Clinical validation of RCSMS: a rapid and sensitive CRISPR-Cas12a test for the molecular detection of SARS-CoV-2 from saliva

TLDR: In this paper, the authors evaluated the performance of RCSMS, a locally adapted variant of DETECTR, to ascertain the presence of SARS-CoV-2 in saliva samples from 276 patients in two hospitals in Peru (current status over a total of 350 samples).

Abstract: Early detection of SARS-CoV-2 using molecular techniques is paramount to the fight against COVID-19. Due to its high sensitivity and specificity, RT-qPCR is the “gold standard” method for this purpose. However, its technical requirements, processing time and elevated costs hamper its use towards massive and timely molecular testing for COVID-19 in rural and socioeconomically deprived areas of Latin America. The advent and rapid evolution of CRISPR-Cas technology has boosted the development of new pathogen detection methodologies. Recently, DETECTR -a combination of isothermal RT-LAMP amplification and Cas12a-mediated enzymatic detection-has been successfully validated in the Netherlands and the USA as a rapid and low-cost alternative to RT-qPCR for the detection of SARS-CoV-2 from nasopharyngeal swabs. Here, we evaluated the performance of RCSMS, a locally adapted variant of DETECTR, to ascertain the presence of SARS-CoV-2 in saliva samples from 276 patients in two hospitals in Lima, Peru (current status over a total of 350 samples). We show that a low-cost thermochemical treatment with TCEP/EDTA is sufficient to inactivate viral particles and cellular nucleases in saliva, eliminating the need to extract viral RNA with commercial kits, as well as the cumbersome nasopharyngeal swab procedure and the requirement of biosafety level 2 laboratories for molecular analyses. Our clinical validation shows that RCSMS detects up to 5 viral copies per reaction in 40 min, with sensitivity and specificity of 93.8% and 99.0% in the field, respectively, relative to RT-qPCR. Since CRISPR-Cas biosensors can be easily reprogrammed by using different guide RNA molecules, RCSMS has the potential to be quickly adapted for the detection of new SARS-CoV-2 variants. Notably, estimation of its negative and positive predictive values suggests that RCSMS can be confidently deployed in both high and low prevalence settings. Furthermore, our field study validates the use of lateral flow strips to easily visualize the presence of SARS-CoV-2, which paves the way to deploy RCSMS as a “point of care” test in environments with limited access to state-of-the-art diagnostic laboratories. In sum, RCSMS is a fast, efficient and inexpensive alternative to RT-qPCR for expanding COVID-19 testing capacity in low- and middle-income countries.

Figures

Table 8: Clinical and epidemiological data of the samples used for clinical validation. Of 276 individuals, their distributions according to gender, age group, time of illness and symptoms are shown.

Figure 3:

Figure 2:

Figure 7:

Figure 6:

Table 9: Results of analytical and clinical (field) evaluations using the RT-qPCR and RT-LAMP / CRISPR-Cas tests. The degree of concordance is shown using the Kappa-Cohen index considering a 95% confidence interval (95% CI).

Full text

Clinical validation of RCSMS: a rapid and sensitive CRISPR-Cas12a test for the
molecular detection of SARS-CoV-2 from saliva
Joaquín Abugattás Núñez del Prado
1
, Angélica Quintana Reyes
1
, Juan Blume La Torre
1
,
Renzo Gutiérrez Loli
1
, Alejandro Pinzón Olejua
2
, Elena Rocío Chamorro Chirinos
3
, Félix
Antonio Loza Mauricio
4
, Jorge L. Maguiña
5
, Julio Leon
6
, Piere Rodríguez Aliaga
7
* y Edward
Málaga Trillo
1
*
1
Facultad de Ciencias y Filosofía, Universidad Peruana Cayetano Heredia, Lima, Perú
2
Department of Computer Science, Christian-Albrecht University of Kiel, Germany
3
Hospital Nacional Guillermo Almenara Yrigoyen, EsSalud, Lima, Perú
4
Hospital Nacional Edgardo Rebagliati Martins, EsSalud, Lima, Perú
5
Instituto de Evaluación de Tecnologías en Salud e Investigación (IETSI), EsSalud, Lima, Perú
6
IMS RIKEN Center for Integrative Medical Sciences, Japan
7
Department of Biology, Stanford University, California, USA
*Correspondence to: Edward.Malaga@upch.pe and piererod@stanford.edu
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256081doi: medRxiv preprint
NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.

