Candidate adenoviruses 40 and 41: fastidious adenoviruses from human infant stool

Adenovirus 41 growth in semi-permissive cells shows multiple-hit kinetics

Adenovirus type 41 infection of semi-permissive Chang conjunctival cells, monitored by fluorescent focus assay, followed multiple-hit kinetics. In non-permissive human cells, type 41 showed infectivity with two-hit kinetics in the presence of type 2. Type 41 infectivity was seen to be directly proportional to input concentration (one-hit) only in 293 cells, a continuous human line expressing Ad 5 E1 products.

Enteric adenovirus 41 (Tak) requires low serum for growth in human primary cells

It had been postulated that due to lack of growth of enteric adenovirus 41 (Ad41) on human primary cells and its growth on Graham-293 cells there was a defect in the Ad41 E1A region. However, we found as a result of careful evaluation of Ad41 growth on several primary cell lines (HEK, WI-38, or Detroit 551) that efficient viral multiplication is possible if the serum concentration in the medium used postinfection (p.i.) is kept between 0.2 and 1%. In contrast, only slight growth of Ad41 occurs in infected cells maintained in 5% serum and virtually no viral replication is found in infected cells cultivated in medium with 10% serum. The serum inhibitory effect appears limited to primary cells because no difference in Ad41 replication, as assayed by accumulation of Ad41 DNA, was found in infected continuous cell lines (HEp-2, 293) cultivated p.i. in either 1 or 10% FBS. Also, this effect appears specific for enteric adenoviruses, such as Ad41, since conventional adenoviruses, such as Ad5, grow well in both 1 and 10% FBS. The above results show that Ad41 can grow in a variety of primary cell lines, under specific culture conditions. In addition, we found that Ad41-infected primary cells grown in medium containing 0.2% serum had an increase in synthesis of the 70-kDa heat shock protein (HSP70) at about 6 hr p.i. and also Ad41 was able to complement the Ad5 E1A deletion mutant dl312. These results show that the E1A function of Ad41 is not impaired in infected cells.

Pathogenesis of gut virus infection

In summary, the pathogenesis of many gut virus infections remains uncertain. However, human and animal studies indicate that the majority of gut viruses infect villous enterocytes. Viruses appear to have different affinities for enterocytes at different sites on the villus. Infection of enterocytes leads to cell death, extrusion into the lumen, and villous atrophy when the rate of cell production in the crypts cannot keep pace with the rate of enterocyte loss. This results in a reduced surface area as well as impairment of digestive and absorptive functions. This may also result in a net secretory state. All these changes, along with others such as reduced enzymatic activity and reduced epithelial integrity, may contribute to the induction of an acute but transient malabsorptive diarrhoea which may persist until the digestive/absorptive functions of the enterocyte are restored. However, if colonic compensation is sufficient to handle the increased fluid load, diarrhoea may not be evident. The roles of villous ischaemia, altered countercurrent exchanger of altered immune responses still remain uncertain and require further investigation.

Enteric adenoviruses

Human adenoviruses are classified into 47 serotypes and six subgenera (A-F) with different tropisms. In recent years adenovirus type 40 (Ad40) and 41 (Ad41) of subgenus F have been shown to be causative agents in enteric infections, which is second in importance only to rotaviruses as a cause of infantile gastroenteritis. Infection with EAds occurs worldwide and has been associated with 4-17% of cases of diarrhoea in children. AD40 and Ad41 primarily affect young children less than 2 years of age and occur throughout the year. The clinical characteristics include watery diarrhoea accompanied by vomiting, low grade fever and mild dehydration. A distinct feature of EAds infection is the protracted diarrhoea (mean 8.6 and 12.2 days for Ad40 and Ad41, respectively). Respiratory symptoms are infrequent. Serotypes Ad40 and Ad41 differ from all other (established) adenoviruses by being unable to replicate in conventional cell cultures. These fastidious viruses only grow in selected cell lines, 293 cells being the most commonly used. In spite of the difficulty of isolating Ad40 and Ad41, they can be directly identified and typed by ELISA and solid-phase immune electron microscopy. The amount of viral DNA in stool specimens is sufficient for identification by DNA restriction and dot-blot assays. The recent development of highly sensitive and specific monoclonal antibody-based ELISAs enable accurate diagnosis of adenovirus gastroenteritis in routine work and make possible the evaluation of the role of the enteric adenoviruses in diarrhoeal disease in the developing countries.

Laboratory identification of adenoviruses associated with gastroenteritis in Canada from 1983 to 1986

A retrospective analysis of adenovirus serotypes associated with gastroenteritis involved the examination of 143 stool specimens collected between 1983 and 1986 from symptomatic patients whose stools were positive for adenovirus by electron microscopy. The virus isolates obtained from 140 of the specimens were typed according to the SmaI cleavage pattern of the viral DNA and by neutralization with specific antisera. The predominant types were adenovirus type 31 (Ad31) (18%), Ad40 (16.9%), and Ad41 (38%), which together accounted for more than 70% of the isolates. The remaining virus isolates were typed as Ad1, 2, 3, 5, 7, and 12. DNA restriction analysis proved to be better than serum neutralization for identification of the enteric adenovirus serotypes in stool specimens. HindIII cleavage identified four Ad41 variants, none of which had a HindIII restriction pattern identical to that of the prototype strain Tak. Over the time period of the study, the incidence of Ad40 showed an overall decrease accompanied by an increased incidence of Ad41, while the incidence of Ad31 was relatively stable.

Adenovirus transformed monkey cell lines permissive to enteric adenovirus type 40

Plasmids containing the E1 regions of adenovirus serotypes 3 and 5 were transfected into primary Rhesus monkey kidney cells. The presence of viral DNA sequences was detected in transformed cell lines. All these cell lines expressed the E1A proteins. In addition, Ad5 transformed cells, have the E1B 21 kDa protein located in the nuclear membrane. These cell lines were permissive to the enteric adenovirus serotype 40 but not to serotype 41.

