Article pubs.acs.org/ac
Label-Free Biochemical Analytic Method for the Early Detection of Adenoviral Conjunctivitis Using Human Tear Biofluids Samjin Choi,*,†,‡,§ Sung Woon Moon,∥ Jae-Ho Shin,∥ Hun-Kuk Park,†,‡,§ and Kyung-Hyun Jin*,∥ †
Department of Biomedical Engineering, College of Medicine, Kyung Hee University, Seoul 130-701, Korea Healthcare Industry Research Institute, Kyung Hee University, Seoul 130-701, Korea § Department of Medical Engineering, College of Medicine, Kyung Hee University, Seoul 130-701, Korea ∥ Department of Ophthalmology, Kyung Hee University, Seoul 130-701, Korea ‡
S Supporting Information *
ABSTRACT: Cell culture and polymerase chain reaction are currently regarded as the gold standard for adenoviral conjunctivitis diagnosis. They maximize sensitivity and specificity but require several days to 3 weeks to get the results. The aim of this study is to determine the potential of Raman spectroscopy as a stand-alone analytical tool for clinical diagnosis of adenoviral conjunctivitis using human tear fluids. A drop-coating deposition surface enhanced Raman scattering (DCD-SERS) method was identified as the most effective method of proteomic analysis in tear biofluids. The proposed DCDSERS method (using a 2-μL sample) led to Raman spectra with high reproducibility, noise-independence, and uniformity. Additionally, the spectra were independent of the volume of biofluids used and detection zones, including the ring, middle, and central zone, with the exception of the outer layer of the ring zone. Assessments with an intensity ratio of 1242−1342 cm−1 achieved 100% sensitivity and 100% specificity in the central zone. Principal component analysis assessments achieved 0.9453 in the area under the receiver operating characteristic curve (AUC) as well as 93.3% sensitivity and 94.5% specificity in the central zone. Multi-Gaussian peak assessments showed that the differences between these two groups resulted from the reduction of the amide III α-helix structures of the proteins. The presence of adenovirus in tear fluids could be detected more accurately in the center of the sample than in the periphery. The DCD-SERS technique allowed for high chemical structure sensitivity without additional tagging or chemical modification, making it a good alternative for early clinical diagnosis of adenoviral conjunctivitis. Therefore, we are hopeful that the DCD-SERS method will be approved for use in ophthalmological clinics in the near future. uman tear fluids play an important role in health, particularly proper functioning of the eyes.1,2 Tears are information-rich biological materials, and systemic ocular diseases can lead to variations in the biochemical constitution of these fluids.3 Since teardrops contains shedded microbes and can be collected using noninvasive methods, their metabolites are a good target for the early detection of viral or fungal keratitis.4,5 Changes in the content and concentration of tears are the result of the body’s defense mechanism against microbes or other foreign material at the ocular surface.6−8 Ocular diseases result in distinctive changes in the protein composition of the aqueous layer of tears, and tracking these changes could indicate the presence of ocular diseases. Previous studies have evaluated the potential of teardrops as a diagnostic fluid for various ocular diseases using conventional laboratory techniques including enzyme-linked immunosorbent assay (ELISA), direct immunofluorescence assay (IFA), microscopy of stained or cultured samples, and polymerase chain reaction (PCR). For example, lysosome, a protective protein in tears, has been used as a marker to detect the presence of herpes simplex virus,9 and contact lens-induced peripheral ulcers have been shown to increase levels of two persistent neutrophil chemoattractants, leukotriene B4 (LTB4) and platelet activating factor (PAF) in tears.10 Locally released
H
© 2014 American Chemical Society
neuropeptides have also been shown to modulate protective and inflammatory reactions in the eyes,11 and ocular neuromediators have been linked to dry eye diseases.12 However, there were three common limitations in these earlier studies. The first was the requirement for a conjunctival smear and the substantial debris resulting from corneal scraping. This procedure causes patient discomfort and can increase susceptibility to secondary infections.13,14 The second limitation was a time-consuming analytical process. Although ELISA, IFA, and PCR techniques require only a few hours to analyze specimens, it can take much longer to receive analytical results in clinical situations. The third was a limit on the amount of tear fluids that can reasonably be collected from one patient. The total protein concentration (TPC) of tear fluids is 7-fold lower than that of blood serum.15 Therefore, highly sensitive molecular techniques including immunoassays,16 electrophoresis,17 and mass spectrometry18 can be useful in tear fluid analysis. However, only small volumes of tear fluids can be obtained from normal, nonirritated eyes.19 Most studies Received: May 10, 2014 Accepted: October 21, 2014 Published: October 21, 2014 11093
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
■
used diluted tear fluids or eye flush fluids instead of pure tear fluids which decreased the reliability of the results. Therefore, more sophisticated assessment techniques with low sample requirements are needed to reliably analyze tear proteins. Raman spectroscopy is a vibrational spectroscopic technique that can be used for sample identification and quantitation. The interaction between vibrating molecules or excited electrons in the sample results in up- and down-shifts in the energy of monochromic laser photons. The elastic scattering photons are filtered out and the rest of the photons are dispersed into the photon detector, which compiles molecular information about the photons in a certain system. Spectral analysis is done with computerized systems and determines the behavior of the sample at the molecular level.20−23 A Raman spectrum is a powerful fingerprint of the biochemical composition of various materials.24 Recently, nondestructive Raman spectroscopy was identified as a powerful diagnostic tool for the investigation of tear fluids and their chemical components.25−30 However, the Raman spectrum of liquid tears has a low signal-to-noise ratio due to inherently weak Raman scatter resulting from infrequent inelastic scattering. Therefore, various methods for enhancing the Raman scattering level in tears have been introduced. Among these methods, drop-coating deposition-surface enhanced Raman scattering (DCD-SERS) is considered the most effective for analyzing the proteins in tear fluids (Table S1, Supporting Information). However, although multivariate analysis improved the results, those studies did not show clear discrimination between normal and infected tear fluids due to low signal-to-noise ratios in Raman spectra as well as low reproducibility of the tear collection procedures. Although some of the findings of this study were consistent with Filik and Stone26 and Kuo et al.,28 the advanced findings were stable, noise-independent Raman spectral signal acquisitions with relatively higher detection rates in the center and periphery zones when compared to previous studies, suggesting that this technique is a stand-alone analytical tool in clinical situations. The most common causes of conjunctivitis are viral or bacterial infections and allergic reactions. Since all conjunctivitides initially present as redness, tear production, and discomfort, it is difficult to distinguish between allergic conjunctivitis, herpetic conjunctivitis, and chlamydia inclusion conjunctivitis. However, prompt diagnosis is imperative for prescription of accurate and appropriate treatments to prevent further infections and corneal damage.31 Infectious conjunctivitis can be caused by various serological subtypes of human adenovirus.32 Cell culture has been the gold standard for adenoviral conjunctivitis diagnosis; however, this assessment takes 3 weeks and has a high failure rate. A novel PCR technique has now replaced cell culture for diagnosis, increasing sensitivity and specificity.33,34 However, this diagnosis method still takes several days to complete. Although PCR has improved diagnosis of infective conjunctivitis, there is still a need for an even quicker method. The purpose of this study is to evaluate the potential of optical DCD-SERS detection as a diagnostic tool for adenoviral conjunctivitis using human tear fluids. The rule-based diagnostic system supported by a multivariate statistical algorithm and a segmentation algorithm enhances the analytical potential for the early detection of adenoviral conjunctivitis. The results of this preliminary study will be used as control data to characterize different causes of conjunctivitis and suggest that DCD-SERS is a stand-alone analytical tool in clinical situations.
