SCANNING VOL. 37, 1–5 (2015) © Wiley Periodicals, Inc.
Quantitative Analysis on Collagen of Dermatofibrosarcoma Protuberans Skin by Second Harmonic Generation Microscopy SHULIAN WU,1 YUDIAN HUANG,2 HUI LI,1 YUNXIA WANG,1 AND XIAOMAN ZHANG1 1
Key Lab of OptoElectronic Science and Technology for Medicine, Ministry of Education, Fujian Provincial Key Lab of Photonic Technology, College of Photonic and Electronic Engineering, Fujian Normal University, Fuzhou, Fujian, China 2 Pathology Department of the Fuzhou First Affiliated Hospital of Fujian Medical University, Fuzhou, Fujian, People‘s Republic of China
Summary: Dermatofibrosarcoma protuberans (DFSP) is a skin cancer usually mistaken as other benign tumors. Abnormal DFSP resection results in tumor recurrence. Quantitative characterization of collagen alteration on the skin tumor is essential for developing a diagnostic technique. In this study, second harmonic generation (SHG) microscopy was performed to obtain images of the human DFSP skin and normal skin. Subsequently, structure and texture analysis methods were applied to determine the differences in skin texture characteristics between the two skin types, and the link between collagen alteration and tumor was established. Results suggest that combining SHG microscopy and texture analysis methods is a feasible and effective method to describe the characteristics of skin tumor like DFSP. SCANNING 37:1–5, 2015. © 2014 Wiley Periodicals, Inc. Key words: Dermatofibrosarcoma protuberans skin, second generation microscopy, texture feature
Contract grant sponsor: National Natural Science Foundation of China; Contract grant number: 61178089; Contract grant sponsor: Program for Changjiang Scholars and Innovative Research Team in University; Contract grant number: IRT1115; Contract grant sponsor: Natural Science Foundation of Fujian Province; Contract grant number: 2014J01226. Conflicts of interest: None. Shulian Wu and Yudian Huang contributed equally to this article. Address for reprints: Hui Li, College of Photonic and Electronic Engineering, Fujian Normal University, Fuzhou 350007, People’s Republic of China. E-mail: [email protected] Received 1 August 2014; revised 24 September 2014; Accepted with revision 6 October 2014 DOI: 10.1002/sca.21172 Published online 4 November 2014 in Wiley Online Library (wileyonlinelibrary.com).
Introduction Dermatofibrosarcoma protuberans (DFSP) is a rare type of skin cancer (Mendenhall et al., 2004; Llombart et al., 2013). It usually appears as a thickened area of the skin on the trunk areas, such as the chest and shoulders. However, DFSP can also occur on the limbs and on the head and neck in some cases (Mark et al., 1993). It develops in the deep layer of the skin (usually occur on the dermis layer). Though the process of DFSP skin grows slowly, it has an extensive and a high recurrence rate if misdiagnosed or incomplete removed by surgery. DFSP skin rarely spreads to other parts of the body, so it can be cured if completely removed with a wide margin of normal tissue. Trauma or injury of the skin may be considered as the predisposing factors for DFSP skin. The pathogenesis of this type skin cancer is still not clear at present. (Bowne et al., 2000; Rubin et al., 2002; McArthur, 2004). The rare occurrence of DFSP resulted in being delayed diagnosis. Moreover, this disease may be mistaken for many other skin conditions (such as some benign diseases). So, in clinic, if DFSP is suspected, a piece of abnormal skin is removed and biopsied to confirm the diagnosis. The golden standard for cutaneous diagnosis is histopathological examination with hematoxylin and eosin staining, but this technique is invasive and time consuming. Considering the disorder in collagen, the alteration of collagen in the dermis to obtain structural and diagnostic information could be visualized by high resolution microscopy systems (Rubin et al., 2002; McArthur, 2004; Mendenhall et al., 2004). In recent years, second harmonic generation (SHG) microscope has become a powerful tool for studying in many biomedical research fields, including dermatological skin applications with its many advantages over traditional optical imaging, including high resolution, penetration depth, inherent optical sectioning, and low phototoxicity (Denk et al., 1990; Campagnola, 2011; Ghazaryan et al., 2012; Yew et al., 2013). It has reported
2
SCANNING VOL. 37, 1 (2015)
that the detectable depth of multiphoton microscopy system has achieved 1.6 mm in the mouse cortex using the excitation wavelength of 1280 nm (Kobat et al., 2011). SHG microscope is a nondestructive imaging tool that allows clear visualization of collagen fibers throughout the connective tissue, and without staining or pretreatment is required (Koehler et al., 2009; Chen et al., 2012; Toki et al., 2013). In the study, a noninvasive real-time imaging tool is required to monitor the structure alteration of collagen in dermis, and then to character quantitatively the molecular dynamics in DFSP skin process. To the best of our knowledge, the study utilized SHG microscope to assess DFSP collagen texture has been not available in previous studies. In this study, SHG microscope was applied to obtain images of the normal and DFSP skin in the same parameters condition, and subsequently, the images were analyzed by mathematical analysis software to describe the changes on spectrum characteristic and texture feature of dermal collagen. The connection of collagen between DFSP skin and normal skin were established.
