3. RESULTS

3.1 Screening of rodents from Slovakia

↓32

A total of 1733 rodents representing nine species were collected between 1995 and 2001 in Slovakia (Figure 7). Among them, 1032 animals belonged to Murinae subfamily and 701 to Arvicolinae subfamily of the Muridae family. The three most common species, A. flavicollis, A. agrarius and Clethrionomys. glareolus, comprised 83.5% of captures. However, A. agrarius was found exclusively in East Slovakia whereas A. flavicollis and C. glareoluswere distributed over the whole country.

Antibodies reactive with hantavirus antigen (PUUV, HTNV and/or DOBV) were detected in six species of Muridae rodents (Table 6). However, in subsequent analyses we were able to detect viral RNA by RT PCR only in four species, A. agrarius (DOBV), A. flavicollis (DOBV), C. glareolus (PUUV) and M. arvalis (TULV). Prevalence of antibodies reactive to hantavirus antigen was highest in A. agrarius (13.7%), followed by M. arvalis (11.6%)and A. flavicollis(3.9%). However, the seroprevalence was highly variable, especially in case of M. arvalis, where 12 out 15 positive mice were captured during years 1995-96. Nevertheless, the data show that DOBV circulating in Apodemus mice represents the most common hantavirus in Slovakia.

↓33

Of 984 rodents tested in the western Slovakia, 36 (3.7%) were sero-positive. In Eastern Slovakia we found 75 (10.0%) positives among 749 rodents (Table 6). This difference was significant (chi-square = 28.65) but can be explained by geographical distribution of A. agrarius which represents the most frequently infected species in Slovakia and can be found only in Eastern part of the country. After excluding the data of A. agrarius from the analysis, the distribution was not significant (chi square = 0.04). 39 individual sites were surveyed in Western (11 sites) and Eastern part (28 sites) of Slovakia. Antibody prevalence at individual sites varied from 0.0 to 26.0%. Three sites with highest antibody prevalence, Slanske Nove Mesto (26.0%), Cerhov (24.4%) and Rozhanovce (21,7%), are all in Eastern Slovakia, what can be also explained by presence of A. agrarius at those localities. Hantavirus antibody prevalence was 7.58% in males and 5.07% in females. This difference was significant (chi-square = 4.53).

Figure 7: Map of Central Europe.

The green dots indicate the regions of trapping activities in Slovakia (Chapter 3.1). The red dot indicates the place where the first patient-associated DOBV-Aa sequence have been obtained (Chapter 3.5). The blue dot indicates the area where a TULV-associated HFRS case has been reported and TULV sequences from Microtus arvalis rodents have been obtained (Chapter 3.7).

Table 6: Prevalence of antibodies to HTNV, PUUV and/or DOBV among wild rodents in Slovakia, 1995-2001.

Eastern Slovakia

Western Slovakia

Slovakia (total)

Subfamily/

No.

No.

%

No.

No.

%

No.

No.

%

Species

Tested

Pos.

Pos.

Tested

Pos.

Pos.

Tested

Pos.

Pos.

Murinae

A. agrarius

467

64

13.7

0

-

-

467

64

13.7

A. flavicollis

119

2

1.7

290

14

4.8

409

16

3.9

A. microps

42

1

2.4

1

0

0

43

1

2.3

A. sylvaticus

1

0

0

103

1

1.0

104

1

1.0

Mus musculus

2

0

0

6

0

0

8

0

0.0

Micromys minutus

1

0

0

0

0

-

1

0

0.0

SUBTOTAL

632

67

10.6

400

15

3.8

1032

82

7.9

Arvicolinae

Clethrionomys glareolus

58

2

3.4

513

12

2.3

571

14

2.5

Microtus arvalis

59

6

10.2

70

9

12.9

129

15

11.6

Pitymys subterraneus

0

-

-

1

0

0.0

1

0

0.0

SUBTOTAL

117

8

6.8

584

21

3.6

701

29

4.1

TOTAL

749

75

10.0

984

36

3.7

1733

111

6.4

3.2 Genetic analysis of complete S and M segment sequences from distinct Dobrava hantavirus subtypes of Apodemus agrarius and A. flavicollis

3.2.1 Comparison of S segment sequences

↓34

We have established the total nucleotide sequence of the S segment of the first A. flavicollis – derived DOBV strain from Central Europe, DOBV/Esl/400Af/98 (Esl/400Af). In contrast to the length of 1,704 nts as determined for A. agrarius - derived DOBV/Esl/856Aa/97 (Esl/856Aa) and DOBV/Esl/862Aa/97 (Esl/862Aa) (Sibold et al., 2001), the S segment of the Esl/400Af strain was found to encompass only 1,673 nt.

Table 7 (upper part) shows the nucleotide and amino acid percent identities between the N-encoding ORF sequences and deduced N proteins of different representatives of the DOBV species, respectively. Within the DOBV strains, Esl/856Aa and Esl/862Aa exhibited the lowest nucleotide sequence similarities (about 85 - 87 %) to the two DOBV-Af strains (Esl/400Af and DOBV/Slovenia-Af) as well as to the two DOBV/Saaremaa-Aa strains (Saa/160v and Saa/90Aa; Nemirov et al ., 1999). The East European DOBV/Kurkino-Aa strains (Kur/44Aa and Kur/53Aa; Plyusnin et al ., 1999b) are somewhat more related to Esl/856Aa and Esl/862Aa (about 90 % nt identity). On the amino acid level, Esl/856Aa and Esl/862Aa showed the lowest similarity to Saaremaa-Aa strains, being even lower than to the DOBV-Af strains. As expected, Esl/400Af exhibited high nt and aa similarities to DOBV/Slovenia-Af (Slo/Af) but lower similarity to the DOBV-Aa strains (Esl/856Aa, Esl/862Aa, Kur/44Aa, Kur/53Aa, Saa/160v, and Saa/90Aa).

The single ORF (nt 36 - 1,325) for the viral Nprotein encodes a putative protein of 429 aa length. In contrast to PUUV and TULV, no putative ORF2 in the S segment of the DOBV-Aa or DOBV-Af strains analysed was found. In general, the N-terminal half of the N protein (approximately 200 aa) is more conserved than the C-terminal half. Cysteine residues at aa positions 219, 244, 293, 315 and 319 are highly conserved; their functional importance is also suggested by conservation in all hantaviruses analysed here (aa 219, 244 and 315), in all Muridae-derived hantaviruses (aa 293), and in all Old World hantaviruses (aa 319).

↓35

The 3’ non-coding region (3’NCR) of the S segment of strain Esl/400Af is 348 nt long and appears highly similar to Slo/Af but most different from Esl/856Aa and Esl/862Aa (Table 7, lower part). The 3’NCR sequences of Esl/856Aa and Esl/862Aa showed identitiesof about 84-85% to the other DOBV-like strains (no matter whether of A. agrariusor A. flavicollis origin). Interestingly, the 3´NCR of Esl/856Aa and Esl/862Aa is longer than that of Esl/400Af; this is due to an 32 nt long insertion identified 28 bp downstream the stop codon in both Esl/856Aa and Esl/862Aa strains. The insert is (imperfectly) repetitive with its downstream neighbouring tract. Besides East Slovakian strains Esl/856Aa and Esl/862Aa, no other available DOBV S segment sequences contain this insertion in their 3´NCR.

Altogether, whereas S segments of the two DOBV-Af strains were found to be very similar, the DOBV-Aa strains exhibited higher diversity. The Central and East European strains were found to be rather dissimilarfrom the Saaremaa-Aa strains.

Table 7: S segment nucleotide and amino acid percent identity of DOBV strains, HTNV, and SEOV

ORF

Strain*

Esl/856Aa

Esl/862Aa

Kur/53Aa

Kur/44Aa

Saa/160V

Saa/90Aa

Esl/400Af

Slo/Af

HTNV

SEOV

Esl/856Aa

-

99.6

90.5

90.2

87.0

87.4

85.0

86.7

74.0

73.5

Esl/862Aa

99.5

-

90.4

90.2

86.9

87.3

85.0

86.7

74.0

73.6

Kur/53Aa

98.6

98.6

-

99.7

87.8

87.7

87.6

88.1

74.2

74.2

Kur/44Aa

98.1

98.1

99.5

-

87.5

87.4

87.4

87.8

74.0

74.1

Saa/160V

96.7

96.7

96.2

95.8

-

98.3

87.5

87.9

73.3

72.5

Saa/90Aa

96.5

96.5

96.0

95.5

99.7

-

87.2

88.0

73.1

72.8

Esl/400Af

97.6

97.6

97.2

96.7

97.2

96.9

-

95.3

73.8

72.5

Slo/Af

97.9

97.9

97.2

96.7

97.2

96.9

99.0

-

74.2

73.6

HTNV

82.9

82.9

83.4

83.2

82.5

82.2

82.7

82.7

-

73.9

SEOV

80.8

80.8

81.3

80.8

79.9

79.9

80.1

80.1

82.2

-

3´NCR

Strain*

Esl/856Aa

Esl/862Aa

Kur/53Aa

Kur/44Aa

Saa/160V

Saa/90Aa

Esl/400Af

Slo/Af

HTNV

SEOV

Esl/856Aa

-

100

85.3

85.3

83.6

84.1

84.4

83.6

51.3

40.7

Esl/862Aa

-

85.3

85.3

83.6

84.1

84.4

83.6

51.3

40.7

Kur/53Aa

-

100

89.0

90.0

91.8

90.3

53.7

40.6

Kur/44Aa

-

89.0

90.0

91.8

90.3

53.7

40.6

Saa/160V

-

98.7

89.6

89.9

53.6

40.6

Saa/90Aa

-

90.2

89.9

53.8

40.6

Esl/400Af

-

96.6

54.4

40.1

Slo/Af

-

54.7

39.7

HTNV

-

46.3

SEOV

-

The S segment ORF (upper part) and 3´NCR (lower part) were analysed. The identity values were calculated using Clustal method. The percentage differences for nucleotide (above the diagonal) and amino acid (below the diagonal) sequences are presented.
*For strain abbreviations, see Table 5.

3.2.2 Comparison of M segment sequences

↓36

Total M segment nucleotide sequences have been determined for the sympatrically occurring strains Esl/862Aa and Esl/400Af from East Slovakia. Their lengths were found to be 3,643 and 3,648 nt, respectively. Both strains encode a putative glycoprotein precursor (GPC) consisting of 1,135 aa.