Article
EXPERIMENTAL SECTION
Human Study. Human tear biofluids were collected from eight healthy patients (33 ± 8 years) and eight patients with adenoviral DNA-positive by PCR (36 ± 14 years) in the Kyung Hee University Medical Center. Adenoviral conjunctivitis severity was classified into one of two stages: mild or severe. Informed consent was obtained from each subject in the Kyung Hee University Medical Center. All procedures involving humans adhered to the Declaration of Helsinki were approved by the Ethical Committee of Kyung Hee University College of Medicine (KMCIRB1401-02). Preparation of Human Tears. Tear fluids were collected from the nasoinferior conjunctival sac without additional stimulation, using a 4 mm diameter, 10 mm long, preweighed, polyester-fiber rod (Transorb Wick, Filtrona, Richmond, VA). After a maximum of 5 min, the rods were withdrawn from the eye and placed into the end of a micropipet tip inside a 4 mm diameter, 1.5 mL Eppendorf tube. Tear fluids were extracted from the saturated rods by centrifuging at 8 000 rpm for 15 min. The rods and pipet were carefully removed and the tear fluid was aspirated. Tear samples were stored in an Eppendorf tube sealed with Parafilm (Pechiney Plastic Packaging Company, Chicago, IL) at −70 °C for 93%) and showed the highest detection rate of normal tear fluids (98%) in the R zone (Table S8, Supporting Information). Also, the specificity of the PCA biomarkers was 95% in the C zone, 91% in the M zone, 86% in the T zone, and 76% in the R zone. These findings suggest that the presence of adenovirus can be detected more accurately in the C zone than the R zone. This indicates that the differences between the C and R zones from the same dried tear can be used as a novel marker to detect the presence of adenovirus. The loading profiles of PC1 and PC2 explain approximately 98% of the total variation in the DCD-SERS spectra. A manually drawn linear separating line (red dotted lines in Figure 2) clearly showed the difference between the normal and adenoviral conjunctivitis groups with regards to AUC values and performance of the three PCA biomarkers. Therefore, this rule-based (database) classification system, supported by a multivariate statistical method, showed enhanced early detection of adenoviral conjunctivitis. These results show that the DCD-SERS spectral range is sufficient for discrimination between the normal and adenoviral conjunctivitis-infected tear biofluids. Therefore, in order to extract multiple characteristic peaks from the DCD-SERS 11096
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
Article
Figure 3. Representative original DCD-SERS spectra (1200−1500 cm−1) and their 10-Gaussian decomposition results for the normal (A, C) and adenoviral conjunctivitis-infected tear fluids (B, D) in the C (A, B) and R (C, D) zones. Decomposed 10-Gaussian functions were used to discriminate the normal and adenovirus-infected tear fluids.
spectral geometry, the multi-Gaussian model was used as the third biomarker. A total of 10 Gaussian functions (1- to 10Gaussian peaks) were extracted from one DCD-SERS spectra of 1200−1500 cm−1, in all four zones of a dried teardrop. Figure 3 shows representative multi-Gaussian decomposition results for the normal and adenoviral conjunctivitis groups in the C and R zone. The curve-fitted DCD-SERS spectra integrated with the 10 decomposed Gaussian functions was consistent with the original DCD-SERS spectra. Normal tear fluids (Figure 3A,C) indicated that DCD-SERS intensity at 1242 cm−1 (amide III β-sheet) was stronger than that at 1342 cm−1 (C−H deformation) in each zone. However, adenoviral conjunctivitis-infected tear biofluids demonstrated that DCDSERS intensity at 1242 cm−1 was lower than that at 1342 cm−1 in the C zone (Figure 3B), while the intensities of the two DCD-SERS peaks were almost identical in the R zone (Figure 3D). Each segmented Gaussian function illustrated the biochemical properties of biofluids in each peak. Four characteristic features (area, intensity, Raman shift, and halfwidth) were calculated from each Gaussian function (Table S9, Supporting Information). Four DCD-SERS assignments including P2 (1242 cm−1), P3 (1275 cm−1), P5 (1342 cm−1), and P10 (1448 cm−1) were selected to analyze the adenoviral infection induced in the protein structure of tear biofluids. The results of four MGP biomarkers for area characteristic feature are summarized in Table S10 (Supporting Information). The
normal tear biofluids showed a 2-fold increase of the amide III β-sheet in the R zone (0.2659 ± 0.0908) compared to the C zone (0.1314 ± 0.0187), while the adenoviral conjunctivitis tear biofluids showed opposite patterns. These changes led to a significant decrease in amide III α-helix vibration in the noninfected group (P < 0.001) and a significant increase in this vibration in the infected group (P < 0.01). C−H deformation vibrations at 1342 cm−1 showed the same patterns, with significant differences between the two groups, while those at 1448 cm−1 demonstrated no significant difference between the two groups. There are no significant Raman shifts for C−H deformation vibrations or α-helix and β-sheet vibrations in the amide III band. The amide III band involves multiple vibrations including C−N stretching and N−H plane bending of the peptide bond as well as Cα−C stretching and CO plane bending.40 Furthermore, the complex vibrational response of proteins in the 1200−1350 cm−1 range makes the interpretation of the amide III band very difficult. However, previous studies shed some light on our findings.41 The reduction of the amide III α-helix and the elevation of the amide III β-sheet in adenoviral conjunctivitis-infected tear biofluids indicates morphological changes in the secondary structures of proteins. This change involves interactions between intercellular reactive oxygen species (ROS) and biological macromolecules. In general, ROS such as OH, O2−, and H2O2 are regulated at controlled rates. However, some evidence suggests that the 11097
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
Article
obtained in as little as 10 min, this product is very expensive and therefore ill-suited for use in developing countries with limited medical resources. Although Raman spectroscopy methods for diagnosis require an initial capital investment, they are generally more cost-effective than PCR and can be easily applied to a variety of diseases, as verified by previous studies, making them more powerful diagnostic tools for ophthalmological diseases. There were some limitations to this study, including limited sample size, lack of biological analysis, and lack of consideration for the presence of surface-active compounds such as eye lipids, and further studies on virusspecific conjunctivitis including herpes simplex conjunctivitis and varicella-zoster blepharconjunctivitis, allergic conjunctivitis, and chlamydia conjunctivitis are necessary to expand the applications of this method.
