The potential of SERS as an AST methodology in clinical settings

2.5.1 Overview of pioneering work for AST determination using label-free SERS

The desire to create bacteria identification SERS based instrument capable of rapid pathogen detection and identification was pioneered and first report showing bacterial discrimination using SERS was presented by Goodacre and Jarvis in 2004, mentioned in excellent review by the same group [48]. The technique has been further developed by the dedicated SERS groups mentioned in the previous paragraph “SERS – the method of choice” involved in bacterial research.

However, out of the research groups involved in SERS bacterial research it was the Wang group that, for the first time, introduced the term SERS-AST in September 2016 by Scientific Reports, Liu et al. [38]. Thus, it is highlighting that the SERS technique is capable of AST determination.

In the study, the group from the National Taiwan University and the Academica Sinica exploited previous knowledge and followed the previously published SERS papers targeting bacteria [59], [60]. Wang’s group [38] described in detail how susceptible strain of bacteria reacts when exposed to an antibiotic. The group found out that the intensity of specific SERS biomarkers is significantly decreased in 2 h after the reaction with antibiotic, which is illustrated in Figure 4. These findings have been further exploited for rapid AST test. As was previously mentioned, the group termed this procedure as SERS-AST [38] to emphasize the full capacity of SERS.

In order to emphasize the speed of the introduced SERS-AST technique authors presented a schematic comparison between the SERS-AST method and the standard broth dilution (BD). Briefly, the two methods start with the two standard steps – (i) overnight bacterial culture and, (ii) subsequent adjustment of bacteria concentration. However, the following step in the SERS method – inoculation with different antibiotic – took only 2 h, which is much faster than the BD method with an overnight inoculation. Authors discovered that, when a susceptible strain of bacteria is exposed to an antibiotic, the intensity of specific biomarkers in its SERS spectra drops evidently in a few hours. Here, as the SERS “marker” spectrum at 730 cm⁻¹ was selected—the SERS biomarker signal of gram-positive bacteria. Note that 730 cm⁻¹ line has been recently used by other group and this will be discussed further in this review.

SERS signal for samples treated with the antibiotic was compared with that of reference samples which were without the added antibiotic. This large study included experiments with four different bacteria Staphylococcus aureus, Escherichia coli, Acinetobactercter baumannii, and Klebsiella pneumoniae which were determined by SERS-AST method and compared with the standard agar dilution (AD) method.

Two years later, in 2018, the Wang group published a follow up research [61]. In this subsequent study, the attention of the group shifted towards the analysis of the biomolecules responsible for bacterial SERS biomarkers found earlier Gram-positive S. aureus (SERS marker at 733 cm⁻¹) and Gram-negative E. coli (SERS marker at 660, 725 or 742 cm⁻¹) which have been identified as purine derivative metabolites involved in bacterial purine salvage pathways. Using ultra performance liquid chromatography/electrospray ionization-mass spectrometry (UPLC/ESI-MS), the time dependences of the concentrations of these molecules were measured during cells starvation. Note that bacteria were incubated in water – which is essentially the situation facing the bacteria when they are going through the washing and SERS measurement procedures used in previous SERS-AST study [38]. Surprisingly, a single S. aureus and E. coli cells were found to release millions of adenine and hypoxanthine into the water environment in 1 h. Based on the SERS spectra and quantitative UPLC/ESI-MS measurements of the bacterial supernatants, Wang research group confirmed that the molecules responsible for the biomarkers in the SERS spectra of S. aureus are adenine, guanine, and AMP; while that for E. coli are adenine, hypoxanthine, guanine, and xanthine.

To be noted is that, there are two possible mechanisms for the accumulation of nucleobases in the bacterial supernatants during water incubation period. The nucleobases could be released from live cells in response to stress, or they could be leaked out from physically ruptured dead cells. To differentiate between these two mechanisms, the viability of bacterial cells after different water incubation time was performed in this study. Moreover, these results confirmed findings presented in the following paper published by Premasiry et al. [55] in 2016.

Now we return to 2016 once again – specifically to Liu et al. in 2016 [38], Premasiri et al. [55] investigated the contribution of the dominant molecular species contributing to the SERS spectra of bacteria excited with a laser at 785 nm. They identified metabolites of purine degradation: adenine, hypoxanthine, xanthine, guanine, uric acid, and AMP. These results provide the biochemical basis for further development of SERS as a rapid bacterial diagnostic and illustrate how SERS can be applied more generally for metabolic profiling as a probe of cellular activity. Which is why we demonstrate this very interesting finding in more details.

