Accessible LAMP-Enabled Rapid Test (ALERT) for detecting SARS-CoV-2
Ali Bektaş
1*
, Michael F. Covington
1,8
, Guy Aidelberg
2
, Anibal Arce
3, 5
, Tamara Matute
3,5
, Isaac Núñez
3,5
, Julia Walsh
4
, David
Boutboul
6
, Ariel B. Lindner
2
, Fernán Federici
3, 5
, Anitha Jayaprakash
1,7
1
Oakland Genomics Center, 355 30th St. Oakland, CA, 94609, USA
2
Université de Paris, INSERM U1284, Center for Research and Interdisciplinarity
(CRI), F-75006 Paris, France
3
Institute for Biological and Medical Engineering, Schools of Engineering, Biology and Medicine, Ponticia Universidad
Católica de Chile, Santiago 7820244, Chile
4
School of Public Health, University of California Berkeley, Berkeley, CA, 94720, USA
5
ANID – Millennium Sci-
ence Initiative Program- Millennium Institute for Integrative Biology (iBIO), Santiago, Chile, FONDAP Center for Genome Regulation.
6
Clinical Immunology
Department, U976 HIPI, Hôpital Saint Louis, Université de Paris, Paris, France
7
Girihlet Inc, 355 30th St, Oakland, CA, 94609, USA
8
Amaryllis Nucleics,
355 30th St, Oakland, CA, 94609, USA
Abstract:
The COVID-19 pandemic has highlighted bottlenecks in large-scale, frequent testing of populations for infections. PCR-based
diagnostic tests are expensive, reliant on expensive centralized labs, can take days to deliver results, and are prone to backlogs and
supply shortages. Antigen tests, that bind and detect the surface proteins of a virus, are rapid and inexpensive but suffer from high false
negative rates. To address this problem, we have created an inexpensive, simple, and robust 60-minute Do-It-Yourself (DIY) workow
to detect viral RNA from nasal swabs or saliva with high sensitivity (0.1 to 2 viral particles/µl) and specicity (>97% True Negative Rate)
utilizing reverse transcription loop-mediated isothermal amplication (RT-LAMP).
Our workow, ALERT (Accessible LAMP-Enabled Rapid Test), incorporates the following features: 1) Increased shelf-life and
ambient temperature storage by using wax layers to isolate enzymes from reaction, 2) Improved specicity by using sequence-specic
QUASR reporters, 3) Increased sensitivity through use of a magnetic wand to enable pipette-free concentration of sample RNA and cell
debris removal, 4) Quality control with a nasopharyngeal-specic mRNA target, and 5) Co-detection of other respiratory viruses, such as
Inuenza B, by duplexing QUASR-modied RT-LAMP primer sets.
The exible nature of the ALERT workow allows easy, at-home and point-of-care testing for individuals and higher-throughput
processing for centralized labs and hospitals. With minimal effort, SARS-CoV-2-specic primer sets can be swapped out for other targets
to repurpose ALERT to detect other viruses, microorganisms or nucleic acid-based markers.
*Corresponding author: alibektas@berkeley.edu
Introduction
The COVID-19 pandemic has brought the eld of molecular
diagnostics into the spotlight. Since the initial release of sequence
data for SARS-CoV-2 on January 10, 2020 (Genbank Accession #
MN908947), a positive-sense single-stranded RNA of approximately
29.8 kb, RT-PCR assays have been designed by the WHO, the
US CDC, the Chinese CDC, Institut Pasteur and others [1]. An
unprecedented demand for RT-PCR diagnostics has put a strain on
every aspect of conducting these laboratory-based assays. Despite
the availability of quickly developed protocols (e.g. [2], [3], [4], [5]),
shortages in materials (e.g., swabs, reagents, and consumables)
and of infrastructure (e.g., approved facilities, technicians, and
equipment) have prevented the efcient testing, tracing, and
isolation of infectious individuals. These shortages have led to a
global public health crisis of unforeseen consequences, which is
further exacerbated in the global south.
