Label-Free Colorimetric Detection of Gelatinases on Nanoporous

Jan 12, 2008 - lens of the microscope was coupled to a bifurcated fiber optic cable. .... networks of gelatin, which may have pores large enough for t...
0 downloads 0 Views 326KB Size
Download PDF
Anal. Chem. 2008, 80, 1468-1473

Label-Free Colorimetric Detection of Gelatinases on Nanoporous Silicon Photonic Films Lizeng Gao, Njideka Mbonu, Liangliang Cao, and Di Gao*

Department of Chemical and Petroleum Engineering, University of Pittsburgh, Pittsburgh, Pennsylvania 15261

We report the development of a sensor platform for detection of gelatinases based on porous silicon photonic films. The sensor is made by spin-coating gelatin, a substrate protein to gelatinases, onto the porous silicon, which forms a thin, uniform, and smooth gel layer where samples can be directly spotted. The digestion products of gelatin by the active gelatinase present in the sample are able to enter the pores and induce color changes that can be detected by the naked eye. Using this sensor, we have demonstrated the detection of matrix metalloproteinase-2 (MMP-2)san important gelatinase closely associated with tumor aggressiveness and metastatic potentialswith concentrations varying from 0.1 to 1000 ng/ mL in samples with volumes as small as 1 µL. The detection limit of this method, in terms of the minimum quantity of active MMP-2 in the sample that can be detected, is 2 orders of magnitude lower than what has been reported for zymography. Gelatinase is a proteolytic enzyme that hydrolyzes gelatin into its subcompounds including polypeptides, peptides, and amino acids. In human, two important gelatinases that belong to the matrix metalloproteinase (MMP) family, namely, MMP-2 and MMP-9, have been found to be closely associated with tumor aggressiveness and metastatic potential.1,2 Both MMP-2 and MMP-9 are zinc-dependent neutral endopeptidases that are capable of degrading extracellular matrix components and play an important role in tumor growth, invasion, and metastasis.3 The expression of these two gelatinases, although low or undetectable in most normal tissues, is substantially increased in the majority of malignant tumors. Overexpression of gelatinases in malignant tissues in comparison to adjacent normal tissues has been observed in a variety of tumor types, including lung, colon, breast, and pancreatic carcinomas.4-10 In addition, patients with cancer are found to have elevated levels of gelatinases in plasma and * To whom correspondence should be addressed. E-mail: [email protected]. (1) Nikkola, J.; Vihinen, P.; Vuoristo, M. S.; Kellokumpu-Lehtinen, P.; Kahari, V. M.; Pyrhonen, S. Clin. Cancer Res. 2005, 11 (14), 5158-66. (2) Fingleton, B. Curr. Phar. Des. 2007, 13 (3), 333-46. (3) Chambers, A. F.; Matrisian, L. J. Natl. Cancer. Inst. 1997, 89 (17), 126070. (4) Davidson, B.; Goldberg, I.; Liokumovich, P.; Kopolovic, J.; Gotlieb, W. H.; Lerner-Geva, L.; Reder, I.; Ben-Baruch, G.; Reich, R. Int. J. Gynecol. Pathol. 1998, 17 (4), 295-301. (5) Liabakk, N. B.; Talbot, I.; Smith, R. A.; Wilkinson, K.; Balkwill, F. Cancer Res. 1996, 56 (1), 190-6. (6) Hashimoto, K.; Kihira, Y.; Matuo, Y.; Usui, T. J. Urol. 1998, 160 (5), 18726.

