Common promoter deletion is associated with 3.9-fold differential transcription of ovine CCR5 and reduced proviral level of ovine progressive pneumonia virus

Introduction

Ovine progressive pneumonia virus (OPPV), also known as maedi-visna virus, is a common lentiviral pathogen of U.S. sheep. Almost half of flocks contain at least one infected animal, and overall prevalence is estimated at 26% (Cutlip et al. 1992). OPPV results in persistent, lifelong infection and the virus can be transmitted throughout the lifespan (De Boer et al. 1979). Infected sheep display symptoms which include varying degrees of dyspnoea, mastitis, cachexia, arthritis and/or encephalitis (Narayan et al. 1988). Through these clinical symptoms and histopathological changes in the mammary gland, lung, carpal synovial membranes, and central nervous system, OPPV impacts on the health of the infected animal and the development of lambs nursed by infected ewes.

It has been repeatedly observed that sheep breeds have differing odds of infection with OPPV, which suggests that resistance to infection may have a genetic basis (Gates et al. 1978; Cutlip et al. 1986; Houwers et al. 1989; Snowder et al. 1990). More recently, it has been demonstrated that host control of provirus levels post-infection may also have a genetic basis (Herrmann-Hoesing et al. 2008b). The same study identified Ovar-DRB1 as one potential gene involved in host containment of OPPV (Herrmann-Hoesing et al. 2008b). However, the size of the breed differences suggested that other genes are involved, but these have yet to be identified.

Logical candidates for additional genes involved in responses to OPPV may be resistance genes for related viruses, including human immunodeficiency (HIV). Both HIV and OPPV are macrophage-tropic lentiviruses that cause lifelong persistent infection in the host (Gendelman et al. 1986; Gorrell et al. 1992; Thormar 2005; Alkhatib & Berger 2007). Chemokine (C-C motif) Receptor 5 (CCR5) has been implicated as a source of genetic resistance to HIV. Individuals with the delta-32 deletion have no functional CCR5 protein on the cell surface and are highly resistant to HIV infection (Kaslow et al. 2005; Alkhatib & Berger 2007). Furthermore, other CCR5 sequence variants in regulatory regions have also been shown to influence CCR5 transcription, translation and acquired immune deficiency syndrome (AIDS) progression in humans (McDermott et al. 1998; Bream et al. 1999; Kostrikis et al. 1999; Kaslow et al. 2005; Jin et al. 2008). Sequence variants of ovine CCR5 could likewise impact OPPV infection in sheep.

As CCR5 may play a role in OPPV infection, this study determined the genomic sequence of sheep CCR5. Genetic variants were examined in a discovery group containing several different breeds, and the specific roles of the genetic variants were scrutinized for potential mechanisms by which they might influence OPPV. Further, association with proviral levels was examined in Idaho range sheep that were naturally exposed to OPPV. These initial steps have allowed the investigation of the role of CCR5 variants in OPPV infection in sheep.

Materials and methods

Results

A 13.3-kb ovine genomic sequence encompassing CCR5 and surrounding regulatory regions was produced from a Rambouillet ram (GenBank accession FJ008056.1). Ovine CCR5 had two exons with conserved boundaries, and the entire open reading frame was located in the second exon. The open reading frame encoded a CCR5 protein of 352 residues with 95.5% amino acid identity to bovine CCR5, 83.5% to human, 82.4% to canine and 79.0% to mouse. An alignment showing amino acid sequence differences between CCR5 from sheep and several other mammals is shown in Fig. S1.

Fifty-seven sequence variants were observed in 5.2 kb of the CCR5 gene region (dbSNP accessions ss105325111–ss105325167) in the marker discovery animal set. Of these, six were insertion–deletion variants and 51 were single nucleotide polymorphisms (SNPs). Four of the SNPs resulted in amino acid substitutions of leucine for serine at residue 38, phenylalanine for isoleucine at residue 263, alanine for serine at residue 331, and valine for alanine at residue 338 (Table 1). One deletion (g.5436_5439delAATG) eliminated four base pairs from an octamer (Oct) protein binding site in the promoter region at nucleotide positions 5436–5439 in GenBank accession FJ008056.1 (marker 3; Fig. 1). Comparison to cow, pig, horse, human and dog revealed that the insertion was the ancestral allele, with the deletion observed only in sheep.