Time-resolved fluoroimmunoassay for enteric adenoviruses using the europium chelator 4,7-bis(chlorosulfophenyl)-1,10-phenanthroline-2,9-dicarboxylic acid

A time-resolved fluoroimmunoassay (TR FIA) was developed for the direct detection of adenovirus types 40 (Ad40) and 41 (Ad41) in stool specimens by using a monoclonal antibody (5D8/2C2) which recognizes both Ad40 and Ad41 but does not cross-react with other adenovirus serotypes. In this assay, the detector antibody is biotinylated directly rather than labeled with europium, and the fluorescent signal is generated on a solid phase in the presence of excess europium (Eu3+). The strength of the signal is dependent on the amount of a Eu3+ chelator [4,7-bis(chlorosulfophenyl)-1,10-phenanthroline-2,9-dicarboxylic acid (BCPDA)]-streptavidin complex bound by the biotinylated detector antibody (5D8/2C2). In a pilot study with 41 specimens, this TR FIA demonstrated a maximum sensitivity and specificity of 88% compared with SmaI restriction analysis of adenovirus isolates from the same specimens. TR FIA using the europium chelator BCPDA represents a feasible approach for the direct identification of specific adenovirus serotypes in stool specimens.

Detection of adenovirus type41 in stool samples by a latex agglutination method

We have developed a simple agglutination (LA) method for the detection of enteric adenovirus (EAd) in stool samples from infants with acute gastroenteritis. Ad type 41 (Ad41) was detected with high sensitivity and specificity by a slide agglutination test using latex particles coated with antiAd41 antibody (LA-antiAd41). The agglutination of LA-antiAd41 with Ad41 on a glass slide was evident macroscopically within 2 min. The sensitivity of the LA method was four times higher than that of the EM method.

Differential growth of human enteric adenovirus 41 (TAK) in continuous cell lines

Differing reports exist about the replication of human enteric adenoviruses (EnAds) in various cell lines. There was a suggestion that EnAds are defective, do not grow on primary human diploid cells, and behave like Ad host-range mutants, i.e., they require early gene products from other Ad types for efficient growth. Thus, initially the Graham-293 cell line, which contains the E1 region (E1A, E1B) of Ad5, was thought to be an ideal host for EnAds, because it provided the needed functions. Our findings, however, question this contention and show that Ad41 strain TAK, cultured on 293 cells, rapidly loses its infectivity (greater than 90% on the first and 100% by the second passage). In contrast to the results with 293 cells, we found that Ad41 strain TAK can be serially grown to high titers on several continuous cell lines: namely, HeLa, HI407, or HEp-2 cells. In order to investigate the basis for the rapid loss of the Ad41 infectivity upon passaging in 293 cells, Ad41 virions were purified from 293, as well as from HEp-2 cells, and their composition was analyzed. When structural proteins of the complete virions (rho = 1.34 g/cm3) were compared by Western blot analysis using polyclonal antibodies to HEp-2-grown virus, only traces of protein V (Mr 46,000 Da) could be detected in particles from 293 cells. In contrast, Ad41 particles obtained from HEp-2 cells exhibited a strong band at the position of protein V. Further, if polyclonal antibodies to 293-grown Ad41 were used in the Western blot, no protein V band was detected in HEp-2-grown virus. Finally, we note that new protein bands (Mr 25,000-35,000 Da) could be observed upon Western blot analysis of 293-derived complete and incomplete Ad41 particles. All of these observations taken together suggest that the low infectivity of Ad41 particles, prepared from 293 cells, could be due to a defect in assembly.

Development and evaluation of monoclonal antibody-based immune electron microscopy for diagnosis of adenovirus types 40 and 41

Immune electron microscopy based on monoclonal antibodies was developed and evaluated for diagnosis of adenovirus type 40 and adenovirus type 41 directly from clinical specimens. One adenovirus type 40 monoclonal (5-8) and one adenovirus type 41 monoclonal (5-15) were found to react to high titre with homotypic but not heterotypic antigen. These monoclonals were tested on a coded batch of 20 stools which contained adenovirus type 40 or adenovirus type 41. The results showed that 5/6 adenovirus type 40 and 13/14 adenovirus type 41 strains were correctly serotyped but one strain of each type failed to react with either serum. A wide variation in the numbers of virions bound to positive grids was observed. A further coded batch of 27 specimens, a mixture of subgenus F (i.e. type 40 or 41) or non-subgenus F adenoviruses, was then tested. There was complete serotype concordance with reference results for 16/19 subgenus F strains and all 8 non-subgenus F adenoviruses gave negative results. However, three subgenus F adenoviruses also gave negative results. In conclusion, monoclonal antibody-based immune electron microscopy accurately distinguished adenovirus type 40 from adenovirus type 41 and both viruses from other adenovirus serotypes in clinical specimens and will therefore be useful in the diagnosis of adenovirus gastroenteritis, but some strains may be missed, presumably because of antigenic variation in surface epitopes.

Sequence characterization of the adenovirus 40 fiber gene

The fiber gene of human adenovirus type 40 has been characterized. The 6.1-kbp EcoRI fragment C of the Ad40 genome, from map units 74 through 92, was cloned and the right-most 2.8 kbp from 84 map units was sequenced. By analogy with Ad2, this region would be expected to contain the gene specifying the Ad40 fiber polypeptide. Sequencing revealed an open reading frame of 1641 bases on the r-strand, the first 53 bases of which had marked homology with the corresponding L5 (fiber) regions of Ad2 (77.2%), Ad5 (75.0%), and Ad3 (64.3%). In addition, base positions 1114 to 1146 of this open reading frame had 85% homology with base positions 1198 to 1230 of the Ad2 fiber gene. The predicted polypeptide sequence of 547 amino acids showed marked homology with the Ad2, Ad5, and Ad3 fiber polypeptides in two regions, in the first 55 amino acids from the N-terminus and from amino acids 372 through 382. Analysis of hydrophobic amino acid positions revealed a repeating pattern of approximately 15 residues between positions 42 and 374, with 21 repeats. The sequence of the Ad40 polypeptide thus fits the model of Green et al. [1983), EMBO J. 2, 1357-1365) for the structure of the adenovirus fiber, but is 35 amino acids shorter than the Ad2 fiber polypeptide, with one less 15-residue repeat in the shaft region. According to this model, the regions of highest homology between the Ad40 fiber polypeptide and those of Ad2, Ad5, and Ad3 correspond to the tail of the shaft and the base of the knob. The results of this analysis are in agreement with previously published EM data on the fiber length of subgroup F adenoviruses.

