JOURNAL OF CLINICAL MICROBIOLOGY, Jan. 1989, p. 120-125 Vol. 27, No. 10095-1137/89/010120-06$02.00/0Copyright C 1989, American Society for Microbiology

Monoclonal Antibody Solution Hybridization Assay for Detection ofHuman Immunodeficiency Virus Nucleic Acids

RAPHAEL P. VISCIDI,1* COLLEEN O'MEARA,2 HOMAYOON FARZADEGAN,3 AND ROBERT YOLKEN2

Departments of Medicine' and Pediatrics,2 Johns Hopkins University School of Medicine, and Department ofEpidemiology, Johns Hopkins University School of Hygiene and Public Health,3 Baltimore, Maryland 21205

Received 18 April 1988/Accepted 19 September 1988

ln this report we describe a novel, nonisotopic hybridization assay for the measurement of viral RNA inbiological samples. The assay involved a solution-phase reaction between a biotinylated DNA probe and RNAtarget sequences. Labeled hybrids were detected in an immunoreaction by using a solid-phase anti-biotinantibody and an enzyme-labeled monoclonal antibody specific for DNA-RNA hybrids. This monoclonalantibody solution hybridization assay was compared with an antigen-capture immunoassay for the detection ofhuman immunodeficiency virus in 436 cell culture samples from 60 seropositive patients. The sensitivity andspecificity of the hybridization assay were 93.5 and 94.6%, respectively. Detection of human immunodeficiencyvirus solely by hybridization in the initial sample but not subsequent samples from seven cultures may reflectdetection of virus that was present in the patients' lymphocytes but did not replicate in vitro. Since the assaymethod is adapatable to the detection of either RNA or DNA, it could provide a means for the detection of awide range of viral nucleic acids.

Nucleic acid hybridization assays have great promise asdiagnostic methods in virology (20). The vast informationencoded in viral genetic sequences, the specificity of basepair interactions, and the kinetic and thermodynamic prop-erties of the reaction between complementary polynucleo-tide chains make possible the construction of highly sensi-tive and specific assays for the detection and identification ofviral nucleic acids. In addition, recent advances in recombi-nant DNA technology and solid-phase oligonucleotide syn-thesis make it feasible to produce large quantities of well-defined, stable, and inexpensive nucleic acid probes asreagents. However, currently available techniques have notbeen widely applied to the diagnosis of human viral infec-tions. Limitations of these techniques include the reliance onhigh-energy isotopic probes, the inefficient binding of targetnucleic acids to solid-energy probes, the inefficient bindingof target nucleic acids to solid-phase surfaces, the prolongedincubation times required to achieve maximal sensitivity,and the inability to objectively quantify results in a practicalmanner.To overcome the problems of current hybridization tech-

niques, methods have been developed for the linkage tonucleic acids of biotin and other markers which can bemeasured by means of enzymatic and fluorescence detectionsystems (4, 5, 8, 9, 15, 17, 18). The development of noniso-topic probes, however, solves only one of the problemsassociated with the application of hybridization assays forthe detection of viral agents in biological samples. We havealso developed assay formats in which hybridization reac-tions can be performed in a liquid phase environment toenhance the rate of the reactions and to minimize theinterference due to extraneous material present in biologicalfluids. A second objective of our work has been to developassays that generate products that can be quantified bymeans of simple instrumentation. In particular, we havebeen interested in assays that can be performed in standard96-well microdilution plates, since instrumentation for theperformance and quantitation of reactions in microdilution

* Corresponding author.

plates is widely available and since clinical microbiologistsand researchers are familiar with this format. In this reportwe describe a novel hybridization assay system whichaddresses many of the problems encountered with currentlyavailable techniques.

MATERIALS AND METHODS

Nucleic acid reagents. The recombinant plasmid pGEM-ARV containing a 2.9-kilobase fragment spanning nearly allof the polymerase gene and the 3' end of the gag gene of amolecular clone of the LAV strain of human immunodefi-ciency virus type 1 (HIV-1) was the generous gift of HowardGendelman and Malcolm Martin of the Laboratory of Mo-lecular Microbiology, National Institute of Allergy and In-fectious Diseases. The recombinant plasmid pCBIII21 con-taining a subgenomic cDNA fragment of the coxsackievirusB3 (Nancy) genome was kindly provided by Steve Tracey.The plasmids were amplified by bacterial growth, recoveredby lysis of cellular pellets, and purified by CsCl-ethidiumbromide density centrifugation by established methods (12).