Table 8(upper part) shows the degree of nucleotide and deduced amino acid sequence identity between the glycoprotein-encoding ORFs of these two strains and the two other DOBV strains for which complete M segment sequences are known, DOBV/Saaremaa/160v (Saa/160v) and DOBV/Slovenia-Af (Slo/Af). Strain Esl/400Af resembles the DOBV/Slovenia-Af prototype sequence (93.0% nt and 99.2 % aa identity), whereas greater differences were found between Esl/862Aa and Saa/160v (87.3% nt and 96.1% aa identity) and, even more pronounced, between Esl/862Aa and the DOBV-Af strains (only about 82.5 % nt and 94.1 % aa identity).

Similar relationships could be deduced from the comparative analysis of the 3´NCR of the M segment (Table 8, lower part). However, the M segment 3´NCR sequences exhibit a higher variability than the S segment 3´NCRs of the strains analysed.

↓37

In contrast to the S segment analysis, where the Central European DOBV-Aa strains exhibited a similar level of identity with Saaremaa-Aa strains and DOBV-Af strains each, the M segment of our Central European strain Esl/862Aa showed higher similarity to Saa/160v than to the DOBV-Af strains. In addition, the M segments of the DOBV-Af strains appeared more similar to each other than were the DOBV-Aa M segments to one another.

The putative aa sequences of the GPC of Esl/Aa862 and Esl/Af400 showed the typical sequence motifs of hantavirus glycoproteins. At the N-terminus of the GPC a signal peptide of 18 aa could be predicted. The GPC of both strains contain the highly conserved WAASA motif responsible for co-translational cleavage of G1 and G2. The number (n = 60; 5.3% of total aa) and localisation of cysteine residues are identical for all four compared DOBV-like sequences, 43 of them can be found in all available hantavirus GPC sequences. The YRTL motif, a potential internalisation signal for the endoplasmic reticulum (ER), was found in the cytoplasmic domain of G1 in all analysed DOBV sequences. KHKKS, a potential ER retention signal in the cytoplasmic domain of G2, is modified in all DOBV sequences to KHKRS.

Table 8: M segment nucleotide and amino acid percent identity of DOBV strains, HTNV, and SEOV

ORF

Strain*

Esl/862Aa

Saa/160V

Esl/400Af

Slo/Af

HTNV

SEOV

Esl/862Aa

-

87.3

82.6

82.5

70.5

70.8

Saa/160V

96.1

-

82.2

82.4

71.2

70.6

Esl/400Af

94.1

94.3

-

93.0

71.0

70.6

Slo/Af

94.1

94.5

99.2

-

70.8

70.7

HTN

77.0

76.9

77.7

77.5

-

72.3

SEO

76.8

76.4

77.3

77.5

76.8

-

3´NCR

Strain*

Esl/862Aa

Saa/160V

Esl/400Af

Slo/Af

HTNV

SEOV

Esl/862Aa

-

79.7

62.5

63.1

37.8

44.2

Saa/160V

-

62.0

63.6

41.3

44.2

Esl/400Af

-

91.9

40.5

43.2

Slo/Af

-

42.3

44.5

HTNV

-

39.1

SEOV

-

The M segment ORF (upper part) and 3´NCR (lower part) were analysed. The identity values were calculated using Clustal method. The percentage differences for nucleotide (above the diagonal) and amino acid (below the diagonal) sequences are presented.
*For strain abbreviations, see Table 5.

3.2.3 Phylogenetic trees and proof of reassortment

↓38

The molecular phylogeny of complete S segment nucleotide sequences from the DOBV strains by maximum likelihood (ML) analysis clearly indicated that Central European DOBV-Aa strains form a well-supported monophyletic group with the Russian Kurkino-Aa strains whereas the strain Esl/400Af forms a monophyletic group together with Slo/Af (Figure 8). Using the complete S segment sequence for analysis, the PUZZLE support for position of the Saaremaa-Aa strains within the tree was slightly below the 70% threshold limit. Therefore we repeated the analysis on the basis of ORF nucleotide sequences (Figure 8, upper right) and aa sequences of the N proteins (data not shown) of the same virus strains. These analyses unambiguously grouped with high support Saaremaa-Aa sequences together with the DOBV-Af strains and well-apart from the Central European and Russian DOBV-Aa lineages. This shows that the S segment of Saaremaa-Aa strains is related to the S segment of A. flavicollis -derived DOBV strains. This result is supported also by presence of five characteristic amino acid exchanges in the N protein where both Saaremaa-Aa strains exhibit the same amino acid as Esl/400Af and/or Slo/Af in contrast to Central European and Russian strains (data not shown). In the opposite direction, only at three positions Saaremaa-Aa strains exhibit aa residues different from those of DOBV-Af strains and identical to those of East and Central European DOBV-Aa strains (see following chapter).

Figure 9 shows the phylogenetic tree based on available complete M segment sequences of DOBV strains and other hantaviruses. In addition to our two newly determined complete sequences Esl/862Aa and Esl/400Af only two other complete M segment sequences were available for analysis. However, the results clearly show that the Esl/862Aa strain forms one well-supported group with Saa/160v whereas the strain Esl/400Af demonstrated, as expected, a close relationship to the Slo/Af prototype. This is also supported by presence of 38 aa exchanges in the GPC where Esl/862Aa and Saa/160v carry identical amino acid residues in contrast to both DOBV-Af strains.

The difference in the phylogenetic placement of Saaremaa-Aa in respect to S and M segments shows that this virus was involved in reassortment processes during its evolution; its S segment is more related to DOBV-Af whereas its M segment resembles the Central European strain Esl/862Aa.

↓39

To investigate the possibility that this contradiction of S- and M-segment derived trees could have been caused simply by absence of additional sequences (e.g., Kurkino-Aa strains) in the M but not S segment ML analysis, we repeated the S segment analysis without involving sequences of Kurkino-Aa strains. Nevertheless, also under these conditions the phylogenetic grouping of the S segments of Saaremaa-Aa strains with DOBV-Af remained unchanged (data not shown) confirming the above mentioned results.

Figure 8: Phylogenetic ML trees from the complete S segment nucleotide sequences of DOBV and further hantaviruses, computed with TREE-PUZZLE.

The main tree is based on sequences of the complete S segments. The insert shows the subtree of the complete ORF sequences for the DOBV-like strains. The sources of hantavirus sequences are given in Table 5.

Figure 9: Phylogenetic ML tree of the available complete M segment nucleotide sequences of different members of the DOBV lineage and further hantaviruses computed with TREE-PUZZLE.

The sources of hantavirus sequences are given in Table 5.

3.2.4 Recombination analysis

↓40

To study more precisely the evolutionary history of the two genomic segments we performed bootscanning analyses on the alignments of the S and M segment ORFs for the viruses Esl/400Af, Esl/856Aa, and Esl/862Aa. In all initial bootscanning analyses (data not shown) as well as all phylogeny reconstructions, Esl/400Af formed a highly supported subtree together with Slo/Af, as did Esl/856Aa with Esl/862Aa. So we assigned the former as close relatives to one group ”DOBV-Af” and the latter to a group here called ”Esl/8xxAa”.

Subsequent bootscanning analyses of the Esl/8xxAa group (S segment, not shown) or Esl/862Aa (M segment, Figure 10) as query, showed the following principal results. While Esl/862Aa revealed a high relationship to Saa/160v in most parts of the M segment ORF, the relationship of Esl/8xxAa to Kurkino-Aa strains throughout the S segment ORF gained only low support values (Figure 10). To elucidate the specific evolutionary histories of different regions in the S and M segments of the Esl/8xxAa viruses we split the full alignments into subalignments guided by the main cross points of a 65% threshold in the bootscan diagrams. For the M segment analysis these split points are marked above the bootscan diagram in Figure 10. The subalignments were then used to reconstruct ML phylogenies using quartet puzzling. The resulting DOBV subtrees were then further examined. The M segment subtrees are given in the upper part of Figure 10.

The bootscan analysis of Esl/8xxAa in respect to the S segment ORFs gave support for the Esl/8xxAa – Kurkino-Aa monophyly in most parts of the sequence although with very low support values. The values almost never reached 90% and were mostly below 60% (data not shown). Changes of the evolutionary history were suggested by substantial drops of the curve associated with Kurkino-Aa for the regions from nt 400-600 and nt 730-1080. The ML phylogenies reconstructed from the subalignments of the S segment ORFs revealed changes in the evolutionary history along the sequences. While the trees grouped together Esl/8xxAa and Kurkino-Aa against DOBV-Af and Saaremaa-Aa in most areas of the alignment, the tree topology could not be resolved in the first 225 nt of the ORF, where the Esl/8xxAa – Kurkino-Aa subtree forms a multifurcation with DOBV-Af and Saaremaa-Aa. In the area from nt 400 - 600 a change in the topology was found where DOBV-Af branched off at the root of the Esl/8xxAa – Kurkino-Aa subtree. However, one should note that the support (64 %) for this was not very high. In contrast to the loss of support for the Esl/8xxAa – Kurkino-Aa monophyly in the bootscan in the area from 730-1080 nt, the phylogenetic tree maintains this monophyly with high support of 82%.

↓41

This discrepancy between bootscanning and TREE-PUZZLE as well as the low overall support for the Esl/8xxAa – Kurkino-Aa monophyly in the bootscanning, can be explained as follows. While the ML quartet puzzling analysis and the resulting PUZZLE support values made use of the full data set, the bootscanning is based on bootstrapping from windows of a certain size, here 200 nt. Bootstrapping draws pseudosamples with replacement from the alignment columns by also discarding alignment columns. The low divergence of the S segments of 29.9 - 51.7 % constant sites in the subalignments (for comparison, M segments have 18.4 - 41.4 % constant sites) can lead to a loss of information in the bootstrapping which results in less resolved trees and subsequently in low values in the bootscanning. For this reason the S segment ORFs are less suitable for bootscanning analysis. However, it is still a good tool to find split points of the alignment for further phylogenetic analysis.

In the analysis of the M segment the resulting picture gave more clear results. First, the support values in the bootscanning reached up to 100% (cf. above and Figure 10). Furthermore the main DOBV topology divided into Esl/862Aa with Saa/160v on the one hand and Slo/Af with Esl/400Af on the other yielding high support (92-100%). In two areas of the ORF this topology changed bringing together Saa/160v and the DOBV-Af group. While this change in grouping of M segment regions had moderate support (62%) in the area 1971-2211 nt, the support in the area 810-1059 nt with 81% was very high.

Homologous recombination events could explain the creation of this significant nucleotide sequence exchange. However, this issue needs to be further evaluated, taking into account biological properties of M segment gene products which are currently not very well characterised.

↓42

Figure 10: Recombination analysis for the M segment ORFs.

