A clinical ablative fractional laser (AFL) with adjustable laser power and skin coverage was used to generate microchannel (MC) arrays in the skin. The laser illumination at 5% and 15% skin coverage generated a 9×9 or 14×14 MC array, respectively, in the low dorsal skin of BALB/c mice as illustrated in . Four different laser conditions were tested, including a laser power of 2.5mJ or 5.0 mJ and 5% or 15% skin coverage, simplified as 2.5mJ 5%, 2.5mJ 15%, 5.0mJ 5% or 5.0mJ 15%, respectively. Skin recovery was evaluated following each laser illumination. Among the four laser conditions, only the mildest AFL (2.5mJ 5%) gave rise to complete skin re-epithelialization within 1~2 days as shown by skin histological analysis (). The conical-shaped MCs each spanned from the skin surface to the dermal tissue with 71±7 μm in diameter (d0, low panel, ) and 40±3 μm in depth (d0, upper panel, ). The microthermal zones (MTZs) were larger and clearly visible around MCs, outlined by a black dashed line (d0, ), each with 120±5 μm in diameter and 113±17 μm in depth as a result of heat dissipation. Yet, total laser-ablated area represented as little as 1.7% of the skin surface involved. In other words, the majority of the skin tissue in the array was spared from the laser-mediated ablation, which might be the reason for fast re-epithelialization and growth of new tissue into MCs in 24 hrs (d1, ). Indeed, most of MCs were healed and filled with newly synthesized tissue that was easily distinguished from the surrounding tissue (d1, ). In the following day, all MCs were completely restored to normal, indistinguishable from those surrounding tissues (d2, ). An increase in skin coverage to 15% generated MCs in a similar size, but resulted in delayed skin recovery, probably due to a reduced ratio of healthy to ablative tissues (data not shown). Likewise, a high laser power of 5.0mJ increased the size and depth of each MC, causing delayed skin recovery at both 5% and 15% skin coverage (data not shown). A further increase in laser energy to 10.0mJ or higher caused instant skin shrinkage and damage even at the smallest skin coverage (data not shown). The histological studies suggest that skin recovery ability is controlled by a ratio of ablative and non-ablative tissues and AFL 2.5mJ/0.5% can generate MCs with a quick skin recovery.
Quick recovery of AFL-treated skin
Similarly, while mild tape stripping (1, 15) effectively ablated SC layer with little damage on the epidermal layer underneath the SC layer, full restoration of SC layer was observed within 1 or 2 days (). A slightly invasive tape stripping (3, 6) damaged not only SC layer but also the upper epidermal tissue (), leaving only one-cell thickness of epidermal tissue, in contrast to several-cell thickness of epidermal tissue in control skin. This slightly invasive treatment delayed restoration of epidermal tissue (d1, low panel, ) and thickened epidermal tissue (d2, low panel, ), concurrent with dermal infiltration of large amounts of inflammatory cells (upper panel, ), indicative of incomplete skin recovery by day 2. More harsh tape stripping such as 10-12 strokes with a tape change every other stripping caused skin redness and damage instantly due to damage of the dermal tissue (data not shown). The studies corroborate that tape stripping (1, 15) was safe with a quick recovery, while tape stripping (3, 6) was slightly invasive with delayed skin recovery.
Transcutaneous delivery of SRB and MB
SRB, a hydrophilic model drug about 560Da in mass, was initially used to evaluate transcutaneous delivery in the skin treated with AFL or tape stripping. As shown in , the compound entered laser-treated skin through the MCs 30 min after topical application of a SRB-coated gauze patch, as evidenced by the presence of strong pink-colored SRB only in AFL-treated skin, but not in control skin, with a highest density in laser-generated MCs. Even in the vicinity of each MC, SRB was higher than that in control skin, presumably resulting from drug dispersing from the MC. There was only a trace amount of SRB in the control and tape stripping-treated skin, with a slightly more SRB in the skin treated by slightly invasive tape stripping (3, 6) as compared to the mild tape stripping (1, 15) (). Bright-field microscopy and fluorescence imaging of skin sections at 5 and 30 minutes corroborated that strong SRB signal was radically diffusing from laser-generated MCs into the surrounding dermal tissue in 5 minutes after patch application, and continuously spreading over the entire dermal tissue in 30 minutes (). In contrast, SRB was presented only on the skin surface in control or in the epidermis at these time points in tape stripping (1, 15) group. Although tape stripping (3, 6) did facilitate penetration of the drug to the upper dermal tissue in 30 minutes, the efficacy was much lower than AFL.