ABSTRACT
Early detection of SARS-CoV-2 using molecular techniques is paramount to the fight against
COVID-19. Due to its high sensitivity and specificity, RT-qPCR is the "gold standard" method
for this purpose. However, its technical requirements, processing time and elevated costs
hamper its use towards massive and timely molecular testing for COVID-19 in rural and
socioeconomically deprived areas of Latin America. The advent and rapid evolution of
CRISPR-Cas technology has boosted the development of new pathogen detection
methodologies. Recently, DETECTR -a combination of isothermal RT-LAMP amplification and
Cas12a-mediated enzymatic detection- has been successfully validated in the Netherlands
and the USA as a rapid and low-cost alternative to RT-qPCR for the detection of SARS-CoV-
2 from nasopharyngeal swabs. Here, we evaluated the performance of RCSMS, a locally
adapted variant of DETECTR, to ascertain the presence of SARS-CoV-2 in saliva samples
from 276 patients in two hospitals in Lima, Perú (current status over a total of 350 samples).
We show that a low-cost thermochemical treatment with TCEP/EDTA is sufficient to inactivate
viral particles and cellular nucleases in saliva, eliminating the need to extract viral RNA with
commercial kits, as well as the cumbersome nasopharyngeal swab procedure and the
requirement of biosafety level 2 laboratories for molecular analyses. Our clinical validation
shows that RCSMS detects up to 5 viral copies per reaction in 40 min, with sensitivity and
specificity of 93.8% and 99.0% in the field, respectively, relative to RT-qPCR. Since CRISPR-
Cas biosensors can be easily reprogrammed by using different guide RNA molecules, RCSMS
has the potential to be quickly adapted for the detection of new SARS-CoV-2 variants. Notably,
estimation of its negative and positive predictive values suggests that RCSMS can be
confidently deployed in both high and low prevalence settings. Furthermore, our field study
validates the use of lateral flow strips to easily visualize the presence of SARS-CoV-2, which
paves the way to deploy RCSMS as a “point of care” test in environments with limited access
to state-of-the-art diagnostic laboratories. In sum, RCSMS is a fast, efficient and inexpensive
alternative to RT-qPCR for expanding COVID-19 testing capacity in low- and middle-income
countries.
Key words: Perú, saliva test, coronavirus, COVID-19, RCSMS, DETECTR, RT-PCR, SARS-
CoV-2
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256081doi: medRxiv preprint

