1.

Introduction

Optical methods as a tool for clinical functional imaging of physiological conditions have been of great interest to the biomedical optics community.^1–3 Over the past decade, noninvasive or minimally invasive methods of biomedical diagnostics and therapy utilizing spectroscopy and imaging techniques, including optoacoustic (OA) imaging, optical coherence tomography (OCT), confocal nonlinear microscopy and high speed imaging, in vivo fluorescent flow cytometry, and reflectance spectroscopy have seen widespread proliferation in biology and medicine.^2,4–7 It is commonly understood that strong light scattering in skin tissue layers greatly diminishes light penetration through skin and limits the scope of these methods by reducing their contrast and penetration depth.^2,8 By reducing this unwanted scattering, the performance of biomedical optical imaging techniques can be significantly improved.

Throughout published literature, one may find several hundred chemical compounds as potential skin penetration enhancers for different tasks, such as drug delivery, diagnostics, imaging biological objects, and visualization in vivo.^9–17 For instance, the use of hyperosmotic biocompatible agents, such as glycerol, dimethylsulfoxide (DMSO), and polyethylene glycol (PEG) is well known to induce the effect of optical clearing (OC).^18–22 One of the pioneering studies in OC demonstrated about a 50% increase in the optical transmittance of rat skin after a short (5 to 10 min) administration of anhydrous glycerol.¹⁹ Recently, optical penetration depth and image contrast of photoacoustic microscopy enhanced by OC were measured as function of glycerol concentration.⁷ The maximum increase of the image brightness of 165% was achieved with topical application of 60% glycerol in water. OC of biological tissues may be achieved utilizing two main mechanisms—tissue dehydration and refractive index matching.²² Most of the previously reported methods of optical clearance utilized tissue dehydration induced by the optical clearing agents (OCA) interacting with external layers of tissue.^3,14,22–28 Khan et al.²⁹ used a PPG and PEG mixture to reduce dermal optical scattering and increase penetration depth of OCT by 17% in human skin in vivo. Tuchin³⁰ reviewed available methods of tissue OC in various tissue structures.

Exogenous hyperosmotic chemicals used as OCAs penetrate into tissues and displace water, thereby reducing refractive index mismatch between scatterers, such as collagen fibrils, cell organelles, and interstitial fluids or cytoplasm.^14,27 The OCA list includes well-known chemical compounds, e.g., glucose, glycerol, dextrose, fructose, sucrose, sorbitol, xylitol, propylene and polypropylene glycol (PPG), PEG, butylene glycol, DMSO, ethanol, oleic acid, sodium lauryl sulfate, azone, and thiazone.^3,14 Typically, OCAs can be classified as being part of three main groups—sugars, alcohols, and electrolyte solutions.²⁷ Some agents, such as DMSO and PEG, can be used both as OCAs and clearing enhancers, or used together with chemical or physical enhancers to increase diffusion across stratum corneum and other skin layers.^14,23,26 Various methods for enhancing penetration of OCA into tissue, such as physical methods, chemical-penetration enhancers, and combinations of physical and chemical methods, can be utilized.³

As presented in the lower part of Fig. 1, molecular agents may undergo three potential pathways when penetrating skin. The first is the intercellular route through the stratum corneum, i.e., through the structured intercellular lipid bilayer.³¹ This transport process involves sequential diffusion and partitioning between the polar head group regions and the long alkyl chains of the lipids.³² The second is the transcellular route, which can be utilized by small hydrophilic substances.¹¹ The third minor route is through hair follicles, associated sebaceous glands, and sweat ducts.

Fig. 1

A diagram showing routes of skin penetration by OCA, adopted from Ref. 11: (1) through the sweat ducts; (2) directly across the stratum corneum; and (3) via the hair follicles. Permeation routes directly through the stratum corneum include intercellular diffusion through the lipid lamellae, intracellular or transcellular diffusion through both the keratinocytes and lipid lamellae, and diffusion through appendages, i.e., hair follicles and sweat ducts.