1468 Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

urine, compared with healthy subjects.11 Therefore, the levels of gelatinases detected in patients provide important information for cancer diagnosis, prognosis, and adjuvant therapy monitoring.12 Currently existing technologies to detect gelatinases in cancer research include zymography,13 enzyme-linked immunosorbent assay,14 western blot,15,16 and immuno-histochemistry.16,17 Although all of these methods have shown promising sensitivity and specificity for the detection of gelatinases, they are cumbersome, time-consuming, and expensive and require large quantities of samples and reagents. As a result, there has been considerable interest in the development of alternative technologies for rapid, sensitive, and inexpensive detection of gelatinases.18,19 Since the discovery of strong visible photoluminescence at room temperature,20,21 porous silicon (PS) has attracted a lot of attention in Si-based optoelectronics.22-24 Several research groups have studied the PS multilayer structures with a focus on the optical properties of controlled layer stacks that are applicable as interference filters.24 By varying the refractive index sinusoidally through the PS multilayered stacks, photonic structures ap(7) La Rocca, G.; Pucci-Minafra, I.; Marrazzo, A.; Taormina, P.; Minafra, S. Br. J. Cancer 2004, 90 (7), 1414-21. (8) Gonzalez-Avila, G.; Iturria, C.; Vadillo, F.; Teran, L.; Selman, M.; PerezTamayo, R. Pathobiology 1998, 66 (1), 5-16. (9) Tutton, M. G.; George, M. L.; Eccles, S. A.; Burton, S.; Swift, R. I.; Abulafi, A. M. Int. J. Cancer 2003, 107 (4), 541-50. (10) Yang, X.; Staren, E. D.; Howard, J. M.; Iwamura, T.; Bartsch, J. E.; Appert, H. E. J. Surg. Res. 2001, 98 (1), 33-9. (11) Moses, M. A.; Wiederschain, D.; Loughlin, K. R.; Zurakowski, D.; Lamb, C. C. Cancer Res. 1998, 58 (7), 1395-9. (12) Turpeenniemi-Hujanen, T. Biochimie 2005, 87 (3-4), 287-97. (13) Ratnikov, B. I.; Deryugina, E. I.; Strongin, A. Y. Lab. Invest. 2002, 82 (11), 1583-90. (14) Kubben, F. J.; Sier, C. F.; van Duijn, W.; Griffioen, G.; Hanemaaijer, R.; van de Velde, C. J.; van Krieken, J. H.; Lamers, C. B.; Verspaget, H. W. Br. J. Cancer 2006, 94 (7), 1035-40. (15) Lee, A. Y.; Akers, K. T.; Collier, M.; Li, L.; Eisen, A. Z.; Seltzer, J. L. Proc. Natl. Acad. Sci. U.S.A. 1997, 94 (9), 4424-9. (16) Cai, K. Q.; Yang, W. L.; Capo-Chichi, C. D.; Vanderveer, L.; Wu, H.; Godwin, A. K.; Xu, X. X. Mol. Carcinog. 2007, 46 (2), 130-43. (17) Kamat, A. A.; Fletcher, M.; Gruman, L. M.; Mueller, P.; Lopez, A.; Landen, C. N., Jr.; Han, L.; Gershenson, D. M.; Sood, A. K. Clin. Cancer Res. 2006, 12 (6), 1707-14. (18) Harris, T. J.; von Maltzahn, G.; Derfus, A. M.; Ruoslahti, E.; Bhatia, S. N. Angew. Chem., Int. Ed. 2006, 45 (19), 3161-5. (19) Pieper-Fu ¨ rst, U.; Kleuser, U.; Sto ¨cklein, W. F.; Warsinke, A.; Scheller, F. W. Anal. Biochem. 2004, 332 (1), 160-7. (20) Canham, L. T. Appl. Phys. Lett. 1990, 57, 1046-8. (21) Cullis, A. G.; Canham, L. T. Nature 1991, 353 (6342), 335-38. (22) Lee, J. S.; Cho, N. H. Appl. Surf. Sci. 2002, 190 (1), 171-75. (23) Kim, D. A.; Im, S. I.; Whang, C. M.; Cho, W. S.; Yoo, Y. C.; Cho, N. H.; Kim, J. G.; Kwon, Y. J. Appl. Surf. Sci. 2004, 230 (1-4), 125-30. (24) Berger, M. G.; Arens-Fischer, R.; Thonissen, M.; Kruger, M.; Billat, S.; Luth, H.; Hilbrich, S.; Theisz, W.; Grosse, P. Thin Solid Films 1997, 297 (1), 237-40. 10.1021/ac701870y CCC: $40.75

