Abstract

Background. Palivizumab is a US Food and Drug Administration–approved monoclonal antibody for the prevention of respiratory syncytial virus (RSV) lower respiratory disease in high-risk infants. Motavizumab, derived from palivizumab with enhanced antiviral activity, has recently been tested in humans. Although palivizumab escape mutants have been generated in the laboratory, the development of resistant RSV in patients receiving palivizumab has not been reported previously.

Methods. We generated palivizumab and motavizumab escape mutants in vitro and examined the development of resistant mutants in RSV-breakthrough patients receiving immunoprophylaxis. The effect of these mutations on neutralization by palivizumab and motavizumab and in vitro fitness was studied.

Results. Antibody-resistant RSV variants selected in vitro had mutations at position 272 of the fusion protein, from lysine to asparagine, methionine, threonine, glutamine, or glutamate. Variants containing mutations at positions 272 and 275 were detected in breakthrough patients. All these variants were resistant to palivizumab, but only the glutamate variant at position 272 demonstrated resistance to motavizumab. Mixtures of wild-type and variant RSV soon lost the resistant phenotype in the absence of selection.

Conclusions. Resistant RSV variants were detected in a small subset (∼5%) of RSV breakthrough cases. The fitness of these variants was impaired, compared to wild-type RSV.

Respiratory syncytial virus (RSV) is a member of the Pneumovirus genus in the Paramyxoviridae family. The RSV genome consists of a negative-sense, nonsegmented, single strand of RNA encoding 10 proteins [1]. RSV is an enveloped virus, and its antigenicity is determined by 2 transmembrane glycoproteins, the attachment glycoprotein (G) and the fusion (F) protein. RSV is classified into A and B subgroups, originally based on antigenic differences in the G protein [2].

RSV is the most serious respiratory pathogen in infants and young children, causing annual epidemics of bronchiolitis and pneumonia worldwide [1, 3, 4]. Overall, RSV infection is responsible for ∼2% of the hospitalization rate among infants <1 year of age [5]. This rate increases at least 4–5-fold among children at high risk of severe RSV disease, including premature infants and those with chronic lung disease of prematurity, immunodeficiency, or complicated congenital heart disease [6–12]. RSV infection in infants and children can cause lung function deterioration that may be sustained for months after the acute illness, and in some circumstances, years of recurrent wheezing or asthma may ensue [13, 14].

To date, prevention of RSV disease has only been achieved by the passive administration of RSV-specific immunoglobulin. Prophylaxis with palivizumab (MedImmune), a humanized monoclonal antibody (mAb) that is directed against the RSV F protein, can significantly reduce the rate of RSV-related hospitalizations in high-risk infants [15, 16]. Motavizumab (MEDI-524; MedImmune), an enhanced mAb developed by affinity maturation of palivizumab [17–19], is in clinical development. Nonclinical studies demonstrated that motavizumab was more effective than palivizumab at neutralizing RSV in vitro. In addition, at equivalent serum and lung levels, motavizumab was shown to be superior to palivizumab at reducing RSV infection in both the upper and lower airways of cotton rats [18]. Motavizumab and palivizumab bind antigenic site A, a highly conserved region on the RSV F protein between amino acids 258 and 275 [20].

Similar to other RNA viruses, replication of RSV depends on an RNA polymerase that lacks proofreading and repair capability, resulting in a relatively high mutation rate. This mutability could increase the potential for the generation of resistant mutants under selective drug pressure, such as antibody prophylaxis. In vitro development of RSV A mutants resistant to palivizumab has been reported previously. Beeler and Coelingh [20] isolated RSV monoclonal antibody resistant mutants (MARMs) containing phenotypic amino acid variations at positions 262, 275, and 276 of the F protein with use of the murine precursor to palivizumab, mAb1129. Sullender et al [21–23] also isolated palivizumab MARMs containing mutations at positions 268 and 272. The potential for resistance to occur was also explored during the clinical development of palivizumab. In a prospective study using a binding assay that was predictive of palivizumab neutralization, the investigators showed that palivizumab bound to all 25 RSV isolates collected from patients actively receiving palivizumab [24]. However, the number of samples was small in this study, and the assay was suboptimal for detecting minor drug-resistant viral populations. Recently, nucleotide sequence analysis of RSV isolates collected directly from nasal wash specimens from infants who received palivizumab and still developed acute lower tract respiratory infection revealed an F protein mutation at position 272 from lysine (K) to glutamate (E). Although the susceptibility of this variant to neutralization by palivizumab could not be determined because it did not propagate in cell culture [25], it was suggested that this K272E variant would most likely be less susceptible to neutralization by palivizumab, because multiple in vitro-selected palivizumab MARMs contain mutations at this position. To date, the data available on the rate of emergence of clinical resistant variants during treatment are limited.