The observed SERS spectra of bacteria result from above mentioned metabolites that have been secreted during the sample preparation and manipulation – bacterial strains were cultivated overnight before subculture and harvested during the log phase by centrifugation of 2 mL of culture. Consequently, washed four times with 2 mL of deionized Millipore or distilled water. The concentration of these molecules is largest in regions closest to the cells from which they are secreted. However, the metabolites are found in the supernatant solution as well.

Researchers found that the supernatant spectra were nearly identical to the corresponding bacterial cell SERS spectrum. Consequently, the main finding was that bacterial SERS signatures are not due to structural bacterial cell wall features. Thus, they must arise from small molecules sufficiently water soluble at biological concentrations, which have been secreted from the bacterial cells and collect in the exogenous regions of these organisms.

Very interesting results are the best fits of the 10 bacterial SERS spectra shown in Figure 5 resulting from a linear combination of the six purine component illustrated in Figure 6. The best fits are displayed in Figure 7. Here, nearly all vibrational features and their relative intensities, seen in each of these bacterial SERS spectra, are captured by this fitting procedure.

Figure 5:

Surface-enhanced Raman spectroscopy (SERS) spectra of 10 bacterial species excited at 785 nm on Au NP substrates. The spectra are averages of four to six individual spectra and normalized to the intensity of the strongest feature in each spectrum. Gram-positive (plus sign) and gram-negative (minus sign) types are indicated. Reprinted with permission from Premasiri et al. [55]: Springer Nature, Copyright (2016).

Figure 6:

Surface-enhanced Raman spectroscopy (SERS) spectra of 20 μM aqueous solutions of the indicated purine components of bacterial SERS spectra. The spectra have been offset for viewing and are normalized to the maximum peak intensity of the 20 μM adenine solution. AMP adenosine monophosphate. Reprinted with permission from Premasiri et al. [55]: Springer Nature, Copyright (2016).

Figure 7:

Empirically determined best fits (red) of the bacterial spectra (black) shown in Figure 5 to a linear combination of purine (adenine, hypoxanthine, xanthine, guanine, uric acid, and AMP) surface enhanced Raman scattering (SERS) spectra shown in Figure 6. Excellent fits are obtained for all bacterial SERS spectra. Reprinted with permission from Premasiri et al. [55]: Springer Nature, Copyright (2016).

It is apparent that (i) adenine makes the overwhelmingly dominant contribution to the Streptococcus pneumoniae TIGR4 and S. aureus NCTC 8325 SERS spectra, (ii) hypoxanthine makes the largest molecular contribution to the SERS spectra of Bacillus anthracis Sterne, Enterococcus faecium DO, and Enterococcus faecalis ATCC 29212, and (iii) high uric acid contributions are observed only in the Pseudomonas putida S16 and A. baumannii ATCC 17978 SERS spectra.

In the consequent study of Premasiri et al. [62] SERS spectra of 12 bacterial strains (urinary tract infection (UTI) clinical isolates) grown in urine were targeted [62]. Again, spectra for selected bacteria are shown to be primarily due to seven purine components: adenine, hypoxanthine, xanthine, guanine, AMP, uric acid, and guanosine (Figure 8). Researchers followed the steps shown in Figure 9. As can be seen from the schematic diagram in the last step 1 μL of this resulting bacterial suspension was pipetted directly onto the SERS substrate and allowed to dry for ∼5 min before SERS spectral acquisition.

Figure 8:

The relative contribution of each of the seven purines found to contribute to the bacterial surface enhanced Raman scattering (SERS) spectra is summarized in this bar graph. Strain-specific differences are evident, it can be seen how the pattern of contributing purines is more different between the four species than between the strains of a given species. Reprinted with permission from Premasiri et al. [62]: Springer Nature, Copyright (2017).

Figure 9:

A summary of the sample preparation procedure for the acquisition of a surface enhanced Raman scattering (SERS) spectrum from a human urine sample spiked with 10⁵ cfu/mL bacteria. A four-stage gravitation filtration system (glass wool and sequential 30-, 10-, and 5-um pore size nylon filters) removed nonbacterial solid materials followed by centrifugation to achieve enrichment. The entire procedure took ∼50 min. Reprinted by permission from Premasiri et al. [62]: Springer Nature, Copyright (2017).