Complementation of enteric adenovirus type 40 for lytic growth in tissue culture by E1B 55K function of adenovirus types 5 and 12

The enteric adenovirus type 40 strain Dugan (Ad40) cannot be passaged in HeLa cells, but will grow in 293 cells, which express Ad5 E1 functions. To determine the reason for this limited host range, KB cell lines expressing Ad2 E1A, E1B, or E1A + E1B (L. E. Babiss, C. S. H. Young, P. B. Fisher, and H. S. Ginsberg, 1983, J. Virol. 46, 454-465) have been tested for their ability to support Ad40 replication. Only cell lines which supply E1B functions, but not those expressing E1A alone, are permissive for Ad40, suggesting that Ad40 may require some function supplied by E1B or induced in E1B-containing cells. In coinfection assays Ad40 complements Ad5 dl312 (delta E1A) but not Ad5 dl313 (delta E1B) and is itself complemented by dl312 but not by dl313. Mutants of Ad2 and Ad12 with lesions in E1B 55K or 19K protein have been used to further delineate the requirements for Ad40 growth in HeLa cells. For mutants lacking 55K function there is minimal complementation in either direction, whereas those lacking only the 19K product are able to complement Ad40.

Evaluation of a commercial monoclonal antibody-based enzyme immunoassay for detection of adenovirus types 40 and 41 in stool specimens

A commercial monoclonal antibody-based enzyme immunoassay (Adenoscreen; Mercia Diagnostics Ltd., Guildford, United Kingdom) for the detection of adenovirus types 40 and 41 in stool specimens was evaluated. Two assay modes were tested. In the first, 177 stool samples were screened for the presence of adenovirus type 40 or 41 (assay mode 1). Virus was detected in 79 of 82 specimens positive for adenovirus type 40 or 41 by a polyclonal antibody-based immune electron microscope test, giving a sensitivity of 96.3%. The enzyme immunoassay was negative in 91 of 95 stool samples which contained either other adenovirus serotypes or other viruses or were virus negative. The specificity was thus 95.8%. The positive and negative predictive values of this assay against immune electron microscopy were 95.2 and 96.8%, respectively, and the diagnostic accuracy was 96.0%. Viruses from the three false-negative enzyme immunoassay stool samples were verified as adenovirus type 40 or 41 by restriction enzyme analysis, monoclonal antibody-based immune electron microscopy, or both. Two of the three false-negative stool samples were subsequently concentrated by ultracentrifugation, and one of the two stool samples was then positive by enzyme immunoassay. The third false-negative virus was typed as adenovirus type 41 in the second (serotyping) enzyme immunoassay mode. The four enzyme immunoassay false-positive stool samples all contained other adenovirus serotypes (two were type 2, and two were type 5), but no cross-reactivity was seen with other strains of these serotypes and the results probably reflected simultaneous excretion of adenovirus type 40 or 41 with other adenovirus serotypes. In the second assay mode viruses from 15 stool samples were serotyped. The results by enzyme immunoassay (4 were type 40 and 11 were type 41) correlated completely with previous results from restriction endonuclease analyses. The commercial enzyme immunoassay system showed excellent sensitivity and specificity for the detection of adenovirus types 40 and 41 in stool specimens and will make an important contribution to the accurate diagnosis of adenovirus gastroenteritis.

New genome types of adenovirus types 1, 3, and 5 isolated from stools of children in Brazil

During an epidemiological survey made in São Paulo (Brazil), fecal specimens were periodically collected from 100 randomly chosen babies from birth to the age of 18 months. The stools, routinely collected each month and also collected each time a child presented any sign of disease, were screened for the presence of adenoviruses. Sixteen adenovirus strains, isolated from the stools of healthy and ill children, were characterized by restriction enzyme analysis. Five isolates were from subgenus A, five were from subgenus B, four were from subgenus C, and two were from subgenus D. All but two showed some restriction patterns different from those of the 42 human adenovirus prototypes and all the genome types described up to now. No fastidious adenovirus (subgenus F, serotypes 40 and 41) was encountered in the stools examined. We report here the restriction enzyme analysis of isolates of subgenera B and C. The following new designation genome types are proposed: Ad3e1 (subgenus B) and Ad1d, Ad5a1, and Ad5a2 (subgenus C).

[Presence of rotavirus and adenovirus in fecal samples of children with gastroenteritis, in the city of Goyania]

In an attempt to detect rotavirus and adenovirus prevalence among other enteropathogens (bacteria and parasites) in diarrhoea, three hundred fecal samples originating from children living in Goiânia city (Goiás state, Brazil) were analysed. Rotavirus was found to be the only pathogen in 47 cases, and associated with other infectious agents in 21 cases. 97.0% positive samples of rotavirus showed an electrophoretic pattern characteristic of subgroup II. Adenovirus was found in 7 cases, and associated with other microorganisms in 1 case. Three methods were applied for virological analyses: enzyme immunoassay for rotavirus and adenovirus (EIARA), polyacrylamide gel electrophoresis (PAGE) and immunoelectron microscopy (IEM). The concordance among the three methods was 92.8%, PAGE and EIARA agreed in 95.8%, and IEM and EIARA agreed in 100.0%.

Genome type analysis of adenovirus 31, a potential causative agent of infants' enteritis

DNA restriction analysis with 7 endonucleases was performed on 21 adenovirus type 31 strains, isolated from faeces of infants from 1974 to 1987 in different parts of Germany. The strains could be grouped into five genome types, differing from the prototype in two to four endonucleases. BglII was the most discriminative enzyme. The D2 genome type was found 16 times; all others only once or twice. No serological differences were found between the strains.

Genome type analysis of adenoviruses: isolates from one year from the Hannover area

Adenoviruses (AV), isolated from 138 children during the year 1981 in the Hannover area, were studied by DNA restriction analysis with the enzymes BamHI, BglII, BstEII, EcoRI, HindIII, KpnI, and SmaI and compared with the respective prototypes. Varying fragment patterns were depicted and genome types analyzed. Prototype-like strains of AV1 and 5 were not found. Types 2, 5, 1, 3, and 7 showed decreasing genetic variation in that order. Altered restriction sites were physically mapped on the genome; they appeared to be randomly distributed. The high genetic variability of AV2 and 5 is remarkable for this study population, limited in time and space. Clinical and epidemiological data were also presented in relation to serotypes.