The lower part shows the bootscan diagram using Esl/862Aa as query, and the other viruses grouped into DOBV-Af, Saaremaa-Aa (Saa/160v), and HTNV. The upper part shows the ML DOBV subtrees using TREE-PUZZLE from the subalignments split at a 65% threshold in the bootscan diagram. At the branches the PUZZLE support values for the subtrees are given. Where neighbouring subalignment resulted in the same tree topology only one with ranges of support values is shown.

3.2.5 Identification of putative host-specific differences in the virus-coded N and GPC

When directly comparing the N protein sequences of Esl/856Aa and Esl/862Aa on one hand and Esl/400Af on the other we have found a total of 9 aa exchanges, 8 conservative and 1 non-conservative ones according to Dayhoff exchange groups (Dayhoff et al ., 1978). Including the Russian Kurkino-Aa strains and Slo/Af in this comparison reduced the putatively host-specific genetic differences to 7 aa exchanges (Figure 11A). Furthermore, when including Saaremaa-Aa in the DOBV-Aa group, only 3 aa differences (out of 7) have been found between the DOBV-Aa strains on the one hand and the DOBV-Af strains on the other (S13N, I295L, and R356K) which may represent host-specific residues (Figure 11A). According to Dayhoff exchange groups, all three aa exchanges represent conservative exchanges. The Saaremaa-Aa S segment has been included into this analysis despite our finding that it is phylogenetically related to the DOBV-Af S segment (see above). Nevertheless, the virus underwent its (at least rather recent) evolution in A. agrarius and should have accumulated mutations which did allow replication in this host.

Position 295 is located in the highly variable part of N protein (Plyusnin et al., 1996) whereas the other two exchanges (pos. 13, 356) had occurred in rather conserved regions of N (Arikawa et al., 1990). Position 13 lies in the major immunodominant region of N proteins from other hantavirus species which includes B- and T-cell epitopes (de Carvalho et al., 2001; Gott et al., 1997; Jenison et al., 1994; Lundkvist et al., 2002b; Van Epps et al., 1999).

↓43

The comparison with N protein aa sequences of other hantavirus species at position 13 showed amino acids always different from DOBV-Aa and DOBV-Af strains. At the other positions ( aa 295 and 356) the same aa residues as in one of the DOBV lineages could be found. ( data not shown). At aa 356 of all hantaviral N proteins the basic aa residues R or K have been found to be conserved which might be caused by the nucleic acid binding function of the C-terminal region previously shown for PUUV-N protein (Gott et al ., 1993)

In a direct comparison of glycoprotein (G1/G2) aa sequences of strains Esl/862Aa and Esl/400Af we found 66 aa differences. When including the only available complete M sequences of Saa/160v and Slo/Af in the analysis, 38 aa exchanges (27 conservative and 11 non-conservative ones) remained which exist between the two DOBV-Aa strains on the one side and the two DOBV-Af strains on the other (Figure 11B). Moreover, when including GPC aa sequences of other hantaviruses, four positions (at aa 14, 15, 230, 335) were found, where every hantavirus species has a specific aa residue. The host specificity of aa 14 and 15 may be due to their functional role at the C-terminal end of the signal peptide of the GPC.

Sequence determination of more DOBV strains from different regions of Europe should enable a further specification of host-dependent genetic differences within the virus species.

↓44

Figure 11: Putative host specific amino acid exchanges in the N protein (Figure 11A) and the glycoproteins (Figure 11B) of DOBV strains from A. flavicollis vs. A. agrarius hosts.

Upper parts show aa residues at sites where DOBV-Aa and DOBV-Af differ. Gray background indicates DOBV-Af-like residues.
Lower parts: Triangles indicate putative host specific aa exchanges. Black triangles indicate non-conservative, gray triangles conservative aa exchanges (according to Dayhoff exchange groups).
Figure 11A: Large triangles represent aa exchanges where Saaremaa-Aa strains exhibit the same amino acid as other DOBV-Aa strains. Small triangles symbolise exchanges, where Saaremaa strains resemble the DOBV-Af strains. Typical hantavirus N protein regions (Arikawa et al., 1990; Plyusnin et al., 1996) are marked. Figure 11B: The positions of the G1 and G2 glycoproteins as well as other functionally important domains (Plyusnin et al., 1996) are shown.

3.3 Genetic diversity of DOBV on single geographical locus

3.3.1 DOBV in Rozhanovce locality, Eastern Slovakia

As a part of long-term epizootiologic survey, small rodents were trapped on Rozhanovce locality in Eastern Slovakia during three trapping nights in September and October of 2001. This locality was selected on the basis of reported human cases of HFRS. The trapping area was situated in a pheasants breeding station.

57 rodents were trapped. Identified rodent species included A. agrarius (n=42), A. flavicollis (n=9) and C. glareolus (n=6). Antibodies reactive with DOBV antigen were found only in A. agrarius rodents, however, the antibody prevalence was very high. 11 out of 42 (26.2%) mice were found sero- and RT PCR-positive. Six of them, designed Esl/33Aa, Esl/193Aa, Esl/194Aa, Esl/197Aa, Esl/200Aa, and Esl/201Aa, were selected for cloning and sequencing of partial S (position 357-955 nt, 559 nt in length when excluding the primer sequences) and M (1674-1990 nt, 271 nt when excluding the primer sequences) segments.

↓45

In addition, we have determined the complete S segment nucleotide sequence of strain Esl/33Aa from Rozhanovce and of strains Esl/29Aa and Esl/81Aa detected in neighbouring localities Sebastovce and Botany, respectively. Esl/29Aa and Esl/81Aa S segments were found to be 7 nt shorter (1697 nt) than sequence of Esl/862Aa, due to an 8 nt deletion and an 1 nt insertion located in 3’ NCR. Esl/33Aa harboured one additional nt deletion, thus reaching 1696 nt in length.

Sequence comparison of these six DOBV sequences originating from Rozhanovce with other DOBV sequences showed interesting results. From S segment sequence identity matrix (Table 9) is obvious that these six sequences originating from one trapping site represent two distinct groups of sequences (Esl/33Aa, Esl/193Aa, Esl/200Aa and Esl/201Aa on one hand, Esl/194Aa and Esl/197Aa on the other hand) exhibiting relatively high sequence diversity of 5.4 – 5.7% while the sequence identity within these groups is reaching 98.6 – 100%.

Similar relationships could be deduced from the comparative analysis of the partial M segment sequences (Table 10). However, the overall diversity of studied sequences is lower, most likely because of short sequenced fragment. Interestingly, strains Esl/194Aa and Esl/197Aa exhibited higher similarity to previously identified DOBV strains from Eastern Slovakia, Esl/856Aa and Esl/862Aa, than to the other Rozhanovce derived strains. However, both groups represent unique sequences. Hence, the possibility that only one group represents real DOBV sequences from Rozhanovce and the other one is only a laboratory contamination by some other DOBV strains previously identified in our laboratory could be excluded.

↓46

Maximum Likelihood phylogenetic analysis of S segment partial sequences (Figure 12) confirmed these observations. The sequences originating from rodents captured on a single trapping place do not form a monophyletic group. Whereas the strains Esl/33Aa, Esl/193Aa, Esl/200Aa and Esl/201Aa form one group and cluster with strains from neighbouring trapping localities, the strains Esl/194Aa and Esl/197Aa form well supported group with strains Esl/856Aa and Esl/862Aa originating also from Eastern Slovakia but from another locality (Novy Ruskov). We could see the same branching topology also on phylogenetic tree based on partial M segment sequences (data not shown).

These interesting observations do not agree with the geographical clustering, which is usually observed in hantaviruses of the same genotype and originating from the same rodent host. An explanation, which is in agreement with the virus-host coevolution concept (Plyusnin and Morzunov, 2001), consists in the simultaneous presence of two phylogenetically distinct subpopulations of A. agrarius in Rozhanovce locality, carrying distinct DOBV strains.

Table 9: S segment nucleotide percent identity of Rozhanovce-originating and other DOBV strains

Strain*

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

1.

Esl/33Aa

-

99.2

98.9

99.2

99.1

98.7

94.4

94.4

94.2

94.2

90.8

90.6

85.1

85.1

84.2

84.6

2.

Esl/193Aa

-

99.6

100

99.4

99.1

94.2

94.2

93.7

93.7

90.5

90.3

85.1

85.1

83.8

84.6

3.

Esl/200Aa

-

99.6

99.1

98.7

93.9

93.9

93.3

93.3

90.1

89.9

84.7

84.7

83.8

84.6

4.

Esl/201Aa

-

99.4

99.1

94.2

94.2

93.7

93.7

90.5

90.3

85.1

85.1

83.8

84.6

5.

Esl/29Aa

-

99.2

93.9

93.9

93.3

93.3

90.5

90.3

85.1

85.1

84.2

84.9

6.

Esl/81Aa

-

93.5

93.5

93.0

93.0

90.6

90.5

85.3

85.3

84.4

85.1

7.

Esl/194Aa

-

100

99.1

99.1

89.9

89.8

86.7

87.1

85.3

85.8

8.

Esl/197Aa

-

99.1

99.1

89.9

89.8

86.7

87.1

85.3

85.8

9.

Esl/856Aa

-

100

89.8

89.6

86.5

87.2

84.4

85.6

10.

Esl/862Aa

-

89.8

89.6

86.5

87.2

84.4

85.6

11.

Kur/53Aa

-

99.8

87.4

87.1

87.2

87.2

12.

Kur/44Aa

-

87.2

86.9

87.1

87.1

13.

Saa/160V

-

98.2

87.4

88.0

14.

Saa/90Aa

-

86.9

88.0

15.

Esl/400Af

-

96.2

16.

Slo/Af

-

An S segment partial sequences (377-935 nt) were analysed. The identity values were calculated using Clustal method. The percentage differences for nucleotide sequences are presented. *For strain abbreviations, see Table 5.

↓47

Table 10: M segment nucleotide percent identity of Rozhanovce-associated and other DOBV strains

Strain*

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

1.

Esl/33Aa

-

99.6

99.6

99.6

98.1

98.1

98.5

88.1

83.0

81.9

2.

Esl/193Aa

-

100

100

98.5

98.5

98.8

87.8

83.3

81.5

3.

Esl/200Aa

-

100

98.5

98.5

98.8

87.8

83.3

81.5

4.

Esl/201Aa

-

98.5

98.5

98.8

87.8

83.3

81.5

5.

Esl/194Aa

-

100

99.6

88.5

82.6

80.8

6.

Esl/197Aa

-

99.6

88.5

82.6

80.8

7.