Materials and Methods Samples
Ten DFSP skin samples were obtained from ten patients undergoing excision surgical procedures aged 51–62 years old. Eleven normal skin samples were collected during aesthetic plastic surgery, from patients aged 48 years old to 60 years old. For comparison, healthy skin specimens were also collected from the same parts of body with DFSP samples. Normal and DFSP skin samples were investigated to demonstrate the possibility of SHG microscopy to characterize the pathological state and evaluate therapeutic efficacy. Our protocol was in accordance with the standard approved by the Institutional rules governing clinical investigation of human subjects in biomedical research, and informed consent
was obtained from the patients. Tissue samples were snap-frozen in liquid nitrogen (–196˚C) immediately after excision from patients. Before the experimental, the samples were cut into 60 mm thickness and then placed between the microscope slide and cover glass. Moreover, a small amount of PBS solution was dripped into the tissue specimen for avoiding dehydration or shrinkage during the imaging process. The samples were processed with the cover glass on the microscope objective, and images of the dermis layer in different position in each sample were obtained.
Second Harmonic Generation Microscopic System
The SHG microscopic system used in this study is a commercialized technique contains a high-throughput scanning inverted Axiovert 200 microscope (LSM 510 META; Zeiss, Germany) and mode-locked femtosecond Ti :sapphire laser (110 fs, 76 MHz), tunable from 700 to 980 nm (Mira 900-F; Coherent, America), as described previously (Wu et al., 2011; Zhu et al., 2013) . Briefly, the excitation wavelength at 810 nm was chosen in this experiment and the average power arrived the sample was 10 mW. All the images were obtained by SHG microscope with an oil immersion objective of Plan– Apochromat 63 (Numerical aperture ¼ 1.4, Zeiss) and of 512 512 pixels with a size of 921 mm 921 mm. The SHG signal was detected at 405 nm in center with a bandwidth of 20 nm. Ten images from region of interest on different positions in each sample were selected for quantitative analysis. Collagen Images Processing Method
The collagen fibril bundles orientation was extracted by the Fast Fourier transform (FFT) method, which is offered from a powerful operating system (online openaccessed at http://rsb.info.nih.gov/ij). The FFT is a good method for using an image to calculate the
Fig 1. The FFT analysis models (a) the plot of FFT in SHG image; (b) 3D images the FFT plot.
S. Wu et al.: Quantitative analysis on collagen of dermatofibrosarcoma
3
Fig 2. The morphological of collagen in different type skin, (a) the normal skin, the collagen bundle is dense, and the texture is distinct; (b) the DFSP skin, the loss collagen caused the texture is obscure. The error bar is 20 mm.
collagen-fibril bundle orientation. So, many researchers have widely used this software to handle experimental data or images in their studies (van Zuijlen et al., 2003; Wu et al., 2011). FFT mainly represents all frequencies present in an image by using the power plot, which shows the values from the orthogonal direction of the collagen-fibril bundles in detail, and the plot will have an elliptical behavior because of a set of aligned collagen-fibril bundles in an image. The plot of FFT image will show a circular behavior in an image with randomly oriented collagen-fibril bundles. The ratio of short to long axe of the generated power plot of images was used to estimate the collagen orientation index (OI), which was calculated with the expression: OI ¼ [1– (short/long ratio)], which was shown in Figure 1(a) (Wu et al., 2011). The OI value of perfectly random tissue is “0,” which indicated isotropic behavior, and the FFT plot image was circular. A Fourier analysis with parallel collagen orientation yielded an elongated power that resulted in a longer orientation index (the maximum is “1”). Furthermore, a parameter collagen bundle packing (BP), which is expressed as BP ¼ 512 (1/h), where h is the pixel distances between the centers of gravity of two first-order maxima in 3D images were shown in Figure. 1 (b) (Wu et al., 2011). Ten images of each sample in different region were selected for quantitative analysis. The experimental data were analyzed by statistical test using SPSS 15.0 software (SPSS Inc., USA). Statistical significances of the data in different skin types were evaluated by T-test, which is used to determine whether or not significant differences exist between the means of two independent groups. In our study, the T-test analysis was used to analyze the variance of collagen intensity, collagen orientation index, and collagen bundle packing with different type’s skin. The differences were considered statistically significant at P < 0.05.
