JOURNAL OF CLINICAL MICROBIOLOGY,0095-1137/01/$04.0010 DOI: 10.1128/JCM.39.2.498–505.2001

Feb. 2001, p. 498–505 Vol. 39, No. 2

Copyright © 2001, American Society for Microbiology. All Rights Reserved.

Rapid Typing of Human Adenoviruses by a General PCRCombined with Restriction Endonuclease Analysis

ANNIKA ALLARD,* BO ALBINSSON, AND GORAN WADELL

Department of Virology, Umeå University, Umeå, Sweden

Received 30 May 2000/Returned for modification 8 September 2000/Accepted 13 November 2000

We have developed a system for rapid typing of adenoviruses (Ads) based on a combination of PCR andrestriction endonuclease (RE) digestion (PCR-RE digestion). Degenerated consensus primers were designed,allowing amplification of DNA from all 51 human Ad prototype strains and altogether 44 different genomevariants of Ad serotypes 1, 3, 4, 5, 7, 11, 19, 40, and 41. The 301-bp amplimer of 22 prototype strainsrepresenting all six subgenera and the genome variant was selected as a target for sequencing to look forsubgenus and genome type variabilities. The sequences obtained were used to facilitate the selection of specificREs for discrimination purposes in a diagnostic assay by following the concept of cleavage or noncleavage ofthe 301-bp amplimer. On the basis of these results, a flowchart was constructed, allowing identification ofsubgenus B:2 and D serotypes and almost complete distinction of subgenus A, B:1, C, E, and F serotypes.Application of the PCR-RE digestion system to clinical samples allowed typing of 34 of 40 clinical samplespositive for Ad. The genome type determined by this method was identical to that obtained by traditional REtyping of full-length Ad DNA. The remaining six samples were positive only after a nested PCR. Therefore, toreduce the risk of false-negative results, samples scored negative by the PCR-RE digestion system should beevaluated by the described nested PCR. Used in combination, the PCR-RE digestion method and the nestedPCR provide a reliable and sensitive system that can easily be applied to all kinds of clinical samples whenrapid identification of adenoviruses is needed.

The 51 different serotypes of human adenoviruses (Ads) areclassified into six subgenera (subgenera A to F) on the basis ofseveral biochemical and biophysical criteria (33, 49). Typing ofAds has so far mainly been of epidemiological interest. How-ever, the improved knowledge about the differences in viru-lence among the several types has increased the medical valueof typing. Cases of severe acute respiratory illness and alsofebrile illness with cardiopulmonary failure in infants andyoung children have been described (9, 30). These syndromeshave frequently been associated with subgenus B Ads, prefer-entially genome variants of Ad type 7, with a high rate ofmortality. In adults Ad serotype 2 infections have been shownto be important in the pathogenesis of left-ventricle failure(34). Ads are also among the agents that take advantage of animpaired or destroyed immune system to set up persistent andgeneralized infections in the immunocompromised host, infec-tions that sometimes result in death (20).

Diagnosis of Ad infections is currently based on virus isola-tion in cell culture, antibody studies, or antigen detection byimmunofluorescence (50). However, the need for rapid andsensitive detection methods has led to PCR being the most wellestablished among all other methods. Primer systems for de-tection of Ads in general have frequently been used during thelast 10 years (5, 6, 10, 31, 32, 35). Ad serotyping is based onneutralization or hemagglutination inhibition (17), but it canalso be done by sequencing (29, 44). Genome typing can bedone by restriction endonuclease (RE) analysis of full-lengthAd DNA (2, 47) and with subgenus- or type-specific PCRprimers (5, 24, 35, 37, 38). Recently, different methods have

suggested how RE analysis and PCR can be combined tofacilitate the typing procedure (7, 24, 39, 45). However, thesemethods have been incomplete or limited, with results some-times difficult to interpret. This prompted us to develop a moreextensive PCR-RE digestion typing method with a simple andclear-cut final readout. On the basis of hexon sequence data,we have developed new primers corresponding to a conservedregion of the hexon gene upstream of the surface loop l1. Thesequences of the Ad products framed by these primers wereheterogeneous enough to allow discrimination between sub-genera and even between serotypes by RE cleavage. We de-scribe here a flowchart of RE digestions that can be used withnonnested PCR products for typing of human Ads. This typingsystem can be valuable when it is of importance to excludetypes with more pronounced virulence, such as the members ofsubgenus B and Ad type 2 (Ad2) of subgenus C, but also Ad31of subgenus A, which have frequently been isolated from im-munocompromised hosts (20). In addition, the flowchart offersfurther possibilities for discrimination of more virulent ge-nome variants of types 3 and 7 from among prototype strains.The method is intended for characterization of Ads both inclinical samples and in cell culture fluids.

MATERIAL AND METHODS

Virus strains. All prototype strains except those of Ad16, Ad40, Ad41, andAd48 to Ad51 were originally obtained from the American Type Culture Col-lection (ATCC). The prototype strain of Ad16 was a gift from R. Wigand,Homburg, Germany. Ad40 reference strain Hovi-X and Ad41 prototype strainTak were originally characterized at Bilthoven, The Netherlands (12). D. Schnurrof the Viral and Rickettsial Disease Laboratory, Berkeley, Calif., kindly donatedstrains of Ad48 and Ad49 (40). Candidate Ad strains of Ad50 and Ad51 (13)were kindly donated by J. C. deJong, Erasmus University, Rotterdam, TheNetherlands. The genome variants Ad1D7, Ad1D10, Ad2D5, Ad2D6, Ad2D7,Ad2D25, Ad2D36, Ad2D63, and Ad5D38, classified by the system of Adrian et

* Corresponding author. Mailing address: Department of Virology,Umeå University, S-901 85 Umeå, Sweden. Phone: 46-90-785 2815.Fax: 46 90 129905. E-mail: Annika.Allard@climi.umu.se.