To test whether the Oct site deletion interfered with transcription, CCR5 transcripts from peripheral blood leucocytes of nine heterozygous, OPPV- and MCFV-free animals were reverse transcribed, cloned and counted. A preliminary chi-squared test for homogeneity between the animals was not significant (P > 0.05), indicating that none of the animals had significantly differing relative transcription of the two alleles. Therefore, the pooled counts were used to estimate the proportion of transcripts from each allele. A significant 3.9-fold reduction (P < 0.0001; 95% CI: 3.3 to 4.6-fold) in transcription from the deletion allele was observed relative to the ancestral allele (Fig. 2).

To test the whole range of observed CCR5 variation for association with OPPV provirus levels, 16 sequence variants including the Oct site deletion and 15 SNPs were chosen by haplotype tagging to genotype the complete OPPV animal set (Table 1). The genotypes for all sequence variants were in agreement with Hardy–Weinberg proportions (P > 0.05). No sequence variants were associated with differential odds of infection (P > 0.05). However, one haplotype was associated with reduced OPPV levels among positive animals (P = 0.0072; Fig. 3). The haplotype included markers 3, 4, 15 and 16 in tight linkage disequilibrium with each other (r² > = 0.99; D′ > = 0.99 for all pairwise comparisons). The Oct site deletion allele occurred at a frequency of 33.3% in the overall animal set, with individual breed frequencies of 29.8% in Rambouillets, 25.4% in Polypays and 45.9% in Columbias. Out of the set of 351 animals, there were 36 homozygotes for the Oct site deletion, 162 heterozygotes and 153 homozygotes with the reference allele.

Discussion

Chemokine (C-C motif) Receptor 5 is an important gene in the immune response, and this study examined common sequence variants in the CCR5 gene region. Analyses of gene variants, their role in gene expression, and their association with control of OPPV will be discussed below. Because CCR5 impacts the recruitment of leucocytes to combat infection with diverse types of pathogens, the CCR5 variants described here may influence the outcomes of infections with many pathogens of sheep.

To our knowledge, this study provides the first finished genomic sequence for the ovine CCR5 gene region. The gene structure and similarity to other mammalian CCR5 genes were similar to a recently reported bovine CCR5 sequence (Blumerman et al. 2007). More than 50 individual sequence variants were observed in the open reading frame and surrounding regions. One variant was a 4-base pair deletion which removed a transcription factor binding site for Oct proteins (Fig. 1).

The Oct binding site disrupted by this deletion has been shown to be functionally important in the human CCR5 promoter, and it is conserved among mammals. The site is bound by at least two transcription factors, the more ubiquitous Oct-1 and the lymphoid-specific Oct-2 (Moriuchi et al. 1997). Binding by Oct-2 has been shown to upregulate CCR5 transcription and the ability of the protein to fuse with HIV-1 Env, a measure of cell surface CCR5 protein expression (Moriuchi & Moriuchi 2001). Further, there are Oct sites in the same region of the CCR5 promoter in at least seven other mammalian species, including cow, pig, horse, dog, rhesus macaque, chimpanzee and human. The sequence-level conservation in the promoter was lower than the open reading frame, as one would expect for transcription factor binding sites, but the consistent location of Oct sites in CCR5 promoters and the demonstrated biological importance suggested the deletion allele might have direct functional importance for sheep and may reduce expression from the sheep deletion allele.

As the sheep deletion removed a transcription factor binding site in the promoter, we tested relative expression from the ancestral vs. deletion haplotypes. In sheep heterozygous for the Oct site deletion, the deletion haplotype had 3.9-fold reduced expression (P < 0.0001; 95% CI 3.2 to 4.6-fold) compared to the ancestral haplotype (Fig. 2). This reduced expression from the sheep deletion allele was consistent with the observation that binding of this site by Oct-2 enhanced CCR5 expression (Moriuchi & Moriuchi 2001), and suggests that the deletion of the Oct site may be a functional mutation. The possibility exists that other sites on the same haplotype, but further outside CCR5, could also play a role as repressors that could function at a distance. However, the functional importance of the human CCR5 promoter, the conservation of Oct site in the CCR5 promoters, and the greatly reduced expression from the deletion haplotype collectively suggest that this deletion may play a functional role in CCR5 expression.

Several regulatory sequence variants have been shown to influence CCR5 and AIDS progression in humans. The most comparable to the sheep deletion reported here is 59029G, also known as −2459 A/G, which reduces CCR5 transcription by 45% and reduces HIV replication (McDermott et al. 1998; Salkowitz et al. 2003). In the homozygous form, 59029G/G delays the progression time until clinical AIDS by 3.8 years (McDermott et al. 1998). Neither 59029G nor any of the other regulatory sequence variants in human CCR5 change the same Oct site as the sheep deletion (McDermott et al. 1998; Bream et al. 1999; Kostrikis et al. 1999; Jin et al. 2008), but the presence of similar kinds of changes in both human and CCR5 suggest there may be similar effects on lentiviral infection.

