TEXT

Top Abstract Text References

Respiratory syncytial virus (RSV) is the major cause of severe diseases such as bronchiolitis and pneumonia in infants and young children. Since all attempts to vaccinate them with attenuated or killed virus have failed, passive immunization has been studied as an alternative for the protection of high-risk infants. Maternal antibodies have a protecting role as they induce in infants resistance to serious disease (10). Moreover, appearance of secretory immunoglobulin A (IgA) correlates with a decrease in virus shedding (13). Prophylactics and therapeutic studies have been performed to confirm the protecting role of antibodies. Protection by passive administration of monoclonal antibodies, anti-RSV polyclonal antisera, or recombinant human Fab has been shown with animal models (4, 15, 18, 20). A humanized monoclonal antibody to RSV was also shown to prevent and clear infection in mice (19). Intranasal administration of a neutralizing monoclonal antibody of isotype A protected mice from upper and lower respiratory tract infection (22). Clinical trials have also been performed with high-risk infants. Infusions of RSV immunoglobulin decreased the incidence of both upper and lower respiratory tract infections as well as prevented severe RSV disease (6-8).

The F protein of RSV is responsible for fusing the virus and cell membranes. Antigenic sites on the F protein have been determined by different approaches. The principal neutralizing domain on the F protein seems to be included in the amino acid sequence 190 to 289, since most of the neutralizing monoclonal antibodies recognize it (23). Assessed for neutralizing activity among a panel of antibodies to fusion protein (23), RS-348 has the highest neutralizing activity. The Ig variable domains are encoded by several fragments (V[D]J) that rearrange during B-cell differentiation. Framework regions separate complementarity-determining regions (CDRs), which are hypervariable regions of Ig interacting with the antigen. At first, we identified the CDR sequences of RS-348 antibody. Then we looked at which CDR, if any, was involved in the generation of a protective immunity.

Monoclonal antibody. RS-348 was produced as previously described (2). The parental fusion partner was Sp2/O cells, which is a nonsecreting mouse myeloma cell line. Its epitope was defined within amino acid sequence 190 to 289 on F protein. RS-348 has a neutralizing specificity for subgroup A strains (23) and inhibits the fusion due to RSV.

Production of V_H and V_L genes and identification of CDRs. mRNA was isolated from 10⁶ hybridoma cells (QuickPrep Micro mRNA purification kit; Pharmacia) and used as a template for reverse transcription. V_H and V_L genes were amplified with the Recombinant Phage Antibody System (Pharmacia). DNA sequences were derived by subcloning V_H and V_L genes and also by direct sequencing of PCR products. The deduced amino acid sequences of the CDRs were defined by alignment with other V_H and V sequences (9). They were then prepared as synthetic peptides (Table 1). Each of them was designed as a sequence of about 20 amino acids in length. If necessary, amino acids from the framework on each side of the CDR were added to reach this length or to facilitate the synthesis. An additional cysteine was added at both ends to obtain cyclic structures. The peptides were synthetized by a solid-phase method using N-Fmoc-protected, Dhbt or Pfp ester-activated amino acids on a polystyrene (PEG-PS; PerSeptive Biosystems, Framingham, Mass.) resin (25). After cleavage and side-chain deprotection, the cyclic form of the peptide was obtained in aqueous solution by spontaneous oxidation of the Cys thiol groups under strong agitation overnight of a 25-µmol aliquot (1 mg/ml) deprotonated by NaOH (final pH, 8 to 8.5). The linear form was kept protonated at a pH of 5.5 to 6.5, and if insoluble, it was mixed with degassed phosphate-buffered saline (pH = 7.4) under bubbling nitrogen in order to avoid cyclization. The peptide solutions were then rapidly lyophilized.