498

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

al. (1), together with Ad7b, Ad7c, Ad7h, Ad7i, and Ad7j, were kindly donated byA. Kajon (22, 23, 30). Ad genome variants 3a, 3a2, 3c, and 3d (28), 4a, 4a1, 4b,4p1, 4p2, and 4p3 (26), and 11a, 11b, and 11c (27) were kindly donated by Q. G.Li, Umeå, Sweden. Ad19a was described by Wadell and de Jong (48). Genomevariants of Ad40 (D2, D5, D9, and D11) and Ad41 (D2, D6, D10, D11, D12,D13, D14, D15, D18, D22, and D24), classified by van der Avoort et al. (46), werekindly donated by Alistair Kidd, Umeå, Sweden. Ad41 strain D389 was describedby Allard et al. (4).

Viral isolation. Ad isolates were obtained by inoculation of tubes of A549 cellsor 293 cells (Ad40 and Ad41) with prototype strains or clinical specimens. Cellswere grown in Dulbecco’s minimum essential medium supplemented with 5%fetal calf serum, glutamine, and antibiotics and maintained with Dulbecco’sminimum essential medium supplemented with 2% fetal calf serum. Stocks ofeach virus isolate were grown in the cells and kept frozen at 270°C for furtheranalysis.

DNA extraction. Each isolate was used to inoculate confluent monolayers ofA549 or 293 cells in 75-cm2 plastic flasks. When an extensive cytopathic effectwas evident (48 to 72 h postinfection), intracellular full-length viral DNA wasextracted as described by Shinagawa et al. (42). DNA to be used in PCR andPCR-RE analysis was purified by the QIAamp Blood Mini kit protocol (QIA-GEN GmbH, Hilden, Germany).

Clinical samples. Forty specimens confirmed to be Ad positive by cell culture,immunofluorescense, or PCR in routine diagnostic work during the period fromNovember 1994 to April 1995 were included in a retrospective study. Materialsanalyzed included 13 stool samples, 9 eye and 3 throat swab samples, 7 naso-pharyngeal aspirate samples, 1 bronchoalveolar lavage sample, 3 cerebrospinalfluid (CSF) samples, 1 serum sample, 1 urine sample, and 2 vesicle materials(Table 1). The specimens were stored at 270°C until viral DNA was purified withthe QIAamp Blood Mini kit.

PCR amplification. The outer primer pair, hex1deg (59-GCC SCA RTG GKCWTA CAT GCA CAT C-39) and hex2deg (59-CAG CAC SCC ICG RAT GTCAAA-39), which created a 301-bp product, was used for both sequencing anddiagnosis. The nested primer pair, nehex3deg (59-GCC CGY GCM ACI GAIACS TAC TTC-39) and nehex4deg (59-CCY ACR GCC AGI GTR WAI CGMRCY TTG TA-39), produced an amplimer of 171 bp. All primer sequences arefound between base pair position 21 and position 322 in the coding region of thehexon gene (3). One-step amplifications were carried out in 100-ml reactionmixtures containing 50 mM KCl, 10 mM Tris-HCl (pH 9.0), 3 mM MgCl2, 0.1%Triton X-100, each deoxynucleoside triphosphate at a concentration of 100 mM,each primer at a concentration of 0.5 mM, and 1 U of Taq DNA polymerase(Promega Corporation, Madison, Wis.). A total of 30 ml of QIAgen-eluted DNAwas added to each reaction mixture. The reaction tubes were placed in an MJResearch PTC-200 thermal cycler and were held at 94°C for 3 min, immediatelyfollowed by 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min. The finalcycle had a prolonged extension time of 5 min. One-tenth of the PCR mixturewas subjected to nested PCR in an identical mixture but with nested primers.After amplification was completed, 10 ml of the single or the nested PCR productwas electrophoresed in 2% NuSieve GTG plus 1% SeaKem ME agarose (FMCBioproducts, Rockland, Maine) in 0.53 Tris-borate-EDTA buffer (pH 8.0) at 10V/cm for 90 min. The bands were visualized by staining with 0.25 mg of ethidiumbromide per ml and inspection under UV light.

RE analysis. (i) Amplified products. Aliquots of 5 to 15 ml of the one-stepPCR mixture containing approximately 1 mg of the amplimer were digested for2 to 3 h with 5 U of different endonucleases under conditions specified by themanufacturers (New England Biolabs, Beverly, Mass., and Promega Corpora-tion). Enzyme digests were analyzed on 2% NuSieve GTG plus 1% SeaKem MEagarose gels (FMC Bioproducts) in 0.53 Tris-borate-EDTA buffer (pH 8.0) at 6V/cm for 2 to 3 h.

(ii) Genomic adenovirus DNA. A total of 2 mg of full-length genomic DNAfrom each viral DNA preparation was used for specific Ad typing. The DNA wasdigested with 10 U of the restriction enzymes BamHI, SmaI, and EcoRI or BglIIfor 4 to 5 h and electrophoresed in 1% SeaKem ME agarose at 2 V/cm for 16 h.

Sequencing of PCR products. The one-step PCR products of unsequenced Adtypes 1, 6, 8, 9, 10, 11, 13, 14, 18, 19, 19a, 21, 31, 34, 35, 37, and 44, together withpreviously sequenced Ad types 3, 4, 7, 12, 40, and 41, were separately ligated intothe pT7 Blue T-vector (Novagene) and transformed to Novablue competentcells. Plasmids from positive recombinants were purified by the QIAgen midipurification protocol. The Pharmacia AutoRead Sequencing kit was used for thesequencing reactions, together with 59 fluorescein-labeled primers (T7 promoterand U19 reverse primers). The sequencing reactions were loaded on a 6% 7 Murea acrylamide gel on a Pharmacia Automated Laser Fluorescent DNA se-quencer. The MegAlign program of DNAStar (DNAStar Inc., Madison, Wis.)was used for sequence alignment and to infer phylogenesis.

