0 INSTITUT PASTEUR/ELSEVIER Res. Viral.

Paris 1997 1997, 148, 11-15

Use of magnetic beads versus guanidium

thiocyanate-phenol-chloroform RNA extraction
followed by polymerase chain reaction for the rapid,
sensitive detection of enterovirus RNA
F. Beaulieux, D.M. See, I. Leparc-Goffart, M. Aymard and B. Lina (*)
Luboratoire de Virologie, Centre national de Rc;fe’rencedes Ente’rovirus et de I’Hkpatite A,
Domaine Rockefeller, Faculte’ de Mt!decine Lyon Grange Blanche, 69373 Lyon Cedex 08 (France)

SUMMARY

The current study compares the sensitivity of RNA extraction using magnetic beads
versus that of a standard extraction method. Streptavadincoated magnetic beads were
labelled with a biotinylated, enterovirus-specific oligonucleotide. RNA was extracted
using labelled beads or guanidium thiocyanate-phenol-chloroform from 1, 0.1 and 0.01
TCID /lOOul of stock coxsackievirus types A9 and B3, echovirus type 11, enterovirus
type 70 and poliovirus type 1. Each strain was tested three times. RNA extraction using
magnetic beads was >50% faster than the standard method. The RNA was amplified
using RT-PCR, and the products were detected using agarose gel electrophoresis; 6/15
and 7/15 samples at an initial concentration of 0.01 TClD,/lOOuI were detected using
magnetic beads or standard extraction, respectively. Negative-stain electron micros-
copy was used to determine that 0.01 TCID,/lOOul of coxsackievirus 83 contained
approximately 3 genomes. Thus, use of magnetic beads labelled with an enterovirus-
specific oligonucleotide was less toxic, more rapid and as sensitive as the current stan-
dard RNA extraction method.

Key-words: RNA, PCR, Enterovirus; Dynabeads, Extraction, Methods.

INTRODUCTION occur (Grist et al., 1978). Furthermore, evidence

suggests that persistent infections may be asso-
Enteroviruses are responsible for a large num- ciated with diabetes mellitus (Helfand et al.,
ber of human diseases (Rotbart, 1989). Most are 1993, idiopathic dilated cardiomyopathy (Koide
asymptomatic and self-limited. Clinically mild et al., 1992), postpoliomyelitis syndrome (Dala-
syndromes include exanthems, upper respiratory kas and Illa, 1991), amyotrophic lateral sclerosis
infections and pleuritis. However, life-threatening (Swanson et al., 1995), polymyositis (Dalakas,
acute infections such as meningitis, encephalitis, 1995) and the chronic fatigue syndrome (Clem-
neonatal sepsis and myopericarditis may also ents et aI., 1995).

Submitted September 4, 1996, accepted October 2, 1996.

(*) Corresponding author: Laboratoire de Virologie, Domaine Rockefeller, 8, avenue Rockefeller, 69373 Lyon Cedex 08, France.
12 F. BEAULIEUX ET AL.