A time-course study found that the amount of SRB in the skin increased linearly to 5.83±0.44, 15.17±1.73, 10.36±0.72 or 31.37±4.72 μg/cm2 after 30 minutes of patch application to the skin treated with AFL 2.5mJ 5%, 2.5mJ 15%, 5.0mJ 5%, or 5.0mJ 15%, respectively. This represented a 12, 30, 21 or 63-fold increase compared to control group (0.50±0.17 μg/cm2, , p<0.001). The amount of SRB in tape stripping (1, 15) and (3, 6) groups was only 1.09±0.24 or 3.09±0.17 μg/cm2, respectively, and was without statistic significance compared to the control group at this time point. The amount of SRB in the skin peaked at 9.62±1.06 μg/cm2 in 2.5mJ 5% group and 15.91±1.10 in 5.0mJ 5% group after 6 hrs of patch application, while SRB amount was a highest level at 25.18±2.90 μg/cm2 in 2.5mJ 15% group and 31.37±4.72 μg/cm2 in 5.0mJ 15% group in 1 or 0.5 hr of patch application, respectively. Thus, a peaking time appeared to inversely correlate with MC density, while the efficiency might be controlled by a combination of MC density and laser power. SRB content in the skin diminished gradually after the peak in all laser groups (), presumably due to entry of the compound into the circulation system through lymphatic or capillary networks in the dermis.
In comparison, tap stripping displayed much diminished efficacies and lacked an overt peak in which a high level of SRB in the skin was around 2.20±0.23 μg/cm2 in tape stripping (1, 15) group or 4.81±1.23 μg/cm2 in tape stripping (3, 6) group spanning from 3 to more than 9 hrs after patch application. There was no significant difference between tape stripping (1, 15) and control group or between the two tape stripping groups at each time point, although tape stripping (3, 6) increased SRB delivery significantly at 3 and 6 hrs as compared to control group (p<0.05).
A similar, superior AFL-enhanced transcutaneous drug delivery was also demonstrated with methylene blue (MB), a hydrophilic photosensitizer with 320Da molecular weight. As shown in , a highly intense blue-colored MB was seen in a pattern matching laser-generated MCs after 30 minutes of patch application. MB was diffusing from the lateral surface of the MCs and sufficiently penetrated deep into the dermal tissue in AFL group (). In contrast, only a trace amount of MB was presented on skin surface of tape stripping (1, 15) and control groups, which was much less than MB deposited on skin surface treated with tape stripping (3, 6) or AFL (). Consistent with this was good penetration of MB into the epidermis in tape stripping (3, 6) group, whereas no MB was able to enter the control skin and only a small amount of MB reached the epidermis after mild tape stripping (). Quantification of skin MB at varying times showed that MB levels sharply rose in AFL-treated skin, peaking at 10~15 μg/cm2 under 2.5mJ/15%, 5mJ/5% and 5mJ/15% conditions or 8 μg/cm2 under 2.5mJ/5% at 6 or 9 hrs after patch application, respectively (). This high level of MB delivery was in marked contrast to a slow deposition of MB with a maximal level less than 1.0 μg/cm2 in mild tape stripping and control skins (p<0.001). Although more invasive tape stripping (3, 6) significantly enhanced MB delivery at and after 3 hrs as compared to tape stripping (1, 15) or control group (p<0.05), the efficacy was far lower than AFL in all the conditions tested. These imaging and quantitative studies of two different drug systems suggest clearly that AFL is superior to tape stripping to facilitate transcutaneous drug delivery.
Transcutaneous delivery of OVA
We next addressed whether AFL could also enhance transcutaneous delivery of a model vaccine antigen, ovalbumin (OVA) that is much higher in mass (~45 kDa), impermeable to the skin. When Texas Red-conjugated OVA (TR-OVA) was coated onto a patch and topically applied to laser-treated skin for 0.5 or 3 hrs, TR-OVA was observed, after 0.5 hr of patch application, in both the epidermis and in the upper dermal tissue surrounding MCs in laser-treated skin (). It continued spreading horizontally and vertically into the entire dermal tissue in 3 hrs ().
