key: cord-0737441-ua7o1k0r
authors: Sitjar, Jaya; Der-Liao, Jiunn; Lee, Han; Tsai, Huey-Pin; Wang, Jen-Ren; Liu, Ping-Yen
title: Challenges of SERS technology as a non-nucleic acid or -antigen detection method for SARS-CoV-2 virus and its variants
date: 2021-03-09
journal: Biosens Bioelectron
DOI: 10.1016/j.bios.2021.113153
sha: 076fe52992313f55219b074b978e00700520c657
doc_id: 737441
cord_uid: ua7o1k0r

The COVID-19 pandemic has caused a significant burden since December 2019 that has negatively impacted the global economy owing to the fact that the SARS-CoV-2 virus is fast-transmitting and highly contagious. Efforts have been taken to minimize the impact through strict screening measures in country borders in order to isolate potential virus carriers. Effective fast-screening methods are thus needed to identify infected individuals. The standard diagnostic methods for screening SARS-CoV-2 virus have always been to perform nucleic acid-based and serological tests. However, with each having drawbacks on producing false results at very early or later stage after symptoms onset, supplementary techniques are needed to back up these tests. Surface-enhanced Raman spectroscopy (SERS) as a detection technique has continuously advanced throughout the years in terms of sensitivity and capability to detect ultralow concentration of analytes ranging from single molecule to pathogens, to present as a highly potential alternative to known sensing methods. SERS technology as a candidate for an alternative and supplementary diagnostic method for the viral envelope of SARS-CoV-2 virus is presented, comparing its pros and cons to the standard methods and what other aspects it could offer that the other methods are not capable of. Factors that contribute to the detection effectivity of SERS is also discussed to show the advantages and limitations of this technique. Despite its promising capabilities, challenges like sources of SARS-CoV-2 virus and its variations, reliable SERS spectra, mass production of SERS-active substrates, and compliance to regulations for wide-scale testing scenario are highlighted.

symptom onset and declines at almost the same rate of its increase. 113 114 Figure 2 . General data of the infectivity on human, in %, is illustrated as the reference, plot (1). The start 115 of viral shedding and the start in decline of live virus are respectively pointed, plots (2) and (3). The 116 estimated effectivity of detection, in %, with respect to the timeline of infection (days to symptom onset) 117 through SERS with live virus (plot (4)), and with dead virus (plot (5)) are compared with that through 118 PCR by nasopharyngeal swab, plot (6) and sputum, plot (7). The stages of infection timeline 119 corresponding to the diagnostic tools -SERS technology, nucleic acid-based test, and serological tests are 120 also shown above based on their scope of applicability. 121

The nucleic acid-based method relies on viral genetic material. As long as a sufficient amount of virus 122 particle can be collected from the sample, it can be extensively tested from 5 days before the onset of 123 symptoms to 14 days after the onset of symptoms, as shown in Figure 2 , plot (1). As viral shedding starts 124 that may make it accurate for early diagnosis, even during asymptomatic periods (Afzal, 2020; He et al., 127 2020). Another advantage of RT-PCR is its versatility in sample types. It has been found that 128 nasopharyngeal swabs, sputum, stool and isolated viruses from the respiratory tract can all be used in this 129 diagnostic method (Sethuraman et al., 2020) . 130

The use of nucleic acid-based tests has its drawbacks -there is a need for sophisticated equipment 131 and costly reagents (i.e., primers, enzymes, buffers, and polymerases) that must be replenished. In nasopharyngeal swab sample, it was found that viral RNA could be detected within the first week of 135 symptoms, but the specific signals declined at around 6-7 days after the onset of symptoms, as shown in 136 In clinically suspected cases that are negative on nucleic acid-based tests, serological testing (i.e., 143

antigen-based immunoassays) can be used as a confirmatory test to supplement the results, however, it is 144 proven to be accurate only if the patients have developed immunity to the infected virus (Meng et al., 145 2020; Serrano et al., 2020) . With the start in the decline of live virus, seroconversion peaks around 7~14 146 days after the onset of symptom, as shown in Figure 2 , the effectiveness and sensitivity of nucleic acid-147 based tests subsequently decline, which makes serology-based diagnosis most suitable for testing within response. It has been reported that the simultaneous use of two antigens to detect IgM, IgA, and IgG will protein is the most conservative component in the virus classification, they are also prone to provide false 159 positive results to detect other coronaviruses other than the SARS-CoV-2 virus, i.e., the antigen used in 160 ELISA is most likely to react with antibodies of other coronaviruses (Wechselberger et al., 2020; Younes 161 et al., 2020) . Since this technology is only suitable for the late stage of the disease, the test results should 162

not be used to screen for asymptomatic suspicious individuals, so it is not a suitable tool to control the 163 spread of the virus; on the other hand, it is more suitable for confirming infected patients that show 164 COVID-19 symptoms. Moreover, the antibody development of each infected patient is different. The 7 to 165 14 days after the occurrence of seroconversion is only the average of the collected data. Certain 166 conditions may be contrary to this, leading to invalid tests and possibly providing false results (Stowell 167 and Guarner, 2020). 168