Nucleotide sequence accession numbers. The GenBank accession numbers forthe new sequences reported from this article are as follows: AF161559 for Ad1,AF161560 for Ad6, AF161561 for Ad8, AF161562 for Ad9, AF161563 for Ad10,AF161564 for Ad13, AF161565 for Ad19, AF161566 for Ad19a, AF161567 forAd37, AF161568 for Ad44, AF161569 for Ad4, AF161570 for Ad11, AF161571for Ad14, AF161572 for Ad21, AF161573 for Ad34, AF161574 for Ad35,AF161575 for Ad18, and AF161576 for Ad31.

RESULTS

Primers. New degenerate primers were designed by modi-fying already described oligonucleotides (5, 6) correspondingto a conserved region within the hexon gene. By comparingsequence data from previously published sequences with newlypublished sequences of Ad3, Ad7, and Ad16 of subgenus B(36), together with Ad12 (43) and Ad48 (11) of subgenus Aand subgenus D, respectively, the new outer primers (primershex1deg and hex2deg) were designed. Degenerate base posi-tions were introduced into the primer sequences to allow an-nealing to all 51 known prototypes including the new serotypesAd50 and Ad51, together with 44 different genome types (datanot shown). The modification was done to increase the homol-ogy to the hexon nucleotide sequences of Ad subgenus Bmembers. The hexon primers used earlier (5) have been shownto require a lower annealing temperature for subgenus B com-pared to the temperatures required for the other subgenera,allowing amplification of this particular subgenus; this is mostlydue to mismatches at the 39 ends (9, 14, 31, 32). The nesteddegenerate primers were designed by comparing sequencedata from this work with the newly published sequence data forAd types 3, 7, 12, 16, and 48. These primers, nehex3deg andnehex4deg, were used in this study for confirmation of the Adtypes in six clinical specimens that were not positive by aone-step PCR but that created a fragment of 171 bp.

Sequencing results. The new degenerate hexon primers,hex1deg and hex2deg, were used to amplify DNAs of severalAd prototype strains for sequencing purposes. The one-stepPCR products of Ad types 12, 18, and 31 (subgenus A), Adtypes 3, 7, 11, 14, 21, 34, and 35 (subgenus B), Ad1 and Ad6(subgenus C), Ad types 8, 9, 10, 13, 19, 19a, 37, and 44 (sub-genus D), Ad4 (subgenus E), and Ad40 and Ad41 (subgenus F)were selected. The length of the amplimer produced was 301bp for each of the 23 different types sequenced. Our sequenc-ing results for Ad types 3, 7, 12, 40, and 41 were identical tothose for the same prototype strains published previously,which led us to not publish the data. However, the sequencingresult for Ad4p presented here is not in concordance with therecently published sequence of the same strain, Ad4 prototypestrain RI-67, ATCC (Pring-Akerblom et al., 1995, EMBL da-tabase accession number X84646). (i) The sequence of Pring-Akerblom et al. has a G at position 99 in the hexon openreading frame (ORF), whereas our sequence has a C, whichintroduces a SmaI site in the former sequence. However, SmaIcould not digest either the DNA of the Ad4 prototype strain(ATCC) or the DNA from six genome variants of Ad4. (ii) Atransposition of two nucleotides, nucleotides 194 and 195, inthe hexon ORF is also noted when the sequence of Pring-Akerblom et al. (GC) is compared with ours (CG). NucleotidesG and C at that position would contribute to a BlpI site, whichcannot be demonstrated in a digestion experiment with thisenzyme either in the prototype strain or in the six different

VOL. 39, 2001 RAPID TYPING OF HUMAN ADENOVIRUSES 499

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

genome variants. On the basis of these results we suggest asecond version of the Ad4 prototype hexon ORF sequence atnucleotides 46 to 299 (GenBank accession number AF161569).

Although the amplified products originated from a con-served part of the hexon-coding gene, the sequencing data forthe 26 prototype strains and genome type 19a were heteroge-neous enough to allow discrimination between subgenera andeven between different serotypes (Fig. 1). Bailey and Mautner(8) have previously described trees that represent other Adgene sequences and that also show a branching pattern of the

serotypes that is in good correlation with the classification intothe six subgenera (subgenera A to F).

RE analysis. The sequencing data were further analyzed forthe presence of distinctive RE cleavage sites to discriminatebetween Ad subgenera and, if possible, between different types(GCG Map program, version 9.1). The selection of enzymeswas based on their cleavage or noncleavage of the 301-bp PCRproduct (i.e., 253 bp with the primer sequences subtracted), astrategy that offers a simple and user-friendly detection system.When alternatives were present, the cost of the enzymes and

TABLE 1. Summary of patient data including age, sample source, and symptoms together with PCR and typing resultsfor 40 clinical samplesa

Positivesamples

no.

Patientage Patient sample Symptoms

One-stepPCRresult

Type byPCR-REdigestionanalysisb

Type by REanalysis of full-

length Ad DNAc

NestedPCRresult

1 9 mo Feces Diarrhea Positive 7 gen. var. No sample left ND2 22 yr Throat swab sample Fever, pharyngeal

conjuctivitis headacePositive 7 gen. var. 7b ND

3 22 yr Conjunctiva Pharyngeal conjunctivitis Positive 7 gen. var. 7b ND4 2.5 yr Feces Diarrhea Positive 40 40 ND5 40 yr Conjunctiva Keratoconjunctivitis Positive 4 4a ND6 39 yr Eye sukb sample Fever, pharyngeal

conjunctivitisPositive 7 gen. var. 7b ND

7 22 mo Feces Fever, conjunctivitis,bronchitis, otitis

Positive 41 41 ND

8 29 yr Conjunctiva Conjunctivitis Positive 7 gen. var. 7b ND9 2.5 yr Throat sample Fever, exanthema,