production of ROS increases in infected eyes. This increase would lead to changes in many cellular components including DNA/RNA, proteins, lipids, and carbohydrates. In particular, ROS elevation could degenerate hydrogen bonds, disulfide bonds, and carbon−sulfur bonds, which play important roles in the maintenance of protein structure. The reduction of the amide III α-helix by adenoviral conjunctivitis infection resulted in microenvironmental changes in chemical C−H bonds. This explains the decreased intensity of the DCD-SERS peak at 1448 cm−1 (C−H deformation in DNA/RNA, proteins, lipids, and carbohydrates) and the increased intensity of the DCD-SERS peak at 1342 cm−1 (C−H deformation in proteins). Each Gaussian function decomposed by multiple Gaussian models clearly showed the difference between the normal and adenovirus-infected tear biofluids. This suggests that MGP markers determined by Gaussian segmentation techniques could be used to quantitatively and qualitatively monitor and detect the presence of adenovirus. It is well-known that conventional cytologic examination and microbial culture are insufficient for early and accurate diagnoses of adenovirus conjunctivitis. Although PCR-based diagnosis is improving, it is a laboratory method and not a clinically applicable one. Its time requirement is too high for clinics that require rapid diagnoses, and PCR requires high-level techniques to collect samples. Alternatively, Raman spectrum determination of tear biofluids is a rapid test with no significant costs after the initial investment in Raman spectroscopy equipment. This study showed the potential of Raman spectroscopy as a simple, accurate, and reproducible diagnostic tool (Figure S8, Supporting Information). Previous studies have described the preclinical applications of Raman spectroscopy in detail,22,28,29,42 and some reports have investigated tear fluid analysis using Raman spectroscopy. Filik and Stone26 showed that a 1.5 μL tear from a healthy human subject could be used to detect a Raman spectrum with a high signal-to-noise ratio using the DCDR method. They attempted to analyze the distribution of the major tear components, including proteins, urea, bicarbonate, and lipids, in a dried teardrop through the correlation between the PCs of PCA analysis. Consistent with the findings in this study, Filik and Stone26 and Kuo et al.28 showed that the central area of dried teardrops are the best fingerprint zones for Raman spectra due to their low variation. Kuo et al.29 also confirmed that there are detectable differences in Raman intensity between noninfectious ulcerative keratitis and infectious ulcerative keratitis (a 1.5 μL tear sample deposited on gold thin film) using the PCA analytic method. However, they were unable to determine the cause of this difference. The composition of human tear fluids is extremely complex, and infection increases their intricacy. Although this study did not clarify the reasons for differences between adenoviral conjunctivitis-infected and normal tear biofluids, it did confirm that changes in composition are due to developments on the surface of the eyeball (Figure S9, Supporting Information). Early diagnosis of adenoviral conjunctivitis will not only prevent infection and drug abuse caused by misdiagnoses but will also decrease the number of adenoviral conjunctivitis patients. Recently, a new PCR-based technique has been developed to diagnose adenoviral conjunctivitis that involves conventional cell culture methods. RPS AdenoPlus, based on the mechanism of lateral-flow immunochromatography, received Food and Drug Administration (FDA) approval in 2011 and is now on the market.33,43 Although results can be
■
CONCLUSIONS
The present study demonstrated that optical DCD-SERS can be used to obtain high-quality Raman spectra for early clinical detection of infectious adenoviral diseases from human tear fluids. The resulting spectra showed highly reproducible, noiseindependent, and uniform characteristics. Also, the spectra were independent of the amount of tear fluids (1−8 μL) and detection zones analyzed, with the exception of the outer layer of the ring zone, and remained unchanged over a period of weeks. Because of limitations of tear collections, a 2 μL droplet was used to acquire high signal-to-noise ratio Raman spectral signals for further studies. The prominent Raman peaks and assignments in the range of 417−1782 cm−1 were interpreted. On the basis of morphological differences between normal and adenovirus-infected tear fluids in the range of 1200−1400 cm−1, three adenovirus-specific DCD-SERS bands, including an amide III β-sheet at 1242 cm−1, a C−H deformation in proteins at 1342 cm−1, and a C−H deformation in DNA/RNA, proteins, lipids, and carbohydrates at 1448 cm−1, were identified as markers of infection by adenovirus. Three clinical biomarkers, including AC biomarker, PCA biomarker, and MGP biomarker, were tested with a total of 200 DCD-SERS spectral signals in four zones of a dried teardrop. The results indicate that the proposed biomarkers are superior markers for the early detection of adenovirus-infected tear fluids. The presence of adenovirus in human tear fluids could be detected more accurately in the C, M, and T zones than the R zone. This zone-dependent difference also indicated that the differences between two zones could be used as a novel marker to detect the presence of adenovirus. The logarithmic AC method demonstrated an easily implementable biomarker with high sensitivity and specificity, but low accuracy, while the rule-based PCA method showed multiple evaluable biomarker determined by a multivariate statistical algorithm with relatively high performance. The MGP biomarker was revealed to produce more accurate analytical results but required a time-consuming procedure. In terms of stability and efficiency, the rule-based (database) PCA biomarker is the most promising assessment for clinical use. Furthermore, the graphic-user representation of the PCA biomarker could be helpful for ophthalmologists to monitor and detect the presence of adenovirus at an early stage of infection. Finally, the DCD-SERS Raman technique allows for high chemical structure sensitivity without additional tagging and chemical modification. This advantage makes DCD-SERS Raman technology an excellent tool for early detection of adenoviral conjunctivitis. Therefore, we hope that 11098
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
Article