Traces of residual RNA in the purified plasmid prepara-tions were degraded by alkali denaturation as follows. A25-1tl aliquot of 0.15 M NaOH was added to 50 ,ul of a0.5-mg/ml solution of the plasmid in 10 mM Tris hydrochlo-ride (pH 7.4)-i mM EDTA (TE buffer). After incubation at37°C for 16 h, the solution was neutralized with 25 pI of 1 MHCl and 25 ,ul of 1 M Tris hydrochloride (pH 8). A 12.5-,usample of 3 M sodium acetate (pH 7) was added, and theDNA was precipitated with 3 volumes of cold 95% ethanol.The dried pellet was dissolved in TE buffer.

Biotinylation of probes. The plasmids were labeled withbiotin by the sodium bisulfite-ethylenediamine method withconditions previously shown to yield the optimum biotincontent for efficient hybridization of the probe to targetsequences (18). Briefly 1 M sodium bisulfite-3 M ethylene-diamine adjusted to pH 6 with concentrated HCI was freshlyprepared, and hydroquinone was added to a final concentra-tion of 1 mg/ml. Since the bisulfite-catalyzed transaminationof cytosine residues is single strand specific, the probe washeat denatured and cooled on ice before the reaction was

120

on May 9, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

HIV MONOCLONAL SOLUTION HYBRIDIZATION ASSAY

initiated by the addition of 9 volumes of the bisulfite-aminesolution to 1 volume of single-stranded nucleic acid. Thereaction mixture was incubated for 3 h at 42°C and thenextensively dialyzed against 5 mM sodium phosphate buffer(pH 8.5). Amine-modified probe nucleic acid in 0.1 Msodium phosphate buffer (pH 8.5) was biotinylated by incu-bation for 1 h at room temperature with 10 mM biotinyl-e-aminocaproic acid N-hydroxysuccinimide ester. The esterwas freshly prepared as a 0.2 M stock solution in N',N'-dimethylyformamide and added dropwise to the modifiednucleic acid to start the reaction. Unreacted biotin ester wasremoved by extensive dialysis against 150 mM NaCl-10 mMsodium phosphate buffer (pH 7)-i mM EDTA. Labeledprobes were stored at 4°C until used.

Extraction of nucleic acids. A cell culture suspension (240,ul) containing approximately 1 x 106 to 5 x 106 cells ml wasfrozen and thawed three times in an Eppendorf tube to lysethe cells. Proteinase K and sodium dodecyl sulfate (SDS)were added to final concentrations of 200 ,ug/ml and 0.5%,respectively, and the reaction mixture was incubated for 1 hat 37°C. The sample was extracted with an equal volume ofphenol and then with an equal volume of chloroform-isoamylalcohol (24:1). Sodium acetate was added to the final aque-ous layer to a 0.3 M concentration, and 3 volumes of cold95% ethanol was added. Nucleic acids were precipitated at-70°C for 30 min and then were recovered by centrifugationat 10,000 x g for 10 min at 4°C. The dried pellet wassuspended in 120 ,ul of diethylpyrocarbonate-treated distilledwater containing 0.5% SDS to protect extracted RNA fromdegradation by RNases. Serial dilutions were prepared in thesame buffer.Monoclonal antibody solution hybridization assay. Hybrid-

ization buffer (120 ,uI) containing 1 ptg of biotinylated probeper ml in 4x SSC (lx SSC is 150 mM NaCl plus 15 mMsodium citrate [pH 7])-20 mM HEPES (N-2-hydroxyethyl-piperazine-N'-2-ethanesulfonic acid) buffer (pH 7)-2 mMEDTA was added to an equal volume of the samples. Themixtures were heated to 100°C for 3 min to denature theprobe, and the hybridization reaction was performed for 3 hat 68°C in a water bath. After completion ofthe hybridizationreaction, the mixtures were cooled to room temperature, and24 pl of 10% Triton X-100 was added. The addition of TritonX-100 was necessary to form micelles containing SDS andthus prevent the high SDS concentration from desorbingantibody from the coated microdilution plates used in thenext step of the assay.For the immunoreaction 100 pli of the hybridized sample