Current methods for the diagnosis of entero- virus type 1 Sabin were propagated in green mon-
virus infections, such as cell culture and sero- key kidney (BGM) cells as previously described
(Melnick, 1990).
logical evaluation, are limited because they are
time-consuming and insensitive. Several anti-
enteroviral agents are under investigation RNA extraction
(Woods et al., 1989), and evidence from animal
models (See and Tilles, 1992) suggests that Guunidium thiocyunate-phenol-chloroform
early intervention may be required to success- Stock virus supernatants were harvested, and
fully treat enteroviral infections in humans. Fur- serial lo-fold dilutions were performed in order to
thermore, timely diagnosis can be important to achieve 1, 0.1, and 0.01 TCID,,/lOOpl. Serial dilu-
prevent unnecessary therapy for other diseases tions from 3 different supematants were used to per-
clinically similar to enteroviral disease. Thus, form 3 assays for each virus. RNA was extracted by
standard methods (Leparc et al., 1994). Briefly,
the rapid diagnosis of enteroviral meningitis, a IOOpl of each dilution was mixed with 1 ml RNA
condition which is usually self-limited, can PLUS (Bioprobe Systems, France) and incubated for
avert the use of antibiotics for suspected bacte- 5 min at 4”C, 200~1 of chloroform was added, the
rial disease. Furthermore, costly hospitalizations mixture was vortexed vigorously for 15 s and then
for observation or empiric treatment can be incubated for 5 min at 4°C. After centrifugation at
12,OOOg for 15 min, the RNA-containing aqueous
avoided. phase was added to 5OOpl of isopropanol and pre-
Polymerase chain reaction (PCR) amplifica- cipitated for 1 h. After centrifugation at 12,000 g for
tion is a sensitive method for the detection of 15 min, the pellet was washed twice with 75 % etha-
nol and then resuspended in 201.~1of DEPC water.
specific nucleic acid sequences. Probes have RNase inhibitor (1 ~1) was then added.
been developed which hybridize to homologous
regions of enteroviral genomes, permitting
detection of most enteroviral serotypes (Fuchs et Magnetic beads
al., 1993). Thus, diagnosis of enteroviral infec-
1) Labelling with biotinylated oligonucleotide (Muir
tions using PCR with specific primers could be et al., 1993)
promising. However, the clinical use of PCR is
The mixture (IOOpl) containing 1OOmg of mag-
limited partly because of the time-consuming netic beads (Dynal, France) was placed in an
nature of sample preparation. In the current Eppendorf tube, and 6~1 containing 600pmol of a
study we describe a rapid, non-toxic method of 5’-biotinylated oligonucleotide from a highly con-
RNA extraction using magnetic beads. Com- served sequence in the 5’-non-coding region of the
enterovirus genome were added (sequence: 5’-ATT
bined with subsequent PCR amplification and GTC ACC ATA AGC AGC CA-3’). The mixture
gel electrophoresis, results with a sensitivity of was incubated for 15 min at room temperature and
3 genomes/lOO ~1 can be obtained within then washed 3 times with 200ml of 6xSSPE using
5 hours. a magnetic concentrator. The labelled magnetic
beads were resuspended in 1 ml of 12xSSPE. They
were either used immediately or stored for up to
48h at -20°C.
MATERIALS AND METHODS
2) Extraction and purification of the RNA
The diluted stock virus supematant (IOOpl) (each
Samples assay was performed 3 times) was added to 300~1 of
RNA PLUS solution for 5 min at 4°C. The oligonu-
Reference viral strains were obtained from the cleotide-conjugated magnetic bead solution (400 ~1)
National Reference Centre (Lyon, France). was added, and the mixture was incubated at 4°C for
Coxsackievirus A9, coxsackievirus B3 Nancy 5 min. Enteroviral RNA-magnetic bead conjugates
strain, echovims type 11, enterovirus 70 and polio- were purified using the magnetic concentrator, then

RT-PCR = reverse-transcription polymerase chain reaction.

 ENTEROVIRUS RNA DETECTION BY MAGNETIC BEADS 13

washed 2 times with 6xSSPE. They were then 12 3 4 5 6 7 8 9 10 11 12 13

resuspended in 20~1 of DEPC water. RNase inhibi-
tor (1~1) was added. A
300bp
cDNA synthesis and PCR

The extracted RNA (5~1) was used to produce

cDNA ; seminested PCR was performed as previ-
ously described (Leparc et al., 1994). A negative
control (DEPC water) was included in each PCR
run. Ten microlitres of the PCR product were run on
a 2% agarose electrophoresis gel and visualized by B
ethidium bromide staining under UV light. 300bp

Sensitivity

To determine the sensitivity of the assay in terms Fig. 1. Extraction of RNA using magnetic beads (A) or
of genomes/lOOpl, IOOpl of stock CVB3 at 1 guanidium thiocyanate-phenol-chloroform (B).
TCID.JlOOpl were mixed with 1OOpl containing Lane 1 =MW marker, lane 2=positive control, lane 3
lo4 llposomes (112nm; Ted Pella, USA). One =negative control, lane 4=coxsackievirus type A9 0.1
microlitre of the mixture was stained with 2 % phos- TCID/,,, lane 5=coxsackievirus type A9 0.01 TCID,,,
photungstic acid (Ted Pella, USA) for 80s. mounted lane 6=coxsackievirus type B3 0.1 TCID,,, lane 7=cox-
onto 400-mesh copper grids, air-dried and examined sackievirus type B3 0.01 TCID,,, lane 8=echovirus
under a “Zeiss CR-lo” electron microscope. The type 11 0.1 TCID/,,, lane 9=echovirus type 11 0.01
numbers of liposomes and viral particles were TCID,,, lane lO=enterovims type 70 0.1 TCID,,, lane 11
counted for 50 fields under high power (x60,000), =enterovirus type 70 0.01 TCID,,, lane 12=poliovirus
type 1 0.1 TCID,,, lane 13 =poliovirus type 1 0.01
and the numbers were averaged. The ratio of the
TCID 50. PCR products were run on a 2 % agarose
complete viral particles to liposomes was used to electrophoresis gel and visualized under UV light.
calculate the number of genomes/ 100 ~1. Bands=300 bp.