Results and Discussion Morphological Alteration in Collagen
The image of collagen morphology where is the strongest intensity in detectable depth in normal skin and DFSP skin is shown in Figure 2, respectively. The normal collagen had a number of compact collagen fibrils. The denser collagen fibrils formed into the big collagenous and thickening collagen bundles. The collagen arrangement was more organized from the Figure 2 (a). In contrast, From the Figure 2 (b), it shown that DFSP skin had thin collagen fibrils, the collagen matrix was slightly looser, and the collagen bundle appeared less organized, which implies loss of collagen structures. It demonstrated that the collagen had been destroyed during the process of DFSP. This
Fig 3. The SHG intensity with emission wavelength in different skin. In centre wavelength of second harmonic generation in 405 nm, the intensity of normal skin is much higher than the DFSP skin.
4
SCANNING VOL. 37, 1 (2015)
method was suitable for visualizing the precise status of the dense matrix and the bundles of the collagen. The level of collagen injured also can be distinct observed according to the SHG images.
Collagen Intensity
Previous reports have indicated that the intensity of change of SHG is a good predictor of collagen structural disruption (Sun et al., 2006). In this study, the SHG spectrum of skin collagen between normal and DPSF skin were obtained by lambda scanning mode of SHG
microscope, and the average data were normalized to the maximal peak intensity, as shown in Figure 3. The excitation laser wavelength was 810 nm, and thus, the peak intensity of SHG was at 405 nm. From the figure, the normalized intensity of SHG in normal skin was approximately 1, whereas the value of DFSP skin was only close to 0.4 at the emission wavelength of 405 nm. The intensity value in normal skin was about 2.5 times compared with the intensity value in the DFSP skin. It indicated that the intensity value of SHG decreased sharply from normal to DFSP skin. The phenomenon indicated that the skin was injured, and the collagen fribrils was destroyed, and then results in the loss of large collagen fibrils. The collagen fibrils were sparse at regular intervals distance of DFSP skin, whereas the collagen fibrils were dense in normal skin. As shown in Figure 4, the normalized value of the collagen intensity at 0.5 was considered as the reference point to analyze the collagen intensity changes with the collagen distance. More than 95% of collagen fibril intensities were more than 0.5 in the normalized value and the fibrils were side by side in normal skin. However, in the DFSP skin, more than 95% of collagen fibril intensities were less than 0.5. The intensity of DFSP collagen was much lower than that in normal skin. Moreover, the distances among collagen fibrils in non-normal skin were wider than those in normal skin according to the figure. The intensity of SHG signals varied greatly with distance in the two skin states.
Collagen Texture Feature
FFT method was employed to evaluate the OI and BP of collagen by Image J software. The results of OI is displayed in Figure 5. From the figure, we could see that
Fig 4. The SHG intensity from collagen images. (a) the DFSP image and its collagen intensity. (b) the intensity of collagen with distances between normal skin and DFSP skin. The intensity of normal skin is much higher than the DFSP skin. They have clear dividing line. It indicates the gap between two collagen bundles.
Fig 5. Bar chart demonstrating mean the value of collage OI ( SD) between normal skin and DFSP skin. P > 0.05. It means that the value has not statistical significance in two type skin.
S. Wu et al.: Quantitative analysis on collagen of dermatofibrosarcoma
5
demonstrated that the OI value was almost unchanged between normal skin and DFSP skin, whereas the BP value was higher in DFSP skin than in normal skin. This finding indicated that collagen degenerates in all directions and that collagen fibril disperses with the degeneration of collagen. But the distance of collagen fibril was obvious in DFSP skin. These results demonstrated that the SHG microscopy combined the texture analysis method could help us understand more information about DFSP.
