The benefits of saliva sampling for frequent and massive COVID-19 testing

Saliva tests for detection of SARS-CoV-2 RNA and antigens are becoming widely available lately. What are the advantages and disadvantages of sampling saliva over the nasal swabs? Saliva sampling simply involves spitting into a collection container. Recently, the U.S. Food and Drug Administration (FDA) also authorized the first diagnostic test with the option of home-collected saliva samples by using the Spectrum Solutions LLC SDNA-1000 Saliva Collection Device. The collected samples are then sent to a lab for further processing and analysis. Therefore, saliva sampling is much simpler and less uncomfortable than nasal swab sampling. This makes taking a sample at home or point of care much easier and more practical. It does not require collection by trained and protected medical personnel wearing personal protective equipment, thus reducing a considerable risk to healthcare workers. In contrast to nasal swab sampling, this approach is not affected by global shortages of swabs and personal protective equipment. In general, saliva sampling should permit more widespread and frequent testing. It is, of course, important that the test be as reliable and sensitive as the nasal swab test, but these appear to be reasonably similar [(1-3).

The current gold standard for COVID-19 diagnosis is real-time reverse transcription polymerase chain reaction (RT-PCR) detection of SARS-CoV-2 from collected samples. Concomitant with the advent of saliva sampling, techniques to simplify the detection of viral RNA have been introduced by eliminating the need for specific equipment, thermal cyclers. This makes it far more adaptable in resource poor settings, which often don’t have the relatively expensive PCR thermal cyclers. One such technique is reverse transcription loop-mediated isothermal amplification (RT-LAMP), which has been previously used to detect other viruses, including Zika and Ebola. A typical RT-LAMP assay takes place at a constant 63°C and the presence of viral RNA generates a color change in as little as a half an hour. A recent modification of the technique that uses inhibitors of salivary RNAs has been claimed to detect a single copy of viral RNA. These methods are obviously not specific to saliva tests, but use of saliva samples should facilitate a mass testing of SARS-CoV-2.

Several new developments in testing for viral RNA combine LAMP isothermal amplification with a technique called lateral flow, in which amplified samples are applied to a strip and allowed to flow along the strip. Amplified viral cDNA is detected by application of a CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats)-CAS12 complex designed to bind to a specific viral sequence. When the CRISPR-CAS12 complex finds and binds its target, it releases a chromophore, which is visualized directly or fluorescently as a specific band on the strip. The CRISPR-CAS12 DETECTR is claimed to have a sensitivity of 95% and specificity of 100% compared to the CDC RT-PCR test (4) despite a concern of false negative results. In addition, STOP (SHERLOCK Testing in One Pot) uses similar technology. STOP has a detection limit of 100 viral RNA copies. Results from both tests are obtained in an hour or less, and analyses of the strips are obviously simple. Positive results are indicated by an exhibition of specific band on the strip.

Of course, all these tests, even though they can use point of care collection, require analyses in laboratories. A true point of care test would be one that could be used without need for laboratory involvement, similar to home pregnancy tests. Several of these kinds of tests are in development. For example, a test would involve placing a drop of saliva onto a device, the size of a quarter, and plugging it in to a smart phone. DNA aptamers on the device bind to viral proteins and then are detected by voltage generated by room temperature electron tunneling. Another potential test uses a microfluidic chip in a cartridge and isothermal amplification. Results can be read and uploaded to a smart phone.

The actual impact of these new technologies still need to be ascertained, yet, they provide a snapshot of innovative testing, not only for SARS-CoV-2, but  for other pathogens. Rapid point of care tests could ultimately be used for screening a large group of people (i.e., airline passengers, concert attendees) especially if they can be linked to smart phones. This could be an important interventional strategy in preventing transmission of the virus and in preparing for future pandemics.

 

1. L. Azzi et al., Saliva is a reliable tool to detect SARS-CoV-2. J Infect 81, e45-e50 (2020).
2. M. Baghizadeh Fini, Oral saliva and COVID-19. Oral Oncol 108, 104821 (2020).
3. K. K. To et al., Consistent detection of 2019 novel coronavirus in saliva. Clin Infect Dis, (2020).
4. J. P. Broughton et al., CRISPR-Cas12-based detection of SARS-CoV-2. Nat Biotechnol 38, 870-874 (2020).