© 2008 American Chemical Society Published on Web 01/12/2008

proximating rugate filters (with high reflectivity in a narrow spectral region) can be fabricated.24,25 The rugate peak position of the PS multilayered photonic structures can be significantly shifted by changes in the refractive index of the PS matrix, which induces a color change visible to the naked eye. With appropriate chemical modification, sensors based on such PS rugate filters have been used to detect biomolecules, explosives, and chemical warfare agents.26,27 The optical property and the porous structure of the PS rugate filters also provide a convenient platform to detect proteases. By tuning the fabrication process, the pore size of the PS film can be tailored to around 10 nm, which is small enough so that very few protein molecules can enter the pore, but large enough for the digestion products of the protein molecules (by the protease) to infiltrate the pores. Thus, by coating the PS film with a thin layer of a substrate protein, proteases can be assayed by spotting the sample on the protein layer and observing the color change of the PS film. Orosco et al. in Sailor’s group28 have used such an approach to assay pepsin A using zein as the substrate protein. In their experiments, the PS film was made hydrophobic by methylation to prevent water from entering the pores, and the hydrophobic protein zein was dissolved in methanol before it was spin-coated on PS films. Following their procedure, the rugate peak of the PS film was observed to shift 20-35 nm after zein was digested by pepsin A in various concentrations and activities. This red-shift was attributed to the introduction of protein fragments, produced by digestion of zein by pepsin A, to the pores with different fractional filling factors.28 However, this approach cannot be directly applied to detect proteases with water-soluble substrate proteins, where water (instead of methanol) solutions of proteins need to be spin-coated on PS films. In order to coat the PS film with water solutions of proteins, the PS film needs to be hydrophilic, and therefore, water contained in the sample may infiltrate the film during the detection procedure. Infiltration of water to the PS film induces a red-shift of more than 50 nm for the rugate peak between 400 and 700 nm, which is more significant than the red-shift (20-35 nm) induced by partial filling of the pores with protein fragments as observed in Orosco’s experiments.28 As a result, color differences between spots loaded with proteases in different concentrations and the blank control sample may not be detectable by the naked eye. Therefore, new processes need to be developed to find wide use of the PS platform for the detection of proteases, particularly those that digest watersoluble substrate proteins such as gelatinases, for cancer research. In this paper, a sensor platform for detection of gelatinases is developed based on the PS photonic structures using gelatin as the substrate protein. Gelatin is a natural protein that dissolves in water at temperatures above ∼30 °C. Its water solution gels when the temperature drops back to room temperature (∼23 °C). This property provides a convenient approach to coat the PS film with a thin, uniform, and smooth gelatin layer in gel form by spincoating gelatin water solution on PS at ∼42 °C followed by cooling (25) Cunin, F.; Schmedake, T. A.; Link, J. R.; Li, Y. Y.; Koh, J.; Bhatia, S. N.; Sailor, M. J. Nat. Mater. 2002, 1 (1), 39-41. (26) Schmedake, T. A.; Cunin, F.; Link, J. R.; Sailor, M. J. Adv. Mater. 2002, 14 (18), 1270-72. (27) Li, Y. Y.; Cunin, F.; Link, J. R.; Gao, T.; Betts, R. E.; Reiver, S. H.; Chin, V.; Bhatia, S. N.; Sailor, M. J. Science 2003, 299 (5615), 2045-7. (28) Orosco, M.; Pacholski, G.; Miskelly, G. M.; Sailor, M. J. Adv. Mater. 2006, 18 (11), 1393-1396.

