Search tips
Search criteria 


Logo of jcmPermissionsJournals.ASM.orgJournalJCM ArticleJournal InfoAuthorsReviewers
J Clin Microbiol. 2010 July; 48(7): 2330–2336.
Published online 2010 April 26. doi:  10.1128/JCM.01224-09
PMCID: PMC2897471

Development and Evaluation of a Simple Assay for Marburg Virus Detection Using a Reverse Transcription-Loop-Mediated Isothermal Amplification Method[down-pointing small open triangle]


Marburg virus (MARV) causes a severe hemorrhagic fever in humans with a high mortality rate. The rapid and accurate identification of the virus is required to appropriately provide infection control and outbreak management. Here, we developed and evaluated a one-step reverse transcription-loop-mediated isothermal amplification (RT-LAMP) assay for the rapid and simple detection of MARV. By combining two sets of primers specific for the Musoke and Ravn genetic lineages, a multiple RT-LAMP assay detected MARV strains of both lineages, and no cross-reactivity with other hemorrhagic fever viruses (Ebola virus and Lassa virus) was observed. The assay could detect 102 copies of the viral RNA per tube within 40 min by real-time monitoring of the turbidities of the reaction mixtures. The assay was further evaluated using viral RNA extracted from clinical specimens collected in the 2005 Marburg hemorrhagic fever outbreak in Angola and yielded positive results for samples containing MARV at greater than 104 50% tissue culture infective doses/ml, exhibiting 78% (14 of 18 samples positive) consistency with the results of a reverse transcription-PCR assay carried out in the field laboratory. The results obtained by both agarose gel electrophoresis and naked-eye judgment indicated that the RT-LAMP assay developed in this study is an effective tool for the molecular detection of MARV. Furthermore, it seems suitable for use for field diagnostics or in laboratories in areas where MARV is endemic.

Marburg virus (MARV) is the causative agent of a severe hemorrhagic fever in humans with a high mortality rate. After the first documented outbreak of Marburg hemorrhagic fever (MHF) in Germany and the former Yugoslavia in 1967, several sporadic outbreaks have been reported in central African countries (19, 20). The largest outbreak of MHF occurred in Uige Province in Angola from 2004 to 2005 and had a mortality rate of 90% among 252 cases (25). As with Ebola virus (EBOV), transmission of MARV is associated with close contact with infected individuals, particularly their body fluids. Therefore, diagnosis early in the course of an MHF outbreak is essential to control infection and to prevent further transmission (3, 4, 5).

Virus isolation, transmission electron microscopy, immunohistochemistry, antigen-capture enzyme-linked immunosorbent assay (ELISA), IgG or IgM virus-specific antibody-capture ELISA, and reverse transcription-PCR (RT-PCR) have been used for the laboratory diagnosis of MARV (9, 20, 23). As the viral load in the serum of individuals infected with filoviruses could be as high as 109 copies per milliliter (26, 28), molecular detection methods based on viral protein or genome sequences have taken over as the first-choice diagnostic techniques for use with clinical specimens. Both antigen-capture ELISA and real-time RT-PCR have been used to detect MARV in field laboratories during some MHF outbreaks, as these methods can yield a positive result rapidly and specifically, but they require expensive apparatuses and sophisticated techniques.

MARV is a member of the family Filoviridae, along with EBOV, and is a single, nonsegmented, negative-sense RNA virus. The virus genome is almost 19 kb in length and encodes seven viral proteins. In contrast to the genus Ebola virus, which includes four definite species, Zaire Ebola virus (ZEBOV), Sudan Ebola virus (SEBOV), Ebola virus Reston (REBOV), and Ivory Coast Ebola virus (ICEBOV), and a putative species, Bundibugyo Ebola virus (27), the genus Marburg virus consists of a single species, Lake Victoria Marburg virus (LVMARV) (6). Comparative sequence analyses of the GP and VP35 genes or the full-length genome of MARV strains showed that there are two distinct lineages within LVMARV and a difference of approximately 20% between the two lineages at the nucleotide sequence level (2, 21, 25). The Ravn (Kenya, 1987) (12) and 09DRC99 (Democratic Republic of Congo [DRC], 1999) strains (2) comprise a distinct minor lineage (Ravn lineage), and the other strains, including the high-virulence strains isolated in Angola, comprise the major genetic lineage within the genus Marburg virus, represented by the Musoke strain (Kenya, 1980) (Musoke lineage) (22). Thus, the development of MARV detection methods based on nucleic acid amplification and applicable to all known MARV isolates has been limited. Therefore, it is necessary to establish a method for MARV detection based on a rapid and simple molecular detection technique adapted to the sequence variants of each strain.

