JOURNAL OF CLINICAL MICROBIOLOGY, May 2007, p. 1581–1587
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
Vol. 45, No. 5
Development of a Loop-Mediated Isothermal Amplification Assay
for Rapid Detection of BK Virus?
Bipin Raj Bista,1Chandra Ishwad,1Robert M. Wadowsky,2Pradip Manna,3Parmjeet Singh Randhawa,4
Gaurav Gupta,1Meena Adhikari,5Rakhi Tyagi,5Gina Gasper,1and Abhay Vats1*
Department of Pediatrics, Children’s Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, Pennsylvania 152131; Department of Pathology,
Children’s Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, Pennsylvania 152132; Department of Molecular Diagnostics,
ViraCor Laboratories, 1210 NE Windsor Drive, Lee’s Summit, Missouri 640863; Department of Pathology, University of
Pittsburgh Medical Center, 200 Lothrop Street, Pittsburgh, Pennsylvania 152134; and Department of
Molecular Virology, Atharva BioSciences, New Delhi 110029, India5
Received 16 May 2006/Returned for modification 6 July 2006/Accepted 14 December 2006
Loop-mediated isothermal amplification (LAMP) is a novel method for rapid amplification of DNA. Its advan-
tages include rapidity and minimal equipment requirement. The LAMP assay was developed for BK virus (BKV),
which is a leading cause of morbidity in renal transplant recipients. The characteristics of the assay, including its
specificity and sensitivity, were evaluated. BKV LAMP was performed using various incubation times with a variety
of specimens, including unprocessed urine and plasma samples. A ladder pattern on gel electrophoresis, typical of
successful LAMP reactions, was observed specifically only for BKV and not for other viruses. The sensitivity of the
assay with 1 h of incubation was 100 copies/tube of a cloned BKV fragment. Additionally, a positive reaction was
visually ascertained by a simple color reaction using SYBR green dye. BKV LAMP was also successful for urine and
plasma specimens without the need for DNA extraction. Due to its simplicity and specificity, the LAMP assay can
potentially be developed for “point of care” screening of BKV.
BK virus (BKV) can cause renal allograft nephropathy
(BKVAN), which is now recognized as one of the major man-
ifestations of the various syndromes associated with polyoma-
viruses (4–6). Currently, 1 to 10% of renal transplant recipients
are diagnosed with BKVAN, which can lead to graft loss in 20
to 80% of patients (3, 7, 13, 19). Recent studies have also
suggested a viruria prevalence of approximately 30% and a
viremia prevalence of 7 to 10% in patients with renal trans-
plants (2, 4–6, 16, 19). Over the last several years, we devel-
oped and brought into clinical practice a quantitative real-time
(TaqMan) PCR assay for BKV (16, 19). We have also shown
that screening by monitoring of viral load in urine can predict
patients at risk for development of BKVAN (16). The corner-
stones of BKVAN management are an early diagnosis followed
by a reduction in immunosuppression with or without antiviral
therapy (4–7, 19). Thus, different PCR techniques have now
become a valuable tool for screening and monitoring for active
BKV infection (7, 16, 19). However, PCR assays require con-
siderable operator skill and expensive equipment. Rapid diag-
nostic tests that can be performed in the outpatient clinic
without the need for complex procedures (like DNA extrac-
tion) or expensive equipment can potentially improve our ca-
pability to screen for this important transplant complication.
Recently, a novel, specific, and rapid technique for amplifi-
cation of DNA under isothermal conditions was reported (14).
The technique, termed loop-mediated isothermal amplifica-
tion (LAMP), requires a set of conditions and primers different
from those used for PCRs (8, 12, 14, 15). LAMP assays have
now been reported for several pathogens (8, 9, 12, 14, 15, 18,
20). The reaction typically occurs over 30 to 60 min under iso-
thermal conditions with temperatures ranging from 60 to 65°C
and can be conducted with a simple heating block. Thus, the
thermal-cycling needs of a PCR are avoided. In this study, we
developed and characterized LAMP primers and amplification
protocols for a BKV assay (U.S. patents pending). We report the
details of this new technique and the assay characteristics.
MATERIALS AND METHODS
Patient samples. Urine and blood samples obtained from renal transplanta-
tion patients and stored at ?80°C were utilized for development and testing of
BKV LAMP assays and for comparison with PCR assays. The patients signed an
informed consent, and the studies were approved by the Internal Review Board
of the University of Pittsburgh Medical Center and the Children’s Hospital of
Pittsburgh (CHP). Anonymous samples were obtained from ViraCor Laborato-
ries and the microbiology laboratories of CHP for various viruses.