 

View All GVN SARS-CoV-2 Perspectives

Can a single amino acid mutation in the spike protein affect the infectivity and immunogenicity of SARS-CoV-2?

Recently, a great deal of attention has become focused on a specific SARS-CoV2 mutant in which amino acid residue 614 of the spike protein is changed from aspartic acid to the less bulky and more neutral glycine. This mutant, D614G, reported by Korber et al., has become increasingly predominant after first seemingly proliferating in Europe, then spreading rapidly elsewhere. Korber et al. suggested the mutant is more infectious, based on higher viral RNA titers from patients who were infected with it and its rapid prevalence, but it does not appear to be more pathogenic, based upon the clinical pictures of the patients. Residue 614 is not in the receptor binding domain. They suggested that the mechanism of enhanced infectivity could be due to glycine not forming a hydrogen bond with the neighboring spike protein subunit, allowing the subunits to dissociate more readily and thereby facilitating virion fusion with the cell membrane.  There is a stunning virtual reality visualization of this on YouTube.  Korber et al. also reported evidence for recombination between genomes carrying both this mutation as well as other mutations, indicating recombination and simultaneous infection of cells with more than one genotype. Although suggestive, no proof was presented for an actual increase in infectivity by the mutant.

Since that report, several other studies have addressed the issue of infectivity more directly. One caveat related to many of these reports is that they use lentiviral particles pseudo typed with coronavirus envelope proteins. Viral entry is measured by a co-transduced indicator gene, such as luciferase. Although this is thought to faithfully mimic coronaviral entry, it is an inherently artificial system. The general consensus of all these studies is that the G614 variant enters cells expressing the ACE2 receptor better than the D614 variant even though the variable residue is not in the receptor binding domain, that there is no difference in clinical outcome, that infection with the G614 variant results in higher viral RNA titers in nasal swabs, and that there is not a great deal of difference in antibody neutralization (which is good news for vaccine development).

Let’s consider the individual reports separately. Ozono et al. used the lentiviral pseudovirus method to sample five different naturally occurring mutations in the spike protein, including D614G, to characterize their behavior relative to the reference genotype. Their entry characteristics varied from having a lesser to a greater ability to enter cells expressing ACE2 and the protease TMPRSS2 cells, which greatly facilitates SARS-CoV2 entry. Significantly, the D614G mutant showed the most efficient entry. Interestingly, SARS-CoV1 was much more efficient at entry than was SARS-CoV2. They performed an in silico structural analysis that suggested that the SARS-CoV1 spike trimer has a more open configuration that would result in greater accessibility to the ACE2 receptor by the receptor binding domain. They also tested COVID 19 antisera from patients infected with the D614 variant, and showed no detectable differences in neutralization between the D614 and G614 variants.

Hu et al. also used a lentiviral pseudovirus system to analyze the D614G variants. As with Korber et al., they found the G614 variant to be globally distributed. Like Ozono et al., they found about a 2.5-fold increase in entry efficiency by the G614 variant, perhaps due to more efficient protease cleavage of its spike protein. Unlike Ozono et al., they found that a minority of COVID 18 antisera failed to neutralize the G614 variant to the same extent as the D614 variant. However, it is not clear with which variants the serum donors were infected.

Wagner et al. used a more natural but messier approach. They looked at viral loads, as measured by RT-PCR, and clinical status of patients in Washington state infected with either the G614 or D614 variants. They found that patients infected with the G614 variant had higher nasal viral RNA loads, but did not have a more severe clinical picture. The age of the patients infected with G614 skewed slightly younger (~3 years). They also found that G614 became increasingly more prevalent in Washington state over time.

Lorenzo-Redondo et al. (1) reported on patients in Chicago infected with one of what they called three clades of SARS-CoV2. Clade 1, which was introduced from Washington, contains the G614 phenotype, while clades 2 and 3 have the D614 phenotype. The origin of clade 2 was ascribed to Illinois, while clade 3 was introduced from New York. Clade 1 had higher viral loads than clade 2, in agreement with Wagner et al. Interestingly, when bronchial alveolar lavage samples were tested, there was little difference in viral RNA titers between the two clades, suggesting that the increased titers were specific to upper airway tissue. This could be a factor increased spread.

