (Tátrai et al. Citation1996). A prior report also suggested that NC induces Th cell differentiation to Th1 cells through the induction of CD11 expression on antigen-presenting cells such as macrophages (Park et al. Citation2018). In addition, two intracellular inflammation-inducing pathways have been reported by in vitro studies using human-derived immune cells (Despres et al. Citation2019; Wang et al. Citation2019). In one pathway, the NLRP3 inflammasome is activated by cathepsin B released into the cytoplasm from lysosomes following NC-induced membrane damage, and the other pathway involves the upregulation of an IL-1β precursor through NF-κB activation due to increased ROS production in mitochondria. These signals ultimately induce IL-1β activation by the NLRP3 inflammasome. In recent years, research on the mechanism of inflammation induced by NC has progressed.
It is necessary to consider the presence of impurities and their effects when discussing respiratory exposure studies using cotton dust or cellulose insulation (Davis Citation1993; Morgan et al. Citation2004; Morgan Citation2006).
According to previous studies (Moon et al. Citation2011; Stefaniak et al. Citation2014), cellulose is not cleaved in macrophage lysosomes with a pH of 4.5–5.0. In vitro evaluation revealed NC persistence in alveolar macrophages cultured for 9 months on an artificial alveolar epithelial cell layer. It is considered that the strength of hydrogen bonding in cellulose fibers is related to the high persistence of cellulose. However, there are no reports evaluating the residual amount of NC in the lungs after a long period of in vivo exposure. In addition, the FDA and JECFA reports did not mention the toxicity of non-NC and its derivatives following inhalation exposure (FDA Citation1973; Elder, Citation1986).
Meanwhile, MWCNTs have been examined in many i.t. administration studies (Ema et al. Citation2012; Fujita et al. Citation2016). From these reports, it is known that i.t. administration at doses of 0.5–1 mg/kg or higher induces persistent inflammation in rodents (Kobayashi et al. Citation2017), and the no observed adverse effect level (NOAEL) and LOAEL for inhalation exposure were 0.1–0.25 and 0.2 mg/m3, respectively. Inflammation is caused by cell membrane damage induced by ROS (Chen et al. Citation2019), and signals are transmitted mainly through NF-κB, NLRP3 inflammasome, and TGF-β1 (Dong and Ma Citation2015). Conversely, MWCNTs differ from NC in that they alter the expression of Th2 cytokines and IgE.
In view of these findings, NC and MWCNTs similarly induce inflammation through oxidative stress. However, when comparing the results of i.t. administration in rodents, the dose required to induce persistent inflammation is extremely different. The dose is estimated to be approximately 2–4 mg/kg or higher for NC, compared with 0.5–1 mg/kg for MWCNTs (Kobayashi et al. Citation2017).
In a prior study, non-NC induced stronger responses of inflammatory markers such as PGE2 and leukotriene compared with MWCNTs and asbestos (Davis Citation1993). These markers are biosynthesized from arachidonic acid using phospholipids in the cell membrane as a raw material, suggesting that non-NC has strong irritant effects on the cell membrane. However, in reports in which animals were administered asbestos or MWCNTs at the same dose as NC, no marker clearly exhibited a stronger response to NC than to asbestos or MWCNTs (Munk et al. Citation2016; Ilves et al. Citation2018; Park et al. Citation2018), and similar results were obtained in an in vitro study that simulated the respiratory system (Clift et al. Citation2011).
Considering the previous reports on NC as references for future studies, although many findings were obtained, some study designs were insufficient for evaluation. There were three specific problems.
Many studies were performed using only one dose, and dose dependency, which is important for toxicity evaluation, was unknown.
The observation period was short in many studies, and only the bolus-specific response was evaluated.
The toxicological significance of the observed changes was not clarified because only blood and gene expression analyses were performed, and clinical signs and histopathological evaluations were not performed.
Influence of differences in physico-chemical properties
NC has different effects on organisms depending on its physico-chemical properties, and in particular, width, length, and aspect ratio are considered important indicators for evaluation.