References
Fig 6. Bar chart demonstrating mean the value of collage BP ( SD) in normal skin and DFSP skin, P < 0.05. It means that the value has statistical significance in two type skin.
the collagen-fibril bundles orientation index was not greatly significant between normal and DFSP skin. The value of OI is 0.65 0.09 for normal skin, whereas the value is 0.61 0.07 for DFSP skin. The results indicated that the alignment of collagen fibrils between normal skin and DFSP skin has not great statistical significance, which may be attributed to the older age of the patients with higher collagen OI in chronological aging skin (Wu et al., 2011). The distribution of collagen in two types skin had more ellipses shape in FFT analysis because the samples were obtained from the older people. However, the collagen BP differed greatly between two types skin from the Figure 6. The value of BP was 18.4 1.1 for normal skin, whereas the value was 23.5 0.9 for DFSP skin. These results confirmed that the collagen fibril bundles gradually decreased in all directions with dermis lesion progression.
Conclusions In this study, SHG was employed to evaluate the collagen alterations of normal skin and DFSP skin. The collagen structure, collagen quantity, and SHG signal intensities were obtained for analyzing the normal skin and DFSP skin. The outcomes indicate that the collagen of DFSP skin was degenerated, which was obvious reflected in the thinner collagen fibril, the weaker collagen density, and the larger collagen distance. Moreover, some collagen texture, such as the collagen orientation indices, and collagen fibril bundle were extracted from SHG images for quantitative charactering the alteration of collagen texture features between the two types skin through the FFT analysis. Our results
Bowne WB, Antonescu CR, Leung DH, et al. 2000. Dermatofibrosarcoma protuberans. Cancer 88:2711–2720. Campagnola P. 2011. Second harmonic generation imaging microscopy: applications to diseases diagnostics. Anal. Chem 83:3224–3231. Chen X, Nadiarynkh O, Plotnikov S, Campagnola PJ. 2012. Second harmonic generation microscopy for quantitative analysis of collagen fibrillar structure. Nat Protoc 7:654–669. Denk W, Strickler JH, Webb WW. 1990. Two-photon laser scanning fluorescence microscopy. Science 248:73–76. Ghazaryan AA, Hu PS, Chen SJ, Tan HY, Dong CY. 2012. Spatial and temporal analysis of skin glycation by the use of multiphoton microscopy and spectroscopy. J Dermatol Sci 65:189–195. Kobat D, Horton GN, Xu C. 2011. In vivo two-photon microscopy to 1.6-mm depth in mouse cortex. J. Biomed. Opt. 16:1–4. Koehler MJ, Preller A, Kindler N, Elsner P, et al. 2009. Intrinsic, solar and sunbed-induced skin aging measured in vivo by multiphoton laser tomography and biophysical methods. Skin.Res.Technol. 15:357–363. Llombart B, Serra-Guille´n C, Monteagudo C, Lo´pez Guerrero JA, Sanmartı´n O. 2013. Dermatofibrosarcoma protuberans: a comprehensive review and update on 2013. Elsevier pp 13– 28. Mark RJ, Bailet JW, Tran LM, et al. 1993. Dermatofibrosarcoma protuberans of the head and neck: a report of 16 cases. Arch. Otolaryngol.—Head & Neck Surg 119:891. McArthur G. 2004. Molecularly targeted treatment for dermatofibrosarcoma. Elsevier. p 30–36. Mendenhall WM, Zlotecki RA, Scarborough MT. 2004. Dermatofibrosarcoma protuberans. Cancer 101:2503–2508. Rubin BP, Schuetze SM, Eary JF, et al. 2002. Molecular targeting of platelet-derived growth factor B by imatinib mesylate in a patient with metastatic dermatofibrosarcoma protuberans. J. Clin. Oncol. 20:3586–3591. Sun Y, Chen WL, Lin SJ, et al. 2006. Investigating mechanisms of collagen thermal denaturation by high resolution secondharmonic generation imaging. Biophys. J. 91:2620–2625. Toki F, Honkura N, Shirakata Y, et al. 2013. Second harmonic generation reveals collagen fibril remodeling in fibroblastpopulated collagen gels. Cell Struct Funct 38:227–236. van Zuijlen PP, Ruurda JJ, van Veen HA, et al. 2003. Collagen morphology in human skin and scar tissue: no adaptations in response to mechanical loading at joints. Burns 29:423–431. Wu S, Li H, Yang H, et al. 2011. Quantitative analysis on collagen morphology in aging skin based on multiphoton microscopy. J. Biomed. Opt. 16:040502–0405 02. Yew E, Rowlands C, So PT. 2013. Application of Multi-Photon Microscopy in Dermatological Studies: A Mini-Review. J. Innov. Opt. Health Sci. 06:1330010. Zhu X, Lin L, Yu H, et al. 2013. Visualization of Epidermal and Dermal Alteration in Papulonodular Mucinosis by Multiphoton Microscopy. Scanning 35:22–27.