Characterization of fastidious adenovirus types 40 and 41 by DNA restriction enzyme analysis and by neutralizing monoclonal antibodies

The DNA of 48 strains of adenovirus type 40 (Ad40) and of 128 strains of adenovirus type 41 (Ad41), isolated between 1971 and 1986 from various countries, was characterized by restriction enzyme analysis using nine and ten restriction endonucleases respectively. Five new DNA variants of Ad40 and 18 new DNA variants of Ad41 were detected. Most of the restriction sites which differed among the various DNA variants appeared to be distributed at random over the entire length of the viral genomes of the two serotypes. The number of restriction sites by which two DNA variants differed from each other was used as a measure of their relatedness. Several clusters of closely related DNA variants were observed for each of the two serotypes. The 35 DNA variants of Ad40 and Ad41 were used to test monoclonal antibody preparations for their range of reactivity in a neutralization assay. One monoclonal antibody (5-8), raised against Ad40 strain Dugan, showed type-specific neutralization of all 11 Ad40 DNA variants tested. Six monoclonal antibodies, raised against Ad41 strain Tak, neutralized different proportions of the variants of Ad41. Two of these preparations (1-21 and 3-19) neutralized all 24 Ad41 DNA variants, while a third (1-23) reacted with only 12 Ad41 variants. Three other monoclonal antibody preparations (3-10, 3-18, 7-14) reacted specifically with only 6 of these 12 variants. The patterns of reactivity with the monoclonal antibody preparations correlated with the presence or absence of a HindIII restriction site at 56 map units and of an EcoRI restriction site at 52 map units on the Ad41 DNA. This region of the adenovirus DNA codes for the hexon protein, which is known to contain the type-specific neutralizing antigenic determinants.

Human viral gastroenteritis

During the last 15 years, several different groups of fastidious viruses that are responsible for a large proportion of acute viral gastroenteritis cases have been discovered by the electron microscopic examination of stool specimens. This disease is one of the most prevalent and serious clinical syndromes seen around the world, especially in children. Rotaviruses, in the family Reoviridae, and fastidious fecal adenoviruses account for much of the viral gastroenteritis in infants and young children, whereas the small caliciviruses and unclassified astroviruses, and possibly enteric coronaviruses, are responsible for significantly fewer cases overall. In addition to electron microscopy, enzyme immunoassays and other rapid antigen detection systems have been developed to detect rotaviruses and fastidious fecal adenoviruses in the stool specimens of both nonhospitalized patients and those hospitalized for dehydration and electrolyte imbalance. Experimental rotavirus vaccines have also been developed, due to the prevalence and seriousness of rotavirus infection. The small, unclassified Norwalk virus and morphologically similar viruses are responsible for large and small outbreaks of acute gastroenteritis in older children, adolescents, and adults. Hospitalization of older patients infected with these viruses is usually not required, and their laboratory diagnoses have been limited primarily to research laboratories.

Genome variants of human adenovirus types 40 and 41 (subgroup F) in Japan

Nineteen strains of fecal adenovirus were typed as adenovirus type 40 (Ad40) by restriction enzyme analysis of the DNA, using SmaI. The DNAs of the strains were further compared by digestion with four other enzymes, namely, HindIII, EcoRI, BglI, and BamHI. At least two different restriction profiles were found to exist for BamHI. Of the 19 strains studied, 17 had the same combination of profiles for all of the enzymes. Thirteen strains of fecal adenovirus were typed as Ad41 by restriction enzyme analysis of the DNA, using SmaI. The DNAs of the strains were further compared by digestion with four other enzymes, as for Ad40. At least two different restriction profiles were found to exist for each enzyme. Of 13 strains, 9 had the same cleavage sites for all of the enzymes, indicating a close relationship. Two strains were genome variants which deviated from the common genome on digestion of the DNA with HindIII. One strain was a genome variant which deviated from the common genome on digestion of the DNA with HindIII and EcoRI. The last one had a combination of restriction sites which was not identical to those of other strains on digestion of the DNA with SmaI. All 13 Ad41 strains studied were easily distinguished from Ad40 strains by using any of the five restriction enzymes.

Comparison of cytopathogenicity, immunofluorescence and In situ DNA hybridization as methods for the detection of adenoviruses

Three different methods were compared for their efficiency at detection of adenoviruses. The samples examined for viral analysis consisted of concentrates prepared from raw sewage, chosen as providing a representation of the spectrum of viruses being intestinally shed from a large population at any given time. When using one single cell line, HEp-2, the overall numbers of adenoviruses detected using cytopathogenicity and immunofluorescence were roughly equal. In situ hybridization was approx. 40% more sensitive than either of these other methods as determined by average virus titers for the different samples, and also proved to be better by means of a nonparametric comparison. The 293 cell line was approx. 5 times more sensitive for detecting adenoviruses by cytopathogenicity as compared with the HEp-2 cell line, but proved unsuitable in our hands for quantitatively detecting indigenous adenoviruses by immunofluorescence. The relative number of indigenous adenoviruses present in the sewage concentrates we examined was, on average, 94-fold greater than that of enteroviruses. Assay of enteroviruses was performed by plaque assay in the BGM cell line.

One-year prospective cross-sectional study to assess the importance of group F adenovirus infections in children under 2 years admitted to hospital

A 1-year prospective cross-sectional study of 363 children under 2 years of age admitted to hospital was undertaken to assess the importance of group F adenovirus infections. Faeces obtained within 48 hours of admission from 97 patients with and 266 patients without diarrhoea were screened by electron microscopy. Viruses were identified by morphological criteria, and all adenoviruses seen were retested by immune electron microscopy to identify group F serotypes. Group F adenoviruses (4 infections) were second in frequency to rotaviruses (16 infections), and both viruses were significantly associated with diarrhoea (P = 0.005 and 0.00001 respectively, chi-squared test). All four group F infections occurred in children with diarrhoeal disease aged between 1 and 6 months and were numerically as important as rotavirus (three infections) in this group. Rotavirus infections occurred significantly more frequently in the 7-24-month age group with diarrhoea (11 v.0 infections, P = 0.001, chi-squared test). Nosocomial infection occurred with group F adenovirus as well as rotavirus. The finding that group F adenoviruses occur as frequently as rotaviruses in diarrhoeal disease that results in hospital admission in children between 1 and 6 months of age could have important implications for preventative strategies.