was added to duplicate wells of a black polystyrene 96-wellmicrodilution plate (microFluor 'B' U-plate; Dynatech no.011-010-7201). Black plates were used to minimize nonspe-cific fluorescence. The plates were prepared by incubationovernight at 4°C with 100 pl ofgoat antibody to biotin (SigmaChemical Co.) diluted to 1 ,ug/ml in 0.06 M carbonate buffer(pH 9.6). Before use the plates were washed five times withphosphate-buffered saline (PBS) containing 0.05% Tween 20(PBST). The hybridized samples were incubated on the plateovernight at room temperature. After five washes withPBST, 100 pul of,-D-galactosidase-labeled Fab' fragments ofa moncolonal antibody to DNA-RNA duplexes diluted to0.065 ,ug/ml in PBST-0.5% gelatin-0.5% normal mouseserum was added to the wells. The P-D-galactosidase-labeledFab' fragments of the monoclonal antibody to DNA-RNAhybrids was the generous gift of Robert J. Carrico, MilesLaboratories, Inc., Elkhart, Ind. The binding properties ofthe monoclonal antibody and the preparation of the labeledFab' fragments have been described previously (2, 21). After

an incubation for 2 h at 37°C, the plate was again washed fivetimes with PBST, and 100 pil of enzyme substrate solution-0.1 mM 4-methylumbelliferyl-p-D-galactoside in PBS con-taining 1 mM MgCl2-50 ,ug of bovine serum albumin wasadded. The plate was incubated for 40 min at room temper-ature, and the amount of fluorescent umbelliferone gener-ated by the enzymatic reaction was measured in a microplatefluorometer (Dynatech Microfluor). A positive test wasdefined as a mean fluorescence activity greater than themean fluorescence activity plus three standard deviations ofa negative culture fluid.Dot hybridization assay. A cell culture sample (400 ,ul) was

digested with proteinase K, extracted with phenol-chloro-form, and precipitated with ethanol as described above. Thedried pellet was suspended in 100 ,ul of 10 mM Tris hydro-chloride (pH 7.4)-i mM EDTA. Three volumes of 10x SSC-6.15 M formaldyde was added, and the sample was heated to65°C for 15 min. Serial dilutions were prepared in 10x SSC,and 100 pul was applied to a nitrocellulose filter (BA-85;Schleicher & Schuell Co.) by using a 96-well manifoldapparatus. The filter was air dried and baked in a vacuumoven at 80°C for 30 min. Prehybridization was performed for4 h at 37°C in 50% formamide-3x SSC-10x Denhardtsolution-0.1% SDS-0.1 mM EDTA-10 mM HEPES buffer(pH 7.4)-100 ,ug of salmon sperum DNA per ml. Thehybridization reaction was performed overnight at 37°C inthe same solution containing, in addition, 10% dextransulfate and the heat-denatured pGEM-ARV plasmid probelabeled with 32P by nick translation (2.5 x 106 CPM/ml ofhybridization buffer; specific activity, 5 x 107 cpm/,ug). Atthe completion of the hybridization reaction the filters werewashed three times each with 2x SSC-0.1% SDS (10 min atroom temperature), O.lx SSC-0.1% SDS (30 min at 45°C),and 0.1 x SSC (10 min at room temperature). Autoradiogra-phy was performed for 6 days at -70°C with Kodak XAR-5film and Cronex quanta III intensifying screens.HIV cultures. In vitro viral isolation was accomplished by

a modification of the method recommended by the Centersfor Disease Control. Peripheral mononuclear cells fromstudy subjects were separated from 30 ml of heparinizedblood on Ficoll-Hypaque gradients and washed three timesin PBS. Then 107 cells were suspended in 250-cm2 cultureflasks in 30 ml of RPMI 1640 medium containing 10% fetalcalfserum, 2 mM glutamine, 1,000 U ofpenicillin per ml, 100,ug of streptomycin per ml, and 1% interleukin-2. Lympho-cytes were routinely cocultivated with 3 x 107 normalhuman mitogen-stimulated peripheral mononuclear cells.These normal peripheral mononuclear cells were obtainedfrom healthy volunteer donors who were negative for anti-body to HIV and hepatitis B virus. They were treated with 3,ug of PHA-P (Difco Laboratories) per ml for 72 h and thenwashed before being used for virus culture. Twice a week 5x 10" PHA-P-stimulated normal human peripheral mononu-clear cells from a different donor, suspended in 10 ml offreshmedium, were added to the culture. Cultures were assayedtwice a week for the presence of HIV antigens by immu-noassay and HIV nucleic acids by hybridization assay.Samples for testing were drawn from the lower layer of theculture medium. The suspensions, containing cellular debris,were heated to 56°C for 30 min and then stored at -20°Cuntil tested. A culture was considered positive if HIVantigens were detected in two or more cell culture samples.Enzyme immunoassays for HIV antigens. Black polysty-

