Previous Article | Next Article ![]()
Journal of Virology, October 2003, p. 11268-11273, Vol. 77, No. 20
0022-538X/03/$08.00+0 DOI: 10.1128/JVI.77.20.11268-11273.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
ApoGene Biotechnologie, D-86567 Hilgertshausen, Germany,1 Institut für Tierzucht und Genetik, Veterinärmedizinische Universität Wien,2 Ludwig-Boltzmann-Institut für Immuno-, Zyto- und Molekulargenetische Forschung Wien, A-1210 Vienna, Austria3
Received 20 May 2003/ Accepted 21 July 2003
|
|
|---|
families were revealed. Novel open reading frames (ORF) in the amplified proviral fragment were found in one OERV ß family and two OERV
families. Hybrid OERV produced by putative recombination events were not detected. Quantitative analysis of the OERV sequences in the ovine genome revealed no relevant variations in the endogenous retroviral loads of different breeds. Expression analysis of different tissues from fetal and pregnant sheep detected mRNA from both gammaretrovirus families, showing ORF fragments. Thus, the release of retroviruses from sheep cells cannot be excluded. |
|
|---|
(C-type) genera (36) and have been found in multiple copies in all mammals. Mutations led to the loss of function of most of the proviral loci (7). In addition to the extensive studies in mice (7), full-length ERV have been identified in several species, including humans and pigs (1, 14, 23, 35). The potential involvement of biologically active ERV in physiological and pathological processes in the host has been discussed intensively (19, 28). In vitro and in vivo experiments have shown the potential of cross-species infection of ERV from several species (reference 5 and references therein). In pigs, retroviral recombination is thought to be involved in generating novel retroviral genomes with unknown consequences for host tropism and pathology (18, 20, 25). The use of genetically modified animals in xenotransplantation has been suggested to enhance the potential risk of ERV transmission (3). In addition, ERV contamination has been found in biomedical products derived from different species, including pigs (12, 32) and chicken (15), but no infectious particles have been detected. Apart from the purification and virus inactivation steps performed in the isolation procedure of transgenic products, this implies that the principal potential risk of ERV exists for biotechnological products derived from transgenic animals.
In sheep (Ovis aries), ERV have been found that are highly homologous to the exogenous jaagsiekte sheep retrovirus (exJSRV) (38) causing ovine pulmonary carcinoma (OPC) and to the related enzootic nasal tumor virus (ENTV) (8). Both JSRV and ENTV belong to the ovine betaretroviruses. In the genomes of domestic breeds and wild sheep, 15 to 20 endogenous JSRV (enJSRV) copies were identified by Southern blot analysis (11). Examination of enJSRV loci revealed intact genes for several loci, but no replication-competent provirus has been found. mRNA expression of enJSRV has been detected in uterus, gut, and lung tissues. Despite the high homology between exJSRV and enJSRV, differences in the env gene of the endogenous copies resulted in different tissue tropism (26). Retroviral vectors pseudotyped with exJSRV and ENTV env genes have been used to infect human cells and other mammalian cells in vitro (2, 29). The potential role of enJSRV during pregnancy is under examination (31). In addition to the intensively investigated ß-type enJSRV, two pro-pol sequences (GenBank accession no. X99931 and X99932) were described and classified as
ovine endogenous retroviruses (OERV); however, they showed no open reading frame (ORF). The sequence assigned GenBank accession no. X99931 was later determined to have originated from mink (mink endogenous retrovirus I [MiEVI]) (34).
Because of the use of sheep for various biotechnological purposes, we precisely examined the OERV ß and
pro-pol nucleotide sequences. ORF were identified for two OERV
families. mRNA expression of both families was found in several tissues of fetal and pregnant individuals.
Amplification and phylogenetic analysis of the OERV sequences.
OERV pro-pol sequences in genomic DNA of the Bergschaf breed were amplified and analyzed as previously described (18). Using six degenerate primer pairs (13, 34), 20 clones were sequenced bidirectionally for each primer pair. Of the 120 clones examined by BLAST search, 101 were of retroviral origin and showed a length of 0.7 to 1.0 kb. Seventeen clones were assigned to the OERV ß family, 84 clones were assigned to the OERV
family, and no clone grouped with spumaretroviruses. From the 101 OERV clones amplified in this study, four pairs of 0.9-kb clones showed the identical sequence, thereby indicating a low rate of polymerase errors in the PCR procedure as previously described (20).