Genome typing of adenovirus strains isolated from conjunctivitis in Japan, Australia, and the Philippines

DNA restriction endonuclease analysis for genome typing of adenovirus (Ad) DNA was carried out on a total of 65 Ad isolates including serotypes Ad4, Ad7, Ad8, Ad11, Ad19, and Ad37 from patients with epidemic keratoconjunctivitis and acute conjunctivitis obtained in Japan from 1982 to 1986, Australia from 1973 to 1986, and the Philippines in 1984. All 4 isolates of Ad7 in Australia were Ad7b. Four of 6 Ad11 isolates obtained in Japan were typed as Ad11 prototype (Ad11p), and the remaining were identified to be new genome types, designated tentatively as Ad11c and Ad11d. An isolate of Ad11 obtained in Australia was typed as Ad11c. Nine Ad8 isolates in Australia and in the Philippines were typed as Ad8p, but 11 Ad8 isolates in Japan were Ad8b. Thirteen Ad19 isolates were identified as Ad19a. All 3 isolates of Ad37 in Japan and three isolates in Australia before 1982 were typed as Ad37p, however, 5 isolates in Australia after 1983 were identified as a new genome type, designated as Ad37d. In Japan, 10 isolates of Ad4 were identified as Ad4a.

Restriction enzyme analysis of faecal adenoviruses in Newcastle upon Tyne

Adenovirus DNA was isolated directly from virus-containing stools and digested with restriction endonucleases. The resulting fragments were separated by polyacrylamide gel electrophoresis (PAGE) and visualized by silver staining. This enabled us to assign most of the viruses detected to subgenus, serotype and, sometimes, unique strains. Although less sensitive than electron microscopy, the method allowed more information about the infecting virus to be obtained and no cultivation was necessary. Comparison with culture also allowed dual infections to be recognized. A 2-year survey of faecal adenoviruses in Newcastle upon Tyne showed that type 41 (strain 41a) was the predominant type and strain 41p was not recorded. Heterogeneity in strain 41a was also noted as found elsewhere. Adenovirus type 40 was common prior to 1985 but was absent during the last 2 years.

Suppression of viral replication by guanidine: a comparison of human adenoviruses and enteroviruses

A comparison was made between the relative sensitivities of laboratory strain human adenoviruses and enteroviruses, and recently isolated human enteroviruses, to the presence of guanidine hydrochloride in cell culture media. The concentration of guanidine hydrochloride used was 100 micrograms per ml. Representatives of all six human Adenovirus subgenera were unaffected in their replication at this concentration of guanidine. The different human Enterovirus types examined varied in their sensitivity, with suppression ranging from less than 1 to 3 log10 units for laboratory strains, and from 2 to 7 log10 units for recently isolated viruses. The findings suggest a novel role for antiviral drugs; serving as an adjunct in facilitating selective isolation of specific virus groups which may be present as part of mixed viral populations.

Helper function of adenovirus 2 for adenovirus 41 antigen synthesis in semi-permissive and non-permissive cells

Using a fluorescent focus assay, complementation and interference effects of Ad2 and Ad41 on each other in mixed infection were investigated. Ad2 provided a helper function for Ad41 late antigen synthesis in cells normally non-permissive for Ad41 growth (HEF cells), and enhanced Ad41 late antigen synthesis in semi-permissive Chang conjunctival cells. The degree of complementation by Ad2 was dependent on its input concentration. In addition, interference by Ad41 on Ad2 replication was seen in HEF cells. The degree of interference by Ad41 was dependent on the relative time of infection by each serotype. The complementation results in HEF cells suggest an absolute dependency of Ad41 on an adenovirus helper function in these cells. The results presented are consistent with the postulated helper function provided in trans by 293 cells, which are transformed by Ad5 early region 1.

Incidence of enteric adenoviruses among children in Thailand and the significance of these viruses in gastroenteritis

In countries with temperate climates, enteric adenoviruses have been shown to be a substantial cause of pediatric gastroenteritis. To determine the incidence of adenovirus infection in a tropical climate, stools were collected from children under age 7 during a 1-year period at an outpatient clinic in Bangkok, Thailand. Stools from 1,114 children with gastroenteritis and from 947 children without gastroenteritis were tested. Each stool was tested for adenovirus group antigen and for specific enteric adenovirus types (Ad40 and Ad41) by monoclonal antibody enzyme immunoassays. We found that 4.4% (49 of 1,114) of children with gastroenteritis and 1.8% (17 of 947) of children without gastroenteritis were positive for adenovirus group antigen. In tests for specific enteric adenovirus types, 2.0% (22 of 1,114) of the tests were positive in children with gastroenteritis and 0.6% (6 of 947) were positive in children without gastroenteritis. There was a significant correlation (P less than 0.02) of gastroenteritis with nonenteric adenovirus types (27 of 1,114) as well as with specific enteric adenovirus types (P less than 0.01). By comparison, 19.7% of children with gastroenteritis and 0.7% of those without gastroenteritis were positive for rotavirus infection. In the adenovirus-infected children with gastroenteritis, there were coinfections with rotavirus only in those with nonenteric adenovirus infection (7 of 27 children). There were no significant differences in the association of bacterial or parasitic infections with either enteric or nonenteric adenovirus infections in either group of children studied. These data demonstrate that Ad40 and Ad41 are causes of gastroenteritis in this population, but among the spectrum of diarrheal etiologies, they may be proportionately less important than they are in countries with temperate climates.