pharyngitisPositive 7 gen. var. 7b ND

10 15 mo CSF Fever, stiff neck,photophobia

Positive 31 31 ND

11 2.5 yr Feces Fever, exanthema Positive 7 gen. var. 7b ND12 13 mo Feces Diarrhea, vomiting Positive 31 31 ND13 1 yr NPA Fever Positive 2, 5, or 6 No sample left ND14 72 yr BAL Pneumonia, Positive 2, 5, or 6 No sample left ND15 2 yr NPA Fever, cough Positive 2, 5, or 6 No sample left ND16 3 mo Feces Diarrhea, edema Positive 2, 5, or 6 No sample left ND17 22 yr Vesicles Stomatitis, fever Positive 1 No sample left ND18 20 yr Vesicles Genital ulcers Positive 2, 5, or 6 No sample left ND19 61 yr CSF Neuropathy Positive 1 No sample left ND20 3 mo NPA Lower respiratory

infectionPositive 2, 5, or 6 No sample left ND

21 74 yr NPA Fever, meningitis Positive 2, 5, or 6 No sample left ND22 4 yr Urine ? Negative ND No sample left Positive23 38 yr Eye swab sample Conjunctivitis Positive 3 gen. var. 3a ND24 35 yr Eye swab sample Conjunctivitis Negative ND No sample left Positive25 35 yr Throat swab sample Pharyngitis Positive 7 gen. var. 7b ND26 9 yr NPA Fever, cough, Negative ND No sample left Positive27 46 yr NPA Fever, cough Negative ND No sample left Positive28 24 yr Eye swab sample Pharyngeal conjunctivitis Negative ND No sample left Positive29 19 yr Eye swab sample Pharyngeal conjunctivitis Negative ND No sample left Positive30 6 mo NPA Bronchitis, cough Positive 2, 5, or 6 No sample left ND31 61 yr Feces ? Positive 2, 5, or 6 No sample left ND32 2 yr Feces Gastroenteritis Positive 41 41 ND33 1 yr Feces Gastroenteritis Positive 31 31 ND34 16 mo Feces Gastroenteritis Positive 41 41 ND35 4 mo Feces Diarrhea, vomiting Positive 40 40 ND36 3 yr Feces Diarrhea Positive 7 gen. var. 7b ND37 44 yr Serum ? syndrome Positive 3 gen. var. No sample left ND38 36 yr CSF Guillain-Barre(?),

pneumonitisPositive 2, 5, or 6 5 D2 ND

39 2.5 yr Feces Enteritis Positive 1 No sample left ND40 33 yr Eye swab sample ? Positive 3 gen. var. 3a ND

a Abbreviations: NPA, nasopharyngeal aspirits; BAL, broncoalveolar lavage sample; gen. var., genome variant; ND, not determined.b Typing results with the suggested enzymes in Fig. 2 and Fig. 3 with the one-step PCR product as a template.c Results of traditional typing, which involves the evaluation of specific migration patterns on agarose gels of full-length Ad DNA cleaved with different REs.

500 ALLARD ET AL. J. CLIN. MICROBIOL.

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

their availability on the market were taken into considerationto direct the choice. Conceivable enzymes were tested with the301-bp amplimer of all 51 Ad prototype strains together withthe amplimers of 44 different genome variants of the virus tocreate a flowchart that was as simple as possible but that stilldiscriminated the types. A compilation of all the digestionsthen resulted in a scheme of the most suitable REs to be used(Fig. 2). Hence, cleavage by SalI and/or TaqI means that theamplified product originates from subgenus F and that furthertyping can be done by the use of HinfI to discriminate betweenAd40 and Ad41. If the product is not cleaved by SalI or TaqI,the second RE to be used is AvaII, which determines whatbranch is to be followed next; the left branch, representingsubgenera A and B, or the right branch, representing subgen-era C, D, and E. Consequently, noncleavage by AvaII leads tothe left branch, with HaeII being a third choice of RE. HaeIIdiscriminates subgenus A from subgenus B. Complete distinc-tion of subgenus A serotypes can further be done by digestionwith PvuI and AluI. Subgenus B can be divided into two clus-ters, clusters B:1 and B:2, with NarI. Cluster B:2 represents adead-end and virus cannot be further typed with this system,whereas cluster B:1 can be completely typed apart from diffi-culties in discriminating Ad21 from new type Ad51. By diges-tion with TaqI, Ad21 and Ad51 can be separated from Ad3,Ad7, and Ad16. Furthermore, Ad16 can be identified by ScaIcleavage. The final distinction between Ad3 and Ad7 must bedone by following a second flowchart (Fig. 3). This scheme issomewhat complicated since amplimers of the prototypes (3p

and 7p) respond differently to digestion compared to the re-sponses of the genome types of these two types. Cleavageinstructions are presented in Fig. 3.

By following the right branch in Fig 2, noncleavage of theamplimer by SmaI can distinguish subgenus E from subgeneraC and D. CspI can discriminate between subgenera C and D,and subgenus C can finally be partially typed with MslI. Sub-genus D leads to a second dead-end since the similarity in theamplimer sequences makes it impossible to further type thissubgenus with the REs available today.

It is very important to use the RE within the flowcharts byfollowing the correct direction from the top to the bottom, asrecommended in the text. Several enzymes have their recog-nition sites in the sequences of more than one subgenus ortype, and when introduced too early, the results will be mis-leading. For example, the RE SmaI that is used in the flow-chart to discriminate subgenera C and D from subgenus E willalso digest subgenus F members and Ad18 of subgenus A.

Products representing Ad31 will pass the flowchart withoutbeing cleaved by any enzyme. To confirm that no inhibitors willprevent cleavage of the DNA, HinfI can be used as a DNAquality control. This RE will cut Ad31, and furthermore, it isinexpensive and is already present within the flowchart.

