Triggering of human parainfluenza virus 3 fusion protein (F) by the hemagglutinin-neuraminidase (HN) protein: an HN mutation diminishes the rate of F activation and fusion

Abstract

For human parainfluenza virus type 3 and many other paramyxoviruses, membrane fusion mediated by the fusion protein (F) has a stringent requirement for the presence of the homotypic hemagglutinin-neuraminidase protein (HN). With the goal of gaining further insight into the role of HN in the fusion process, we developed a simple method for quantitative comparison of the ability of wild-type and variant HNs to activate F. In this method, HN/F-coexpressing cells with red blood cells (RBC) bound to them at 4 degrees C are transferred to 22 degrees C, and at different times after transfer 4-guanidino-neu5Ac2en (4-GU-DANA) is added; this inhibitor of the HN-receptor interaction then releases all reversibly bound RBC but not those in which F insertion in the target membrane or fusion has occurred. Thus, the amount of irreversibly bound (nonreleased) RBC provides a measure of F activation, and the use of fluorescently labeled RBC permits microscopic assessment of the extent to which F insertion has progressed to fusion. We studied two neuraminidase-deficient HN variants, C28a, which has two mutations, P111S and D216N, and C28, which possesses the D216N mutation only. C28a but not C28 exhibits a slow fusion phenotype, although determination of the HNs' receptor-binding avidity (with our sensitive method, employing RBC with different degrees of receptor depletion) showed that the receptor-binding avidity of C28a or C28 HN was not lower than that of the wild type. The F activation assay, however, revealed fusion-triggering defects in C28a HN. After 10 and also 20 min at 22 degrees C, irreversible RBC binding was significantly less for cells coexpressing wild-type F with C28a HN than for cells coexpressing wild-type F with wild-type HN. In addition, F insertion progressed to fusion more slowly in the case of C28a HN-expressing cells than of wild-type HN-expressing cells. Identical defects were found for P111S HN, whereas for C28 HN, representing the 216 mutation of C28a, F activation and fusion were as rapid as for wild-type HN. The diminished fusion promotion capacity of C28a HN is therefore attributable to P111S, a mutation in the stalk region of the molecule that causes no decrease in receptor-binding avidity. C28a HN is the first parainfluenza virus variant found so far to be specifically defective in HN's F-triggering and fusion promotion functions and may contribute to our understanding of transmission of the activating signal from HN to F.

FIG. 1.

Plaque enlargement of C28a, C28, and wild-type (wt) viruses. The generation of plaques and determination of their areas at the indicated times (abscissa) after infection were done as described in Materials and Methods. The values for plaque areas (ordinate) are means ± standard deviation of results for 10 to 20 plaques.

FIG. 2.

Release of wild-type, C28a, C28, and P111S HN from receptor. RBC were bound to cells expressing wild-type (wt), C28a, C28, or P111S HN at 4°C. The cells were then transferred to 37°C, and RBC release was quantified as described in Materials and Methods at the indicated times (minutes) (abscissa). The amount of released RBC is given as a percentage of total bound RBC (ordinate).

FIG. 3.

Quantitation of HN receptor binding for wild-type and variants. RBC with different degrees of receptor depletion were prepared by treatment with various amounts of bacterial neuraminidase (see abscissa) as described in Materials and Methods. Aliquots of these and control (undepleted) RBC preparations were used to quantify HAD on cell monolayers transfected with constructs expressing wild-type (WT, open circles), C28 (open squares), C28a (open triangles), P111S (open diamonds), or ZM1 (hash mark) HN. The extent of binding of each of the “depleted” RBC is expressed (ordinate) as a percentage of the control (i.e., of the amount of untreated, nondepleted RBC bound on cells expressing the corresponding HN). The points are means of results for triplicate monolayers, with bars denoting standard deviation.

FIG. 4.

Release by 4-GU-DANA of RBC bound to HN-expressing cells. RBC were bound to cells expressing wild-type (wt), C28a, C28, or P111S HN at 4°C. The cells were then placed at 22°C or 37°C in the presence of 10 mM 4-GU-DANA; released RBC, quantitated 15 min later, are expressed as a percentage of total bound RBC.

FIG. 5.

F activation mediated by wild-type (wt) HN compared with that mediated by C28a HN, C28 HN, and P111S HN. The experimental strategy is described in the text. Cells expressing the indicated HNs and transfected with F were allowed to adsorb RBC at 4°C. They were then transferred to 22°C, and 4-GU-DANA was added either before (0 min) or 10, 20, or 60 min after transfer to 22°C. (A) The amount of RBC that remained bound 15 min after the addition of 4-GU-DANA was determined. These amounts, expressed as a percentage of total bound RBC (ordinate), are shown as a function of time at 22°C prior to the addition of 4-GU-DANA (abscissa). (B) Photographs were taken under a fluorescence microscope of the R18-labeled RBC on cell monolayers at the same time points as those shown in panel A.

FIG. 5.

F activation mediated by wild-type (wt) HN compared with that mediated by C28a HN, C28 HN, and P111S HN. The experimental strategy is described in the text. Cells expressing the indicated HNs and transfected with F were allowed to adsorb RBC at 4°C. They were then transferred to 22°C, and 4-GU-DANA was added either before (0 min) or 10, 20, or 60 min after transfer to 22°C. (A) The amount of RBC that remained bound 15 min after the addition of 4-GU-DANA was determined. These amounts, expressed as a percentage of total bound RBC (ordinate), are shown as a function of time at 22°C prior to the addition of 4-GU-DANA (abscissa). (B) Photographs were taken under a fluorescence microscope of the R18-labeled RBC on cell monolayers at the same time points as those shown in panel A.