Challenges for assessing carbon nanomaterial toxicity to the skin

Abstract

This manuscript reviews a number of issues that must be dealt with to assess carbon nanomaterial interactions with the skin in the context of potential toxicity. The potential pathway for dermal absorption of carbon nanomaterials is discussed. The few existing studies assessing carbon nanomaterial toxicity to skin are reviewed. This paper addresses potential confounding factors in dealing with the experimental design of nanomaterial toxicity studies and their interpretation. Certain standard cytotoxicity assays that are well suited to assess chemical toxicity may generate conflicting results when carbon materials are assessed. This was demonstrated in an experimental study comparing carbon effects on human keratinocyte cytotoxicity assessed by transmission electron microscopy, neutral red and MTT cell viability assays, as well as irritation assessed by release of the cytokine IL-8. Four sources of carbon black particles were assessed. Conflicting results were obtained across all cytotoxicity endpoints potentially secondary to the adsorbing properties of carbon interfering with viability markers in the assay systems. These data suggest that a single cytotoxicity assay should not be relied upon in assessing carbon nanomaterial toxicity and that carbon black may not be optimal control particles for assessing nanomaterial toxicity in epidermal cell culture systems due to the wide range of responses seen between the carbon black varieties.

Introduction

Nanomaterial toxicology remains a challenge with regard to conducting a comprehensive safety evaluation of nanomaterials. As large scale production of nanomaterials increases, so does the threat of adverse health effects in humans and the potential for environmental damage [1], [2]. An important route of exposure is by the skin, which could lead to skin cancer, skin sensitization, skin irritation, or produce systemic effects after absorption. Little information is available on nanomaterial toxicology, and as the field grows there is an additional need to standardize and categorize the nanomaterials so that results can be compared across studies and data can be obtained for risk assessment. However, to determine and understand the toxic effects of nanomaterials, strategies and interpretation of the data must be done correctly and assumptions taken into consideration. The purpose of this paper is to provide a review along with original research data to explain the potential problems working with carbon and carbon nanomaterials in skin.

Skin is unique because it is a potential route of both occupational and/or environmental exposure to nanoparticles and also provides an environment within the avascular epidermis where particles could potentially lodge and not be susceptible to removal by phagocytosis. The skin or integument is the largest organ of the body and can serve as one of the principal portals of entry by which nanomaterials can enter the body. It also has a relative large surface area for exposure. Skin is considered to be the barrier between the well-regulated “milieu interieur” and the outside environment. The structure of skin is heterogeneous, yet is functionally integrated to yield a dynamic organ that has a myriad of biological tasks far beyond its role as a barrier to the external environment.

Information is lacking in describing the human health and environmental implications of manufactured nanomaterials. Data is needed for cutaneous hazard analysis after topical exposure that could occur during the engineering or manufacturing process, application, or waste management of these materials. Currently, there is a lack of information on whether nanomaterials or nanoparticles can be absorbed across the skin’s stratum corneum barrier or whether systemically administered particles can accumulate in dermal tissue. The tendency for carbon nanomaterials to traverse the skin is a primary determinant of its dermatotoxic potential. That is, the nanomaterials or nanoparticles must penetrate the uppermost stratum corneum layer in order to gain entrance to the viable epidermis and exert toxicity in the lower cell layers. The quantitative prediction of the rate and extent of percutaneous penetration (into skin) and absorption (through skin) of topically applied nanomaterials is complicated because the processes which drive nanoparticles into skin may be different than those which govern chemicals.

Major problems now exist in assessing skin absorption and skin toxicity of nanomaterials, the first being how to actually conduct the experiments. For instance, there is great difficulty in obtaining the quantity and quality (characterization) of nanomaterials to conduct appropriate in vivo studies. Due to their small size it would be difficult to determine their location in skin or within the systemic circulation. They would be diluted out throughout the entire body if systemically absorbed, or may become lodged in major organs, further reducing the ability to detect them. The need for new detection techniques to localize within cells or tissues is needed, especially if one wants to quantitate the amount present. Access to radiolabeled or fluorescently tagged nanomaterials would make localization within tissues and cells easier. However, this could alter the characteristics of the carbon nanomaterial in question. Therefore, it is important that full characterization of the nanomaterials be conducted and described in order to facilitate comparisons between studies. To study the direct interactions with cells, in vitro models may be more appropriate to estimate the in vivo starting dose for toxicity and absorption and to gain some insight as to the mechanisms of toxicity.

