This content is current only at the time of printing. This document was printed on 28 March 2017. A current copy is located at http://apvma.gov.au/node/19436
You are here
Transcript for Distinguished Professor Nancy Monteiro-Riviere
Safety Implications of Nanoparticles and Skin
This presentation was delivered at the APVMA’s science feature session on 15 October 2015.
I would like to share with you some of the research we've conducted over the years relating to dermal absorption and dermal penetration of nanomaterials. I'm going to talk a lot about a lot of these little nanomaterials but it's going to be sort of like a Julia Childs, you're going to just see the end product. There's a lot of opportunity for contact, whether you're synthesising these nanomaterials in a research lab or during a manufacturing process. In this case you can see people wearing proper PPE. However, if you look closely to the glove and the lab‑coat area, there's an interface and bare skin is exposed. If you look closely, there are particles on there.
Whether you're talking about occupational exposure, again, if you notice they have a gap with particles, or cosmetics or sunscreens, everyone is concerned, is there a safety issue? Yes, skin is an important root of exposure. I'm interested in skin absorption and toxicity of engineered nanomaterials. I want to know, can these nanoparticles cross the skin? Would these particles preferentially locate in the stratum corneum lipids? Are they toxic using keratinocytes? Can they gain access to tissue spaces? We also have a profuse model which we've shown it coming out, it extravagates out through the endothelial cells. Are there species differences? Can they penetrate abnormal skin? Also, do solvents influence nanoparticle penetration?
This is a generic skin and I call it generic because on this side I have close, the human skin where eccrine opens to the surface where most domestic species, the apocrine sweat gland and it opens opposite sebaceous into the hair follicle. But skin is a very complex and dynamic organ and has many functions. If you look at the skin here you have an epidermis and a dermis. When you topically apply a substance it either can go between the cells, through the cells, through the opening of a hair follicle or an opening of a sweat duct. But however in these keratinocytes they can function as exogenous transducers. They can convert exogenous stimuli and then release cytokines that set up an inflammatory‑type reaction. Also you have these linegerin cells.
If you look at this rate limiting barrier, picture, this is Peter Elias's model of "brick and mortar" where each brick represents the acellular dead keratinocyte and the mortar between the bricks represents your lipid. Most chemicals come through this torturous pathway, through the intercellular space. This is the morphological counterpart representing in the lipids here, the aqueous and lipid domains.
It was once thought you had to conjugate nanomaterials in order to get them uptake to cells. We know that that's not true anymore and most cell types tend to pick up nanomaterials. We first started work with multiwall carbon nanotubes and you can see in multiple vesicles of the human epidermal keratinocyte these large tubes. We found them as big as 3.6 microns in the cells. Within 24 hours, 59% of the cells uptake these cells, these tubes, and by 48 hours you've got almost 84% of them have update. They can readily get in without being conjugated.
We also know that even with bucky‑amino‑acids that they also can uptake through the cells. As you can see here, you've got sort of an endocytic process and they're phagocytising these nanomaterials and bring them in. As a toxicologist I'm interested in the mechanism of toxicity and which specific endocytic process do these particles penetrate.
We had conducted a study with 24 different inhibitors where we evaluated the seven major pathways, whether they come in through cyto‑skeletals, caveolae, clathrin, macropinocytosis, gene protein‑coupled receptors, melanosome pathways because they are keratinocytes and keratinocytes can engulf melanosomes and also scavenger receptor of the bone density lipo receptor pathway.
My graduate student did a great job with 24 different inhibitors and figured the set. This is only one of his papers. We used quantum dots as a marker because they naturally fluoresce. We did this with 565s and 655 and we mapped out all these different pathways.
What we showed here is that, here's a control with a keratinocyte picking up the normal quantum dots. With fucoidan and lipodensil, lipoprotein we almost suppress the quantum dot uptake. We knew the scavenger receptor, low density lipoprotein receptor played a role. We also showed that with the gene protein coupled receptor related pathways also played the role by U‑73122 which then completely suppressed it.
We went on and studied absorption and penetration. Many times these two words are used interchangeably but they shouldn't be because they're actually very different. Absorption relates to the amount of chemical that you topically apply on the surface of the skin of the stratum‑corneum that goes down through the granulosum, the spinosum, the vesselae, through the basement membrane, that can get picked up in the dermal papillary capillaries. That's the whole process of absorption, going through the skin, getting in to where the blood vessels are to have the systemic effect.