rene microdilution plates (Dynatech Laboratories, Inc., Alex-andria, Va.) were coated with 100 pil of human anti-HIVserum diluted 1:20,000 in 0.06 M carbonate buffer (pH 9.6)

VOL. 27, 1989 121

on May 9, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

122 VISCIDI ET AL.

0 V....5,.........>.. tin

Hybridizotion ofRNA target withbiotinyloted DNA probe

Incubation ononti-biotin pla

)-goloctosidos

e...biotin

Reaction withte 6-goloctosidose

labeled monoclonalantibody to RNA-DNA hybrids

Substrote--* PRODUCT4e )-galoctosidose

. ...

biotin

Addition offluorescentsubstrate

FIG. 1. Scheme of the monoclonal antibody solution hybridization assay.

per well. Control wells were coated with normal humanserum. The plates were incubated overnight at 4°C. Beforeuse and after each incubation step, the plates were washedfive times with PBST; 0.1 volume of 0.5% Tween 20 wasadded to cell culture samples, and 100 ,u was added to eachof two test and control wells. The plates were incubatedovernight at room temperature. Biotinylated human anti-HIV immunoglobulins (100 iil; 1 Fg/ml) in PBST-0.5%gelatin containing 0.5% normal human serum was added toall the wells, and the plates were incubated for 2 h at 37°C.Biotinylated immunoglobulin bound in the preceding stepwere detected in a 1-h incubation at 37°C with 100 ,ul of acomplex of strepavidin (0.5 p.g/ml) and biotinylated P-D-galactosidase (0.125 U/ml; Vector Laboratories Inc., Bur-lingame, Calif.) in PBST-0.5% normal human serum. Thesubstrate, 0.1 mM 4-methylumbilliferyl 3-D-galactoside inPBS containing 1 mM MgCl and 50 Fg of bovine serumalbumin per ml, was then added to the wells, and theenzymatic activity was measured in a microfluorometer. Thenet fluorescence activity was calculated by subtracting themean activity of the control wells from the mean activity ofthe test wells. A positive test was defined as a net fluores-cence activity greater than the net fluorescence activity plusthree standard deviations of a negative culture fluid.The Abbott Diagnostics antigen-capture assay for the

detection of HIV antigens was performed according to themanufacturer's instructions for the overnight protocol.

RESULTS

The design of the monoclonal antibody solution hybridiza-tion assay is depicted in Fig. 1. The assay format was basedon a hybridization reaction in solution between a biotinyla-ted DNA probe and complementary viral RNA target se-quences. Labeled hybrids were then bound in an immuno-reaction to the wells of a microdilution plate coated with ananti-biotin antibody. After removal of unbound nucleic acidsby washing, the amount of probe-virus nucleic acid hybridsbound to the solid phase was measured by reaction withenzyme-labeled monoclonal antibody directed at DNA-RNAhybrids and the appropriate enzyme substrate. We foundthat this format could detect as little as 1 pg (0.05 pg per50-.l reaction volume) of target RNA per ml with a modelsystem composed of complementary homopolymeric nucleicacids (Fig. 2).We applied the monoclonal antibody solution hybridiza-

tion assay for the detection of HIV RNA in cell culturesamples. The optimum conditions for the assay were deter-mined in preliminary experiments with a cell culture sampleinoculated with a wild-type HIV isolate and a control sample

from an uninoculated culture. Nucleic acids were extractedfrom lysed cells, and a solution hybridization reaction wasperformed for 3 h with 0.5 p.g of a biotinylated probe per ml.The incubation temperature, which was 24°C below thecalculated thermal denaturation temperature (Tm) of a per-fectly matched hybrid, constitutes a moderately high strin-gency hybridization reaction for DNA-RNA hybrids. Al-though the signal could be increased with higher probeconcentrations and longer hybridization times, this did notalter the sensitivity of the assay for detection of HIV RNA(data not shown). Labeled hybrids formed between thebiotinylated DNA probe and viral RNA sequences were

2.1Flw Units Lu 10 A

-_ te. m tu n le 5.1 I *.u .1 u et

Concentration pg/el- ly Ig Cslo

Fluo Ibits Lo u B

c.cc.a\

c.c .