Conventional RT-PCR-based assays, which require a level of
training, are neither rapid, inexpensive, nor highly scalable without
the use of cost-prohibitive equipment. To address the challenge
of diagnostics at scale there has been an increased emphasis
on simpler detection techniques conducted at the point-of-care
or even in the convenience of one’s own home. Most of these
methods can be broadly categorized into two groups in relation
to their targets, nucleic acids or proteins (i.e. antigen). Amongst
the rapid methods targeting nucleic acids, reverse transcription
loop-mediated isothermal amplication (RT-LAMP)[6] has risen in
prominence during the course of the COVID-19 pandemic. Beyond
its isothermal nature, which eliminates the need for complex
instruments such as thermocyclers necessary for PCR, RT-LAMP
is attractive since it produces an immense amount of amplication
products allowing visual detection by the naked eye via diverse
methods [7]. Since the emergence of COVID-19, multiple primer
sets for SARS-CoV-2 [8] [9] and variations on the RT-LAMP method
have been published [10] and some have received emergency use
authorization by the FDA (Color and Mammoth Biosciences). While
the RT-LAMP workow delivers results faster than RT-PCR and is
highly scalable, it still requires a trained technician to perform the
assay as well as cold shipping and storage of reagents. We have
developed the Accessible LAMP-Enabled Rapid Test (ALERT) to
address some of the shortcomings.
Materials And Methods
Controls and Reference Material
SARS-CoV-2 synthetic RNA controls were purchased from
Twist Biosciences (San Francisco, CA, SKU:102019) and Inuenza
B virus RNA from ATCC (Catalogue Number VR-1813D).
SARS-CoV-2 virus was cultured at the Innovative Genome
Institute’s BSL3 facility at University of California Berkeley and
inactivated in RNAShield (gift of Xammy Hu Nguyenla) at a 2.5e5
PFU/ml concentration. Samples were spiked with virus at units
ranging between 125 PFU and 1000 PFU per RT-LAMP reaction
These samples were processed in BSL2-dedicated rooms following
strict biosafety guidelines.
RNA positive controls for the reactions performed in Chile
were created by in vitro transcription using the Hi-Scribe kit
from NEB (catalogue E2040S) from an amplicon obtained by
PCR amplication of the IDT control for Sars-CoV-2 (catalogue
10006625) using the following primers: NT7_Fw (CGA AAT TAA
TAC GAC TCA CTA TAG GGG CAA CGC GAT GAC GAT GGA
TAG) and T7_Nter_Rv (ACT GAT CAA AAA ACC CCT CAA GAC
CCG TTT AGA GGC CCC AAG GGG TTA TGC TAG TTA GGC
CTG AGT TGA GTC AG).
In vitro transcribed RNA was treated with DNAse I (NEB
M0303S) for 15 minutes at 37°C before purication with Qiagen
RNeasy kit, and serially diluted to the concentrations used.
Sample Collection
Human nasal samples were collected with either
nasopharyngeal (NP) swabs, nasal mid-turbinate swabs (NMT),
or a nasopharynx ush thru (NFT) method. NFT samples were
collected by using a 5 ml dropper lled with water squirted into a
nostril with the head tilted back and the water passing through the
nasopharynx and subsequently spit into a 50-ml Falcon tube.
For collecting nasal mid-turbinate samples, we used a variety
. CC-BY-NC-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 preprint this version posted February 20, 2021. ; https://doi.org/10.1101/2021.02.18.21251793doi: 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.
of swabs due to supply chain shortages, ranging from Q-tip cotton
swabs to ocked nylon nasopharyngeal swabs. We observed that
some sample collection swabs, such as those supplied by Zymo
Research Corporation (C105250), had blue uorescence emission
when excited by UV light thus interfering with the interpretation of
results following our crude sample lysis protocols. We currently use
sterile, ocked oral swabs with a 30mm breakpoint compatible with
1.5ml microcentrifuge tubes (SNT Biotech, www.sntbiotech.com).