Consequently, partial least squares-discriminant analysis (PLS-DA) classification treatment was employed which enabled strain level identification with >95% sensitivity and >99% specificity. It should be noted that here no growth or cell culturing is required, which in turn is time demanding step at traditional bacterial identification methods. Consequently, UTI diagnosis can be accomplished in less than an hour via SERS with antibiotic specificity. Thus, presented SERS-based technology could be used at POC as a fast (under 1 h), growth-free, relatively low cost diagnostic tool.

Independently, the Munich group by Kubryk et al. [56] in 2016 also confirmed the origin of the SERS band around 730 cm⁻¹ used in paper by Liu et al. [38]. The main driving force behind these experiments was to clarify the origin of the band which appears in SERS spectra of different bacteria. The mentioned band was usually assigned to glycosidic ring vibrations, adenine or to CH₂ deformation. In this illustrative study Kubryk employed a stable isotope approach with cells grown on unlabeled (¹²C, ¹⁴N) and labeled (¹³C, ¹⁵N) carbon and nitrogen sources in different combinations. The position of selected bands of polysaccharides, phospholipids and adenine-related substances depend on their different molecular structure. Thus, determination of the spectra origin of stable isotope labeled bacteria was possible. For this four samples of E. coli were cultivated on unlabeled (¹²C, ¹⁴N) and labeled (¹³C, ¹⁵N) carbon and nitrogen sources in different combinations and the spectral changes could be recorded. Authors concluded that the band at 730 cm⁻¹ can be assigned to adenine (in-plane ring breathing mode). It was noted that another compounds may also contribute to this band which contains adenine such as FAD, NAD, etc.) as well as different products of the purine degradation pathway (e.g. hypoxanthine).

The report by Weiss et al. 2019 [57], which is a joint paper by Niessner (Munich) and Wagner (Vienna) groups involved in SERS investigations, further exploited the origin of microbial SERS signals and parameters which affect reproducibility of SERS spectra. SERS signals from six phylogenetically diverse microorganisms representing different cell compositions were analyzed to investigate the variability and reproducibility of SERS technique in different physiological conditions of the cells.

Authors also focused on the strong signal at 730 cm⁻¹ in SERS spectra of E. coli cells and concluded that the SERS intensity of the marker represents the starvation-induced release of substances related to the metabolic activity [38, 56, 61, 62]. Thus, metabolically inactive and dormant cells would not release purine derivatives which explain that slow growing microorganisms might not exhibit the characteristic SERS signal at 730 cm⁻¹. It was concluded that the effect of physiological state of cells must be taken into the account; otherwise an incomplete understanding may give rise to pitfalls in the characterization and evaluation, and to misinterpretations.

This identification of metabolically inactive and dormant cells via SERS could work in the favor of recent findings by Dengler Haunreiter et al. [15]. Here, resistance to antibiotic evolved over time and in turn S. epidermidis bacteremia was observed despite the full antibiotic treatment in a patient who developed a break-through bacteremia. Antimicrobial tolerance could be achieved by a small subpopulation of bacteria dormant cells which can survive the antibiotic treatment. Here, as mentioned above, SERS should be able to detect this subpopulation of dormant cells and also different subpopulations of cells such as resistant, tolerant, susceptible, and persistent.

The exhaustive study [63] illustrates the quantitative differentiation of bacteria labeled with isotopes. When bacteria is incorporated with varying isotopic ratios the SERS spectral bands displayed clear redshifts that were used as the basis for quantitative characterization of bacteria at the molecular level.

The findings of this study also highlight the importance of the choice of protocols (e.g. microbial sample preparation and NP synthesis) and instruments (laser wavelength) used for the bacterial SERS analysis, which may consequently reveal or suppress specific spectral features. For example, Kubryk and colleagues [56] used hydroxylammonium chloride solution as the reduction agent and employed a 633 nm laser to acquire SERS spectra of isotopically labeled bacterial cells. These authors reported major shifts occurring at 733 and 1330 cm⁻¹, whilst in this study using sodium borohydride as the reduction agent and a 532 nm laser, the peak at 733 cm⁻¹ was, surprisingly, not detected.