In summary, the currently applied detection techniques are supplementary to each other to some 169 extent, that is, nucleic acid-based detection can be applied in the early infection stage. As the patient 170 develops immunity, the detection will eventually lead to a decline in effectiveness, thereafter, serological 171 tests can be used effectively. Both tests are subject to timeframe restrictions, an appropriate method can 172 be employed according to the specific conditions and stages of the disease. Nevertheless, having a 173 diagnostic test that can cover the entire timeframe of the disease would not only eliminate the need to 174 switch between timeframe-restricted tests, but also minimize the uncertainties from using diagnostic tests 175 that may produce false positive results (Giri et al., 2020) . 176

Aside from the aforementioned SARS-CoV-2 diagnostic tests, some studies have also attempted to 178 explore alternative detection technologies. For examples, the use of field-effect transistor (FET), which 179 uses graphene sheets conjugated with SARS-CoV-2 antibody to coat the transistor; then a nasopharyngeal 180 swab sample is placed on the sensing area of the transistor, and the generated electricity signals are used 181 to match and confirm the presence of viruses (Seo et al., 2020) . By using antibodies to detect viruses, the 182 technique combines electrical induction with immunoassays, which means that this method is only 183 suitable for patients who have already been found to have antibodies, making the detection tool suitable 184

for later-stage infections. 185

In another study, electrical signals were also used monitor the presence of SARS-CoV-2 virus 186 wherein exhaled breath was used as the sample. The detection platform has multiple sensors composed of unwanted particles may interfere with analyte collection in addition to the fact that there could be 191 instances wherein the virus particles in VOCs from exhaled breath may not be at a detectable amount. spherical structure, and is predominantly composed of spike proteins (Sivashanmugan et al., 2013) . In the 282 said study, peaks with highest intensities corresponded to vibrational modes of adenine and tyrosine, 283 which could be attributed to the tyrosine-histidine interactions that are responsible for the formation of the 284 neuraminidase amino acid sequences on the viral envelope. These studies demonstrate that SERS 285 technique is capable of detecting pathogens regardless of size -S. aureus being around 500~1000 nm, 286 their membrane surface, each Raman-active component is expected to have at least a characteristic peak 288 in the SERS spectrum, which may correspond to a particular vibrational mode such as what is shown in 289 Table 2 with a variety of enveloped viruses. 290

SERS technique is gaining more attention for its potential in analyte detection, even at low 292 concentrations, and cases wherein this technology is applicable would require systems that are not only 293 accurate and sensitive but also selective, to be able to detect specific analytes out of a complex specimen 294 

Substrate design is crucial in SERS detection systems as parameters involved predominantly dictate analyte molecule could also lead to a change in the SERS spectra since vibrational modes would then 315 consequently change, which results to the appearance, disappearance, and/or shift in Raman peaks process, it is also important to consider the suitability of the substrate material in terms of its stability in 329 an environment or condition where it is to be applied. Each plasmonic metal exhibit their own 330 characteristic properties that are not found in other metals; for example, gold is a biocompatible material 331 but provides a lower signal enhancement than silver, which on the other hand is less stable due to it being 332 prone to oxidation (Liu et al., 2020). Studies have thus explored the use of both of these metals to take 333 advantage of combining these properties, giving synergistic effects. The interaction of analyte with the 334 substrate material also contributes to the total enhancement brought by the substrate, as illustrated on 335 In Figure 3 