Sample Lysis
We experimented with different lysis buffers. Our Proteinase
K extraction method begins by inserting a freshly used NP or NMT
swab into 300 µl of PK Lysis Solution and agitating the swab for
30 seconds. 75 µl of the resulting solution is used and, depending
on the experiment, spiked with inactivated SARS-CoV-2. Samples
were heated at 55°C for 5 min for lysis and 95°C at 10 min for
inactivation. The PK Lysis Solution contained ~15 mAU/ml and a
10 ml stock was prepared by diluting 0.25 ml Proteinase K (Qiagen,
19131) with 9.75 ml water.
In order to process samples already in solution, such as
PBS, nasal swabs (in an unknown solution), and saliva sent by
the XPrize foundation (www.xprize.org/prizes/covidtesting) we
used QuickExtract DNA extraction solution (Lucigen, catalogue #
QE09050) following the protocol by Kellner et al. (2020).
For our DIY workow, our buffer of choice is TCEP/EDTA
following the protocol outlined by Rabe and Cepko (2020) of a
modied HUDSON buffer (Myhrvold et al. 2018). In contrast to
other protocols where a concentrated 100X HUDSON buffer is
added to a liquid sample such as saliva we used a 1X solution
(2.5mM TCEP, 1mM EDTA) for the direct addition of an NMT swab.
Nucleic acid extraction
RNA from NP, NMT, and NFT samples was isolated using
both a QIAamp Viral RNA Mini Kit (Qiagen Catalogue # 52904)
according to the manufacturer’s instructions as well as with a
TRIzol/chloroform method[11]. NP and NMT swabs were inserted
into 1ml of PBS and vortexed prior to extraction while 1 ml of NFT
samples was used directly for extraction.
RNA Isolation with Paramagnetic Beads and Magnetic Wand
For paramagnetic bead-based RNA capture, 0.6 volumes of
a 1X bead solution was added to the lysed sample and uniformly
suspended. After a 5-minute incubation at room temperature the
samples containing the beads were placed on an appropriately
sized magnetic rack and beads were pelleted by the magnet for 5
minutes. All liquid was removed, and a brief wash was conducted
with 100 µl of 85% ethanol. Following the removal of ethanol,
sample tubes were removed from the magnetic rack and the
beads were resuspended in 10-20 µl of water. Depending on the
cartridge design, 5-10 µl of beads in water were directly added
to the cartridge. The 1X bead solution is prepared by combining
one part RNAClean XP beads (Beckman Coulter Life Sciences,
A63987) with ve parts of a concentrated Bead Binding Buffer.
The composition of the Bead Binding Buffer [12] is 20% PEG
8000 (Sigma-Aldrich, P1458), 10 mM Tris HCl, pH 8.0 (Invitrogen,
15568025), 0.05% Tween 20 (Thermo Scientic, J20605AP), 5 mM
Sodium Azide (MP Biomedicals, 0210289125), Sodium Chloride
(Fisher Chemical, S271-500).
To enable a DIY workow, we 3D printed the sheaths for
the anged magnetic wand with a $450 3D printer (Adventurer
3; Flashforge) in as few as 10 minutes using about $0.02 worth
of 1.75 mm diameter PLA lament (Flashforge) per wand. These
anged sheaths accommodate a $0.72 cylindrical grade N42
neodymium magnet (1-inch-long x 1/16-inch diameter; D1X0, K&J
Magnetics, Inc.). The diameter and vertical position of the ange
can be adjusted to accommodate a diverse array of sample tubes.
The diameter of the rest of the sheath is optimal for depositing the
beads in RT-LAMP reaction cartridges that have been prepared
in 0.2 ml strip tubes. If larger tubes are used for the RT-LAMP
reaction cartridge, the diameter of the sheath can be increased
to accommodate more powerful, larger diameter magnets. Using
these designs as a prototype, the magnetic wand sheaths can also
be produced in bulk with injection-molded low-retention plastics.
Capturing the beads with magnetic wands followed a similar
protocol to using a magnetic rack but without the use of pipettes.
After the 5-minute room temperature incubation the magnetic wand
was inserted into the sample/bead solution. A further 5-minute
room temperature incubation was sufcient to allow RNA-bound
beads to be attracted by the wand. The wand, now with beads,
was removed from the sample. At this stage an optional bead wash
can be performed by briey dipping (~2 seconds) the magnetic
wand into a tube with 85% ethanol lled to the same volume as the
sample/bead solution. If an ethanol wash was performed, we let the
beads on the magnetic wand air dry without over-drying beads (~2-
5 minutes). The magnetic wand was placed in a tube containing
eluant (in this case, our RT-LAMP cartridges with 10 µl water on
the topmost layer of wax). The magnet was removed from the wand
and the wand was swirled to release beads into the eluant.