Carbon nanomaterial absorption through the skin may not be similar to chemical absorption because these materials have been engineered differently, and can be of different sizes (single-walled carbon nanotubes, multi-walled carbon nanotubes, and fullerenes (buckyballs)), and are not homogeneous like chemicals. Also, the surface engineering of nanoparticles can make it more compatible with the biological milieu. The physiochemical properties of carbon nanomaterials may affect its dermatotoxic potential or its ability to penetrate the skin and elicit a toxicological response. This could also affect their rate of absorption, direct cellular interactions, or uptake in skin cells. In addition, nanomaterials occur in a range of surface modifications, sizes, shapes, and compositions thereby making standardization a challenge. Also, all of these properties will effect how they traverse the stratum corneum and interact with underlying cells.

Anatomically, nanomaterial absorption may occur through several routes: the majority of lipid soluble particles may move through the intercellular lipid pathway between the stratum corneum cells (intercellular), through the cells (transcellular), or through the hair follicle or sweat ducts (transappendageal), Fig. 1[3].

Carbon nanomaterials first have to penetrate the stratum corneum layer, the outermost layer of the epidermis that consists of several layers of completely keratinized dead cells. These proteinaceous keratinocytes are embedded in an extracellular lipid matrix composed primarily of sterols, other neutral lipids and ceramides. The intercellular lipids (40–50% ceramides, 20–27% cholesterol, 10% cholesterol esters, and 10–12% free fatty acids [4], [5], and 1–2% triglycerides [6]) are derived from the lamellar granules of the stratum granulosum layer to form the complex barrier. When carbon nanomaterials are applied topically to the skin, they would have to penetrate this torturous lipid pathway that prevents both the penetration of substances from the environment and the insensible loss of body water by surface evaporation. They must be able to penetrate this complex barrier and traverse through the viable epidermal layers and epidermal–dermal junction (basement membrane) to gain access to the capillaries within the papillary layer of the dermis in order to get into the systemic circulation. Of course, disease or occupational conditions that cause damage to the stratum corneum barrier (e.g. abrasion, solvent exposure) may abrogate these protective functions.

If particles penetrate the stratum corneum cells and become lodged within the viable epidermal cell layers of the skin, they may enter the keratinocytes directly or trigger the production of pro-inflammatory cytokines or initiate other sequela. If carbon nanomaterials trigger the immune system, the manifestations seen will be dependent upon the type of immunologic response elicited (e.g. cellular versus humoral, acute hypersensitivity, etc.). It should be stressed that immune cells (e.g. Langerhans cells, lymphocytes, mast cells) may modulate the reaction or the keratinocytes themselves may initiate this response. In fact, keratinocytes were once thought to produce only keratin and mucopolysaccharides, but recent studies have shown that they can produce growth factors, chemotactic factors, and adhesion molecules. Keratinocytes may act as the key immunocyte in the pathophysiology of allergic contact and irritant contact dermatitis. This is especially relevant to nanomaterials that may be poorly absorbed. There are several potential pathways through which skin contact with carbon nanomaterials may trigger immunologic manifestations. Direct irritation of keratinocytes by these materials may also initiate this cytokine cascade without involvement of the immune system, blurring the distinction both between direct and indirect cutaneous irritation, as well as between local and systemic toxicity.

Fullerenes (molecular structures made up of 60 or more carbon atoms) having a size less than 100 nm may have beneficial effects and numerous studies have also indicated that specific functionalized fullerenes may be therapeutically useful in the treatment of a number of diseases. However, in reviewing the fullerene skin literature, many studies have stressed the toxicologic response. There is a little information regarding skin toxicity but no literature pertaining to the absorption of carbon nanomaterials through human skin. The dermal toxicity of topical administration of 200 μg of fullerene to mouse skin over 72 h found no effect on either DNA synthesis or ornithine decarboxylase activity. Repeated application of fullerenes to mouse skin after initiation with dimethlybenzanthracene (DMBA) for 24 weeks did not result in skin tumor formation, but promotion was observed with 12-0-tetradecanoylphorbol-13-acetate (TPA) resulting in benign skin tumors [7].

In vitro studies using ¹⁴C-labelled underivatized C₆₀ exposed to immortalized human keratinocytes depicted cellular incorporation of the label uptake at various times. By 6 h, approximately 50% of the radiolabeled was taken up but it was unclear whether particles actually entered the cell or were associated with the cell surface. These investigators also found no effect of C₆₀ on the proliferation of immortalized human keratinocytes or fibroblasts [8]. Water-soluble fullerenes were assessed in human carcinoma cells and dermal fibroblasts for their toxicity. Studies have shown that water-soluble functional groups on the surface of fullerenes can decrease the toxicity of pristine C₆₀. This least derivatized and most aggregated form of C₆₀ was more toxic than the highly soluble derivatives such as C₃, ${\mathrm{Na}}_{2''–''3}^{+}$[C₆₀O_7–9(OH)_12–15]^(2–3)-, and C₆₀(OH)₂₄[9]. There have been conflicting reports as to the potential toxicity of fullerenes such as C₆₀. While C₆₀ itself has essentially no solubility in water, it has been shown to aggregate with either organic solvent inclusion or partial hydrolysis to create water-soluble species n-C₆₀. These aggregates have exceptionally low mobility in aqueous solutions but have been proposed to have high cellular toxicity.