It should be pointed out that all the above results are in the form of preprints. In addition, the methods used to measure entry directly are somewhat artificial. However, taken together directly from patient data, it seems that G614 may in fact be more capable of spreading, perhaps because of more facile entry into cells, perhaps due to better proteolytic processing because of a more open quarternary structure. Fortunately, this mutation does not seem to worsen the clinical outcome of infection, nor does it seem to abrogate recognition by most neutralizing antibodies.

 

References

 

1. R. Lorenzo-Redondo et al., A Unique Clade of SARS-CoV-2 Viruses is Associated with Lower Viral Loads in Patient Upper Airways. medRxiv, (2020).
2. Tang, Leyan & Schulkins, Allison & Chen, Chun-Nan & Deshayes, Kurt & Kenney, John. (2020). The SARS-CoV-2 Spike Protein D614G Mutation Shows Increasing Dominance and May Confer a Structural Advantage to the Furin Cleavage Domain. 10.20944/preprints202005.0407.v1.
3. Grubaugh, N.D., Hanage, W.P., Rasmussen, A.L., Making sense of mutation: what D614G means for the COVID-19 pandemic remains unclear, Cell (2020), doi: https:// doi.org/10.1016/j.cell.2020.06.040.

View All GVN SARS-CoV-2 Perspectives

The question of where viruses involved in spillover zoonoses originate is always important and fascinating. Identifying their origin is critical in understanding how zoonotic epidemics originate, and in preparing preventive strategies for future epidemic. Fascinating, because identification, like solving a difficult whodunit, involves a series of forensics including those from epidemiology, molecular genetics, phylogenetics, and the study of reservoir and intermediate hosts (i.e., wild and domestic animals). Finding the immediate animal precursor virus is normally difficult; indeed, coronaviruses compound the difficulty because they have a predilection for recombination (1), constantly creating novel viruses, and making it necessary to identify more than one ancestral virus; unless, one has the rare good fortune of finding the exact parental recombinant. Of course, after the ancestral virus(es) has been identified, there remains the question of where and when did it first get into the human leading to the secondary transmission.

In this time of inexpensive and rapid genomic sequencing, the entire genome of the virus causing COVID-19 was rapidly obtained and published, identified as a beta coronavirus (2), and designated as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). It was distantly related to SARS-CoV-1, basically refuting the idea that it had simply mutated from SARS-CoV1. A much more closely related virus (>96% over its entire genome) was sequenced from horseshoe bats living in a cave in Yunnan province, called CoV RaTG13b (3). Large stretches of the CoV RaTG13 were virtually identical to SARS-CoV-2; although, other stretches appeared to be sufficiently distinct as to rule out a direct ancestral relationship. Some attention has also been focused on pangolins, an endangered species that also carry coronaviruses, and that is imported illegally into China in great numbers for food and traditional medicine. One coronavirus isolated from a Malayan pangolin confiscated in Guangdong province was >90% identical to that of SARS-CoV-2 over its whole genome, but was clearly not a direct precursor (4).

Recently a phylogenetic analysis has provided a reasonably convincing explanation for how SARS-CoV-2 was originally formed. The study shows that the virus is likely a triple recombinant, containing genetic elements of two bat coronaviruses and a pangolin coronavirus. Likely recombinational break points are before and after the angiotensin I Converting Enzyme 2 (ACE2) receptor binding motif in the spike protein of SARS-CoV-2. The receptor binding motif of SARS-CoV-2 is virtually identical to that of an isolate from pangolins. The upstream portion of the ORF1A region appears to have been acquired from a second bat coronavirus. Interestingly, the authors present data that SARS-CoV-1 is also the product of multiple recombination events involving different bat coronaviruses. Another study has analyzed in-depth evolutionary history of bat coronaviruses, and their ability to jump species to better understand how and where zoonoses are most likely to occur. Importantly, the SARS-CoV-2 spike protein displays a unique feature, namely the presence of a furin cleavage site insertion (PRRA) at the junction of two subunits of the S protein. Neither the bat beta coronaviruses, nor the pangolin beta coronaviruses sampled thus far have polybasic cleavage sites.  No animal coronavirus has been identified that is sufficiently similar to have served as the direct progenitor of SARS-CoV-2; although, the diversity of coronaviruses in bats and other species is massively under sampled (6).