Reverse transcription-loop-mediated isothermal amplification (RT-LAMP) is a promising technique for nucleic acid amplification (16). The use of this method has recently been demonstrated for the diagnosis of several human pathogenic RNA viruses (11, 14, 18). This technique is based on the principle of strand-displacing DNA synthesis by the Bst DNA polymerase with six distinct primers that recognize a total of eight independent sites. cDNA synthesis by avian myeloblastosis virus reverse transcriptase and DNA amplification were performed in one step under isothermal conditions (60° to 65°C), thereby obviating the need for a thermal cycler. Moreover, LAMP of positive samples could be performed simply with real-time monitoring of the turbidities of the reaction mixtures as well as naked-eye judgment with addition of a fluorescent substance (calcein) to the reaction mixture (24).

In the present study, we developed a MARV-specific RT-LAMP assay which is highly specific for MARV and shows no cross-reactivity with the viral RNA of closely related EBOV strains. The assay could detect 102 RNA molecules per tube. The RT-LAMP assay does not require the use of sophisticated equipment or highly skilled personnel and can provide accurate results within a short time frame. These characteristics make this assay potentially useful for the clinical diagnosis of MARV infection in a field laboratory.


Viruses and RNA extraction.

All virus strains used in this study, Lassa virus (LASV) strains Josiah and Pinneo, ZEBOV strains Mayinga'76 and Kikwit'95, SEBOV strain Boniface, REBOV strain Reston, and ICEBOV strain Cote d'Ivoire, and MARV strains Musoke, Ozolin, Ravn, and Angola, were propagated in Vero E6 cells, as described previously (15). With the exception of the MARV Angola strain, which was isolated from clinical material from the outbreak in Uige, Angola, all viruses were kindly provided to the National Microbiology Laboratory (NML), Public Health Agency of Canada (PHAC), by the Special Pathogens Branch of the Centers for Disease Control and Prevention (at that time, from T. G. Ksiazek and P. E. Rollin) and the Virology Division of the U.S. Army Medical Research Institute of Infectious Diseases (at that time, from P. B. Jahrling and T. W. Geisbert). Viral RNAs were extracted manually from virus suspensions using a QIAamp viral RNA minikit (Qiagen, Hilden, Germany). All infectious materials were handled in the biosafety level 4 facility of NML, PHAC.

Standard transcript RNA.

The 3′ end of the virus genome was amplified by RT-PCR using Musoke-specific forward (5′-AGACACACAAAAACAAGAGA-3′) and reverse (5′-CTTGGATGGG/CGCCAGGCATC-3′) primers and Ravn-specific forward (5′-AGACACACAAAAACAAGAGATGATG-3′) and reverse (5′-CTTTGGACGGGCGC/CAAGCATC-3′) primers. Fragments of the predicted size (~5.1 kb) were cloned into the vector pGEM-3Zf(+) (Promega, Madison, WI), and partial viral genome clones were obtained (pGEM-Mus1 and pGEM-Rav1, respectively). The nucleoprotein (NP)-coding sequence of each strain was amplified by PCR using primers MusNP (5′-ATGGATTTACACAGTTTGTTGGAGTTGGG-3′) and MusNP+ (5′-CTACAAGTTCATCGCAACATGTCTCCTTTC-3′) and the template pGEM-Mus1 and primers RavNP (5′-ATGGATTTACATAGTTTGCTAGAATTAGG-3′) and RavNP+ (5′-TTACAAGTTCATAGCAACATGCCTCCTCTC-3′) and the template pGEM-Rav1. Fragments of the expected size (2,088 bp each) were purified with a gel extraction kit (Qiagen) and subcloned into pGEM-3Zf(+) in inverse orientation relative to the orientation of the T7 promoter sequence. The negative-sense NP RNAs of the Musoke strain (Mus-NP) and the Ravn strain (Rav-NP) were synthesized in vitro using each of the subclones of the NP gene with T7 RNA polymerase (Promega). The transcripts were extracted using an RNeasy minikit (Qiagen) and resuspended in 50 μl of diethyl pyrocarbonate (DEPC)-treated water. The RNA concentration was determined by measuring the optical density at 260 nm (OD260), and the RNA was diluted with DEPC-treated water to achieve the appropriate concentrations.

Primer design.

MARV-specific primers for RT-LAMP were designed based on the sequences of the NP gene, and two sets of lineage-specific primers for Musoke and Ravn were designed. The complete MARV NP-coding sequences available in the GenBank database were aligned using DNA analysis software (GENETYX, Tokyo, Japan) to identify a conserved region and determine a consensus sequence for each lineage (Fig. (Fig.1).1). Potential target regions in the consensus sequence for Musoke lineage were analyzed using the LAMP primer design support software program PrimerExplorer (version 3; Net Laboratory, Tokyo, Japan;, and the Musoke lineage-specific primers were designed automatically. Ravn lineage-specific primers were designed based on the Musoke lineage primers by replacement with nucleotides complementary to the Ravn lineage consensus sequence. The lineage-specific RT-LAMP assay required a set of six primers: two outer primers (F3 and B3), a forward inner primer (FIP), a reverse inner primer (BIP), a forward loop primer (LF), and a reverse loop primer (LB). FIP consisted of F1c complementary to the F1 sequence, a TTTT spacer, and the F2 sequence. BIP consisted of B1c complementary to the B1 sequence, a TTTT spacer, and the B2 sequence. Primer LF was commonly used for the Musoke (Mus) and Ravn (Rav) lineage-specific RT-LAMPs. A multiple RT-LAMP reaction was performed by combining a total of 11 primers specific for each lineage. The locations and sequences of the oligonucleotide primers are shown in Fig. Fig.11 and Table Table1.1. All the primers used were oligonucleotide purification cartridge-purified primers and were purchased from Hokkaido System Science (Sapporo, Japan).