Concerning previous reports, it was reported that changes in inflammatory marker levels in BALF and blood differ depending on the surface modification (Ilves et al. Citation2018) and length (Park et al. Citation2018) of NC. Specifically, the cause of the different findings was attributed to the aggregability of the fibers due to their different lengths (Yanamala et al. Citation2014) and interactions with the cell membrane and intracellular molecules due to the different numbers of OH groups on the fiber surface (Hadrup et al. Citation2019). In addition, several in vitro experiments using lung-derived cells have been reported, and differences in lignin content (Yanamala et al. Citation2016), aggregability due to differences in NC dispersion (Menas et al. Citation2017), interfiber interactions in cells due to fiber length (Endes et al. Citation2015), and surface modification (Jimenez et al. Citation2017) are mentioned as the causes of different findings. However, in some cases, changes observed in vitro were not reproduced in vivo (Ilves et al. Citation2018), and the toxicological significance of changes in gene expression and inflammatory marker levels is unknown.
The response of immune cells is also affected by the length (Wang et al. Citation2019), hydrophilicity (Jelinková et al. Citation2002), and zeta potential (Despres et al. Citation2019) of NC. In other in vitro evaluations, reports found that surface charge affects the uptake of NC into human embryonic kidney-derived cells (Mahmoud et al. Citation2010) and intracellular inflammatory signal transduction pathways activated by NC differ because of differences in surface modification (cationic or anionic) groups in the human monocytes or mouse macrophages (Lopes et al. Citation2017).
Even for MWCNTs, different physico-chemical properties have different effects on organisms. It is known that higher aspect ratio of MWCNTs increases the prominence of inflammation and the retention in the organisms (Kobayashi et al. Citation2017), and very short MWCNTs increase the clearance of the lungs (Muller et al. Citation2009). In addition, MWCNTs of certain widths exhibit strong inflammation and carcinogenicity in the mesothelium (Toyokuni Citation2013). Therefore, as mentioned in previous reports, the difference in the effect on organisms due to differences in properties is a characteristic common of fibrous nanomaterials (de Lima et al. Citation2012).
Contrarily, rodents are often used in studies, but no data clearly revealed differences due to strain or sex within the same study. Moreover, even among different studies using the same cellulose derivative including non-NC, no data to date have revealed differences linked to strain or sex.
Differences in the effects of cellulose derivatives on organisms due to differences in physico-chemical properties such as surface modification, length, and aggregability are commonly recognized for in vivo and in vitro studies. The main cause is the action of the test substance to the cell, and the reaction in cells differs depending on the properties of cellulose. Therefore, in the safety evaluation of NC, it is considered important to evaluate the properties in detail in advance and clarify both the characteristics of the properties and the study results.
Conclusion and future study
In this review, we discussed knowledge and problems related to the previous inhalation toxicity studies of NC and identified important points for future evaluations.
There were three major findings obtained from previous inhalation toxicity studies of NC:
Differences in the reaction of organisms due to differences in the physico-chemical properties of NC
Cell damage is induced by a mechanism mediated by oxidative stress
No serious changes in the effects of NC were observed from MWCNTs in the same study or at the same dose
Conversely, the evaluation was insufficient in some studies. It is necessary to pay attention to the following four points for NC inhalation assessment.
Because the influence on organisms depends on the physico-chemical properties of NC, the properties of the test substance should be evaluated in detail. Specifically, it is necessary to classify the features by evaluating indicators such as length, width, aspect ratio, surface modification, zeta potential, viscosity, pH, bending rate, and rigidity. Compared with the reports on MWCNTs, many reports on NC did not describe the length or width. NC is derived from natural products, and in particular, it is believed that measurement is often difficult because of the entanglement of fibers in CNF. However, considering the results of MWCNTs studies, length and width are particularly important indicators of the impact on organisms.