Typically, the OC process takes more than an hour,³³ but it can be accelerated with chemical or physical enhancers.³ We propose hyaluronic acid (HA) as a potential chemical enhancer because of its ability to change physiological properties of skin (stratum corneum) through water redistribution. Hyaluronan or HA is a nonsulphated, straight chain glycosaminoglycan polymer of the extracellular matrix. It can reach several thousands of sugar units (carbohydrates) in length.³⁴ The biological functions of HA include maintenance of the elastoviscosity of liquid connective tissues, control of tissue hydration and water transport, and numerous receptor-mediated roles in cells.^35,36

The structure of the disaccharide is energetically very stable.³⁷ HA is highly nonantigenic and nonimmunogenic, and has poor interaction with blood components.^38,39 The unique viscoelastic nature of HA along with its biocompatibility has led to its use in a number of clinical applications, including the supplementation of joint fluid in arthritis, in eye surgery, and enhancement of healing and regeneration of surgical wounds.^40–43 HA has been investigated for delivery of different agents for various administration routes, including ophthalmic, nasal, pulmonary, parenteral and topical drugs, and gene delivery.^42,44–46 Dermatological and cosmetic applications using of HA have been extensively probed.⁴⁷ HA-based formulations may also be capable of protecting the skin against ultraviolet irradiation due to the free radical scavenging properties of HA.⁴⁸ Radio-labeled HA was found not only to penetrate the skin of nude mice and humans, but also to aid the transport of diclofenac to the epidermis.⁴⁹ The results³⁸ show that HA is not cytotoxic and shows good biocompatibility. Studies^38,49 collectively imply that HA’s influence on the local delivery of chemical compounds to the skin depends upon the skin characteristics and species of skin used, with the compounds depot forming in deeper layers of mouse skin (dermis) compared with humans (epidermis).

HA hydrogels and nanocomposites were used in different experimental and preclinical models.^34,50,51 Amphiphilic HA derivatives bearing hydrophobic indocyanine green dye and hydrophilic PEG were synthesized through the use of condensation and copper-catalyzed click cyclization reactions.⁵² The repair effect of HA and vaseline was demonstrated in dermal blood vessel study for peripheral vascular diseases.⁵³ Two kinds of fluorescent HA analogs, one serving as a normal imaging agent and the other used as a biosensitive contrast agent, were developed for the investigation of HA uptake and degradation.⁵⁴ An increase in HA synthesis, cellular uptake, and metabolism occurs during the remodeling of tissue microenvironments following injury.⁵⁵ In recent years, HA has emerged as a promising candidate for intracellular delivery of various therapeutic and imaging agents because of its innate ability to recognize specific cellular receptors.⁵⁶ Injectable hydrogels based on HA and PEG were designed as biodegradable matrices for cartilage tissue engineering.⁵⁷ It should be noted that hydrogels are not OCA, and in fact they can increase optical scattering of tissues.⁵⁸

In this report, we demonstrate the use of aqueous solutions of HA without modification and functionalization as potential enhancers for delivery of OCAs, such as PPG and PEG, and their mixtures, through pig skin. The aim of the present study is to demonstrate OC of skin after using chemical pretreatment with HA.

2.

Materials and Methods

2.5.

Optoacoustic Imaging System

Imaging experiments reported here were made with a commercial 3-D laser optoacoustic imaging system (LOIS-3-D) developed for preclinical research by TomoWave Laboratories (Houston, Texas) and introduced in our earlier publications.^4,62–65 Laser output consisted of Alexandrite laser (757 nm, $50\mathrm{ns}/\text{pulse}$) and Nd:YAG (1064 nm, $9\mathrm{ns}/\text{pulse}$), which was manufactured as a single unit with coaxial output beams (Quanta Systems, Solbiate Olona, Italy).⁶² OA tomography was performed using the imaging module of LOIS-3-D, as shown in Fig. 2. The stainless steel tank holds a 115-deg arc-shaped array of 96 wide-band ultrasonic transducers connected to our custom-designed electronics and our quad fiberoptic bundle, which was proven to be a viable light delivery system for small animal imaging (see Fig. 2).^62,63 The pulsed laser energy was about 6.5 mJ out of each of the four fiberoptic illuminators, which corresponded to the total effective optical fluence of $2.5\mathrm{mJ}/{\mathrm{cm}}^2$.The system has low noise equivalent pressure of 2 Pa without averaging,⁶⁴ which permits reliable visualization of the OA contrast generated by variation of the tissue optical absorption of $0.12{\mathrm{cm}}^{-1}$ with 4-pulse averaging, making the total image acquisition time 2 min.⁶⁵