FIG. 1.
Alignment of MARV NP gene sequences and positions of primers for RT-LAMP assay. Primers were designed according to a conserved region of the NP gene (nucleotides 693 to 897, GenBank accession number DQ217792) determined by use of the NP genes of Musoke- ...
LAMP primers for MARV detection

RT-LAMP reaction and restriction enzyme analysis.

The RT-LAMP reaction was performed in 25-μl reaction mixtures with an RNA amplification kit (Eiken Chemical Co., Ltd., Tokyo, Japan), in accordance with the manufacturer's protocol. The reaction mixture contained 1.6 μM each primers FIP and BIP, 0.2 μM each outer primers F3 and B3, 0.8 μM each loop primers LF and LB, and 1 μl of RNA extract (0.2 ng RNA). The reaction mixture was incubated in a heat block at 63°C for 45 or 60 min and was then heated at 80°C for 5 min to terminate the reaction. To distinguish the amplified DNA products of a Musoke-lineage strain from those of a Ravn-lineage strain, aliquots of 1 μl of the products in a 10-μl reaction mixture were digested with 20 units of SmlI at 55°C or 40 units of XbaI at 37°C for 30 min. Aliquots of 1 μl of each amplified product or 10 μl of digested DNA products were analyzed on a 3% agarose gel, followed by staining with ethidium bromide and visualization on an imaging analyzer (LAS-3000; FujiFilm, Tokyo, Japan). For visual detection, 1 μl of fluorescent reagent (Eiken Chemical) was added to the RT-LAMP reaction mixture, and the incubated reaction tube was irradiated with UV light (254 nm) using a handheld irradiator (UVP, San Francisco, CA). For real-time monitoring of RT-LAMP amplification, the reaction mixtures were incubated at 63°C and examined by spectrophotometric analysis using a real-time turbidimeter (LA-200; Teramecs, Kyoto, Japan). A turbidity value of >0.1 was considered a positive result.

RT-PCR and TaqMan RT-PCR.

The RT-PCR assay for filoviruses was performed in 25-μl reaction mixtures containing 1 μl of RNA template using a One-Step RT-PCR kit (Qiagen). The cycling profiles and the primers targeting a region of the L gene described by Zhai et al. (29) were used. The TaqMan RT-PCR assay was performed using a One Step PrimerScript RT-PCR kit (Perfect Real Time; Takara Bio, Shiga, Japan) and a SmartCycler II system (Cepheid, Sunnyvale, CA). Amplifications were carried out in 25-μl reaction mixtures containing 1 μl of the target virus RNA, 0.5 μM each sense and antisense primers, and 0.2 μM TaqMan probe targeting a region of the nucleoprotein, as described by Weidmann et al. (28). The number of RNA copies in viral RNA extracts of each MARV strain was calculated from the standard curve prepared with the in vitro-transcribed RNA standard and the cycle threshold (CT) value obtained with the TaqMan RT-PCR.

Clinical specimens.

A total of 24 clinical specimens from whole blood (9 specimens), serum (3 specimens), oral swabs (11 specimens), and breast milk (1 specimen) were obtained from 16 individuals with suspected MARV infection during the outbreak in Uige, Angola, from 2004 to 2005. The RNAs were extracted directly from the clinical specimens using a QIAamp viral RNA minikit, according to the manufacturer's instructions.

RT-PCR diagnostic assays.