There was no dermis-penetration of TR-OVA in both tape stripping and control groups in the initial 0.5 hr of patch application, although some TR-OVA penetration of the upper dermis was detected in 3 hrs after patch application with more invasive tape stripping (3, 6) ().
The amount of TR-OVA in the skin climbed precipitously in laser-treated skin following patch application and reached a peaking level in 3 hrs in all AFL-treated groups (), which was significantly higher than that obtained with tape stripping or control group (, p<0.001). In comparison with the control group, laser treatment increased TR-OVA transcutaneous delivery by approximate 8~15 times within 3 hrs (p<0.001), whereas slightly invasive tape stripping (3, 6), but not mild tape stripping (1, 15), increased TR-OVA delivery only by less than 4 times in the same period of time (p<0.05). There was no significant difference between the two tape stripping groups or between each tape stripping and control groups at each time point (). Thus, AFL is able to facilitate transcutaneous delivery of protein antigen OVA in a much more efficient manner than tape stripping.
Antigen uptake following OVA delivery
We went on to address whether the photothermal effect of AFL could compromise the capacity of skin APCs for processing and presentation of antigens in the vicinity of the MCs, in view of an importance of these processes in induction of a potent immune response [23
]. To this end, Alexa Fluor 647-conjugated OVA (AF647-OVA) was coated onto a gauze patch and topically applied to one ear after AFL exposure at the mildest condition in MHC II-EGFP transgenic mice in which epidermal Langerhans cells (LCs) and dDCs were GFP-labeled [22
]. Ear skin is relatively thin and less tolerable to photothermal effect and thus the mildest AFL was applied in this analysis. The other ear was left untreated before patch application as a control. As shown in , AF647-OVA appeared to sufficiently enter laser-generated MCs as reflected by the same pattern of OVA distribution as AFL-generated MC arrays, in contrast to a background level of AF647-OVA seen in the control ear. With the facilitation of the MCs, OVA could travel radically into the vicinity of MCs. It is conceivable that LCs in the epidermis and DCs in the dermis both could capture OVA in AFL-treated skin should their functions not be adversely affected by AFL treatment. When the control ear was imaged by intravital confocal microscopy, focusing first on the epidermis (, left) and then on the dermis (, right), OVA-uptake by LCs or dDCs was hardly seen even with an extensive search (, top panel). OVA fluorescence in the epidermal layer was mostly out of focus and probably not truly co-localized with LCs due to its poor penetration of epidermis as suggested by histology data (). The dermis was largely devoid of OVA except for some non-specific fluorescence debris. In contrast, antigen (Ag) positive LCs and dDCs were readily visible in AFL-treated ear (middle panel) in the vicinity of the MCs. One enlarged antigen-captured single APC demonstrated the presence of OVA inside the cell as evidenced by merged yellow fluorescence of green (APCs) and red (OVA) (lower panel), in contrast to the red fluorescence of OVA outside APCs.
Antigen uptake and LN-drainage following OVA delivery
Besides local antigen uptake, transportation of free or antigen-captured APCs to the draining lymph node is equally important for stimulating a potent immune response [23
]. We found a high amount of OVA in auricular lymph nodes draining AFL-treated ear and little OVA in those lymph nodes draining control ear (). The data confirmed that functions of afferent lymphatic vessels and APCs were well preserved in the vicinity of the MCs in AFL-treated skin.
AFL-enhanced transcutaneous immunization
In accordance with the sufficient local antigen uptake and transportation of antigen to the draining lymph nodes, AFL treatment augmented the production of OVA-specific antibody by 28-534 times at weeks and 53-545 times at 4 weeks after patch-based OVA immunization as compared to the control group (). In particular, AFL 5.0mJ 15% stimulated a highest level (1068±445) of antibody production compared to other AFL conditions (124±50, 56±14, and 71±29 in 2.5mJ 5%, 2.5mJ 15%, and 5.0mJ 5% conditions, respectively) (, p<0.001). Tape stripping also increased OVA-specific antibody production by 5-8 times as compared to control group, but it was significantly lower than all AFL conditions tested (p<0.001, ), approximately 4-82 times lower at 2 weeks and 8-105 times lower at 4 weeks than AFL treatment groups. To our surprise, there was no significant difference between the two tape stripping groups. We conclude that AFL treatment followed by topical application of vaccine-coated patch can sufficiently deliver vaccines into the skin and greatly potentiate a vaccine-specific immune response.
Immune responses induced by AFL-facilitated transcutaneous delivery of OVA