That broadly confirms the finding of Jarvis et al. [48] pointing out that one of the fascinating aspects of bacterial SERS is the radical difference in chemical information observed when experimental parameters are altered. Here, the spectra obtained using laser at 785 nm, with a citrate reduced silver colloid and NaCl aggregating agent were compared to the spectrum of E. coli this time acquired with green excitation laser at 532 nm and a borohydride reduced silver colloid [48]. The spectra were entirely different, despite the fact that the cells have very similar surface chemistries. Thus, slight change in the experimental parameters could well be the reason why the peak at 733 cm⁻¹ was not detected in the finding by Chisanga et al. [63]

2.5.2 Recent research focused on lab-on-a-chip applications suitable for POC applications

In the following section we will return to the Taiwan group which subsequently published a follow up of the previous findings by Chang et al. in 2019 [64]. Thus, the SERS-AST method was again revisited by dedicated Wang group. To address the problems related to the antibiotic susceptibility, a microfluidic system integrating membrane filtration and dedicated SERS-active substrate (MF-SERS) was developed. The main aim of this study was to perform on-chip bacterial enrichment, metabolite collection, and in situ SERS measurements for AST. In this study E. coli was used as the prototype bacterium with detection limit of 10³ CFU/mL. The bacteria and secreted metabolites are enclosed during bacterial trapping, metabolite filtration, and SERS detection, thus minimizing possible contamination and human errors. Here, bacteria were trapped at the filter which in turn enabled only secreted metabolites to flow through the membrane filter to the SERS detection zone. The operation of the MF-SERS system is illustrated in Figure 10 and can be divided into four steps: (1) Filtration on membrane filter, (2) washing, (3) incubation, and (4) detection. First, the bacterial solution was injected into the filtration zone at 2 mL/min and bacteria were trapped on the membrane filter. An amount of 1.5 mL of deionized water was then injected to the chamber to force the culture medium to pass through the filter. The trapped bacteria then released metabolites in the chamber filled with deionized water for 20 min (incubation at room temperature). Consequently, only the metabolite solution in the chamber was subsequently compelled through the filter and guided into the SERS detection zone for SERS measurements. The total operational time was 30 min long.

Figure 10:

MF-SERS system: (A) Schematic diagram of the microfluidic system and device; (B) scanning electron microscopy (SEM) image of SERS substrate (scale bar, 200 nm); and (C) photo of the microfluidic device. Reprinted with permission from [64], Copyright (2019) American Chemical Society.

Presented MF-SERS system improved three issues in the previous SERS-AST method: (1) Entailing prolonged culture time to obtain a large number of bacteria, (2) human errors and contamination owing to manual operations, and (3) large SERS signal variation due to nonuniform and unstable molecular adhesion on the SERS substrate.

Finally, the authors claim that employing miniature size and well-confined microenvironment allows the integration of multiple bacteria processes for bacterial enrichment, culture and determination of AST. Thus, MF-SERS systems can potentially be assembled in a parallel for determination of AST and minimum inhibitory concentration (MIC) of different antibiotics in clinical samples (e.g. whole blood or urine). The SERS substrates in this study were made of arrays of silver NPs embedded in porous anodic aluminum oxide (AAO) nanochannels.

In the middle of 2020, again, the Taiwan group came up with two excellent manuscripts published in the prestigious journals [58], [65]. The researchers followed previously explored topic of rapid antibiotic susceptibility testing of bacteria based on findings that SERS spectra of bacteria originate primarily from the metabolites of purine degradation rather than the cell wall structures: Adenine, hypoxanthine, xanthine, guanine, uric acid, and AMP [61].

This time they dedicated their research to a protocol based on SERS in order to obtain consistent AST results from clinical blood-culture samples within 4 h [65]. This novel methodology could be used for clinical diagnostics to the benefit of patients as well as to the economy for timely administration of appropriate antibiotic therapy. The study was conducted in the National Taiwan University Hospital (NTUH). Here, again, authors stressed a linear increase in the risk of mortality for patients with e.g. bacteremia (bacteremia could be in most cases associated with sepsis) for each hour delay in antibiotic administration. Because of the lack of timely microbiological evidence, antibiotics are usually forced to start empirically, rather than precisely upon a specific target.