Fig. 2

The imaging module of the laser OA imaging system, LOIS-3-D (TomoWave Laboratories, Houston, Texas), which includes (1) quad fiberoptic bundle of the light delivery system, (2) arc array of ultrasonic transducers, (3) computer controlled rotational stage, and (4) video camera.

The phantom consisted of four garolite rods for support, angled target tubes (tubes with GNR and nickel sulfate), and a layer of skin of 2.5 cm height creating a rectangular shell centered around the section of the tubes, where the focus of the arc would be (see Fig. 3). A custom-made bracket supported two polytetrafluoroethylene (PTFE) tubes with internal diameter $510\mu \mathrm{m}$ and wall thickness of 150 mm. One tube was filled with GNR nanoparticle suspension and the other was filled with nickel sulfate solution in water. Nickel sulfate was used with the same optical density as GNR to enable quantitative measurements of changes in the OA image brightness of the tubes covered with pig skin that occur after OC of the skin. Figure 3 shows a picture of the phantom prepared for imaging.

Fig. 3

Preparation of pig skin sample for OA scanning with LOIS-3-D. Skin is wrapped around a holder containing PTFE tubes filled with contrast agents.

OA contrast of GNR and nickel sulfate was compared for images obtained before and after application of 0.5% solution of HA (pretreatment) and low molecular weight 100% PPG-425 or 100% PEG-400 and PPG-PEG mixture $1:1$. Imaging experiments were repeated three times. Signal processing was similar to that described in our previous publications using deconvolution of transducer impulse response and a bandpass filter.⁶⁶ Image reconstruction was made with a filtered back projection algorithm utilizing a half-time data set.⁶⁷ 3-D visualization was done with LOIS-View, a software package developed at TomoWave based on the open software development kit source developed by 3-D slicer.⁶⁸ We created a segmentation module of LOIS-3-D software that allowed one to draw regions of interest along the axial slices of the reconstructed scan, interpolate, and connect them. This segmented region was statistically analyzed by measuring mean and median brightness values and their standard deviation. Each tube was segmented into a region of interest that was a cylinder of 0.6 mm in diameter and height of 5 mm. The average brightness of OA images was measured in the central area of the images illuminated more evenly than the adjacent areas. The image processing parameters such as scalar opacity, color mapping, and gradient opacity were held constant between all objects and scans such that one can directly compare the image brightness across all images in a qualitative manner and assign their colors using a linear color palette. From lowest to highest, OA contrast is black to red, with voxels deemed noise colored black (transparent).

3.

Results

First, we demonstrate the OC effect after application of clearing agents without HA. Figure 4 shows consecutive photographs of skin samples treated with PPG and PEG solutions for 10, 20, 45, 90, and 120 min. We observed significant OC in these samples after $\sim 90$ and 120 min of exposure to PPG and PEG mixtures, respectively. In the course of experiments, it was discovered that when the same clearing agents were used in combination with HA pretreatment, a similar clearing effect was achieved in a significantly shorter period of time.

Fig. 4

Dynamics of OC after treatment of pig skin with control (water and PBS) and active (PEG and PPG) clearing agents. Time values are in minutes.

Figure 5 shows the comparison of measured unscattered transmission spectra from skin samples after clearing procedures with and without HA pretreatment. The recorded spectra were not normalized by the spectral profile of the illumination source. We observed that application of only PPG, PEG, and their mixture for 60 min led to a relatively small clearing effect [Fig. 5(a)]. With HA pretreatment for 30 min, followed by treatment with PPG, PEG, and their mixture for 15 and 45 min [Figs. 5(b) and 5(c)], OC was significantly improved. The experiment showed that pretreatment with HA enhanced OC of skin even after a relatively short PPG–PEG mixture application. After 45 min, the transmission of unscattered light through skin samples increased at least eight times.