NC may easily aggregate due to its physico-chemical properties (de Lima et al. Citation2012; Yanamala et al. Citation2014), and it can be contaminated with bacteria and fungi. It is important to ensure the dispersibility of NC using a device with high stirring force and confirm sterilization. This is because large dust particles and endotoxins can induce emphysema, and these effects cannot be distinguished from those of the test substance (Milton et al. Citation1990), and the aggregation state of the test sample affects the results (Catalán et al. Citation2017). Common sterilization methods include filtration, γ-ray irradiation, and autoclaving. However, NC is a fiber that cannot be filtered, and γ-ray irradiation causes the main chain to break or crosslink (Glegg and Kertesz Citation1957). In most previous studies, sterilization using an autoclave was selected, but depending on the properties of NC, heat resistance may not be high (Noguchi et al. Citation2017). Therefore, the addition of a preservative also should be considered depending on the properties of NC.
Animal experiments should be performed as much as possible because in vitro results may not be reproducible in vivo (Ilves et al. Citation2018). In vivo studies should be conducted using multiple doses to confirm the dose dependency of the test substance. In the case of i.t. administration in particular, a study period should be as long as possible. Biochemical inflammatory marker levels are temporarily increased immediately after i.t. administration before decreasing, whereas histopathological changes worsen over time. Therefore, markers in blood and BALF and histopathological changes should be measured to comprehensively evaluate the effects on the organisms. It is important to select the indicators for evaluation. In addition to ILs, LDH, and TNF-α described in many reports, some reports used BALF MIP-1α and blood SAA3 protein.
Though it is presumed that the gastrointestinal mucosal epithelium does not absorb NC in terms of size (Kimura et al. Citation1994), evaluation of biokinetics is considered extremely important for confirming the uptake (transition from epithelial cells to blood and lymph nodes) and persistence in the alveoli of NC. In the biokinetics evaluation, some oral administration studies used non-NC derivatives (Elder Citation1986; Bar et al. Citation1995) or MWCNT (Deng et al. Citation2007) in which a modifying group was linked to an isotope such as 14C, but it is suggested that the modifying group can detach from the cellulose chain for some reason (Bar et al. Citation1995). In particular, NC has a large surface area relative to its weight. Thus, more modifying groups are detached and absorbed into the body, and the proportions of isotopes detected in urine, organs, and exhaled breath may be higher than those for non-NC. For this reason, attention should be paid to obtained values. To eliminate such a possibility, it is effective to use an isotope labeled in cellulose itself, but high cost is a problem. Methods in which a fluorescent or dye linked to a modifying group are simple, inexpensive, and rapid. In these methods, because the surface modification and physico-chemical properties of NC differ from those of the original test substance, it is necessary to confirm whether there is no change in the reaction with the cell membrane or intracellular molecules as well as the aggregation properties. Although the effect of differences in the length of NC on its uptake into lung-derived cells using the fluorescent dye rhodamine B has been reported (Endes et al. Citation2015), no quantitative evaluation was described. Because NC has low absorbance and low thermal stability, it is difficult to adapt previous substance detection methods to in vivo analysis, and thus, it is necessary to select or develop an appropriate method for quantitative evaluation.
This review focused on papers on pulmonary toxicity of NC. No paper found that inhaled NC induces tumors in in vivo study (Catalán et al. Citation2017), but fibrous nanomaterials, MWCNTs, and asbestos, were reported to induce mesothelial injury and tumorigenesis (Toyokuni, Citation2013). The induced reaction is considered to greatly differ between cellulose and asbestos in vivo (Pinto et al. Citation2016). However, because of the dearth of reports, it is necessary to evaluate tumor induction by NC through routes other than the actual exposure route, such as i.p. administration, as performed for other fibrous nanomaterials.