Quantitative Analysis on Collagen of Dermatofibrosarcoma Protuberans Skin by Second Harmonic Generation Microscopy SHULIAN WU,1 YUDIAN HUANG,2 HUI LI,1 YUNXIA WANG,1 AND XIAOMAN ZHANG1 1
Key Lab of OptoElectronic Science and Technology for Medicine, Ministry of Education, Fujian Provincial Key Lab of Photonic Technology, College of Photonic and Electronic Engineering, Fujian Normal University, Fuzhou, Fujian, China 2 Pathology Department of the Fuzhou First Affiliated Hospital of Fujian Medical University, Fuzhou, Fujian, People‘s Republic of China
Summary: Dermatofibrosarcoma protuberans (DFSP) is a skin cancer usually mistaken as other benign tumors. Abnormal DFSP resection results in tumor recurrence. Quantitative characterization of collagen alteration on the skin tumor is essential for developing a diagnostic technique. In this study, second harmonic generation (SHG) microscopy was performed to obtain images of the human DFSP skin and normal skin. Subsequently, structure and texture analysis methods were applied to determine the differences in skin texture characteristics between the two skin types, and the link between collagen alteration and tumor was established. Results suggest that combining SHG microscopy and texture analysis methods is a feasible and effective method to describe the characteristics of skin tumor like DFSP. SCANNING 37:1–5, 2015. © 2014 Wiley Periodicals, Inc. Key words: Dermatofibrosarcoma protuberans skin, second generation microscopy, texture feature
Contract grant sponsor: National Natural Science Foundation of China; Contract grant number: 61178089; Contract grant sponsor: Program for Changjiang Scholars and Innovative Research Team in University; Contract grant number: IRT1115; Contract grant sponsor: Natural Science Foundation of Fujian Province; Contract grant number: 2014J01226. Conflicts of interest: None. Shulian Wu and Yudian Huang contributed equally to this article. Address for reprints: Hui Li, College of Photonic and Electronic Engineering, Fujian Normal University, Fuzhou 350007, People’s Republic of China. E-mail: [email protected] Received 1 August 2014; revised 24 September 2014; Accepted with revision 6 October 2014 DOI: 10.1002/sca.21172 Published online 4 November 2014 in Wiley Online Library (wileyonlinelibrary.com).
Introduction Dermatofibrosarcoma protuberans (DFSP) is a rare type of skin cancer (Mendenhall et al., 2004; Llombart et al., 2013). It usually appears as a thickened area of the skin on the trunk areas, such as the chest and shoulders. However, DFSP can also occur on the limbs and on the head and neck in some cases (Mark et al., 1993). It develops in the deep layer of the skin (usually occur on the dermis layer). Though the process of DFSP skin grows slowly, it has an extensive and a high recurrence rate if misdiagnosed or incomplete removed by surgery. DFSP skin rarely spreads to other parts of the body, so it can be cured if completely removed with a wide margin of normal tissue. Trauma or injury of the skin may be considered as the predisposing factors for DFSP skin. The pathogenesis of this type skin cancer is still not clear at present. (Bowne et al., 2000; Rubin et al., 2002; McArthur, 2004). The rare occurrence of DFSP resulted in being delayed diagnosis. Moreover, this disease may be mistaken for many other skin conditions (such as some benign diseases). So, in clinic, if DFSP is suspected, a piece of abnormal skin is removed and biopsied to confirm the diagnosis. The golden standard for cutaneous diagnosis is histopathological examination with hematoxylin and eosin staining, but this technique is invasive and time consuming. Considering the disorder in collagen, the alteration of collagen in the dermis to obtain structural and diagnostic information could be visualized by high resolution microscopy systems (Rubin et al., 2002; McArthur, 2004; Mendenhall et al., 2004). In recent years, second harmonic generation (SHG) microscope has become a powerful tool for studying in many biomedical research fields, including dermatological skin applications with its many advantages over traditional optical imaging, including high resolution, penetration depth, inherent optical sectioning, and low phototoxicity (Denk et al., 1990; Campagnola, 2011; Ghazaryan et al., 2012; Yew et al., 2013). It has reported
2
SCANNING VOL. 37, 1 (2015)
that the detectable depth of multiphoton microscopy system has achieved 1.6 mm in the mouse cortex using the excitation wavelength of 1280 nm (Kobat et al., 2011). SHG microscope is a nondestructive imaging tool that allows clear visualization of collagen fibers throughout the connective tissue, and without staining or pretreatment is required (Koehler et al., 2009; Chen et al., 2012; Toki et al., 2013). In the study, a noninvasive real-time imaging tool is required to monitor the structure alteration of collagen in dermis, and then to character quantitatively the molecular dynamics in DFSP skin process. To the best of our knowledge, the study utilized SHG microscope to assess DFSP collagen texture has been not available in previous studies. In this study, SHG microscope was applied to obtain images of the normal and DFSP skin in the same parameters condition, and subsequently, the images were analyzed by mathematical analysis software to describe the changes on spectrum characteristic and texture feature of dermal collagen. The connection of collagen between DFSP skin and normal skin were established.