Construction of Wax-layered Cartridges
For a sampling and process control target sequence, we
chose the BPI fold containing family A, member 1 gene (BPIFA1)
as its expression is nasopharyngeal specic [37]. We designed
a LAMP primer set for this gene that traverses an exon-exon
boundary, making the reaction specic to BPIFA1 mRNA. The
SARS-CoV-2 [8][10] and Inuenza B [13] primers were adopted
from the literature but modied for the sequence specic QUASR
(Quenching of Unincorporated Amplication Signal Reporters)
reporting system [38] (Supplemental Table 1).
1-step RT-LAMP
1-step RT-LAMP cartridges using the NA/NB primer set were
prepared by adding 27 µl of RT-LAMP primer mix to the bottom of
a 0.75 ml cryogenic tube (Micronic, Netherlands) followed by 200
µl of melted silicone wax (Siltech D-222, melting point 37°C). After
solidication of this wax layer a small divot was made with a 1 ml
micropipette tip, deposited into which was 3 µl of an enzyme mix
consisting of 1 µl (15 units) of WarmStart RTx (NEB, Ipswich, MA,
M0380) and 2 µl (16 units) of Bst 3.0 (NEB, Ipswich, MA, M0374).
The enzyme droplet was capped by a further addition of 300 µl
of the same silicone wax. In an alternative construction of this
cartridge, we added a layer of 150 µl cold mineral oil (consumer
grade) between the enzyme droplet and silicone wax to shield it
from the thermal shock of hot wax.
For the single wax layer RT-LAMP cartridges using this NA/
NB primer set, the primer mix, prepared in Isothermal Amplication
Buffer (I) (NEB Ipswich, MA, B0537S) contained 2.1 µM each of
NB-FIP and NB-BIP, 0.32 µM each of NB-F3 and NB-B3, 1.3 µM
each of NB-LB and NB-LF-Tx and 2.2 µM of NB-LF-Q. MgSO4
concentration was at 7.8 mM and dNTPs at 1.8 mM. 27 µl of the
primer mix was used as the bottom layer of reaction cartridges
anticipating a 35 µl nal reaction volume together with 1 µl
WarmStart Rtx (NEB, Ipswich, MA, M0380), 2 µl Bst 3.0 (NEB,
Ipswich, MA, M0374) and 5 µl of sample.
For direct addition of paramagnetic beads using the magnetic
wand, we used the NM primer set (Supplemental Table 1) in a 30 µl
nal reaction volume cartridge consisting of (from bottom to top) a
16.5 µl primer mix, a layer of parafn wax (IGI 1250A melting point
of 61.4°C), and an enzyme mix of 1.5 µl (22.5 units) of WarmStart
RTx (NEB, Ipswich, MA, M0380) and 2 µl (16 units) of Bst 2.0
(NEB, Ipswich, MA, M0374). The enzyme was trapped between
two layers of solid wax by adding a nal 4 mm solid, lower-melting-
point silicone wax (Siltech, Silwax D-222, melting point of 37°C)
and melting it down briey at 40°C for 1 min. When the magnetic
2
. CC-BY-NC-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 preprint this version posted February 20, 2021. ; https://doi.org/10.1101/2021.02.18.21251793doi: medRxiv preprint
wand was used, 10 µl of molecular biology grade water for bead
deposition was added above this wax layer. (Figure 1a and 1b).
For one-step cartridges using this NM primer set, the 16.5
µl primer mix was prepared with Isothermal Amplication Buffer (I)
(NEB Ipswich, MA, B0537S) with 2.9 µM NM-FIP-F and NM-BIP,
4.35 µM NM-FIP-Q, 1.5 µM NM-LF and NM-LB, and 0.4 µM NM-F3
and NM-B3. MgSO4 was at 12.8 mM and dNTPs at 2.5 mM.