In addition, there are few NC reports regarding carcinogenicity and immunotoxicity though MWCNTs (especially MWCNT-7 classified as 2B by the International Agency for Research on Cancer) exhibited carcinogenicity and immunotoxicity in many reports (Takagi et al. Citation2012). It is common to judge genotoxicity from the results of multiple tests. Regarding the combinations of tests used (battery tests), in addition to the Ames test, in vivo micronucleus tests, other in vivo tests, and in vitro genotoxicity tests using mammalian cells are defined in the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use S2 guidelines (van der Laan and DeGeorge Citation2013). However, there are reports of skepticism in applying the Ames test to fibrous nanomaterials, especially in terms of cell permeability (Ema et al. Citation2012; Toyokuni Citation2013; Pitkanen et al. Citation2014), so it is considered important to select or develop an appropriate study design. One of the characteristic results of genotoxicity studies is a non-dose-dependent increase in the numbers of comet-positive cells and micronuclei (Catalán et al. Citation2017; Ventura et al. Citation2018), and both papers noted the possibility that high-concentration NC was aggregated and eliminated early in the body or not taken up into cells. The same phenomenon has been observed for surface-modified CNTs (Ventura et al. Citation2018). There is no report stating that such a phenomenon was observed in the genotoxicity studies alone. However, because NC induces oxidative stress in cells, it is considered that chromosomal abnormalities due to adducts and DNA damage reflect the cytotoxicity of NC with the highest sensitivity. Therefore, the dose or concentration is important in studies in which NC is directly taken up by cells.
In conclusion, although NC has many promising uses as an additive or structural element, the substance has health concerns because it is a type of nanomaterial. Whether this new material will be widely used in society in the future depends on evaluations of its toxicity using reliable study designs and assessments of the risk together with the results of exposure assessment studies.
Disclosure statement
The authors declare no conflicts of interest.
Additional information
Funding
References
- Adamis Z, Tátrai E, Honma K, Ungváry G. 1997. In vitro and in vivo assessment of the pulmonary toxicity of cellulose. J Appl Toxicol. 17(2):137–141.
- Bar A, Van Ommen B, Timonen M. 1995. Metabolic disposition in rats of regular and enzymatically depolymerized sodium carboxymethylcellulose. Food Chem Toxicol. 33(11):901–907.
- Catalán J, Rydman E, Aimonen K, Hannukainen KS, Suhonen S, Vanhala E, Moreno C, Meyer V, Perez DD, Sneck A, Forsström U, et al. 2017. Genotoxic and inflammatory effects of nanofibrillated cellulose in murine lungs. Mutagenesis. 32(1):23–31.
- Chen M, Sun Y, Liang J, Zeng G, Li Z, Tang L, Zhu Y, Jiang D, Song B. 2019. Understanding the influence of carbon nanomaterials on microbial communities. Environ Int. 126:690–698.
- Clift MJD, Foster EJ, Vanhecke D, Studer D, Wick P, Gehr P, Rothen-Rutishauser B, Weder C. 2011. Investigating the interaction of cellulose nanofibers derived from cotton with a sophisticated 3D human lung cell coculture. Biomacromolecules. 12(10):3666–3673.
- Cullen RT, Searl A, Miller BG, Davis JM, Jones AD. 2000. Pulmonary and intraperitoneal inflammation induced by cellulose fibres. J Appl Toxicol. 20(1):49–60.
- Davis J. 1993. The need for standardising testing procedure for all products capable of liberating respirable fibers: the example of materials based on cellulose. Br J Ind Med. 50(2):187–190.
- de Lima R, Oliveira FL, Rodrigues MC, Abreu BM, Yamawaki PC, Vieira IJ, Teixeira EM, Corrêa AC, Caparelli MLH, Fernandes FL. 2012. Evaluation of the genotoxicity of cellulose nanofibers. Int J Nanomed. 7:3555–3565.
- Deng X, Jia G, Wang H, Sun H, Wang X, Yang S, Wang T, Liu Y. 2007. Translocation and fate of multi-walled carbon nanotubes in vivo. Carbon. 45(7):1419–1424.
- Despres HW, Sabra A, Anderson P, Hemraz UD, Boluk Y, Sunasee R, Ckless K. 2019. Mechanisms of the immune response cause by cationic and anionic surface functionalized cellulose nanocrystals using cell-based assays. Toxicol in Vitro. 55:124–133.
- Dong J, Ma Q. 2015. Advances in mechanisms and signaling pathways of carbon nanotube toxicity. Nanotoxicology. 9(5):658–676.