8.4 ..

1.2 ..

lm lm 33 O 3.3 1 0.33 0.1 onCtUConcentration pg/Il-r l, Iet E2 ply à

FIG. 2. Detection of poly(I) and poly(A) with biotinylatedpoly(dC) in a monoclonal antibody solution hybridization assay. (A)Titration of poly(I) from 1,000 to 0.1 pg/ml. (B) Titration of poly(I)from 33 to 0.1 pg/ml.

J. CLIN. MICROBIOL.

on May 9, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

HIV MONOCLONAL SOLUTION HYBRIDIZATION ASSAY

A

2 3 4 5RECIPROCAL DILUTION (loglo)

*1s-.0 1.5 2.0

'R-.C-irPROCCAi DILUTION (LOG10)

TABLE 1. Comparison of monoclonal antibody solutionhybridization assay and antigen-capture enzyme immunoassay for

detection of HIV in 436 cell culture samples

Monoclonal antibody No. of antigen-capture immunoassay resultssolution hybridization

assay result Positive Negative Total

Positive 116 17 133Negative 8 295 303Total 124 312 436

tive individuals. A total of 436 cell culture samples from 60separate in vitro isolation attempts from seropositive indi-viduals were tested in the monoclonal antibody solutionhybridization assay and in a highly sensitive enzyme immu-noassay for`the detection of HIV antigens (19). A compan-

son of the results obtained with the two assays for theindividual samples is shown in Table 1 and Fig. 4. Comparedwith results with the antigen-capture assay, the sensitivityand specificity of the hybridization assay for the 436 cellculture samples were 93.5 and 94.6%, respectively. Therewere eight samples that were positive in the antigen assay

and negative in the hybridization assay. These samples came6 from five cultures that were subsequently positive for the

presence of HIV nucleic acids a median of 4 days (range, 3to 11 days) after HIV antigens were detected. Converselythere were 17 samples that were positive in the hybridizationassay and negative in the antigen-capture assay. Six sampleswere from cultures that were positive for the presence ofHIV nucleic acids a median of 4 days (range, 3 to 17 days)before HIV antigens were detected. In general viral nucleicacids and antigens were detected simultaneously from posi-tive cultures. Nine discordant samples were accounted forby the initial samples from nine separate cultures. The firstsample from a culture was obtained after the patient'slymphocytes had been incubated for 3 to 4 days. For seven

FIG. 3. Detection of HIV nucleic acids in serial dilutions of cellculture samples. (A) Monclonal antibody solution hybridizationsamples: *, positive culture; O, negative culture. (B) Dot hybrid-ization assay: 1, positive culture; 2, negative culture.

detected in a sandwich immunoassay with a solid-phase goatanti-biotin antibody and ,B-D-galactosidase-labeled Fab' frag-ments of a monclonal antibody to DNA-RNA duplexes. Theassay was capable of detecting viral RNA at a i0-' dilutionof a cell culture inoculated with an HIV isolate (Fig. 3A).The sensitivity of the assay was compared with that of a dothybridization assay with a 32P-labeled probe. The nucleicacids from extracted cell culture samples were denaturedwith formaldehyde and immediately applied to a nitrocellu-lose filter with a 96-well manifold apparatus. Hybridizationwas performed with the pGEM-ARV probe labeled with 2p

(specific activity, 5 x 107 cpmlug). Viral RNA could only bedetected at a 10-2 dilution of a positive culture on an

autoradiogram developed after 6 days (Fig. 3B). The cellculture sample inoculated with an HIV isolate was alsoassayed for the presence of HIV proteins by the enzymeimmunoassay for HIV antigens and by a commercial anti-gen-capture assay (Abbott Laboratories) (7). The limit ofdetection of viral antigens in both assays was a 10-3 5

dilution.We then evaluated the performance of the monoclonal

antibody solution hybridization assay to detect virus inperipheral mononuclear cells cultured from HIV-seroposi-

