Sensitive and selective detection of prostate-specific antigen using a photonic crystal nanolaser

The detection of low-concentration biomarkers is expected to facilitate the early diagnosis of severe diseases, including malignant tumors. Using photonic crystal nanolaser sensors, we detected prostate-specific antigen (PSA) from a concentration of 1 fM, which is difficult to detect by conventional enzyme-linked immunosorbent assay. The signal intensity and stability were improved by using a surfactant (i.e., ethanolamine). Even when a contaminant such as bovine serum albumin was mixed into the PSA sample, thereby increasing the concentration of the contaminant ten billion times, it was still possible to maintain a high level of detection. ©2016 Optical Society of America OCIS codes: (280.1415) Biological sensing and sensors; (170.4580) Optical diagnostics for medicine; (230.5298) Photonic crystals; (140.3945) Microcavities. References and links 1. M. Lončar, A. Scherer, and Y. M. Qiu, “Photonic crystal laser sources for chemical detection,” Appl. Phys. Lett. 82(26), 4648–4650 (2003). 2. S. Kita, K. Nozaki, S. Hachuda, H. Watanabe, Y. Saito, S. Otsuka, T. Nakada, Y. Arita, and T. Baba, “Photonic crystal point-shift nanolasers with and without nanoslots—design, fabrication, lasing, and sensing characteristics,” IEEE J. Sel. Top. Quantum Electron. 17(6), 1632–1647 (2011). 3. S. Kita, S. Hachuda, S. Otsuka, T. Endo, Y. Imai, Y. Nishijima, H. Misawa, and T. Baba, “Super-sensitivity in label-free protein sensing using a nanoslot nanolaser,” Opt. Express 19(18), 17683–17690 (2011). 4. S. Hachuda, S. Otsuka, S. Kita, T. Isono, M. Narimatsu, K. Watanabe, Y. Goshima, and T. Baba, “Selective detection of sub-atto-molar Streptavidin in 10-fold impure sample using photonic crystal nanolaser sensors,” Opt. Express 21(10), 12815–12821 (2013). 5. D. Takahashi, S. Hachuda, T. Watanabe, Y. Nishijima, and T. Baba, “Detection of endotoxin using a photonic crystal nanolaser,” Appl. Phys. Lett. 106(13), 131112 (2015). 6. H. Abe, M. Narimatsu, T. Watanabe, T. Furumoto, Y. Yokouchi, Y. Nishijima, S. Kita, A. Tomitaka, S. Ota, Y. Takemura, and T. Baba, “Living-cell imaging using a photonic crystal nanolaser array,” Opt. Express 23(13), 17056–17066 (2015). 7. M. K. Brawer, “Prostate-specific antigen,” Semin. Surg. Oncol. 18(1), 3–9 (2000). 8. H. Yu, E. P. Diamandis, A. F. Prestigiacomo, and T. A. Stamey, “Ultrasensitive assay of prostate-specific antigen used for early detection of prostate cancer relapse and estimation of tumor-doubling time after radical prostatectomy,” Clin. Chem. 41(3), 430–434 (1995). 9. Y. D. Ivanov, V. M. Govorun, V. A. Bykov, and A. I. Archakov, “Nanotechnologies in proteomics,” Proteomics 6(5), 1399–1414 (2006). 10. F. Vollmer, S. Arnold, and D. Keng, “Single virus detection from the reactive shift of a whispering-gallery mode,” Proc. Natl. Acad. Sci. U.S.A. 105(52), 20701–20704 (2008). 11. X. Fan, I. M. White, S. I. Shopova, H. Zhu, J. D. Suter, and Y. Sun, “Sensitive optical biosensors for unlabeled targets: a review,” Anal. Chim. Acta 620(1-2), 8–26 (2008). 12. N. Yang, X. Chen, T. Ren, P. Zhang, and D. Yang, “Carbon nanotube based biosensors,” Sens. Actuators B Chem. 207, 690–715 (2015). 13. J. P. Kim, B. Y. Lee, J. Lee, S. Hong, and S. J. Sim, “Enhancement of sensitivity and specificity by surface modification of carbon nanotubes in diagnosis of prostate cancer based on carbon nanotube field effect transistors,” Biosens. Bioelectron. 24(11), 3372–3378 (2009). 14. G. Wu, R. H. Datar, K. M. Hansen, T. Thundat, R. J. Cote, and A. Majumdar, “Bioassay of prostate-specific antigen (PSA) using microcantilevers,” Nat. Biotechnol. 19(9), 856–860 (2001). #264557 Received 4 May 2016; revised 30 May 2016; accepted 1 Jun 2016; published 3 Jun 2016 © 2016 OSA 13 Jun 2016 | Vol. 24, No. 12 | DOI:10.1364/OE.24.012886 | OPTICS EXPRESS 12886 15. J. Zhao, Y. Zhang, R. Gao, and S. Liu, “A new sensitivity improving approach for mass sensors through integrated optimization of both cantilever surface profile and cross-section,” Sens. Actuators B Chem. 206, 343– 350 (2015). 16. D. Erickson, S. Mandal, A. H. J. Yang, and B. Cordovez, “Nanobiosensors: Optofluidic, electrical and mechanical approaches to biomolecular detection at the nanoscale,” Microfluid. Nanofluidics 4(1-2), 33–52 (2008). 17. C. RoyChaudhuri, “A review on porous silicon based electrochemical biosensors: Beyond surface area enhancement factor,” Sens. Actuators B Chem. 210, 310–323 (2015). 18. M. Narimatsu, S. Kita, H. Abe, and T. Baba, “Enhancement of vertical emission in photonic crystal nanolasers,” Appl. Phys. Lett. 100(12), 121117 (2012). 19. K. Watanabe, Y. Kishi, S. Hachuda, T. Watanabe, M. Sakemoto, Y. Nishijima, and T. Baba, “Simultaneous detection of refractive index and surface charges in nanolaser biosensors,” Appl. Phys. Lett. 106(2), 021106 (2015). 20. K. Matsumoto, N. Konishi, T. Samori, E. Kimura, M. Doi, S. Kato, and Y. Yuki, “ELISA for a complexed antigen with a monoclonal antibody blocking reaction with the free antigen-assay-specific for complexed prostate-specific antigen,” J. Immunol. Methods 234(1-2), 99–106 (2000). 21. M. Sakemoto, Y. Kishi, K. Watanabe, H. Abe, S. Ota, Y. Takemura, and T. Baba, “Cell imaging using GaInAsP semiconductor photoluminescence,” Opt. Express 24(10), 11232 (2016). 22. T. Watanabe, H. Abe, Y. Nishijima, and T. Baba, “Array integration of thousands of photonic crystal nanolasers,” Appl. Phys. Lett. 104(12), 121108 (2014). 23. J. Kim, P. Seidler, L. S. Wan, and C. Fill, “Formation, structure, and reactivity of amino-terminated organic films on silicon substrates,” J. Colloid Interface Sci. 329(1), 114–119 (2009).


