Quantitative detection of acyclovir by surface enhanced Raman spectroscopy using a portable Raman spectrometer coupled with multivariate data analysis

Highlights

• After optimization, Ag nanoparticles reduced by citrate were used to acyclovir analysis.

• Acyclovir’s characteristic peak obtained by the three methods involved are basically consistent.

• Assignment to vibrations of each characteristic peak.

• Partial least squares model is more suitable for the quantification of acyclovir.

Abstract

Acyclovir (ACV) is a synthetic antiviral agent with serious side effect, particularly its nephrotoxicity, so this study was to explore the ultrasensitive detection of ACV by surface-enhanced Raman scattering (SERS). The enhancement capability of nanoparticles prepared by different chemical reduction were compared, and Ag nanoparticles reduced by citrate are the most propriate enhanced substrate for acyclovir. In addition, comparison between prominent SERS-enhanced bands and the precise mode descriptions predicted through density functional theory (DFT) simulations is used to understand the mechanisms between ACV and metallic surface. 130 different levels of ACV concentrations in a range from 10⁻¹∼10^-7 were used to build quantitative prediction models by two different modeling methods, partial least-squares (PLS) regression and artificial neural network (ANN). Under the optimal conditions, the performance of the PLS model was much better than ANN. The results demonstrated that SERS imaging with multivariate analysis holds great potential for the sensitive and cost effective clinic test of ACV and its metabolites in biological fluids.

Introduction

Acyclovir (ACV), a typical nucleoside antiviral agent derived from guanosine, due to its remarkable therapeutic effect, is generally chosen as the first-line clinical treatment against herpes simplex virus, hepatitis B virus, varicella zoster virus [1]. The drug has its potential in the prophylaxis of cytomegalovirus infections in immunologically compromised patients [2]. Accordingly, tens of millions of patients worldwide have received these antivirals.

Unfortunately, neither the parenteral nor the oral administration of the currently available formulations of ACV is able to result in suitable concentrations of the agent reaching at target sites therefore require the administration of high doses, up to 1.2 g/day [3]. This administration will result in several adverse events and systemic toxicity such as nephrotoxicity. The side effect caused by improper dosage or overdose of ACV is getting more and more attention. Therefore, methods for the quantitative detection of ACV are in great need.

In view of the prominence of ACV in clinical applications, various methods have been developed to determine ACV, including high performance liquid chromatography [4] spectrophotometry [5] and liquid chromatography-tandem mass spectrometry [6]. Although chromatography-based analytical methods have high separation power, sensitivity and/or selectivity, those approaches have several limitations which include lengthy and tedious sample preparation procedures, cumbersome analysis process, difficulties in real-time monitoring and expensive equipment. The colorimetric analysis method is faster, but is more likely to destroy the sample [7]. In the field of spectroscopic methods, our research group used near-infrared spectroscopy coupled with multivariate data analysis methods such as PLS regression and ANN in quantitation and fast measurement of blood glucose in vivo [8,9]. Another possible method is Raman detection which has been used widely to investigate the structural variations in macromolecules [10]. Although it performed outstandingly in providing “fingerprint-like” details about numerous components in a multifaceted system [11], due to only one photon out of one million photons undergoes Raman scattering, the standard Raman scattering signal is too weak to reach the ultra-low concentration of detection. Hence, surface-enhanced Raman spectroscopy (SERS) technique is introduced as an efficacious method to analyze ACV selectively and sensitively.

SERS can maximally enhance the Raman signal of the analyte, even the detailed vibration modes of molecules adsorbed on the noble metal surface can be detected, which greatly extends the application of Raman spectroscopy [[12], [13], [14]]. There are several monographs and reviews that introduced the mechanism of SERS, two principles have been widely recognized so far —— electromagnetic (EM) enhancement and chemical enhancement (CE) [15,16]. In the abovementioned work, metal colloids, in particular Ag and Au, were most commonly used as substrates in SERS. At present, SERS has been widely pursued in various fields such as: explosive detection [17], pesticide and veterinary drug residues in foods [18], additives [19], and drug metabolites in biological samples [20].

Due to the complex nature of SERS phenomena, there is no universal substrate that can be applied for detection of all different chemicals. In this study, AgNPs (Ag nanoparticles) and AuNPs (Au nanoparticles) which prepared by different chemical reduction were investigated. The temporal stability, uniformity and reproducibility of the substrate were evaluated. Secondly, density functional theory (DFT) calculations are engaged to help with the ACV assignments of vibrational modes and to complement the study of ACV adsorption onto nanoparticle surface. Furthermore, SERS coupling with PLS and ANN was applied to build forecast model and explore the feasibility of predicting trace ACV. The results could be considered as a basis for further development of rapid and highly sensitive SERS imaging for non-destructive investigation of drug metabolites in biological fluids.