Depending on the primer mixes, 5 µl (for NA/NB) or 10 µl (for
NM) of sample was added to the top layer. Cartridges containing the
NA/NB set were incubated in a 63°C dry block with a heated lid for 1
hr. Cartridges containing the NM primer set, and where the enzyme
was packaged between waxes of two different melting points, tubes
were rst heated to 64°C to allow the mixing of all reagents and
then cooled to 55°C for a 12.5 min reverse transcription reaction,
followed by a 64°C LAMP incubation for 45 min to 80 min.
For the BPIFA1 primer set, standard RT-LAMP primer
concentrations were followed: 1.6 µM FIP and BIP, 0.4 µM LF and
LB-F, 0.2 µM F3 and B3, and 0.6 µM LB-Q. 7 mM MgSO4 and
1.4 mM dNTPs were used in a 25 µl reaction in 1X Isothermal
Amplication Buffer (I) (NEB Ipswich, MA, B0537S) with 0.75 µl
WarmStart Rtx (NEB, Ipswich, MA, M0380) and 1.5 µl Bst 2.0 (NEB
Ipswich, MA, M0538L). BPIFA1 RT-LAMP reactions were incubated
at 63°C for 30 min.
A wax-layered cartridge that is ready-to-run following sample
addition presents a challenge towards the homogenous mixing of
all of its components owing to a pipette-free operation. We found
that the addition of a 2 mm glass bead to the top of the tube allows
for efcient and thorough mixing by both allowing for reagents to
fall through wax layers as well as displacing the bottommost primer
mix.
2-Step RT/LAMP
Cartridges separating the reverse transcription (RT) step
from the loop-mediated isothermal amplication step (LAMP)
were constructed into 8-tube strips using two separate waxes with
distinct melting points. The 10 µl LAMP primer mix, at the bottom
of the cartridge, was sealed from the RT primer mix above it with a
layer of parafn wax (IGI 1250A; melting point of 61.4°C). This seal
was created by adding a roughly 3 mm diameter piece of solid wax
above the LAMP primer mix and melting it at 65°C for 1 min. After
solidication of this wax layer at room temperature, 7 µl of RT primer
mix was added above it. Two drops of a lower melting point silicone
wax (Siltech, Silwax D-222; melting point of 37°C), heated to 50°C
in a water bath, was added on top of the RT primer mix using a 1
ml micropipette and allowed to solidify at room temperature. Finally,
an enzyme mix consisting of 1 µl (15 units) WarmStart RTx (NEB,
Ipswich, MA, M0380) and 2 µl (16 units) of Bst 3.0 (NEB, Ipswich,
MA, M0374) was added as the top layer of the cartridge (Figure 2a-
c).
For the 2-step detection cartridges, we made two separate
primer mixes corresponding to the reverse transcription (RT Mix)
and loop-mediated isothermal amplication (LAMP Mix) steps.
The RT Mix contained the F3 outer primer from the N-A primer
set [8] together with the B3 primer from the N-B primer set, which
we found to improve the detection limit compared to using only
the NB-B3 primer. This may result from DNA polymerase activity
stemming from the inclusion of Bst 3.0 during the RT step. NA-F3
and NB-B3 were at 0.54 mM along with 12.9 mM of MgSO4 and
1.07 mM of dNTPs in NEB Isothermal Amplication Buffer (I) at a
2X concentration. In each cartridge 7 µl of this mix was used in
anticipation of a 15 µl reverse transcription reaction together with 1
µl WarmStart Rtx (NEB, Ipswich, MA, M0380), 2 µl Bst 3.0 (NEB,
Ipswich, MA, M0374), and 5 µl of sample.
The LAMP Mix was composed of the NB primer set published
by Zhang et al. (2020) but modied for the sequence-specic
QUASR reporting system. This mix, with the NEB Isothermal
Amplication Buffer (I) (NEB Ipswich, MA, B0537S) at a 1X
concentration, had primer concentrations of 4 µM NB-FIP and NB-
BIP, 0.625 mM NB-F3 and N-B-B3, 2.5 µM NB-LB and NB-LF-Tx,
and 5 µM of NB-LF-Q. MgSO4 concentration was at 6 mM and
dNTPs at 3.5 mM. The concentration of the LAMP mix is higher
than standard reactions; the nal LAMP reaction volume will be 25
µl when all reagent mixes are combined at the bottom of the tube.