Viral DNA extraction. Viral DNA was extracted from blood samples by using
a QIAmp DNA mini kit (QIAGEN Inc., CA). A QIAmp viral RNA kit was used
to extract viral DNA from urine samples, as we have found that this method
provides a better yield for urinary viral DNA extraction than the use of a DNA
kit (16, 17). The DNA was finally eluted in a 60-?l volume of elution buffer and
stored at ?20°C until further use.
LAMP reaction. The LAMP reaction was conducted, following methods de-
scribed by Notomi et al. and Nagamine et al. with minor modifications (12, 14).
Primers specific for BKV were designed and synthesized (Integrated DNA Tech-
nologies, IA). The six primers consisted of one pair each of forward and reverse
outer primers (BKVTF3 and BKVTB3), forward and reverse inner primers
(BKVTFIP and BKVTBIP), and forward and reverse loop primers (BKVTloopF
and BKVTloopB). The LAMP reaction mixtures (final volume, 25 ?l) contained
these three primer pairs in a 1:8:4 ratio, i.e., 0.2 ?M each of the outer primers,
1.6 ?M each of the inner primers, and 0.8 ?M each of the loop primers. The
other components of the reaction mixture were 2.5 ?l of 10? Bst DNA poly-
merase reaction buffer (New England Biolabs, MA), 1 ?l of 8U/?l Bst DNA
polymerase (New England Biolabs Inc., MA), 2 mM of MgSO4(2 ?l), 5 ?l of
Betaine (Sigma-Aldrich, MO), and 5 ?l of a target sample. The Bst DNA
polymerase reaction buffer (1?) contained 20 mM Tris-HCl (pH 8.8, 25°C), 10
* Corresponding author. Mailing address: Department of Pediatrics,
Children’s Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA
15213. Phone: (412) 692-5182. Fax: (412) 692-7443. E-mail: abhay.vats
?Published ahead of print on 21 February 2007.
mM KCl, 10 mM (NH2)SO4, 2 mM MgSO4, and 0.1% Triton X-100. The
reaction mixtures were incubated at a temperature of 63°C for various periods of
time, ranging from 30 to 120 min.
LAMP product detection. As part of the assay development, the LAMP prod-
ucts were initially detected by several methods, including electrophoresis using
agarose gel (1.2%) with UV light transillumination and photography using the
KODAK EDAS gel documentation system (Eastman Kodak Company, Roch-
ester, NY). A portable UV transilluminator was also used (FastGel system;
Cambrex BioSciences Inc., Rockland, ME) for LAMP product visualization. In
addition, the products were detected visually by a color change after addition of
SYBR green I dye to tubes containing LAMP products. Ten microliters of SYBR
green I dye (100?) (Cambrex BioSciences Inc., Rockland, ME) was added to
each tube containing LAMP products. The LAMP products were also detected
spectrophotometrically (SmartSpec 3000; Bio-Rad Laboratories, CA) at wave-
lengths of 494 to 521 nm. After the correlation was established between various
detection methods, the SYBR green dye method was used for evaluation of
Amplification targets and LAMP equipment. BKV DNA was used to stan-
dardize the LAMP reaction and to determine the specificity of the assay. DNA
extracted from cytomegalovirus (CMV), Epstein-Barr virus (EBV), human her-
pesvirus 6 (HHV-6), adenovirus (Ad), herpes simplex virus 1 and 2 (HSV-1 and
HSV-2), varicella-zoster virus (VZV), and other polyomaviruses (simian virus 40
[SV40] and JC virus [JCV]) was used to determine the specificity of the BKV
LAMP reaction. CMV-, EBV-, HHV-6-, Ad-, HSV-1-, HSV-2-, VZV-, and
SV-40-positive patient blood or urine samples were obtained from the microbi-
ology laboratories of CHP and ViraCor. JCV was obtained from the American
Type Culture Collection (ATCC number 45027). A plasmid containing a cloned
target sequence of BKV (pBKVT3) was used to determine the sensitivity of the
reaction. The negative controls used for various reactions included no-template
controls and no-enzyme controls. The LAMP reaction was primarily carried out
with a heating block (LAB-LINE; Barnstead International, Dubuque, IA). The
reaction was also performed with conventional PCR (MyCycler; Bio-Rad Lab-
oratories, CA) and real-time PCR (Mini-Opticon; Bio-Rad Laboratories, CA)
Cloning of BKV DNA. To determine the sensitivity of the BKV LAMP assay,
we generated and quantitated a plasmid containing the target BKV sequence
(pBKVT3). Briefly, a 217-bp target DNA sequence spanning the region from
nucleotide position 4676 to nucleotide 4893 of the BKV genome was amplified
by PCR using the same outer primers as those used in the LAMP reaction
(BKVTF3 and BKVTB3). The amplified product was then cloned into a TOPO
cloning vector, using a TA cloning kit according to the manufacturer’s instruc-
tions (Invitrogen, CA). The vector was used to transform XL1-Blue competent
Escherichia coli cells (Stratagene, CA). The transformed cells were incubated
overnight and the colonies with the insert (using blue-white distinction) further
grown. The cloned insert was isolated from the cells by use of a Fast Plasmid mini
kit (Eppendorf, NY). The presence of the positive clone was tested by digestion
of the plasmid DNA by EcoRI, followed by gel electrophoresis and sequencing.