Fig. 5

Unscattered light transmission spectra through skin samples measured after different OC procedures: (a) without HA pretreatment, using only PPG, PEG, and their mixture as clearing agents for 60 min, and (b, c) with HA pretreatment for 30 min, followed by treatment with the same agents for 15 and 45 min, respectively.

To quantify increase in the skin transmission after OC, we compared the transmission spectra of cleared skin samples shown in Fig. 5(c) relative to a control, which was measured from a sample pretreated with HA solution for 30 min and then treated by an inactive PBS solution for 45 min. Figure 6 shows photographs of the skin samples before and after the clearing procedure (a) and the normalized transmission spectra (b). This figure demonstrates a $\sim 47$-fold increase of the optical fluence transmitted through skin. To demonstrate potential benefits of skin clearing, we performed a series of OA imaging experiments of a test phantom presented.

Fig. 6

(a) Photographs of pig skin samples before and after OC procedure consisting of 30-min HA pretreatment and 45-min treatment with PBS control, PPG, PEG, and their mixture. (b) Normalized unscattered light transmission spectra after OC procedures. The spectrum of unscattered light transmitted through HA-PBS treated skin was used as a reference.

Both two-component GNR and nickel sulfate solutions were used as optical contrasts in these experiments. Absorption spectra of pegylated GNR with maximum LSPR at 757 and 1064 nm are shown in Fig. 7. By adjusting concentrations of these components in a mixture, it is possible to tune the absorption coefficients at these wavelengths to desired values.

Fig. 7

UV–VIS–NIR spectra of pegylated gold nanorods with maximum LSPR at 757 nm (Alexandrite laser) and 1064 nm (Nd:YAG laser) and their mixture adjusted for the same optical density for both wavelengths. The mixture solution demonstrated LSPR peaks very similar to each GNR measured separately.

Figure 8 shows the absorption spectra of a two-component GNR solution matching absorbance of the nickel sulfate solution at both 757- and 1064-nm laser excitation wavelengths.

Fig. 8

Absorption spectra of a two-component GNR solution and nickel sulfate solution with matched absorbance at 757 and 1064 nm.

Figure 9 shows a photograph of the phantom positioned for scanning (a) and the reconstructed images of the test tubes (b). Skin in the test phantom was subject to the same OC procedure that begins with 30-min pretreatment with HA and is followed by 45-min treatment with the PEG–PPG mixture. OA images of the phantom were acquired before and after the OC procedure. Projections of 3-D volumes represent segmented portions of tubes filled with GNR and ${\mathrm{NiSO}}_4$, which were reconstructed after scanning with 757- and 1064-nm lasers Fig. 9(b). Figure 9 shows a photograph of the phantom and the ultrasound probe positioned for scanning (a), and the reconstructed images of test tubes filled with contrast agents before and after the clearing procedure (b).

Fig. 9

(a) Photograph of the experimental setup showing the arc-shaped ultrasonic probe and phantom in their vertical orientation during OA scanning. (b) 3-D image projections of a segmented portion of GNR and ${\mathrm{NiSO}}_4$ tubes reconstructed after OA scanning with 757 and 1064 nm. A significant contrast increase can be seen after pretreatment with HA as opposed to the image obtained before the treatment.

We found that application of the OC procedure proposed in this study resulted in significant enhancement in the average pixel brightness for both contrast agents and for both wavelengths.

4.

Discussion

PEG and PPG are chemicals that fall into the broad chemical classification known as “alcohols” as any organic chemical that contains an $─\mathrm{OH}$ (oxygen/hydrogen) functional group. The US Food and Drug Administration (FDA) approved PEGs for medical applications. Propylene glycol is among the most thoroughly tested and well understood chemicals. It has been affirmed by the FDA as a “generally recognized as safe” compound and is permitted for use in food and drug products. FDA permits use of the term “alcohol-free” for cosmetics that contain alcohols other than ethanol. HA is approved by the FDA for many different clinical applications from dermal fillers to injectable substations and viscosupplements.