Esl/862Aa

-

88.5

82.6

80.8

8.

Saa/160v

-

83.3

83.0

9.

Esl/400Af

-

94.0

10.

Slo/Af

-

An M segment partial sequence (1700-1970 nt) was analysed. The identity values were calculated using Clustal method. The percentage differences for nucleotide sequences are presented. *For strain abbreviations, see Table 5.

Figure 12: Phylogenetic ML tree based on the partial S segment nt sequences (377-935 nt) of Rozhanovce-originating and other DOBV strains, computed with TREE-PUZZLE.

Strains with identical sequence are represented by single branch. HTNV strain 76-118 was used as an outgroup. For graphical presentation, to obtain a better resolution of the picture in the analysed Rozhanovce clades, the length of the longest branch indicated by two vertical lines (HTNV as an outgroup) was shortened by 0.1 of evolutionary distance. For strain name abbreviations and sequence references see Table 5.

3.3.2 Rodent genetics

In order to find out whether there are two distinct subpopulations of A. agrarius in Rozhanovce locality, we carried out a sequence analysis of rodent genetic material using two phylogenetic mitochondrial markers, 12S rRNA gene (in collaboration with Dr. Plötner, Museum of Natural History, Berlin, Germany) and control region of mitochondrial DNA (D-loop). For DNA extraction and PCR amplification, we used material from four DOBV-positive individuals from Rozhanovce locality (Apa194/01 and Apa197/01, which carry DOBV Rozhanovce group II; Apa200/01 and Apa201/01, which carry DOBV Rozhanovce group I), and individuals of A. agrarius and A. flavicollis from other three selected localities in Slovakia (Table 11, Figure 13).

↓48

We have established the nucleotide sequence of 12S rRNA genetic marker (372 nt) of 10 A. agrarius and six A. flavicollis mice captured in Slovakia. The sequence identity matrix and multiple sequence alignment of polymorphic sites are shown in Table 12andTable 13, respectively. There was a significant diversity of 7.2 –7.8% between the A. agrarius and A. flavicollis sequences. However, the variability within the species is so low that there was not enough phylogenetic information to construct a reasonable phylogenetic tree.

In a second attempt to study the evolutionary history of A. agrarius population from Slovakia, we have sequenced a part of mitochondrial DNA (1167 nt) including D-loop of all A. agrarius animals mentioned above. Despite relatively long sequences studied, we could find only few variable sites (n=13) in the sequence alignment (Table 14) and the sequence identities were reaching 99.3 – 100% (Table 15).

Altogether, the analysis of both 12S rRNA and D-loop genetic markers revealed lack of sequence variability within the A. agrarius mice captured in Eastern Slovakia. This suggests that according to these markers all these individuals represent a single population. Moreover, the analysis of 12S rRNA gene showed that this marker is suitable for distinguishing at least A. agrarius and A. flavicollis species and also confirmed that the individuals used in this study were previously well identified according to the morphologic criteria used.

↓49

Table 11: Apodemusmice used for amplification and sequencing of mitochondrial markers 12S rRNA gene and D-loop

Sample

Locality

Geographic location

Apodemus agrarius

1.

Apa194/01

Rozhanovce

Eastern Slovakia

2.

Apa197/01

Rozhanovce

Eastern Slovakia

3.

Apa200/01

Rozhanovce

Eastern Slovakia

4.

Apa201/01

Rozhanovce

Eastern Slovakia

5.

Apa24/01

Sebastovce

Eastern Slovakia

6.

Apa25/01

Sebastovce

Eastern Slovakia

7.

Apa95/99

Krasny Brod

Eastern Slovakia

8.

Apa99/99

Krasny Brod

Eastern Slovakia

9.

Apa64/01

Botany

Eastern Slovakia

10.

Apa65/01

Botany

Eastern Slovakia

A. flavicollis

11.

Apf95/01

Zahorska Ves

Western Slovakia

12.

Apf96/01

Zahorska Ves

Western Slovakia

13.

Apf46/01

Sebastovce

Eastern Slovakia

14.

Apf47/01

Sebastovce

Eastern Slovakia

15.

Apf110/99

Krasny Brod

Eastern Slovakia

16.

Apf111/99

Krasny Brod

Eastern Slovakia

The sample names consist of abbreviation of rodent species (Apa stands for A. agrarius, Apf for A. flavicollis) / year of trapping. For better geographic location of trapping sites, see Figure 13.

Figure 13: Map of Slovakia with trapping localities for virus and rodent genetics studies.

Table 12: 12s rRNA gene nucleotide sequence percent identity of Apodemus mice trapped on selected localities of Slovakia

Sample

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

1.

Apa194/01

-

99.7

100

100

100

100

100

99.7

100

100

92.7

92.7

92.7

92.7

92.7

92.4

2.

Apa197/01

-

99.7

99.7

99.7

99.7

99.7

99.4

99.7

99.7

92.4

92.4

92.4

92.4

92.4

92.2

3.

Apa200/01

-

100

100

100

100

99.7

100

100

92.7

92.7

92.7

92.7

92.7

92.4

4.

Apa201/01

-

100

100

100

99.7

100

100

92.7

92.7

92.7

92.7

92.7

92.4

5.

Apa24/01

-

100

100

99.7

100

100

92.7

92.7

92.7

92.7

92.7

92.4

6.

Apa25/01

-

100

99.7

100

100

92.7

92.7

92.7

92.7

92.7

92.4

7.

Apa95/99

-

99.7

100

100

92.7

92.7

92.7

92.7

92.7

92.4

8.

Apa99/99

-

99.7

99.7

92.4

92.4

92.4

92.4

92.4

92.2

9.

Apa64/01

-

100

92.7

92.7

92.7

92.7

92.7

92.4

10.

Apa65/01

-

92.7

92.7

92.7

92.7

92.7

92.4

11.

Apf95/01

-

100

99.4

100

100

99.7

12.

Apf96/01

-

99.4

100

100

99.7

13.

Apf46/01

-

99.4

99.4

99.7

14.

Apf47/01

-

100

99.7

15.

Apf110/99

-

99.7

16.

Apf111/99

-

12S rRNA mitochondrial gene sequences (372 nt) were analysed. The identity values were calculated using Clustal method. The percentage differences for nucleotide sequences are presented.

↓50

Table 13: Multiple alignment of a partial 12S rRNA gene sequence (372 nt) from Apodemus sp. mice captured in Slovakia; only polymorphic sites are shown.

polymorphic sites

 

7

8

11

18

43

44

83

111

118

139

142

143

187

188

204

205

206

211

269

270

271

281

282

287

288

301

309

337

349

356

Apa194/01

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa197/01

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

G

T

T

A

G

T

Apa200/01

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa201/01

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa24/01

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa25/01

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa64/01

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa65/01

C

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa95/99

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apa99/99

T

G

C

T

T

A

C

A

C

T

-

T

T

A

T

T

T

A

A

T

T

T

T

A

A

T

T

A

G

T

Apf95/01

T

A

T

C

C

T

T

C

A

C

A

T

A

T

C

A

-

T

T

A

C

C

-

T

A

C

C

G

A

A

Apf96/01

T

A

T

C

C

T

T

C

A

C

A

T

A

T

C

A

-

T

T

A

C

C

-

T

A

C

C

G

A

A

Apf46/01

T

A

T

C

C

T

T

C

A

C

A

C

A

T

C

A

-

T

T

A

C

C

-

T

A

C

C

G

G

A

Apf47/99

T

A

T

C

C

T

T

C

A

C

A

T

A

T

C

A

-

T

T

A

C

C

-

T

A

C

C

G

A

A

Apf110/99

T

A

T

C

C

T

T

C

A

C

A

T

A

T

C

A

-

T

T

A

C

C

-

T

A

C

C

G

A

A

Apf111/99

T

A

T

C

C

T

T

C

A

C

A

C

A

T

C

A

-

T

T

A

C

C

-

T

A

C

C

G

A

A

Table 14: Multiple alignment of mitochondrial DNA (1167 nt) including D-loop of A. agrariusmice captured in Slovakia; only polymorphic sites are shown.

polymorphic sites

 

22

214

240

271

326

719

776

781

788

789

816

841

1085

 

Apa194/01

A

-

G

C

C

G

T

G

T

T

A

A

A

 

Apa197/01

A

T

A

T

T

G

T

A

A

C

A

A

G

 

Apa200/01

-

-

A

C

C

T

C

A

A

C

A

T

A

 

Apa201/01

-

-

A

C

C

G

T

A

A

C

A

T

A

 

Apa24/01

A

-

A

C

C

G

T

A

A

C

A

A

A

 

Apa25/01

A

-

A

C

C

G

T

A

G

C

A

A

A

 

Apa95/99

A

-

A

C

C

G

T

G

A

C

G

A

A

 

Apa99/99

A

-

A

C

C

G

T

G

A

C

G

A

A

 

Apa64/01

A

-

A

C

C

G

T

A

A

C

A

A

A

 

Apa65/01

A

-

A

C

C

G

T

A

A

C

A

A

A

 

Table 15: D-loop nucleotide sequence percent identity of A. agrariusmice trapped on selected localities of Slovakia

Sample

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

1.

Apa194/01

-

99.3

99.3

99.4

99.6

99.6

99.6

99.6

99.6

99.6

2.

Apa197/01

-

99.3

99.4

99.6

99.5

99.4

99.4

99.6

99.6

3.

Apa200/01

-

99.8

99.6

99.5

99.4

99.4

99.6

99.6

4.

Apa201/01

-

99.8

99.7

99.6

99.6

99.8

99.8

5.

Apa24/01

-

99.9

99.8

99.8

100

100

6.

Apa25/01

-

99.7

99.7

99.9

99.9

7.

Apa95/99

-

100

99.8

99.8

8.

Apa99/99

-

99.8

99.8

9.

Apa64/01

-

100

10.

Apa65/01

-

3.4 Isolation of DOBV from A. agrarius captured in East Slovakia

3.4.1 Virus isolation and titration

↓51

In order to obtain a cell culture isolate of DOBV representing the Central European DOBV-Aa genetic lineage, tissues of sero- and RT-PCR-positive A. agrarius were used for virus isolation attempts. 10% suspensions of homogenised tissues (liver of Esl/34Aa , lung of Esl/35Aa) were inoculated onto confluent Vero E6 cells in three 25 cm2 flasks each. Cells were passaged at days 14, 39 and 47 post infection (p.i.). After the third passage at day 47 p.i., in the flasks of both samples some hantavirus antigen-positive cells were found by IFA using a DOBV-specific human convalescent serum. After the fourth passage (day 59), most cells were found to be positive. One isolate, originating from the field strain DOBV/East Slovakia/34AaV/2001 was then further passaged and designated Slovakia (short name SK/Aa, where SK stands for Slovakia and Aa indicates for better clarity the natural host).