All of the above has bearing on the evolution of virus. No clear answer for the origin of the virus has been concluded prior to human infection. At first, it seemed most likely to have emerged from the wet market in Hunan, but follow-up analyses of the market and the animals present have not yielded proof. This quickly gave rise to conspiracy theories that the virus was generated in a virology lab in Wuhan.  The possibility that the virus could be man-made was fairly convincingly refuted by Anderson et al., (6) and is not supported by academic data. What then was the direct source of the virus, and when did the virus first enter humans?

Given the prevalence of recombination among bat coronavirus and the myriad of possibilities for the introduction of SARS-CoV-2, it is probable that sample collection from diverse hosts and genomic sequencing analysis need to be continue to identify the origin of SARS-CoV-2. However, global spread of these viruses and their potential for cross-species transmission makes it clear that it is possible to have waves of COVID-19 outbreaks. The important thing is how well we be prepared for the next one.

1. B. Hu et al., Discovery of a rich gene pool of bat SARS-related coronaviruses provides new insights into the origin of SARS coronavirus. PLoS Pathog 13, e1006698 (2017).
2. F. Wu et al., A new coronavirus associated with human respiratory disease in China. Nature 579, 265-269 (2020).
3. P. Zhou et al., A pneumonia outbreak associated with a new coronavirus of probable bat origin. Nature 579, 270-273 (2020).
4. T. Zhang, Q. Wu, Z. Zhang, Probable Pangolin Origin of SARS-CoV-2 Associated with the COVID-19 Outbreak. Curr Biol 30, 1346-1351 e1342 (2020).
5. V. D. Menachery et al., A SARS-like cluster of circulating bat coronaviruses shows potential for human emergence. Nat Med 21, 1508-1513 (2015).
6. K. G. Andersen, A. Rambaut, W. I. Lipkin, E. C. Holmes, R. F. Garry, The proximal origin of SARS-CoV-2. Nat Med 26, 450-452 (2020).
7. N. Wang et al., Serological Evidence of Bat SARS-Related Coronavirus Infection in Humans, China. Virol Sin 33, 104-107 (2018).

 

View All GVN SARS-CoV-2 Perspectives

There has been some very interesting news in the last several days that relate to vaccine development. The first of these is a report of results on the Phase 1 trials of the Moderna mRNA vaccine.  Most importantly, the vaccine appeared to be safe; no major adverse events were reported. Vaccines developed binding antibodies to the spike protein of SARS-CoV-2 at levels seen in immunized people, and of those eight who were tested, all had neutralizing antibodies to the virus. This is hopeful; neutralizing antibodies are generally a critical part of a successful immune response. However, researchers measured virus-recognizing antibodies in 25 participants, and detected levels similar to or higher than those found in the blood of people who have recovered from COVID-19. In general, most people who have recovered from COVID-19 without hospitalization do not produce high levels of neutralizing antibodies. Therefore, it is uncertain whether the responses are enough to protect people from infection. In addition, the data of this study has not been published yet. The report also begs the question, what about T cell responses? These will likely be important for long term immunity. he Oxford vaccine, an adenoviral vectored vaccine expressing the spike protein of SARS-CoV-2, wsa also addressed in the previous vaccine post.  In rhesus monkey challenge, a significantly reduced viral load in bronchoalveolar lavage fluid and respiratory tract tissue of vaccinated animals challenged with SARS-CoV-2 compared with control animals. Importantly, vaccinated rhesus macaques were protected from pneumonia without evidence of immune-enhanced disease. However, the scientists found similar levels of viral titers in animals’ noses between unvaccinated and vaccinated groups, probably due to a high dose of challenge virus used in this study.  This has raised a concern whether the vaccination can provide the protection against transmission. The outcome of ongoing clinical trial will provide a better answer whether the vaccine can protect humans from COVID-19.  Eli Lilly is proceeding with a passive immunization trial using monoclonal neutralizing antibodies to SASR-CoV2. It will be interesting to see the results.