INTRODUCTION
After sweeping across Asia and Europe, COVID-19 took main stage in Latin America, where
socioeconomic disparity and the saturation of public health systems have contributed to
alarming infection and mortality rates, particularly in Perú and Brazil. Very conservative
estimates from Perú’s National Death Information System (SINADEF) and Ministry of Health
(MINSA) show that the impact of COVID-19 on the Peruvian population (~30 MM) is dramatic,
totaling at least 1,730,000 cases, 58,604 deaths and 161,200 excess deaths associated with
COVID-19 as of March 30, 2021 (1, 2). The prolonged health emergency has particularly
affected vulnerable populations such as indigenous communities, urban and rural areas with
low-income, refugees and migrants.
Given the rapid spread of the disease and the arrival of new pandemic waves, it remains urgent
to implement early and massive SARS-CoV-2 testing strategies alongside with contact tracing
and isolation (3). Molecular tests must be sensitive, rapid and easily accessible to timely
identify and treat individuals at high risk of transmitting the infection and efficiently apply partial
lockdowns, border closures, and travel restrictions. In addition, molecular testing must be
integrated into a genomic surveillance system that tracks the appearance and distribution of
viral mutations. In sum, obtaining high-quality diagnostic data is essential to correctly monitor
the evolution of the epidemic, ensure the success of public health strategies and the transition
to a new normality.
Currently, ~600 COVID-19 diagnostic kits are commercially offered worldwide (4). These tests
are based on the detection of viral genes, viral proteins, or antibodies generated by the human
immune system against the virus. While antibody tests are recommended for seroprevalence
and epidemiological studies, the “gold standard” among COVID-19 molecular tests is the
quantitative reverse transcription-polymerase chain reaction (RT-qPCR) (5). Recently, viral
antigen detection tests have added speed and accessibility to COVID-19 molecular testing,
although their lower sensitivity relative to RT-qPCR limits their utility to individuals with high
viral loads and risk of transmission (6). In developing countries, RT-qPCR testing at the
population level is restricted by poor access to adequate equipment, supplies and
infrastructure, as well as by the need for trained personnel and high biosafety standards; in
fact, most diagnostic kits in Latin America are produced in the developed world (7). Hence, the
local development and distribution of molecular detection methods are key to the fight against
COVID-19 in these countries.
DETECTR is a recently developed molecular test that uses the Nobel-prize winning CRISPR-
Cas technology to detect SARS-CoV-2 RNA extracted from nasopharyngeal swabs, with high
efficiency (8). This method couples two reactions: i) the first reaction performs simultaneous
reverse transcription and loop-mediated isothermal amplification (RT–LAMP) to generate DNA
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256081doi: medRxiv preprint

copies from viral RNA and amplify them exponentially at a single constant temperature
(between 60 and 65°C), using three pairs of nested primers and a DNA polymerase with
displacement and replication activities (9), and ii) the second reaction exploits the CRISPR-
Cas12a recognition system to target and detect the RT-LAMP-amplified viral gene sequence,
unleashing the collateral DNAse activity of Cas12a and leading to the indiscriminate cut of
single-stranded DNA reporter molecules that produce either fluorescent or
immunochromatographic readouts upon cleavage (8). A significant advantage of DETECTR
over other methods is that it requires neither sophisticated equipment nor specialized
personnel, and it is a low-cost alternative with sensitivity and specificity comparable to RT-
qPCR (6). Altogether, RT-LAMP and CRISPR-Cas12a reactions have a processing time of
~40 minutes, very convenient compared to the 2-4 hours required for the RT-qPCR reaction
and its even longer data processing time (6). Another important advantage of DETECTR over
RT-qPCR is its ability to confirm the presence of SARS-CoV-2 RNA not only via standard
fluorescence but also by using lateral flow strips, which simply requires replacing the
fluorophore quencher with biotin in the reporter molecules. This results in an inexpensive
visualization format, similar to a pregnancy test, easy to apply and interpret (8). Importantly,
various molecular tests that similarly combine RT-LAMP and CRISPR-Cas have successfully
been used to detect Zika, Dengue and HIV viruses in humans, and some coronaviruses in
animals (10).
COVID-19 Nucleic Acid Tests (NATs) -such as DETECTR or RT-qPCR- have two major
limitations related to sample collection. On one hand, the of nasopharyngeal swab procedure
is uncomfortable for patients, risky for sample collection personnel, and generates the need
for swabs and costly viral transport media. On the other hand, the extraction of high-quality
viral RNA in standard laboratories requires commercial kits, expensive reagents and long
processing times. Interestingly, recent studies have shown that respiratory samples can be
directly lysed and used efficiently for molecular diagnostics without commercial RNA extraction
kits (11-14). This simple incubation method uses standard low-cost laboratory equipment and
is based on a simple thermochemical reaction that guarantees 1) inactivation of RNases in the
sample and 2) lysis of viral particles, thus ensuring the release of viral genetic material and
rendering the sample non-infectious (11-14). This is of great relevance for COVID-19
molecular testing in low and middle-income countries because it eliminates the need for highly
trained personnel and biosafety level 2 facilities.
In the present study, we report the development, standardization and clinical validation of a
new procedure that combines the efficiency and robustness of DETECTR (i.e. RT-LAMP and
CRISPR-Cas coupled reactions) with the rapid and low-cost TCEP/EDTA thermochemical
method for viral RNA preparation from saliva samples. We show that our integrated system -
called Rapid Coronavirus-Sensitive Monitoring from Saliva, RCSMS- successfully detected
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256081doi: medRxiv preprint