Initially, two quantitative RT-PCR (Q-RT-PCR) assays were used that targeted regions of the polymerase (L) gene (MARVLF, TTATTGCATCAGGCTTCTTGGCA; MARVLR, GGTATTAAAAAATGCATCCAA [GenBank accession number AY358025; bp 13321 to 133517]) and the glycoprotein (GP) gene (MARVGPF, AAAGTTGCTGATTCCCCTTTGGA; MARVGPR, GCATGAGGGTTTTGACCTTGAAT [GenBank accession number AY358025; bp 6131 to 6355]). Later, an assay that targeted a region of NP (MARVNPF; TGAATTTATCAGGGATTAAC; MARVNPR, GTTCATGTCGCCTTTGTAG [GenBank accession number AY358025; bp 967 to 1146]) was used in place of the GP gene assay. MARV RNA was detected using a LightCycler RNA amplification SYBR green I kit (Roche). Briefly, 5 μl of RNA was added to 20 μl of a master mixture containing 1× SYBR green I mix, 5 mM MgCl2, 0.6 μM forward and reverse primers, and 0.5 μl of the enzyme mix. Q-RT-PCR assays were run on Smartcycler II thermocyclers. A reverse transcriptase step at 50°C for 20 min and a 2-min inactivation step at 94°C were followed by 40 cycles at 94°C for 15 s, 50°C for 30 s, and 72°C for 30 s, where a single datum point was taken. Melt curve analysis was performed to confirm the identities of the amplification products. Samples were considered positive if they produced melting point-confirmed amplification products in both assays. The amplification products were returned to NML and directly sequenced using both forward and reverse primers. A titrated virus stock, previously isolated from a MARV-positive blood sample, was used to provide a standard curve between the 50% tissue culture infective dose (TCID50) per milliliter and the CT value. The titers of the other samples were calculated on the basis of this standard curve. Aliquots of 1 μl of each RNA extract were tested by RT-LAMP assay using Musoke lineage-specific primers.


Specificities and sensitivities of RT-LAMP assay.

We initially examined the target region for RT-LAMP, which is longer than 250 nucleotides and which is highly conserved among the NP genes of the strains of each of the Musoke and Ravn lineages (Fig. (Fig.1).1). A set of six primers specific for the Musoke or Ravn lineage was designed for this target region, as described in Materials and Methods. To amplify all of the MARV strains simultaneously in one tube, we performed a multiple RT-LAMP (mRT-LAMP) assay with a combination of two sets of primers specific for both the Musoke and Ravn lineages (Table (Table11).

To evaluate the specificities of the mRT-LAMP assay, aliquots of 0.2 ng of RNA extract of each virus-infected cell culture were tested. Eleven virus strains, nine filoviruses (four MARV strains, two ZEBOV strains, one SEBOV strain, one REBOV strain, and one ICEBOV strain), and two arenaviruses (two LASV strains) (see Materials and Methods), were examined. The reaction was monitored by agarose gel deposition and naked-eye visualization, following incubation of the reaction mixture at 63°C for 60 min. The mRT-LAMP specifically amplified products with typical ladder-like patterns from all four MARV strains tested (Fig. (Fig.2,2, upper panel). No cross-reaction with the RNA of EBOV and LASV was observed in any RT-LAMP assay. In addition, we have confirmed by a BLAST homology search that the sequences of the primers for MARV do not have any homology with the genes of any other hemorrhagic fever viruses, including EBOV, LASV, Crimean-Congo hemorrhagic fever virus, Rift Valley fever virus, yellow fever virus, and dengue virus. The assay also did not show any cross-reactivity with these viruses, since the assay did not show any positivity even for EBOV, which belongs to the same virus family as MARV, indicating that the MARV-specific primers demonstrated a high degree of specificity for MARV. Corresponding positive results could be seen visually by UV irradiation of the reaction mixture containing fluorescence reagents (Fig. (Fig.2,2, lower panel). The assay with the Musoke and the Ravn lineage-specific primers reacted only with MARV Musoke-lineage strains Musoke, Ozolin, and Angola and the Ravn strain, respectively (data not shown).

FIG. 2.
Specificities of mRT-LAMP assay for detecting MARV. The assay was performed at 63°C for 60 min using a mixture of Musoke and Ravn lineage-specific primers. Amplification was detected by electrophoretic analysis on a 3% agarose gel (upper ...

To confirm that the products were amplified from the target region, the RT-LAMP products were digested with restriction endonucleases SmlI and XbaI, which recognized the sequences on the amplification products conserved among the Musoke- and Ravn-lineage strains, respectively (Fig. (Fig.1).1). The mRT-LAMP products from standard RNAs with the corresponding Mus-NP and Rav-NP sequences were digested with SmlI and XbaI, respectively. As shown in Fig. Fig.3,3, the sizes of the fragments were consistent with the sizes predicted for each strain (105 bp and 140 bp for Mus-NP, 107 bp and 138 bp for Rav-NP). The same results were seen for the products of RT-LAMP with viral RNAs, indicating that the mRT-LAMP reaction was specific.

FIG. 3.
Restriction enzyme analysis of LAMP products. Products were obtained by LAMP in a reaction using multiple primers with the standard RNAs of the NP-coding sequence of the Musoke strain (lanes 1 and 2) or the Ravn strain (lanes 3 and 4) or viral RNA of ...