In each cartridge 10 µl of this LAMP mix was used.
The 2-step cartridges (Figure 2a) had 5 µl of sample added to
the enzyme droplet on the top layer of lower melting point silicone
wax, incubated in a thermal cycler at 55°C for 2 minutes allowing
for the enzyme droplet and sample to fall through liquid wax and
meet the RT primer mix. Cartridges were removed from incubation
and the wax allowed to briey solidify in order to allow for mixing of
reagents by manual shaking. After mixing, cartridges were further
incubated for 20 min at 55°C for the RT step to take place (Figure
2b). After this step, the temperature was raised to 63°C for 2 min to
allow the RT reaction to fall through the now liquied parafn wax
layer and meet the LAMP primer mix. The cartridges were again
removed from incubation, wax layers allowed to solidify, and the
RT/LAMP Mixes, now together under a layer of solid silicone and
another of parafn wax, mixed briey by vortex. The LAMP reaction
was then allowed to progress for 45 min at 63°C (Figure 2c).
Duplex RT-LAMP reactions
The nal 25 µl reactions, prepared with 1X Isothermal
Amplication Buffer (I) (NEB Ipswich, MA, B0537S) had 1.6 mM of
both FIP and BIP primer pairs, 0.2 mM of Inuenza B or BPIF1A F3
and B3 primer pairs, 0.25 µM of the NB-F3 and NB-B3 primer pair,
1 µM of both LF and LB primer pairs (LF or LB being modied with
a uorophore) and 2 mM of the corresponding LF or LB quencher
primer, 6 mM MgSO4, 1.6 mM dNTPs, 1 µl (15 units) of WarmStart
RTx (NEB, Ipswich, MA, M0380), and 2 µl (16 units) of Bst 3.0
(NEB, Ipswich, MA, M0374). Reactions were incubated at 63°C for
45 min.
Visualization
End-point reactions were visualized on a UV Transilluminator.
Alternatively, for reactions utilizing uorescein tagged
oligonucleotides we also used the low-cost (<$2) open-source
GMO Detective uorescence detector ([35] and https://github.com/
MakerLabCRI/GMODetective-Detector).
Incubation
Incubations (55°C and 63°C or 64°C) were performed in a
thermalcycler. We have also successfully performed all reactions
in a dry block incubator with a heated lid (Benchmark Scientic,
BSH200-HL).
Analytical Validation for 2-Step RT/LAMP: XPRIZE Blinded
Prociency Test
As a semi-nalist in the XPRIZE Covid Testing challenge [14]
we had the opportunity to participate in a blinded prociency test
alongside 218 other teams. Two separate 96-well sample plates
were mailed to our laboratory in Oakland, CA, with one consisting
of 86 synthetic RNA samples spiked into water (SARS-CoV-2
and 15 other viruses to test cross-reactivity) shipped on dry ice,
and the other plate consisting of 67 samples of different matrices
(phosphate buffered saline (PBS), nasal swabs resuspended in
an unknown buffer, and saliva) spiked with SARS-CoV-2 particles
from Zeptometrix shipped with cold packs. The synthetic RNA plate
was kept at -80ºC and the Zeptometrix plate was kept at 4°C until
analysis. For RNA extraction of the XPRIZE Rapid Covid Testing test
samples, 200 µl of QuickExtract DNA Extraction Solution (Lucigen,
QE09050) was added to 200 µl of each sample. Samples were
incubated for 5 minutes at room temperature prior to inactivation of
the QuickExtract solution at 95ºC for 10 minutes. The samples were
cooled on ice to room temperature prior to adding 240 µl of bead
3
. CC-BY-NC-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 preprint this version posted February 20, 2021. ; https://doi.org/10.1101/2021.02.18.21251793doi: medRxiv preprint
Room temperature
64°C
1 2 3 4 5 6 7 8
9 10 11 12 13 14 15 16
17 18 19 20 21 22 23 24
a. b. c.
d.
e.
f.
g.
h.
a.
b.
