2.1.

Fractional Photothermolysis Treatment Procedures

The shaved dorsal skin of nude mice (National Laboratory Animal Center, Taipei, Taiwan) was treated by FP through a 1550-nm erbium-doped fiber laser with a roller handpiece (Fraxel® laser SR 1500, Reliant Technologies, Palo Alto, California, USA). Exposures of 10 and 30 mJ, two illumination conditions, were performed (one pass) on mice to create MTZ densities of $326/{\mathrm{cm}}^2$ and $108/{\mathrm{cm}}^2$, respectively. These two energy settings were selected to study the recovery process in superficial dermis (10 mJ) and deeper dermis (30 mJ). In addition, we did not observe adverse side effects to the mice as the MTZs remained in the dermis. These two settings resulted in the same treatment level (coverage area). At Days 0, 7, and 14 following FP treatment, $\sim 1{\mathrm{cm}}^2$ of the skin specimen was excised from the dorsal region of each mouse and stored in 10% formalin solution (Sigma-Aldrich, St. Louis, Missouri) for imaging and histological examinations. The control samples were excised from the nude mouse without FP treatment.

2.2.

Multiphoton Autofluorescence and Second Harmonic Generation Imaging

The MAF and the SHG images of skin were acquired through the use of a homemade multiphoton microscopic system as previously reported.¹⁸ Briefly, skin samples were imaged with a multiphoton laser scanning microscope and with a PlanFlour $20\times /\mathrm{NA}$ 0.75, water-immersion objective lens (Nikon, Tokyo, Japan). The MAF and the SHG signal were excited at 780 nm by a diode solid-state laser-pumped mode-locked femtosecond Ti:sapphire laser (Tsunami; Spectra Physics, Mountain View, California, USA) and the emitted autofluorescence and the SHG signal were detected at the wavelength range of 470 to 635 nm and 380 to 400 nm, respectively. Emitted photons were detected by photomultiplier tubes (R7400P; Hamamatsu, Hamamatsu, Japan). Due to the unevenness of the skin’s inherent structure, we determined the skin surface by first detecting the skin epithelium through the MAF imaging. The skin surface is then defined to be about 10 μm below the topmost layer of the imaged epithelium skin. The images were acquired at depths of approximately 0, 20, 40, 60, 80, and 100 μm from the surface of the skin. All acquired optical images were $228\times 228{\mu \text{m}}^2$ in area with a resolution of $256\times 256\text{pixels}$. To visualize skin samples on a large scale, a specimen translation stage (H101, Prior Scientific Instruments, Cambridge, United Kingdom) was used to translate the specimen after the acquisition of each small area of optical image. In all, 16 adjacent images ($4\times 4$) were acquired and assembled into a larger scale image ($912\times 912{\mathrm{m}}^2$).

2.3.

Image Processing and Analysis

The MPM images were analyzed and reconstructed by use of ImageJ 1.47v (National Institutes of Health, Bethesda, Maryland, USA). This program allows visual inspection of various morphologic features and fluorescence patterns of cellular components within the skin specimens. The SHG signal of skin was determined by setting a threshold at each image to exclude the counts located at the lowest 5% SHG intensity. This value was selected since we found that the SHG signal from one frame is $\sim 20$ times that of the background. Therefore, we used 5% as the cutoff in analyzing the SHG intensity. In this manner, the SHG signal of each image was presented by the percentage of all pixel counts, which is indicative of the second harmonic generating collagen that is present. The data presented are determined from at least five independent images.

2.4.

Histological Examination

Full-thickness dorsal skin was excised and prepared for comparable histological examination. The histological preparation includes specimen fixation in 10% buffered formaldehyde solution followed by dehydration using ethanol. Subsequently, the specimen was embedded in paraffin wax and stained with hematoxylin and eosin.

3.1.

Fractional Photothermolysis Treatment-Induced Damage Revealed by Multiphoton Imaging

To visualize the photothermal damage on mouse skin, we performed ex vivo SHG and MAF imaging immediately after the FP treatment at 10 and 30 mJ. As shown in Figs. 1(a)–1(c), these en face multiphoton images were acquired at intervals of 20-μm depth from the skin surface to a 100-μm depth. Compared to control images, the disappearance of the SHG signal in the multiphoton images as indicated by the yellow arrow, revealed FP treatment-induced damage on skin collagen. The larger damaged area and several smaller damaged areas were observed in FP-treated specimens at 30 and 10 mJ, respectively, which correspond to the Fraxel® laser setup ($108/{\mathrm{cm}}^2$ MTZs at 30 mJ and $326/{\mathrm{cm}}^2$ MTZs at 10 mJ). Localized melting and structural changes (i.e., MTZs) of mouse skin can be seen in the corresponding histological images [Figs. 1(d)–1(f)] following FP treatment, indicating that the disappearance of the SHG signal in the multiphoton images is due to FP treatment-induced collagen damage.