The sensitivity of the RT-LAMP assay for MARV was determined by testing serial 10-fold dilutions of viral genomic RNAs of MARV. The viral genomic RNAs of the Musoke and Ravn strains, ranging from 104 to 100 copies per reaction volume, were tested in three separate runs for each dilution on a real-time turbidity-monitoring device. The detection limit of the mRT-LAMP assay was 102 copies per tube for both the Musoke and Ravn strains, and positive results were obtained in 20 to 35 min (Table (Table2).2). The detection limits were further confirmed by testing two additional strains, Angola and Ozolin (data not shown). The assay was equivalent or more sensitive than conventional RT-PCR specific for filoviruses, but real-time RT-PCR using TaqMan probes targeting the NP gene of MARV could detect a lower viral RNA copy number in the test using the same diluted viral RNA template. For further analysis of the sensitivities of the assay, the dose-response curves for RT-LAMP were determined by testing serial 10-fold dilutions of in vitro-transcribed standard RNAs (standard transcript RNAs). The mRT-LAMP using a total of 11 primers showed RNA standard dose-response curves similar to those of ordinary RT-LAMP using 6 primers specific for Musoke- or Ravn-lineage strains (Fig. (Fig.4).4). The mRT-LAMP assay showed 100% sensitivity in detecting more than 103 copies of Mus-NP and 104 copies of Rav-NP standard transcript RNAs per reaction. The borderline analytical detection limit for mRT-LAMP was 102 RNA copies per reaction. The average times for detection of 102 copies were 39.9 min with a 50% positivity rate for Mus-NP RNA and 34.8 min with a 37.5% positivity rate for Rav-NP RNA. Real-time monitoring of the turbidity could determine most of the positive results in less than 45 min in the reaction using multiple primers or lineage-specific primers.

FIG. 4.
RNA dose-response curve of RT-LAMP assays. Tenfold serially diluted transcript RNAs ranging from 1 to 100,000 copies were tested in eight replicate assays each. The RT-LAMP reaction was carried out for 60 min on a real-time turbidity-monitoring device. ...
Comparison of sensitivities of the mRT-LAMP and RT-PCR methods for detection of MARV RNAa

Evaluation of RT-LAMP assay with clinical specimens.

A total of 24 clinical specimens obtained from the outbreak in Angola were used for evaluation of the RT-LAMP assay, and the results were compared with those of the real-time RT-PCR assay performed in the field laboratory. The RT-LAMP reaction was performed at 63°C for 45 min using Musoke lineage-specific primers with 1 μl of RNA extracted from the clinical samples, followed by detection of the LAMP products by agarose gel electrophoresis or visually with fluorescence of the reaction mixture (Table (Table3).3). A concordance of 78% (14 of 18 samples) of positive results between the RT-LAMP assay and the real-time RT-PCR was observed. All six samples negative by RT-PCR were also negative by RT-LAMP. The sensitivity of the RT-LAMP assay with clinical specimens was approximately 104 TCID50s/ml. In this study, each sample was tested in duplicate, and the samples that showed positive results in both reactions were finally determined to be MARV positive. Two RNA samples from the specimens containing the viruses at 3.5 × 104 and 8 × 103 TCID50s/ml showed one positive reaction. The RT-LAMP-negative samples were retested by extension of the incubation to 60 min; however, the results did not change. These negative samples were tested in duplicate using the assay deployed during the Angola outbreak; both replicates of three of four samples were positive but produced higher CT values, indicating that they had lower levels of viral RNA. The other sample produced a positive signal in only one replicate.

Evaluation of the RT-LAMP assay with clinical specimens from MARV outbreak in Angola, 2004-2005

In this study, we did not carry out the mRT-LAMP with clinical samples, since sufficient amounts of viral RNA from the clinical specimens did not remain. However, we suppose that similar results could be obtained from the mRT-LAMP, since the sensitivities of the mRT-LAMP for the in vitro-transcribed viral RNA were almost equivalent to those of the lineage-specific RT-LAMP (Fig. (Fig.44).


Phylogenetic analysis indicated that MARV strains could be separated into two genetic lineages, the minor Ravn lineage and the major Musoke lineage, and strains of both lineages are pathogenic in humans (2, 25). In the outbreak in Durba, DRC, multiple genetically distinct viruses of both lineages were thought to have been introduced into the human population (2). Thus, it is necessary to establish clinical diagnostic methods for detection of the genetic diversity within the genus Marburg virus.

Degenerate primers are often used for the simultaneous detection of diverged sequences by PCR or other nucleotide amplification methods. We initially attempted to use primers with several mixed bases for the simultaneous detection of strains within the main and minor lineages, but these did not work for rapid amplification. The advantage of LAMP using multiple primers for the simultaneous detection of distinct genetic sequences, such as those of Babesia bovis and Babesia bigemina and norovirus genogroups I and II, has previously been reported by other groups (7, 10). The viral genome sequences were highly conserved among the strains in the respective MARV lineages. Therefore, we adopted the mRT-LAMP assay using two sets of primers designed according to the Musoke or Ravn lineage consensus sequence. There was no significant loss of specificity or sensitivity with the mRT-LAMP assay when the results were compared to those of lineage-specific RT-LAMP assays (Fig. (Fig.22 and and44).

