Identification of the receptor binding domain of the mouse mammary tumor virus envelope protein

Abstract

Mouse mammary tumor virus (MMTV) is a betaretrovirus that infects rodent cells and uses mouse transferrin receptor 1 for cell entry. To characterize the interaction of MMTV with its receptor, we aligned the MMTV envelope surface (SU) protein with that of Friend murine leukemia virus (F-MLV) and identified a putative receptor-binding domain (RBD) that included a receptor binding sequence (RBS) of five amino acids and a heparin-binding domain (HBD). Mutation of the HBD reduced virus infectivity, and soluble heparan sulfate blocked infection of cells by wild-type pseudovirus. Interestingly, some but not all MMTV-like elements found in primary and cultured human breast cancer cell lines, termed h-MTVs, had sequence alterations in the putative RBS. Single substitution of one of the amino acids found in an h-MTV RBS variant in the RBD of MMTV, Phe(40) to Ser, did not alter species tropism but abolished both virus binding to cells and infectivity. Neutralizing anti-SU monoclonal antibodies also recognized a glutathione S-transferase fusion protein that contained the five-amino-acid RBS region from MMTV. The critical Phe(40) residue is located on a surface of the MMTV RBD model that is distant from and may be structurally more rigid than the region of F-MLV RBD that contains its critical binding site residues. This suggests that, in contrast to other murine retroviruses, binding to its receptor may result in few or no changes in MMTV envelope protein conformation.

FIG. 1.

Structural alignment of MMTV and F-MLV envelope proteins. The β-strands, α-helices, and 3₁₀ helical turns identified in the F-MLV crystal structure are underlined and numbered according to reference . The variable regions that change with the tropism of the different MLVs (VR regions) are marked by dotted lines above the sequences. The putative MMTV RBS is shaded, and the HBDs in MMTV and F-MLV are boxed. Amino acids in F-MLV and MMTV identified by mutation analysis to be critical to receptor interaction are in bold (1, 10, 30). N-linked glycosylation sites are in italics. An arrowhead marks the Arg codon at position 246 used to define the end of the proline-rich region in the MMTV SU.

FIG. 2.

Three-dimensional model of MMTV and F-MLV SU proteins. (A) Structure of the F-MLV RBD depicted from the crystal coordinates (Protein Data Base 1AOL) (15). (B) Model of the MMTV RBD generated with SwissModel. (C) Modified model illustrating potential disulfide bonds between cysteines 62 and 73 and cysteines 133 and 156 of MMTV. The five amino acids constituting the putative RBS of MMTV (residues 40 to 44) and the amino acid residues identified by mutation analysis to be involved in the F-MLV SU-receptor interaction (see Fig. 1) are shown as space-filled red atoms. The black arrows indicate the N-linked glycosylation sites in F-MLV and putative sites in MMTV, which are depicted as space-filled green atoms. The cysteine residues with the potential for disulfide bonding are shown as space-filled yellow atoms; the red and white arrows point to cysteine residues with the potential for stabilizing the putative VRA loop and VRC loops with disulfide bonds, respectively. Abbreviations: N, amino terminus; C, carboxyl terminus; HBD, heparin-binding domain. The structures in panels A and B were depicted with RasMol 2.7.1.1 (5, 41). The diagram in panel C was drawn from the RasMol depiction in panel B.

FIG. 3.

Amino acid sequence comparison of the SUs of MMTV(C3H), MMTV(RIII), and h-MTVs found in human tissues and cells. Boldfaced amino acid designations indicate changes unique to h-MTVs relative to all known exogenous and endogenous MMTVs. The boxed sequences show the amino acids used to make the RBS-GST fusion protein.

FIG. 4.