Cell culture supernatants harvested from 3rd –5th passages were titrated using a focus assay. Whereas the titers of third and fourth passage supernatants were extremely low (< 103 focus forming units, FFU/ml), the fifth passage supernatant had markedly higher titer (4.5 x 105 FFU/ml). A cell culture supernatant harvested from fifth passage (day 51) was therefore regarded as a first virus stock (stock A).

Since hantaviruses do not naturally form plaques in Vero E6 cell culture allowing plaque purification of the virus we had to follow an alternative approach based on limiting dilution to obtain a homogenous virus stock., Ten wells of six-well plates were inoculated with the stock A diluted to 0.5 FFU/well. After ten days incubation, supernatant samples were collected and the cells were examined using the chemiluminiscent focus assay. Four out of ten wells were found to be antigen-positive. The supernatant from one of the positive wells was then used to infect new Vero E6 cells and produce new viral stock (B) for further analyses. Assuming that the positive wells were infected with a single virus particle, we obtained a homogenous virus stock equivalent to a plaque-purified virus. The titre of stock B was determined to be 5.0 x 105 FFU/ml.

3.4.2 Sequence analysis of SK/Aa genomic segments

↓52

The complete S segment sequence of the new isolate SK/Aa was determined to be 1697 nt in length containing a single ORF (nt 36 - 1,325) for the viral Nprotein which encodes a putative protein of 429 aa length. In addition, the complete S segment nucleotide sequences of wild strains Esl/29Aa and Esl/81Aa were obtained from the lung tissues of A. agrarius animals and were determined to be of the same length and coding structure.

As expected, the sequences of Esl/29Aa and Esl/81Aa were found to be very similar to the new isolate (98.9 - 99.0% nt and 99.7% aa identities). When compared with other DOBV sequences, previously determined DOBV-Aa sequences from East Slovakia (Esl/856Aa and Esl/862Aa) showed the highest similarity (93.3 nt and 99.3% aa identity). However, the S segment of the new isolate was found to be 7 nt shorter than the sequences of Esl/856Aa and Esl/862Aa strains, due to an 8 nt deletion and an 1 nt insertion located in the 3’ NCR. The most dissimilar DOBV strain on nt level, Esl/400Af (84.4% nt identity), is from the same geographical region of Slovakia but belongs to the DOBV-Af lineage. On the deduced N protein aa sequence level, the Saa/160V isolate showed the highest diversity to SK/Aa, even higher than the virus isolates of DOBV-Af lineage, Slo/Af and AP/Af (Table 16, upper part). Five cysteine residues were found in the N protein aa sequence (at aa positions 203, 244, 293, 315 and 319), all of them are highly conserved and were found in all N protein sequences analysed here (Table 16).

The total M segment nucleotide sequence of SK/Aa was found to be 3,643 nt in length. It contains a single ORF (nt 41 to 3445) encoding the putative GPC of 1,135 aa in length. Table 16 (lower part) shows the degree of M segment nucleotide and deduced amino acid sequence identity between the DOBV isolates. The DOBV-Af isolate Slo/Af was found to be the most dissimilar (only 81.5% nt and 93.4% aa identity) what could explain the distinct serological behaviour of SK/Aa and Slo/Af in cFRNT (see below). The GPC aa sequence of the of SK/Aa showed the typical sequence motifs of hantavirus glycoproteins described above for Esl/862Aa and Esl/400Af.

↓53

In addition, a partial L segment sequence of 541 nt length (nt position 109-649 according to the co-ordinates in the L segment of DOBV AP/Af, GenBank accession number AJ410617) from the SK/Aa isolate was determined. The comparison with the only available complete DOBV L segment sequences showed relatively low nt sequence identities of 86.3 and 85.0% to the Saa/160V and AP/Af strains, respectively. In contrast, percent-identity values of the corresponding amino acids (180 aa, aa positions 25-204) were very high, reaching 97.7 % towards both Saa/160V and AP/Af strains. This shows that most of the nucleotide exchanges represented silent mutations.

It should be mentioned that the S, M, and L segment sequences determined for SK/Aa did not encompass any known sequence from any virus strain handled in our laboratory. This clearly shows that no contamination problem had occurred.

3.4.3 Phylogenetic analysis

To avoid the phylogenetic analysis being disturbed by conflicting phylogenetic signal due to potential recombination events, both SK/Aa complete S and M segment sequences were first screened by similarity plots and bootscanning. No significant recombination signals, which could disturb construction of the phylogenetic trees, were found (data not shown).

↓54

Complete coding sequences, allowing unambiguous aligning on amino acid level, were used to construct Maximum likelihood trees assuming Tamura-Nei evolutionary model with gamma distributed rate heterogeneity over sites. In the S segment ML phylogenetic tree (Figure 14a), the new isolate formed a well-supported monophyletic group with all DOBV-Aa strains except Saaremaa strains which clustered together with the DOBV-Af strains (see below). As expected, all other Slovakian strains were the nearest relatives of the SK/Aa. The new strains SK/Aa, Esl/29Aa and Esl/81Aa formed a sister group to the previously characterised strains Esl/856Aa and Esl/862Aa. A. flavicollis-derived strains from Slovakia (Esl/400Af), Greece (AP/Af) and Slovenia (Slo/Af) formed the second cluster sharing the common ancestor with the Saaremaa strains.

In the M segment ML tree (Figure 14b), SK/Aa again clustered with the Esl/862Aa strain. However, in contrast to the S segment analysis, SK/Aa shared a common ancestor also with the Saa/160Aa strain from Estonia. The strains Esl/400Af, AP/Af, and Slo/Af formed again a strongly supported monophyletic group representing the DOBV-Af lineage.

Phylogenetic analysis of partial L segment sequences (541 nt) showed results consistent with the S segment analysis. AP/Af strain clustered with the Saa/160V strain while SK/Aa formed an outgroup to this well-supported cluster (Figure 14a). When using only 374 nt long sequences (nt position 157-530), the available partial sequence of the original DOBV isolate, Slo/Af, could be included in the alignment. Also in this case, SK/Aa was the most ancestral sequence and Saa/160V formed a well supported monophyletic group with the AP/Af and Slo/Af strains (data not shown).

↓55

To enable a direct comparison of the phylogenetic trees based on S, M, and L segment sequences, the S and M segment phylogenetic analysis was repeated only with sequences of those strains which were also used for construction of the L segment phylogenetic tree. Nevertheless, clustering of the Saa/160V strain with the DOBV-Af strains in S and L segment phylogeny (Figure 15a, b) but with DOBV-Aa strains in M segment phylogeny (Figure 15c) remained unchanged.

Table 16: Complete S and M segment nucleotide and amino acid sequence identities of SK/Aa with other DOBV, HTNV and SEOV isolates.

strains

% identity with virus strain:

SK/Aa

Saa/160V

Slo/Af

AP/Af

HTNV 76-11

SEOV SR11

S segment / N protein

SK/Aa

-

86.2

85.7

85.2

70.8

68.1

Saa/160V

96.9

-

88.3

87.3

69.7

66.2

Slo/Af

97.6

97.4

-

96.6

70.7

67.4

AP/Af

98.1

97.4

99.5

-

70.5

67.6

HTNV 76-11

83.6

82.5

82.5

82.9

-

67.5

SEOV SR11

81.1

79.9

79.9

80.4

82.2

-

M segment / GPC

SK/Aa

-

86.8

81.5

81.5

69.0

69.6

Saa/160V

95.7

-

81.3

81.5

69.6

69.5

Slo/Af

93.4

94.1

-

93.4

69.1

69.4

AP/Af

93.5

94.1

98.5

-

69.7

69.7

HTNV 76-11

77.0

76.9

77.1

77.3

-

71.3

SEOV SR11

76.7

76.4

77.2

77.1

76.8

-

The percentage differences for nucleotide (above the diagonal) and amino acid (below the diagonal) sequences are presented.

Figure 14: Maximum likelihood phylogenetic trees of DOBV strains based on (a) complete S and (b) complete M segment ORF nucleotide sequences (corresponding to S segment nt sequence position 36 - 1,325 and M segment nt position 41 to 3,445 of SK/Aa, respectively).

DOBV isolates are in bold. Three proposed DOBV lineages are marked by grey boxes. For abbreviations and accession numbers, see the Material and methods section. The trees were computed with the TREE-PUZZLE. The values at the tree branches are the PUZZLE support values.

↓56

Figure 15: Maximum likelihood phylogenetic trees of DOBV virus isolates based on (a) partial (541 nt, position 109 - 649) L segment sequences, (b) complete S ORF, and (c) complete M ORF nucleotide sequences (corresponding to S segment nt position 36 - 1,325 and M segment nt position 41 to 3,445, respectively).

The trees were computed with the TREE-PUZZLE package. The values at the tree branches are the PUZZLE support values.

3.4.4  In vitro evolution of virus during passaging

Isolation of hantaviruses on Vero E6 cells is a tedious process including blind passages during several weeks. It is believed that during this time, the virus has to undergo adaptation to cell culture resulting in changes of nucleotide and amino acid sequences. We were interested to determine whether the nucleotide sequence of the S segment underwent changes during the virus isolation procedure in cell culture. For this purpose we compared the sequences of the original hantaviral RNA in the rodent tissue (Esl/34Aa) and of the cell-culture adapted isolate SK/Aa. In the complete S segment we found one single nt exchange G (rodent tissue) → A (isolate) at position 162 resulting in an amino acid exchange Ala (GCA) → Thr (ACA) at aa position 43 of viral N protein. Only this single mutation was found in the S segment sequence even after 11 passages of virus growth. Interestingly, amino acid Ala (GCA) at position 43 found in Esl/34Aa is also present in all other DOBV N protein sequences including the Vero E6-adapted isolates Slo/Af, Saa/160V and AP/Af. On the other hand, Thr at this position is harbored by N proteins of many Hantaan virus (HTNV) and Seoul virus (SEOV) strains and most of the other hantaviruses.

3.4.5 Antigenic characterisation of the DOBV-SK/Aa isolate

First antigenic characterisation was performed by IFA with a panel of eight mAbs directed against HTNV, SEOV and ANDV N protein. The obtained pattern for the SK/Aa and Slo/Af isolates was identical (Table 17, upper part). Both isolates reacted with Eco2, E5/G6, C116D11 and 2E2 mAbs (all raised against HTNV) while both did not react with B5D9, C24B4 (both against HTNV), R31 (against SEOV), and 5C2/E10 (against ANDV). In the cases where the respective epitope locations in the N protein are known a comparison of the SK/Aa and Slo/Af sequences at these positions showed no or only a few aa exchanges which could all be considered as conservative according to Dayhoff et al. (1978) (Table 17, lower part).