In vitro biological activity of CDRs. Peptides derived from all CDRs of RS-348 were assessed for their ability to neutralize the Long strain of RSV (Fig. 1 and 2). Briefly, virus (5 × 10³ PFU) was mixed for 1 h at 37°C with serially diluted (15 to 0.875 µg) peptide in either a cyclic or linear form. Monolayers of HEp-2 cells in six-well plates were then infected. Four days later, syncytia were counted after neutral red staining. Results were then expressed as percentages of infectivity compared to a control well without peptide. Among the six CDRs, only CDR3 of the heavy chain (PEP3H) presented a neutralizing activity (Fig. 1). The linear form was more efficient than the cyclic form at low concentrations. At higher concentrations, the level of inhibition obtained with both forms was the same. PEP3H reproduced a subgroup specificity in neutralizing assays, as no activity was observed against 9320 strain (subgroup B). The conformation, linear or cyclic, did not influence this result. An inhibition fusion assay was also carried out with peptide PEP3H. Cells were first infected with the virus (5 × 10³ PFU), and then 6 h later diluted peptide was added. A low (25%) but reproducible fusion-inhibiting activity was observed with 60 µg of PEP3H (data not shown). PEP2L and PEP3L, under linear form only, caused enhanced infection. This suggests that binding of these CDRs to fusion protein reinforced its fusing activity. Conformation of these sequences was important as cyclic forms did not give the same results. It seems that fusion protein activity may be modulated by single association with peptidic sequences, leading to an enhancement or a decrease of the RSV infectivity (Fig. 2). No cytotoxic effect due to peptides was observed.

View larger version (14K): [in this window] [in a new window]   FIG. 1.   Neutralizing capacity of linear (square) or cyclic (circle) PEP3H on subgroup A strain (solid symbols) or subgroup B strain (open symbols).

View larger version (15K): [in this window] [in a new window]   FIG. 2.   Neutralizing capacities of CDR peptides in either linear (A) or cyclic form (B) against Long strain of RSV.

4.5

In vivo biological activity of CDR3 of V_H: prophylactic assays with mice. In order to determine whether CDR3 can give some protection, we chose to run protection assays with mice. We used linear peptide in in vivo assays as the level of neutralization was slightly higher than with cyclic peptide. Four groups of seven to eight female 10-week-old BALB/c mice (IFFA/Credo, L'Arbresle, France) were used. These animals weighed about 20 g each. Fifty microliters of either PBS or 50 µl of ascitic fluid containing RS-348 (dilution, 1/25; 5.4 µg of Ig) or linear CDR3 peptide (PEP3H; 140 µg) was inoculated intranasally per mouse to one of three respective groups of mice under ketamine-xylamine anesthesia. A peptide (NEDFGLLGTTLLNLDAG) derived from amino acid sequence 60 to 75 of rotavirus VP6 protein was used also as a control (PEP control). Twenty-four hours later, the mice were inoculated intranasally under a second anesthesia with 3.75 × 10⁴ PFU of RSV Long strain diluted in basal medium Eagle. At 5 days postinfection, the mice were put down and their lungs harvested. Lung homogenates, diluted 1/10 (wt/vol), were titrated for RSV by plaque assay on HEp-2 cells after neutral red staining. Virus titers were expressed as PFU per gram of tissue. The differences between groups were tested for significance by multiple comparison tests (Duncan, Scheffé, Student-Newman-Keuls, and Tukey-Kramer). Performance of these four tests allowed us to check agreement among the statistical results. A single dose of 140 µg of PEP3H was very effective in the prevention of RSV infection. Indeed, intranasal administration of PEP3H 24 h before infection caused a significant decrease in virus in the lungs of RSV-infected mice. In these conditions, virus titer in the lungs was reduced by 2.3 log₁₀. For RS-348, we obtained a reduction of 1.56 log₁₀. No significant difference between PEP3H and RS-348 was observed. The peptide used as control did not reduce RSV replication at all. All statistical tests led to the same conclusions. Classical  level at 1% was chosen (Table 2). To check the absence of residual neutralizing activity which could interfere at the time of lung harvesting, we mixed (1/1 [vol/vol]) a PEP3H-treated lung with an infected lung which had not received any peptide. No neutralizing activity was obtained. Thus, resistance of lung to RSV infection was not due to residual in vitro neutralizing activity. Therefore, we have demonstrated for the first time that a single CDR can prevent progression of RSV infection in mice. Up to now, all prophylaxis studies on RSV infection in rodents have been performed with complete Ig or Fab. A protective effect with IgG has been obtained by different routes of administration, i.e., intravenous (18), intraperitoneal (19), or intranasal (14). Secretory IgA was not effective when administered intraperitoneally, suggesting that the antibody was not transported to the lungs. A reduction could even be obtained when the antibodies, IgG (16) or IgA (22), were administered intranasally several days before challenge. When polyclonal Ig was used, no difference between intranasal or parenteral administration was observed for the prophylactic effect (5), while the therapeutic effect was higher when the treatment was given intranasally (16). It seems therefore that intranasal administration is a better way to obtain protection of the lower respiratory tract. On the other hand, the choice of Ig isotype seems less important, as IgA and IgG are equally efficacious in protecting the airways from viral infection (12).