Acute gastroenteritis is one of the most common diseases and causes of death worldwide. Rotaviruses, caliciviruses, adenoviruses and astroviruses have been recognized as the main viral agents responsible for this condition. There is, however, a large diagnostic gap, resulting in the etiologic agent not being identified in approximately a third of all specimens, and it has been suggested that other viruses, including picornaviruses, are involved [1].

Aichi virus (AiV) was first discovered in 1989 during a search for an infectious agent in an outbreak of oyster-associated non-bacterial gastroenteritis [2]. The isolated virus was later classified as the first member of a novel genus, Kobuvirus, of the family Picornaviridae [3]. Kobuviruses have since then been associated with diarrhea not only in humans but also in cattle [4, 5] and swine [6, 7].

Since its discovery, AiV has been detected with an incidence of 0.9-4.1%, primarily when studying outbreaks of diarrhea in children or young adults [8–12]. AiV seems to be circulating worldwide, and detection of AiV have been reported in Asia [13], South America [14], Europe [8–10, 14] and Africa [11]. Previous studies have suggested oysters, or other seafood, to be the source of AiV infections [2, 8, 15–17], but its occurrence has also been reported in sewage [15, 18], which indicates that kobuviruses might use the fecal-oral route of infection, like several other picornaviruses. Studies in Japan [19], European countries [14, 20, 21] and Tunisia [22] show high seroprevalence, with antibodies against AiV in at least 80% of the adult population, a result that stands in contrast with the low incidence reports. The incongruity between rare isolation and high seroprevalence suggests that the virus might circulate without causing any symptoms or might cause weakly symptomatic infections that do not require medical attention. It is also possible that it co-circulates with other enteric viruses that cause gastroenteritis and has therefore not been detected during screening of patients [8, 11, 14]. The high seroprevalence data suggest a general exposition of the virus, but the importance of AiV in human diseases needs to be further studied.

Studies of AiV have primarily been performed using reverse transcription followed by PCR of a 519-bp fragment from the 3CD junction, a genomic region coding for the viral protease 3Cpro and the RNA polymerase 3Dpol [23].

In this paper, we report the first detection of AiV in Sweden. In addition, this is the first paper describing the presence of AiV in elderly patients.

RNA was isolated from 221 stool samples from hospitalized patients suffering from diarrhea at Linköping University Hospital, Sweden, who were negative for adenovirus, calicivirus and rotavirus. The isolation was done from January to June 2009 without taking the age of the patients into consideration. Nucleic acid was extracted from 200 μl of a 10% particle-cleared fecal suspension using a MagAttract Viral RNA M48 kit (QIAGEN) according to the manufacturer’s instructions and eluted in a volume of 75 μl. Extracted viral RNA was reverse transcribed by random priming using an Applied Biosystems TagMan reverse transcriptase kit according to the manufacturer′s protocol. The screening for AiV was performed using a nested PCR, first amplifying 519 bp in the 3CD junction with primers 6261 (5′-ACACTCCCACCTCCCGCCAGTA-3′) and 6779 (5′-GGAAGAGCTGGGTGTCAAGA-3′) as described previously [23]. Next, a nested PCR amplifying a 223-bp segment within the 3CD junction was carried out using primers C94b (5′-GACTTCCCCGGAGTCGTCGTCT-3′) and 246 k (5′-GACATCCGGTTGACGTTGAC-3′) [13]. Additional PCR to amplify VP1 was performed using primers CapE (5′-CTAGTCGGACCCACACCGC-3′) [24] and R3636 (5′-GCAAGAGAGCTGGAAGT-3′) [8] on samples positive for AiV. The nucleotide sequences of the PCR products were determined using an ABI Prism 3130 automated sequencer and BigDye chemistry (Applied Biosystems). The sequences obtained were analyzed using the Sequencher, version 4.6, software package (Gene Codes Corporation). AiV sequences were deposited in GenBank with the accession numbers JN584163-65 for the 3CD junction and JN584162 for the VP1 sequence. Nucleotide sequences were aligned using ClustalW [25], and phylogenetic signals were evaluated by likelihood mapping [26]. The phylogenetic relationships between the different AiV strains based on the 3CD junction and VP1 were constructed using the maximum-likelihood method as implemented in PhyML [27] and evaluated by non-parametric bootstrap analysis consisting of 1000 pseudo-replicates [28]. The general time-reversible (GTR) substitution, determined by the Modeltest program, version 3.7 [29], with gamma-distributed rate of heterogenicity among nucleotide positions was used for the analysis. The phylogenetic relationship was also examined by the neighbor-joining method as implemented in MEGA, version 4.0 [30]. Trees were visualized in MEGA.

In this report, we describe the first cases of AiV in Sweden and, to our knowledge, the first report of AiV associated with diarrhea in elderly people. Our findings that 1.4% of the samples were positive for AiV are in accordance with previous observations [8–10, 31].