↓57

Nine convalescent sera of HFRS patients from Germany and Slovakia were used to compare the neutralising antibody titers against the new isolate SK/Aa as a representative of the DOBV-Aa lineage and the prototype strain Slo/Af from the DOBV-Af lineage. All sera from German HFRS patients have been serotyped as anti-DOBV specific by FRNT (Sibold et al., 2001; chapter 3.6).

Four patient's sera showed four- and one serum sixteen-fold higher reciprocal titer to SK/Aa than to Slo/Aa. In contrast, two sera exhibited four-fold higher reciprocal titer to Slo/af strain. Two sera reacted at equal end point titers with both viruses (Table 18).

Table 17: Reactivity of DOBV isolates Slovakia and Slovenia in IFA with the panel of anti-N protein monoclonal antibodies.

Virus

Monoclonal antibody / raised against / epitope region

 

Eco2

E5/G6

C16D11

2E2

B5D9

C24B4

R31

5C2/E10

HTNV

HTNV

HTNV

HTNV

HTNV

HTNV

SEOV

ANDV

1-103

166-175

conf.

n.a.

1-45

conf.

n.a.

40-59

Slovakia

+

+

+

+

-

-

-

-

Slovenia

+

+

+

+

-

-

-

-

aa exchanges in epitope region

S13N
T43A
V50I
S56G

no

n.a.

n.a.

S13N
T43A

n.a.

n.a.

T43A

V50I

S56G

↓58

Table 18: Comparison of DOBV isolates SK/Aa and Slo/Af with respect to their neutralisation by convalescent sera from Central European HFRS patients.

serum # a

c-FRNT b

SK/Aa

Slo/Af

1

10,240

640

2

10,240

2,560

3

2,560

640

4

640

160

5

640

160

6

2,560

2,560

7

640

640

8

2,560

10,240

9

640

2,560

a Serum samples of HFRS patients clinically characterized elsewhere (sera ##1-6, Sibold et al., 2001; serum #7, Klempa et al., 2004; sera #8 and #9, our unpublished data). Sera ##1-7 were taken from patients in Germany, #8 and #9 in Slovakia.
b Reciprocal end-point titers are given as determined by c-FRNT.

3.5 First genetically confirmed DOBV infection in Central Europe

3.5.1 Clinical description

A 19-year-old male (patient H169) from Ratzeburg (North Germany) was admitted to the local hospital with a five day history of fever (< 39°C), back- and abdominal pains, nausea, vomiting and diarrhoea (two days). Physical examination showed results within normal limits besides occurrence of hepatosplenomegaly. The laboratory tests exhibited very high levels of urea (BUN), serum creatinine and C-reactive protein. Proteinuria and hematuria could be observed (Table 19). He was treated with Paracetamol and with antibiotics (Cotrimoxazol and Ciprofloxacin). With the diagnosis of acute renal failure, after three days (nine days after onset of disease) he was transferred to the University hospital in Lübeck, where he underwent three courses of hemodialysis. Initial complications of hypertension (< 150 mmHg), leucocytopenia and thrombocytopenia were reported. However, the biochemical values gradually returned to normal and the patient was discharged from hospital in good physical conditions five days after third dialysis, about 16 days after onset of disease.

Acute hantavirus infection was identified after the transfer of patient to the University Hospital in Lübeck. By using the POC Hanta test (Erilab Ltd, Finland), for detecting of IgM antibodies to various hantaviruses, IgM antibodies to hantavirus were found in patient’s serum taken nine days after onset. The same serum was then examined for the presence of IgM, IgA and IgG antibodies against DOBV and PUUV by in-house ELISA. The titers of IgM, IgA and IgG antibodies to DOBV were determined to be 1:25600, 1:12800 and 1:1200 whereas PUUV-specific antibodies could not be detected. The antibody titer of 1:6400 was detected by in-house immunofluorescence assay (IFA) against DOBV-infected cells (strain Slo/Af). Blood, serum and urine samples taken the same day were also used for detection of hantaviral RNA by RT-PCR.

3.5.2 DOBV infection confirmed by sequence analysis

↓59

RNA for RT-PCR was extracted from patient’s serum, urine, EDTA-blood and blood stored in AVL Buffer (Qiagene Viral RNA Kit). DOBV S segment specific nested RT-PCR produced a DNA band of expected size (599 nt, nt position 357-955) only for the sample from AVL-blood (Figure 16). Nucleotide sequence of this fragment was determined (557 nt, nt position 378-934 when excluding the primer sequences) and designated DOBV/H169 (H169).

The sequence analysis showed a clear uniqueness of the patient’s sequence, sharing the highest similarity of only 87.4-87.7% nt identity with DOBV-Aa strains from East Slovakia. However, the similarity to DOBV-Af strains was only slightly lower, 86.3-87.0% nt identity. Within the DOBV species, strain Saa/90Aa showed the lowest similarity to the patient sequence, representing only 86.1% nt identity. Although the sequence diversity to other DOBV sequences was very high, most of the nucleotide exchanges represented silent mutations, because percent identity values of the corresponding amino acids (71 aa, aa position 115-185) were very high. Interestingly, the most similar strain appeared to be our new DOBV-Aa isolate SK/Aa (99.4%) (Table 20).

In ML phylogenetic tree based on the 557 nt S segment sequence (nt position 378-934), H169 clustered within the DOBV species and shared a common ancestor with the DOBV-Aa lineage. However, the PUZZLE support for this position of H169 was slightly below the 70% threshold limit (Figure 17). The use of an only 385 nt long fragment (nt position 378-762) allowed us to include the only DOBV-Aa sequence from Slovenia, Slo/9Aa, in the phylogenetic analysis. Interestingly, when we used this shorter alignment including Slo/9Aa, the tree topology remained unchanged but the statistical support, particularly for position of H169, was improved (Figure 18). Two main branches within the DOBV species could be determined. The first well-supported group was formed by DOBV sequences originated from A. flavicollis rodents (Slovenia, Slovakia), HFRS patients from Greece and by Saaremaa-Aa strains from Estonia. The second monophyletic group consisted of the DOBV-Aa strains from Slovakia, Slovenia, and Russia and of our new sequence H169. The node giving rise to this group is relatively short (0.01175) although the obtained PUZZLE support of 74% exceeded the 70% threshold limit. H169 represents the most basal strain within this group.

↓60

Table 19: Selected laboratory and clinical findings of the H169 patient with acute DOBV infection

Parameter

Patient H169

Sex

male

Age, years

19

Max. serum creatinine (<97 µmol/l)

1294

Max. BUN (<49.9 mg %)

378.6

Max. extent of albuminuria (<0.3g/d)

5

Days of albuminuria

7

Erythrocyturia (<4/µl)

300

Max. white blood count (<9/nl)

6.4

Min. platelet count (>150 /nl)

98

C-reactive protein (<8.2 mg/l)

42

Max. systolic blood pressure, mmHg

150

Oedema

no

Days of fever

5

Days of oliguria

0

Dialysis treatments (number)

3

Normal range values are depicted in brackets, pathological findings are in bold.

Figure 16: Results of diagnostic DOBV S segment specific PCR with samples from patient H169 visualised by electrophoresis in 1.0% agarose gel after 20 min of staining in ethidium bromide (1 μg/ml) .

Starting material for RNA extraction:
1. serum, 2. EDTA-blood, 3. urine, 4. AVL-blood (560μl AVL Buffer + 140μl blood), 5. Diluted #4 (560μl AVL Buffer + 140μl sample #4), 6. negative control ( DEPC water), 7. positive control (DOBV/Slo/Af RNA).

Table 20: Partial S segment nucleotide (599 nt, nt position 357-955) and amino acid (71 aa, aa position 115-185) percent identity of patient-derived sequence H169, other DOBV strains, HTNV, and SEOV

Strain

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

1.

H169

-

87.6

87.4

87.7

87.4

87.2

86.5

86.1

86.7

86.3

86.7

87.0

87.0

86.7

86.5

70.7

70.3

2.

Esl/862Aa

98.9

-

93.7

99.1

89.7

89.5

86.5

87.2

84.3

85.6

85.6

85.0

84.7

84.7

84.5

70.7

71.6

3.

SK/Aa

99.4

99.4

-

94.2

90.6

90.4

84.9

84.9

84.0

84.7

84.7

84.0

85.0

85.0

84.5

70.7

71.0

4.

Esl/194Aa

98.9

100

99.4

-

89.9

89.7

86.7

87.0

85.2

85.8

86.1

85.2

84.9

84.9

85.0

70.9

71.6

5.

Kur/53Aa

98.3

98.3

98.9

98.3

-

99.8

87.4

87.0

87.2

87.2

87.2

87.2

87.7

87.7

87.0

71.9

70.9

6.

Kur/44Aa

97.8

97.8

98.3

97.8

99.4

-

87.2

86.8

87.0

87.0

87.0

87.0

87.6

87.6

86.8

71.8

70.9

7.

Saa/160v

96.7

96.7

97.2

96.7

96.2

95.6

-

98.2

87.4

87.9

87.6

86.8

86.8

86.8

86.3

70.7

71.2

8.

Saa/90Aa

96.7

96.7

97.2

96.7

96.2

95.6

100

-

86.8

87.9

87.6

86.5

86.5

86.5

85.9

70.9

70.7

9.

Esl/400Af

98.3

98.3

98.9

98.3

97.8

97.2

97.2

97.2

-

96.2

95.8

96.4

94.7

94.7

96.4

70.0

69.1

10.

Slo/Af

97.2

97.8

97.8

97.8

96.7

96.2

97.2

97.2

98.9

-

99.2

96.0

95.3

95.3

96.2

70.5

69.4

11.

DOB-Af1

97.2

97.8

97.8

97.8

96.7

96.2

97.2

97.2

98.9

100

-

96.4

94.9

94.9

96.2

70.1

69.6

12.

DOB-PR

98.3

98.3

98.9

98.3

97.8

97.2

97.2

97.2

100

98.9

98.9

-

95.6

95.3

98.7

70.1

70.0

13.

DOB-EA

98.3

98.3

98.9

98.3

97.8

97.2

97.2

97.2

100

98.9

98.9

100

-

99.6

95.8

71.0

69.8

14.