10,000 r

1000

w

usz

w

Cro

-Jwa

100

I0

__

3o

*-0000

oe e.:S"

"O..S.

f-

__à#r 30

s

ANTIGEN ANTIGENPOSITIVE NEGATIVE

FIG. 4. Comparison of monoclonal antibody solution hybridiza-tion assay and antigen-capture enzyme immunoassay for detectionof HIV in 436 cell culture samples. The results for the monoclonalantibody solution hybridization assay are expressed as the meanfluorescence activity of the sample minus the mean fluorescenceactivity plus three standard deviations of a negative culture fluid.

10,000

1000Uf)H

z70

L1JzhLJw

C/)!U 100

70-J1-

10

VOL. 27, 1989 123

on May 9, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

124 VISCIDI ET AL.

TABLE 2. Comparison of monoclonal antibody solutionhybridization assay and antigen-capture enzyme

immunoassay for detection of HIV in 60 cell culturesfrom HIV-seropositive individuals

Monoclonal antibody No. of antigen-capture immunoassay resultssolution hybridization

assay result Positive Negative Total

Positive 29 8 37Negative 0 23 23Total 29 31 60

of these cultures the first sample tested was the only positivesample. For two cultures, although the initial positive sam-

ple was followed by samples that were negative in thehybridization assay, the cultures became positive by boththe hybridization assay and the antigen-capture assay afteran additional 14 to 17 days of cultivation.A comparison of the results obtained with two assays, the

monoclonal antibody solution hybridization assay and theantigen-capture immunoassay, for the 60 cultures is shownin Table 2. HIV antigens were detected by enzyme immu-noassay in one or more of the samples from 29 cultures. Thenumber of positive samples obtained from each culture wasas follows. Three or more positive samples were obtainedfrom 23 of the 29 cultures, two positive samples wereobtained from 3 cultures, and a single positive sample wasobtained from the remaining 3 cultures. HIV nucleic acidswere detected by the monoclonal antibody solution hybrid-ization assay in all 29 cultures that were positive by antigendetection. Three or more positive samples were obtainedfrom 22 of the 29 cultures, two positive samples wereobtained from 5 cultures, and a single positive sample wasobtained from the remaining 2 cultures. Of the 31 antigen-negative cultures, 8 cultures were scored as positive by thehybridization assay based on the results of a single positivesample. For seven of these cultures the positive sample wasthe first one tested, and no additional positive results wereobtained on multiple, subsequent samplings of the cultures.None of the samples that were positive in the HIV hybrid-ization assay but negative for the presence of HIV antigensreacted with pCBIII21, a recombinant pBr322 plasmid con-

taining a cDNA fragment of the coxsackie B3 genome.

DISCUSSION

For hybridization assays to achieve their potential as

practical methods for the diagnosis of viral pathogens,systems must be developed that overcome the limitations ofexisting technology. The format of the monoclonal antibodysolution hybridization assay involves a step in which a

hybridization reaction is performed in solution with a non-isotopic probe and a second step in which labeled DNA-RNA hybrids are detected in a solid-phase immunoreactionsimilar to a traditional enzyme immunoassay. The advantageof such an assay design for viral diagnosis is that it combinesthe high sensitivity and specificity of a hybridization reactionwith the technical conveniences of widely available enzymeimmunoassay systems.The monoclonal antibody solution hybridization assay is

shown here to be applicable for the detection of retroviralRNA. When all of the tested samples are considered, theconcordance between the results with an antigen-captureassay and the hybridization assay was approximately 95%.The time of appearance of detectable viral antigens andnucleic acids coincided for most viral cultures. The discrep-