Agarose gel electrophoresis is a reliable method for detecting LAMP products, By this method, ladder-like patterns are observed, and the sizes of fragments are determined by digestion of the products with a restriction endonuclease (16). Nevertheless, agarose gel electrophoresis is less suitable for field laboratories. The LAMP assay with real-time turbidity monitoring or naked-eye judgment should be applied as a screening test for MARV infection due to its rapidity and simplicity. In real-time turbidity monitoring, mRT-LAMP-positive results could be observed within approximately 20 min in reactions with 104 copies of viral RNA (Table (Table2).2). In addition, the amplification reaction with naked-eye judgments showed specificities consistent with those by detection by agarose gel electrophoresis in the reaction with viral RNAs (Fig. (Fig.22).

Recently, RT-PCR and TaqMan RT-PCR assays that could potentially detect all known MARV strains have been reported (8, 17, 23, 28, 29). Our RT-LAMP assay had sensitivity equivalent to or greater than that of RT-PCR in parallel tests using the same template RNAs. However, TaqMan RT-PCR could provide positive results with a lower copy number than the mRT-LAMP assay (Table (Table2).2). The mRT-LAMP reactions were positive at approximately 102 copies of viral RNA or the in vitro-transcribed RNA per reaction within 35 or 45 min, and the sensitivities of the assay did not differ significantly among the MARV strains. These results indicate that the RT-LAMP assay has a specificity and a sensitivity sufficiently high to detect MARV genomes.

In the evaluation test with clinical specimens obtained from the outbreak in Angola, the RT-LAMP assay using Musoke lineage-specific primers showed good performance with regard to concordance and specificity, even with visual detection using a fluorescent reagent (Table (Table3).3). This will make the RT-LAMP assay more highly beneficial than RT-PCR or other nucleic acid amplification methods for use as a field diagnosis technique, as the only apparatus required for the assay is a simple device, such as a water bath or a heat block, that furnishes a constant temperature of 63°C. It has been reported that LAMP was less influenced by PCR inhibitors in blood components (1, 13), and with its high sensitivity, the assay could be used with clinical specimens, such as blood or body fluids. In this study, the MARV-positive rate of the assay was less than that of RT-PCR performed in the field laboratory in Uige, Angola. Four RT-PCR-positive samples with titers less than 104 or 103 TCID50s/ml were negative by RT-LAMP. There are several possible reasons for this. First, the quality of viral RNA in the clinical specimens did affect the differences in sensitivity between RT-LAMP and RT-PCR. We used clinical samples that had been repeatedly frozen and thawed before this study, and therefore, the viral RNA was likely partially degraded, as indicated by the higher CT values obtained when the samples were retested using the field RT-PCR assay. Second, the field RT-PCR assay had higher sensitivities than the RT-LAMP assay used for the clinical specimens. The retest with the field RT-PCR exhibited positive results even with the clinical samples negative by RT-LAMP. In addition, the RT-PCR assay used in the field laboratory utilized a sample volume of 5 μl, whereas the LAMP assay used 1 μl, which would likely have also contributed to the negative results for these samples.

The assay developed in this study is highly specific and rapid for the molecular diagnosis of MARV infections. However, the specificities and sensitivities of the assay must be continuously assessed and improved due to the emergence of unknown MARV strains in future outbreaks. The assay does not require the use of sophisticated equipment or highly skilled personnel. The cost of the assay per sample is lower than that of RT-PCR or ELISA, and the assay can provide accurate results in a short time frame. These characteristics make it potentially useful for the clinical diagnosis and management of MARV outbreaks in areas where it is epidemic.


We thank Hideki Ebihara, Institute of Medical Sciences, University of Tokyo, Tokyo, Japan, and Ayato Takada, Hokkaido University, Sapporo, Japan, for their contributions to the present study.

This work was supported by grants from the Ministry of Health, Labor, and Welfare of Japan, the Japan Society for the Promotion of Science, and the Japan Science and Technology Agency (JST). The study was further supported by the National Microbiology Laboratory of the Public Health Agency of Canada.


[down-pointing small open triangle]Published ahead of print on 26 April 2010.