(22) Sengupta, A.; Mujacic, M.; Davis, E. Anal. Bioanal. Chem. 2006, 386, 1379−1386. (23) Lopez-Diez, E. C.; Winder, C. L.; Ashton, L.; Currie, F.; Goodacre, R. Anal. Chem. 2005, 77, 2901−2906. (24) Lee, Y. J.; Jung, G. B.; Choi, S.; Lee, G.; Kim, J. H.; Son, H. S.; Bae, H.; Park, H. K. PLoS One 2013, 8, e79761. (25) Reyes-Goddard, J. M.; Barr, H.; Stone, N. Photodiagn. Photodyn. Ther. 2008, 5, 42−49. (26) Filik, J.; Stone, N. Anal. Chim. Acta 2008, 616, 177−184. (27) Filik, J.; Stone, N. Analyst 2007, 132, 544−550. (28) Kuo, M. T.; Lin, C. C.; Liu, H. Y.; Chang, H. C. Invest. Ophthalmol. Vis. Sci. 2011, 52, 4942−4950. (29) Kuo, M. T.; Lin, C. C.; Liu, H. Y.; Yang, M. Y.; Chang, H. C. Invest. Ophthalmol. Vis. Sci. 2012, 53, 1436−1444. (30) Zhang, D.; Xie, Y.; Mrozek, M. F.; Ortiz, C.; Davisson, V. J.; Ben-Amotz, D. Anal. Chem. 2003, 75, 5703−5709. (31) Schwab, I.; Dawson, C. R. In General Ophthalmology, 13th ed.; Vaughan, D., Asbury, T., Riordan-Eva, P., Eds.; Appleton & Lange: Norwalk, CT, 1992; pp 96−124. (32) González-López, J. J.; Morcillo-Laiz, R.; Muñoz-Negrete, F. J. Arch. Soc. Esp. Oftalmol. 2013, 88, 108−115. (33) Sambursky, R.; Tauber, S.; Schirra, F.; Kozich, K.; Davidson, R.; Cohen, E. J. Ophthalmology 2006, 113, 1758−1764. (34) Cooper, R. J.; Yeo, A. C.; Bailey, A. S.; Tullo, A. B. Invest. Ophthalmol. Vis. Sci. 1999, 40, 90−95. (35) Jung, G. B.; Bae, Y. M.; Lee, Y. J.; Ryu, S. H.; Park, H. K. Appl. Surf. Sci. 2013, 282, 161−164. (36) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Nature 1997, 389, 827−829. (37) Gorr, H. M.; Zueger, J. M.; McAdams, D. R.; Barnard, J. A. Colloids Surf. B: Biointerfaces 2013, 103, 59−66. (38) Marin, A. G.; Gelderblom, H.; Lohse, D.; Snoeijer, J. H. Phys. Rev. Lett. 2011, 107, 085502. (39) Hendarto, E.; Gianchandani, Y. B. J. Micromech. Microeng. 2013, 23, 075016. (40) Sun, W.; Zhao, Q.; Zhao, M.; Yang, B.; Cui, C.; Ren, J. J. Agric. Food Chem. 2011, 59, 11070−11077. (41) Chen, P.; Zhang, F.; Lin, L.; Bai, H.; Zhang, L.; Tang, G. Q.; Fang, H.; Mu, G. G.; Gong, W.; Liu, Z. P.; Han, Z. B.; Zhao, H.; Han, Z. C. Stem Cells in Clinic and Research; Gholamrezanezhad, A., Ed.; InTech: Rijeka, Croatia, 2011. (42) Ellis, D. I.; Goodacre, R. Analyst 2006, 131, 875−885. (43) RPS. AdenoPlus, http://www.rpsdetectors.com/in/products/ adenoplus/.
a DCD-SERS-based Raman device can be developed for implementation in ophthalmological clinics within a few days.
■
ASSOCIATED CONTENT
S Supporting Information *
Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
*Phone: +82 2 961 0290. Fax: +82 2 6008 5535. E-mail: [email protected]. *E-mail: [email protected]. Author Contributions
Samjin Choi and Sung Woon Moon contributed equally to this paper. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by a grant from Kyung Hee University in 2013 (Grant KHU-20131086). REFERENCES
(1) Karns, K.; Herr, A. E. Anal. Chem. 2011, 83, 8115−8122. (2) Ohashi, Y.; Dogru, M.; Tsubota, K. Clin. Chim. Acta 2006, 369, 17−28. (3) Zavaro, A.; Samra, Z.; Baryishak, R.; Sompolinsky, D. Doc. Ophthalmol. 1980, 50, 185−199. (4) Kaji, Y.; Hiraoka, T.; Oshika, T. Graefes. Arch. Clin. Exp. Ophthalmol. 2009, 247, 989−992. (5) Fukuda, M.; Deai, T.; Higaki, S.; Hayashi, K.; Shimomura, Y. Semin. Ophthalmol. 2008, 23, 217−220. (6) Pramod, N. P.; Dhevahi, E.; Sudhamathi, K.; Kannan, K.; Thyagarajan, S. P. Ocul. Immunol. Inflamm. 1999, 7, 61−67. (7) McNamara, N. A.; Andika, R.; Kwong, M.; Sack, R. A.; Fleiszig, S. M. Curr. Eye Res. 2005, 30, 517−525. (8) Cao, Z.; Saravanan, C.; Goldstein, M. H.; Wu, H. K.; Pasricha, G.; Sharma, S.; Panjwani, N. Arch. Ophthalmol. 2008, 126, 348−352. (9) Eylan, E.; Ronen, D.; Romano, A.; Smetana, O. Invest. Ophthalmol. Vis. Sci. 1977, 16, 850−853. (10) Thakur, A.; Willcox, M. D. Exp. Eye Res. 1998, 67, 9−19. (11) Sacchetti, M.; Micera, A.; Lambiase, A.; Speranza, S.; Mantelli, F.; Petrachi, G.; Bonini, S.; Bonini, S. Mol. Vis. 2011, 17, 47−52. (12) Lambiase, A.; Micera, A.; Sacchetti, M.; Cortes, M.; Mantelli, F.; Bonini, S. Arch. Ophthalmol. 2011, 129, 981−986. (13) Hidalgo, F.; Melón, S.; de Oña, M.; Do Santos, V.; Martínez, A.; Cimadevilla, R.; Rodríguez, M. Eur. J. Clin. Microbiol. Infect. Dis. 1998, 17, 120−123. (14) Kaufman, H. E.; Azcuy, A. M.; Varnell, E. D.; Sloop, G. D.; Thompson, H. W.; Hill, J. M. Invest. Ophth. Vis. Sci. 2005, 46, 241− 247. (15) Ng, V.; Cho, P. Graefes. Arch. Clin. Exp. Ophthalmol. 2000, 238, 571−576. (16) Willcox, M. D.; Lan, J. Clin. Exp. Optom. 1999, 82, 1−3. (17) Bjerrum, K. B. Acta Ophthalmol. Scand. 1999, 77, 1−8. (18) Grus, F. H.; Podust, V. N.; Bruns, K.; Lackner, K.; Fu, S.; Dalmasso, E. A.; Wirthlin, A.; Pfeiffer, N. Invest. Ophthalmol. Vis. Sci. 2005, 46, 863−876. (19) Fullard, R. J.; Snyder, C. Invest. Ophthalmol. Vis. Sci. 1990, 31, 1119−1126. (20) Barhoumi, A.; Zhang, D.; Tam, F.; Halas, N. J. J. Am. Chem. Soc. 2008, 130, 5523−5529. (21) Taleb, A.; Diamond, J.; McGarvey, J. J.; Beattie, J. R.; Toland, C.; Hamilton, P. W. J. Phys. Chem. B 2006, 110, 19625−19631. 11099
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Label-Free Biochemical Analytic Method for the Early Detection of Adenoviral Conjunctivitis Using Human Tear Biofluids Samjin Choi,*,†,‡,§ Sung Woon Moon,∥ Jae-Ho Shin,∥ Hun-Kuk Park,†,‡,§ and Kyung-Hyun Jin*,∥ †
Department of Biomedical Engineering, College of Medicine, Kyung Hee University, Seoul 130-701, Korea Healthcare Industry Research Institute, Kyung Hee University, Seoul 130-701, Korea § Department of Medical Engineering, College of Medicine, Kyung Hee University, Seoul 130-701, Korea ∥ Department of Ophthalmology, Kyung Hee University, Seoul 130-701, Korea ‡
S Supporting Information *
ABSTRACT: Cell culture and polymerase chain reaction are currently regarded as the gold standard for adenoviral conjunctivitis diagnosis. They maximize sensitivity and specificity but require several days to 3 weeks to get the results. The aim of this study is to determine the potential of Raman spectroscopy as a stand-alone analytical tool for clinical diagnosis of adenoviral conjunctivitis using human tear fluids. A drop-coating deposition surface enhanced Raman scattering (DCD-SERS) method was identified as the most effective method of proteomic analysis in tear biofluids. The proposed DCDSERS method (using a 2-μL sample) led to Raman spectra with high reproducibility, noise-independence, and uniformity. Additionally, the spectra were independent of the volume of biofluids used and detection zones, including the ring, middle, and central zone, with the exception of the outer layer of the ring zone. Assessments with an intensity ratio of 1242−1342 cm−1 achieved 100% sensitivity and 100% specificity in the central zone. Principal component analysis assessments achieved 0.9453 in the area under the receiver operating characteristic curve (AUC) as well as 93.3% sensitivity and 94.5% specificity in the central zone. Multi-Gaussian peak assessments showed that the differences between these two groups resulted from the reduction of the amide III α-helix structures of the proteins. The presence of adenovirus in tear fluids could be detected more accurately in the center of the sample than in the periphery. The DCD-SERS