The main aim of this study was to apply SERS for rapid AST from clinical blood-culture samples, which in turn, is a major step for SERS application in clinical microbiology. This is an extremely challenging task because of the complex nature of clinical blood samples and pretreatment procedure of samples to separate the blood components from the bacteria has to be applied. Thus, the red blood cells containing hemoglobin were lysed using ACK (Ammonium–chloride–potassium) and consequently separated from „blood-culture isolates“ along with other constituents in blood samples affecting SERS spectra – providing that targeted bacteria are left viable – using dedicated procedure. In the experiments the group followed the protocol described in the previous SERS study on reference strains and pure clinical isolates where the spectral peaks located at 730 and 724 cm⁻¹, respectively, were identified and adopted as the SERS biomarkers of S. aureus and E. coli, based on purine metabolites. The release of purines strongly depends on applied antibiotic – e.g. the purine biosynthesis pathway of susceptible E. coli strain is influenced by antibiotics while that of resistant strain is unaffected. Thus, the SERS biomarkers were used for bacteria AST determination. Bacterial samples were pipetted on the active SERS substrate based on two-dimension Ag NP array embedded in nanochannels of anodic alumina [66] and dried on a hot plate in order to perform SERS measurements. Laser at 632.8 nm was used for excitation.

The second manuscript was published in 2020 by Lab on a Chip journal dedicated to microfluidic applications [58]. Here, again, authors addressed the problem of conventional AST which usually requires a prolonged bacterial culture time and a labor-intensive sample pretreatment process. They report on the development of a microfluidic microwell device which integrates SERS so that a rapid and high-throughput AST can be achieved. Presented results show that the Microwell-SERS system can successfully encapsulate bacteria in a miniaturized microwell with a greatly increased effective bacteria concentration which in turn enables a 2 h AST on susceptible and resistant E. coli and S. aureus. This instrument integrates a SERS substrate (Microwell-SERS system) for bacteria isolation, followed by the enrichment and in situ AST shown in Figure 11. As a result SERS spectra of a bacteria-secreted metabolite after antibiotic treatments can be monitored and measured. Compared to the NP-based SERS probe, the nanostructure-based SERS substrate is believed to have a higher sensing uniformity and reproducibility since it has a more uniform nanostructure arrangement and would not suffer from NP aggregation issue.

Figure 11:

The Microwell-SERS system: (A) Schematic diagram of the system, including Raman microscopy and the Microwell-SERS device; (B) photo of the microwell device attached with a SERS substrate with the SERS-active surface (in purple) facing downward; (C) bright-field image of the microwells rehydrated with blue food dye (scale bar, 200 μm); (D) fluorescence image of 106 CFU mL⁻¹ GFP-transfection bacteria confined in nine microwells (scale bar, 50 μm). Reprinted with permission from [58], Copyright © Royal Society of Chemistry 2017.

Bacteria secreted metabolites were also reported by Andrei et al. [67] on generation of reproducible SERS spectra for complex structures of bacteria and their aggregation. The studies demonstrate that densely packed bacteria deposited on ultrathin silver films (thermally evaporated Ag films) were characterized with the secretion of adenosine triphosphate (ATP) for S. aureus which enabled discrimination between the various strains. Silver thin films were deposited using a home-made thermal evaporator and a 7-nm-thick. Ag film was chosen for the study of bacteria.

The main finding was that the signature of S. aureus is dominated by the signature of adenosine and/or ATP shown in Figure 12 and the SERS features of E. coli are mainly due to the proteins from the pili/flagella or outer membrane and to the lipopolysaccharides, which contain adenine part in the lipid layer.

Figure 12:

Zoom of surface enhanced Raman scattering (SERS) spectra of S. aureus, ATP (in red) and adenosine (in green). Reprinted with permission from Andrei et al. [67]: Springer Nature, Copyright (2017).

In the case of S. aureus, a different response is observed as a function of the surface concentration of bacteria, giving rise to two distinct clusters in principal component analysis (PCA). This separation highlights the secretion of ATP at high concentrations (most likely generated outside the bacteria), whereas the poor stability of bacteria at low concentrations allows for detecting the cell wall signature and/or cell lysis. To conclude, these substrates, in spite of their simple, robust, fast and easy technique of production, meet the challenges for fast single-cell analysis of bacteria using SERS.

2.7.1 Sandwich SERS assay

Out of the successful applications of label-based SERS for bacterial diagnostics, the report by Kears et al. [77] (prof Goodacre group) is an excellent study of three bacterial pathogens (E. coli, Salmonella typhimurium, and methicillin-resistant S. aureus). These species were successfully isolated and detected, with the limit of detection at 10 CFU/mL Moreover, the group presented SERS spectra obtained from multiple pathogen detection.