2000pfu 1000pfu 500pfu 250pfu
125pfu 0pfu
2000pfu 1000pfu 500pfu 250pfu
125pfu 0pfu
Non-template water control
0pfu 125pfu 250pfu 500pfu
200cp Twist 100cp Twist Non-template water
Figure 1. Easy to use ALERT Cartridges. One Step reaction-ready cartridge layering. Cartridge layers at room temperature (A) and at 64°C (B).
Demonstration of magnetic wand. Wand accumulates beads (C and D) and is placed into the reaction cartridge where the magnet is removed to
release beads (E and F). Reagents have dyes added for demonstration purposes: primer mix is orange and the enzyme droplet is black G. Detection
of SARS-CoV-2 using Proteinase K, magnetic wand extraction and a one-step cartridge, NMT swabs spiked with inactivated SARS-CoV-2 H. Same
tubes visualized with the GMO Detective.
0.4ng IB RNA 2000cp Twist 4000cp Twist Non-template water control
9 10 11 12 13 14 15 16
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
9 10 11 12 13 14 15 16
17 18 19 20 21 22
1 2 3 4 5 6 7 8
Room temperature 63°C55°C
a.
b.
c.
a.
b.
c.
a. b. c.
d.
e.
500cp 100cp 50cp
25cp Non-template water control
500cp 100cp 50cp
25cp Non-template water control
0.2ng IB RNA+
NP RNA 2000cp Twist 4000cp Twist NT Water Cnt.
NP RNA +
50cp Twist 25cp Twist Non-template water control
Non-template water control
500pfu
nCoV2
Figure 2. Two Step reaction-ready cartridge layering. Cartridge layers at room temperature (A), 55°C (B) and at 64°C (C). D. 1-Step RT-LAMP LoD
conrmed to be at ~500 copies of Twist Biosciences SARS-CoV-2 control as reported by Zhang et al. 2020 E. 2-step RT/LAMP cartridge with increased
sensitivity with LoD down to 25 copies of Twist Biosciences SARS-CoV-2 control
4
. CC-BY-NC-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 preprint this version posted February 20, 2021. ; https://doi.org/10.1101/2021.02.18.21251793doi: medRxiv preprint
solution as described above. After thoroughly mixing the beads into
the samples, beads were allowed to bind RNA for 5 minutes at
room temperature prior to placing the samples on a magnetic rack
for 5 minutes to pellet the RNA-bound beads. The supernatant was
removed from each sample and each set of remaining beads was
washed for 5 minutes with 850 µl 85% ethanol. After removal of
ethanol, the beads were allowed to air dry briey. RNA was eluted
from beads with 10 µl of nuclease-free water and added to the
2-step RT/LAMP cartridges as described above.
Clinical Validation for 2-Step RT/LAMP
In order to validate our detection cartridges with true clinical
samples from patients diagnosed for the presence of absence
of SARS-CoV-2 using established RT-PCR methods we shipped
our v1 2-step RT/LAMP cartridges, at ambient temperature, to
locations willing to share clinical samples. At the Hôpital Saint
Louis (Paris, FR), 18 samples, determined to be positive with RT-
PCR methods approved by local agencies, (Ct values ranging
between 13 and 37.9) were analyzed alongside a negative water
control and Coronavirus NL63 for cross-reactivity. At the Ponticia
Universidad Católica de Chile (Santiago, CL), 10 negative patient
samples and 19 samples, determined to be positive with RT-PCR
using the N1 and N2 primer set designed by the CDC, (Ct values
ranging between 20.25 and 35.34) were analyzed. These clinical
samples were obtained from anonymous patients that attended
the outpatient service of Red Salud UC-CHRISTUS (Santiago,
Chile). All methods were performed in accordance with the relevant
guidelines and regulations. All procedures were approved by the
Ethics Committee of the Ponticia Universidad Católica de Chile.