The pBKVT3 clone was quantitated using UV spectrophotometry at 260 nm
(SmartSpec 3000; Bio-Rad Laboratories, CA). A series of 10? dilutions, span-
ning 10 to 10E7 copies/tube of the clone, was used to test the sensitivity of the
Real-time PCR. The real-time PCR technique was used to quantify the BKV
DNA as previously described (16, 19) and to compare the sensitivities of the
LAMP and PCR assays. PCRs were performed using the Takara PCR protocol
(TakaraMirus Bio Inc., WI). The reaction mixtures (20 ?l) contained 10 ?l of
Takara master mix, 0.67 ?l each of 10 mM forward and reverse primers, 0.67 ?l
of 10 mM probe, and 5 ?l of the sample. Thermal cycling and quantitation were
performed using a Bio-Rad Mini-Opticon machine (Bio-Rad Laboratories, CA).
The reaction mixture was cycled as follows: 95°C for 30 seconds and 40 cycles of
two-step PCR, with each cycle consisting of 95°C for 1 s, followed by 60°C for 5 s,
with plate reading after each cycle. The primers and probe used for the real-time
PCR are located in the cloned plasmid (pBKVT3), and the sequences were as
follows: forward, GGACCCACCATTGCAGAGTTT; reverse, AGAGCCCT
TGGTTTGGATAGATT; probe, 6-FAM (6-carboxyfluorescein)-5?-AAGC
CAAACCACTGTGTGAAGCAGTCAAT 3?-TAMRA (6-carboxytetramethyl-
rhodamine). The real-time measurement of BK viral load in clinical specimens
also utilized an internal control. Each specimen was spiked with a universal
internal control (UIC) prior to DNA extraction. The UIC contains a partial gene
sequence for green fluorescence protein that is absent in viral-pathogen and
mammalian genomes. To assess for PCR inhibition, a separate real-time PCR
assay was performed with UIC-specific primers and probe. Real-time PCR in-
hibition was defined by the absence of UIC target amplification.
Assay design. The LAMP primers were selected from a
conserved region of the BKV genome as determined by whole-
genome sequencing of several viral strains obtained from dif-
ferent BKV-positive patient samples and are shown in Table 1.
The locations and sequences of the different primers are shown
in Fig. 1. All primers were designed by manual inspection of
TABLE 1. LAMP primers
Primer used for BKV LAMP Sequence No. of bases
Outer forward and backward
GGACCCACCATTGCAGAGTTT (outer forward primer)
TCTTTGCCCAGATACCCTGT (outer primer)
Inner forward and backwarda
BKVTBIP (B1 ? B2c)
BKVTFIP (F1c ? F2)
Loop forward and backward
aThe F1c and B2c sequences are complementary and reverse to the F1 and B2 regions shown in Fig. 1 and are shown as underlined. The BKVTBIP primer had two
degenerate bases in three positions (W ? A/T and Y ? C/T).
FIG. 1. Segment of large T-antigen gene sequence (nucleotide po-
sitions 4676 to 4893) showing the locations and orientations of various
primers used in the BKV LAMP assay.