Introduction
Photonic crystal nanolasers [1,2] are semiconductor lasers consisting of an air-bridge semiconductor slab with a tiny cavity region and surrounding air holes.They are fabricated by a simple process and are operated easily by room temperature photopumping.Moreover, they function as biosensors; when a nanolaser chip is immersed in an analyte including biomolecules and the biomolecules attach to the exposed cavity, the laser wavelength redshifts (shifts to the long wavelength side) because of changes, for example, in the environmental index.Since the total footprint of one nanolaser is only 20 × 20 μm 2 , their high-throughput, cost-effective mass-production and large-scale integration will be possible.Previously, we reported on the highly sensitive detection of common proteins (including the chemisorption of bovine serum albumin (BSA) from sub-pM regime [3] and specific binding of biotin and streptavidin from sub-aM regime [4]) as well as the high-speed quantification of environmental toxin [5] and imaging living cells [6].However, we have not yet detected the antibody-antigen reaction of real biomarkers, such as those actually used in a medical diagnosis and clinical practice.The first aim of this study is to confirm the sensing of such biomarkers using nanolasers.
For this sensing, we targeted the prostate-specific antigen (PSA; 33-34 kDa) [7] because it is a well-known biomarker in the blood for prostate cancer.If we neglect its high falsepositive rate due to fluctuations between people and health conditions, the critical concentration for a positive diagnosis has been determined to be 4 ng/mL (~100 pM) in human blood.Therefore, a more effective use of PSA may be the detection of the recurrence of prostate cancer.After surgery of prostate cancer, the PSA concentration becomes a very low level, but it increases again when the recurrence occurs.It has been reported that an assay with a detection limit of 100 pg/mL (~3 pM) can identify the recurrence approximately one year earlier than other patients who do not get this checkup [8].To date, such lowconcentrations of biomarkers have been evaluated by bio-chemical methods such as the enzyme-linked immunosorbent assay (ELISA).A typical detection limit of a standard ELISA kit for PSA is 0.3 pM, which is applicable to identifying the recurrence.However, a higher sensitivity is demanded to ensure a sufficient diagnosis margin and monitor the beginning of the recurrence.In general, ELISA shows detection limits for various antigen proteins in the sub-pM to sub-nM regime, depending on the performance of their antibodies, and requires a complicated procedure and relatively high associated costs [9].To solve these problems, various biosensors have been studied, including optical resonators and interferometers [10,11], ion-sensitive field effect transistors [12,13], micro-mechanical systems [14,15], and so on.They have also been summarized in some review papers [16,17].Although these options are potentially more cost effective, most of them are limited to detection limits between sub-nM and 10 nM.Although some of them achieved detection limits considerably lower than 1 pM, they are not stable and reproducible; no reports have discussed this issue.Therefore, the second aim of this study is to demonstrate the high sensitivity of the nanolaser sensor for antibody-antigen reactions (using PSA as an example) with statistical data showing the stability.We conducted experiments not for pure analyte of PSA but also for a mixture that contains PSA and high concentrations of a contaminant.Here, we tested the inclusion of a surfactant, ethanolamine (EA), for the purpose of improving the sensitivity and stability.