- Eichhorn SJ, Dufresne A, Aranguren M, Marcovich NE, Capadona JR, Rowan SJ, Weder C, Thielemans W, Roman M, Renneckar S, et al. 2010. Review: current international research into cellulose nanofibres and nanocomposites. J Mater Sci. 45(1):1–33.
- Elder R. 1986. Final report on the safety assessment of hydroxyethylcellulose, hydroxypropylcellulose, methylcellulose, hydroxypropyl methylcellulose and cellulose gum. J Am Col Toxicol. 5:1–59.
- Ellakkani MA, Alarie Y, Weyel D, Karol MH. 1987. Chronic pulmonary effects in guinea pigs from prolonged inhalation of cotton dust. Toxicol Appl Pharmacol. 88(3):354–369.
- Ellakkani MA, Alarie YC, Weyel DA, Mazumdar S, Karol MH. 1984. Pulmonary reactions to inhaled cotton dust: an animal model for byssinosis. Toxicol Appl Pharmacol. 74(2):267–284.
- Ema M, Imamura T, Suzuki H, Kobayashi N, Naya M, Nakanishi J. 2012. Evaluation of genotoxicity of multi-walled carbon nanotubes in a battery of in vitro and in vivo assays. Regul Toxicol Pharmacol. 63(2):188–195.
- Endes C, Mueller S, Kinnear C, Vanhecke D, Foster EJ, Petri-Fink A, Weder C, Clift MJ, Rothen-Rutishauser B. 2015. Fate of cellulose nanocrystal aerosols deposited on the lung cell surface in vitro. Biomacromolecules. 16(4):1267–1275.
- Farcas MT, Kisin ER, Menas AL, Gutkin DW, Star A, Reiner RS, Yanamala N, Savolainen K, Shvedova AA. 2016. Pulmonary exposure to cellulose nanocrystals caused deleterious effects to reproductive system in male mice. J Toxicol Environ Health Part A. 79(21):984–997.
- Food and Drug Administration 1973. Select Committee on GRAS Substances. [accessed 13 May 2019]. https://www.accessdata.fda.gov/scripts/fdcc/?set=SCOGS&sort=Sortsubstance&order=ASC&startrow=1&type=basic&search=cellulose.
- Fujita K, Fukuda M, Endoh S, Maru J, Kato H, Nakamura A, Shinohara N, Uchino K, Honda K. 2016. Pulmonary and pleural inflammation after intratracheal instillation of short single-walled and multi-walled carbon nanotubes. Toxicol Lett. 257:23–37.
- Glegg RE, Kertesz ZI. 1957. Effect of gamma‐radiation on cellulose. J Polym Sci. 26(114):289–297.
- Hadrup N, Knudsen KB, Berthing T, Wolff H, Bengtson S, Kofoed C, Espersen R, Højgaard C, Winther JR, Willemoës M, et al. 2019. Pulmonary effects of nanofibrillated celluloses in mice suggest that carboxylation lowers the inflammatory and acute phase responses. Environ Toxicol Pharmacol. 66:116–125.
- Ilves M, Vilske S, Aimonen K, Lindberg HK, Pesonen S, Wedin I, Nuopponen M, Vanhala E, Højgaard C, Winther JR, et al. 2018. Nanofibrillated cellulose causes acute pulmonary inflammation that subsides within a month. Nanotoxicology. 12(7):729–746.
- Jelinková M, Briestenský J, Santar I, Ríhová B. 2002. In vitro and in vivo immunomodulatory effects of microdispersed oxidized cellulose. Int Immunopharmacol. 2(10):1429–1441.
- Jimenez AS, Jaramillo F, Hemraz UD, Boluk Y, Ckless K, Sunasee R. 2017. Effect of surface organic coatings of cellulose nanocrystals on the viability of mammalian cell lines. Nanotechnol Sci Appl. 10:123–136.
- Kimura T, Sudo K, Kanzaki Y, Miki K, Takeichi Y, Kurosaki Y, Nakayama T. 1994. Drug absorption from large intestine: physicochemical factors governing drug absorption. Biol Pharm Bull. 17(2):327–333.