1582BISTA ET AL.J. CLIN. MICROBIOL.
the sequence data. Briefly, the outer primers (F3 [BKVTF3]
and B3 [BKVTB3]) were located outside two inner primers
(labeled F2 [BKVTF2] and B2 [BKVTB2]) in the large-T-
antigen gene region of the BKV genome. The forward internal
primer (BKVTFIP) comprises a fusion primer with the com-
plementary F1 sequence (F1c) and the direct F2 sequences
(14). Similarly, the reverse internal primer (BKVTBIP) com-
prises the direct B1 sequence and the complementary B2 se-
quence (B2c) (Table 1). To increase the amplification effi-
ciency, two loop primers (forward [BKVTloopF] and reverse
[BKVTloopB]) were also designed. The loop primers signifi-
cantly improved the performance of the assay and led to a
considerable reduction in the time required for amplification
of the target. The LAMP reaction was carried out at a constant
temperature of 63°C and generally was positive after various
periods of time (30 to 120 min) with the addition of loop
primers to the reaction mixture. Without the use of loop prim-
ers, the same reactions required periods 4 to 5 times longer
(i.e., a 30-min reaction required over 2 or 3 h). Also, several
samples with lower viral loads could not be amplified at all
without the addition of loop primers.
Assay specificity and sensitivity. A positive BKV LAMP
reaction typically required incubation for 60 min and revealed
a ladder pattern with many bands of different sizes on agarose
gel electrophoresis (Fig. 2A, BKV lane). The specificity of the
BKV LAMP assay was evaluated by testing the assay on several
DNA viruses that can be seen in transplant patients (i.e., EBV,
CMV, HHV-6, Ad, HSV-1, HSV-2, and VZV) and the two
related polyomaviruses (i.e., JCV and SV40). No ladder pat-
tern was seen with the other (non-BK) viruses or for the neg-
ative control (Fig. 2A and B). The sensitivity of the LAMP
reaction was determined using 10? serial dilutions of the
pBKVT3 clone. The BKV LAMP sensitivity was found to be
100 copies per tube (Fig. 2C, lane 7) when an incubation
period of 60 min was employed.
Detection of LAMP products by alternate methods. Besides
gel electrophoresis, we also determined two additional meth-
ods, i.e., visual inspection and spectrophotometry, to detect a
positive reaction. Upon addition of the SYBR green I dye to
tubes after the LAMP reaction, the color changed to yellowish
green in a positive reaction (Fig. 3A, tube 1) and remained
reddish orange (the color of the unbound dye) in the negative
reactions (Fig. 3A, tube 2). Although the intensities of the
color changes appeared to be similar for a range of viral loads,
the color changes occurred earlier in samples with higher
loads. All the samples that were positive by gel electrophoresis
were also positive by visual detection of color change (and vice
versa). The visual detection of a positive reaction was further
improved by using UV light from a conventional as well as a
portable blue light transilluminator, which demonstrated a
bright green fluorescence in positive reactions (Fig. 3B and C).
Finally, the positive reaction was also detected by a spectro-
photometer in a quantitative manner.
LAMP assay with clinical samples. Detection of BKV DNA
by LAMP was done with 49 clinical blood and urine samples
obtained from renal transplant patients. Real-time PCR assays
for BKV were also done with the same samples for compari-
son. As there was a good correlation between various detection
methods, positive-reaction detection with clinical samples was
performed by visual observance of color change with SYBR
green and UV transillumination. A summary of the results and
the comparison between real-time PCR and LAMP is shown in
Table 2. Plasma samples required approaches slightly different
from those used for urine samples for LAMP to work success-
fully as described below.
Plasma samples. We analyzed 21 plasma samples from 18
different patients, and four of the plasma samples belonged to
a single BKVAN patient (no. 18). These four samples were
obtained over a period of 3 months and showed loads ranging
from 2,264 to 333,200 copies/ml. LAMP was performed with
extracted as well as unextracted plasma samples under two
separate conditions, i.e., after heating at 99°C for 10 min and
without heating (i.e., samples freshly obtained without any
processing). The unprocessed and unextracted plasma samples
did not give positive reactions with LAMP. However, the same
samples were successfully amplified by LAMP after the plasma
was heated. Approximately 15 to 20 ?l of supernatant was
obtained after 50 ?l of plasma was heated, of which 5 ?l was
used for each reaction (Fig. 4A). The LAMP assay was nega-
tive for samples that were negative by real-time PCR and also for
three samples with viral loads of ?1,500 copies/ml (Table 2). The
samples with viral loads of ?5,000 copies/ml required the LAMP
reaction to be performed for a longer period (120 min).