Device and method
Based on the description presented in our earlier study [2], the photonic crystal nanolaser consists of a triangular lattice of air holes (230-240 nm diameter and 500 nm pitch) in a GaInAsP single-quantum-well slab (180 nm total thickness), and 2-4 adjacent air holes shifted to form an H0-type cavity.A nanoslot (NS, 30-60 nm width) is added to this cavity, which reduces the temperature dependence of the laser wavelength, narrows the spectral width, and improves the sensing resolution and stability.In addition, an asymmetric modification to the air hole diameter is applied, which narrows the radiation pattern of the laser emission in the vertical direction [18].The device was fabricated using e-beam lithography of the air hole pattern, HI inductively coupled plasma etching of the GaInAsP layer, HCl selective wet etching of InP, and atomic layer deposition of ~3-nm thick ZrO 2 for chemical stabilization [19].Figure 1 shows an example of a fabricated nanolaser array.The device was photopumped by pulsed laser light at 0.98 μm through a 50 × objective lens.The laser emission was directed into optical fiber through the same lens and its wavelength was measured using optical spectrum analyzer (OSA; Advantest Q8383) with a resolution limit of 0.1 nm.All nanolaser wavelengths in the array were automatically measured with a computer-controlled stage and the results were evaluated statistically.Here we summarize the materials used for sample preparation and functionalization.We obtained 3-aminopropyl triethoxysilane (APTES), PSA derived from human semen, a mouse antibody derived from mouse serum, and a total-PSA ELISA kit from Sigma-Aldrich Co. LLC.Glutaraldehyde (GA), BSA, and polyoxyethylene sorbitan monolaurate (Tween20) were all purchased from Wako Pure Chemical Industries, Ltd.The PSA monoclonal antibody (clone No. 4D10) [20] was purchased from Mikuri Immune Laboratory, and EA was purchased from Merck Ltd.Japan.Deionized water with an electrical resistance of 18.3 MΩ was used as pure water for the dilution, rinse, and as an environmental medium in the sensing.
In usual biosensing experiments, proteins are dissolved in a buffer solution to maintain the activity of proteins and stabilize the immobilization condition.However, we actually used pure water here, because we already optimized these conditions using water in previous experiments and then realized recently that the sensing characteristics can be changed by ions included in buffers [19,21].We should optimize the condition for buffers, which will be a near future work.
Figure 2(a) illustrates the procedure for surface modification and PSA sensing.The nanolaser chip was first modified with 0.05% APTES diluted with toluene at 25°C for 5 min and then washed with acetone and methanol, respectively.Next, the chip was soaked in a 2.5% GA solution diluted with pure water at 37°C for 1 h and rinsed in pure water for 10 min.The same immersion conditions were also applied to the subsequent processes.The chip was then immobilized with an antibody in water.The anti-PSA antibody was used for the target detection and the mouse antibody was used for the negative control.We used a 10 μg/mL (67 nM) concentration for the anti-PSA antibody and a 2-4 μg/mL (13-27 nM) for the mouse antibody as optimum concentrations, as discussed later.The chip was also soaked in a 1 M EA solution simply diluted with pure water at room temperature.The effect of this process will also be discussed later.We then submerged the chip in solutions representing a wide range of PSA concentrations, rinsed in pure water, and measured the nanolaser wavelengths in pure water at 25°C.To prepare pure PSA samples, the most concentrated PSA solution was systematically diluted with pure water by pipetting repeatedly.Each PSA concentration was calibrated using the commercial ELISA kit and microplate reader, ChroMate® 4300 (Awareness Technology Inc.), with a detection limit of 0.3 pM.The same dilution process was repeated to produce lower concentrations, and their concentrations were simply estimated from the dilution rate.For the preparation of impure PSA samples, a second set of the same PSA solutions were prepared and then mixed with the BSA contaminant and Tween 20 as a surfactant, where their final concentrations were 1 μM or 10 μM and 0.1 vol%, respectively.These samples were compared to know if these additives affected the PSA detection.