The disaccharide subunit of HA consists of d-glucuronic acid and d-N-acetyl-glucosamine linked together. Both the glucuronic acid and amino sugar are related to the glucose family. Glucose in its beta configuration allows all of its bulky groups to be in sterically favorable equatorial positions, while all of the small hydrogen atoms occupy the less sterically favorable axial positions. Thus, the structure of the disaccharide is energetically very stable.^35,36 HA is one of the most hydrophilic molecules in nature and can be described as natural moisturizer. As mentioned above, the HA polymer in solution assumes a stiffened helical configuration that can be attributed to hydrogen bonding between the hydroxyl groups along the chain. As a result, a coil structure is formed that traps water $\sim 1000$ times higher than its own weight.³⁴ HA can be obtained by self-assembling of amphiphilic graft copolymers with different polymer molecules, such as polylactic acid, poly-lactide-coglycolic acid chains of PEG and PPG, and others.^69–71 The amphiphilic properties of HA are base for both hydrophilic and lipophilic pathways of penetration into skin. The amphiphilic HA derivatives dissolved in water form self-assemblies in which the near-infrared dyes were tightly packed and arranged to form dimers or H-aggregates. By utilizing HA derivatives as contrast agents, tumors were clearly visualized by optical imaging as well as by OA tomography.⁵²

Because skin is composed of alternating lipophilic and hydrophilic layers, amphipathic compounds that possess both properties may be absorbed by the skin more effectively. Hydration is the most widely used and safest method to increase skin penetration for both hydrophilic and lipophilic permeants. Ideally, amphipathic compounds should minimize the lipophilic–hydrophilic interface to overcome physiological forces in the skin. Studies^19,72 have demonstrated that the injection of hyperosmotic agents into rat dermis could significantly reduce light scattering. However, penetration of glycerol and PEG through intact skin is very minimal and extremely slow because these agents are hydrophilic and penetrate the lipophilic stratum corneum poorly. At the same time, it has been reported^29,33 that OC potential was significantly higher for the lower molecular weight combined PPG- and PEG-based prepolymer mixture as compared to either glycerol or prepolymer alone.

Lipophilic PPG- and hydrophilic PEG-based polymers have a refraction index of 1.47, which is close to that of dermal collagen.²⁹ The OC synergistic effect was shown when all three components, OCA PEG-400 and the two chemical-penetration enhancers thiazone and 1,2-propanediol, were applied together with physical massage.⁷³ These results were confirmed by application of low molecular weight PEG-400 in vivo.⁷³ Our data demonstrate a synergistic effect of OC when an HA, PEG-400, and PPG-425 mixture was used. The unscattered transmission of light through pig skin is significantly higher for the combined PPG- and PEG-based prepolymer mixture than PPG and PEG alone [Fig. 5(a)]. This effect is significantly enhanced if pretreatment of skin with HA was made before application of the polymers [Figs. 5(b) and 5(c)]. The combined effect of the PPG- and PEG-based mixture on light penetration through skin was also evaluated in OA images (Fig. 9). As we found experimentally,⁶⁵ water suspension of GNR with aspect ratio (AR) 3.3 has LSPR around 757 nm, and GNR with AR near 6.0 have LSPR around 1064 and concentrations of GNR-757 and GNR-1064 ~250 and 150 pM for optical density 1.0, respectively. In the current experiments we used GNR concentration of 400 pM for both wavelengths, which corresponds to about $10\mu \mathrm{g}/\mathrm{mL}$ GNR-757 and about $7\mu \mathrm{g}/\mathrm{mL}$ of GNR-1064. At the same time, concentration of ${\mathrm{NiSO}}_4$ was $112\mathrm{mg}/\mathrm{mL}$ or 11,000 greater than GNR-757 and 16,000 times greater than GNR-1064. Comparing effective concentrations of GNR and ${\mathrm{NiSO}}_4$ for similar OD has shown that the molar concentration of GNR is $1.5\times {10}^8$ smaller (0.4 nM and 0.6 M, respectively, see Fig. 9). The sensitivity of LOIS-3-D for GNR visualization in the tubes is about 440 fmole in the tube segment volume of $1.11{\mathrm{mm}}^3$. The tubes were illuminated with combined effective optical fluence of about $2.5\mathrm{mJ}/{\mathrm{cm}}^2$. This is at least eight times smaller than the maximum safe level of optical fluence permitted by American National Standards Institute ($20\mathrm{mJ}/{\mathrm{cm}}^2$). This data demonstrated promising results for preclinical investigation of OCA 3-D high-sensitivity OA tomography by LOUIS 3-D with GNR as contrast agents for multiwavelength applications.