Materials and Methods Samples
Ten DFSP skin samples were obtained from ten patients undergoing excision surgical procedures aged 51–62 years old. Eleven normal skin samples were collected during aesthetic plastic surgery, from patients aged 48 years old to 60 years old. For comparison, healthy skin specimens were also collected from the same parts of body with DFSP samples. Normal and DFSP skin samples were investigated to demonstrate the possibility of SHG microscopy to characterize the pathological state and evaluate therapeutic efficacy. Our protocol was in accordance with the standard approved by the Institutional rules governing clinical investigation of human subjects in biomedical research, and informed consent
was obtained from the patients. Tissue samples were snap-frozen in liquid nitrogen (–196˚C) immediately after excision from patients. Before the experimental, the samples were cut into 60 mm thickness and then placed between the microscope slide and cover glass. Moreover, a small amount of PBS solution was dripped into the tissue specimen for avoiding dehydration or shrinkage during the imaging process. The samples were processed with the cover glass on the microscope objective, and images of the dermis layer in different position in each sample were obtained.
Second Harmonic Generation Microscopic System
The SHG microscopic system used in this study is a commercialized technique contains a high-throughput scanning inverted Axiovert 200 microscope (LSM 510 META; Zeiss, Germany) and mode-locked femtosecond Ti :sapphire laser (110 fs, 76 MHz), tunable from 700 to 980 nm (Mira 900-F; Coherent, America), as described previously (Wu et al., 2011; Zhu et al., 2013) . Briefly, the excitation wavelength at 810 nm was chosen in this experiment and the average power arrived the sample was 10 mW. All the images were obtained by SHG microscope with an oil immersion objective of Plan– Apochromat 63 (Numerical aperture ¼ 1.4, Zeiss) and of 512 512 pixels with a size of 921 mm 921 mm. The SHG signal was detected at 405 nm in center with a bandwidth of 20 nm. Ten images from region of interest on different positions in each sample were selected for quantitative analysis. Collagen Images Processing Method
The collagen fibril bundles orientation was extracted by the Fast Fourier transform (FFT) method, which is offered from a powerful operating system (online openaccessed at http://rsb.info.nih.gov/ij). The FFT is a good method for using an image to calculate the
Fig 1. The FFT analysis models (a) the plot of FFT in SHG image; (b) 3D images the FFT plot.
S. Wu et al.: Quantitative analysis on collagen of dermatofibrosarcoma
3
Fig 2. The morphological of collagen in different type skin, (a) the normal skin, the collagen bundle is dense, and the texture is distinct; (b) the DFSP skin, the loss collagen caused the texture is obscure. The error bar is 20 mm.
collagen-fibril bundle orientation. So, many researchers have widely used this software to handle experimental data or images in their studies (van Zuijlen et al., 2003; Wu et al., 2011). FFT mainly represents all frequencies present in an image by using the power plot, which shows the values from the orthogonal direction of the collagen-fibril bundles in detail, and the plot will have an elliptical behavior because of a set of aligned collagen-fibril bundles in an image. The plot of FFT image will show a circular behavior in an image with randomly oriented collagen-fibril bundles. The ratio of short to long axe of the generated power plot of images was used to estimate the collagen orientation index (OI), which was calculated with the expression: OI ¼ [1– (short/long ratio)], which was shown in Figure 1(a) (Wu et al., 2011). The OI value of perfectly random tissue is “0,” which indicated isotropic behavior, and the FFT plot image was circular. A Fourier analysis with parallel collagen orientation yielded an elongated power that resulted in a longer orientation index (the maximum is “1”). Furthermore, a parameter collagen bundle packing (BP), which is expressed as BP ¼ 512 (1/h), where h is the pixel distances between the centers of gravity of two first-order maxima in 3D images were shown in Figure. 1 (b) (Wu et al., 2011). Ten images of each sample in different region were selected for quantitative analysis. The experimental data were analyzed by statistical test using SPSS 15.0 software (SPSS Inc., USA). Statistical significances of the data in different skin types were evaluated by T-test, which is used to determine whether or not significant differences exist between the means of two independent groups. In our study, the T-test analysis was used to analyze the variance of collagen intensity, collagen orientation index, and collagen bundle packing with different type’s skin. The differences were considered statistically significant at P < 0.05.
