**1. Introduction**

The lateral flow immunoassay (LFIA) is a common analytical platform for the point-of-care testing of medical diagnostics and environmental monitoring because of its rapidity and simplicity. The LFIA provides clear advantages, including the availability of results within a few minutes, the small volume of an analyzed sample, and inexpensive and user-friendly point-of-care testing [1]. The LFIA combines immunochemical reactions with a chromatography principle. It relies on interactions between an analyte and pre-immobilized recognition elements initiated by the addition of a liquid sample. The LFIA result is a signal at the test line generated by a nanodispersed reporter used. Despite all the advantages mentioned above, the widespread use of LFIAs has been limited by their insu fficient sensitivity. Significant e ffort has been devoted to improving LFIA sensitivity, including the use of alternative labels and detectors, as well as the addition of amplification stages [2,3].

To improve the sensitivity of the immunoassay, integration of the LFIA and surface-enhanced Raman spectroscopy (SERS) was proposed. Because of a simple and cost-effective synthesis, gold and silver nanoparticles are the most common SERS substrates [4]. Typically, nanostructured substrates are functionalized with Raman reporter molecules to produce strong and characteristic peaks in SERS spectra, thus enabling quantitative detection of target analytes. The effectiveness of the SERS-based LFIA technique has been confirmed in numerous recent studies [5–8]. In addition to common AuNPs or latex beads, magnetic and fluorescent particles are used as labels in LFIAs. QDs are used as labels because of their unique optical properties, such as high fluorescence, broad and continuous distributed excitation, photostability, and proven immunoassay effectiveness [9,10]. LFIAs with magnetic and photoluminescent labels showed improved sensitivity for a wide range of analytes [11–14]. Among other markers applied in LFIA, carbon nanoparticles can be mentioned [15,16]. Compared to other labels, carbon nanoparticles are easily detected visually, which contributes to reducing the detection limit of the analyte.

A survey of the literature shows there have been many works published on new immunoassay markers, but they do not go beyond the description of the effectiveness at detecting a particular analyte or report a comparison of the results with conventional gold nanoparticle-based LFIAs. These regularities are poorly transformed into other objects of research. Therefore, the assessment of the test systems with the same reagents that vary according to the kind of marker and readout technique applied will provide more information.

During the study, we explored three approaches to improving LFIA sensitivity. To verify the effectiveness of the proposed methods, we selected the antimicrobial lincomycin (LIN), which is a product of *Streptomyces lincolnensis* bacteria. The known varieties of methods for quantitative detection of LIN include mainly microbiological and chromatographic techniques [17]. The use of accurate chromatographic methods is a common practice to identify and quantify antibiotics in different matrices. Although chromatography–mass spectrometry is a highly sensitive and efficient method, its use requires sample pretreatment, costly equipment, and specially trained personnel [18,19]. Recently, other techniques have also been reported for the determination of LIN in foodstuffs [20,21]. Numerous studies have reported the use of the enzyme-linked immunosorbent assay (ELISA) and LFIA for monitoring LIN residues [22–24]. However, despite the availability of the techniques to control antibiotics, there is grea<sup>t</sup> demand for the development of highly sensitive alternative ways of (a) achieving simple pretreatment procedures (reduce it to dilution eliminating the matrix effect) and (b) minimizing the risk of long-term consumption of contaminants at concentrations below threshold levels.

In this study, the same bioreagents were used to compare different labels and readout systems in a competitive LFIA for LIN. An increase in competitive LFIA sensitivity is possible by reducing the concentration of immunoreagents; however, this decrease is limited by the ability to detect the analytical signal. Beyond the optimization of reagen<sup>t</sup> concentrations, improving the signal-generating elements and readout techniques are other effective strategies to achieve increased assay sensitivity. Moreover, the integration of sensitive detection techniques with LFIA allows for a reduction in immunoreagent consumption.

The current study is a systematic investigation using LFIA integrated with different labels (AuNPs and QDs) and readout techniques (colorimetry, fluorescence, and SERS) to detect LIN. AuNPs were implemented both for traditional colorimetric detection and for coupling to SERS readouts. The quantitative detection of LIN was performed by registering the colorimetric or fluorescence intensity of AuNPs or QDs, respectively, captured on the test line. To design a SERS-based LFIA, AuNPs functionalized with 4-mercaptobenzoic acid (4-MBA) and coupled with anti-LIN monoclonal antibodies (AuNPs–MBA–Ab) were used as a SERS reporter bioprobe. In this case, a conventional LFIA procedure was followed by registration of Raman spectra from the test line.

#### **2. Materials and Methods**