at room temperature. The gelled gelatin layer, however, is permeable to water. As a result, when samples in water-based buffer are loaded onto the gelatin-coated PS film, water penetrates the gelatin layer instantly and induces the color change. This makes it difficult to detect the gelatinase in the sample directly from the color of the PS film, because little difference in color can be observed between the spots loaded with different concentrations of gelatinase and the blank control. To overcome this problem, two detection schemes have been developed. The first scheme is based on a color gradient observed during the chip drying process on spots loaded with a series of control samples containing different concentrations of gelatinases, which can be used for the semiquantitative detection of gelatinases with concentrations spanning 4 orders of magnitude. The second scheme is based on the employment of glycerol after the chip is completely dried, which produces two distinct colors indicating whether the gelatinase concentration is below or above a critical value. By the use of either of these two detection schemes, active MMP-2 with a concentration of as low as 0.1 ng/mL has been detected. Sample volumes as small as 1 µL can be analyzed, and no toxic chemical agents are needed during the detection. The detection limit of 0.1 pg MMP-2 (0.1 ng/mL in 1 µL sample) is significantly lower than what has been reported for zymography, in terms of the minimum quantity of active MMP-2 present in the sample that can be detected. EXPERIMENTAL SECTION Materials and Chemicals. Boron-doped p-type (100)-oriented silicon wafers with a resistivity of less than 1 mΩ‚cm were purchased from Siltronix (France) and diced into 2 cm × 2 cm chips by a diamond pen. MMP-2 from human fibroblasts and gelatin from porcine skin type A were purchased from Sigma Aldrich. Glycerol and hydrofluoric acid (48% w/w in water) were purchased from EMD Chemicals, Inc. Tris, CaCl2, NaCl, ZnCl2, and ethanol were purchased from Fisher Scientific. Deionized (DI) water (18 MΩ‚cm, Barnstead International) was used in all experiments to prepare buffers. All chemicals were used as received without further purification. Preparation of PS Films. The photonic PS films were prepared by anodic etching of degenerated Si chips, using a previously published process27 after modification. The etching solution consisted of a 3:1 (v/v), 48% aqueous HF-ethanol solution. A Teflon etch cell that exposed 1.6 cm2 of the Si chip was employed. The current density was modulated with a sinusoidal waveform ranging between 38.5 and 192.3 mA/cm2 with a period of 4 s. The etching was performed for 20 cycles, generating a multilayered optical structure known as the rugate filter. After being rinsed with ethanol and dried with N2, the sample was oxidized in an ultraviolet ozone cleaner (Jelight, Inc.) for 5 min. The PS film fabricated by this procedure has a rugate peak of ∼518 nm in air. Coating Gelatin on PS Films. A water solution of 1% gelatin (w/w) was prepared by dissolving gelatin powder in DI water at 42 °C. A volume of ∼100 µL gelatin solution was loaded onto the PS film after the Si chip was mounted on a spin-coater. The Si chip was spun at a speed of 300 rpm for 10 s and then at a speed of 5000 rpm for 2 min. The thin gelatin layer on the PS film was then cooled to room temperature (∼23 °C) in air, which formed a thin, uniform, and smooth gel layer on top of the PS film. Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

1469

Figure 1. Schematic process for the sensor fabrication and detection of gelatinases based on PS photonic structures. The color of the chip is labeled on top of each step in parentheses.