Penetration relates to the amount of chemicals or particles that could be topically applied to the skin that may reside in the skin somewhere in one of the layers or just stay in the stratum corneum and could potentially set up a cuteaneous response. The reason I say that is linegerhen cells and genetic processes come very close up to the granulosum corneum interface. If a nanoparticle gets there and gets taken up by these endorytic processes there's a potential for an immunologic response.
The two ways of measuring dermal absorption which many of you are probably familiar is either using Bob Bronaugh, his flow through chamber, or Thomas Franz static diffusion cell chamber. Sometimes it's hard to compare these studies with other investigators because many people will use different types of skin. We use surgical‑abdominoplasty human skin, we use pig skin because it's anatomically, physiologically, pharmacologically similar to humans and much better than the rodent species. When you topically apply a substance there you can measure what comes through over time and with different media, maybe in the receptors. Some people use phosphate buffered saline. Others add bovine serum albumen. This makes different variables and sometimes very difficult to compare between investigators.
For more defined cells, it doesn't have a continuous flow underneath, it has a static so you can get a build up of whatever you're looking for. This is the actual apparatus for the flow through. That's fine when you're talking about a radial labelled compound or a compound that you can then localise by HPLC or GC that's coming through into the receptor. But when you're working with nanoparticles it's a different ball game. It has to somehow be fluorescently tagged or then you're changing the whole physicochemical structure of the nanomaterial. Or if you put a FITC label, you've got to make sure you put an amino acid sequence on your nanomaterial to keep it away from the nanomaterial, but the FITC has to have a certain distance so they don't quench when you're looking under, say, a confocal.
We started off with naturally fluorescent quantum dots. They have a cadmium selenide core and a zinc sulphide shell which is only 4.6nanometres in size. When you change this by adding a hydrophilic coding such as a carboxyl‑amine or PEG you then increase its size. So we studied a spherical and elliptical‑type quantum dot and again, these sizes will change. Anything you do to this particle, it will alter its physicochemical composition and therefore be almost like a different particle.
We evaluated this in dermatome pig skin on the flow through diffusion cells, looked at whether charge or size and shape made a difference as far as penetration. You could see here with the 565s, by eight hours all three charges here were penetrated. With the 655, those were the different shape, those were the elliptical ones, we had good penetration at eight hours with the PEG, with the amine, but it took 24 hours before we could see penetration here of the carboxyl group.
Then that led us ours, size and shape, we pretty much knew it was the surface charge, could it be size? We went on and got some nail shape nanomaterials and this was obtained, this is 621s and this was obtained from Dr Vicki Colvin and William Yu. She's from the lab at Rice University with a Nobel Laureate for buckyballs which Smalley was.
We took these quantum dots here and we topically applied it into pig skin on flow through diffusion cells at different concentration over different time points to look for penetration. You could see primarily with the 621s they sat up right up here on the surface and did not penetrate but in areas of hair follicles you can see little hotspots. This is very typical with a lot of nanoparticles and it has been thought that maybe hair follicles may play a role. A hair follicle is just an invagination of the epidermis into the dermis and if picture take at a tangential section, we've shown all these particles are still binding in that stratum corneum, invagination.
Ultra‑structurally, we looked at these 621s and we could see here, between the stratum cornea cells, you could see the particles. They also went through that intracellular torturous route that I was talking about. Similar to a lot of the chemicals and you can still see it retained its shape in the lipid milieu.
Many times you don't see them as nicely as I showed. That's one of hundreds. We had crap to look at sometimes. A lot of these particles can appear differently and as a morphologist by training, I like to do serial sections. If you take a section through here you would probably see particles on the top and the bottom. At other times remember they follow this torturous pathway so if you take it down here you would see them in all cell layers. That can explain why some people just seem them at the top or just at the bottom, so you have to take that into consideration.
Another thing we're interested in is that people in the African Rift Valley can walk around barefoot in the mud and when they do that they tend to get particles from the soil into the inguinal lymph node. That triggers a condition called Podoconiosis. That condition is a non‑filarial elephantiasis. It was thought that because of the jumping and standing and not wearing shoes, that this particle can then migrate up to the lymph node. We designed a study with a biomedical engineering student where we took dermatome skin and we placed it on a rotating motor with a door hedge side which she took the motor axe rotates and then lifts this little flexing device at 20 flexes a minute where we have the dermatome skin. We kept on doing this for every 15 minutes and then put this on the diffusion cells to assess the penetration with chronic motion, which would mimic a working and an occupational exposure, maybe doing a repetitive movement, repetitive motions.