DOB-SZ

98.3

98.3

98.9

98.3

97.8

97.2

97.2

97.2

100

98.9

98.9

100

100

-

95.5

71.2

69.8

15.

DOB-TD

98.3

98.3

98.9

98.3

97.8

97.2

97.2

97.2

100

98.9

98.9

100

100

100

-

70.7

69.8

16.

SEOV

75.6

75.6

76.2

75.6

76.2

75.6

75.1

75.1

76.2

75.1

75.1

76.2

76.2

76.2

76.2

-

68.7

17.

HTNV

77.2

76.7

77.2

76.7

77.2

77.2

77.2

77.2

77.8

76.7

76.7

77.8

77.8

77.8

77.8

76.2

-

The identity values were calculated using Clustal method. The percentage identities for nucleotide (above the diagonal) and amino acid (below the diagonal) sequences are presented.

↓61

Figure 17: Phylogenetic ML tree based on the partial S segment nucleotide sequences (378-934 nt) of H196 and other DOBV strains, computed with TREE-PUZZLE.

Figure 18: Phylogenetic ML tree based on the partial S segment nucleotide sequences (378-762 nt) of H196 and other DOBV strains, computed with TREE-PUZZLE.

3.6 First report of DOBV-Af associated HFRS cases in Slovakia

Most of the HFRS cases due to DOBV infection in Slovakia are occurring in the Eastern part of the country where A. agrarius is dominating over A. flavicollis in rodent population. Several DOBV-Aa strains and only one DOBV-Af strain were also directly detected in local Apodemus populations. Therefore, it was interesting to investigate two DOBV-associated HFRS cases from West Slovakia, where A. flavicollis but not A. agrarius is present in nature.

↓62

A 13-year-old boy (patient A) from Trencin (West Slovakia) had a three-day history of fever (<39.0 °C), abdominal pain and vomiting. He was treated as an outpatient with Amoxycillin (1g tid po) and Bactisubtil® (dried Bacillus cereus spores, 105 mg/d po). Because of persisting abdominal pains and nausea, diarrhoea and urticaria on skin, he was admitted to the hospital eight days after the onset of symptoms. The laboratory tests (Tables 21 and 22) exhibited very high levels of urea (>40mmol/L), serum creatinine (>900 μmol/L), and C-reactive protein (34.5 mg/L). With the diagnosis of acute renal failure, he was transferred to intensive care unit where he underwent two courses of hemodialysis (at days 9 and 10 after onset). After dialysis the biochemical values gradually returned to normal and patient recovered within five days. He was discharged with normal laboratory and clinical findings 14 days after hospitalisation, 22 days after onset of illness.

A 11-year-old boy (patient B), the brother of patient A, was admitted to the hospital one day later than his brother. Four days before admittance, the patient suffered from fever (38.7°C), nausea, vomiting, abdominal pain and petechiae on face and neck. Laboratory tests (Tables 21 and 22) revealed high levels of urea (22.6 mmol/L, serum creatinine (290 μmol/L) and 10-fold elevation of liver enzymes. Because of similar symptoms of his brother, he was transferred to intensive care unit with a diagnosis of a hepatorenal syndrome at the same day as his brother. However, dialysis was not necessary. Under parenteral rehydratation treatment, his clinical and laboratory parameters gradually improved to normal values. During the hospitalisation, he developed bronchopneumonia which was treated with parenteral antibiotics. He was discharged from hospital in good physical condition 21 days after onset of disease.

With routine diagnostic methods, no evidence for acute bacterial, fungal, or viral infections was found. When patients were discharged from hospital, the etiology of disease was still unknown. Patients reported that they had been regularly performing water sports on a natural water canal near Trencin. After intervention of the local epidemiologist (Dr. Stankovicova, Trencin), the convalescent sera were sent to our collaboration partner, Institute of Virology, Bratislava, Slovakia, as late as 17 months after the disease. IgG antibodies in sera of both patients were tested by nucleocapsid-protein specific ELISA using PUUV and HTNV antigens (Progen GmbH) and found to be positive. By in-house immunofluorescence assay (IFA), antibody titers were found to be higher against DOBV and HTNV than against PUUV suggesting DOBV infection (Table 23).

↓63

This virological diagnosis was then confirmed by FRNT. We have determined the neutralising activities of sera #1 against the following viruses; DOBV, HTNV, SEOV, and PUUV. An at least fourfold higher reciprocal titer was found against DOBV strain Slo/Af when compared with the other viruses (Table 23). However, DOBV strain originating from A. agrarius was not available for us at that time. After the isolation of a DOBV-Aa strain from Slovakia the comparison with DOBV-Aa could be done. Sera obtained 43 months after onset of disease were analysed and showed fourfold higher reciprocal titer against Slo/Af than against SK/Aa (Table 23) indicating that DOBV-Af was responsible for the infection of both patients.

Table 21: Clinical and laboratory data of two DOBV-Af patients according to criteria for estimating severity of the various phases of HFRS

Phase

Observation

Criteria for severity*

Patient

Mild +

Moderate ++

Severe +++

A

B

Febrile

Max. temperature

<39

39-40

>40

+

+

Days temp. over 38 °C

1-5

5-7

>7

+

+

Petechiae, flushing

0/+

++/+++

++++

+

+

Max WBC count

<14.500

14.5-30x103

>30.000

+

++

Hypotensive

Max. haematocrit (%)

<50

51-56

>57

+

+

Min. systolic BP (mmHg)

96-120

81-95

<80

+

+

Hours of hypotension

<24

24-48

>48

+

+

Min. platelets

>90.000

4-9x104

<40.000

+

+

Oliguric

Min. haematocrit

>45

35-44

<35

++

++

Max. systolic BP (mmHg)

<140

141-170

>170

++

+

Days of hypertension

<1

1-2

>2

++

+

Max. BUN (mg %)

20-79

80-149

>150

+++

++

Days of proteinuria

<4

4-5

>5

+++

+

Diuretic

Max. daily urine vol.

<3.4l

3.5-4.9l

>5l

+

+

Criteria for estimating severity of the various phases of HFRS according to (Lee and van der Groen, 1989).

Table 22: Further laboratory and clinical data of the Trencin patients of acute DOBV infection

Parameter

Patient

A

B

Sex

male

male

Age, years

13

11

Max. serum creatinine (<102µmol/l)

930

304

Serum creatinine at discharge

85

64

Min. platelet count (>140 Gpt/l)

189

137

Max. platelet count (<440 Gpt/l)

377

366

Max. white blood count (<10.8 Gpt/l)

9.4

16.9

C-reactive protein (<8.2 mg/l)

34.5

27.37

GOT (<0.49 µmol/l/s)

0.76

2.63

GPT (<0.57 µmol/l/s)

1

2.33

Max. extent of proteinuria (<0.15g/d)

0.256

0.258

Max. systolic blood pressure, mmHg

150

130

Systolic blood pressure at discharge, mmHg

110

115

Oedema

no

no

Days of fever

3

3

Days of oliguria

0

0

Dialysis treatments (number)

2

0

Normal range values are depicted in brackets, pathological findings are in bold.

↓64

Table 23: Detection of hantavirus-specific antibodies in serum samples of HFRS patients from West Slovakia

Seruma

ELISA b

IFA c

c-FRNT d

HTNV

DOBV

PUUV

HTNV

DOBV

PUUV

DOBV-Af

DOBV-Aa

HTNV

SEOV

PUUV

TULV

Patient A

 

#1

pos.

n.d.

pos.

6400

6400

400

2560

n.d.

40

640

40

40

#2

pos.

pos.

pos.

n.d.

25600

n.d.

10240

2560

n.d.

n.d.

n.d.

n.d.

Patient B

 

#1

pos.

n.d.

neg.

6400

6400

200

2560

n.d.

160

160

40

40

#2

pos.

pos.

pos.

n.d.

12800

n.d.

2560

640

n.d.

n.d.

n.d.

n.d.

a Sera were taken 17 months (#1) and 43 months (#2) after onset of disease
b s/c ratios > 1.5 were taken to be positive. The IgG screening ELISA (Progen GmbH, Heidelberg, Germany) was used.
c Reciprocal end-point titers are given. Vero-E6 cells used in IFA were infected by HTNV (strain 76-118), DOBV (strain Slovenia) or PUUV (strain Sotkamo) .
d Reciprocal end-point titers are given. The chemiluminescence focus reduction neutralisation test was undertaken by use of DOBV-Af (strain Slovenia), DOBV-Aa (strain SK/Aa), HTNV (strain 76-118), SEOV (strain 80-39), PUUV (strain Sotkamo), and TULV (strain Moravia).

3.7 Occurrence of renal and pulmonary syndrome in a region of North-East Germany where Tula hantavirus circulates

3.7.1 HFRS clinical case associated with TULV infection

3.7.1.1 Clinical description

A male 43-year-old patient from a rural region near Cottbus (Northern German Plain) was admitted to the hospital. Since the previous day he had suffered from fever (40.0 0C), chills, headache and left-thoracic, breathing-dependent pain. The X-ray examination showed an infiltration within the left pulmonary midzone. The laboratory tests exhibited elevated levels of serum creatinine (116 μmol/L ≅ 1.3 mg/dL), C-reactive protein (91.55 mg/L) and a deviation to the left in the differential blood count. Platelet count, serum bilirubin and transaminase values were found to be normal. However, proteinuria, erythrocyturia, urobilinogenuria and bilirubinuria have been detected in urine. During clinical course, the proteinuria reached a value of 4.55 g/d. No oliguria was observed but a moderate polyuria did occur at day 7 and the following days. With routine diagnostic methods, no evidence for acute bacterial, fungal, or viral infections (with the exception of elevated hantavirus antibody titers, see below) was found to explain the nephritis and pneumonia. A spontaneous remission of clinical symptoms and laboratory values was observed during supportive treatment of the patient. The man had not visited other countries during the previous years. However, he reported that he had frequently seen and trapped rodents in his rural living place.

At day 15 the patient was discharged from the clinic, however, at day 21 he was hospitalized again with symptoms of fever (up to 40 0C), unproductive cough and pain in the left thorax. Diagnostic radiology now demonstrated an infiltration in the basolateral segment of left lower lobe, whereas the primary infiltration in the left pulmonary midzone was no longer detectable. Again, elevated levels of serum creatinine and C-reactive protein were found, but this time no proteinuria occurred. Ribavirin treatment (RebetolR 1g/d) was started and maintained over two weeks. Within one week after the second admission (day 28 after onset of illness) the patient was free of fever and biochemical values gradually returned to normal. At day 36 after onset the patient was finally discharged from the clinic. In the follow-up period no clinical or laboratory deviations were observed. In particular, there was a complete remission of the pulmonary infiltration.