FIG. 2. Specificity and sensitivity of the BKV LAMP assay. (A) The
specificity of the BKV LAMP assay is shown in the positive reaction,
visible as a ladder-like pattern in lanes with BKV DNA only. A neg-
ative reaction was obtained when the target of amplification was other
non-BKV DNA viruses, i.e., EBV, CMV, Ad, HHV-6, HSV-1, HSV-2,
VZV, or no-target control (NTC). (B) The LAMP reaction was neg-
ative for two related polyomaviruses, JCV and SV40, and amplified
only BKV. (C) The BKV LAMP reaction was carried out with the
serial dilutions of cloned BKV plasmid (pBKVT3), and the sensitivity
of the reaction was determined to be 100 copies per tube.
VOL. 45, 2007 LAMP ASSAY FOR BK VIRUS1583
Urine samples. LAMP was performed with 28 urine samples
from 27 patients, including 8 renal transplant recipients with
asymptomatic viruria, 14 patients with biopsy-confirmed
BKVAN, and 5 patients with no viruria. We initially analyzed
the urine samples with and without (i.e., freshly obtained with-
out any processing) DNA extraction from 10 patient samples
(7 viruria and 3 BKVAN). Unlike for plasma, BKV LAMP
performed successfully even when freshly voided (unprocessed)
urine was added directly to the reaction mixture (Fig. 4B). The
reactions were positive for unprocessed urine samples from
four patients whose urine viral loads were ?50,000 copies per
ml (Fig. 4B). Interestingly, the assay was negative for two
samples (Fig. 4B, lanes 1 and 3) with viral loads of approxi-
mately 60,000 and 40,000 copies/ml even after incubation for
up to 2 h. All 14 BKVAN patients were positive by the LAMP
assay performed with unprocessed urine, including a patient
with a low viral load of 30,500 copies per ml (Table 2). These
results show that the sensitivity of LAMP is approximately
100,000 copies/ml for urinary samples, with variable perfor-
mance at lower urinary viral loads.
In recent years, the detection of BK viremia and viruria by
real-time PCR has become an important tool for identification
of patients at risk for developing BKVAN (4, 6, 7, 19). Despite
TABLE 2. Comparison of LAMP and real-time PCR results for
clinical plasma and urine samplesa
(no. of copies/ml)
aAsympt viruria, asymptomatic viruria; neg, negative; pos, positive.
FIG. 3. Visual detection of LAMP products. (A) The BKV LAMP
products were detected after addition of SYBR green I dye. The tube
with a positive reaction (tube 1) shows a color change to yellowish
green, which can be distinguished from the reddish orange color of a
negative reaction. (B) The same positive sample, when visualized un-
der UV transillumination, shows a bright green fluorescence in tube 1
(positive reaction) compared to what is observed for tube 2 (negative
reaction). (C) A portable blue light illuminator that can be used in a
clinic setting for transillumination of LAMP products. A 0.5-ml PCR
tube is placed on it to show the relative size of the transilluminator.
1584BISTA ET AL.J. CLIN. MICROBIOL.
the clinical utility of PCR-based techniques, they have the
inherent disadvantage of being time-consuming as well as re-
quiring considerable operator skill and expensive equipment.
We demonstrate that the LAMP technique can be used to
amplify BKV DNA under isothermal conditions with high
specificity and efficiency. It has a minimal equipment require-
ment and can be accomplished in 1 hour or less. Moreover, we
have also shown that the LAMP reaction can be performed
even with unprocessed urine samples and minimally processed
(i.e., heat-treated) plasma samples without the need for DNA
extraction. Thus, this technique can potentially be developed
to screen for virus in “point of care” settings, such as outpa-
The LAMP method is based on autocycling strand displace-
ment DNA synthesis performed by the Bst DNA polymerase
enzyme and was first described by Notomi et al. (14). The
principle underlying LAMP is complex, and excellent details of
the technique are available in two recent publications (12, 14).
Briefly, the LAMP assay requires isothermal strand displace-
ment and the amplicons are stem-loop DNA structures with
several inverted repeats of the target. The amplicons are
formed after multiple rounds of DNA amplification and have
structures with multiple loops. Typically, LAMP utilizes four
specifically designed primers (two pairs of outer and inner
primers) and isothermal Bst DNA polymerase, which has
strand displacement activity. Nagamine et al. reported a fur-
ther improvement by introducing another set of primers, called
loop primers, in the reaction (12). Loop primers hybridize to
the stem-loops that are formed in the initial phases of LAMP
and act by providing additional priming for the DNA polymer-
ase for strand displacement (12). The use of these additional
two primers (a total of six primers) further improves the am-
plification efficiency of the reaction, as originally reported by
Nagamine et al. and also confirmed in our studies. The ampli-
cons are typically seen as multiple bands on agarose gel elec-
trophoresis, in contrast to the single band seen with PCR.