What we saw here, now this is with a very small nanomaterial, this is a bucky‑amino acid with a FITC label and a nuclear localisation signal, and what we saw here only at 60 minutes, that we could see penetration of this small 3.5 nanometre particle through the epidermis. By 90 minutes we saw a gradual penetration into the dermis. Of course, we did this for 24 hours and it was just more of an exaggerated response.
We went on and evaluated different species. Everyone likes to look at the rat and a rat is a very poor model for dermal absorption because it's usually very, very permeable. Normal rat skin is only one to two cell layers in thickness versus human and pig which is four to five cell layers. If you tape strip to remove that rate limiting barrier, which is the stratum corneum, you need to do it 10 times where in a pig you've got to do it almost 30 times. Then with flexion there's not much difference but with abrasion 60 times with sandpaper you can remove this epithelium to see if we could increase penetration.
We did this on rodents then put them on diffusion cells but I want to point out that even though you clipped an animal, we usually clipped 24 hours prior to putting it on the cells, the stubs of the hair starts growing. This is where, when you place particles, they bind to the hair so they're really not good contact here to the skin itself. Flexion sort of evenly disperses it out but with tape stripping you removed all that hair, you removed that rate‑limiting barrier, so you have the particle directly, topically on skin. Also through abrasion you're sitting right on pretty much the dermis.
We evaluated these and in a nutshell only with abrasion did we see any penetration, at 24 hours you could see here, which is understandable because you've got no epithelium, and with the 565s the same thing occurred at 24 hours with abrasion you could see penetration. I want to show you the area here of the fluorescence, it lights up the hair follicle so again, maybe hair follicles do play a role but again it is also in the outline of the follicle which if you cross‑section it is still out in that stratum corneum layer.
We did human skin abdominoplasties which we dermatome within two hours. It's fresh skin, it's then tape stripped, not tape stripped and placed on our diffusion cells. We thought, well, without the corneum you probably would get good penetration. We've got a great barrier because even without the stratum corneum, at eight and 24 hours we saw penetration? No, we just had it sitting here on the surface, it did not penetrate into the epidermis.
Again, a human skin, same thing with all the models, the pig and the rat, if there is an area of the hair follicle which is an invagination here, it takes down that stratum corneum and what lights up is the stratum corneum. When you take a tangential section what you get is really in that stratum corneum.
We also did these studies with different turpenes for solvents. Can solvents help, cause, increase penetration with a lot of chemicals? In this case we did it with ethanol, menthol, eucalyptol and limonene. What we saw here after 24 hours, we first pre‑treated this human skin and then topically applied the quantum dots and exposed it for eight and 24 hours. We could see it pretty much sat on the surface of all these different solvents. We've also conducted in vivo studies where we looked at a C60 which is a bucky‑amino acid. We did through single liquid extractions and then HPLC, either isolating the C60 but we put this as C60 in several different solvents like chloroform, cyclohexane, toluene and saturated mineral oil. When you tape strip we had to go to about the first five or six tape strips, we found great penetration mainly with the chloroform.
Let's talk about silver. We looked at silver and silver in culture is difficult because silver nanoparticles with dissolute. We've now moved on to gold but even back in Hippocrates time they realised the value of silver and that had antimicrobial properties and they used it for their wine, to preserve their wine. We know silver can be used in all types of photography, in jewellery. It's on your keyboards to prevent growth of bacteria, it's on your stain‑resistant clothing. Due to new resistant strains of bacteria there's also that need and we're searching for, is an alternative so we don't get this bacterial resistant. Silver nanoparticles has been a way that has been used in different types of things like in paints and textiles to prevent the growth of bacteria.
The Germans tend to have a lot of GI discomfort and they tend to drink colloidal gold. Once one drinks colloidal gold silver you get argyria. This is a condition where the silver nanoparticles then undergoes dissolution and the ions, not the primary particles but the secondary, resides in the skin. The skin that's exposed to UV light then causes this condition.
We examined silver at several different sizes, 20, 50, 80 unwashed, 20, 50, 80 washed and 25 and 35 carbon‑coated. While we always conduct our own measurements, you should never take the manufacturers suggestion because it's usually never correct, you always use your Malvern Zetasizer or whatever and look at your dynamic light scattering. We also do it by TEM. And we also do a zeta potential. What we've shown was we did all toxicity studies. The unwashed, when we washed it several times and evaluated the supranae and we did show toxicity. We also found five parts per million of formaldehyde. The question arose, was the toxicity really due to the nanoparticle or was it due to the five parts per billion of formaldehyde? When you washed them they did not cause toxicity and when they were carbon‑coated they did not show toxicity to keratinocytes.