MMP-2 Sample Preparation and Detection Procedure. Samples with different concentrations (ranging from 0.01 ng/mL to 10 µg/mL) of MMP-2 were made by diluting MMP-2 stock solution (100 µg/mL) with a water-based reaction buffer containing 50 mM Tris-HCl, pH 7.5, 10 mM CaCl2, 150 mM NaCl, and 0.5 mM ZnCl2. The blank control sample contains only the reaction buffer. A volume of 1 µL of each sample was spotted onto the gelatin-coated PS film using a pipet. The chip was then stored in a humid chamber (made by placing water-rinsed filter paper inside a Petri dish) at room temperature (∼23 °C) for 12 h to allow the digestion of gelatin by the gelatinase. After the digestion was finished, the chip was removed from the humid chamber and dried in air (∼23 °C and 40% relative humidity). The color change on each sampling spot was observed and recorded by a spectrophotometer during the drying process. Detection of MMP-2 in the samples can be made during this drying process. An alternative detection method was to allow the chip to completely dry in air (∼23 °C and 40% relative humidity) for 1 h followed by loading of a droplet of glycerol (∼1 µL) onto each sampling spot. Data Acquisition and Analysis. All the colors reported in this paper refer to the colors observed on the chip when it was illuminated by white light generated from a tungsten light source. The spectra of the reflected light were recorded using an Ocean Optics HR4000 high-resolution spectrophotometer fitted with an optical microscope (Motic PSM-1000 Probescope). The objective lens of the microscope was coupled to a bifurcated fiber optic cable. One end of the cable was illuminated by the tungsten light source. The light was focused by the objective lens, and a round spot of approximately 1.5 mm in diameter on the PS film was illuminated along the surface normal direction. The reflected light was transmitted through the bifurcated fiber and recorded by the spectrophotometer at the other end of the fiber, in the wavelength range of 400-1000 nm and with a spectral acquisition time of 100 ms. Each experiment was repeated five times. The peak position of the spectrum was observed to differ by less than 10 nm in the repeated experiments. Safety Considerations. Caution should be taken to avoid skin and eye contact with the HF-ethanol solution during the PS film fabrication. To the best of our knowledge, the rest of the fabrication and detection procedure presents no serious hazards. However, when the assay is used in conjunction with unknown 1470

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

Figure 2. As-fabricated PS films. (A) SEM image. (B) Optical image.

Figure 3. Spectra of the reflected light taken from the PS film after it was (from bottom to top) fabricated and dried in air, loaded with water, loaded with glycerol, coated with gelatin and dried in air, coated with gelatin (dried) and loaded with glycerol, coated with gelatin (dried) and loaded with water. Spectra are offset along the y axis for clarity.

biological samples, all proper government safety protocols should be followed. RESULTS AND DISCUSSION A schematic process for the fabrication of the sensor and detection of MMP-2 is shown in Figure 1. The PS film had pores of 10-20 nm in diameter, as shown in the scanning electron microscopy (SEM) image (Figure 2A). The color of the PS film under white light illumination was initially green (Figure 2B) and changed when the pores were filled by solvents with different refractive indices (RI). The color change is presented in Figure 3 using the spectra of the reflected light taken from the PS film. The spectrum taken from the as-fabricated PS film showed a peak around 518 nm, corresponding to the observed green color in