Haplotype tagging markers were chosen to efficiently test for genotypic differences in OPPV proviral levels related to any of the variants observed (Table 1). Recent studies have shown that factors such as breed and age are important for OPPV (Herrmann-Hoesing et al. 2008b), so all analyses in this study accounted for those factors explicitly in the association model. After accounting for these factors, the Oct deletion haplotype was associated with reduced proviral levels among OPPV-positive sheep (Fig. 1). Homozygotes for this haplotype had approximately half the OPPV proviral levels compared to insertion homozygotes or heterozygotes, depending on the specific insertion genotype to which it was compared (P < 0.01; Fig. 3).

The lower proviral levels in mature sheep homozygous for the Oct site deletion suggest superior immune containment of the virus post-infection, but the exact mechanisms at work with OPPV remain to be elucidated. Reduced expression of CCR5 could result in reduced chemotaxis of macrophages and other leucocytes to the site of infection (Locati et al. 2002; Ma et al. 2005), which could also slow the rate of cellular infection relative to infected cell death. CCR5 could also function as a coreceptor for OPPV, as it does with HIV (Alkhatib & Berger 2007), and reduced coreceptor availability could slow the rate of cellular infection relative to the rate of infected cell death, which would result in lower proviral levels. Finally, reduced CCR5 expression could lead to changes in cell-mediated immunity that could enhance killing of infected cells (Dolan et al. 2007). The many functional roles of CCR5 suggest a combination of mechanisms as the most likely explanation, and further experiments will be necessary to elucidate the exact role of the Oct site deletion and reduced CCR5 expression in OPPV infection.

Future experiments could validate the Oct site deletion as a selectable genetic marker to reduce the proviral loads of OPPV. If so, the high allele frequency (33%) suggests that rapid genetic progress could be made toward deletion homozygotes that may have lower proviral levels if infected with OPPV. Further, the fact that the allele frequency is high in all breeds examined suggests that rapid genetic progress could be made in many breeds without the need for time-consuming introgression of the allele. However, the Oct site deletion will need to be validated before it can be recommended for wide use in marker-assisted selection.

Beyond its role in OPPV, the fundamental nature of CCR5’s roles in the immune system are demonstrated by the wide range of viral pathogens for which CCR5 variants have been shown to play large roles in resistance/resilience. CCR5 is a critical regulator of viral load and mortality for infections with such diverse infectious diseases as respiratory infection with influenza A and sexually transmitted infection with herpes simplex virus-2 (Dawson et al. 2000; Thapa et al. 2007). CCR5 is also important in inflammation and the development of symptomatic disease for West Nile virus and respiratory syncytial virus (Hull et al. 2003; Lim et al. 2008). Further, CCR5 variants have been associated with differences in odds of clearance vs. chronic infection with hepatitis B (Thio et al. 2007).

The important roles of CCR5 have also been highlighted in responses to bacteria, protozoa and parasitic worms. CCR5 has been associated with Mycoplasma pneumoniae infection in humans (Ungvari et al. 2007). It has also been shown to function as an innate immune receptor for mycobacterial Hsp70 (Floto et al. 2006). CCR5 has been associated with parasite burden of the parasitic protozoa Leishmania major, Trypanosoma cruzi and Toxoplasma gondii (Sato et al. 1999; Hardison et al. 2006; Khan et al. 2006). Finally, CCR5 knockout mice have recently been shown to have enhanced lethality and larger granuloma size when infected with the trematode worm Schistosoma mansoni (Souza et al. 2008), which suggests that CCR5 could play a role in resistance to parasitic worms in sheep. The large variety of pathogens for which CCR5 is important suggests that a sequence variant with large impact on CCR5 function in sheep could play roles in resistance/resilience to a wide range of sheep pathogens.

In conclusion, these results demonstrate a common deletion in the promoter region of ovine CCR5. The location of the deletion suggested differential transcription as a mechanism of action, and 3.9-fold reduced transcription was confirmed from the deletion allele. Homozygotes for the deletion had approximately half the proviral levels of OPPV compared to insertion homozygotes or heterozygotes. The large difference in CCR5 expression and the many immunological functions of CCR5 suggest that this deletion could have roles in resistance or resilience for many diverse pathogens of sheep.