The RE SalI digests all genome variants of subgenus Ftested so far except genome type D9 of Ad40. TaqI can dis-criminate all known genome types of subgenus F but alsocleaves Ad21 and Ad51. However, the TaqI restriction profilesof Ad21 and Ad51 are clearly different from that of the sub-

FIG. 1. Phylogenetic analysis of 27 different Ad types showing the subgenus relationship between the 253-bp sequence at positions 46 to 299within the hexon gene. The view is unbalanced, and the scale beneath the tree measures the distance between sequences. Published sequences fromthe GenBank database of Ad2 (J01917), Ad3 (X76549), Ad5 (X02997), Ad7 (X76551), Ad12 (X73487), Ad16 (X74662), Ad40 (X51782), Ad41(X51783), and Ad48 (U20821) are also included.

VOL. 39, 2001 RAPID TYPING OF HUMAN ADENOVIRUSES 501

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

genus F Ads. Ad40 and Ad41 amplimers representing allknown genome types are cleaved by TaqI, yielding two frag-ments of 191 and 110 bp, and the Ad21 and Ad51 amplimersare cleaved, yielding two fragments of 231 and 70 bp (Fig. 4).

The digestion of the Ad16 amplimer with ScaI yields a smallfragment of 41 bp, together with a larger fragment of 260 bp.The efficiency of this enzyme has, in our hands, not alwaysbeen 100%, so the 260-bp product will comigrate together withthe undigested 301-bp amplimer, giving twin bands on the gel.However, the interpretation of the result is still simple.

Tests with clinical samples. The enzymes indicated in Fig. 2and Fig. 3 were repeatedly tested (more than four times) withall 51 prototype strains and 44 genome variants (data notshown). As a final test, 40 Ad-positive samples were added tothe study to estimate the use of the typing system in diagnosticwork. The assay was performed directly with clinical samples,and 34 of the 40 samples were positive by a one-step PCR andcould be typed by following the RE cleavage schemes. The sixadditional samples became positive when the general nestedPCR primers were used, creating a 171-bp product too short tobe used in the typing system presented here. Thirteen mem-bers of subgenus C were found, represented by 10 samples oftype Ad2, Ad5, or Ad6 and 3 samples of type Ad1. Two of thesubgenus C-positive samples were CSF samples from adults.The second most common subgenus found was subgenus B,represented by nine genome types of serotype Ad7, togetherwith three genome types of serotype Ad3. The viruses in twosamples with type Ad40 and three samples with type Ad41represented subgenus F. Type Ad31 of subgenus A was foundin CSF from a child and in fecal samples from two differentadults. Finally, type Ad4, the only member of subgenus E, wasfound in the eye of a 40-year-old woman. No member ofsubgenus D was identified. All the results for the positive

clinical samples together with patient data are summarized inTable 1. Some examples of the restriction profiles of the PCRproducts derived from clinical samples are given in Fig. 5.

Traditional RE typing of Ad genomic DNA. To confirm theresults of typing by PCR-RE digestion, traditional RE typingwas performed with full-length genomic viral DNA. The con-firmation was limited because of restricted amounts of mate-rial, and therefore, only 20 of the original samples positive bya one-step PCR could be included. The specimens were inoc-ulated onto cells, and all of them displayed a cytopathic effectcharacteristic of Ad infection. Ad-specific DNA was preparedfrom the positive cultures, and a type-specific migration pat-tern was obtained by BamHI digestion (Fig. 6). These resultswere further confirmed by digestion with SmaI, EcoRI, or BglII(data not shown). All the results obtained were in concordancewith the PCR-RE typing results and also introduced morespecific information about genome variants such as 3a, 4a,5D2, and 7b (Table 1).

DISCUSSION

In 1 day the identification of Ad isolates for epidemiologicalsurveillance can be done by using the PCR-RE digestionmethod. The identification is generally based on genomic typ-ing by DNA analysis with REs or neutralization tests, whichboth involve cell cultivation, which consumes about 2 weeks of

FIG. 2. Flowchart of REs for typing of PCR products framed by thegeneral hexon primers hex1deg and hex2deg. The symbol of a pair ofscissors indicates that the product will be cleaved by the enzyme inquestion, whereas no symbol indicates an uncleaved product. See Re-sults for further information on the use of the flowchart.

FIG. 3. Flowchart of REs used for partial typing of prototypes andgenome types of Ad3 and Ad7, subgenus B1. The symbol of a pair ofscissors indicates that the product will be cleaved by the enzyme inquestion, whereas no symbol indicates an uncleaved product.

502 ALLARD ET AL. J. CLIN. MICROBIOL.

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

time. A typing method based on PCR offers a great possibilityto identify fastidious Ads or Ads in specimens toxic to cellcultivation, since modern DNA extraction techniques adaptedto all kinds of human tissue and fluids are available today. Ifnecessary, this PCR-RE digestion method can be combinedwith more laborious systems to extend the typing capacity toinclude subgenus B2 and subgenus D (35, 37, 39, 44).

The Ad types within a subgenus are similar in their tropismspathogenic potentials, tendencies to cause a latent infection,and levels of occurrence or reactivation in immunosuppressedindividuals. Therefore, in the clinical setting, identification ofAds by determination of subgenus alone is often but not alwayssufficient when a more virulent strain must be excluded. Asystem of subgenus typing by PCR with the VA RNA generegions as a target has been described; the system can be usedwith further differentiation with REs to confirm the subgenusidentity (24). A multiplex PCR for subgenus-specific detectionwith selective primers from the loop l4 gene region of thehexon has also been published (38). However, in these twosystems, together with other PCR typing systems (39, 45),

discrimination between the different types is performed bycomparing the migration of small fragments of similar sizes ina gel, a result that can ocassionally be very difficult to interpret,despite the use of high-resolution gels or high-quality molec-ular size markers. The PCR-RE digestion method describedhere provides an intelligible flowchart with a simple and clear-cut final readout, namely, cleavage or noncleavage of a PCRamplimer.