Results and Discussion Morphological Alteration in Collagen
The image of collagen morphology where is the strongest intensity in detectable depth in normal skin and DFSP skin is shown in Figure 2, respectively. The normal collagen had a number of compact collagen fibrils. The denser collagen fibrils formed into the big collagenous and thickening collagen bundles. The collagen arrangement was more organized from the Figure 2 (a). In contrast, From the Figure 2 (b), it shown that DFSP skin had thin collagen fibrils, the collagen matrix was slightly looser, and the collagen bundle appeared less organized, which implies loss of collagen structures. It demonstrated that the collagen had been destroyed during the process of DFSP. This
Fig 3. The SHG intensity with emission wavelength in different skin. In centre wavelength of second harmonic generation in 405 nm, the intensity of normal skin is much higher than the DFSP skin.
4
SCANNING VOL. 37, 1 (2015)
method was suitable for visualizing the precise status of the dense matrix and the bundles of the collagen. The level of collagen injured also can be distinct observed according to the SHG images.
Collagen Intensity
Previous reports have indicated that the intensity of change of SHG is a good predictor of collagen structural disruption (Sun et al., 2006). In this study, the SHG spectrum of skin collagen between normal and DPSF skin were obtained by lambda scanning mode of SHG
microscope, and the average data were normalized to the maximal peak intensity, as shown in Figure 3. The excitation laser wavelength was 810 nm, and thus, the peak intensity of SHG was at 405 nm. From the figure, the normalized intensity of SHG in normal skin was approximately 1, whereas the value of DFSP skin was only close to 0.4 at the emission wavelength of 405 nm. The intensity value in normal skin was about 2.5 times compared with the intensity value in the DFSP skin. It indicated that the intensity value of SHG decreased sharply from normal to DFSP skin. The phenomenon indicated that the skin was injured, and the collagen fribrils was destroyed, and then results in the loss of large collagen fibrils. The collagen fibrils were sparse at regular intervals distance of DFSP skin, whereas the collagen fibrils were dense in normal skin. As shown in Figure 4, the normalized value of the collagen intensity at 0.5 was considered as the reference point to analyze the collagen intensity changes with the collagen distance. More than 95% of collagen fibril intensities were more than 0.5 in the normalized value and the fibrils were side by side in normal skin. However, in the DFSP skin, more than 95% of collagen fibril intensities were less than 0.5. The intensity of DFSP collagen was much lower than that in normal skin. Moreover, the distances among collagen fibrils in non-normal skin were wider than those in normal skin according to the figure. The intensity of SHG signals varied greatly with distance in the two skin states.
Collagen Texture Feature
FFT method was employed to evaluate the OI and BP of collagen by Image J software. The results of OI is displayed in Figure 5. From the figure, we could see that
Fig 4. The SHG intensity from collagen images. (a) the DFSP image and its collagen intensity. (b) the intensity of collagen with distances between normal skin and DFSP skin. The intensity of normal skin is much higher than the DFSP skin. They have clear dividing line. It indicates the gap between two collagen bundles.
Fig 5. Bar chart demonstrating mean the value of collage OI ( SD) between normal skin and DFSP skin. P > 0.05. It means that the value has not statistical significance in two type skin.
S. Wu et al.: Quantitative analysis on collagen of dermatofibrosarcoma
5
demonstrated that the OI value was almost unchanged between normal skin and DFSP skin, whereas the BP value was higher in DFSP skin than in normal skin. This finding indicated that collagen degenerates in all directions and that collagen fibril disperses with the degeneration of collagen. But the distance of collagen fibril was obvious in DFSP skin. These results demonstrated that the SHG microscopy combined the texture analysis method could help us understand more information about DFSP.