- Kobayashi N, Izumi H, Morimoto Y. 2017. Review of toxicity studies of carbon nanotubes. J Occup Health. 59(5):394–407.
- Lopes VR, Sanchez-Martinez C, Strømme M, Ferraz N. 2017. In vitro biological responses to nanofibrillated cellulose by human dermal, lung and immune cells: surface chemistry aspect. Part Fibre Toxicol. 14(1):1.
- Mahmoud KA, Mena JA, Male KB, Hrapovic S, Kamen A, Luong JH. 2010. Effect of surface charge on the cellular uptake and cytotoxicity of fluorescent labeled cellulose nanocrystals. ACS Appl Mater Interf. 2(10):2924–2932.
- Menas AL, Yanamala N, Farcas MT, Russo M, Friend S, Fournier PM, Star A, Iavicoli I, Shurin GV, Vogel UB, et al. 2017. Fibrillar vs crystalline nanocellulose pulmonary epithelial cell responses: cytotoxicity or inflammation?. Chemosphere. 171:671–680.
- Milton DK, Godleski JJ, Feldman HA, Greaves IA. 1990. Toxicity of intratracheally instilled cotton dust, cellulose, and endotoxin. Am Rev Respir Dis. 142(1):184–192.
- Moon RJ, Martini A, Nairn J, Simonsen J, Young-Blood J. 2011. Cellulose nanomaterials review: structure, properties and nanocomposites. Chem Soc Rev. 40(7):3941–3994.
- Morgan DL. 2006. NTP Toxicity Study Report on the atmospheric characterization, particle size, chemical composition, and workplace exposure assessment of cellulose insulation (celluloseins). Toxic Rep Ser. 74:1–62.
- Morgan DL, Su YF, Dill JA, Turnier JC, Westerberg RB, Smith CS. 2004. Chemical and physical characteristics of cellulose insulation particulates, and evaluation of potential acute pulmonary toxicity. Am J Ind Med. 46(6):554–569.
- Muhle H, Ernstb H, Bellmann B. 1997. Investigation of the durability of cellulose fibres in rat lungs. Ann Occup Hyg. 41:184–188.
- Muller J, Delos M, Panin N, Rabolli V, Huaux F, Lison D. 2009. Absence of carcinogenic response to multiwall carbon nanotubes in a 2-year bioassay in the peritoneal cavity of the rat. Toxicol Sci. 110(2):442–448.
- Munk M, Camargo LS, Quintão CC, Silva SR, Souza ED, Raposo NR, Marconcini JM, Jorio A, Ladeira LO, Brandão HM. 2016. Biocompatibility assessment of fibrous nanomaterials in mammalian embryos. Nanomedicine. 12(5):1151–1159.
- Nagato LK, Lourenço MG, Cadete RA, Leite-Júnior JH, Koatz VL, Rocco PR, Faffe DS, Zin WA. 2008. Microcrystalline cellulose induces time-dependent lung functional and inflammatory changes. Respir Physiol Neurobiol. 164(3):331–337.
- Noguchi Y, Homma I, Matsubara Y. 2017. Complete nanofibrillation of cellulose prepared by phosphorylation. Cellulose. 24(3):1295–1305.
- Park EJ, Khaliullin TO, Shurin MR, Kisin ER, Yanamala N, Fadeel B, Chang J, Shvedova AA. 2018. Fibrous nanocellulose, crystalline nanocellulose, carbon nanotubes, and crocidolite asbestos elicit disparate immune responses upon pharyngeal aspiration in mice. J Immunotoxicol. 15(1):12–23.
- Pinto FCM, De-Oliveira A, De-Carvalho RR, Gomes-Carneiro MR, Coelho DR, Lima SVC, Paumgartten FJR, Aguiar J. 2016. Acute toxicity, cytotoxicity, genotoxicity and antigenotoxic effects of a cellulosic exopolysaccharide obtained from sugarcane molasses. Carbohydr Polym. 137:556–560.