Results
We have developed a streamlined process for detecting
nucleic acids of RNA viruses, demonstrated as a proof-of-principle
here for the detection of SARS-CoV-2 from nasal mid-turbinate
swabs. Our system incorporates a room temperature stable lysis
buffer (a modied HUDSON buffer or Proteinase K) for the lysis
of virions as well inactivation of RNAses, followed by bead-based
sample concentration in which the beads are captured with a
magnetic wand developed for pipette-free testing. The RT-LAMP
reagents are packaged into a reaction vessel constructed with
layers of waxes in-order to increase shelf-life at room temperature,
provide physical stability for a just-add-sample operation as well as
to minimize risk of cross-contamination.
We were able to achieve a substantial improvement in the
LoD of the rst RT-LAMP primer set published [8] for SARS-CoV-2
by separating the RT reaction from the LAMP reaction using waxes
of different temperatures (2-step cartridges) (Figure 2d and 2e). To
facilitate simpler production and usage of cartridges we ultimately
switched to a 1-step design (Figure 1) using another published
primer set [10] with improved LoD.
Lysis
We used Proteinase K to lyse NMT swabs from 11 distinct
individuals (asymptomatic and presumed to be uninfected with
SARS-CoV-2) spiked with inactivated SARS-CoV-2 ranging from 250
to 1,000 PFU. Lysed samples were concentrated using RNAClean
XP paramagnetic beads and SARS-CoV-2 RNA detected using
a master mix containing the NM primer set. All spiked samples
presented uorescence associated with amplication while non-
spiked samples did not (Figure 4a).
In order to reduce cost and eliminate the 55°C incubation
step of Proteinase K we used the modied HUDSON buffer made
of TCEP and EDTA [15]. NMT swabs were deposited into tubes
with 300 µl 1X HUDSON buffer and spiked with inactivated SARS-
CoV-2 at 0, 125, 250 and 750 PFU. After a brief vortex, samples
were incubated for 5 min at 95ºC. We concentrated these samples
with paramagnetic beads and deposited 10 µl of a 20 µl suspension
of beads in water into a one-step cartridge containing the NM
primer set. All spiked samples presented uorescence associated
with amplication while non-spiked samples did not (Figure 4b).
Magnetic Wand Sample Deposition
In order to produce a streamlined extraction method
compatible with an at-home test we used our 3D-printed magnetic
wand to directly deposit paramagnetic beads into RT-LAMP
reaction cartridges. In its simplest conguration, the magnetic
wand consists of a cylindrical neodymium rare-earth magnet that
slides into a 3D-printed sheath closed on one side. A typical bead
pulldown using a conventional magnetic rack or plate incorporates
a 5-minute incubation with the paramagnetic beads. To make
the magnetic wand more user-friendly with such incubations,
we designed a version of the magnetic wand that incorporates a
ange that rests on the rim of the sample tube and allows the tip
of the wand to be optimally positioned in the sample. The general
workow for magnetic wand use is shown in Figure 3.
To demonstrate the magnetic wand, we used it to transfer
RNA-bound paramagnetic beads into a one-step reaction cartridge
with 10 µl of molecular biology grade water at the top to elute the
beads into. NMT swabs in Proteinase K were spiked with 125 to
1,000 PFU of inactivated SARS-CoV-2 in duplicate and processed
according to our Proteinase K extraction and bead purication
protocol. Paramagnetic beads were recovered with our wands and
directly deposited into the 10µl of water after a brief dip in 100 µl
85% ethanol. All concentrations of virus down to 250 PFU gave
a bright signal while one of the two repeats of 125pfu appeared
Figure 3. Sample concentration with magnetic wand and paramagnetic wand. The paramagnetic beads bind to viral RNA in the lysed sample. The magnetic
wand is used (1) to collect these RNA-bound beads away from the rest of the sample volume and (2) to transfer them to the LAMP reaction cartridge. To
release the beads from the magnetic wand, the magnet is slid from the 3D-printed sheath. In the absence of a magnetic eld, the paramagnetic beads are
no longer magnetized and are able go into solution in water at the top of the LAMP reaction cartridge.
5
. CC-BY-NC-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 preprint this version posted February 20, 2021. ; https://doi.org/10.1101/2021.02.18.21251793doi: medRxiv preprint