technique allowed for high chemical structure sensitivity without additional tagging or chemical modification, making it a good alternative for early clinical diagnosis of adenoviral conjunctivitis. Therefore, we are hopeful that the DCD-SERS method will be approved for use in ophthalmological clinics in the near future. uman tear fluids play an important role in health, particularly proper functioning of the eyes.1,2 Tears are information-rich biological materials, and systemic ocular diseases can lead to variations in the biochemical constitution of these fluids.3 Since teardrops contains shedded microbes and can be collected using noninvasive methods, their metabolites are a good target for the early detection of viral or fungal keratitis.4,5 Changes in the content and concentration of tears are the result of the body’s defense mechanism against microbes or other foreign material at the ocular surface.6−8 Ocular diseases result in distinctive changes in the protein composition of the aqueous layer of tears, and tracking these changes could indicate the presence of ocular diseases. Previous studies have evaluated the potential of teardrops as a diagnostic fluid for various ocular diseases using conventional laboratory techniques including enzyme-linked immunosorbent assay (ELISA), direct immunofluorescence assay (IFA), microscopy of stained or cultured samples, and polymerase chain reaction (PCR). For example, lysosome, a protective protein in tears, has been used as a marker to detect the presence of herpes simplex virus,9 and contact lens-induced peripheral ulcers have been shown to increase levels of two persistent neutrophil chemoattractants, leukotriene B4 (LTB4) and platelet activating factor (PAF) in tears.10 Locally released
H
© 2014 American Chemical Society
neuropeptides have also been shown to modulate protective and inflammatory reactions in the eyes,11 and ocular neuromediators have been linked to dry eye diseases.12 However, there were three common limitations in these earlier studies. The first was the requirement for a conjunctival smear and the substantial debris resulting from corneal scraping. This procedure causes patient discomfort and can increase susceptibility to secondary infections.13,14 The second limitation was a time-consuming analytical process. Although ELISA, IFA, and PCR techniques require only a few hours to analyze specimens, it can take much longer to receive analytical results in clinical situations. The third was a limit on the amount of tear fluids that can reasonably be collected from one patient. The total protein concentration (TPC) of tear fluids is 7-fold lower than that of blood serum.15 Therefore, highly sensitive molecular techniques including immunoassays,16 electrophoresis,17 and mass spectrometry18 can be useful in tear fluid analysis. However, only small volumes of tear fluids can be obtained from normal, nonirritated eyes.19 Most studies Received: May 10, 2014 Accepted: October 21, 2014 Published: October 21, 2014 11093
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
■
used diluted tear fluids or eye flush fluids instead of pure tear fluids which decreased the reliability of the results. Therefore, more sophisticated assessment techniques with low sample requirements are needed to reliably analyze tear proteins. Raman spectroscopy is a vibrational spectroscopic technique that can be used for sample identification and quantitation. The interaction between vibrating molecules or excited electrons in the sample results in up- and down-shifts in the energy of monochromic laser photons. The elastic scattering photons are filtered out and the rest of the photons are dispersed into the photon detector, which compiles molecular information about the photons in a certain system. Spectral analysis is done with computerized systems and determines the behavior of the sample at the molecular level.20−23 A Raman spectrum is a powerful fingerprint of the biochemical composition of various materials.24 Recently, nondestructive Raman spectroscopy was identified as a powerful diagnostic tool for the investigation of tear fluids and their chemical components.25−30 However, the Raman spectrum of liquid tears has a low signal-to-noise ratio due to inherently weak Raman scatter resulting from infrequent inelastic scattering. Therefore, various methods for enhancing the Raman scattering level in tears have been introduced. Among these methods, drop-coating deposition-surface enhanced Raman scattering (DCD-SERS) is considered the most effective for analyzing the proteins in tear fluids (Table S1, Supporting Information). However, although multivariate analysis improved the results, those studies did not show clear discrimination between normal and infected tear fluids due to low signal-to-noise ratios in Raman spectra as well as low reproducibility of the tear collection procedures. Although some of the findings of this study were consistent with Filik and Stone26 and Kuo et al.,28 the advanced findings were stable, noise-independent Raman spectral signal acquisitions with relatively higher detection rates in the center and periphery zones when compared to previous studies, suggesting that this technique is a stand-alone analytical tool in clinical situations. The most common causes of conjunctivitis are viral or bacterial infections and allergic reactions. Since all conjunctivitides initially present as redness, tear production, and discomfort, it is difficult to distinguish between allergic conjunctivitis, herpetic conjunctivitis, and chlamydia inclusion conjunctivitis. However, prompt diagnosis is imperative for prescription of accurate and appropriate treatments to prevent further infections and corneal damage.31 Infectious conjunctivitis can be caused by various serological subtypes of human adenovirus.32 Cell culture has been the gold standard for adenoviral conjunctivitis diagnosis; however, this assessment takes 3 weeks and has a high failure rate. A novel PCR technique has now replaced cell culture for diagnosis, increasing sensitivity and specificity.33,34 However, this diagnosis method still takes several days to complete. Although PCR has improved diagnosis of infective conjunctivitis, there is still a need for an even quicker method. The purpose of this study is to evaluate the potential of optical DCD-SERS detection as a diagnostic tool for adenoviral conjunctivitis using human tear fluids. The rule-based diagnostic system supported by a multivariate statistical algorithm and a segmentation algorithm enhances the analytical potential for the early detection of adenoviral conjunctivitis. The results of this preliminary study will be used as control data to characterize different causes of conjunctivitis and suggest that DCD-SERS is a stand-alone analytical tool in clinical situations.