Essentially, main approach was a sandwich SERS assay in which they combine SERS active Ag NPs functionalized with bacteria specific recognition molecules (antibodies) with lectin functionalized magnetic NPs enabling the capture and isolation of bacteria through magnetic separation (Figure 18). It is the use of lectin functionalized magnetic NPs which makes this separation step novel and efficient in capturing and isolating bacteria from a sample matrix.

Figure 18:

Schematic illustrating the single-plex and multiplex detection assay. Assay format: (A) Lectin (Con A) functionalized silver coated magnetic nanoparticle(Ag@MNPs) will bind to bacteria, and the presence of the magnet will allow for magnetic separation of the bacteria from the sample matrix. (B) Surface enhanced Raman scattering (SERS) active silver nanoparticles (AgNPs) functionalized with a biorecognition molecule (antibody; Ab) and a unique SERS reporter are added. The mixture is shaken for 30 min before application of a magnet for a further 30 min and collection of the sample. Any unbound matrix is gently removed and the sample subsequently resuspended in deionized water. (C) The sample is then interrogated with a 532 nm laser beam and a SERS signal obtained (green spectrum). When no target is present, the functionalized AgNPs will be washed away; thus, they will not bind to bacteria so a minimum SERS signal is obtained (red spectrum). (D) Multiplexing step: 3× AgNP conjugates each functionalized with a different Raman reporter and an antibody (which is specific for a bacterial pathogen) are added together with three bacterial pathogens and Con A (which binds to all three bacteria) functionalized Ag@MNPs. In the same way as the single-plex systems, magnetic separation allows for the samples to be concentrated and analyzed via a 532 nm laser. A SERS spectrum is obtained which contains characteristic peaks from the three Raman reporters and thus can be used to confirm the targets are present. Reprinted with permission from [77], Copyright (2017) American Chemical Society.

The total analysis time was ∼1 h. This is comparable with other bionanosensors which have been developed, however this approach is unprecedented because it can detect multiple pathogens simultaneously in this time.

Raman reporters used for multiple pathogen detection:

1. 7-dimethylamino-4-methylcoumarin-3-isothiocyanate (DACITC) – S. typhimurium, with a characteristic peak at 535 cm⁻¹

2. 4-(1H-pyrazol-4-yl)-pyridine (PPY) – MRSA, with a characteristic peak at 955 cm⁻¹,

3. Malachite green isothiocyanate (MGITC) – E. coli, with a characteristic peak at 1616 cm⁻¹

2.7.2 Pathogens in blood

In a subsequent study, Li et al. [78] reported on AST of pathogens in blood. This study is based on the magnetic separation of bacteria from the blood samples using polyethyleneimine-modified magnetic microspheres (Fe₃O₄@PEI) to capture and enrich the pathogens rapidly upon receiving the blood culture bottle. The first step was the creation of the spectral library of pathogens and their drug-resistant strains as reference spectra in order to identify pathogens separated from the blood samples. In the second step, the PEI-modified magnetic microspheres were used to separate the bacteria from clinical samples, and the magnetic microspheres/bacterial complexes were cultured overnight on normal blood and drug-resistant culture plates to form a single colony. In the third step, the target single colony was selected from the culture medium and mixed with the high performance SERS reinforcing material (AgNPs) and then added to the Si wafer. After the mixture was naturally air dried, SERS spectrum was detected, and the bacterial spectrum was compared with the standard spectrum measured in our first step to determine whether it was the corresponding pathogenic bacteria.

Using this methodology the drug resistance of S. aureus, A. baumannii, P. aeruginosa and their resistant strains (obtained from 77 common pathogenic bacteria in clinical blood infection) can be identified and monitored quickly. The successful implementation of this detection scheme could provide a new solution for the rapid detection of clinically important pathogenic bacteria and drug susceptibility test of pathogens which in turn could replace currently used methods.

The main driving force of another excellent research jointly performed by Munich group (Niessner and Haisch) by Yuan et al. [23] was to develop a SERS sandwich strategy for the sensitive detection and discrimination of E. coli, P. aeruginosa, and S. aureus directly in blood samples.