For further classification of the amplified pro-pol fragments, clones showing more than 90% sequence identity were grouped into families, and the families were named according to the results of studies in pigs (27). The least defective clone was chosen as representative of the families. Thus, the clones were put into two closely related betaretrovirus families (OERV ß2 and ß3 families) containing 2 and 15 clones and to nine families that grouped with gammaretroviruses (OERV
1 to
9 families) with 1 to 22 members. None of the clones was assigned to the betaretrovirus enJSRV, where 15 to 20 copies per genome have been found (11). As examination of the hybridization site of the forward primers used revealed a conserved 3-nucleotide (nt) gap in the pro gene of enJSRV, enJSRV pro-pol sequences were amplified with specific primers (GenBank accession no. AF136224, nt 2752 to 2772 and nt 3720 to 3700, respectively) (26) and designated OERV ß1. Phylogenetic analysis was performed with different commonly used algorithms (maximum likelihood, most parsimony, and neighbor joining) by using the PHYLIP package (http://evolution.genetics.washington.edu/phylip.html) for both the nucleotide and protein sequences.
OERV ß. As described for human ERV K and the mouse mammary tumor virus, JSRV showed a different reading frame for pro and pol (26). Sequence analysis of the 11 0.9-kb OERV ß1 clones resulted in the correct ORF for both pro and pol, including the different reading frame for both genes in all amplified fragments. They represented at least eight different loci. Further phylogenetic analysis of the JSRV family assigned 2 and 7 of the 11 clones to the previously described enJSRV subfamilies A and B, respectively (26). In addition, two clones, clones j18 and j111 (GenBank accession no. AY266331 and AY266332), grouped with the exJSRV subfamilies where they were most closely related to exJSRV type II, which is represented by the full-length clone with GenBank accession no. AF105220 (Fig. 1).
![]() View larger version (61K): [in a new window] |
FIG. 1. Analysis of the amplified OERV ß1 clones. (A) Consensus of 500 neighbor-joining tree with bootstrap values above 50% indicated at the branch nodes. (B) Alignment of the 927-nt pro-pol fragments of OERV ß1 clones. This analysis revealed that the classification was based on differences of only a few polymorphic nucleotides (<1% of the investigated sequence data). The polymorphic nucleotides (n = 74; 8%) are depicted, and their positions in the alignment of the data set are shown above the sequences. Mutant nucleotides (shaded) are indicated. All sequences had the identical reading frame and, therefore, are intact. The classification of the GenBank entries to the endogenous and exogenous subfamilies (enJSRV, endogenous jaagsiekte sheep retrovirus; exJSRV, exogenous JSRV; ENTV, enzootic nasal tumor virus) is indicated according to Palmarini et al. (26). Clones j18 and j111 were submitted to GenBank (accession no. AY266331 and AY266332, respectively).
|
|
View this table: [in a new window] |
TABLE 1. Approximate copy numbers of OERV ß and sequences per haploid genome in sheep breeds by Southern blot analysis
|
ORF-containing OERV
1 and
2 families.
Two
families were found to contain clones with an ORF in the amplified pro-pol region. For OERV
1 (n = 18), two ORF-containing clones, clones p113 and p17 (GenBank accession no. AY193896 and AY193897, respectively), which differed in 12 nt (99% identity) were found. Both sequences showed high identity (98%) to the previously published PERV
8 (GenBank accession no. AF511112) which differed in a defined TA gap from all amplified OERV
1 clones (18). The 16 non-ORF-containing members of OERV
1 showed various mutations in the amplified pro-pol region. Sequence comparison assigned the 18 OERV
1 clones to at least 12 independent loci in the ovine genome (data not shown). Southern blot analysis revealed about 25 copies per haploid genome for all breeds investigated (Table 1).
OERV
2 family contained 22 sequences including the ORF-containing clone am27 (GenBank accession no. AY193900). Clone am27 was identical in 887 of 934 nt (95% identity) to MiEVI (GenBank accession no. X99931), which was originally described as ovine endogenous retrovirus I (OvEVI) (34). The remaining 21 clones of this family had nonsense and/or frameshift mutations. More than 100 copies per haploid genome were detected for OERV
2 (Table 1).