Genomic comparison of adenovirus type 3 isolates from patients with acute conjunctivitis in Japan, Australia, and the Philippines

A total of 53 Ad3 isolates obtained from patients with acute conjunctivitis in Japan, Australia, and the Philippines during 1973 to 1986 were analyzed for their genome types with 4 restriction endonucleases, BamHI, BglII, HindIII, and SmaI. Two new genome types designated tentatively as Ad3f and Ad3g were identified by combination of BamHI and BglII in the isolates. The changes of restriction sites and sizes of restriction fragments in newly recognized Ad3f and Ad3g were located at the similar regions reported in other Ad3 genome types by O'Donnell et al (1986) on physical maps of the Ad3 prototype strain GB genome. In Japan, 46 Ad3 isolates obtained from 1983 to 1986 were either Ad3f or Ad3g. Yearly alternation of predominance of both genome types were observed in the northern part of Japan during the period. In Australia, two genome types Ad3p and Ad3f were found in 6 isolates, and the former was observed in the 3 isolates obtained before 1981 and the remaining 3 isolates were obtained after 1983. In the Philippines, the only isolate obtained in 1984 was Ad3p.

Isolation and propagation of enteric adenoviruses in HEp-2 cells

Eighty-two stool samples from children with gastroenteritis in Canada, England, and Thailand which had been shown to contain adenovirus antigen (by a group-specific enzyme-linked immunosorbent assay) or adenovirus particles (by electron microscopy) or both, were tested for primary isolation of enteric adenoviruses in HEp-2 and Graham 293 cells. Graham 293 cells are known to support the replication of enteric adenovirus types (Ad40 and Ad41) on primary isolation, whereas HEp-2 cells reportedly do not. Of the 82 adenovirus isolates, 73 could be typed as Ad40 or Ad41 by type-specific monoclonal antibody enzyme-linked immunosorbent assay and by analysis of SmaI endonuclease digests. Of these 73, 30 (41%) could be isolated in HEp-2 cells, which included 43% (9/21) of those typed as Ad40 and 40% (21/52) of those typed as Ad41. On the basis of these results, the growth characteristics of adenoviruses in HEp-2 cell cultures, commonly used to distinguish enteric from nonenteric adenovirus types, are not valid for either diagnosis or epidemiological studies. For the samples studied here, use of these nondefinitive criteria would result in underestimation of the incidence of enteric adenoviruses in viral gastroenteritis.

Comparison of enteric adenovirus infection in various human cell lines

Growth of the fastidious enteric adenoviruses 40 and 41 was compared in different human cell lines. Purified virions were used to infect the following cell lines: A549; KB; Chang's conjunctiva; 293; HeLa. Both types of enteric adenovirus were infectious for each cell line, with the exception of adenovirus 40 in HeLa cells. Relatively low infectious titers were obtained from each cell type following infection with adenovirus 40 (TCID50 average = 10(-1.5)), whereas adenovirus 41 replicated to significantly higher titer (TCID50 average = 10(-3.0)). For both viruses, the highest infectious titers were obtained with A549 and KB cells. A time course experiment performed to quantitate the amount of hexon present in A549 and KB cells infected with each virus indicated that while the kinetics of accumulation were similar for both viruses, the concentration of type 41 hexon was significantly greater than that for type 40 in either cell line. The concentration of type 41 hexon was similar in each cell type; for type 40, a greater concentration of hexon was obtained in the A549 cell line than in the KB cells. The results indicate the distinct replication characteristics exhibited by adenovirus 40 are not due to a restriction in a specific host cell, and, because purified virions were used, not attributable to interference that might occur with co-infection from multiple viruses present in the same clinical specimen. We conclude the differences observed in the replication of these viruses are independent of host cell type and are associated, uniquely, with each virus.

Characterization of adenovirus type 40 E1 region

The left-most 3.9 kb of adenovirus type 40 (Ad40) DNA has been sequenced using cloned viral DNA fragments. The Ad40 E1 region is deduced to code for at least four polypeptides, 221 and 249 amino acids as E1A products in addition to 166 and 475 amino acids as E1B products. E1B polypeptides share about 50% homology with well-defined adenovirus types, 2/5, 7, and 12, throughout the E1B sequences. E1A homology of Ad40 to these types is relatively lower than that of E1B, while highly conserved regions of E1A are retained to a certain level in Ad40 as well. Activity for morphological transformation of Ad40 E1A on 3Y1 cells is considerably lower when compared to that of Ad5 and Ad12 E1A genes. Transient chloramphenicol acetyltransferase (CAT) expression assay shows that Ad40 E1A has a trans-acting function, though lower than that of other E1A genes, on adenovirus early promoter. The Ad40 E1A promoter also holds only a little cis-acting activity in 3Y1 cells. Lower activities of both Ad40 E1A promoter and certain E1A functions may explain in part the difficulty in propagation of Ad40.

Physical organization of the enteric adenovirus type 41 early region 1A

Enteric adenovirus types 40 and 41 (Ad40 and Ad41), representing subgenus F, differ from all other human adenoviruses by being so fastidious that productive replication does not occur in conventional established cell lines. They are dependent of the Ad5 early regions E1A and E1B since they can not grow in HEK cells, only in 293 HEK cells transformed by Ad5 E1. The overall genetic organization of Ad41 E1A is similar to the E1A region of other characterized human adenoviruses but it is slightly shorter, comprising 1350 bp. The inverted terminal repeat (ITR) at the 5' end of both Ad40 and AD41 consists of 163 nucleotides, being similar to the ITR of Ad12 (subgenus A) and longer than the ITRs of adenoviruses of subgenera B, C, and E. The early mRNA products (12 and 13 S) can be translated into a 222-amino acid (aa) and a 251-aa tentative protein, respectively. In a comparison of the Ad41 251-aa protein with corresponding peptides of Ad12, Ad7, Ad5, and Ad4, three conserved amino acid sequences CS1-CS3 can be found. In the second conserved domain CS2, which is particularly acidic, the homology is very high within all five serotypes compared. Only one among eight conserved amino acids differs in the Ad41 251-aa protein. Within CS1 and CS3 which exhibit a hydrophilic and a hydrophobic character, respectively, the amino acid composition of the Ad41 protein is less conserved than the corresponding regions in all other analyzed adenovirus types. Ten of 16 conserved amino acids in CS1 are shared by Ad41 and 18 of 23 conserved amino acids in CS3 are shared by Ad41.