Multifunctional physiologically active HA can change properties not only of the stratum corneum, but also the dermis.⁴⁹ It is known that removal of the stratum corneum dramatically increases skin permeability.⁷⁴ However, removal of the full epidermis increases skin permeability by twice as much, and the epidermis can also offer a significant permeability barrier.⁷⁴ The swelling of scatterers in the dermis can lead light scattered in a more forward direction.⁷³ Alternatively, hygroscopic OCA-induced dehydration may decrease the interfiber spacing, yielding more tightly packed fibers, which allows more constructive interference between scatterers, thereby decreasing the scattering coefficient and increasing light penetration as a result.⁷³

Physical and chemical enhancement of transdermal transport of OCA was discussed in the literature.¹⁵ Ultrasound was used as a physical enhancer of OCA diffusivity, DMSO was used as a chemical enhancer, and glycerol combined with PEG-400 was used as the clearing agent. It was found that simultaneous application of ultrasound, DMSO, and OCA allowed one to reach the upper limit of skin OC. In our current report, we demonstrated that mixture of highly lipophilic PPG and hydrophilic PEG polymers with support by amphiphilic HA demonstrates an optimal base for coupling optical cleaner gel. Alternating lipophilic and hydrophilic layers of skin may absorb amphipathic compounds that possess both properties more effectively. Presented in this report are two steps; the order of OC with pretreatment of skin with HA before OCA application confirmed this mechanism.

HA as an enhancer for OC is unique because it is a biocompatible and naturally biodegradable molecular agent with amphiphilic properties. HA derivatives bearing both hydrophobic (indocyanine green) and hydrophilic (PEG) components were developed, but these substances did not possess the capability of enhancing penetration of light through skin.⁵² To transport drugs effectively through the cellular membrane and deliver them into the intracellular environment, several smart carrier systems based on HA composites were developed containing both synthetic or natural polymers.⁵⁶ We speculate that the method of HA pretreatment described in this work may be also effective for drug delivery. In recent years, HA has emerged as a likely candidate for intracellular delivery of various therapeutic and imaging agents because of its innate ability to recognize specific cellular receptors that are overexpressed on the membrane of abnormal cells.⁵⁶ HA-based conjugates and hydrogels are made through conjugation of HA and functional group molecules.⁵⁶ Hydrogels as HA-based OC composites for increased penetration of light require further investigation in the skin models because properties of HA and OCA may change when incorporated into a hydrogel. However, the potential role of HA in water solutions as an endogenous helper for efficient delivery OCA into skin and tissue is highly promising.

In summary, we proposed and tested a method for enhanced OC of skin and demonstrated its utility for OA imaging. Topical pretreatment of skin for about 30 min with 0.5% HA in aqueous solution offers effective delivery of low molecular weight OCAs such as PPG, PEG, and their mixture. The tissue clearing effect was tested by measuring unscattered light transmission through treated tissue samples, and by detecting increased contrast in OA imaging of test targets through the treated skin. Our formulations for the OA coupling medium were recognized as being nontoxic, hypoallergenic, nonirritant, nonphototoxic, and pharmacologically inert, and they possess high chemical and physical stability.