References
Fig 6. Bar chart demonstrating mean the value of collage BP ( SD) in normal skin and DFSP skin, P < 0.05. It means that the value has statistical significance in two type skin.
the collagen-fibril bundles orientation index was not greatly significant between normal and DFSP skin. The value of OI is 0.65 0.09 for normal skin, whereas the value is 0.61 0.07 for DFSP skin. The results indicated that the alignment of collagen fibrils between normal skin and DFSP skin has not great statistical significance, which may be attributed to the older age of the patients with higher collagen OI in chronological aging skin (Wu et al., 2011). The distribution of collagen in two types skin had more ellipses shape in FFT analysis because the samples were obtained from the older people. However, the collagen BP differed greatly between two types skin from the Figure 6. The value of BP was 18.4 1.1 for normal skin, whereas the value was 23.5 0.9 for DFSP skin. These results confirmed that the collagen fibril bundles gradually decreased in all directions with dermis lesion progression.
Conclusions In this study, SHG was employed to evaluate the collagen alterations of normal skin and DFSP skin. The collagen structure, collagen quantity, and SHG signal intensities were obtained for analyzing the normal skin and DFSP skin. The outcomes indicate that the collagen of DFSP skin was degenerated, which was obvious reflected in the thinner collagen fibril, the weaker collagen density, and the larger collagen distance. Moreover, some collagen texture, such as the collagen orientation indices, and collagen fibril bundle were extracted from SHG images for quantitative charactering the alteration of collagen texture features between the two types skin through the FFT analysis. Our results
Bowne WB, Antonescu CR, Leung DH, et al. 2000. Dermatofibrosarcoma protuberans. Cancer 88:2711–2720. Campagnola P. 2011. Second harmonic generation imaging microscopy: applications to diseases diagnostics. Anal. Chem 83:3224–3231. Chen X, Nadiarynkh O, Plotnikov S, Campagnola PJ. 2012. Second harmonic generation microscopy for quantitative analysis of collagen fibrillar structure. Nat Protoc 7:654–669. Denk W, Strickler JH, Webb WW. 1990. Two-photon laser scanning fluorescence microscopy. Science 248:73–76. Ghazaryan AA, Hu PS, Chen SJ, Tan HY, Dong CY. 2012. Spatial and temporal analysis of skin glycation by the use of multiphoton microscopy and spectroscopy. J Dermatol Sci 65:189–195. Kobat D, Horton GN, Xu C. 2011. In vivo two-photon microscopy to 1.6-mm depth in mouse cortex. J. Biomed. Opt. 16:1–4. Koehler MJ, Preller A, Kindler N, Elsner P, et al. 2009. Intrinsic, solar and sunbed-induced skin aging measured in vivo by multiphoton laser tomography and biophysical methods. Skin.Res.Technol. 15:357–363. Llombart B, Serra-Guille´n C, Monteagudo C, Lo´pez Guerrero JA, Sanmartı´n O. 2013. Dermatofibrosarcoma protuberans: a comprehensive review and update on 2013. Elsevier pp 13– 28. Mark RJ, Bailet JW, Tran LM, et al. 1993. Dermatofibrosarcoma protuberans of the head and neck: a report of 16 cases. Arch. Otolaryngol.—Head & Neck Surg 119:891. McArthur G. 2004. Molecularly targeted treatment for dermatofibrosarcoma. Elsevier. p 30–36. Mendenhall WM, Zlotecki RA, Scarborough MT. 2004. Dermatofibrosarcoma protuberans. Cancer 101:2503–2508. Rubin BP, Schuetze SM, Eary JF, et al. 2002. Molecular targeting of platelet-derived growth factor B by imatinib mesylate in a patient with metastatic dermatofibrosarcoma protuberans. J. Clin. Oncol. 20:3586–3591. Sun Y, Chen WL, Lin SJ, et al. 2006. Investigating mechanisms of collagen thermal denaturation by high resolution secondharmonic generation imaging. Biophys. J. 91:2620–2625. Toki F, Honkura N, Shirakata Y, et al. 2013. Second harmonic generation reveals collagen fibril remodeling in fibroblastpopulated collagen gels. Cell Struct Funct 38:227–236. van Zuijlen PP, Ruurda JJ, van Veen HA, et al. 2003. Collagen morphology in human skin and scar tissue: no adaptations in response to mechanical loading at joints. Burns 29:423–431. Wu S, Li H, Yang H, et al. 2011. Quantitative analysis on collagen morphology in aging skin based on multiphoton microscopy. J. Biomed. Opt. 16:040502–0405 02. Yew E, Rowlands C, So PT. 2013. Application of Multi-Photon Microscopy in Dermatological Studies: A Mini-Review. J. Innov. Opt. Health Sci. 06:1330010. Zhu X, Lin L, Yu H, et al. 2013. Visualization of Epidermal and Dermal Alteration in Papulonodular Mucinosis by Multiphoton Microscopy. Scanning 35:22–27.