RESULTS AND DISCUSSION

trophoresis can be accomplished in 5 h. Thus,
The current study describes a rapid method for same-day results can be reported. The use of
extraction of enteroviral RNA from cell culture seminested PCR increases sensitivity and obvi-
supematants using magnetic beads labelled with a ates the need for time-consuming hybridization
specific capture oligonucleotide. The extraction detection techniques. Seminested PCR techniques
process requires only lo-20 min per specimen have a substantial risk of carryover contamina-
compared with 2 h for guanidium thiocyanate- tion. However, negative controls were used with
phenol-chloroform. The sensitivity of the two each assay in the current study, and there were no
methods was comparable : 0.1 TCID,, was false-positive results.
detected using both methods for each sample One TCID,,/lOO yl CVB3 contained an aver-
(fig. 1). After PCR amplification of 0.01 TCID,,, age of 2,035 + 342 incomplete virus and 312 +
detectable bands were observed for 6/l 5 and 7/15 46 complete (nucleic acid-containing) particles
samples after magnetic bead or guanidium thioc- (fig. 2) by electron microscopy. Thus, 0.01
yanate-phenol-chloroform extraction, respec- TCID,,/lOO~l (the lower limit of the assay)
tively (table I). Finally, the magnetic bead assay contained an average of 3 genomes/lOOpl.
did not require the use of the toxic reagents chlo- However, the actual sensitivity of the assay in
roform and isopropanol. terms of genomes/lOOpl could be slightly dif-
RNA extraction followed by cDNA produc- ferent, given two variables: (i) the yield of RNA
tion, seminested PCR amplification and gel elec- extraction could be < lOO%, and/or (ii) some of
14 F. BEAULIEUX ET AL.

Table I. Comparison of RNA extraction by magnetic beads versus standard methodology.

Concentration of Virus
I TCIDJlOO ~1 0.1 TCID#OO ~1 0.01 TCIDJlOO ~1
Virus Magnetic Standard Magnetic Standard Magnetic Standard
strain beads(*) method (*I bead method beads method

CVA9 +(**I + + + - -
+ + + + - +
+ + + + - -
CVB3 + + + + - +
+ + + + - -
+ + + + + +
El1 + + + + + -
+ + + + - +
+ + + + - -
EV70 + + + + - -
+ + + + + -
+ + + + + +
PVl + + + + - +
+ + + + + +
-
Total (%) l& lie GO 160 4’0 46.7

(*) Method of RNA extraction. See “Materials and Methods” for details.
(**) +(positive) or - (negative) visualization of final PCR product.

the incomplete particles could contain genomic

fragments. Since RNA was extracted from
lOO@ of diluted supematant, the assay could
detect as few as 3 genomes almost 50% of the
time. However, enteroviral RNA was only reli-
ably detected at 0.1 TCID,, (approximately 30
genomes). The negative results at a concentra-
tion of 0.01 TCIDsO could be due to failure to
transfer RNA during the serial dilution process,
or limitations in the amplification process.

The current study used an oligonucleotide

with a conserved sequence on the magnetic beads
in order to capture enteroviral RNA from most
serotypes. Other oligonucleotides can be used to
capture serotype-specific enteroviral strains or
other organisms. Thus, a large variety of micro-
organisms can be detected in conjunction with
Fig. 2. Electron microphotograph (X 120,000) showing amplification using appropriate primers.
an incomplete (arrow) and 2 complete virions (arrow-
heads). (Virions are stained with 2% phosphotungstic In summary, magnetic beads labelled with a
acid). conserved, enterovirus-specific oligonucleotide
 ENTEROVIRUS RNA DETECTION BY MAGNETIC BEADS 15