In this study, a new biosensor based on the sandwich structure employing antimicrobial peptides (AMPs) functionalized magnetic NPs as “capturing” probes for bacteria isolation and gold coated silver decorated graphene oxide (Au@Ag-GO) nanocomposites modified with 4-mercaptophenylboronic acid (4-MPBA) as SERS tags was used. Authors noted that AMPs as a capture element for the SERS detection of bacteria has not yet been reported before. In order to improve the stability of Au@Ag NPs the combination with graphene oxide (GO) was used which in turn makes the SERS active substrate more durable and beneficial for further chemical modification.

If bacteria is present in the blood (bacteremia or sepsis), the sandwich structure is created and SERS spectra can be consequently analyzed. Thus, in the “real-world” experiments the AMPs modified Fe₃O₄NPs@bacteria complex were magnetically separated from the blood, and mixed with 4-MPBA modified Au@Ag–GO nanocomposites, summarized in Figure 19.

Figure 19:

Schematic illustration of the operating procedures for bacterial detection via a surface enhanced Raman scattering (SERS) sandwich strategy, in which AMP modified magnetic Fe₃O₄NPs were utilized in the bacteria capture and 4-MPBA modified Au@Ag–GO nanocomposites were used as SERS tags. (A) AMP modified Fe₃O₄NPs were cultured with a bacterial sample matrix, which included bacteria, blood cells or other interference; (B) the Fe₃O₄NPs@bacteria complex was magnetically separated from the sample matrix; (C) blood cells or any other interference were removed; (D) 4-MPBA modified Au@Ag–GO nanocomposite SERS tags were cultured with the Fe₃O₄NPs@bacteria complex to form a sandwich structure; (E) the Fe₃O₄NPs/bacteria/SERS tags sandwich structure was magnetically separated and detected by the Raman spectrometer; (F) different kinds of bacteria were discriminated according to their Raman “fingerprints”; and (G) 4-MPBA can be used as an IS to correct the SERS intensities [23].

Using suggested methodology, E. coli, S. aureus, and P. aeruginosa were discriminated with limit of detection (LOD) of 10¹ CFU mL⁻¹, respectively. This novel method was further used in the detection of bacteria from clinical patients who were infected with bacteria and the results showed that 97.3% of the real blood samples (39 patients) can be classified effectively. Moreover, authors noted that the AMP modified Fe₃O₄NPs have good antibacterial activities, so that the instrument can be used for three tasks – (i) capture, (ii) discrimination, and (iii) inactivation of bacteria in the storage of blood.

2.7.3 Ultralow detection limits

The label-based approach offers very low detection limit (as low as 1 CFU/mL) however, methodology can be at some instances quite complex as it is presented by You et al. [79]. Here, authors employed gold nanoparticle-coated starch magnetic beads (AuNP@SMBs) that were prepared by in situ synthesis of AuNPs on the surface of starch magnetic beads (SMBs). In the next step the AuNP@SMBs were mixed with 4GS (linker protein) and incubated for 30 min. The specific affinity of the 4x-GBP (gold binding peptide) domain of the linker protein toward the gold surface was utilized for the spontaneous conjugation of 4GS with the AuNPs present on the surface of SMBs. The aggregation of immuno-AuNP@SMBs in the presence of target bacteria, followed by coupling of the captured target bacteria with SERS tags, provided an effective hotspot to amplify the Raman signals. The results suggest that the SERS-based detection system proposed in this study is highly sensitive and can detect target bacteria as low as 1 CFU/mL with excellent specificity.

Li et al. [80] developed a SERS sensor based on aptamers for the sensitive and simultaneous detection of different food pathogens – E. coli O157:H7 and S. typhimurium with high detection sensitivity (<8 CFU/mL). Presented biosensor benefits from the flexible combination of aptamers and Raman reporters in novel SERS tags with possible detection of other foodborne pathogens.

Jabbar et al. [81] proposed an innovative SERS platform for the detection of E. coli bacteria where plasmonics (Ag@PdNPs) have been formed on the gradient porosity silicon (GPSi) by simple immersion plating process. The main aim of this study was to enhance the activity of bimetallic SERS substrate for the detection of a single cell of pathogenic bacteria. The main idea behind this study is the superior structural characteristics of the GPSi in formation of (Ag@PdNPs) bimetallic NPs. The ultralow concentration of about 1 CFU/mL of E. coli was detected.

2.7.4 SERS in combination with different techniques

The research presented by Chisanga et al. [82] has been motivated by rapid detection of Campylobacter jejuni which is a major cause of foodborne gastroenteris worldwide. Samples were tested with a battery of techniques, including ultraviolet, Raman spectroscopy, SERS, and MALDI-TOF MS to provide complementary biomolecular information capable of differentiating bacterial classes. Raman spectroscopy provided whole-bacterium fingerprint, SERS detected structural moieties and bond linkages located specifically on the bacterial cell surface.

