Rotavirus NSP4 interacts with both the amino- and carboxyl-termini of caveolin-1

Abstract

Rotavirus NSP4 plays multiple roles in viral pathogenesis, morphogenesis and replication. We previously reported a direct interaction between full-length NSP4 and the enterotoxic peptide composed of NSP4 residues 114-135 with full-length caveolin-1, the structural protein of caveolae. Caveolin-1 forms a hairpin loop in the cytoplasmic leaflet of plasma membrane caveolae. This unique orientation results in both termini of caveolin-1 exposed to the cytoplasm. The goal of this study was to map the caveolin-1 residues that interact with NSP4 to obtain a more complete picture of this binding event. Utilizing reverse yeast two-hybrid analyses and direct peptide binding assays, the NSP4 binding site was localized to caveolin-1 residues 2-22 and 161-178, at the amino- and carboxyl-termini, respectively. However, NSP4 binding to one of the termini was sufficient for the interaction.

Figure 1

Linear schematic of full-length (FL) and deletion mutants of caveolin-1. The N-terminus, C-terminus, caveolin-1 scaffolding domain (CSD, aa 80–100), hydrophobic domain that traverses the cytofacial leaflet of the plasma membrane (aa 102–134), and the oligomerization domain (60–100) were deleted to determine their role in binding NSP4.

Figure 2

Western blot analyses of NSP4_80–140-, and full-length- and mutant caveolin-1-GAL4 fusion proteins expressed in MaV203 strain. Yeast were co-transformed with each of the fusion protein pairs and the lysates probed with NSP4- or caveolin-1-specific antibodies. Panel A. Untransformed and yeast transformed with the Cav_60–178-GAL4 fusion protein were probed with rabbit antibodies to Cav_65–78. A ~65 kD dimer of the Cav_60–178-GAL4 fusion protein was detected (lane 2); no reactivity was seen in the untransformed yeast lysates (lane 1). Panel B. The other caveolin-1 mutant-Gal4 fusion proteins were similarly tested for expression by Western blot analyses. Rabbit antibodies to Cav_2–22 were utilized to detect the Gal4-full length and -caveolin-1 deletion mutants in which the N-terminus remained intact. The full-length caveolin-1 fusion protein (lane 7) and all of the Cav-1 mutant-GAL4 fusion proteins (lanes 4–6) were detected as dimers. No band was observed in the untransformed yeast (lane 3). Panel C. To verify NSP4_80–140 fusion expression, Western blot analysis similarly was utilized to detect the NSP4 fusion protein in the same yeast lysates. The membrane was probed with anti-NSP4_150–175 and the expected Mr of the NSP4 fusion protein was seen (lane 9) that was absent from untransformed yeast (lane 8). Panel D. Western blot analysis of co-transformed yeast lysates showing the Cav_60–156 fusion protein was expressed as a monomer (~27 kD) and dimer (~54 kD) (lane 11). Panel E. The same co-transformed lysates analyzed in panel D were probed with anti-NSP4_114–135 and showed the expression of the NSP4 fusion protein (lane 13). The positive control (lane 14) was RV infected cell lysates and the negative control (lane 12) was untransformed yeast lysates.

Figure 3

The NSP4_114–135 peptide binding assay confirms the yeast two-hybrid data. NSP4_114–135 was coupled to CnBr-activated sepharose beads and reacted with the transformed InVSc1 yeast lysates expressing full-length or mutant caveolin-1. Bound proteins were pelleted, washed and examined by Western blot analyses. Panel A. Yeast were transformed with an empty plasmid (InVsC1) (lanes 1, 7, 13) or the indicated caveolin-1 mutants. Yeast lysates were reacted with NSP4_114–135-bound sepharose beads (lanes 1–6), beads only (lanes 7–12), or lysates only were added directly to the gel to control for the presence of the mutant proteins (lanes 13–18). The bound proteins were separated by SDS-PAGE, transferred to nitrocellulose, and the membrane was probed with rabbit anti-Cav_2–22. To ensure antibody reactivity, MDCK lysates that normally express caveolin-1 were also tested in the binding assay (lanes 3, 9, 15). Panel B. Western blot analyses of the Cav_60–175 reacted with NSP4_114–135-bound beads. Antibodies to Cav_68–75 were utilized for detection of the caveolin-1 mutant. Lysates only (lane 1), reactivity with beads only (lane 2) and reactivity of Cav_60–175 reacted with NSP4_114–135-bound beads (lane 3) are shown. Panel C. The binding experiments were repeated several times with the same results. The blot in panel C shows the multimeric forms of two of the mutant caveolin proteins that were bound by NSP4_144–135-linked beads. InVSC1 only served as the negative control and MDCK lysates were used as a positive control..

Figure 3

The NSP4_114–135 peptide binding assay confirms the yeast two-hybrid data. NSP4_114–135 was coupled to CnBr-activated sepharose beads and reacted with the transformed InVSc1 yeast lysates expressing full-length or mutant caveolin-1. Bound proteins were pelleted, washed and examined by Western blot analyses. Panel A. Yeast were transformed with an empty plasmid (InVsC1) (lanes 1, 7, 13) or the indicated caveolin-1 mutants. Yeast lysates were reacted with NSP4_114–135-bound sepharose beads (lanes 1–6), beads only (lanes 7–12), or lysates only were added directly to the gel to control for the presence of the mutant proteins (lanes 13–18). The bound proteins were separated by SDS-PAGE, transferred to nitrocellulose, and the membrane was probed with rabbit anti-Cav_2–22. To ensure antibody reactivity, MDCK lysates that normally express caveolin-1 were also tested in the binding assay (lanes 3, 9, 15). Panel B. Western blot analyses of the Cav_60–175 reacted with NSP4_114–135-bound beads. Antibodies to Cav_68–75 were utilized for detection of the caveolin-1 mutant. Lysates only (lane 1), reactivity with beads only (lane 2) and reactivity of Cav_60–175 reacted with NSP4_114–135-bound beads (lane 3) are shown. Panel C. The binding experiments were repeated several times with the same results. The blot in panel C shows the multimeric forms of two of the mutant caveolin proteins that were bound by NSP4_144–135-linked beads. InVSC1 only served as the negative control and MDCK lysates were used as a positive control..

Figure 4

NSP4 is captured by peptides corresponding to the N- and C- termini of caveolin-1, but not by a peptide corresponding to the central region of caveolin-1. Untransformed yeast (lanes 1, 4, 7, 10, 13) and NSP4_HydroMut mutant that has been shown to not bind caveolin-1 (Parr, et al., 2006) (lanes 3, 6, 9, 11) served as the negative controls. Yeast lysates only showed both proteins were expressed (lanes 14–15). The N-terminal, Cav_2–22, and the C-terminal, Cav_161–178, peptides both captured NSP4 from yeast lysates expressing full-length NSP4 (lanes 2 and 5), whereas the peptide corresponding to the centrally located caveolin-1 residues, Cav_76–101, failed to capture NSP4 (lane 8).