were used to efficiently and rapidly extract RNA References

without the use of toxic reagents. The sensitivity Clements, G.B., McGarry, F., Naim, C. & Galbraith, D.N.
was comparable to that of a standard method. (1995), Detection of enterovirus-specific RNA in
Other capture probes could be used to extract serum: the relationship to chronic fatigue. J. Med.
RNA from other organisms. Thus, magnetic bead Viral., 45, 156-161.
Dalakas, M. & Illa, I. (1991), Post-polio syndrome: con-
extraction followed by seminested RT-PCR and cepts in clinical diagnosis, pathogenesis and etiology.
gel electrophoresis could permit the same-day Adv. Neurol., 56, 495-5 11.
detection of a variety of infectious agents. Fur- Dalakas, M.C. (1995). Enterovirus and human neuromus-
cular diseases, in “Human enterovirus infections”.
thermore, the technique holds promise for the (H.A. Rotbart) (pp. 387-398), American Society for
sensitive detection of specific organisms. Microbiology, Washington DC.
Fuchs, F., Leparc, I., Kopecka, H., Garin, D. &
Aymard, M. (1993), use of cRNA diggoxigenin-
labelled probes for detection of enteroviruses in
humans and in the environment. J. Virol. Methods,
Acknowledgements 42, 217-226.
Grist, N.R., Bell, A.J. & Assaad, F. (1978), Enteroviruses
Thanks to F. Najioullah for proofreading the French in human disease. Bog. Med. Viral., 24, 114- 157.
abstract. Helfand, R.F., Gary, H.E., Jr., Freeman, C.Y., Anderson,
L.J. & Pallansch, M.A. (1995), Serologic evidence
of association between Enteroviruses and the onset
of type 1 diabetes mellitus. J. Infect. Dis., 172,
1206-1211.
Koide, H., Kitaura, Y., Deguchi, H., Ukimura, A., Kawa-
Comparaison de deux techniques d’extraction mura, K. & Hirai K. (1992), Genomic detection of
de I’ARN des ent&ovirus, avant PCR, enteroviruses in the myocardium - studies on animal
par l’utilisation de microbilles magnktiques hearts with Coxsackievirus B3 myocarditis and
ou de solvants organiques (isothiocyanate endomyocardial biopsies from patients with myocar-
de guanidium, phbnol et chloroforme) ditis and dilated cardiomyopathy. Jpn. Circ. J., 56,
1081-1093.
Leparc, I., Aymard, M. & Fuchs, F. (1994), Acute, chronic
Des billes marquBesB la streptavidine sont 1iCesg and persistent enterovirus and poliovirus infections :
une sonde biotinylCe, sptciflque des entCrovirus. detection of viral genome by nested PCR amplifi-
L’ARN est extrait par les billes magnetiques ou par cation in culture-negative samples. Mol. Cell. Probe,
une m&bode standard utilisant l’isothiocyanate de 8,487-495.
guanidium, le phenol et le chloroforme. L’Ctude a Melnick, J.L. (1990), Enteroviruses : poliovirus, cox-
tt6 r&liste 2 l’aide de stocks viraux B 1, 0.1 et 0.01 sackieviruses, echoviruses, and newer enterovi-
DITC,p/lOOpl pour les coxackievirus A9 et B3, ruses, in “Virology” (Fields) (pp. 549-605) 2nd.
Raven Press, NY.
l’echovuus 11, l’entirovirus 70 et le poliovirus type 1.
Muir, P., Nicholson, F., Jhetan, M., Neogi, S. & Banat-
Chaque souche a CtC testke 3 fois. L’extraction de vala, J. (1993), Rapid diagnosis of enterovirus infec-
I’ARN par les billes magnCtiquesest plus rapide que tion by magnetic bead and polymerase chain reaction
la technique standard. L’ARN est amplifiC g&e B of enterovirus RNA in clinical specimens. J. Clin.
une RT-PCR et les produits de l’amplification Microbial., 31, 31-38.
gtnique sont dCtectts par Clectrophor&se en gel Rotbart, H.A. (1989), Human Enterovirus infections :
d’agarose; 6/15 et 7/15 Cchantillons & la concentra- molecular approaches to diagnosis and pathogene-
tion initiale de 0.01 DITC ,,/lOOpl sont dCtectCs sis, in “Molecular aspects of picomavirus infection
g l’aide, respectivement, des billes magnktiques ou and detections” (B.L. Semler & E. Ehrenfeld) (pp.
de l’extraction standard. La microscopic Clectro- 242-264). American Society for Microbiology,
Washington, DC.
nique, en coloration nCgative montre que 0.01
See, D.M. & Tilles, J.G. (1992), Treatment of Coxsackie-
DITC,,/lOOpl de coxackievirus B3 contiennent virus A9 myocarditis in mice with WIN 54954. Anti-
approximativement 3 gtnomes. Ainsi, nous avons microb. Agents Chemother., 36.425-428.
montrt que la technique utilisant les billes Swanson, N.R., Fox, S.A. & Mastaglia, F.L. (1995),
magnktiques sur lesquelles est fixCe une sonde Search for persistent infection with poliovirus or
sptcifique des enttrovirns est moins toxique, plus other enterovirus in amyotrophic lateral sclerosis-
rapide est aussi sensibleque la m&ode convention- motor neuron disease. Neuromusc. Disord., 5,
nelle d’extraction des ARNs. 457-465.
Woods, M.G., Diana, G.D., Rogge, M.C., Otto, M.J.,
Mots-cl&s : ARN, PCR, EntCrovirus ; Billes Dutko, F.J. & M&inlay, M.A. (1989), In vitro and in
magnCtiques,Extraction, MCthodes. vivo activity of WIN 54954, a new broad-spectrum
antipicomavirus drug. Antimicrob. Agents Chemo-
ther., 33, 2069-2074.