Moreover, researchers for the first time employed label-free SERS involving simple mixing and in situ synthesis of AgNPs (two different in situ methods for silver nanoparticle (AgNP) production) to probe molecular dynamics of pathogenic bacteria to complement data obtained from Raman and MALDI-TOF MS instruments. Chemometric treatment of data was employed to differentiate biochemical differences and phenotypic information from different locations in the cell–cell wall versus cytoplasm based on cell envelope or intracellular molecular dynamics. In this study, again, particularly for the three mutants highlighted more intense SERS bands at 730 cm⁻¹ (adenine moiety in flavin adenine dinucleotide [FAD]) located in the cell wall were observed. Authors concluded that Raman, SERS, and MALDI-TOF MS are powerful metabolic fingerprinting techniques capable of discriminating clinically relevant cell wall mutants of a single isogenic bacterial strain.

2.7.5 Inhibitory effect of antibiotic

Fu et al. [83] illustrated an elegant approach to rapid AST and MIC determination. For this, the researchers employed combination of aptamer, silver NPs (in situ synthesis), and bacteria to the final product – Bacteria-aptamer@AgNPs. The SERS spectra of E. coli O157:H7 with different concentration of tigecycline and S. aureus with different concentration of vancomycin have been recorded in 2 h. Here, 2 μL of sample (Bacteria-aptamer@AgNPs) suspension was dropped on a fluted glass slide and consequently analyzed.

It should be pointed out that the researchers found interesting behavior when monitoring the SERS spectra peak at 735 cm⁻¹ enhancement. It was observed that the bacteria CFU increased under sub MIC antibiotic concentration. The authors concluded that this is probably due to the fact that antibiotic at sub MIC concentrations stimulate the bacterial reproduction. There is an assumption that antibiotic at sub MIC has a short inhibitory effect on bacterial respiration, bacteria still have higher activity, and the number of bacteria is increasing. This is an important finding because this effect is not always taken into the account in the experiments of other groups using SERS for bacteria study.

2.7.6 Impressive approach inspired by nature

An impressive approach – enlightened by the physiology and predation of octopus – for detection of pathogenic bacteria in case of sepsis was presented by Lv et al. [84]. The captured bacteria (S. aureus) can be identified by the SERS method with a detection limit as low as 10 cells/mL⁻¹. In this work, a facile method based on a novel synthetic method to mimic the octopus-like architecture in order to capture bacterial pathogens was introduced – octopus-like multiarm structure (MAS). Consequently, MAS-ligand conjugates (MAS-Ls) were prepared so that octopus-like polymeric arms and ligands distributed along each arm behave as octopus’ suckers. Manufactured biomimetic arms have a diameter of about 30 nm and are easy to curl, and move around freely, so that target pathogens can be easily captured. The ligands modified on each arm provide multivalent binding sites for the bacterial cell. Thus, the octopus-like multivalent scaffold architecture has enormous potential for the competitive and effective capture of target bacteria. In the final step the MAS-Ls-binding bacteria (MAS-Ls-S. aureus) is characterized by SERS spectra, as illustrated in Figure 20. Here, the peak at 735 cm⁻¹ was employed and for very low concentration of S. aureus of 10 CFU/mL, could still be distinguished, demonstrating this system to be highly sensitive without the interference from other background pathogens present in the sample demonstrated in Figure 20.

Figure 20:

Scanning electron microscopy (SEM) images of (A) MAS-Ls-S. aureus and (B) the local enlarged images of the interaction between the arms and bacteria cell. Pseudocolored cell in image (a) shows the topographic interaction between the captured cell and MAS-Ls. (C) Schematic illustration of the specific capture of S. aureus by MAS-Ls. (D) Enrichment of 102 S. aureus cells by 0.5 mg MAS-Ls in varied volumes. (E) surface enhanced Raman scattering (SERS) detection of target bacteria after capture and separation: SERS spectrum of MAS-Ls binding with S. aureus, L. mono, E. coli, S. flexneri, and blank samples, respectively, with bacteria concentration of 107 CFU mL⁻¹. Reprinted with permission from [84], Copyright (2019) American Chemical Society.