Two recent reports (in Cell and MedRxiv) represent an important beginning to answering this question (or questions). In addition to B cells in producing antibodies, T cell can also play a role in preventing infection. Helper T cells activate B cells to secrete antibodies and macrophages to destroy ingested microbes. Cytotoxic T cells kill infected target cells. These two studies  showed, using peptide pools covering predicted epitopes derived from the viral genome, that most convalescent patients had CD4+ and CD8+ responses to a wide variety of viral proteins, including the spike protein. Other common responses include those against the matrix and nucleoprotein as well as nonstructural proteins, such as nsp3 and 4 and ORFs 3a and 8. This suggests that vaccine inducing an immune response to not only the S protein but also these proteins could enhance its protective efficacy. Interestingly, these reports also showed that 40-60% of people who had never been exposed to SARS-CoV-2 also had T cell responses to the virus, and the authors speculated that this could be due to T cell cross-reactivity with the common seasonal cold coronaviruses. This raises some interesting questions. Could previous exposure to cold coronaviruses afford some protection against SARS-CoV2? Do these activities differ in people with severe disease? Is it predictive; that is, could it be a factor in whether infection results in mild or severe disease? How might the presence or absence of these T cell reactivities affect vaccine responses? In any event, it is likely that elicitation of T cell responses will need to be part of a successful vaccine response.

The flip side of T cell responses is that they could be harmful. A cytokine storm has been invoked by many to account for aspects of the end stage disease. If an overactive immune reaction is important, it seems intuitive it should affect younger people disproportionately, as happened in the Spanish flu pandemic. On the other hand, perhaps people with weaker immune systems don’t mount an effective response until late in the disease and generate an inappropriate inflammatory response as a result. A recent report investigated this possibility. The authors looked at moderate, severe and critical patients. They reported a depletion of CD4+ and CD8+ T cells with a CD11a++ memory/effector phenotype, and survival was correlated with recovery of these cells. Conversely, more severe disease was correlated with an increase in S protein-reactive cells producing inflammatory cytokines. Unsurprisingly, with SARS-CoV-2 the immune system appears to be a two-edged sword.

 

View All GVN SARS-CoV-2 Perspectives

Just to cut to the chase, we just do not know; but we may hope for one year and a half to two years, maybe less. And that will be great, but we have to distinguish between getting early evidence of efficacy and getting a vaccine ready for mass vaccination. But let’s consider background material that suggests it is likely we will indeed have a vaccine.

First, it is helpful to consider different mechanisms of immunity. There are two general types of immunity, innate and adaptive. The innate system is quick and dirty and recognizes general patterns widely shared by pathogens, such as double stranded RNA or bacterial lipopolysaccharides, but not normally present. It is mediated through the activity of cytokines and involves inflammation. The other mechanism, the adaptive immune system, is slower to respond but more precise and is mediated through antibodies and cytotoxic T cells.

Second, there are several kinds of immunization. Transfusion of immune serum or of purified antibodies, called passive immunization, can protect temporarily against infection and disease. This is currently being tried in patients.  Another type of immunization may induce a general innate immune response. Administration of some vaccines appear to induce a temporary immunity against viruses other than their targets, probably through stimulation of innate immunity. In theory, they should be able to protect against a relatively broad array of targets, which might include SARS-CoV2. This is likely the basis for the activities of BCG, which is used as immunotherapy against bladder cancer, for example. Phase 3 trials are currently enrolling subjects in Australia (BRACE) and the Netherlands (BCG-CORONA). More recently, Dr. Konstantin Chumakov (Food and Drug Administration, Office of Vaccines Research and Review) and Dr. Robert Gallo (Institute of Human Virology, University of Maryland School of Medicine, Baltimore) of the Global Virus Network have suggested repurposing polio vaccine for use against SARS-CoV2 through its induction of innate immunity. Finally, targeted vaccines can stimulate both innate and adaptive immunity. Targeted vaccines are the most commonly used vaccines, as they are the most specific and can provide durable protection.

Is a targeted vaccine for SARS-CoV2 possible? Several lines of evidence suggest the answer is yes. Worry number one is whether immune responses to the virus will be protective. Sera from infected rhesus monkeys and infected people can neutralize the virus, indicating it is possible to generate an effective response given an appropriate stimulus.  A second concern is that the virus will mutate rapidly and escape the immunity induced by vaccines. However, available sequence data suggest the overall genetic variability might be low. The virus has an RNA proofreading mechanism that reduces its mutation rate. A hopeful finding is that immune sera from people infected with the SARS1 virus (which is only 79% related at the whole genome level) neutralizes the SARS2 virus, suggesting elicited immunity will be broadly reactive. Similarly, a monoclonal antibody has been reported that neutralizes both SARS-CoV1 and SARS-CoV2.