ancies may reflect variations in sampling or storage condi-tions; however, a more detailed study of the kinetics ofnucleic acid production and gene expression may revealsignificant differences for some isolates. There were eightcultures that were positive by the hybridization assay butnegative by the antigen-capture assay. For seven of thesecultures the only positive sample was the first one tested.These results may reflect the detection of virus that waspresent in the patient's lymphocytes but failed to replicate invitro. Virus is often not recovered by current cultivationmethods from individuals who are known to be infected andsuspected to carry the virus (6, 10, 11, 13). The presence ina patient's lymphocytes of enough virus to be detected byour hybridization assay before viral replication in vitro issupported by the observation that the hybridization assayalso detected virus in the initial samples from two cultureswhich became positive by both the hybridization and theantigen-capture assays after an additional 14 to 17 days ofcultivation. In addition, the ability of other hybridizationtechniques to detect HIV directly in lymphocytes frominfected patients has been reported (3, 14, 16; J. J. Sninsky,S. Kwok, D. Mack, G. Ehrlich, P. Ullrich, B. Poies, et al.,Program Abstr. 3rd Int. Conf. AIDS, F.2.3., p. 207, 1987).The possibility that these results were due to false-positivehybridization reactions cannot be excluded. However, webelieve that this is unlikely because none of the positivesamples reacted with pCBIII21, a recombinant pBr322 plas-mid containing a cDNA fragment of the coxsackievirus B3genome. The specificity of the assay is also supported by theconsistently low level of reactivity observed with a samplefrom a culture of a pool of peripheral mononuclear cells fromseveral seronegative donors.

It is of note that this assay format yields objective resultsand can allow for the performance of quantitative controlreactions. These features are important for the detection ofagents, such as HIV, the diagnosis of which has seriousconsequences for the infected individual. Although themethod that we used required the cultivation of patients'lymphocytes, preliminary analyses reveal that HIV RNAcan also be detected directly in the lymphocytes of infectedindividuals. The sensitivity and specificity of these determi-nations are the subject of current investigations.We have also adapted this format for the detection ofRNA

from double-stranded RNA viruses such as rotaviruses,positive-stranded RNA viruses such as picornaviruses, andnegative-stranded RNA viruses such as influenza viruses(R. P. Viscidi and R. H. Yolken, Abstr. VII Int. Congr.Virol., abstr. no. R.30.5, 1987). Furthermore, it is possiblethat this method could be adapted to the detection of nucleicacids from DNA viruses such as herpesviruses and humanpapillomaviruses by the use of biotin-labeled single-strandedRNA probes and the measurement ofDNA-RNA hybrids bythe above method. Monoclonal antibody solution hybridiza-tion assays thus have the potential of becoming widely usedsystems for the detection of a broad range of microbialnucleic acids in biological specimens.

ACKNOWLEDGMENTSThis work was supported by Public Health Service contract

AI-62556 and grant AI-00625 from the National Institute of Allergyand Infectious Diseases.

LITERATURE CITED1. Allain, J.-P., Y. Laurian, D. A. Paul, D. Senn, and members of

the AIDS-Haemophilia French Study Group. 1986. Serologicalmarkers in early stages of human immunodeficiency virus

J. CLIN. MICROBIOL.

on May 9, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

HIV MONOCLONAL SOLUTION HYBRIDIZATION ASSAY

infection in haemophiliacs. Lancet ii:1233-1236.2. Boguslawski, S. J., D. E. Smith, M. A. Michalak, K. E. Mickel-

son, C. O. Yehle, W. L. Patterson, and R. J. Carrico. 1986.Characterization of monoclonal antibody to DNA-RNA and itsapplication to immunodetection of hybrids. J. Immunol. Meth-ods 89:123-130.

3. Carter, W. A., D. R. Strayer, I. Brodsky, M. Lewin, M. G.Pellegrino, L. Einck, H. F. Henriques, G. L. Simon, D. M.Parenti, R. G. Scheib, R. S. Schulof, D. C. Montefiori, W. E.Robinson, W. M. Mitchell, D. J. Volsky, D. Paul, H. Paxton,W. A. Meyer, Kariko, N. Reichenbach, R. J. Suhadolnik, andD. H. Gillespie. 1987. Clinical, immunological, and virologicaleffects of ampligen, a mismatched double-stranded RNA, inpatients with AIDS or AIDS-related complex. Lancet ii:1286-1292.

4. Chollet, A., and E. H. Kawashima. 1985. Biotin-labeled syn-thetic oligodeoxyribonucleotides: chemical synthesis and usesas hybridization probes. Nucleic Acids Res. 13:1529-1541.

5. Forster, A. C., J. L. McInnes, D. C. Skingle, and R. H. Symons.1985. Nonradioactive hybridization probes prepared by thechemical labelling of DNA and RNA with a novel reagent,photobiotin. Nucleic Acids Res. 13:745-761.