Article
EXPERIMENTAL SECTION
Human Study. Human tear biofluids were collected from eight healthy patients (33 ± 8 years) and eight patients with adenoviral DNA-positive by PCR (36 ± 14 years) in the Kyung Hee University Medical Center. Adenoviral conjunctivitis severity was classified into one of two stages: mild or severe. Informed consent was obtained from each subject in the Kyung Hee University Medical Center. All procedures involving humans adhered to the Declaration of Helsinki were approved by the Ethical Committee of Kyung Hee University College of Medicine (KMCIRB1401-02). Preparation of Human Tears. Tear fluids were collected from the nasoinferior conjunctival sac without additional stimulation, using a 4 mm diameter, 10 mm long, preweighed, polyester-fiber rod (Transorb Wick, Filtrona, Richmond, VA). After a maximum of 5 min, the rods were withdrawn from the eye and placed into the end of a micropipet tip inside a 4 mm diameter, 1.5 mL Eppendorf tube. Tear fluids were extracted from the saturated rods by centrifuging at 8 000 rpm for 15 min. The rods and pipet were carefully removed and the tear fluid was aspirated. Tear samples were stored in an Eppendorf tube sealed with Parafilm (Pechiney Plastic Packaging Company, Chicago, IL) at −70 °C for 93%) and showed the highest detection rate of normal tear fluids (98%) in the R zone (Table S8, Supporting Information). Also, the specificity of the PCA biomarkers was 95% in the C zone, 91% in the M zone, 86% in the T zone, and 76% in the R zone. These findings suggest that the presence of adenovirus can be detected more accurately in the C zone than the R zone. This indicates that the differences between the C and R zones from the same dried tear can be used as a novel marker to detect the presence of adenovirus. The loading profiles of PC1 and PC2 explain approximately 98% of the total variation in the DCD-SERS spectra. A manually drawn linear separating line (red dotted lines in Figure 2) clearly showed the difference between the normal and adenoviral conjunctivitis groups with regards to AUC values and performance of the three PCA biomarkers. Therefore, this rule-based (database) classification system, supported by a multivariate statistical method, showed enhanced early detection of adenoviral conjunctivitis. These results show that the DCD-SERS spectral range is sufficient for discrimination between the normal and adenoviral conjunctivitis-infected tear biofluids. Therefore, in order to extract multiple characteristic peaks from the DCD-SERS 11096
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
Article
Figure 3. Representative original DCD-SERS spectra (1200−1500 cm−1) and their 10-Gaussian decomposition results for the normal (A, C) and adenoviral conjunctivitis-infected tear fluids (B, D) in the C (A, B) and R (C, D) zones. Decomposed 10-Gaussian functions were used to discriminate the normal and adenovirus-infected tear fluids.
spectral geometry, the multi-Gaussian model was used as the third biomarker. A total of 10 Gaussian functions (1- to 10Gaussian peaks) were extracted from one DCD-SERS spectra of 1200−1500 cm−1, in all four zones of a dried teardrop. Figure 3 shows representative multi-Gaussian decomposition results for the normal and adenoviral conjunctivitis groups in the C and R zone. The curve-fitted DCD-SERS spectra integrated with the 10 decomposed Gaussian functions was consistent with the original DCD-SERS spectra. Normal tear fluids (Figure 3A,C) indicated that DCD-SERS intensity at 1242 cm−1 (amide III β-sheet) was stronger than that at 1342 cm−1 (C−H deformation) in each zone. However, adenoviral conjunctivitis-infected tear biofluids demonstrated that DCDSERS intensity at 1242 cm−1 was lower than that at 1342 cm−1 in the C zone (Figure 3B), while the intensities of the two DCD-SERS peaks were almost identical in the R zone (Figure 3D). Each segmented Gaussian function illustrated the biochemical properties of biofluids in each peak. Four characteristic features (area, intensity, Raman shift, and halfwidth) were calculated from each Gaussian function (Table S9, Supporting Information). Four DCD-SERS assignments including P2 (1242 cm−1), P3 (1275 cm−1), P5 (1342 cm−1), and P10 (1448 cm−1) were selected to analyze the adenoviral infection induced in the protein structure of tear biofluids. The results of four MGP biomarkers for area characteristic feature are summarized in Table S10 (Supporting Information). The
normal tear biofluids showed a 2-fold increase of the amide III β-sheet in the R zone (0.2659 ± 0.0908) compared to the C zone (0.1314 ± 0.0187), while the adenoviral conjunctivitis tear biofluids showed opposite patterns. These changes led to a significant decrease in amide III α-helix vibration in the noninfected group (P < 0.001) and a significant increase in this vibration in the infected group (P < 0.01). C−H deformation vibrations at 1342 cm−1 showed the same patterns, with significant differences between the two groups, while those at 1448 cm−1 demonstrated no significant difference between the two groups. There are no significant Raman shifts for C−H deformation vibrations or α-helix and β-sheet vibrations in the amide III band. The amide III band involves multiple vibrations including C−N stretching and N−H plane bending of the peptide bond as well as Cα−C stretching and CO plane bending.40 Furthermore, the complex vibrational response of proteins in the 1200−1350 cm−1 range makes the interpretation of the amide III band very difficult. However, previous studies shed some light on our findings.41 The reduction of the amide III α-helix and the elevation of the amide III β-sheet in adenoviral conjunctivitis-infected tear biofluids indicates morphological changes in the secondary structures of proteins. This change involves interactions between intercellular reactive oxygen species (ROS) and biological macromolecules. In general, ROS such as OH, O2−, and H2O2 are regulated at controlled rates. However, some evidence suggests that the 11097
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
Article
obtained in as little as 10 min, this product is very expensive and therefore ill-suited for use in developing countries with limited medical resources. Although Raman spectroscopy methods for diagnosis require an initial capital investment, they are generally more cost-effective than PCR and can be easily applied to a variety of diseases, as verified by previous studies, making them more powerful diagnostic tools for ophthalmological diseases. There were some limitations to this study, including limited sample size, lack of biological analysis, and lack of consideration for the presence of surface-active compounds such as eye lipids, and further studies on virusspecific conjunctivitis including herpes simplex conjunctivitis and varicella-zoster blepharconjunctivitis, allergic conjunctivitis, and chlamydia conjunctivitis are necessary to expand the applications of this method.