We noticed ‑ here's a TEM of all those particles whether washed and not washed and the sizes. They were very homogeneous and very singular, not so much in large agglomerates like a lot of other particles. We also noticed that when you look for them inside of these cells, whether they're washed or non‑washed and what sizes, again, they all got easily taken up into these cells. Now, what we do in our lab is we do not stain or keratinocytes because when you do the post-fixation with osmium tetroxide, there's enough post‑fixation stain in that osmium to stain your membrane. This way you can find your particles quite readily and they just sort of jump out at you as big black boulders.
We always do EDS on these so no‑one can say anything like it could be a lead precipitator and staining and that's another way you don't want a stain because that avoids that other variable. You can see through EDS that this was silver particles.
These silver particles were also evaluated in vivo. We used pig skin and with the pig, Megan ended up putting topical application of these particles every day for 14 days because at the time there was very little work longer than six hours or 12 hours on skin. To me that's not enough for penetration. Even with 14 days of repetitive application of these silver nanoparticles, they still resided in the stratum corneum and verified via EDS.
Another thing I'd like to talk about which is we may all have concern is titanium and zinc oxide nanoparticles in sunscreens. If you're like me, you're at the beach you get a sunburn but you're not going to stop going out, you're just going to apply more nanomaterial sunscreen on you and go back out in the sun. We wanted to assess the potential for TiO2 and zinc oxide in UVB damaged skin which is red and inflamed and see if they penetrated or had toxicity.
We've been doing UVB studies for about 20 years, working with Sheldon Pinnell at Duke. We did it mainly for anti‑aging. Usually you have to work out your two‑and‑a‑half MED dose and what we did is found between 110 and 120 milo‑joules per centimetre squared. We have a very narrow fibre optic probe in which we can get nice round plus‑two erythema on the pig. Once you find that, the conditions for a plus‑two, the whole pig either was topically applied with these sunscreens and they were dermatomed and then placed onto diffusion cells to look through ICPMS, if any of the ions came through.
We worked with BASF which provided us with different sunscreens and different formulations such as oil and water, water and oil emulsions and also coated. They were coated with hydrated silica and dimethicone and dimethicone co‑polymer. When we did zinc, zinc was either coated or uncoated and if they were, they were coated with a triethoxycaprylylsilane, I hope I'm saying that right.
Anyway, we applied these formulations both in vivo and in vitro, and we also made sure we looked for the rutile nature of the TiO2 by TEM and also did EDS to show we had nice TiO2 particles in the sunscreens. We also looked at zinc oxide. Zinc is always very bold and un‑homogeneous. Then this was evaluated by TEM because you really can't see anything by HN‑ese which we did except you can see residual formulations sitting on the surface with some of the titanium. What we saw here, this is no UVB‑treated skin and this is UVB‑treated skin. TiO2 and the formulation sat primarily on the surface. With UVB‑treated skin we had a little bit more that penetrated down a few layers of the stratum corneum but still was within this non‑viable layer.
With zinc oxide it was quite boring. Zinc always stayed on the surface. Nothing excited, whether it was UVB‑treated or not. But remember the hair follicles, we showed nanoparticles in those invaginations so we were looking for hair follicles which is quite difficult because in pig skin there's only 10 or 11 hair follicles per centimetre squared and when you're looking at one cell at a time. But we did scanning EM here to look for the formulation and you could see it up in this infundibulum area of the follicle and verified it as a formulation.
Same with zinc. We did the same thing and you could see the formulation here, verified with EDS, that we did have zinc that congregated at the orifice of the follicle. We decided we had time‑of‑flight secondary ion mass spec which is a very sophisticated technique. Since we weren't finding the particles we thought, lets look for ions. What you have here is a bismuth gun that then sends out, erodes the surface of the skin, and that eroded species goes up through time‑of‑flight and then gives you a mass‑to‑charge ratio. What you're looking at is not necessarily the particle but you're looking at the ions in the skin.