- Pitkanen M, Kangas H, Laitinen O, Sneck A, Lahtinen P, Peresin MS, Niinimäki J. 2014. Characteristics and safety of nano-sized cellulose fibrils. Cellulose. 21(6):3871–3886.
- Rajinipriya M, Nagalakshmaiah M, Robert M, Elkoun S. 2018. Importance of agricultural and industrial waste in the field of nanocellulose and recent industrial developments of wood based nanocellulose: a review. ACS Sustainable Chem Eng. 6(3):2807–2828.
- Rol F, Belgacem MN, Gandini A, Bras J. 2019. Recent advances in surface-modified cellulose nanofibrils. Prog Polym Sci. 88:241–264.
- Shatkin JA, Oberdörster G. 2016. Comment on Shvedova et al. (2016), “gender differences in murine pulmonary responses elicited by cellulose nanocrystals". Part Fibre Toxicol. 13(1):59.
- Shvedova AA, Kisin ER, Yanamala N, Farcas MT, Menas AL, Williams A, Fournier PM, Reynolds JS, Gutkin DW, Star A, et al. 2016. Gender differences in murine pulmonary responses elicited by cellulose nanocrystals. Part Fibre Toxicol. 13(1):28.
- Silva-Carvalho R, Silva JP, Ferreirinha P, Leitão AF, Andrade FK, Gil da CRM, Cristelo C, Rosa MF, Vilanova M, Gama FM. 2019. Inhalation of bacterial cellulose nanofibrils triggers an inflammatory response and changes lung tissue morphology of mice. Toxicol Res. 35(1):45–63.
- Stefaniak AB, Seehra MS, Fix NR, Leonard SS. 2014. Lung biodurability and free radical production of cellulose nanomaterials. Inhal Toxicol. 26(12):733–749.
- Takagi A, Hirose A, Futakuchi M, Tsuda H, Kanno J. 2012. Dose-dependent mesothelioma induction by intraperitoneal administration of multi-wall carbon nanotubes in p53 heterozygous mice. Cancer Sci. 103(8):1440–1444.
- Tátrai E, Brozik M, Adamis Z, Merétey K, Ungváry G. 1996. In vivo pulmonary toxicity of cellulose in rats. J Appl Toxicol. 16(2):129–135.
- Tayeb AH, Amini E, Ghasemi S, Tajvidi M. 2018. Cellulose nanomaterials—binding properties and applications: a review. Molecules. 23(10):2684.
- Toyokuni S. 2013. Genotoxicity and carcinogenicity risk of carbon nanotubes. Adv Drug Deliv Rev. 65(15):2098–2110.
- van der Laan JW, DeGeorge J Jr. 2013. Global approach in safety testing: ICH guidelines explained. New York (NY): Springer. ISBN 978-1-4614-5950-7.
- Ventura C, Lourenço AF, Sousa-Uva A, Ferreira PJT, Silva MJ. 2018. Evaluating the genotoxicity of cellulose nanofibrils in a co-culture of human lung epithelial cells and monocyte-derived macrophages. Toxicol Lett. 291:178–183.
- Wang X, Chang CH, Jiang J, Liu Q, Liao YP, Lu J, Li L, Liu X, Kim J, Ahmed A, et al. 2019. The crystallinity and aspect ratio of cellulose nanomaterials determine their pro-inflammatory and immune adjuvant effects in vitro and in vivo. Small. 15(42):e1901642.
- Yanamala N, Farcas MT, Hatfield MK, Kisin ER, Kagan VE, Geraci CL, Shvedova AA. 2014. In Vivo evaluation of the pulmonary toxicity of cellulose nanocrystals: a renewable and sustainable nanomaterial of the future. ACS Sustain Chem Eng. 2(7):1691–1698.
- Yanamala N, Kisin ER, Menas AL, Farcas MT, Khaliullin TO, Vogel UB, Shurin GV, Schwegler-Berry D, Fournier PM, Star A, et al. 2016. In vitro toxicity evaluation of lignin-(un)coated cellulose based nanomaterials on human A549 and THP-1 cells. Biomacromolecules. 17(11):3464–3473.