Heparan binding enhances MMTV infection. (A) ΔHBD infection levels. Equal amounts of ΔHBD and wild-type (WT) pEnv_C3H pseudovirus were used for infection. The titer for each virus was calculated and is presented as a percentage of wild-type infection levels. The data represent the averages of five independently performed experiments. (B) Virion proteins in MMTV pseudoviruses. Supernatants from equal numbers of 293T cells cotransfected with pENV_C3H or ΔHBD, pHIT111, and pHIT60 were pelleted by ultracentrifugation through 30% sucrose. Equal volumes of the resuspended pellets (supernatant) or the transfected cell extracts (intracellular) were subjected to SDS-PAGE followed by Western blot analysis with anti-SU antiserum. The arrow shows the SU protein (gp52); the upper band in the extracts is the unprocessed polyprotein. Abbreviations: M, mock infected; V, purified virus. (C) Heparan sulfate treatment. Cells were treated with heparan sulfate as described in Materials and Methods. Data are presented as a percentage of wild-type infection levels without heparan sulfate and are the averages of three independent experiments. Solid bars, wild-type pseudovirus; open bars, ΔHBD.

FIG. 5.

Mutation of a single amino acid in the RBD abolishes infectivity. (Top panel) Virion proteins in MMTV pseudoviruses. Supernatants from equal numbers of 293T cells cotransfected with pENV_C3H or the mutant pENVs (Ser₄₀, Ala₄₀, and Tyr₄₀), pHIT111, and pHIT60 (MLV gag and genome) were pelleted by ultracentrifugation. Then 50 μl of the resuspended pellets was subjected to SDS-PAGE followed by Western blotting analysis with anti-MMTV antiserum. The arrows point to SU (gp52) and TM (gp36). (Bottom panel) Pseudovirus infection of NMuMG cells. Triplicate infections were performed with equal amounts of supernatant. The data are presented as the titer and the standard deviation for each infection.

FIG. 6.

Phe residue at position 40 is required for high-level virus binding to NMuMG cells. Equal amounts of MMTV pseudotypes prepared with either the wild-type (WT) envelope or envelope containing the Ser₄₀ mutation (inset, panel A) were incubated on ice with NMuMG cells. The cells were stained with anti-MMTV antiserum, followed by FITC-labeled secondary antibodies, and subjected to FACS analysis. Dead cells were excluded by their forward scatter/side scatter properties. The experiment was performed three to four times with similar results; a representative experiment is shown. (A) Histograms showing the fluorescence of cells incubated with wild-type (thick line) or Ser₄₀ mutant virus (thin line) or without virus (dotted line). The inset shows a Western blot of 10 μl of each concentrated virus preparation or milk-borne MMTV (MMTV) as a positive control, with polyclonal anti-MMTV SU antiserum used for detection. NV, no virus. (B) Histograms showing the fluorescence of cells incubated with wild-type or mutant virus in cells pretreated with 100 μg of heparan sulfate per ml.

FIG. 7.

Wild-type but not mutant MMTV blocks surface expression of the transferrin receptor. Cells were incubated with virus by spinoculation and then stained with FITC-labeled anti-mouse TfR antibodies. Each experiment was performed four to five times; shown is the MCF value ± standard deviation. Abbreviations: 293T, untransfected 293T cells incubated with MMTV(C3H) pseudotypes; TRH3/no virus, clonal isolate of 293T stably transfected with mTfR1; wt, TRH3 cells incubated with MMTV(C3H)-pseudotyped virus; Ser₄₀, TRH3 cells incubated with the Ser₄₀ mutant; wt/αMMTV, TRH3 cells incubated with MMTV(C3H) pretreated with goat anti-MMTV polyclonal antiserum; αTfR, TRH3 cells incubated with anti-mouse TfR monoclonal antibody and MMTV(C3H). P values were calculated with Student's t test. *, P ≤ 0.05; **, P ≤ 0.005 compared to TRH3 cells not bound to virus.

FIG. 8.

Blocking monoclonal antibodies bind to the RBS. Equal amounts of GST alone (lane 1), the RBS-GST fusion protein (lane 2), or extract from pEnv_C3H-transfected 293T cells (lane 3) were subjected to SDS-PAGE followed by Western blot analysis with monoclonal Black 6, Black 6-5D, polyclonal anti-MMTV SU, or anti-GST antibody.