This typing system is based on cleavage of one-step PCRproducts. Occasionally, a one-step reaction is not sufficient toallow detection of virus DNA. To avoid the risk of false-negative results, a nested system can be used. The general pairof nested primers described herein will produce an internalproduct of 171 bp that can perhaps be used for partial typingsince several REs have their cleavage recognition sites locatedin the central part of the sequences. However, a flowchartbased entirely on RE cleavage of this short fragment would betoo incomplete to be practicable. Instead, cell culture can beperformed prior to PCR amplification if the virus concentra-tion in a sample is too low. However, both nested PCR andcombined cell culture-PCR are powerful tools and can easilydetect persistent Ad infections in stool samples. The Ad that isfound in a sample should therefore not always be consideredthe cause of illness since infections caused by members ofsubgenus C and serotype Ad3 are characterized by a prolongedintermittent excretion in stool (19).

FIG. 4. Comparison of TaqI restriction profiles of amplified DNAof prototype strains of Ad types 21, 51, 40, and 41.

FIG. 5. RE digestion profiles of PCR products of viruses from sixdifferent clinical samples representing subgenera A, B1, C, E, and F.Sections: 1, Ad31, subgenus A, patient 10; 2, Ad7 variant, subgenus B1,patient 2; 3, Ad 4, subgenus E, patient 5; 4, Ad1, subgenus C, patient39; 5, Ad40, subgenus F, patient 4; 6, Ad41, subgenus F, patient 34. SeeTable 1 for patient data. fX174 DNA digested with HaeIII was usedas a molecular size marker. The symbol of a pair of scissors indicatesthat the product will be cleaved by the enzyme in question.

VOL. 39, 2001 RAPID TYPING OF HUMAN ADENOVIRUSES 503

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

The modified primers used here are highly degenerate. Theprimers were tested for homology to all available sequences inthe GenBank database, with no indication of unspecific bind-ing. Furthermore, the two sets of primers, including the onesused for nested PCR, have been used in routine diagnosticwork for 5 years without the introduction of unspecific ampli-fications. The viruses in a limited number of clinical sampleswere typed with the PCR-RE digestion system to evaluate themethod. The comparison with full-length DNA-RE typing inthis study indicated that the PCR-RE digestion method isreliable and can be used as a tool for characterization of Adsin samples in daily routine diagnostic work. Furthermore, inour study the distribution of genotypes into each subgenus wassimilar to the distributions found in an extensive epidemiolog-ical study of 200 Ad isolates from the Stockholm area of Swe-den from 1987 to 1992 (21). All 200 Stockholm isolates weretyped by using RE analysis of full-length Ad DNA. Nine per-cent of the 34 samples contained Ads of subgenus A in ourstudy, whereas 8% of the samples in the Stockholm studycontained Ads of subgenus A. The only member that we foundwas type 31, which also was predominant in Stockholm. Sub-genus B Ads were responsible for 35% of the Ad infectionsfound in Umeå, whereas they were responsible for 25.5% ofthe Ad infections in Stockholm. The higher figure was probablydue to an outbreak of type 7b during this short 6-month periodof sampling. The distribution of members of subgenus C wasvery similar in both studies: 38% in Umeå, and 34% in Stock-holm. No representative of subgenus D was found in Umeå,whereas 22.5% of the infections in Stockholm were caused bysubgenus D, dominated by one nosocomial outbreak of kera-

toconjunctivitis due to adenovirus 19a. Ads of subgenus E werefound in 3% of the typed samples in both areas, whereas Adsof subgenus F caused symptoms in 15% of the patients in ourstudy and 7% of the patients in the Stockholm study.

The importance of Ads as a cause of disseminated diseasehas remained underappreciated, perhaps since Ad infectionshave been diagnosed primarily through the use of cell culture.The fact that cell culture is sometimes insensitive for the de-tection of this virus has hindered recognition of the role thatAds may play in morbidity and mortality in patients with symp-toms not normally associated with this virus (9, 15, 16, 18, 25,41).

When designing primers it is important to introduce se-quences derived not only from prototype strains since proto-type strains and genome types of the same serotype do notexhibit identical sequences and the prototype strains are pre-sumed to circulate more rarely in the population. In spite ofthis fact, we cannot exclude the possibility that there exist innature other Ad strains whose hexon regions do not have thesame restriction properties as those seen in the present study.However, this method can facilitate typing of human Ads in avariety of different sample materials and thereby also contrib-ute to an increased understanding of the pathological impor-tance of these viruses.

ACKNOWLEDGMENTS

We thank J. C. deJong, A. Kajon, A. Kidd, Q. G. Li, D. Schnurr, andR. Wigand for the Ad strains used in the study. We also thank P. Jutofor access to clinical samples and J. Forssell for critical reading of themanuscript.

REFERENCES

1. Adrian, T., J. Sassinek, and R. Wigand. 1990. Genome type analysis of 480isolates of adenovirus types 1, 2, and 5. Arch. Virol. 112:235–248.

2. Adrian, T. H., G. Wadell, J. C. Hierholzer, and R. Wigand. 1986. DNArestriction analysis of adenovirus prototypes 1 to 41. Arch. Virol. 91:277–290.

3. Akusjarvi, G., P. Alestrom, M. Pettersson, M. Lager, H. Jornvall, and U.Pettersson. 1984. The gene for the adenovirus 2 hexon polypeptide. J. Biol.Chem. 259:13976–13979.

4. Allard, A., G. Wadell, K. M. Evander, and G. K. K. Lindman. 1985. Specificproperties of two enteric adenovirus 41 clones mapped within early region1A. J. Virol. 54:145–150.

5. Allard, A., R. Girones, P. Juto, and G. Wadell. 1990. Polymerase chainreaction for detection of adenoviruses in stool samples. J. Clin. Microbiol.28:2659–2667.

6. Allard, A., B. Albinsson, and G. Wadell. 1992. Detection of adenoviruses instools from healthy persons and patients with diarrhea by two-step polymer-ase chain reaction. J. Med. Virol. 37:149–157.