AiV belongs to the genus Kobuvirus, a growing genus in the family Picornaviridae, whose members infect not only humans but also cattle [4, 5], swine [6], canines [32], sheep [33] and mice [34]. Since the discovery of AiV in 1988, the virus, or antibodies against it, has been detected worldwide, although knowledge about its epidemiology is still limited. Seroprevalence studies show that 80-95% of the population in their thirties has antibodies against AiV [14, 20, 21], which indicates a general and frequent circulation of the virus among humans. The reported incidence of AiV is in general low, but is significantly higher in outbreaks of gastroenteritis associated with shellfish [8, 11]. However, the high seroprevalence, even in the very young population, indicates the possibility of other transmission routes. AiV has recently been detected in sewage and sewage-polluted water in Venezuela [18] and Tunisia [15], which indicates a circulation of AiV in the environment.

AiV has been associated with gastroenteritis in children and young adults in Germany [14], France [8], Hungary [10], Tunisia [35], Pakistan [12], China [31] and Finland [9], but also in different age groups where gastroenteritis has most often been associated with food consumption [8, 12, 14, 16, 36]. The age distribution of the patients included in the present report shows that a vast majority of the 221 patients studied were over 65 years of age (Fig. 1; positive samples are indicated by arrows). All patients were hospitalized and had diarrhea, and standard screening for enteric viruses gave negative results. The majority of previous studies have been performed on children and young adults, which are age groups that are rarely hospitalized with diarrhea in Sweden. Therefore, the majority of the samples in this study originate from elderly people. Similar to our study, Drexler et al. [36] recently presented a German study of patients with approximately the same distribution of age as in this paper. AiV was detected in 2% of the patients, with age ranging from children to adults in their forties.

Fig. 1
figure 1

Age distribution of patients included in the screening of AiV. Two hundred twenty-one samples that previously had been reported to be negative for rotaviruses, caliciviruses and adenoviruses were screened for AiV using a nested PCR in the 3CD junction. Arrows indicate patients who were positive for AiV

Full size image

In this report, we identified three Swedish samples that were positive for AiV, all from elderly people. From patient one, an 88-year-old male, we were able to generate sequence data from the 3CD junction as well as the gene coding for the structural protein VP1. This patient was hospitalized because of deep venous thrombosis with secondary lung embolism. After three days of diarrhea and fever, a fecal sample was sent for viral diagnostics. Patient two was a 68-year-old female hospitalized in the intensive care unit because of trauma with multiple fractures, pneumothorax and splenectomy. She was initially constipated and treated with laxatives; a fecal sample was taken because of diarrhea secondary to this treatment. Hence, this patient had a plausible reason for diarrhea other than that of viral origin. Patient three, an 88-year-old male with diabetes type 2 and cardiovascular disease, was hospitalized because of fever, diarrhea, vomiting and skin eruptions two to three days prior to hospitalization. He developed high fever and was diagnosed with generalized herpes zoster and sepsis caused by Gram-negative bacteria.

Yamashita et al. [23] has suggested that AiV isolates can be divided into two groups, genotype A and B, based on sequence analysis of a 519-bp sequence in the 3CD junction. In addition to these groups, a third genotype, genotype C, has been suggested [8], comprising only one isolate so far. The genotype classification has been confirmed, and it has been shown that there is a good correlation between phylogenetic trees from both the well-conserved 3CD and VP1, which has higher genetic diversity [8, 37]. The AiV sequences available indicate that the genotype classification can be based on either genomic region. However, molecular typing of the VP1 sequence is the golden standard used for the classification of picornavirus (sero) types [38–40]. The capsid region of genotype B includes three nucleotide insertions within the VP0 region that are not present in genotype A [3, 14, 24]. Genetic and phylogenetic analysis of the three Swedish AiV strains identified in this study showed that they were members of genotype A (Fig. 2), which is the most common genotype found in Europe and Japan [8]. All isolates were genetically close to each other and to previously reported Asian strains.

Fig. 2
figure 2

Phylogenetic trees representing the relationship of Aichi virus strains. (A) Phylogram based on the ML method using a 225-nt sequence from the 3CD junction (GenBank accession numbers JN584163-65 for isolates detected in this study) (B) Phylogeny of 500 nt of the VP1 gene (GenBank accession number JN584162 for the isolate presented in this study). Both ML trees were evaluated by non-parametric bootstrap analysis with 1,000 pseudo-replicates. Only bootstrap values ≥70% are shown. Scale bars represent the genetic distance (nucleotide substitutions per site). The Swedish AiV isolates are indicated by black dots. Abbreviations in isolate names: AUS, Australia; BGD, Bangladesh; CHN, China; DEU, Germany; FRA, France; IDN, Indonesia; IND, India; JPN, Japan; PAK, Pakistan; SGP, Singapore; SWE, Sweden; THA, Thailand; TUN, Tunisia; and VNM, Vietnam

Full size image

In conclusion, AiV was detected in three elderly hospitalized patients with diarrhea in Sweden with the same incidence as was previously described for children and young adults in other countries. Our results indicate that AiV may be associated with a limited number of diarrhea cases even in elderly patients. The discrepancies between the high seroprevalence and low incidence of detection could be due to AiV infections being mainly asymptomatic, a lack of appropriate detection assays, or a lack of detection due to infections with other viruses. AiV remains rare, and the significance of AiV as a human pathogen is a question that remains to be further investigated.