Figure 2B. The peak shifted to ∼598 nm upon loading DI water (RI ∼ 1.34) to the film and to ∼625 nm upon loading glycerol (RI ∼ 1.47). The red-shift of the spectrum was induced by filling the pores with media that had higher RI, which was consistent with previous reports.29 We then investigated the film color change resulted from the gelatin coating. Initially, the film showed a red color (with a peak around 598 nm in the spectrum) after a water solution of gelatin at ∼42 °C was spin-coated onto the film (Figure 1C). The red color remained if the chip was stored in a humid environment, even after the gelatin gelled at ∼23 °C and formed a smooth and thin coating layer (Figure 1D). The color, however, changed gradually back to green when the gelatin-coated chip was dried in air (∼40% relative humidity; Figure 1E). As shown in Figure 3, the peak position (∼521 nm) of the dried gelatin-coated PS film shifted less than 4 nm, compared with that of the as-fabricated PS film. This indicated that the pores of the PS film were small enough so that very few gelatin molecules were able to enter the pores. The dried gelatin layer on the PS film had different permeability to water and glycerol. When a droplet of water was loaded onto the gelatin-coated PS film, the film color instantly changed from green to red. As shown in Figure 3, the peak in the spectrum shifted from ∼521 nm to ∼592 nm. However, when glycerol was loaded onto the same gelatin-coated PS film, no significant color change was observed, and the peak of the spectrum remained at ∼521 nm. We attribute this result to the much larger size of the glycerol molecules compared with that of water molecules, which prevents glycerol from penetrating the network formed by the gelatin molecules. We repeated this experiment by spin-coating different concentrations (0.5, 1, 1.5, 2, and 2.5 wt %) of gelatin solutions on the PS film using the same coating procedure. The gelatin layers formed by spin-coating 1, 1.5, 2, and 2.5 wt % gelatin solutions were permeable to water but not to glycerol. However, the gelatin layer formed from the 0.5% gelatin solution was permeable to both water and glycerol, and loading glycerol to the film induced an instant color change. This is likely due to the fact that the less concentrated gelatin solution forms less dense networks of gelatin, which may have pores large enough for the glycerol to pass through. Gelatin layers formed by spin-coating 1 wt % gelatin solutions were used in the following experiments. The gelatin-coated PS films can be stored either in a humid chamber or in air before they are used. The color of the films was red if they were stored in a humid chamber and green if they were stored in dry air. When the samples containing different concentrations of MMP-2 in a reaction buffer were spotted onto the gelatin-coated PS film, the sampling spots instantly showed a red color due to water infiltration (Figure 1F), regardless of how the chip was stored. In our experiment, six MMP-2 samples at concentrations of 0.1, 1, 10, 100, 1000, and 10 000 ng/mL, respectively, and one blank control sample (that contained only the reaction buffer) were spotted onto the chip. The chip was then placed in a humid chamber for 12 h to allow digestion of gelatin by the MMP-2 in the sample. When the digestion process was finished and the chip was removed from the humid chamber, the entire chip showed a red color because of infiltration of water (29) Lin, V. S.; Motesharei, K.; Dancil, K. P.; Sailor, M. J.; Ghadiri, M. R. Science 1997, 278 (5339), 840-3.

Figure 4. Detection of MMP-2 on a partially dried chip. (A) An optical image of the chip. (B) Spectra of the reflected light taken from the spots that have been loaded with (from bottom to top) blank control, 0.1, 1, 10, 100, and 1000 ng/mL MMP-2 samples. Spectra are offset along the y axis for clarity.

vapor from the humid chamber into the PS films (Figure 1G), and little difference in color could be observed by the naked eye between sampling spots loaded with different concentrations of MMP-2 and the blank control, except for the two spots loaded with the most concentrated MMP-2 samples (1000 and 10 000 ng/ mL, respectively) which showed a dark violet color. The color on the spots loaded with 0.1, 1, 10, and 100 ng/mL MMP-2 samples that showed little difference from the background, however, became distinguishable during the chip drying process in air. After the chip removed from the humid chamber was dried in air (∼40% relative humidity) for about 15 min, most of the chip gradually turned back to a green color, and distinct colors were observed on the spots loaded with 0.1, 1, 10, and 100 ng/mL MMP-2 samples. Figure 4A presents an optical image of the chip in this stage. The spectra of the reflected light taken from each sampling spot were recorded and shown in Figure 4B. The spectra taken from the spot loaded with the blank control sample and the rest of the chip that had not been spotted with samples show a peak around 550 nm. Compared with the peak around 521 nm in the spectrum taken from the completely dried gelatin-coated PS film, there was a ∼29 nm red-shift, suggesting that the PS film was partially filled with water (i.e., partially dried) in this state. The spectra taken from spots loaded with 0.1, 1, 10, and 100 ng/ mL MMP-2 samples showed peaks around 584, 648, 663, and 672 nm, respectively. The colors of these spots varied from dark green, to pink, and to violet, which were distinguishable to the naked eye. The two spots loaded with the 1000 and 10 000 ng/mL MMP-2 samples, respectively, still showed a dark violet color, corresponding to a peak of ∼690 nm in the spectrum, with little difference between the two spots. Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