production of ROS increases in infected eyes. This increase would lead to changes in many cellular components including DNA/RNA, proteins, lipids, and carbohydrates. In particular, ROS elevation could degenerate hydrogen bonds, disulfide bonds, and carbon−sulfur bonds, which play important roles in the maintenance of protein structure. The reduction of the amide III α-helix by adenoviral conjunctivitis infection resulted in microenvironmental changes in chemical C−H bonds. This explains the decreased intensity of the DCD-SERS peak at 1448 cm−1 (C−H deformation in DNA/RNA, proteins, lipids, and carbohydrates) and the increased intensity of the DCD-SERS peak at 1342 cm−1 (C−H deformation in proteins). Each Gaussian function decomposed by multiple Gaussian models clearly showed the difference between the normal and adenovirus-infected tear biofluids. This suggests that MGP markers determined by Gaussian segmentation techniques could be used to quantitatively and qualitatively monitor and detect the presence of adenovirus. It is well-known that conventional cytologic examination and microbial culture are insufficient for early and accurate diagnoses of adenovirus conjunctivitis. Although PCR-based diagnosis is improving, it is a laboratory method and not a clinically applicable one. Its time requirement is too high for clinics that require rapid diagnoses, and PCR requires high-level techniques to collect samples. Alternatively, Raman spectrum determination of tear biofluids is a rapid test with no significant costs after the initial investment in Raman spectroscopy equipment. This study showed the potential of Raman spectroscopy as a simple, accurate, and reproducible diagnostic tool (Figure S8, Supporting Information). Previous studies have described the preclinical applications of Raman spectroscopy in detail,22,28,29,42 and some reports have investigated tear fluid analysis using Raman spectroscopy. Filik and Stone26 showed that a 1.5 μL tear from a healthy human subject could be used to detect a Raman spectrum with a high signal-to-noise ratio using the DCDR method. They attempted to analyze the distribution of the major tear components, including proteins, urea, bicarbonate, and lipids, in a dried teardrop through the correlation between the PCs of PCA analysis. Consistent with the findings in this study, Filik and Stone26 and Kuo et al.28 showed that the central area of dried teardrops are the best fingerprint zones for Raman spectra due to their low variation. Kuo et al.29 also confirmed that there are detectable differences in Raman intensity between noninfectious ulcerative keratitis and infectious ulcerative keratitis (a 1.5 μL tear sample deposited on gold thin film) using the PCA analytic method. However, they were unable to determine the cause of this difference. The composition of human tear fluids is extremely complex, and infection increases their intricacy. Although this study did not clarify the reasons for differences between adenoviral conjunctivitis-infected and normal tear biofluids, it did confirm that changes in composition are due to developments on the surface of the eyeball (Figure S9, Supporting Information). Early diagnosis of adenoviral conjunctivitis will not only prevent infection and drug abuse caused by misdiagnoses but will also decrease the number of adenoviral conjunctivitis patients. Recently, a new PCR-based technique has been developed to diagnose adenoviral conjunctivitis that involves conventional cell culture methods. RPS AdenoPlus, based on the mechanism of lateral-flow immunochromatography, received Food and Drug Administration (FDA) approval in 2011 and is now on the market.33,43 Although results can be
■
CONCLUSIONS
The present study demonstrated that optical DCD-SERS can be used to obtain high-quality Raman spectra for early clinical detection of infectious adenoviral diseases from human tear fluids. The resulting spectra showed highly reproducible, noiseindependent, and uniform characteristics. Also, the spectra were independent of the amount of tear fluids (1−8 μL) and detection zones analyzed, with the exception of the outer layer of the ring zone, and remained unchanged over a period of weeks. Because of limitations of tear collections, a 2 μL droplet was used to acquire high signal-to-noise ratio Raman spectral signals for further studies. The prominent Raman peaks and assignments in the range of 417−1782 cm−1 were interpreted. On the basis of morphological differences between normal and adenovirus-infected tear fluids in the range of 1200−1400 cm−1, three adenovirus-specific DCD-SERS bands, including an amide III β-sheet at 1242 cm−1, a C−H deformation in proteins at 1342 cm−1, and a C−H deformation in DNA/RNA, proteins, lipids, and carbohydrates at 1448 cm−1, were identified as markers of infection by adenovirus. Three clinical biomarkers, including AC biomarker, PCA biomarker, and MGP biomarker, were tested with a total of 200 DCD-SERS spectral signals in four zones of a dried teardrop. The results indicate that the proposed biomarkers are superior markers for the early detection of adenovirus-infected tear fluids. The presence of adenovirus in human tear fluids could be detected more accurately in the C, M, and T zones than the R zone. This zone-dependent difference also indicated that the differences between two zones could be used as a novel marker to detect the presence of adenovirus. The logarithmic AC method demonstrated an easily implementable biomarker with high sensitivity and specificity, but low accuracy, while the rule-based PCA method showed multiple evaluable biomarker determined by a multivariate statistical algorithm with relatively high performance. The MGP biomarker was revealed to produce more accurate analytical results but required a time-consuming procedure. In terms of stability and efficiency, the rule-based (database) PCA biomarker is the most promising assessment for clinical use. Furthermore, the graphic-user representation of the PCA biomarker could be helpful for ophthalmologists to monitor and detect the presence of adenovirus at an early stage of infection. Finally, the DCD-SERS Raman technique allows for high chemical structure sensitivity without additional tagging and chemical modification. This advantage makes DCD-SERS Raman technology an excellent tool for early detection of adenoviral conjunctivitis. Therefore, we hope that 11098