3.7.1.2 Serological data

↓65

The hantavirus antibody titers against PUUV antigen as determined by IFA (Progen GmbH, Heidelberg, Germany) increased from 1:128 (day 2 after onset), through 1:256 (day 12) to 1:3,200 (day 38) demonstrating an acute hantavirus infection (laboratory Dr. Limbach & Partners, Heidelberg, Germany). The first serum (#1) studied by us was taken 27 days after onset of symptoms. The patient gave his informed consent to these investigations. IgG antibodies were tested by nucleocapsid-protein specific ELISA using PUUV and HTNV antigens (Progen GmbH) and found to be positive. IgM antibodies could not be detected by IgM specific ELISA (Progen GmbH). By in-house immunofluorescence assay (IFA), antibody titers were detected against PUUV- (1:512), TULV- (1:512), HTNV- (1:32) and DOBV-infected cells (1:256). A second serum sample (#2) taken 7 months later confirmed these findings (Table 24). The slightly higher IFA antibody titers against the viruses of the PUUV/TULV group as compared with DOBV/HTNV could indicate an infection by PUUV or a related hantavirus.

The chemilumuniscent focus reduction neutralisation test (c-FRNT), which is performed under biocontainment level-3 conditions, is the only method for fine-typing of sera positive for hantavirus antibodies. By c-FRNT we have determined the neutralising activities of sera #1 and #2 against the following viruses; PUUV, TULV, HTNV, DOBV, and SEOV. As shown in Table 24, an at least fourfold higher reciprocal titer was found against TULV when compared with the other viruses. This supports the assumption that TULV was the virus responsible for the infection of the patient. Nevertheless, it cold not be completely excluded that the infection was caused by a TULV-related, albeit unidentified virus not present in the FRNT virus collection. It was thus important to show that TULV is endemic in this region of Germany.

3.7.2 Detection of TULV in Microtus arvalis from North-East Germany

In the course of our epidemiological studies on hantavirus distribution in Germany, rodents were trapped and screened for hantavirus infection. Infected M. arvalis rodents could be captured in North-East (NE) Germany at two sites a few kilometers far from the home village of the patient. From tissues of three out of 18 tested animals hantaviral RNA could be detected by RT-PCR and the nucleotide sequence of the complete S segment was determined.

↓66

In all three cases (samples no. D5, D17, D63) the S segments were found to be 1,852 nt in length (1,828 nt, when excluding the primer sequences) and revealed an ORF1 encoding a nucleocapsid protein of 430 aa, flanked by a 5‘ NCR of 42 nt and a 3‘ NCR of 517 nt. In addition to the ORF1, all three samples contained the potential second ORF2, which overlaps the nucleocapsid protein ORF1 and is 273 nt long. The ORF2 reading frame is in +1 position with respect to the ORF1, encoding a putative nonstructural protein of 90 aa.

Percent sequence identities on the nucleotide and amino acid level for the three samples in comparison to all other European Tula viruses are given in Table 25. Sequence analysis revealed that their genetic difference followed the geographic distance of the trapping sites: Tula D5 and Tula D17 (trapping sites < 1km apart) showed 99.8% (nt) and 99.7% (aa) identities, whereas Tula D63 (trapping site about 3 km apart) showed identities of 99.4-99.6% (nt) and 99.3-99.5% (aa), respectively. The most similar sequences were found in the Polish strains (91.2-91.5% nt and 98.6-90.0% aa identity). Similarity to all other strains was a approximately on the same level (83.6-84.6 nt, 96.5-97.9 aa identity).

A summary of the amino acid changes between the Tula strains is shown in Table 26. At five positions, unique amino acid changes consistent among the NE German and Polish samples were found that distinguished them from the other Tula isolates. At position 263, aa Tyr could be found only in two of our strains. All other TULV strains show a non-conservative exchange to aa Asn at that position.

↓67

Comparison of the 3‘ NCR of the NE German strains revealed > 99.6% nt identity. Comparison with other Tula strains yielded nt identities of about 80.6%-90.9%. Unique for our samples a consistent two nt insertion was observed immediately downstream of the stop codon (nt 1340-1341), as well as another at nt 1447-1448. The 16 nt long deletion that was shown to be typical for Tula strains from Southern Central Europe (Sibold et al., 1999a) was neither found in the NE German nor the Polish strains.

Including known complete S segment sequences of other TULV strains, a ML phylogenetic tree was constructed (Figure 19). The results show that the NE German strains form a monophyletic group with the Polish strains. TULV strains can be divided into three distinct, well-supported lineages. The first lineage is represented by strains from Russia (East Europe), the second by strains from the Czech Republic and West Slovakia, Croatia and South Germany (Southern Central Europe) and the third by strains from North-East Germany and Poland (North-Central Europe). Regarding the third genetic lineage it seems to be reasonable to conclude that its distribution follows the Northern German Plain which extents to Poland in the East.

Table 24: Detection of hantavirus-specific antibodies in serum samples of the patient

Serum a

ELISA b

IFA c

c-FRNT d

PUUV

HTNV

PUUV

TULV

HTNV

DOBV

PUUV

TULV

HTNV

DOBV

SEOV

#1 (H145)

pos.

pos.

512

512

32

256

< 40

160

< 40

< 40

< 40

#2 (H53)

pos.

pos.

1024

256

64

256

40

160

40

< 40

< 40

a Sera were taken 4 weeks (#1) and 7 months (#2) after onset of disease
b s/c ratios > 1.5 were taken to be positive. IgG screening ELISAs (Progen GmbH, Heidelberg, Germany) were used.
c Reciprocal end-point titers are given. Vero-E6 cells were infected by PUUV (strain Sotkamo), TULV (strain Moravia), HTNV (strain 76-118) or DOBV (strain Slovenia) and used in immunofluorescence assay (IFA).
d Reciprocal end-point titers are given. The chemiluminescence focus reduction neutralization test was undertaken by use of PUUV (strain Sotkamo), TULV (strain Moravia), HTNV (strain 76-118), DOBV (strain Slo/Af), and SEOV (strain 80-39).

↓68

Table 25: Percent sequence identities among North-East German and other European TULV strains based on entire S segment and N protein sequences

Strains *

D5

D17

D63

nt

aa

nt

aa

nt

aa

D5

-

-

99.8

99.7

99.4

99.3

D17

99.8

99.7

-

-

99.6

99.5

D63

99.4

99.3

99.6

99.5

-

-

Poland

91.5-91.6

98.8-99.0

91.4-91.5

98.6-98.8

91.2-91.3

98.6-98.8

Russia

84.4-85.6

96.9-97.4

84.4-85.5

96.5-97.2

84.3-85.4

96.5-97.2

East Slovakia

84.3-84.8

97.6-97.9

84.2-84.7

97.4-97.6

84.1-84.6

97.4-97.6

West Slovakia

84.5

97.4

84.2-84.3

97.2

84.2-84.3

97.2

Czech Republic

83.7-84.6

96.7-97.4

83.6-84.3

96.5-96.9

83.6-84.3

96.5-96.9

Croatia

84.6

97.2

84.6

96.9

84.5

96.9

South Germany

84.2

96.2

84.2

96.0

84.2

96.0

* The European TULV strains used in this analysis were grouped according to their geographical of origin. Acc. numbers of the strains included in these groups are listed in Table 26.

Table 26: Amino acid exchanges in N protein of TULV strains

Acc. No.

polymorphic sites

country of origin

2

3

42

48

60

84

121

127

134

165

234

250

252

254

256

258

263

268

305

312

323

387

404

 

Z30941

S

Q

N

R

E

I

A

I

V

F

E

Q

S

G

E

E

N

I

A

A

I

M

D

Russia

Z30942

S

Q

N

R

E

L

A

I

V

F

E

Q

S

G

E

E

N

I

A

A

I

M

D

Z30943

S

Q

N

R

E

L

A

I

V

F

E

Q

S

G

E

E

N

I

A

A

I

M

D

Z30944

S

Q

N

R

E

I

A

I

V

F

E

Q

S

G

E

E

N

I

A

A

I

M

D

Z30945

S

Q

N

R

E

L

A

I

V

F

E

Q

A

G

E

E

N

I

A

A

I

M

D

Y13979

S

Q

S

R

D

L

S

I

V

F

E

Q

A

G

E

D

N

L

A

A

I

T

D

E Slovakia

Y13980

S

Q

S

R

D

L

S

I

V

F

E

Q

A

G

E

D

N

L

A

A

I

T

D

Z48235

S

Q

S

R

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

A

I

T

D

W Slovakia

Z68191

S

Q

S

R

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

A

I

T

D

Z48573

S

Q

S

G

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

R

I

T

S

Czech Rep.

Z48574

S

Q

S

G

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

R

I

T

S

Z48741

S

Q

S

R

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

R

I

T

D

Z49915

S

Q

S

R

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

R

I

T

D

Z69991

S

Q

S

R

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

R

I

T

D

AJ223601

S

Q

S

R

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

R

I

T

D

AJ223600

S

Q

S

R

D

L

S

T

I

Y

E

S

-

S

E

D

N

L

A

R

I

T

S

AF164094

S

Q

S

R

D

L

S

T

I

Y

E

N

-

S

E

D

N

L

A

A

I

T

D

Croatia

AF164093

S

Q

S

Q

D

L

S

M

I

Y

E

S

-

S

E

D

N

L

A

A

I

T

D

S Germany

D5

S

Q

S

R

D

L

S

I

V

Y

D

Q

G

N

D

E

Y

L

A

A

V

M

D

NE Germany

D17

S

Q

S

R

D

L

S

I

V

Y

D

Q

G

N

D

E

Y

L

G

A

V

M

D

D-63

S

Q

S

R

D

L

S

I

V

Y

D

Q

G

N

D

E

N

L

G

A

V

M

D

Af063892

T

E

S

R

D

L

S

M

V

Y

D

Q

G

N

D

E

N

L

A

A

V

M

D

Poland

Af063897

T

E

S

W

D

L

S

M

V

Y

D

Q

G

N

D

E

N

L

A

A

V

M

D

Only informative sites are shown. Amino acid residues unique for NE German and/or Polish strains are grey boxed. TULV lineages as defined in Figure 19 and recombinant strains from East Slovakia are separated by empty lines.

Figure 19: ML phylogenetic tree of TULV strains based on complete S segment nucleotide sequences

TULV lineages are grey boxed. North-East German strains are in bold.


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.
DiML DTD Version 4.0Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML generated:
20.10.2005