Interaction of parvovirus B19 with human erythrocytes alters virus structure and cell membrane integrity

Abstract

The unique region of the capsid protein VP1 (VP1u) of B19 virus (B19V) elicits a dominant immune response and has a phospholipase A(2) (PLA(2)) activity required for the infection. Despite these properties, we have observed that the VP1u-PLA(2) motif occupies an internal position in the capsid. However, brief exposure to increasing temperatures induced a progressive accessibility of the PLA(2) motif as well as a proportional increase of the PLA(2) activity. Similarly, upon binding on human red blood cells (RBCs), a proportion of the capsids externalized the VP1u-PLA(2) motif. Incubation of B19V with RBCs from 17 healthy donors resulted in extensive virus attachment ranging between 3,000 and 30,000 virions per cell. B19V empty capsids represent an important fraction of the viral particles circulating in the blood (30 to 40%) and bind to RBCs in the same way as full capsids. The extensive B19V binding to RBCs did not cause direct hemolysis but an increased osmotic fragility of the cells by a mechanism involving the PLA(2) activity of the exposed VP1u. Analysis of a blood sample from an individual with a recent B19V infection revealed that, at this particular moment of the infection, the virions circulating in the blood were mostly associated to the RBC fraction. However, the RBC-bound B19V was not able to infect susceptible cells. These observations indicate that RBCs play a significant role during B19V infection by triggering the exposure of the immunodominant VP1u including its PLA(2) constituent. On the other hand, the early exposure of VP1u might facilitate viral internalization and/or uncoating in target cells.

FIG. 1.

Accessibility of the PLA₂ region of B19V VP1u. (A) Schematic representation of the VP1/VP2 proteins of B19V. Filled circles indicate the calcium binding domain, and filled squares indicate the PLA₂ enzymatic core. The sequence of the peptide used for the immunization of rabbits is indicated. (B) Accessibility of the VP1u-PLA₂ motif in native B19V capsids. B19V DNA was quantified following immunoprecipitation with an antibody against capsids (α-caps; MAb 860-55D) or an antibody against the PLA₂ region (α-PLA₂; VP1u 142-163). (C) Accessibility of the VP1u-PLA₂ region after heat treatment. B19V capsids were treated at increasing temperatures for 3 min, cooled on ice, and immunoprecipitated with MAb 860-55D (α-caps) or VP1u 142-163 (α-PLA₂). The 75-kDa band corresponds to an LC-HC dimer resulting from partial denaturation of the antibody (band present without virus). (D) Activation of B19V-PLA₂ enzyme by heat treatment. PLA₂ activity of B19V capsids was measured following exposure to increasing temperatures for 3 min. Error bars represent the deviation from three independent experiments.

FIG. 2.

Proportion of FC and EC from an infected plasma sample. Dot blot hybridizations from CsCl fractions (A) and from dilutions of the pooled fractions (B) are shown. (C) Quantification of B19V-DNA from FC and EC pool fractions. (D) Detection of FC and EC bound to RBCs by immunofluorescence. (E) RBCs were incubated with EC (5 × 10⁴ particles per cell). After 1 h at 4°C, the cells were washed, and B19V was added and further incubated at 4°C for 2 h. Following intensive washing to remove unbound virus, the amount of B19V associated to RBCs was quantified by real-time PCR. Deviations from three independent experiments are indicated.

FIG. 3.

(A) Quantification of B19V binding to RBCs from different donors. B19V was incubated with RBCs from 17 healthy donors. Following washings to remove unbound virus, the amount of virus bound per cell was quantified by real-time PCR. (B) Effect of pretreatment of RBCs with an antibody specific for globoside on B19V binding. RBCs where either untreated (PBS) or treated with anti-globoside IgM (AME-2) or a similar concentration of a nonspecific mouse IgM isotype control before the addition of B19V (2.5 × 10⁴ virions per cell). Deviations from three independent experiments are indicated.

FIG. 4.

VP1u-PLA₂ is exposed on the capsid surface following attachment to P antigen on human RBCs. (A) Following binding to RBCs at 4°C, the supernatant (unbound virus) and cellular fractions (bound virus) were incubated in lysis buffer and centrifuged at 10,000 × g for 10 min. The viral capsids present in the supernatants were immunoprecipitated with a MAb against assembled capsids (α-caps; 860-55D) and a polyclonal antibody against the VP1u-PLA₂ motif (α-PLA₂; VP1u 142-163). The amount of virus present in the immunoprecipitated material was examined by Western blotting. (B) The amounts of virions immunoprecipitated with the antibody against capsids, with the antibody against the PLA₂ motif, or with the preimmune serum from the rabbits used to produce the PLA₂ antibody were quantified by real-time PCR. Error bars represent the deviations from four independent experiments.

FIG. 5.

B19V increases the OF of human RBCs. RBCs (2 × 10⁶) were incubated with increasing amounts of B19V per cell (virus concentrated from serum by ultracentrifugation through 20% sucrose and resuspended in PBS). Following virus attachment for 1 h at 4°C, the cell suspension was incubated in an isotonic buffer containing calcium for 3 h at 37°C. The percentage of hemoglobin release was measured in the supernatant and referred to as direct hemolysis. Subsequently, the RBC pellet was resuspended in a hypotonic NaCl buffer (0.6g/dl) and incubated overnight at 37°C. The percentage of hemoglobin release was measured in the same way and was referred to as OF. Results are mean values from three independent experiments.

FIG. 6.

The VP1u-associated PLA₂ activity is responsible for the increased OF of human RBCs. (A) B19V-PLA₂ activity in the presence of antibodies against VP1u. The PLA₂ activity from heat-treated (60°C for 3 min) B19V (1.2 μg) was measured after incubation with MAb 1418-1 (α-N-VP1u; against the most N-terminal part of VP1u) or with antibody VP1u 142-163 (α-PLA₂; targeting the PLA₂ region). (B) OF of RBCs incubated with B19V under different conditions. RBCs (2 × 10⁶) were incubated with 3 × 10⁴ virions per cell. Following virus attachment for 1 h at 4°C, the cells were incubated in an isotonic buffer with or without calcium or in the presence of antibodies against VP1u for 3 h at 37°C. Following centrifugation, the RBC pellets were resuspended in a hypotonic NaCl buffer (0.6g/dl) and incubated overnight at 37°C. The percentage of hemoglobin release was measured. The graph indicates the average OF value of RBCs incubated with B19V (upper line) or without (w/o) B19V (lower line). B19, B19V; B19-60°C, B19V exposed to heat treatment of 60°C for 3 min. Error bars from three independent experiments are indicated.

FIG. 7.

(A) Quantification of B19V in the plasma and RBC fractions from an individual with a recent B19V (B19) infection (IgG and IgM positive). The heparinized fresh blood sample was centrifuged to separate plasma and RBCs. The RBCs were extensively washed in PBS. B19V DNA was quantified from a volume of 100 μl of plasma or washed RBCs, as specified in Materials and Methods. Infectivity of B19V bound to human RBCs based on NS1 RNA (B) or viral DNA quantification (C) is shown. RBCs were preincubated with B19V. Subsequently, the RBCs were washed intensively with PBS to remove unbound virus. The RBCs were added as intact (in PBS) or lysed cells (in water) to UT7/Epo cells. As controls, similar amounts of B19V alone or in the presence of RBCs that were preblocked with EC were used. NS1 RNA and viral DNA were quantified after 24 h and 84 h, respectively. Error bars from three independent experiments are indicated.