Both OERV
1 and
2 families grouped with leukemia-associated retroviruses and, therefore, are part of the murine leukemia virus-like superfamily which contains ERV of various mammals, including mice, pigs, mink, and monkeys (10, 17, 18, 34, 37). Detailed phylogenetic analysis of the OERV
1 and
2 families revealed significant differences (Fig. 2). The 18
1 clones clustered in three subfamilies named
1A (GenBank accession no. AY193896 and AY193897),
1B (AY193898), and
1C (AY193899) with eight, four, and six members, respectively. In contrast, low bootstrap values and small genetic distances between the ancestral branch nodes in the examination of the OERV
2 clones resulted in weak resolution of the phylogenetic trees.
![]() View larger version (15K): [in a new window] |
FIG. 2. Phylogenetic analysis of the OERV 1 (A) and 2 (B) pro-pol clones amplified in this study. The sequences of the highly homologous PERV 8 (GenBank accession no. AF511112) and MiEVI (GenBank accession no. X99931) were included in the alignments of OERV 1 and 2, respectively. The trees were created by using the maximum-likelihood method, and bootstrap values greater than 60% obtained from 500 genetic distance trees were added to the branch nodes. Fragments harboring an ORF are indicated with an asterisk. Clones shown in bold type were submitted to GenBank (p113, GenBank accession no. AY193896; p17, AY193897; p110, m317, AY193898; am22, AY193899; am27, AY193900; ap23, ap29, AY193901; fm28, m118, AY193902). The genetic distance (0.01) is defined in the PHYLIP package (http://evolution.genetics.washington.edu/phylip.html).
|
3 to
9 families.
Seven additional OERV
families (
3 to
9) were found in our study without revealing ORF-containing clones. Representatives of the OERV
5 (GenBank accession no. AY193905) and
6 (AY193906) families contained only one stop or frameshift mutation. All members of the other OERV families
3,
4, and
7 to
9 showed multiple mutations. The detected copy numbers of the OERV families are given in Table 1. We did not observe the previously described mutant OERV clone OvEVII (GenBank accession no. X99932) (34), which grouped with OERV
9 family, showing 73 and 68% identity on the nucleotide and amino acid levels, respectively.
OERV expression.
To detect expression of the ORF-containing OERV families
1 and
2, six different tissues (heart, kidney, liver, lung, spleen, and fetal thymus) were analyzed in three fetuses and one pregnant sheep of the Bergschaf breed. After extraction of total RNA, potential DNA contamination was removed by treatment with DNase I, and cDNA synthesis was performed with oligo(dT)12-18 primers or random primer hexamers by the standard protocols. Comparable OERV expression patterns were observed in the tissue samples from the three fetuses and the pregnant sheep. OERV
1 fragment expression was mainly detected in cDNA generated with random primers, whereas OERV
2 reverse transcription-PCR (RT-PCR) amplification products were obtained by using both random primers and oligo(dT) primers indicating the presence of polyadenylated transcripts (Fig. 3). Northern blot analysis of the OERV
1 and
2 families did not show any expression signals in both fetuses and the pregnant animal, suggesting low transcription levels (data not shown).
![]() View larger version (66K): [in a new window] |
FIG. 3. RT-PCR expression analysis of OERV 1 and 2 families in six different tissues from fetuses and pregnant sheep. cDNA produced with random primers (RP) and oligo(dT) (dT) was used as the template in the OERV 1 and 2 RT-PCR analysis. As control, non-reverse-transcribed RNA (RNA) was used as a PCR template. The 1-specific primers g1-fw and g1-rv (GenBank accession no. AY193896 and AY193897 and nt 252 to 272 and nt 905 to 885, respectively) and the 2-specific primers g2-fw and g2-rv (GenBank accession no. AY193900, nt 88 to 108 and nt 857 to 837, respectively) were used. The absence of contaminating genomic DNA was proven by the lack of amplification of genomic actin DNA in non-reverse-transcribed RNA (not shown). The amplification of an ß-actin gene fragment (act) from oligo(dT)-primed cDNA was used to show RNA integrity. The lengths of the expected fragments (in base pairs) are given. Amplification products for 1, 2, and act were isolated, cloned, and sequenced to verify the origin of the RT-PCR products. The marker lanes contain 1-kb DNA ladder (Invitrogen, Vienna, Austria). As positive controls, genomic sheep DNA containing 50,000 and 5,000 copies of OERV 1, 2, and ß-actin was used. Water was used as a negative RT-PCR control.
|
1 RT-PCR-amplified clones resulted in the identification of at least six transcriptionally active
1 loci. Two clones amplified from ORF-containing loci were assigned to the OERV
1A subfamily, and another belonged to the
1C subfamily. The other RT-PCR-amplified clones showed multiple mutations and belonged to the
1C subfamily. The
2 RT-PCR products were also verified by sequence analysis. Almost all clones obtained from the RT-PCR analysis were expressed from other loci as defined by the genomic OERV analysis.