SARS-CoV-2 RNA in a set of 276 saliva samples (current status from a total of 350 samples)
from patients from two hospitals in Lima. The clinical diagnostic performance of RCSMS was
validated by comparison with routine RT-qPCR from nasopharyngeal swabs under field
conditions. Together, our results show that RCSMS is a simple and easy-to-implement
molecular test at the primary care level, outside the laboratory, with a great potential to
contribute to the massification of COVID-19 molecular testing in Peru and neighboring
countries.
MATERIALS AND METHODS
Study design
The present study, aimed at validating the use of RCMS to test for the presence of SARS-
CoV-2 in saliva from Peruvian patients, was developed in three stages:
1) Method optimization. To establish the appropriate experimental conditions for the molecular
detection of SARS-CoV-2 via DETECTR, synthetic viral RNA was generated from DNA
templates and taken as input material (see below), using PCR products as amplification control
for DETECTR. This stage evaluated the performance of DETECTR under ideal conditions, in
particular: a) the robustness of the RT-LAMP reaction upon changes in reaction time (between
20 and 30 min) and temperature (gradient between 62 and 68°C), b) the limit of detection
(LOD) of the CRISPR-Cas reaction (serial dilutions of the RNA substrate for RT-LAMP), c) the
robustness of the CRISPR-Cas reaction to changes in the final concentration of Cas12a
enzyme and guide RNAs (0.5, 1, 2 and 3X) in the ribonucleoprotein complex (RNP), length of
guide RNAs (41 and 44 bp), preincubation time for the formation of the RNP complex (10, 20
and 30 min), length of fluorescent and biotinylated ssDNA probes (5 and 8 bp), final
concentration of fluorescent (20, 50 and 100 nM) and biotinylated probes (20, 50, 100, 200,
400, 500 and 600 nM), CRISPR detection time by fluorescence (5-30 min) and
immunochromatography (10, 20 and 30 min), amount of RT-LAMP product (1, 2, 3 and 5 µl)
and d) the repeatability of measurements (three replicas). These evaluations also allowed us
to optimize the efficiency, cost and readout levels of the test, as well as to verify its
reproducibility upon variations in instrumentation and operators.
2) Analytical validation. To assess the analytical specificity of DETECTR, we relied on in vitro
data from a similar study (8) and examined the possibility of cross-reactivity in the primers and
guide RNAs used here. For this, we ran a comparative in silico analysis of the corresponding
regions in common human coronavirus sequences (HCoV-HKU1 (NC_006577.2), HCoV-NL63
(NC_005831.2), HCoV-OC43 strain ATCC VR-759 (NC_006213.1), MERS-CoV
(NC_019843.3), SARS-CoV (NC_004718) and SARS-CoV-2; NC_045512). The potential
ability of the test to detect viral variants currently circulating in Perú was inferred by aligning
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted April 29, 2021. ; https://doi.org/10.1101/2021.04.26.21256081doi: medRxiv preprint

Frequently asked questions

Q1. What are the contributions mentioned in the paper "Clinical validation of rcsms: a rapid and sensitive crispr-cas12a test for the molecular detection of sars-cov-2 from saliva" ?

Here, the authors evaluated the performance of RCSMS, a locally adapted variant of DETECTR, to ascertain the presence of SARS-CoV-2 in saliva samples from 276 patients in two hospitals in Lima, Perú ( current status over a total of 350 samples ). The authors show that a low-cost thermochemical treatment with TCEP/EDTA is sufficient to inactivate viral particles and cellular nucleases in saliva, eliminating the need to extract viral RNA with commercial kits, as well as the cumbersome nasopharyngeal swab procedure and the requirement of biosafety level 2 laboratories for molecular analyses. Since CRISPRCas biosensors can be easily reprogrammed by using different guide RNA molecules, RCSMS has the potential to be quickly adapted for the detection of new SARS-CoV-2 variants. Notably, estimation of its negative and positive predictive values suggests that RCSMS can be confidently deployed in both high and low prevalence settings. Furthermore, their field study validates the use of lateral flow strips to easily visualize the presence of SARS-CoV-2, which paves the way to deploy RCSMS as a “ point of care ” test in environments with limited access to state-of-the-art diagnostic laboratories. 