1. Al-Soud, W. A., and P. Radstrom. 2001. Purification and characterization of PCR-inhibitory components in blood cells. J. Clin. Microbiol. 39:485-493. [PMC free article] [PubMed]
2. Bausch, D. G., S. T. Nichol, J. J. Muyembe-Tamfum, M. Borchert, P. E. Rollin, H. Sleurs, P. Campbell, F. K. Tshioko, C. Roth, R. Colebunders, P. Pirard, S. Mardel, L. A. Olinda, H. Zeller, A. Tshomba, A. Kulidri, M. L. Libande, S. Mulangu, P. Formenty, T. Grein, H. Leirs, L. Braack, T. Ksiazek, S. Zaki, M. D. Bowen, S. B. Smit, P. A. Leman, F. J. Burt, A. Kemp, and R. Swanepoel. 2006. Marburg hemorrhagic fever associated with multiple genetic lineages of virus. N. Engl. J. Med. 355:909-919. [PubMed]
3. Bausch, D. G., J. S. Towner, S. F. Dowell, F. Kaducu, M. Lukwiya, A. Sanchez, S. T. Nichol, T. G. Ksiazek, and P. E. Rollin. 2007. Assessment of the risk of Ebola virus transmission from bodily fluids and fomites. J. Infect. Dis. 196(Suppl. 2):S142-S147. [PubMed]
4. Borchert, M., S. Mulangu, P. Lefevre, A. Tshomba, M. L. Libande, A. Kulidri, J. J. Muyembe-Tamfum, and P. Van der Stuyft. 2007. Use of protective gear and the occurrence of occupational Marburg hemorrhagic fever in health workers from Watsa health zone, Democratic Republic of the Congo. J. Infect. Dis. 196(Suppl. 2):S168-S175. [PubMed]
5. Colebunders, R., H. Sleurs, P. Pirard, M. Borchert, M. Libande, J. P. Mustin, A. Tshomba, L. Kinuani, L. A. Olinda, F. Tshioko, and J. J. Muyembe-Tamfum. 2004. Organisation of health care during an outbreak of Marburg haemorrhagic fever in the Democratic Republic of Congo, 1999. J. Infect. 48:347-353. [PubMed]
6. Feldmann, H., T. W. Geisbert, P. B. Jahrling, H. D. Klenk, S. V. Netesov, C. J. Peters, A. Sanchez, R. Swanepoel, and V. E. Volchkov. 2005. Filoviridae, p. 645-653. In C. M. Fauquet, M. A. Mayo, J. Maniloff, U. Desselberger, and L. A. Ball (ed.), Virus taxonomy. VIIIth report of the ICTV. Academic Press, London, United Kingdom.
7. Fukuda, S., S. Takao, M. Kuwayama, Y. Shimazu, and K. Miyazaki. 2006. Rapid detection of norovirus from fecal specimens by real-time reverse transcription-loop-mediated isothermal amplification assay. J. Clin. Microbiol. 44:1376-1381. [PMC free article] [PubMed]
8. Gibb, T. R., D. A. Norwood, Jr., N. Woollen, and E. A. Henchal. 2001. Development and evaluation of a fluorogenic 5′-nuclease assay to identify Marburg virus. Mol. Cell. Probes 15:259-266. [PubMed]
9. Grolla, A., A. Lucht, D. Dick, J. E. Strong, and H. Feldmann. 2005. Laboratory diagnosis of Ebola and Marburg hemorrhagic fever. Bull. Soc Pathol. Exot. 98:205-209. [PubMed]
10. Iseki, H., A. Alhassan, N. Ohta, O. M. Thekisoe, N. Yokoyama, N. Inoue, A. Nambota, J. Yasuda, and I. Igarashi. 2007. Development of a multiplex loop-mediated isothermal amplification (mLAMP) method for the simultaneous detection of bovine Babesia parasites. J. Microbiol. Methods 71:281-287. [PubMed]
11. Ito, M., M. Watanabe, N. Nakagawa, T. Ihara, and Y. Okuno. 2006. Rapid detection and typing of influenza A and B by loop-mediated isothermal amplification: comparison with immunochromatography and virus isolation. J. Virol. Methods 135:272-275. [PubMed]
12. Johnson, E. D., B. K. Johnson, D. Silverstein, P. Tukei, T. W. Geisbert, A. N. Sanchez, and P. B. Jahrling. 1996. Characterization of a new Marburg virus isolated from a 1987 fatal case in Kenya. Arch. Virol. Suppl. 11:101-114. [PubMed]
13. Kurosaki, Y., T. Sakuma, A. Fukuma, Y. Fujinami, K. Kawamoto, N. Kamo, S. I. Makino, and J. Yasuda. 2009. A simple and sensitive method for detection of Bacillus anthracis by loop-mediated isothermal amplification. J. Appl. Microbiol. 107:1947-1956. [PubMed]
14. Kurosaki, Y., A. Takada, H. Ebihara, A. Grolla, N. Kamo, H. Feldmann, Y. Kawaoka, and J. Yasuda. 2007. Rapid and simple detection of Ebola virus by reverse transcription-loop-mediated isothermal amplification. J. Virol. Methods 141:78-83. [PubMed]
15. Mühlberger, E., A. Sanchez, A. Randolf, C. Will, M. P. Kiley, H. D. Klenk, and H. Feldmann. 1992. The nucleotide sequence of the L gene of Marburg virus, a filovirus: homologies with paramyxoviruses and rhabdoviruses. Virology 187:534-547. [PubMed]