An important advantage of LAMP is that the thermal-cy-
cling needs of the PCR method are completely avoided (12).
Also, the LAMP assay produces a large amount of amplified
product, resulting in easier detection by visual inspection, by
observation of either an increase in turbidity caused by gener-
ation of magnesium pyrophosphate or a color change after
addition of SYBR green I dye (10, 11, 12, 14). The change in
turbidity can also be measured quantitatively and in real-time
mode by using a real-time turbidimeter (8, 9, 15). However, the
increase in turbidity can be difficult to appreciate and we found
it difficult to detect a change in turbidity visually in our exper-
iments. However, visual detection of a red-to-green color
change with SYBR green I was easily accomplished and always
correlated with gel electrophoresis findings. Our ability to vi-
sually detect a positive reaction was significantly improved by
using UV transilluminators, including a portable blue light
transilluminator that recently became available and can be
useful in developing the assay for “point-of-care applications”
settings (Fig. 3C). One relative limitation of the LAMP assay is
that the color change appears at a visual level to be an all-or-
none phenomenon (i.e., a “yes” or a “no”) and did not behave
in a semiquantitative manner, as the various shades of green
for different viral loads were not easily visually distinguishable.
Although we did not intend to design a quantitative or semi-
FIG. 4. BKV LAMP with plasma and urine samples from renal trans-
plant patients. (A) Four plasma samples from a single patient with dif-
ferent viral loads assayed with and without heating of samples. The
LAMP reaction was negative in unprocessed and unheated plasma sam-
ples (lanes 1 to 4) and positive in the plasma only after heat treatment of
the same samples for 10 min (lanes 5 to 8) or after DNA extraction
(CTRL, lane 9). Lane 1, 2,264 copies/ml; lane 2, 333,200 copies/ml; lane
3, 50,480 copies/ml; lane 4, 157,200 copies/ml. Lanes 5 to 8 show the same
four samples after heat treatment. Lane 5, 2,264 copies/ml (heated
plasma); lane 6, 333,200 copies/ml (heated plasma); lane 7, 50,480
copies/ml (heated plasma); lane 8, 157,200 copies/ml (heated plasma);
lane 9, positive control with extracted DNA (333,200 copies/ml).
(B) The LAMP reaction was carried out with unextracted, freshly
voided urine samples from 10 renal transplant patients. The corre-
sponding BKV DNA copies were as follows: lane 1, 58,000 copies/ml;
lane 2, 6,000 copies/ml; lane 3, 40,000 copies/ml; lane 4, 3.4E?9 copies/
ml; lane 5, 1.7E?9 copies/ml; lane 6, 54,000 copies/ml; lane 7, 110,000
copies/ml; lane 8, 390,000 copies/ml; lane 9, 52,000 copies/ml; lane 10,
20,000 copies/ml. (C) The BKV LAMP reaction was carried out with
DNA extracted from the same urine samples as those shown in panel
B. Note that the reaction was now strongly positive in lane 7 and faintly
positive in lane 10, in addition to lanes 4, 5, 6, and 9 in panel B.
VOL. 45, 2007LAMP ASSAY FOR BK VIRUS1585
quantitative LAMP assay, the time required to develop a color
change with SYBR green can possibly be used to get a semi-
quantitative idea of the viral load. The color change occurs
earlier at higher viral loads (i.e., ?30 min incubation for viral
loads of ?10E7 copies) and takes longer for lower loads (60
min for 10,000 to 100,000 copies/ml).
The LAMP assay is more complex to design than PCR
assays. However, once developed, LAMP is much easier to
perform than PCR or real-time PCR and requires a simple
addition of unprocessed urine or plasma to a tube containing a
mixture of primers, enzyme, and buffers, followed by warming
for 30 to 60 min at 60 to 65°C. Hence, the hands-on time of
approximately 1 h or less for LAMP can be comparable to or
even significantly shorter than that required for PCR or real-
time PCR. Another advantage of LAMP is that, in contrast to
PCR, it requires no DNA extraction, which can be a major
obstacle in widespread employment of DNA amplification as-
says. Our LAMP assay performed successfully without DNA
extraction for both urine and plasma samples. Unprocessed,
freshly voided urine is considered to contain PCR inhibitors
but was successfully tested with BKV LAMP assays. Although
there was a loss of some sensitivity with the use of unprocessed
samples (versus what was found for extracted DNA samples),
the assay could still be very useful in clinical situations. The
reasons for the variable performance of the assay for viral
loads of less than 100,000 copies/ml of urine are not clear but
may be multifactorial. One possibility is that the LAMP assay
demands more integrity of the target, owing to the many prim-
ers that need to be annealed to the different regions, than
PCR. Since the LAMP assays were tested on stored samples, it
is possible that there may have been partial degradation of the
target DNA in the region of LAMP primers but not in the
region that is amplified by PCR primers. Another possibility is
the presence of urine-specific inhibitors of LAMP reaction.