CNT are either single-walled (SWCNT) or multi-walled (MWCNT) and have diameters that range from a few to hundreds of nanometers, while their length can be up to a few micrometers. One of the principal attributes of CNT and other nanoparticles that makes their development such a breakthrough is their unique catalytic properties. For example, pure carbon buckytubes are referenced as being capable of reacting with many organic compounds due to their carbon chemistry base. Modifications including end-of-tube (e.g. via reaction with carboxyl groups at open tip ends of carbon nanotubes) or sidewall derivatization would modify their physical properties and alter solubility or dispersion. CNT have also been engineered to be more pharmaceutically compatible. Biomedical applications of nanotubes have been explored for the delivery of genes, drugs, and antigens [10]. When these attributes are deliberately modified, useful products and delivery of effective therapeutic may be obtained.

There are several reports of carbon skin irritation in humans (carbon fiber dermatitis and hyperkeratosis) that suggest that nanoparticles may gain entry into the viable epidermis after topical exposure. Some studies have shown cellular toxicity when unrefined SWCNT were exposed to immortalized non-tumorigenic human epidermal HaCaT cells for 18 h [11] but did not evaluate for inflammatory markers. Our group has demonstrated significant differences in the toxicological response between immortalized versus primary keratinocytes [12]. Therefore, it would be interesting to see if normal human epidermal keratinocytes responded differently. Gene expression profiling conducted on human epidermal keratinocytes exposed to SWCNT (1.0 mg/ml) has shown a profile similar to that of alpha-quartz or silica, considered to be the main cause of silicosis in humans [13].

In addition to skin toxicity, the use of SWCNT has been explored for drug delivery. Drug delivery can be enhanced by many types of chemical vehicles, including lipids, peptides and polyethylene glycol derivatives (PEG). Carbon nanotubes can be filled with DNA or peptide molecules and can serve as a potential delivery system in gene or peptide delivery [14]. Strategies using SWCNT and SWCNT-streptavidin conjugates as biocompatible transporters have shown to be localized within human promyelocytic leukemia (HL60) cells and human T (Jurkat) cells via endocytosis. Functionalized SWCNT exhibited little toxicity to the HL60 cells, but the SWCNT–biotin–streptavidin complex caused extensive cell death [15]. Studies by Pantarotto et al. [16] have also demonstrated that functionalized, water soluble SWCNT derivatives modified with a fluorescent probe can translocate across the cell membrane of human and murine fibroblasts. Carbon nanotubes can be functionalized with different groups that can make them biologically compatible so that they can interact, bind, or be taken up by mammalian skin cells and then have the ability to transport their therapeutic efficacy with minimal toxicity. The translocation pathway or mechanisms still remains to be elucidated.

Our group has shown the dermal toxicity of MWCNT in a primary human keratinocyte model. These MWCNT exhibited a base mode growth, had very little disordered carbon, and were well ordered and aligned [17]. Human neonatal epidermal keratinocytes were exposed to 0.1, 0.2 and 0.4 mg/ml of chemically unmodified MWCNT for 1, 2, 4, 8, 12, 24 and 48 h. Transmission electron microscopy depicted numerous vacuoles within the cytoplasm of epidermal keratinocytes containing MWCNT. The MWCNT had a range of diameters and were up to 3.6 μm in length. Quantitative analysis conducted at 24 h at the 0.4 mg/ml dose demonstrated that 59% of the keratinocytes contained MWCNT, compared to 84% by 48 h. Keratinocyte viability decreased with an increase in MWCNT concentration, and IL-8, an early biomarker for irritation, increased with time and concentration [18]. These data showed that MWCNT, neither derivatized nor optimized for biological applications, were capable of both localizing and initiating an irritation response in skin cells. These initial data are suggestive of a significant dermal hazard after topical exposure to select nanoparticles should they be capable of penetrating through the stratum corneum barrier. Additional studies were conducted with proteomic analysis in human epidermal keratinocytes exposed to MWCNT which showed both an increase and decrease in expression of many proteins relative to controls. Changes seen with protein alterations suggested dysregulation of intermediate filament expression, cell cycle inhibition, altered vesicular trafficking/exocytosis and membrane scaffold protein down-regulation [19], [20]. Currently, our proteomic studies with MWCNT are ongoing. Hat-stacked carbon nanofibers (which are similar to MWCNT in diameter and length) implanted subcutaneously in rats depicted granulation and an inflammatory response that resembled foreign body granulomas. These fibers were also found within the cytoplasm of macrophages. However, investigators reported no severe inflammation, necrosis or degeneration of tissue [21].