The genes encoding the DNA binding protein and the 23K protease of adenovirus types 40 and 41

The adenovirus (Ad) single-stranded DNA binding protein (DBP) is a multifunctional protein. It is thought to consist of two domains, the amino-terminal domain involved in host-range determination and the carboxyl-terminal domain functioning in DNA replication and DNA binding. We have determined the nucleotide sequences of the DBP genes of Ad40 and Ad41, two human adenoviral serotypes that differ significantly from other adenoviruses. Regions of structural and functional importance in the corresponding proteins could be identified by comparison of the amino acid sequences with those of other known DBPs. In addition, the nucleotide sequences of the DBP early promoters, of the 23K protease genes, and of parts of the hexon and 100K protein genes have been determined. It can be deduced from the nucleotide sequences, that the Ad40 and Ad41 DBPs are relatively small (473 and 474 amino acids (a.a.), respectively, versus 529 a.a. for the Ad5 DBP). This is caused by the presence of very small amino-terminal domains of 119 a.a. (Ad40) and 120 a.a. (Ad41), as compared to 173 a.a. for the corresponding Ad5 domain. Only a few amino acids in this domain have been conserved in all known DBPs. The carboxyl-terminal domains show a higher degree of sequence conservation. In this domain, four strongly conserved regions can be identified, one of which might form a metal-binding site. The 23K proteases of both Ad40 and Ad41 show a strong homology to the Ad2 and Ad5 proteins, with the exception of the carboxyl-terminal end of the proteins. The 23K protease gene of Ad41 has an open reading frame that extends beyond the polyadenylation signal, in contrast to the Ad40 gene that ends well in front of the signal.

A study of adenovirus gastroenteritis in the Tokyo area

Probable serotypes of faecal adenoviruses 40, 41, 31, 2 and 1 were found in the Tokyo area by rapid DNA restriction endonuclease analyses directly of stool specimens or after primary culture in Graham 293 cells. Some variations were found in both adenoviruses 40 and 41 by this method.

DNA hybridization for diagnosis of enteric adenovirus infection from directly spotted human fecal specimens

By using a genomic probe, DNA hybridization for adenovirus type 41 (Ad41) showed equivalent sensitivity with a direct spot method from clinical specimens compared with a more laborious DNA phenol extraction procedure. By using this direct spot preparation method, fecal specimens of 67 patients were examined under code for blind testing for the presence of adenovirus by DNA hybridization by using two Ad41 probes (genomic and cloned BglII-D) and an adenovirus type 2 genomic probe. Identical results were obtained with both of the Ad41 probes. Of the fecal specimens from 42 children with adenovirus gastroenteritis studied prospectively (16 of whom had enteric adenoviruses), 13 specimens (81%) were detected by DNA hybridization with a cloned Ad41 BglII-D probe. There were 14 fecal specimens that were positive by electron microscopy (EM) and culture for nonenteric adenovirus, and 2 specimens were positive by DNA hybridization (87% specificity); these 2 specimens may have been from a mixed enteric adenovirus and nonenteric adenovirus infection. None of 26 specimens from age-matched healthy control patients was positive for adenovirus by EM or DNA hybridization. Our data indicated that DNA hybridization gives highly reproducible results. The direct spot technique is the method of choice for specimen preparation in the diagnostic laboratory, since it requires only the simplest manipulations in specimen preparation. By using DNA hybridization with the BglII D fragment of a cloned enteric Ad41, both adenovirus type 40 and Ad41 were detected directly from fecal specimens, but it was less sensitive than EM following direct ultracentrifugation of specimens. The Bg1II-D Ad41 DNA probe was highly specific for enteric adenoviruses, and DNA hybridization with this probe could be a useful diagnostic test for these fastidious adenoviruses.

Comparison of monoclonal time-resolved fluoroimmunoassay with monoclonal capture-biotinylated detector enzyme immunoassay for adenovirus antigen detection

A sensitive time-resolved fluoroimmunoassay (TR-FIA), adapted from TR-FIA procedures already described, was developed with monoclonal antibodies and compared with several enzyme immunoassays (EIAs) for detecting adenovirus antigens in clinical specimens. The most sensitive EIA was an all-monoclonal assay with biotin-labeled detector antibody and streptavidin-peroxidase conjugate. All tests were evaluated with nasopharyngeal aspirate specimens from respiratory illness, with tissue homogenates from patients with systemic infection, and with stool specimens from gastrointestinal illnesses. For respiratory and tissue specimens, the TR-FIA detected adenovirus in 85% of the specimens positive by culture, which was a sensitivity similar to those of the all-monoclonal biotin-avidin EIA (79%) and the polyclonal-capture biotin-avidin EIA (88%). For stool specimens, the TR-FIA detected adenovirus in 100% of the specimens positive by culture, which was a decidedly higher sensitivity than either EIA format (78 and 75%, respectively). The TR-FIA was shown to be an efficient, flexible, and specific test for large numbers of clinical specimens.

Antibody response to enteric adenovirus types 40 and 41 in sera from people in various age groups

A total of 322 single and paired serum samples from children (newborn to 10 years old), young adults (18 to 20 years old), pregnant women and their cord serum samples, and elderly people (more than 70 years old) was tested for antibodies to enteric adenovirus types 40 and 41 by neutralization test. Serum samples were also tested for antibody to the common antigen of adenovirus by enzyme immunoassay. The incidence of antibodies rose gradually through childhood. Antibodies were found in 20% of children between 1 and 6 months old and in 50% of those 37 to 48 months old. Of serum samples from young adults, 48% had antibodies. Antibodies were found in 10% of serum samples from the aged. Of patients with acute gastroenteritis, 19% showed a significant rise in antibody to adenovirus type 40 or 41 or both, and 42% of the same serum samples had a significant rise in antibody to rotavirus by enzyme immunoassay. None of the serum samples tested was negative to adenovirus common antigen.

Transcriptional activation by the E1A regions of adenovirus types 40 and 41

In order to establish whether the poor growth of the two fastidious adenoviruses types 40 and 41 (Ad40 and Ad41) in HeLa cells is due to a reduced trans-activation by the early region 1A (E1A), we have determined the trans-activating effect of this region on the expression of the chloramphenicol acetyltransferase (CAT) gene controlled by the Ad2 E4 promoter. Cotransfection of HeLa cells with plasmids containing the E1A regions of Ad5, Ad40, and Ad41, respectively, and the CAT gene controlled by the Ad2 E4 promoter showed that activation of the E4 promoter by the E1A regions of Ad40 and Ad41 depends on the same sequence elements of the E4 promoter as activation by the Ad5 E1A gene products. The level of activation, however, is significantly lower. This might partly explain the reduced growth in HeLa cells of the two viruses.