the distribution of viruses on the nanostructures. In the illustration, nanocavities serve as the SERS-active 339 substrate, and a virus particle that is bigger than the cavity itself, case (1), would not be able to maximize 340 the effects brought by the cavity as it somehow covers the structure whereas in cases (2) high laser power could be an advantage in some cases such as in the example in Figure 3 (b)-ii, as high 353 power could induce more plasmonic resonance to occur on the substrate. However, high power could also 354 degrade the analyte and thus, weak to no SERS or fake signals are obtained. It is therefore important to 355 also determine the best laser power setting to maximize the SERS effects brought by the laser conditions. the classification of the virus in terms of structure; this way, when the resulting SERS spectra are 358 analyzed for interpretation, it would be convenient to assign the peaks based on the composition of the 359 virus. Figure 3 As much as each of the abovementioned factors impart individual contributions to the total SERS 372 effects, their synergistic contributions should not be disregarded. As an example, a gold nanostructure 373 might perform the best with a virus analyte only if this factor is considered but when combined with a 633 374 nm laser that performed the best in a similar nanostructure that is made of silver, it is possible that the 375 SERS effects in the case of the gold nanostructure and 633 nm laser might be weaker than that when 376 silver is used. Therefore, for the SERS effects to be maximized, experiments should be meticulously 377 designed with careful considerations of the factors mentioned. 378

Furthermore, as illustrated in Figure 3 (d)-i, analytes experience different types of adsorption on the 380 substrate -CM effect which, as previously mentioned is due to the affinities between the analyte and 381 substrate leading to possible charge transfers between these two systems; EM effect which is due to the 382 formation of electromagnetic fields from the localized surface plasmon resonance produced by the 383 plasmonic nanostructures upon exposure to incident radiation; and the van-der Waals-induced effect 384 which is a relatively weak physical adsorption attributed to the electrical interactions between the analyte 385 and substrate when in close proximity to each other. 386 interaction between the analyte molecule and the metal substrate, as there occurs a charge transfer 390 between these components (Prakash, 2020; Trivedi et al., 2020) . Chemisorption of the analyte causes a 391 change in the electronic states of the molecules, causing the absorbance of the analyte to shift and when it 392 is exposed to the incident radiation, its Raman cross section increases, leading to signal enhancement. On 393 the other hand, EM mechanism relies on the localized surface plasmon resonance (LSPR) and the local 394 (graphene, graphene oxide, MoS 2 ) were also found to provide SERS effects (Wolosiuk et al., 2014) . 412

Substrates in colloidal form were found to agglomerate uncontrollably, causing sites of enhancement to 413 be distributed unevenly, resulting to inconsistent measurements throughout the substrate. Immobilized 414 substrates fabricated through sophisticated methods were then developed to control nanostructure 415 distribution in a uniform arrangement to address the problem of low reproducibility. Moreover, as much 416 as the geometry of the nanostructures plays a vital role in the signal enhancement capability of the 417 substrate, it is also important to consider that the gaps between these nanostructures are where hot spots 418 

What has been discussed so far only deals with label-free detection wherein the analyte of interest is 433 directly detected with Raman spectroscopy through the SERS spectra of the analyte itself from its 434 intrinsic molecular properties such as functional groups, molecular weight, charge, and its overall In the detection of complex macromolecules of biological nature such as viruses, bacteria, and protein 444 biomarkers, SERS spectra obtained through label-free detection are most often complicated and peaks 445 coming from the analytes are hard to isolate. However, there are some exceptions to this case as some 446

viruses were reported to be easily detectable even through label-free detection, and this could possibly be 447 attributed to the structure and composition of the virus. Influenza strains were successfully detected on 448

Au/Ag nanorods without the use of any Raman reporters. In the study, peaks found in the resulting SERS 449 spectra were associated with the spike proteins found on the viral envelope -hemagglutinin and 450

neuraminidase. In another case wherein the analyte was the Human Enterovirus 71 (EV71), collection of 451 SERS signals was not possible due to the incompatibility of the analyte with the substrate. It should be 452 considered that EV71 lacks the envelope and spike proteins, unlike the influenza strains as it is classified 453 as a picornavirus, a nonenveloped virus, which is only characterized by a capsid on its surface. In cases, consider using SERS labels to obtain defined signals associated with the virus (Mauriz, 2020) . alternative diagnostic methods such as SERS technology would prove to be significant. 463

Experimental studies performed to detect SARS-CoV-2 virus through SERS technology would Moreover, in actual applications, there are several sources of specimen to be used for testing -it has 481 been reported that samples could be obtained from nasopharyngeal swabs, saliva, and throat swabs. Viral 482 loading could differ from samples taken at different sources from a single subject and this could pose 483 false results in such way that the virus could be present in one location but could be absent in another. 484

Another issue that needs to be addressed is that in actual cases, swab samples does not contain only the some degree could cause a difficulty in the interpretation of the SERS spectra due to the interferences 487 brought by these unwanted chemicals. Thus, it is desirable to have a SERS system that is selective or one 488 that could distinguish between various other viruses, making multiplex detection of viruses possible 489 