The patients with false-negative urine LAMP results did not
have any obvious clinical differences from other patients and
also did not have any proteinuria, which can possibly act as an
inhibitor. However, various drugs, metabolites, chemicals, and
dietary agents are excreted in urine, and some of these may
affect urinary LAMP more than plasma LAMP assays. Further
refinements of technique are clearly needed to improve the
sensitivity and the semiquantitative aspects of the LAMP as-
says as well as to address the issues related to as-yet-unrecog-
nized urinary “inhibitors.” Also, additional studies, including
validation of visual testing with a larger number of clinical
samples and proficiency testing, will be required before these
assays can be employed in routine clinical laboratories.
BKVAN is emerging as an important cause of renal allograft
dysfunction and can resemble acute rejection (AR) both clin-
ically and histologically (3, 6, 7, 13). Differentiation between
BKVAN and AR is important, however, because the treat-
ments for the two conditions are diametrically opposite. Sev-
eral studies have shown that BK viruria is a requirement for
development of BKVAN and generally precedes the appear-
ance of viremia and full-blown nephropathy by several weeks
to months. BKV DNA detection in urine by PCR is also much
more sensitive than viral detection by urinary cytology (decoy
cells) and also viral detection by immunohistochemistry or in
situ hybridization on allograft biopsy (2, 3, 16). However,
asymptomatic BK viruria can be seen in up to 30% of kidney
transplant patients and only approximately 6 to 10% of these
patients develop full-blown BKVAN (13, 19). The presence of
a high viral load is now considered to be more correlated with
risk for development of BKVAN than just the presence of virus
in urine (16, 17). Our previous studies have shown that viral
loads of ?10E7 copies/ml in urine and ?10,000 copies/ml in
blood are reasonable threshold loads above which a renal
transplant patient is more likely to have or develop BKVAN
(16, 17, 19). BKV LAMP showed consistent performance in
our studies at urine loads of ?10E5 copies/ml and for ?2,000
copies/ml of plasma even when performed with unextracted
samples. These detection limits are well below the viral loads
that are considered to be clinically significant. Hence, this assay
could easily be used in point-of-care settings to detect high-risk
patients with viral loads above the critical levels. Demonstra-
tion of BKV in urine and blood can alert a clinician to the
possibility of BKVAN in a patient presenting an asymptomatic
rise in serum creatinine, which is most often attributed to AR
in renal transplant recipients.
In summary, we report our early experience with the devel-
opment and performance characteristics of a BKV LAMP as-
say. It is a novel technique that can possibly be used for rapid
diagnosis of BKV infection not only in laboratories but also in
an outpatient clinic setting. A reduction in immunosuppression
is now considered the first-line therapy for BKV infections in
renal transplant recipients (1, 4, 7). However, physicians cur-
rently depend on PCR results for BKV detection, which might
take 2 to 3 days or longer to arrive in many clinic settings. Thus,
the LAMP assay for BKV can potentially be used to guide
therapeutic decisions in outpatient clinics itself, especially if a
urine sample becomes positive after a short, 30-min incubation
and is accompanied by a plasma sample that shows positivity,
suggesting that the patient is more likely to have a clinically
significant viral load. Further large-scale studies for determi-
nation of the sensitivity, specificity, and clinical utility of this
new method will be needed before this method can find wider
This work was supported by an NIH award (R01 AI060602) to A.V.
1. Brennan, D. C., I. Agha, D. L. Bohl, M. A. Schnitzler, K. L. Hardinger, M.
Lockwood, et al. 2005. Incidence of BK with tacrolimus versus cyclosporine
and impact of preemptive immunosuppression reduction. Am. J. Transplant.