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
Analytical Chemistry
Article
(22) Sengupta, A.; Mujacic, M.; Davis, E. Anal. Bioanal. Chem. 2006, 386, 1379−1386. (23) Lopez-Diez, E. C.; Winder, C. L.; Ashton, L.; Currie, F.; Goodacre, R. Anal. Chem. 2005, 77, 2901−2906. (24) Lee, Y. J.; Jung, G. B.; Choi, S.; Lee, G.; Kim, J. H.; Son, H. S.; Bae, H.; Park, H. K. PLoS One 2013, 8, e79761. (25) Reyes-Goddard, J. M.; Barr, H.; Stone, N. Photodiagn. Photodyn. Ther. 2008, 5, 42−49. (26) Filik, J.; Stone, N. Anal. Chim. Acta 2008, 616, 177−184. (27) Filik, J.; Stone, N. Analyst 2007, 132, 544−550. (28) Kuo, M. T.; Lin, C. C.; Liu, H. Y.; Chang, H. C. Invest. Ophthalmol. Vis. Sci. 2011, 52, 4942−4950. (29) Kuo, M. T.; Lin, C. C.; Liu, H. Y.; Yang, M. Y.; Chang, H. C. Invest. Ophthalmol. Vis. Sci. 2012, 53, 1436−1444. (30) Zhang, D.; Xie, Y.; Mrozek, M. F.; Ortiz, C.; Davisson, V. J.; Ben-Amotz, D. Anal. Chem. 2003, 75, 5703−5709. (31) Schwab, I.; Dawson, C. R. In General Ophthalmology, 13th ed.; Vaughan, D., Asbury, T., Riordan-Eva, P., Eds.; Appleton & Lange: Norwalk, CT, 1992; pp 96−124. (32) González-López, J. J.; Morcillo-Laiz, R.; Muñoz-Negrete, F. J. Arch. Soc. Esp. Oftalmol. 2013, 88, 108−115. (33) Sambursky, R.; Tauber, S.; Schirra, F.; Kozich, K.; Davidson, R.; Cohen, E. J. Ophthalmology 2006, 113, 1758−1764. (34) Cooper, R. J.; Yeo, A. C.; Bailey, A. S.; Tullo, A. B. Invest. Ophthalmol. Vis. Sci. 1999, 40, 90−95. (35) Jung, G. B.; Bae, Y. M.; Lee, Y. J.; Ryu, S. H.; Park, H. K. Appl. Surf. Sci. 2013, 282, 161−164. (36) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Nature 1997, 389, 827−829. (37) Gorr, H. M.; Zueger, J. M.; McAdams, D. R.; Barnard, J. A. Colloids Surf. B: Biointerfaces 2013, 103, 59−66. (38) Marin, A. G.; Gelderblom, H.; Lohse, D.; Snoeijer, J. H. Phys. Rev. Lett. 2011, 107, 085502. (39) Hendarto, E.; Gianchandani, Y. B. J. Micromech. Microeng. 2013, 23, 075016. (40) Sun, W.; Zhao, Q.; Zhao, M.; Yang, B.; Cui, C.; Ren, J. J. Agric. Food Chem. 2011, 59, 11070−11077. (41) Chen, P.; Zhang, F.; Lin, L.; Bai, H.; Zhang, L.; Tang, G. Q.; Fang, H.; Mu, G. G.; Gong, W.; Liu, Z. P.; Han, Z. B.; Zhao, H.; Han, Z. C. Stem Cells in Clinic and Research; Gholamrezanezhad, A., Ed.; InTech: Rijeka, Croatia, 2011. (42) Ellis, D. I.; Goodacre, R. Analyst 2006, 131, 875−885. (43) RPS. AdenoPlus, http://www.rpsdetectors.com/in/products/ adenoplus/.
a DCD-SERS-based Raman device can be developed for implementation in ophthalmological clinics within a few days.
■
ASSOCIATED CONTENT
S Supporting Information *
Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
*Phone: +82 2 961 0290. Fax: +82 2 6008 5535. E-mail: [email protected]. *E-mail: [email protected]. Author Contributions
Samjin Choi and Sung Woon Moon contributed equally to this paper. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by a grant from Kyung Hee University in 2013 (Grant KHU-20131086). REFERENCES
(1) Karns, K.; Herr, A. E. Anal. Chem. 2011, 83, 8115−8122. (2) Ohashi, Y.; Dogru, M.; Tsubota, K. Clin. Chim. Acta 2006, 369, 17−28. (3) Zavaro, A.; Samra, Z.; Baryishak, R.; Sompolinsky, D. Doc. Ophthalmol. 1980, 50, 185−199. (4) Kaji, Y.; Hiraoka, T.; Oshika, T. Graefes. Arch. Clin. Exp. Ophthalmol. 2009, 247, 989−992. (5) Fukuda, M.; Deai, T.; Higaki, S.; Hayashi, K.; Shimomura, Y. Semin. Ophthalmol. 2008, 23, 217−220. (6) Pramod, N. P.; Dhevahi, E.; Sudhamathi, K.; Kannan, K.; Thyagarajan, S. P. Ocul. Immunol. Inflamm. 1999, 7, 61−67. (7) McNamara, N. A.; Andika, R.; Kwong, M.; Sack, R. A.; Fleiszig, S. M. Curr. Eye Res. 2005, 30, 517−525. (8) Cao, Z.; Saravanan, C.; Goldstein, M. H.; Wu, H. K.; Pasricha, G.; Sharma, S.; Panjwani, N. Arch. Ophthalmol. 2008, 126, 348−352. (9) Eylan, E.; Ronen, D.; Romano, A.; Smetana, O. Invest. Ophthalmol. Vis. Sci. 1977, 16, 850−853. (10) Thakur, A.; Willcox, M. D. Exp. Eye Res. 1998, 67, 9−19. (11) Sacchetti, M.; Micera, A.; Lambiase, A.; Speranza, S.; Mantelli, F.; Petrachi, G.; Bonini, S.; Bonini, S. Mol. Vis. 2011, 17, 47−52. (12) Lambiase, A.; Micera, A.; Sacchetti, M.; Cortes, M.; Mantelli, F.; Bonini, S. Arch. Ophthalmol. 2011, 129, 981−986. (13) Hidalgo, F.; Melón, S.; de Oña, M.; Do Santos, V.; Martínez, A.; Cimadevilla, R.; Rodríguez, M. Eur. J. Clin. Microbiol. Infect. Dis. 1998, 17, 120−123. (14) Kaufman, H. E.; Azcuy, A. M.; Varnell, E. D.; Sloop, G. D.; Thompson, H. W.; Hill, J. M. Invest. Ophth. Vis. Sci. 2005, 46, 241− 247. (15) Ng, V.; Cho, P. Graefes. Arch. Clin. Exp. Ophthalmol. 2000, 238, 571−576. (16) Willcox, M. D.; Lan, J. Clin. Exp. Optom. 1999, 82, 1−3. (17) Bjerrum, K. B. Acta Ophthalmol. Scand. 1999, 77, 1−8. (18) Grus, F. H.; Podust, V. N.; Bruns, K.; Lackner, K.; Fu, S.; Dalmasso, E. A.; Wirthlin, A.; Pfeiffer, N. Invest. Ophthalmol. Vis. Sci. 2005, 46, 863−876. (19) Fullard, R. J.; Snyder, C. Invest. Ophthalmol. Vis. Sci. 1990, 31, 1119−1126. (20) Barhoumi, A.; Zhang, D.; Tam, F.; Halas, N. J. J. Am. Chem. Soc. 2008, 130, 5523−5529. (21) Taleb, A.; Diamond, J.; McGarvey, J. J.; Beattie, J. R.; Toland, C.; Hamilton, P. W. J. Phys. Chem. B 2006, 110, 19625−19631. 11099
dx.doi.org/10.1021/ac5025478 | Anal. Chem. 2014, 86, 11093−11099