Fig. 1

En face second harmonic generation (SHG) (red) and multiphoton autofluorescence (MAF) (green) images of untreated (a) and fractional photothermolysis (FP)-treated nude mouse skin at 30 mJ (b) and 10 mJ (c) were acquired from the surface to the depth of 100 μm. Dark regions in (b) and (c), indicated by the yellow arrow, are locations of the microthermal zones (MTZs). The histological image (d) to (f) corresponded to SHG (red) and MAF (green) images of untreated and FP-treated nude mouse skin at 30 and 10 mJ, respectively. The yellow arrow indicated the MTZ location caused by FP treatment. The scale bar in (c) and (f) is 200 μm.

En face MPM images along the $z$-axis direction revealed the structures of MTZs at different depths. Moreover, the averaged areas of MTZs at various depths can be quantitatively evaluated from the en face results. The complete MTZs of five images were circled to estimate the damaged areas. Figure 2 show the averaged areas of MTZs at the depths of 20, 40, 60, 80 μm from the surface of the skin. We excluded the results at the depths of 0 and 100 μm due to the clear edges of MTZs at the skin surface and because the deeper depth are difficult to determine. The result showed a trend such that the area of MTZs decreases as the depth increases, which is most likely due to the decay of pulse energy with the increasing depth.

Fig. 2

MTZs areas from FP treatment at 30 and 10 mJ were estimated from en face SHG images at the depth of 20, 40, 60, and 80 μm.

3.2.

Multiphoton Imaging of Skin Rejuvenation After Fractional Photothermolysis Treatment

To evaluate skin rejuvenation after the FP treatment, ex vivo SHG and MAF images were acquired at Days 0, 7, and 14 post FP treatment. Figure 3(a) shows an en face MPM image at a depth of 40 μm at Days 0, 7, and 14 post FP treatment. The decrease in the area of MTZs with time was observed at both energies of 30 and 10 mJ, indicating collagen regeneration following FP treatment. Compared to the FP treatment of 30 mJ, the smaller area of MTZ was found 7 days following an FP treatment at 10 mJ. In the corresponding histological images [Fig. 3(b)], the large and the small MTZs are visible at Day 0 after FP treatment with 30 and 10 mJ. Similar to our MPM images, the MTZs became less significant with time [Fig. 3(b)]. At Days 7 and 14 after the FP treatment of 10 mJ, it is difficult to identify the FP modified regions from the histological images.

Fig. 3

En face SHG (red) and MAF (green) images of mouse skin were acquired at the depth of 40 μm: normal skin (a); treated skin at Days 0 (b), 7 (c), and 14 (d) post FP treatment at 30 mJ; treated skin at Days 0 (e), 7 (f), and 14 (g) post FP treatment at 10 mJ. Histological images corresponded to SHG (red) and MAF (green) images: normal skin (g); treated skin at Days 0 (h), 7 (i), and 14 (j) post FP treatment at 30 mJ; treated skin at Days 0 (k), 7 (l), and 14 (m) post FP treatment at 10 mJ. The yellow arrow indicates the location of MTZ. The scale bar in (g) and (m) is 200 μm each.

3.3.

Quantitative Second Harmonic Generation Analysis in Evaluating the Effect of Fractional Photothermolysis Treatment on Collagen Rejuvenation

The SHG intensity of skin has been used as an index of collagen structural disruption and skin rejuvenation.^9,19,20 For quantitative SHG analysis, each of three mice underwent FP treatment at 30 and 10 mJ, then at least five ex vivo MPM images were acquired at every time point. The percentage of the SHG area in each image was determined to evaluate collagen rejuvenation. To investigate dermal rejuvenation, the percentage of the SHG area was analyzed on normal and FP-treated skin. In Fig. 4, results at the depths of 20 and 80 μm are shown. Since the thickness of mouse epidermis is around 20 μm,²¹ the results at a depth of 20 μm would represent the physiological changes at the upper dermis, whereas that at a depth of 80 μm represents corresponding changes in the deeper dermis. The percentage of the SHG area of normal skin at the depths of 20 and 80 μm are $85\pm 3\%$ and $81.8\pm 1.3\%$ of all pixel counts, respectively. On Day 0 following FP treatment, the SHG signal at the depth of 20 μm decreased to $51.8\pm 9.8\%$ (30 mJ) and $56.8\pm 2.6\%$ (10 mJ). Subsequently, on Day 14, the percentage of the SHG area increased to $68\pm 2.8\%$ (30 mJ) and $77.7\pm 5.1\%$ (10 mJ). A similar trend was observed at the depth of 80 μm as the percentage of SHG area increased from $49.3\pm 9.3\%$ to $64.6\pm 4.2\%$ (30 mJ) and from $57.2\pm 3.8\%$ to $80\pm 1.3\%$ (10 mJ) on Day 14. These results indicated that the collagen regeneration occurred throughout the dermis and can be investigated by multiphoton imaging. Additionally, compared to the FP treatment at 30 mJ, the SHG signal recovered more quickly at 10 mJ.

Fig. 4

Percentage of SHG area was quantitatively analyzed from en face SHG images at 20 μm (a) and 80 μm (b). The gray bar corresponds to the control result. The black and white bars are the results of 30 and 10 mJ, respectively. * indicates that significant difference exits ($p<0.05$, student’s t-test).