7. Allard, A., A. Kajon, and G. Wadell. 1994. Simple procedure for discrimi-nation and typing of enteric adenoviruses after detection by polymerasechain reaction. J. Med. Virol. 44:250–257.

8. Bailey, A., and V. Mautner. 1994. Phylogenetic relationships among adeno-virus serotypes. Virology 205:438–452.

9. Cardosa, M. J., S. Krishnan, P. Hooi Tio, D. Perera, and S. Chang Wong.1999. Isolation of subgenus B adenovirus during a fatal outbreak of entero-virus 71-associated hand, foot, and mouth disease in Sibu, Sarawak. Lancet354:987–991.

10. Cooper, R. J., C. Y. Yeo, A. S. Bailey, and A. B. Tullo. 1999. Adenoviruspolymerase chain reaction assay for rapid diagnosis of conjunctivitis. Invest.Ophthalmol. Vis. Sci. 40:90–95.

11. Crawford-Miksza, L., and D. P. Schnurr. 1996. Analysis of 15 hexon proteinsreveals the location and structure of seven hypervariable regions containingserotype-specific residues. J. Virol. 70:1836–1844.

12. de Jong, J. C., R. Wigand, A. H. Kidd, G. Wadell, J. G. Kapsenberg, C. J.Muzerie, A. G. Wermenbol, and R.-G. Firtzlaff. 1983. Candidate adenovi-ruses 40 and 41: fastidious adenoviruses from human infant stool. J. Med.Virol. 11:215–231.

13. de Jong, J. C., A. G. Wermenbol. M. W. Verweij-Uijterwaal, K. W. Slaterus,P. Wertheim-van Dillen, G. J. J. van Doornum, S. H. Khoo, and J. C.Hierholzer. 1999. Adenoviruses from AIDS patients, including two new

FIG. 6. BamHI restriction patterns of full-length Ad DNAs ex-tracted from cells infected with viruses from different patient samples.The numbers above each lane indicate the patient number listed inTable 1. Patients 2, 3, 6, 8, 9, 11, 25, and 36 were infected with Ad7b.Patients 10, 12, and 33 were infected with Ad31, patient 38 was in-fected with Ad5 D2, patients 23 and 40 were infected with Ad3a,patient 5 was infected with Ad4a, and finally, patients 4 and 7 wereinfected with Ad40 and Ad41, respectively. Molecular size standardsconsisted of bacteriophage l DNA digested with HindIII and fX174DNA digested with HaeIII.

504 ALLARD ET AL. J. CLIN. MICROBIOL.

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

candidate serotypes: Ad50 and Ad51 of subgenus D and B1, respectively.J. Clin. Microbiol. 37:3940–3945.

14. Echavarria, M., M. Forman, J. Ticehurst, J. S. Dumler, and P. Charache.1998. PCR method for detection of adenovirus in urine of healthy andhuman immunodeficiency virus-infected individuals. J. Clin. Microbiol. 36:3323–3326.

15. Echavarria, M. S., S. C. Ray, R. Ambinder, J. S. Dumler, and P. Charache.1999. PCR detection of adenovirus in a bone marrow transplant recipient:hemorrhagic cystitis as a presenting manifestation of disseminated disease.J. Clin. Microbiol. 37:686–689.

16. Flomenberg, F., E. Gutierrez, V. Piakowski, and J. T. Casper. 1997. Detec-tion of adenovirus DNA in peripheral blood mononuclear cells by polymer-ase chain reaction assay. J. Med. Virol. 51:182–188.

17. Fong, C. K. Y. 1994. Adenoviridae, p. 232–235. In C. D. Hsiungs (ed.),Diagnostic virology, 4th ed. Yale University Press, New Haven, Conn.

18. Forsnes, E. V., M. K. Eggleston, and J. R. Wax. 1998. Differential transmis-sion of adenovirus in a twin pregnancy. Obstet. Gynecol. 91:817–818.

19. Fox, J. P., C. E. Hall, and M. K. Cooney. 1977. The Seattle Virus Watch. VII.Observations of adenovirus infections. Am. J. Epidemiol. 105:362–386.

20. Hierholzer, J. C. 1992. Adenoviruses in the immunocompomised host. Clin.Microbiol. Rev. 5:262–274.

21. Johansson, M. E., M. A. Andersson, and P. A. Thorner. 1994. Adenovirusesisolated in the Stockholm area during 1987–1992: restriction endonucleaseanalysis and molecular epidemiology. Arch. Virol. 137:101–115.

22. Kajon, A., and G. Wadell. 1994. Genome analysis of South American ade-novirus strains of serotype 7 collected over a 7-year period. J. Clin. Micro-biol. 32:2321–2323.

23. Kajon, A. E., A. S. Mistchenko, C. Videla, M. Hortel, G. Wadell, and L. F.Avendano. 1996. Molecular epidemiology of adenovirus acute lower respi-ratory infections of children in the south cone of South America (1991–1994). J. Med. Virol. 48:151–156.

24. Kidd, A. H., M. Jonsson, D. Garwicz, A. Kajon, A. G. Wermenbol, M. W.Verweij, and J. C. de Jong. 1996. Rapid subgenus identification of humanadenovirus isolates by a general PCR. J. Clin. Microbiol. 34:622–627.

25. Kuwano, K., M. Kawasaki, R. Kunitake, N. Hagimoto, Y. Nomoto, T. Mat-suba, Y. Nakanishi, and N. Hara. 1997. Detection of group C adenovirusDNA in small-cell lung cancer with the nested polymerase chain reaction. J.Cancer Res. Clin. Oncol. 123:377–382.

26. Li, Q., and G. Wadell. 1988. The degree of genetic variability among ade-novirus type 4 strains isolated from man and chimpanzee. Arch. Virol.101:65–77.

27. Li, Q. G., J. Hambraeus, and G. Wadell. 1991. Genetic relationship betweenthirteen genome types of adenovius 11, 34, and 35 with different tropism’s.Intervirology 32:338–350.

