Analysis of a Subacute Sclerosing Panencephalitis Genotype B3 Virus from the 2009-2010 South African Measles Epidemic Shows That Hyperfusogenic F Proteins Contribute to Measles Virus Infection in the Brain

Abstract

During a measles virus (MeV) epidemic in 2009 in South Africa, measles inclusion body encephalitis (MIBE) was identified in several HIV-infected patients. Years later, children are presenting with subacute sclerosing panencephalitis (SSPE). To investigate the features of established MeV neuronal infections, viral sequences were analyzed from brain tissue samples of a single SSPE case and compared with MIBE sequences previously obtained from patients infected during the same epidemic. Both the SSPE and the MIBE viruses had amino acid substitutions in the ectodomain of the F protein that confer enhanced fusion properties. Functional analysis of the fusion complexes confirmed that both MIBE and SSPE F protein mutations promoted fusion with less dependence on interaction by the viral receptor-binding protein with known MeV receptors. While the SSPE F required the presence of a homotypic attachment protein, MeV H, in order to fuse, MIBE F did not. Both F proteins had decreased thermal stability compared to that of the corresponding wild-type F protein. Finally, recombinant viruses expressing MIBE or SSPE fusion complexes spread in the absence of known MeV receptors, with MIBE F-bearing viruses causing large syncytia in these cells. Our results suggest that alterations to the MeV fusion complex that promote fusion and cell-to-cell spread in the absence of known MeV receptors is a key property for infection of the brain.IMPORTANCE Measles virus can invade the central nervous system (CNS) and cause severe neurological complications, such as MIBE and SSPE. However, mechanisms by which MeV enters the CNS and triggers the disease remain unclear. We analyzed viruses from brain tissue of individuals with MIBE or SSPE, infected during the same epidemic, after the onset of neurological disease. Our findings indicate that the emergence of hyperfusogenic MeV F proteins is associated with infection of the brain. We also demonstrate that hyperfusogenic F proteins permit MeV to enter cells and spread without the need to engage nectin-4 or CD150, known receptors for MeV that are not present on neural cells.

Keywords: central nervous system infections; measles; viral fusion.

FIG 1

Phylogenetic tree of H gene (nucleotide positions 7496 to 9124) generated by maximum likelihood analysis of MeV from the SSPE case, patients with MIBE, patients who had acute measles infection during the measles epidemic of 2009-2010 in South Africa, and reference sequences. Reference sequences of other MeV genotype B3.1 viruses were retrieved from the NCBI GenBank database and are indicated by accession numbers. Bootstrap values greater than 75% are indicated at the nodes of the tree. The branch lengths are proportional to the evolutionary distances, as shown on the scale.

FIG 2

Location of substitutions within the F protein from CNS-adapted virus. (A) Schematic representation of fusion protein with relevant regions indicated. FP, fusion peptide; S-S, disulfide bond; HRN and HRC, heptad repeat domains at the N terminus or at the C terminus; TM, transmembrane domain; CTD, cytoplasmic domain. (B) Ribbon diagrams of the measles F model structure in prefusion conformation. The F SSPE clinical sample had a total of 6 mutations compared to sequence of the wild-type B3 strain. Residues G168G and E170G map in the HRN domain, while A440P maps close to the HRC. The residue S262G is in the region between HRN and HRC. Residues R520C and L550P are in the CTD region. One additional mutation abolished the F stop codon and resulted in a 29-amino-acid extension of the CTD (long tail). The SSPE residues are indicated in magenta. Additionally, three substitutions (L454W, T461I, and N462K) localized in the HRC domain, and previously described (6, 7) in neuropathogenic strains, are shown in green.

FIG 3

Functional analysis of F/H protein pairs from MeV B3 SSPE clinical sample. The cell-to-cell fusion of 293T cells coexpressing the indicated MeV B3 F protein and MeV B3 H WT (white bars) or MeV B3 H SSPE (black bars) with 293T cells transfected with MeV receptor nectin-4 (A and D), CD150 (B and E), or with an empty vector (C and F) was assessed by a β-Gal complementation assay. The values are expressed as the relative luminescence unit (RLU) averages (with SEM) of results from three independent experiments in triplicate. **, P < 0.01; ***, P < 0.001 (two-way ANOVA and Fisher’s post hoc test). WT, wild-type F; WT-LT, wild-type F with long cytoplasmic tail; SSPE, F bearing 5 mutated amino acids (G168R, E170G, S262G, A440P, R520C, and L550P) and the ablation of the stop codon that leads to a longer cytoplasmic tail. Levels of expression of the different proteins were comparable (data not shown).

FIG 4

Multiply mutated SSPE B3 F has higher fusion activity than any individually mutated F in the absence of MeV host cell receptors. HEK-293T cells were cotransfected with the indicated F protein, α-subunit of β-galactosidase, and wild-type B3 MeV H (A) or B3 MeV SSPE H (B). Transfected cells were subsequently overlain with HEK-293T cells expressing the ω-subunit of β-galactosidase for 6 h to allow fusion. Resulting luminescence from β-galactosidase activity was quantified using TECAN Infinite M1000 Pro. Results depict a representative experiment with means and SEM from three biological replicates. Activity of mutants was compared with wild-type B3 F activity by one-way ANOVA. Fusion activities significantly different (P < 0.05) from that of wild-type B3 F are indicated by an asterisk(s).

FIG 5

Regulation of fusion promotion by receptor binding protein. The cell-to-cell fusion of 293T cells coexpressing uncleaved influenza virus (HA), the indicated F proteins (x axis), and MeV H WT (A), MeV H SSPE (B), or the empty vector pCAGGS (C) was assessed by a β-Gal complementation assay using target cells that do not express MeV H receptor. The values are expressed as the mean RLU (with SEM) of results from three independent experiments in triplicate. ***, P < 0.001 (two-tailed, unpaired t test).

FIG 6

Thermal stability of the MeV wild-type and mutant F proteins. 293T cells expressing the indicated MeV F proteins (x axis) were incubated overnight at 37°C. The cells were then incubated for the indicated times at 55°C and then at 4°C with mouse MAbs recognizing the prefusion conformation of MeV F (21). The values on the y axis represent the percentages of conformational antibody binding and indicate the averages (with SEM) of results from four independent experiments in triplicate. WT, wild-type F; SSPE, F bearing G168R, E170G, S262G, A440P, R520C, and L550P. The L454W F (found in two MIBE clinical samples) (7) was added for comparison.

FIG 7

MeV recombinant viruses spread in the absence of known receptor. (A) Vero cells (without known MeV receptor) were infected with the indicated recombinant viruses expressing EGFP and incubated at 37°C. Infected cells were detected 24 and 72 h postinfection using an epifluorescence Nikon TS2R-FL inverted microscope. (B) Areas of infection, in pixels, were measured using ImageJ software on images randomly acquired from three separate experiments (*, P < 0.05; **, P < 0.01; ***, P < 0.001; each by Mann-Whitney U test, n = 8 at least).