Rapid and sensitive detection of prostate-specific antigen via label-free frequency shift Raman of sensing graphene

Highlights

• The biosensor was first developed based on the specific aptamer and the Raman shift of G peak of graphene.

• The assay allowed to detect PSA with high speed, sensitivity, and simplicity, showing a superior performance.

• The analytic samples can be detected directly without any extensive preparation and label process.

• The strategy exhibited the potential and practical application for determining PSA in the human serum samples.

Abstract

The sensitive and accurate detection of cancer biomarkers is critically important to early clinical diagnosis, disease monitoring, and successful cancer treatment. Here, we first demonstrate an aptamer-based frequency shift Raman approach via sensing of graphene. This biosensor allows the rapid, sensitive, and label-free detection of the acknowledged protein cancer biomarker, prostate-specific antigen (PSA). Monolayer graphene is employed as the Raman substrate, which is highly sensitive to its electronic structure and interface properties. The PSA aptamer can be adsorbed strongly on the surface of substrates through π-π stacking interactions. The vibrational frequency of the G peak of graphene shifted upon the specific binding between the PSA and its aptamer. The corresponding frequency shifts of the G peak are directly correlated with PSA concentrations. The limit of detection is as low as 0.01 ng/mL, with a wide linear range from 0.05 ng/mL to 25 ng/mL. The analytic samples can be detected directly without any extensive preparation and label process. The whole detection is completed in only 30 min. Furthermore, excellent recoveries are acquired to validate the feasibility of this assay in human serum samples. The proposed technology could provide a selective, versatile, and user-friendly strategy for the early detection of cancer biomarkers.

Introduction

Cancer is a major public health problem with high mortality worldwide (Siegel et al., 2019). Therefore, the sensitive and accurate detection of cancer biomarkers is very important to the early diagnosis, monitoring, and treatment of the disease (Wu and Qu, 2015). Prostate-specific antigen (PSA) is a 33–34 kDa glycoprotein and produced primarily by the prostate gland (Lilja et al., 1987). This antigen is a well-known organ-specific biomarker for prostate cancer (PCa) diagnosis. The normal cutoff value for PSA in the serum is below 4 ng/mL and is often elevated in the presence of PCa and other prostate disorders (Lilja et al., 2008; Ma et al., 2014). The successful treatment of these diseases depends mainly on the highly sensitive and selective detection of PSA at early stages.

Raman spectroscopy has emerged as a promising noninvasive detection technique because it provides the chemical fingerprints of the analytes (Austin et al., 2016; McAughtrie et al., 2014). However, the signal of scattering light is quite weak in Raman spectroscopy, limiting the sensitivity of detection. Nevertheless, plasmon-induced surface-enhanced Raman scattering (SERS) can compensate for its deficiency to detect sensitively trace analytes (Schlucker, 2014; Yuan et al., 2017). Typically, gold or silver plasmonic nanoparticles are used for SERS sensors, which can generate strong electromagnetic enhancement (Vigderman et al., 2012). However, SERS presents distinct challenges, such as the uncontrolled agglomeration of plasmonic materials (Osberg et al., 2012) and the expensive and time-consuming extensive functionalization of active Raman tags (Huang et al., 2009). Therefore, a simple, non-plasmonic, and label-free Raman technique should be developed.