Peripheral blood mononuclear cells (PBMC) were chosen for the RNA expression analysis of adult sheep analogous to the results described in pigs (16). mRNA isolated from PBMC of nine adult males and nonpregnant females of three different breeds (Bergschaf, Milchschaf x Romanov, and Tiroler Steinschaf) did not show detectable levels of OERV
1 and
2 transcripts by both Northern blot and RT-PCR analyses. In a further experiment, probes from all OERV ß and
families identified in this study were mixed and hybridized to the Northern blots. Expression was not detected (data not shown).
The extensive search for OERV copies was performed by identification and phylogenetic analysis of 101 independent pro-pol sequences and resulted in several OERV ß and
families as well as in clones harboring an intact ORF. According to our results in pigs (98% PERV
clones, n = 52) (18), use of the degenerate primers resulted in the preferential amplification of OERV
sequences (83%, n = 84 of 101 clones). As no data are available on the phylogenetic relation of the various breeds, the proviral load was determined in a broad spectrum of sheep breeds, including breeds of economic importance (e.g., Merino and milk sheep) and local breeds (e.g., Krainer Steinschaf and Tiroler Steinschaf). No relevant difference in the copy number of the retroviral families in the different breeds was observed. Therefore, integration of the proviruses into the germ line is suggested to have taken place before different breeds of O. aries have evolved. However, additional OERV may be found in the different breeds.
Sequence analysis of three OERV ß enJSRV copies has revealed intact proviral genes in replication-incompetent clones (26). Within the examined 0.9-kb pro-pol region, we identified three OERV
loci harboring an ORF (two OERV
1A and one OERV
2). Compared to the portion of ORF-containing sequences found in PERV
(17%, n = 9) (18), ORF-containing OERV
were shown to occur in a low ratio (4%, n = 3). However, we sequenced only part of the proviral load in the ovine genome, which has a large OERV
2 load. Therefore, further ORF-containing OERV
may exist in the sheep genome, which was confirmed by the detection of additional ORF-containing transcripts.
Phylogenetic examination of the ORF-containing OERV
1 and
2 families revealed significant differences. Three distinct
1 subfamilies (OERV
1A, B, and C) were defined, whereas the reproducibility of the phylogenetic tree of the
2 clones was low. In addition, the genetic distances between the
2 clones were higher than the distances between their ancestors. A possible explanation is the infection of the sheep germ line by at least three different OERV
1 strains, whereas the
2 clones might be the result of multiple integration events into the ovine germ line after a singular infection of an ancestral virus. The occurrence of distinct
1 subfamilies in the pro-pol region raises the question whether the viruses also harbor specific differences in the env gene, which might lead to differences in the host tropism, as in PERV
1 subfamilies (21, 33).
Distinct OERV ß2,
1, and
2 clones are highly homologous to ERV sequences in pigs and mink (96, 98, and 95% identity, respectively). Infection of different species might be caused by the use of a conserved receptor and/or by changes in host tropism due to recombinant env genes. Recently, chimeric functional human tropic PERV
1 caused by putative recombination events have been identified (18, 20, 25). Chimerism has also been found in human ERV (4). In this study, we did not find any recombinant OERV. Thus, it remains to be determined whether there is a preferential occurrence of recombination events in defined ERV families.
Expression of human immunodeficiency virus, mouse mammary tumor virus, and bovine leukemia virus was stimulated by glucocorticoid-responsive elements of the long terminal repeats (6, 24, 30). The detection of retroviral transcripts in fetuses and pregnant females, but not in PBMC of adult males and nonpregnant females in our study correlates with these data. However, activation of OERV
1 and
2 expression by hormones has to be shown.
Nucleotide sequence accession numbers.
The nucleotide sequence data of the representatives of the OERV families have been submitted to GenBank for OERV ß1 (GenBank accession no. AY266331 and AY266332), ß2 (AY193894), ß3 (AY193895),
1A (AY193896 and AY193897),
1B (AY193898),
1C (AY193899),
2 (AY193900 to AY193902),
3 (AY193903),
4 (AY193904),
5 (AY193905),
6 (AY193906),
7 (AY193907),
8 (AY193908), and
9 (AY193909).
|
|
|---|
pro-pol nucleotide sequences. J. Virol. 76:11738-11743.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»