Q2. What is the “gold standard” among COVID-19 molecular tests?

While antibody tests are recommended for seroprevalence and epidemiological studies, the “gold standard” among COVID-19 molecular tests is the quantitative reverse transcription-polymerase chain reaction (RT-qPCR) (5). 

Q3. What is the importance of implementing early and massive testing strategies?

Given the rapid spread of the disease and the arrival of new pandemic waves, it remains urgent to implement early and massive SARS-CoV-2 testing strategies alongside with contact tracing and isolation (3). 

Q4. How many copies of the virus were detected in a single assay?

RCSMS routinely detected 50 copies per 10 µl reaction in triplicate assays, occasionally reaching five copies in one or two replicates. 

Q5. How many false positives were recorded between the two tests?

In addition, seven incongruencies (2.8%) were recorded between the two tests: two false positives (TFP=2.6%) and five false negatives (TFN=2.5%). 

Q6. Why is a sharp decrease of PPV common to all diagnostic tests?

A sharp decrease of PPV in low prevalence settings is common to all diagnostic tests because of the higher probability of obtaining false positives (31). 

Q7. What is the importance of obtaining high-quality diagnostic data?

In sum, obtaining high-quality diagnostic data is essential to correctly monitor the evolution of the epidemic, ensure the success of public health strategies and the transition to a new normality. 

Q8. What is the effect of the RT-LAMP reaction on the ssDNA reporter?

Recognition of viral target sequences by the RNP complex triggers the collateral activity of Cas12a, resulting in the cleavage of ssDNA reporter probes. 

Q9. What is the procedure for a lateral flow strip?

For immunocromatographyic (qualitative) readout, a lateral flow strip is then inserted into the CRISPR-Cas12a reaction tube or well. 

Q10. What is the effect of the NPV of diagnostic tests in high prevalence settings?

the NPV of diagnostic tests is expected to drop in high prevalence settings, as the number of false negatives increase (31). 

Q11. What is the way to ensure a low rate of false negatives?

Ensuring a low rate of false negatives (positive individuals being deemed negative) is especially desirable considering that COVID-19 is a serious, largely asymptomatic and contagious disease, for which early measures are advisable. 

Q12. How many individuals did not record their date of symptom onset?

Ten individuals (3.6% of the total) did not record their date of symptom onset; therefore, their results were excluded from the analysis stratified by time of illness. 

Q13. How many positive samples were found in one of the three rounds of analyses?

three rounds of analyses reproduced these results, with only two positive samples yielding a negative result in one of their three replicates. 

Q14. How does the RCSMS simulation predict a high prevalence?

their statistical simulation for RCSMS using the currently available data (276 out of 350 samples analyzed, Fig. 7), predicts excellent NPVs >95% for prevalence scenarios as low as 1% and as high as 50%. 

Q15. How many copies of the three viral genes were detected?

The limit of detection (LOD) for both viral genes was determined using serial dilutions of the synthetic RNA templates, covering the range from 250 to 1 viral copy per 10 µl RT-LAMP reaction. 

Q16. What is the significance of the PPVs for RCSMS?

These PPV estimates are consistent with the results from their field study, where prevalence reached almost 30%, strengthening the conclusion that RCSMS is particularly well suited for the identification of positive cases during acute pandemic phases, such as the current second waves hitting Perú, India and Brazil. 

Q17. What was the level of concordance between RT-qPCR and RCSMS?

the level of concordance (Cohen's Kappa coefficient) between RT-qPCR and RCSMS was calculated for both analytical and clinical validations, as well as sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), the area under the curve (AUC), accuracy of RCSMS, and PPV/ NPV simulations assuming various prevalence scenarios.