16. Notomi, T., H. Okayama, H. Masubuchi, T. Yonekawa, K. Watanabe, N. Amino, and T. Hase. 2000. Loop-mediated isothermal amplification of DNA. Nucleic Acids Res. 28:E63. [PMC free article] [PubMed]
17. Panning, M., T. Laue, S. Olschlager, M. Eickmann, S. Becker, S. Raith, M. C. Courbot, M. Nilsson, R. Gopal, A. Lundkvist, A. Caro, D. Brown, H. Meyer, G. Lloyd, B. M. Kummerer, S. Gunther, and C. Drosten. 2007. Diagnostic reverse-transcription polymerase chain reaction kit for filoviruses based on the strain collections of all European biosafety level 4 laboratories. J. Infect. Dis. 196(Suppl. 2):S199-S204. [PubMed]
18. Parida, M., K. Horioke, H. Ishida, P. K. Dash, P. Saxena, A. M. Jana, M. A. Islam, S. Inoue, N. Hosaka, and K. Morita. 2005. Rapid detection and differentiation of dengue virus serotypes by a real-time reverse transcription-loop-mediated isothermal amplification assay. J. Clin. Microbiol. 43:2895-2903. [PMC free article] [PubMed]
19. Peterson, A. T., R. R. Lash, D. S. Carroll, and K. M. Johnson. 2006. Geographic potential for outbreaks of Marburg hemorrhagic fever. Am. J. Trop. Med. Hyg. 75:9-15. [PubMed]
20. Sanchez, A., T. W. Geisbert, and H. Feldmann. 2007. Filoviridae: Marburg and Ebola viruses, p. 1409-1448. In D. M. Knipe, and P. M. Howley (ed.), Fields virology, vol. 1, 5th ed. Lippincott Williams & Wilkins, Philadelphia, PA.
21. Sanchez, A., S. G. Trappier, U. Stroher, S. T. Nichol, M. D. Bowen, and H. Feldmann. 1998. Variation in the glycoprotein and VP35 genes of Marburg virus strains. Virology 240:138-146. [PubMed]
22. Smith, D. H., B. K. Johnson, M. Isaacson, R. Swanepoel, K. M. Johnson, M. Killey, A. Bagshawe, T. Siongok, and W. K. Keruga. 1982. Marburg-virus disease in Kenya. Lancet i:816-820. [PubMed]
23. Strong, J. E., A. Grolla, P. B. Jahrling, and H. Feldmann. 2006. Filoviruses and arenaviruses, p. 774-790. In B. Detrick, R. G. Hamilton, and J. D. Folds (ed.), Manual of molecular and clinical laboratory immunology, 7th ed. ASM Press, Washington, DC.
24. Tomita, N., Y. Mori, H. Kanda, and T. Notomi. 2008. Loop-mediated isothermal amplification (LAMP) of gene sequences and simple visual detection of products. Nat. Protoc. 3:877-882. [PubMed]
25. Towner, J. S., M. L. Khristova, T. K. Sealy, M. J. Vincent, B. R. Erickson, D. A. Bawiec, A. L. Hartman, J. A. Comer, S. R. Zaki, U. Stroher, F. Gomes da Silva, F. del Castillo, P. E. Rollin, T. G. Ksiazek, and S. T. Nichol. 2006. Marburg virus genomics and association with a large hemorrhagic fever outbreak in Angola. J. Virol. 80:6497-6516. [PMC free article] [PubMed]
26. Towner, J. S., P. E. Rollin, D. G. Bausch, A. Sanchez, S. M. Crary, M. Vincent, W. F. Lee, C. F. Spiropoulou, T. G. Ksiazek, M. Lukwiya, F. Kaducu, R. Downing, and S. T. Nichol. 2004. Rapid diagnosis of Ebola hemorrhagic fever by reverse transcription-PCR in an outbreak setting and assessment of patient viral load as a predictor of outcome. J. Virol. 78:4330-4341. [PMC free article] [PubMed]
27. Towner, J. S., T. K. Sealy, M. L. Khristova, C. G. Albarino, S. Conlan, S. A. Reeder, P. L. Quan, W. I. Lipkin, R. Downing, J. W. Tappero, S. Okware, J. Lutwama, B. Bakamutumaho, J. Kayiwa, J. A. Comer, P. E. Rollin, T. G. Ksiazek, and S. T. Nichol. 2008. Newly discovered Ebola virus associated with hemorrhagic fever outbreak in Uganda. PLoS Pathog. 4:e1000212. [PMC free article] [PubMed]
28. Weidmann, M., F. T. Hufert, and A. A. Sall. 2007. Viral load among patients infected with Marburgvirus in Angola. J. Clin. Virol. 39:65-66. [PubMed]
29. Zhai, J., G. Palacios, J. S. Towner, O. Jabado, V. Kapoor, M. Venter, A. Grolla, T. Briese, J. Paweska, R. Swanepoel, H. Feldmann, S. T. Nichol, and W. I. Lipkin. 2007. Rapid molecular strategy for filovirus detection and characterization. J. Clin. Microbiol. 45:224-226. [PMC free article] [PubMed]

Articles from Journal of Clinical Microbiology are provided here courtesy of American Society for Microbiology (ASM)