28. Li, Q. G., Zheng, Y. Liu, and G. Wadell. 1996. Molecular epidemiology ofadenovirus types 3 and 7 isolated from children with pneumonia in Beijing.J. Med. Virol. 49:170–177.

29. Li, Q. G., A. Henningsson, P. Juto, F. Elgh, and G. Wadell. 1999. Use ofrestriction fragment analysis and sequencing of a serotype-specific region totype adenovirus isolates. J. Clin. Microbiol. 37:844–847.

30. Mistchenko, A. S., J. F. Robaldo, F. C. Rosman, E. R. R. Koch, and A. Kajon.1998. Fatal adenovirus infection associated with new genome type. J. Med.Virol. 54:233–236.

31. Morris, D. J., A. S. Bailey, R. J. Cooper, P. C. Turner, R. Jackson, G. Corbitt,and A. B. Tullo. 1995. Polymerase chain reaction for rapid detection ofocular adenovirus infection. J. Med. Virol. 46:126–132.

32. Morris, D. J., R. J. Cooper, T. Barr, and A. S. Bailey. 1996. Polymerase chain

reaction for rapid diagnosis of respiratory adenovirus infection. J. Infect.32:113–117.

33. Murphy, F. A., C. M. Fauquet, D. H. L. Bishop, S. A. Ghabrial, A. W. Jarvis,G. P. Martelli, M. A. Mayo, and M. D. Summers (ed.). 1995. Virus taxonomy:classification and nomenclature of viruses. Sixth report, p. 128–133. Spring-er-Verlag, Vienna, Austria.

34. Pauschinger, M., N. E. Bowles, F. J. Fuentes-Garcia, V. Pham, U. Kuhl, P. L.Schwimmbeck, H.P. Schultheiss, and J. A. Towbin. 1999. Detection of ad-enoviral genome in the myocardium of adult patients with idiopathic leftventricular dysfunction. Circulation 10:1348–1354.

35. Pring-Akerblom, P., and T. Adrian. 1994. Type- and group-specific polymer-ase chain reaction for adenovirus detection. Res. Virol. 145:25–35.

36. Pring-Akerblom, P., F. E. J. Trussenaar, and T. Adrian. 1995. Sequencecharacterization and comparison of human adenovirus subgenus B and Ehexons. Virology 212:232–236.

37. Pring-Akerblom, P., T. Adrian, and T. Kostler. 1997. PCR-based detectionand typing of human adenovirus in clinical samples. Res. Virol. 148:225–231.

38. Pring-Akerblom, P., F. E. J. Trijssenaar, T. Adrian, and H. Hoyer. 1999.Multiplex polymerase chain reaction for subgenus-specific detection of hu-man adenoviruses in clinical samples. J. Med. Virol. 58:87–92.

39. Saitoh-Inagawa, W., A. Oshima, K. Aoki, N. Itoh, K. Isobe, E. Uchio, S.Ohno, H. Nakajima, K. Hata, and H. Ishiko. 1996. Rapid diagnosis ofadenoviral conjunctivitis by PCR and restriction fragment length polymor-phism analysis. J. Clin. Microbiol. 34:2113–2116.

40. Schnurr, D., and M. Dondero. 1993. Two new candidate adenovirus sero-types. Intervirology 36:79–83.

41. Schnurr, D., A. Bollen, L. Crawford-Miksza, M. E. Dondero, and S. Yagi.1995. Adenovirus mixture isolated from the brain of an AIDS patient withencephalitis. J. Med. Virol. 47:168–171.

42. Shinagawa, M., A. Matsuda, T. Ishiyama, H. Goto, and G. Sato. 1983. Arapid and simple method for preparation of adenovirus DNA from infectedcells. Microbiol. Immunol. 27:817–822.

43. Sprengel, J., B. Schmitz, D. Heuss-Neitzel, C. Zock, and W. Doerfler. 1994.Nucleotide sequence of human adenovirus type 12 DNA: comparative func-tional analysis. J. Virol. 68:379–389.

44. Takeuchi, S., N. Itoh, E. Uchio, K. Aoki, and S. Ohno. 1999. Serotyping ofadenoviruses on conjunctival scrapings by PCR and sequence analysis.J. Clin. Microbiol. 37:1839–1845.

45. Tiemessen, C. T., and M. J. Nel. 1996. Detection and typing of subgroup Fadenovirus using the polymerase chain reaction. J. Virol. Methods 59:73–82.

46. van der Avoort, H. G. A. M., A. G. Wermenbol, T. P. L. Zomerdijk, J. A. F. W.Kleijne, J. A. A. M. van Asten, P. Jensma, A. D. M. E. Osterhaus, A. H. Kidd,and J. C. de Jong. 1989. Characterization of fastidious adenovirus types 40and 41 by DNA restriction enzyme analysis by neutralizing monoclonalantibodies. Virus Res. 12:139–158.

47. Wadell, G., M.-L. Hammarskjold, G. Winberg, T. M. Varsanyi, and G.Sundell. 1980. Genetic variability of adenoviruses. Ann. N.Y. Acad. Sci.354:16–42.

48. Wadell, G., and J. C. de Jong. 1980. Restriction endonucleases in identifi-cation of a genome type of adenovirus 19 associated with keratoconjuncti-vitis. Infect. Immun. 27:292–296.

49. Wadell, G., A. Allard, M. Evander, and Q. Li. 1986. Genetic variability andevolution of adenoviruses. Chem. Scripta 26B:325–335.

50. Wadell, G., A. Allard, and J. C. Hierholzer. 1999. Adenoviruses, p. 970–982.In P. R. Murray, E. J. Baron, M. A. Pfaller, F. Tenover, and R. N. Yolken(ed.), Manual of clinical microbiology, 7th ed. American Society for Micro-biology, Washington, D.C.

VOL. 39, 2001 RAPID TYPING OF HUMAN ADENOVIRUSES 505

on Novem

ber 26, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from