The secondary structure of the 5' end of the FIV genome reveals a long-range interaction between R/U5 and gag sequences, and a large, stable stem-loop

Abstract

Feline immunodeficiency virus (FIV) is a lentivirus that infects cats and is related to human immunodeficiency virus (HIV). Although it is a common worldwide infection, and has potential uses as a human gene therapy vector and as a nonprimate model for HIV infection, little detail is known of the viral life cycle. Previous experiments have shown that its packaging signal includes two or more regions within the first 511 nucleotides of the genomic RNA. We have undertaken a secondary structural analysis of this RNA by minimal free-energy structural prediction, biochemical mapping, and phylogenetic analysis, and show that it contains five conserved stem-loops and a conserved long-range interaction between heptanucleotide sequences 5'-CCCUGUC-3' in R/U5 and 5'-GACAGGG-3' in gag. This long-range interaction is similar to that seen in primate lentiviruses where it is thought to be functionally important. Along with strains that infect domestic cats, this heptanucleotide interaction can also occur in species-specific FIV strains that infect pumas, lions, and Pallas' cats where the heptanucleotide sequences involved vary. We have analyzed spliced and genomic FIV RNAs and see little structural change or sequence conservation within single-stranded regions of the 5' UTR that are important for viral packaging, suggesting that FIV may employ a cotranslational packaging mechanism.

FIGURE 1.

Minimal free-energy models of the FIV packaging signal RNA. Models were generated by the Mfold server using FIV Petaluma sequence RNA nt 1–511. The two most stable structures are shown. (A) δG = −180.83 kcal/mol; (B) δG = −181.60 kcal/mol.

FIGURE 2.

Biochemical analysis of the FIV packaging signal region RNA. FIV Petaluma strain nt 1–511 were transcribed in vitro and were digested with RNases or chemicals specific for ss (A, T1, CL3, U2, kethoxal) or ds (CV1) nucleotides. Primers were annealed to the RNA and cDNAs were made with AMV RT, incorporating ³³P-dATP. Samples were separated on 10% polyacrylamide gels, alongside cycle sequencing ladders. Sequence is given in the genome sense for clarity. Different primers were used to examine different regions of the RNA. These regions were: A, B: R/U5; C, D: SL1 and PBS; E, F, G, H, I: SL2; J, K: SL3, SL4, gag; L, M: gag. Numbers represent FIV nucleotide numbers shown in Figure 3. Various concentrations of each RNase/chemical or incubation time were used; only lanes where RNAs were minimally cleaved are shown. These were: A, RNase T1 0.2/1 U; B, RNase CL3 1 U and RNase CV1 2 min; C, RNase CL3 0.1/1 U, RNase U2 0.1 U; D RNase A 0.2/2 pg, RNase CV1 30 sec/2 min; E, RNase T1 0.1 U, CV1 30 sec; F, RNase A 0.2/2 pg; G, RNase CV1 5 sec; H, RNase CL3 1 U, RNase U2 0.01/0.1/ 0.5 U; I, kethoxal 200 μg; J, RNase CL3 0.1/0.2 U, RNase U2 0.1/0.2 U; K, RNase CV1 2 min; L, RNase T1 0.1 U, RNase A 0.2 pg; M, RNase CV1 30 sec/2 min. RNase cleavage sites are indicated with arrows (ss nucleases, colored light to dark gray in the order T1, CL3, A, U2) or diamonds (CV1).

FIGURE 3.

Secondary structural model of the FIV packaging signal RNA. Model generated based on minimal free energy as determined by Mfold, and biochemical mapping constraints. Arrows represent cleavage sites for ss RNases A, T1, CL3, or U2 or kethoxal. Diamonds represent CV1 cleavage sites. White, gray, and black indicate cleavage sites shown in Figure 2 and observed in multiple independent experiments (shades of gray represent different RNases). Yellow represents cleavage sites not shown in Figure 2 but observed in multiple independent experiments. Orange shows cleavage sites shown in Figure 2 but not always observed in repeat experiments.

FIGURE 4.

Sequence conservation and base-pair covariation in FIV leader RNA. (A) Secondary structural model of the 5′ end of the FIV Petaluma strain RNA showing sequence conservation between 77 isolates (see Supplemental Fig. 1 for details). Nucleotides in black are 100% conserved, those in gray are >95% conserved, those in orange are 75%–95% conserved, and those in pink are <75% conserved between isolates. Filled circles indicate nucleotides which vary in sequence between isolates but form a Watson-Crick or G–U pair in >95% of mutated isolates, indicating structural conservation. Open circles on SL4 indicate structural conservation so that one base-pair and one internal loop are always present. Black triangles indicate base pairs where both nucleotides are substituted in the same isolate and canonical base pairing is maintained. Structural conservation in SL1 is not shown in A. (B) Sequence alignment of the FIV SL1 sequence. The Petaluma SL1 sequence is shown at the top, and sequences of SL1 variant isolates are shown below alongside the accession numbers. Pink represents unpaired nucleotides. Black indicates nucleotides able to form a base pair between the left-hand and the right-hand sequence.

FIGURE 5.

SL1, SL2, and the LRI sequence is conserved among FIV strains that infect other feline species. (A) Mfold structural prediction of the 5′ end of the Cougar lentivirus genome (accession number EF455603). (B) Mfold structural prediction of the 5′ end of the Cougar lentivirus genome with LRI paired. (C) Mfold structural prediction of the 5′ end of the FIVPleo genome. (D) Mfold structural prediction of the 5′ end of the FIVOma genome. (E) Sequence alignment of the 5′ end of FIV strains that infect cougars, showing conservation of LRI and palindromic regions. (F) Sequence alignment of the 5′ end of FIV strains that infect Pallas' cats (AY713445 and U56928) and lions (EU117991 and EU117992), showing conservation of LRI and palindromic regions.

FIGURE 6.

SLs 1 and 2 are present within spliced and unspliced RNA. Mfold structural predictions of FIV Petaluma 5′ UTR spliced to orf A (A) and vif (B). (C) Biochemical analysis of the SL2 region of “spliced” (nucleotides 1–381) versus “unspliced” (nucleotides 1–511) RNA. In vitro transcribed RNAs were digested with 0 U (−) or 0.2 U (+) RNase U2 and primer extension was performed, incorporating ³³P-dATP and using a primer that annealed 3′of SL2. cDNAs were separated on a 10% polyacrylamide gel alongside a sequencing ladder that was generated using the same primer. DNA sequence is given in the genome sense for clarity.