There are many antibody-based passive immunization approaches with human and animal serum and recovered antibodies in pre-clinical development. There are also currently more than 200 targeted vaccines in widely various stages of development, including at least seven phase 1 or 2 clinical trials, and there are currently at least eight different approaches to developing a vaccine. One approach is simply to inject one or more of the viral proteins. Two other approaches involve injection and cellular take-up of either RNA or DNA encoding one or more viral proteins. Two other approaches involve vaccination with a heterologous viral vector derived from adenovirus, lentivirus, measles, or influenza virus, which has been engineered to express SARS-CoV2 proteins. These viral vectors can be either replication incompetent or able to replicate to various extents. Another type of vaccine vehicle is a virus-like particle (VLP), made artificially from viral proteins but lacking a viral genome. The SARS-CoV2 protein of interest is attached to the outer surface of the VLP. The other two vaccine approaches use whole SARS-CoV2 virus, either in killed form or attenuated so that its ability to replicate is severely compromised.

Obviously, testing for safety and efficacy is critical. Animal testing provides a way to obtain preliminary safety and efficacy data from candidate vaccines relatively quickly, because animals can be challenged with live virus. Drawbacks are that they may not accurately mimic the pathogenesis in humans and their immune responses may differ from that of humans. A study from China reported that vaccination with inactivated virus protected 8 rhesus macaques from viral challenge, but the numbers are small. The vaccinated macaques also developed antibodies able to neutralize different strains of the virus. They observed no harmful effects, including enhancement, in which a vaccine induces an immune response that makes infection more severe.

The gold standard in vaccine efficacy testing is obviously controlled clinical trials in humans. There are two ways to do this. One is with a large cohort of people from a group that has a high attack rate from the virus. Obvious, the more susceptible the test population, the fewer the numbers and shorter the time period required for a definitive answer. The most straightforward way, however, is to challenge vaccinated and unvaccinated unpaid volunteers with virus. This has the advantage that far fewer numbers are needed than in trials using an unchallenged general population, and results are obtained more quickly. However, such an approach is obviously fraught with ethical concerns.

As mentioned above, there are at least seven phase 1 or 2 trials of targeted vaccines now underway, several of which hope to have a vaccine by fall. An RNA-based vaccine by Moderna (clinical trials ID NCT04283461) has been reviewed by the FDA for phase 2 testing, and it is hoped to start phase 3 trials this summer. A Chinese whole killed virus vaccine candidate by Sinopharm (Chinese Clinical Trial Registry IdentifierChiCTR2000031809) has reached phase 2 status. The Oxford Vaccine Group is conducting a phase 1/2 trial with a vaccine (clinical trials ID NCT04324606) based on a chimp adenoviral vaccine vector that reportedly has shown protection in rhesus macaques.  A vaccine developed by Pfizer and BioNTech, called BNT162, is based on modified RNA (clinical trial ID NCT04368728). A phase 1/2 trial is planned in Germany and the United States. Inovio’s vaccine, called INO4800, is a DNA-based vaccine. They are currently planning a phase 1 trial in the United States and a phase 1/2 trial in South Korea (clinical trials ID NCT04336410). CanSino Biologics has an adenovius 5-based vaccine candidate that the company announced has passed phase 1 trials and are initiating a phase 2 trial (clinical trials ID NCT04313127). Sinovac has initiated phase 1 trials with a vaccine candidate based on inactivated virus plus adjuvant (clinical trials ID NCT04352608).

With the large number of approaches, the relative genetic stability of the virus, and hopeful results from animal trials, it seems likely a vaccine will be successfully developed. The big question, of course, is how soon. The best case seems to be by fall of 2020. Pressure to speed up the schedule may spur further considerations of the ethics of human volunteer challenge studies. The development of better antiviral therapies may influence considerations of these types of trials, as it would make them somewhat safer.

 

Read the June 2, 2020 update here

View All GVN SARS-CoV-2 Perspectives