6. Gallo, R. C., S. Z. Salahuddin, M. Popovic, G. M. Shearer, M.Kaplan, B. F. Haynes, T. J. Palker, R. Redfield, J. Oleske, B.Safai, G. White, P. Foster, and P. D. Markham. 1984. Frequentdetection and isolation of cytopathic retroviruses (HTLV-III)from patients with AIDS and at risk for AIDS. Science 224:500-502.

7. Goudsmit, J., D. A. Paul, J. M. A. Lange, H. Speelman, J. VandDer Noordaa, H. J. Van Der Helm, F. DeWolf, L. G. Epstein,W. J. A. Krone, E. C. Wolters, J. M. Oleske, and R. A.Coutinho. 1986. Expression of human immunodeficiency virusantigen (HIV-Ag) in serum and cerebrospinal fluid during acuteand chronic infection. Lancet ii:177-180.

8. Kempe, T., W. I. Sundquist, F. Chow, and S-L Hu. 1985.Chemical and enzymatic biotin-labeling of oligodeoxyribonucle-otides. Nucleic Acids Res. 13:45-57.

9. Langer, P. R., A. A. Waldrop, and D. C. Ward. 1982. Enzymaticsynthesis of biotin-labeled polynucleotides: novel nucleic acidaffinity probes. Proc. Natl. Acad. Sci. USA 79:6633-6637.

10. Levy, J. A., A. D. Hoffman, A. D., S. M. Kramer, J. A. Landis,and J. M. Shimabukuro. 1984. Isolation of lymphocytopathic

retroviruses from San Francisco patients with AIDS. Science225:840-842.

11. Levy, J. A., and J. Shimabukuro. 1985. Recovery of AIDS-associated retroviruses from patients with AIDS or AIDS-related conditions and from clinically healthy individuals. J.Infect. Dis. 152:734-738.

12. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

13. Markham, P. D., S. Z. Salahuddin, M. Popovic, A. Patel, K.Veren, A. Fladager, S. Orndorff, and R. C. Gallo. 1985. Ad-vances in the isolation of HTLV-III from patients with AIDSand AIDS-related complex and from donors at risk. Cancer Res.45:4588s-4591s.

14. Pellegrino, M. G., M. Lewin, W. A. Meyer, R. S. Lanciotti, L.Bhaduri-Hauck, D. J. Volsky, K. Sakai, T. M. Folks, and D.Gillespie. 1987. A sensitive solution hybridization technique fordetecting RNA in cells: application to HIV in blood cells.BioTechniques 5:452-460.

15. Renz, M., and C. Kurz. 1984. A colorimetric method for DNAhybridization. Nucleic Acids Res. 12:3435-3444.

16. Richman, D. D., J. A. McCutchan, and S. A. Spector. 1987.Detecting human immunodeficiency virus RNA in peripheralblood mononuclear cells by nucleic acid hybridization. J. Infect.Dis. 156:823-827.

17. Tchen, P., R. P. P. Fuchs, E. Sage, and M. Leng. 1984.Chemically modified nucleic acids as immunodetectable probesin hybridization experiments. Proc. Natl. Acad. Sci. USA 81:3466-3470.

18. Viscidi, R. P., C. J. Connelly, and R. H. Yolken. 1986. Novelchemical method for the preparation of nucleic acids for non-isotopic hybridization. J. Clin. Microbiol. 23:311-317.

19. Viscidi, R. P., H. Farzadegan, F. Leister, M. L. Francsico, andR. H. Yolken. 1988. Enzyme immunoassay for detection of HIVantigens in cell cultures. J. Clin. Microbiol. 26:453-458.

20. Viscidi, R. P., and R. H. Yolken. 1987. Molecular diagnosis ofinfectious diseases by nucleic acid hybridization. Mol. Cell.Probe 1:3-14.

21. Yehle, C. O., W. L. Patterson, S. J. Boguslawski, J. P. Albarella,K. F. Yip, and R. J. Carrico. 1987. A solution hybridizationassay for ribosomal RNA from bacteria using biotinylated DNAprobes and enzyme-labeled antibody to DNA:RNA. Mol. Cell.Probes 1:177-193.

VOL. 27, 1989 125

on May 9, 2020 by guest

http://jcm.asm

.org/D

ownloaded from