We did this with in vivo and in vitro with all the different formulations. This is normal skin, this is UVB skin but with the treatment of the 630 which is the Ti02 combination and UVB‑treated skin you can see they were quite similar. We always did serial sections of these. These were done off the cryostat and then placed on these special wafers to see if we could localise what were these hotspots. Were they cross‑section of follicles? Of course this would be too high but we could not distinguish what was there.
Same with zinc. Zinc was quite strange again but expected. Sat pretty much on the surface. When we did in vitro we wanted to look at it to see if any of these ions could get absorbed so we did them on the diffusion cells with the UVB‑treated skin. With ICP‑MS analysis for titanium you can see control was just below the level of detection and both formulations look very similar at 6, 12, 20 and 24 hours. The same for the zinc. The zinc, the control was really no different than the two zinc formulations, whether they were coated or not coated we had the same amounts. Like we said, zinc was found in other things in the lab, like in powdered gloves and in stainless steel.
When we looked at the TEM of this, with normal skin you could see you did have penetration in vitro and this came through several layers but with UVB, the TiO2 came through several layers and you can see the nice rutile nature. Zinc again, you're lucky if you could find it. Zinc always sat right here on the surface. When we did TOF‑SIMS, this is the UVB‑treated here, non‑UVB‑treated here, you can see we had a greater ion penetration with in vitro rather than in vivo models. I think that the in vitros, because your temperature on the diffusion cells is a little high, I think around 32 to 37 degrees, but we did see more ion penetration. Again, if you compare non‑UVB to UVB, there wasn't much difference. When you did zinc, zinc ions again, comparing non‑UVB to UVB, you can see again there were very similar higher concentration in the upper layers.
A lot of things to consider when you're assessing nanoparticle penetration, whether it's in cosmetics or in manufacturing or in the lab, that every particle can be very different when it's in a different vehicle like in solvents because it can also change the particle structure and solubility. Some of these formulations can cause agglomeration. A lot of the sunscreens, once you put it in a formulation, that particle in the sunscreen can go up to three, four, five‑hundred microns in size even though it started out at maybe 50 nanometres.
There's differences in different species, different follicular arrangements. There's differences in thickness in both pig, human and rat. Also, you can't really extrapolate what you see with one particle and be bias and say that particle is similar to all particles of that size because every particle is different, particle size, the size distribution, their shape, surface, coating, the PH that's in that particle. Quantum dots are made to be biocompatible so they're in a normal PH. Penetration can be a function of the type of particle. Is it AC60? Is it carbon? Is it a metal oxide? Is it a multiwall carbon nanotube? Species and skin treatment. Whether you flexed the skin. Whether it was taped stripped as mechanical action or whether it was abraded or exposed to UVG light.
The take home message was what the current literature says has shown very minimal penetration of nanoparticles in skin and that the physicochemical properties of those particles are the major determinants and modulate how it can get in through the skin or in skin cells. Right now we are manipulating a lot of this with gold nanoparticles and put them in different physiological concentration of the plasma concentrations like albumen, fibrinogen, the most common ones, and study the uptake intercells, which we show fibrinogen is very weird. It does cause a lot of agglomeration and it's very different. When you stick a particle in the body, say an IV dose, it comes in contact with 3500 different proteins and it depends on what protein that binds to that particle will determine where it gets distributed throughout the body.
There's been no firm conclusion or disease effects of nanoparticles or exposure with skin and we've shown here that a lot of the things can influence particle penetration as far as up to the stratum corneum layer and studies have been conducted on diseased skin like with psoriasis and have shown no additional penetration. There's things to consider such as in vitro models versus in vivo models because you're limited to a lifecell span of about 48 hours on a diffusion cell versus doing in vivo studies. In vivo studies would be very different than say, some of the cell culture studies because in vivo you have to worry about the lipids and the intracellular, how those particles traverse through, versus in a cell culture, you're worried about the media of that culture. Some media has proteins and a lot of times when you deal with cell cultures you have to be careful for what type of proteins, proprietary formulations. You know, they've got the pituitary hormone, they have all this other stuff and that causes agglomeration and also causes a coding on your particle, and all that can also alter the effect of the particle as far as its cellular uptake.
That's something to think about because with skin you don't have to worry about protein coronas but maybe lipid coronas.
I'd like to thank the group here in our lab which is mainly post‑ops and graduate students and support by two EPA star grants, a US Air Force grant, a National Academy Keck Futures grant, an NIH(RO1) and Industrial Support by BASF on the sunscreens. I just want you to remember that not all buckyballs are the same size. I'll be glad to answer questions.
Errors and omissions excepted; check against delivery.