1471

We attribute the color differences observed on the spots loaded with MMP-2 samples in varied concentrations to the infiltration of varied quantities of digestion products into the pores of the PS film. The capability of MMP-2 to digest gelatin into small fragments has been employed in zymography to detect the activity and expression level of MMP-2.30 The products produced from the digestion of gelatin by MMP-2 include polypeptides, peptides, and amino acids. When these molecules are small enough, they may enter the pores and increase the RI of the media inside the pores, which causes the peak of the spectrum to shift to the longer wavelength region. As the chip drying process continued, the colors on the spots loaded with 0.1, 1, 10, and 100 ng/mL MMP-2 samples gradually changed back to green, and the color difference between these spots and the blank control gradually disappeared. When the chip was nearly fully dried, all these spots showed a green color that was indistinguishable from the background color of the chip by the naked eye. The peak position in the spectra taken from these spots showed less than 15 nm difference between each other, and no apparent trend as a function of the MMP-2 concentrations was observed. A plausible explanation for this result is that the digestion products adsorbed on the walls of the pores (after most water inside the pores evaporates) contribute less significantly to the RI of the medium inside the pores compared to when they are mixed with water and stay homogenously inside the pores. Therefore, less color change or red-shift of the spectrum is induced. The spots loaded with 1000 and 10 000 ng/mL MMP-2 samples remained a dark violet color. This is likely because the quantities of the digestion products produced from these high concentration MMP-2 samples are great enough to occupy most of the space inside the pores, and, therefore, evaporation of water does not significantly change the RI of the medium inside the pores. Apparently, the color differences as observed on the spots loaded with MMP-2 samples in varied concentrations during the chip drying process can be used for the semiquantitative detection of MMP-2. In practice, a sample with an unknown concentration of MMP-2 can be assayed with a series of control samples with known concentrations of MMP-2 on the same chip. The concentration of the unknown sample can be read directly by comparing the color of its sampling spot with the color gradient observed on the spots loaded with the control samples. The disadvantage of this method, however, lies in the fact that the color on each spot gradually changes as a function of time. Therefore, the assay needs to be read within a certain period of time, which is likely to be dependent on the drying process of the sample as a function of a set of parameters including temperature, relative humidity, and ventilation of the environment where the chip is dried. In our experiment, when the chip is dried at ∼23 °C and ∼40% relative humidity without forced air flow above the sample, the best time for reading the assay is 15-30 min after the chip is removed from the humid chamber. In order to be able to read the assay after the chip was almost completely dried, we developed another detection scheme by applying glycerol onto the sampling spots. We found that glycerol was able to penetrate the gelatin layer on the sampling spots (30) Okada, Y.; Morodomi, T.; Enghild, J. J.; Suzuki, K.; Yasui, A.; Nakanishi, I.; Salvesen, G.; Nagase, H. Eur. J. Biochem. 1990, 194 (3), 721-30.

1472 Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

Figure 5. Detection of MMP-2 on a completely dried chip after loading glycerol onto each sampling spot. (A) An optical image of the chip. (B) Spectra of the reflected light taken from spots that have been loaded with (from bottom to top) blank control, 1000 ng/mL deactivated MMP-2, 0.01, 0.1, 1, 10, 100, and 1,000 ng/mL MMP-2 samples. Spectra are offset along the y axis for clarity.