Results and discussion
Figure 2(b) shows the laser spectrum at each modification.The full width at half maximum of the spectrum was typically ~0.2 nm, which was mainly determined by the resolution limit of the OSA used.The fluctuation of the peak wavelength during the long-time measurement was ± 0.06 nm and that between repeated immersion, measurement, and extraction of the chip was ± 0.1 nm.Therefore, a shift of more than ± 0.1 nm was considered to be significant in this study.In the same figure, we observed such shifts after immobilization of the antibody and adsorption of PSA.To optimize the immobilization of the antibody, we measured the wavelength shift for the adsorption of 1 pM PSA, changing the concentration of the antibody in the immobilization, as shown in Fig. 2(c), where the EA treatment was omitted.A significant PSA shift was observed when the antibody shift was 0.7-0.8nm.The smaller antibody shift meaning a smaller amount of antibodies immobilized on the nanolaser surface actually decreased the PSA shift.Conversely, when the antibody shift was too large, the PSA adsorption might be limited by steric hindrance and other factors because of excessive adsorption of the antibody.In subsequent experiments, the antibody shift was fixed at 0.8 nm, which corresponded to a concentration of the anti-PSA antibody of 67 nM and that of the control antibody of 1.3-2.7 nM.Next, we compared the shift for various PSA concentrations, between different antibody conditions, and with and without the inclusion of EA, as shown in Fig. 3.Each plot shows an averaged value based on results from 10 to 25 nanolasers measured simultaneously.The inner and outer error bars show the standard error and standard deviation, respectively.With the anti-PSA antibody, the significant shift appeared from a PSA concentration of 100 fM, even without EA.With EA, the shift appeared from 1 fM and the shift was approximately two times greater than that observed at 100 fM without EA.The shift was almost saturated beyond 100 fM.A possible reason is that the saturated adsorption inside the nanoslot has a higher sensitivity than the outside [3].On the other hand, for devices with the control antibody and without antibodies, significant shifts were not observed for any PSA concentrations both with and without the inclusion of EA.However, we did observe some fluctuations in such a selective detection between trials.We performed 4-6 trials on different dates, repeating all procedures for each experimental condition, and each measuring 6-25 nanolasers.The wavelength shifts were distributed, as summarized in Fig. 4. Here, the histogram is normalized for each trial and those for all trials are summed, because the fluctuation was smaller between nanolasers in the same trial and larger between different trials.Significant shifts appeared for the anti-PSA antibody even without the inclusion of EA.The shifts were observed from 1 fM, but the majority of the shifts started from 10 to 100 fM.With EA, the yield was improved and shifts were observed from 1 fM in many nanolasers and trials.However, the inclusion of EA also increased the shift for the control antibody and some nanolasers exhibited a significant shift beyond the fluctuation.In general, EA as a surfactant is expected to suppress the physisorption of proteins by blocking unbounded GA and washing out the physisorbed antibodies and antigen by hydrophilization.However, the shift of the control antibody indicates that in this experiment this was not the case.Another effect of EA is the suppression of excessive hydrophobicity at the nanoscale level of rough surfaces by hydrophilizing the device surface.This allows the antigen to effectively penetrate into holes of the photonic crystal slab and approach the antibody inside the holes and on the backside of the photonic crystal slab.In any case, it may be necessary at this stage to repeat the measurement and the statistical analysis.An important advantage of using nanolasers is that we can integrate a huge number of nanolasers on a chip [22], which allows us to acquire statistical data easily in automatic measurements so that the shift beyond the noise is effectively suppressed.We also detected PSA selectively in the impure sample with EA. Figure 5 shows the wavelength shift for the PSA concentration, where the wavelength in the impure sample without PSA was used as a reference.For the nanolaser with the anti-PSA antibody, the significant shift occurred from the PSA concentration of 1 fM, which is comparable to the case without the contaminant.When the concentration of the contaminant was increased from 1 μM to 10 μM, the overall shift decreased, which might be because of physisorption of the contaminant to the antibody, inhibiting the adsorption of PSA.However, the shift was still sufficient to confirm the selective detection of the PSA adsorption.The selectivity of the PSA detection against the contaminant is evaluated to be ten billion.This selectivity is smaller than that obtained in the detection of biotin-streptavidin specific binding in a sample with BSA [4], but still reasonably large, considering the 5-6 orders of lower affinity constant of the PSA antibody-antigen binding compared with the biotin-streptavidin.In human blood, the total amount of proteins included is 1.2 mM (serum albumin, 0.8 mM).When the detection was conducted using blood, the nanolaser could detect ~100 fM PSA after diluting the sample 100 times.Thus, this performance is still superior to that of the aforementioned ELISA kit.On the other hand, the nanolaser with the control antibody exhibited a blue shift with increasing PSA concentrations.Similar blueshifts were observed repeatedly when the chip was soaked in impure samples.The APTES, which is the base used to immobilize the antibody, is applied to the nanolaser surface by dehydration condensation of the hydroxyl groups and also by physisorption [23].When the chip was soaked in the impure sample, desorption of the physisorbed APTES may occur because of the competitive adsorption of the contaminant; this results in desorption of the antibody and blueshift.