Most of the SERS studies done on a laboratory scale have shown potential to be produced in a larger 492 scale as intended for their applications in the field of chemical sensing. With the recent event of the 493 pandemic that calls for urgent solutions, quick but reliable mass testing is much needed especially in 494 airports and large-scale events wherein the transmission of virus is most likely to occur. This 495 consequently calls for testing supplies that could cater to a large number of people and to meet this 496 demand, a mass production of the materials should be done. 497 SERS-active substrates produced for experimental studies could sometimes follow a complicated 498 fabrication route that requires not only costly materials and sophisticated equipment but could also 499 undergo processes that takes time to proceed, which defeats the purpose of mass production that should 500 be quick but cost-effective. Thus, a process flow in a laboratory scale that could easily be transferred for 501 mass production is highly desirable. With mass testing as the intended objective, SERS substrates should 502 be designed with a device or a platform that could easily process the sample from the test subject. 503

A recent study showed that at least 35% of people are asymptomatic, revealing an increased risk of 505 rapid community spread and the need for widespread testing ("COVID-19 Pandemic Planning Scenarios," 506 2020). With the quick spread of COVID-19, the FDA has begun to issue Emergency Use Authorizations 507 (EUA) to several diagnostic tests for COVID-19. In emergency cases such as the COVID-19 pandemic 508 wherein urgent strategies are needed, medical devices could be utilized to address these issues through a 509 short-term EUA or a regular application of in-vitro diagnostic (IVD) device ("FDA MOU 225-14-017," 510 2020). The challenge posed in both cases is the requirement of performing a minimum number of 511 screenings using cultured or actual samples from test subjects; however, the regulation depends upon 512 countries. And so, for SERS technology to be applied as a potential diagnostic tool in the case of a 513 pandemic, the previously mentioned challenges -source of the virus and mass production of the 514 substrates should be addressed to be able to comply with the regulations that would prepare and approve 515 the use of the device. 516 J o u r n a l P r e -p r o o f as assist the nation in implementing disease prevention tactics and health policies. The FDA primarily 518 serves as a regulatory agency for medical devices, and acquiring the EUA for diagnostic tools 519 ("Emergency Use Authorizations for Medical Devices," 2020) to give emergency approval to COVID-19 520 diagnostic tests allows protocols for a wide range of activities related to the monitoring, diagnostics and 521 treatment of COVID-19 to be implemented urgently. Similar to the EUA approval procedure by the FDA, 522 the Emergency Use Listing (EUL) by the WHO validates in vitro diagnostics (IVD) used for the detection 523 of COVID-19, focusing on IVDs most likely to be used in countries with limited resources for testing. 524

While the EUA is meant to provide accelerated approval for all IVDs that meet requirements, the EUL 525 prioritizes simpler products to support countries most in need. ("COVID-19 Puts the WHO's EUL) to 526

Work," 2020). 527

The current diagnostic tools for SARS-CoV-2 virus and its variants have always been the convention 529 but with the advancement of method, more techniques are being developed to address the limitations 530 brought by these standard methods, such as their applicability to only limited timeframes. In this article, 531

we introduce the development of SERS technology, which makes it a complementary choice suitable for 532 conventional methods. It should be emphasized that in order to produce a SERS-active substrate 533 specifically for SARS-CoV-2 virus detection, its substrate design is a critical factor, because not any 534 SERS substrate has the ability to detect viruses. With the use of a corresponding SARS-CoV-2 535 pseudovirus, there is no need to perform SERS experiments in the BSL-3 laboratory; although 536 convenient, this is only to test the effectiveness of SERS-active substrates in detecting viruses. Identifying 537 positive and negative cases through the SERS system will be able to meet EUA regulations and 538 subsequent IVD applications. It is foreseeable that with the further development of high-throughput 539 manufacturing technology and modular design, SERS-active substrates dedicated to virus detection can 540 be produced on a large scale to meet the high demand during the outbreak. 

COVID-19 Pandemic Planning Scenarios

COVID-19 Puts the World Health Organization's Emergency Use Listing (EUL) to Work

Emergency Use Authorizations for Medical Devices

What Kind of Covid Test Should I Get? Answers on Cost

Diagnostics for SARS-CoV-2 detection: A 619 comprehensive review of the FDA-EUA COVID-19 testing landscape

Diagnostic techniques for COVID-19 and new 640 developments

Performance evaluation of serological assays to 668 determine the immunoglobulin status in SARS-CoV-2 infected patients