Evaluation of three types of cell culture for recovery of adenovirus from clinical specimens

The performance characteristics of HEK, HDF, and MK cells for adenovirus isolation were examined for eye and respiratory tract specimens. HEK cells were superior to HDF and MK cells in terms of both speed of virus detection and sensitivity.

Nucleotide sequence of the adenovirus type 40 inverted terminal repeat: close relation to that of adenovirus type 5

Human adenovirus type 40 (Ad40) is a pathogen that causes acute infantile gastroenteritis. Ad40 has the distinct characteristic of being difficult to propagate in conventional cultured human cells. The nucleotide sequence of the inverted terminal repeat (ITR) of Ad40, which includes the origin of adenoviral DNA replication, was determined using recombinant plasmid DNA. By using our newly developed program to express the ITR homologies simply, we found that the ITR of Ad40, which is 163 nucleotides long, was related most closely to that of adenovirus type 5, which replicates efficiently.

Detection of adenovirus types 40 and 41 in stool specimens by immune electron microscopy

An immune electron microscope (IEM) test was developed that allowed the direct detection of adenovirus type 40 (ad 40) or ad 41 in stools specimens. The polyclonal rabbit antisera used differentiated ad 40 and 41 from other ad serotypes but not from each other. The method was evaluated in a 13 month prospective study of stools from children with gastroenteritis. Seventy-two specimens found to contain ad by conventional electron microscope screening were retested by IEM. Results were typically obtained within 2 hr and showed that 55 (76%) viruses typed as ad 40/41. No ads were recovered from conventional virus isolation attempts on these specimens. Additionally, 39 of these 55 viruses were tested by restriction endonuclease analysis (REA) after growth in 293 cells, and results showed that all produced digest patterns typical of ad 40 (seven cases) or ad 41 (32 cases). Twenty-four percent (17/72) of viruses could not be typed by IEM; 9/17 (53%) yielded ads [ad 1 (1), ad 2 (4), ad 5 (1), ad 6 (1), ad 7 (2)] in routine culture, whereas REA identified the other eight as ad 2 (6), ad 1 (1), and ad 41 (1). The concordance between IEM and the reference methods was therefore 100% specificity and 97.5% sensitivity. The method described allows the clinically useful diagnosis of ad 40/41 infection to be rapidly made and will be a particularly useful technique in laboratories screening faeces by electron microscopy.

Molecular epidemiology of adenovirus type 21 in the Netherlands and the Federal Republic of Germany from 1960 to 1985

After a period of high prevalence in the early 1960s, adenovirus serotype 21 (Ad21) was identified in The Netherlands only very sporadically for more than 20 years. From December 1984 to July 1985, Ad21 was isolated relatively often from hospitalized children living in different parts of The Netherlands. The patients in question suffered from respiratory, gastrointestinal, meningeal, or ocular disorders. An increase in the incidence of Ad21 infections was also observed in the Federal Republic of Germany during this period. The DNAs of 93 isolates of Ad21 were subjected to restriction enzyme analysis with eight endonucleases. All 50 strains isolated in The Netherlands between 1960 and 1963 proved to be DNA variant Ad21/D2/20655/Netherlands/60. This variant has already been described (T. Adrian, R. Wigand, and J. C. Hierholzer, Arch. Virol. 84:79-89, 1985) as typical for the Ad21 strains circulating since 1960. Analysis of the DNAs of the 28 Ad21 strains isolated in The Netherlands or in the Federal Republic of Germany in 1984 and 1985 showed them to belong to two new, closely related DNA variants designated Ad21/D7/1857/Netherlands/84 and Ad21/D8/5398/Netherlands/85. The BglI and KpnI restriction profiles were characteristic for these recent DNA variants.

Direct detection and differentiation of fastidious and nonfastidious adenoviruses in stools by using a specific nonradioactive probe

Restriction fragments of fastidious human adenovirus type 41 (Ad41) were cloned in vector plasmid pBR322. A rapid and sensitive nonradioactive molecular-hybridization technique (M. Renz and C. Kurz, Nucleic Acids Res. 12:3435-3444, 1984) showed that one clone specifically detected fastidious Ad40 and Ad41 (subgenus F) without cross-hybridization with nonfastidious adenoviruses. This clone was mapped in a region of the Ad41 genome corresponding to early transcription unit E1B of Ad2. A number of DNAs from fastidious and nonfastidious adenoviruses were extracted, without cultivation, from stools of children with gastroenteritis and were hybridized with an Ad2 probe and with the cloned probe, allowing the differentiation of the two groups of viruses. This method could detect DNA quantities as low as 10 pg and should be particularly suitable for stool samples containing adenoviral DNA in amounts too low to be detected by staining with ethidium bromide.

Analysis of 15 different genome types of adenovirus type 7 isolated on five continents

A total of 15 different genome types of adenovirus type 7 (Ad7), i.e., Ad7p, Ad7p1, Ad7a, Ad7a1 to Ad7a5, Ad7b, Ad7c, Ad7d, Ad7d1, Ad7e, Ad7f, and Ad7g, were identified among 40 selected strains isolated in Europe, Asia, North America, South America, and Australia by using restriction endonucleases BamHI, BclI, BglI, BglII, BstEII, EcoRI, HindIII, HpaI, SalI, SmaI, XbaI, and XhoI. Eight of them, Ad7p1, Ad7a1 to Ad7a5, Ad7d1, and Ad7g, are newly discovered. All 15 genome types could be distinguished by the four restriction endonucleases BamHI, BclI, BglI, and XbaI. At least four restriction sites differed between Ad7d and Ad7g. Pairwise analyses of comigrating DNA restriction fragments of all 15 Ad7 genome types were performed and presented in a schematic fashion. According to the degree of comigration of DNA restriction fragments, the 15 genome types could be divided into three clusters. Ad7b was the dominant genome type in different parts of the world and may have evolved in China into Ad7d and further to Ad7d1.