Conclusion
We succeeded in sensing the antibody-antigen reaction using photonic crystal nanolasers, employing PSA as an example target.By optimizing the immobilization of the antibody and injecting surfactants, we detected 1 fM PSA from impure samples with a maximum selectivity of ten billion.This suggests that the potential performance of nanolasers as immunosensors for various target antigens is superior to that of standard ELISA techniques.Even limiting the target to PSA, it enables the monitoring of potential recurrences of prostate cancer after surgical procedures.
To improve and stabilize their performance, the sensing mechanism must be investigated further.In particular, the ultrahigh sensitivity cannot be explained only by the refractive index sensing.Another mechanism, such as the sensing of surface charges, will be of great interest for further investigations [19,21].The use of buffer solution instead of water in the immobilization of the antibody and antibody-antigen reaction will be another issue to further stabilize the sensing, although it will require particular optimizations because the sensitivity to surface charge is changed by the buffer solution.Furthermore, the quantification of target molecules is necessary for practical biomarker sensing.Similar to ELISA, one approach is to specify the dilution rate that corresponds to the detection limit concentration.The low-cost fabrication of the chip will be crucial for their disposable use.Provided that the productivity scale is larger than 10,000 chips/day, the chip cost will be no higher than 1 Euro/chip, even including the functionalization procedure, which is smaller than the expendable cost of ELISA.After their use, these chips may have to be retrieved and reused because the semiconductor includes rare metals such as Ga and In, and toxins such as As and P. For reading sensor signals, a pump source, automatic stage, and OSA are necessary.The OSA is a major cost but is comparable to ELISA readers.

Fig. 1 .
Fig. 1.Fabricated 5 × 5 GaInAsP photonic crystal nanolaser array and a magnified scanning electron microscope image around one nanolaser which has a NS at the center.Each nanolser (square area) has two side grooves which accelerate the selective wet etching of InP underneath the GaInAsP layer.

Fig. 2 .
Fig. 2. (a) Surface modification procedure.(b) Laser spectrum at each step of the procedure.(c) Wavelength shift because of PSA as a function of wavelength shift because of PSA antibody.EA was not used.The ± 0.1 nm range of the wavelength shift shows the fluctuation of the nanolaser and measurement.

Fig. 3 .
Fig. 3. PSA sensing (a) without EA and (b) with EA.Circular (red), square (blue), and triangular (black) plots show the wavelength shift with the anti-PSA antibody, mouse antibody as a negative control, and without antibodies, respectively.Gray area shows the undetectable range of the concentration by ELISA.

Fig. 4 .
Fig. 4. Distribution of nanolasers showing the corresponding wavelength shifts in repeated trials.Red, blue, and gray bars show the frequencies that correspond to the redshift, blueshift (shift to the short wavelength side), and no significant shift, respectively.The black line shows the Gaussian fitting with average values indicated by the arrow.

Fig. 5 .
Fig. 5. PSA sensing in the impure sample, where EA was used.Concentrations of the BSA contaminant were (a) 1 μM and (b) 10 μM.Circular (red) and square (blue) plots show the wavelength shift with the PSA antibody and the mouse antibody, respectively.