2. Drachenberg, C. B., J. C. Papadimitriou, H. H. Hirsch, R. Wali, C. Crowder,
J. Nogueira, et al. 2004. Histological patterns of polyomavirus nephropathy:
correlation with graft outcome and viral load. Am. J. Transplant. 4:2082–
3. Drachenberg, C. B., H. H. Hirsch, E. Ramos, and J. C. Papadimitriou. 2005.
Polyomavirus disease in renal transplantation: review of pathological find-
ings and diagnostic methods. Hum. Pathol. 36:1245–1255.
4. Hariharan, S. 2006. BK virus nephritis after renal transplantation. Kidney
5. Hirsch, H. H. 2005. BK virus: opportunity makes a pathogen. Clin. Infect.
6. Hirsch, H. H., and J. Steiger. 2003. Polyomavirus BK. Lancet Infect. Dis.
7. Hirsch, H. H., D. C. Brennan, C. B. Drachenberg, F. Ginevri, J. Gordon,
A. P. Limaye, et al. 2005. Polyomavirus-associated nephropathy in renal
transplantation: interdisciplinary analyses and recommendations. Transplan-
8. Hong, T. C., Q. L. Mai, D. V. Cuong, M. Parida, H. Minekawa, T. Notomi,
et al. 2004. Development and evaluation of a novel loop-mediated isother-
mal amplification method for rapid detection of severe acute respiratory
syndrome coronavirus. J. Clin. Microbiol. 42:1956–1961.
1586BISTA ET AL. J. CLIN. MICROBIOL.
9. Ihira, M., T. Yoshikawa, Y. Enomoto, S. Akimoto, M. Ohashi, S. Suga, et al.
2004. Rapid diagnosis of human herpesvirus 6 infection by a novel DNA
amplification method, loop-mediated isothermal amplification. J. Clin. Mi-
10. Mori, Y., K. Nagamine, N. Tomita, and T. Notomi. 2001. Detection of
loop-mediated isothermal amplification reaction by turbidity derived from
magnesium pyrophosphate formation. Biochem. Biophys. Res. Commun.
11. Mori, Y., M. Kitao, N. Tomita, and T. Notomi. 2004. Real-time turbidimetry
of LAMP reaction for quantifying template DNA. J. Biochem. Biophys.
12. Nagamine, K., T. Hase, and T. Notomi. 2002. Accelerated reaction by loop-
mediated isothermal amplification using loop primers. Mol. Cell. Probes
13. Nickeleit, V., and M. J. Mihatsch. 2004. Polyomavirus allograft nephropathy
and concurrent acute rejection: a diagnostic and therapeutic challenge.
Am. J. Transplant. 4:838–839.
14. Notomi, T., H. Okayama, H. Masubuchi, T. Yonekawa, K. Watanabe, N.
Amino, et al. 2000. Loop-mediated isothermal amplification of DNA.
Nucleic Acids Res. 2:E63.
N. Tangpukdee, K. Y. Yuen, Y. Guan, S. Looareesuwan, and J. S. Peiris. 2006.
Sensitive and inexpensive molecular test for falciparum malaria: detecting Plas-
modium falciparum DNA directly from heat-treated blood by loop-mediated
isothermal amplification. Clin. Chem. 52:303–306.
16. Randhawa, P., A. Ho, R. Shapiro, A. Vats, P. Swalsky, S. Finkelstein, et al.
2004. Correlates of quantitative measurement of BK polyomavirus (BKV)
DNA with clinical course of BKV infection in renal transplant patients.
J. Clin. Microbiol. 42:1176–1180.
17. Randhawa, P., R. Shapiro, and A. Vats. 2005. Quantitation of DNA of
polyomaviruses BK and JC in human kidneys. J. Infect. Dis. 192:504–509.
18. Suzuki, R., T. Yoshikawa, M. Ihira, Y. Enomoto, S. Inagaki, K. Matsumoto,
et al. 2006. Development of the loop-mediated isothermal amplification
method for rapid detection of cytomegalovirus DNA. J. Virol. Methods
19. Vats, A., R. Shapiro, P. Singh Randhawa, V. Scantlebury, A. Tuzuner, M.
Saxena, M. L. Moritz, T. J. Beattie, T. Gonwa, M. D. Green, and D. Ellis.
2003. Quantitative viral load monitoring and cidofovir therapy for the
management of BK virus-associated nephropathy in children and adults.
20. Yoshikawa, T., M. Ihira, S. Akimoto, C. Usui, F. Miyake, S. Suga, et al. 2004.
Detection of human herpesvirus 7 DNA by loop-mediated isothermal am-
plification. J. Clin. Microbiol. 42:1348–1352.
VOL. 45, 2007 LAMP ASSAY FOR BK VIRUS1587