In recent decades, the sensitive quantum mechanical effects (Zhu et al., 2016) and high specific surface area (Wang et al., 2018) of nanoparticles are harnessed to develop analytical techniques of biomarkers. The nanoparticles include plasmonic nanoparticles (Lim and Gao, 2016), metal-organic frameworks (Lian et al., 2018), quantum dots (Haldavnekar et al., 2018), and carbon-based nanomaterials (Teradal and Jelinek, 2017). On the basis of the principle of optical, mechanical, chemical, and/or physical actuation, the detection can be achieved by using nanoparticles interfaced with target biomarkers (Keisham et al., 2016). Graphene is a two-dimensional sheet of sp² carbon hexagonal plane networks (Lee et al., 2008). The excellent properties of graphene are widely favored in sensitive sensors. The thickness of the monolayer graphene is ~0.3 nm, and the vibration modes of this material are extremely active to the analytes on its surface (Malard et al., 2009). Graphene is highly sensitive to doping and its Fermi level because of its high quantum capacitance (Das et al., 2008) and carrier-phonon coupling with phononic energies (Basko et al., 2009). The large interfacial area and rich aromatic ring in the graphene structure provide an ideal platform for the attachment of target analytes through π-π stacking interactions (Björk et al., 2010). Graphene is usually characterized with Raman spectroscopy because of the peculiar dispersion of π electrons in this material (Ferrari, 2007). The Raman spectra of graphene only represent three prominent Raman regions, namely, a G peak near 1580 cm⁻¹ (the in-plane vibration mode), a 2D peak at approximately 2700 cm⁻¹ (second-order overtone of in-plane vibration), and a small disorder-induced D peak at approximately 1350 cm⁻¹ (Ferrari et al., 2006). Given that these three characteristic peaks are associated with phonon vibrational modes (Ferrari and Basko, 2013), they possess very high sensitivity to the electronic, structural, and interfacial properties of graphene (Wu et al., 2014). Recently, Berry's group have investigated the change in the frequency shift of graphene interfaced with a single cell for ultrasensitive mapping and monitoring of cancer cell hyperactivity and membrane dipolarity and discrimination between cancer and noncancer cells (Keisham et al., 2016). However, cancer cells of various origins undergo anaerobic fermentation (excessive accumulation of organic acids) (Kato et al., 2013) and show increased sialic acid expression in their membranes (Büll et al., 2014). These processes can induce cell hyperactivity and enhance the electronegative membrane to rearrange the charge of graphene and change its atomic vibrational energy. Hence, the vibration energy shift can be sensitively modified by interfacing almost all cancer cells via a hole-doping mechanism. The specificity of the previous study is insufficient for the highly desirable discrimination of diverse cancerous cells. Aptamers (artificial single-stranded DNAs or RNAs) can bind different entities with high affinity and specificity (Hermann, 2000) to be utilized broadly in bioanalysis.

In this study, we first proposed a rapid, sensitive, and selective biosensor to detect PSA according to the specific recognition of the aptamer and label-free frequency shift Raman of sensing graphene (Scheme 1). Single-stranded nucleic acids can be strongly adsorbed to the surface of graphene through π-π stacking interactions between the hexagonal graphitic network and the ring structures in the nucleobase (Chang et al., 2010; Liu et al., 2013). The normal mode vibrational frequency of the G peak of graphene shifted to a lower frequency upon the adsorption of aptamer, owing to the addition of electrons from the negatively charged aptamer toward the graphene surface. When the PSA solutions with different concentrations were interfaced with the aptamer-adsorbed graphene, the drastic interaction force for specific binding between PSA and aptamer might cause the deformation and strain effect on graphene and led to a strong blue shift in the G peak. The absolute change in the G peak shift at 1579.85 cm⁻¹ for the aptamer-adsorbed graphene had a direct relationship with the concentration of PSA bound to the aptamer. According to the absolute shift change in the G peak position to detect PSA, the response signal was easier to obtain than the peak intensity above the noise in the Raman spectroscopy for the analyte at low-level concentration. This characteristic further improved detecting sensitivity and accuracy. The limit of detection (LOD) was 0.01 ng/mL with a wide linear range from 0.05 ng/mL to 25 ng/mL, and high specificity and stability. The proposed bioassay displayed excellent recoveries and relative standard deviations (RSDs) for determining PSA in the human serum samples. The bioassay was advantageous because the prior purification of the test samples was not necessary, a low sample volume (10 μL) can be used, and the process can be easily manipulated. Compared with existing PSA assays and commercially available enzyme-linked immunosorbent assay (ELISA) kits, the preparation and measurement were label-free, easy to operate, and were completed in only 30 min. Furthermore, the proposed system can be easily adjusted to extensive applications by just changing the types of aptamers. Thus, this bioassay can facilitate the fabrication of assays for cancer biomarkers and provide a favorable reference for the diagnosis and treatment of diseases at early stages.