digested by MMP-2 with concentrations greater than a critical value (0.1 ng/mL in our experiment). Figure 5A shows an optical image of a chip after glycerol was loaded onto the sampling spots that had been digested with varied concentrations of MMP-2. For this experiment, the chip was dried in air for 1 h after it was removed from the humid chamber where digestion had proceeded for 12 h. A deactivated MMP-2 sample (1000 ng/mL) prepared by heating the sample at 100 °C for 10 min was used as a reference in addition to the blank control sample. A volume of ∼1 µL of glycerol was applied to each sampling spot afterward. The spectra of the reflected light taken from each spot are shown in Figure 5B. It was apparent that the spots digested by 0.1, 1, 10, and 100 ng/mL MMP-2 samples showed a red color, corresponding to a peak between 625 and 635 nm in the spectrum, whereas the spots loaded with 0.01 ng/mL MMP-2, the deactivated MMP-2 sample, and the blank control sample showed a green color, corresponding to a peak around 518 nm in the spectrum. MMP-2 digests gelatin into small fragments, while simultaneously making pores inside the gelatin gels. When the concentration of the MMP-2 sample is greater than a critical value (0.1 ng/mL in our experiment), the pores formed in the gelatin gels are large enough for glycerol to pass through and induce the color change of the PS film. This detection scheme can be used on completely dried gelatin-coated PS films and therefore requires less control on the drying process. It is very effective in detecting whether MMP-2 is present in the sample at a concentration higher

than a critical value. However, it does not provide a scale to quantify the MMP-2 concentration. It is worth noting that by using either of the two detection schemes, MMP-2 with a concentration of as low as 0.1 ng/mL in a 1 µL sample has been detected. The experiments were repeated five times, and the result was consistent in terms of the detection limit. This corresponds to a total amount of 0.1 pg MMP-2 that has been detected. This detection limit is 2 orders of magnitude lower than what has been reported (∼10 pg) for detection of MMP-2 by zymography,31 although both the method presented here and zymography are based on the digestion of gelatin by active MMP-2 for detection. Because the PS platform is much more sensitive than zymography for the detection of the presence of digestion products, the method presented here shows a much lower detection limit than zymography. By the use of the first detection scheme, the MMP-2 concentration can be assayed between 0.1 and 1000 ng/mL. This concentration range that spans 4 orders of magnitude is of special interest for measuring the expression levels of MMP-2 in cancer research.32 Although the incubation time for the digestion of gelatin by MMP-2 is 12 h in our experiment, it can be significantly shortened with the sacrifice of the detection limit. The effect of incubation time on the detection limit is currently under study. (31) Baragi, V. M.; Shaw, B. J.; Renkiewicz, R. R.; Kuipers, P. J.; Welgus, H. G.; Mathrubutham, M.; Cohen, J. R.; Rao, S. K. Matrix Biol. 2000, 19 (3), 26773. (32) Wong, T. S.; Kwong, D. L.; Sham, J. S.; Wei, W. I.; Kwong, Y. L.; Yuen, A. P. Eur. J. Surg. Oncol. 2004, 30 (5), 560-4.

CONCLUSIONS We have developed a sensor platform for the detection of gelatinases based on the PS photonic structures. Gelatin is used as the substrate protein, which forms a thin layer on the PS film. Digestion of gelatin by the gelatinase produces small molecules that are able to enter the pores and induces color changes that can be detected by the naked eye. Detection of such color changes is complicated by the fact that water may penetrate the gelatin layers and enter the pores, which may induce more significant color changes than the protein fragments. To overcome this problem, two detection schemes have been developed: one is based on monitoring the color change during the chip drying process; the other is based on the employment of glycerol after the chip is completely dried. The developed sensor is able to detect gelatinase without the use of costly antibodies or fluorescent chemical agents. The colormetric detection is convenient and visible to the naked eye. No toxic chemical agents are needed during the detection process. Samples with MMP-2 concentrations of as low as 0.1 ng/mL in a 1 µL sample (0.1 pg MMP-2 in total) can been detected. This detection limit, in terms of the minimum quantity of active MMP-2 in the sample that can be detected, is significantly lower than what has been reported for zymography.

Received for review November 28, 2007.

September

5,

2007.

Accepted

AC701870Y

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

1473