In the present study, we describe the in vitro selection and characterization of additional palivizumab MARMs and a novel motavizumab MARM. We also examined amino acid changes in the F protein of RSV isolates collected from RSV-breakthrough patients receiving palivizumab or motavizumab in a large phase 3 clinical study comparing motavizumab with palivizumab (MI-CP110: Study of MEDI-524; for the prophylaxis of RSV disease in high risk children) [26]. The sensitivity of these immunoprophylaxis breakthrough isolates to motavizumab and palivizumab was determined using recombinant RSV (rRSV). Moreover, the effect of these neutralization-resistant mutations on in vitro fitness was assessed.

MATERIALS AND METHODS

Viruses and Cells

HEp-2 cells, RSV A Long and A2 strains were obtained from American Type Culture Collection. BSR T7 cells, a BHK-derived cell line which constitutively expresses the T7 RNA polymerase, was kindly provided by Karl-Klaus Conzelmann [27].

Selection of MARM Variants

All of the MARMs were selected using the same general procedure. Palivizumab (30 μg/mL) or motavizumab (4 μg/mL) was mixed with a stock of RSV A Long and incubated at 37°C for 1 h, and HEp-2 cells were infected at a multiplicity of infection of 10 plaque-forming units (PFUs)/cell, followed by 7 days incubation. The presence of cytopathic effects was assessed daily using light microscopy. On day 7, the contents of each well were harvested, adjusted to the initial mAb concentration, and used to infect freshly seeded HEp-2 cells as described above. The selection process was repeated 2 more times. For the motavizumab MARMs only, the motavizumab concentration was increased to 10 μg/mL in the third round of selection. The MARMs were biologically cloned 2–3 times either by plaque assay or by sequential limiting dilutions in 96-well plates containing HEp-2 cells.

Subjects, Sample Collection, and Genotyping of Resistant Viruses

Infants were recruited using the same high-risk criteria used in the pivotal IMpact trial of palivizumab [16]. Subjects received either motavizumab or palivizumab, each at 15 mg/kg doses given at monthly intervals, with follow-up for 150 days. Samples of nasopharyngeal secretions were collected at the time of each hospitalization for respiratory illness. These samples were tested in a central laboratory for the presence of RSV with use of real-time reverse-transcriptase polymerase chain reaction (RT-PCR). Aliquots of nasal specimens that tested positive for RSV by real-time RT-PCR were used for further genotypic analysis of the F protein. RNA was isolated from nasal specimens with use of the High Pure Viral Nucleic Acid Kit (Roche Diagnostic). Viral cDNA corresponding to the entire F gene was generated using AccuScript High Fidelity 1st strand cDNA synthesis kit (Stratagene), and the F gene was subsequently amplified (primer sequences and conditions are available upon request) using Herculase II Fusion DNA Polymerase (Stratagene). The GenBank accession numbers of the nucleotide sequences obtained in the present study are HQ317232–HQ317243.

Generation of Recombinant Virus

The conditions used for rRSV cDNA construction containing the desired mutations and for rescuing rRSV are described elsewhere [28]. The rRSV collected in the supernatant was biologically cloned 2–3 times in sequential limiting dilutions in 96-well plates containing HEp-2 cells. The viral clones were amplified 2–3 times in HEp-2 cells before characterization.

Microneutralization Assay

The RSV microneutralization assay was performed as described elsewhere [17, 19]. In brief, RSV was incubated with an equal volume of serially diluted antibodies and incubated for 2 h at 37°C. HEp-2 cells were added and cultured for 5 days. The replication of RSV was determined by quantification of expressed F protein by enzyme-linked immunosorbent assay measured at 450 nm (A450). The neutralizing titer (50% inhibitory concentration [IC50]) is expressed as the antibody concentration that reduced the A450 value by 50%, compared to control wells.

Determination of Growth Curves and Fitness of RSV Mutant Variants

To generate growth curves, HEp-2 cells were infected with RSV at a multiplicity of infection of .1 pfu/cell. The culture supernatant was collected twice daily, and viral titers were determined using a median tissue culture infective dose (TCID50) assay in HEp-2 cells [29]. To investigate the replicative capacities of different mutant variants, HEp-2 cells were infected at a total multiplicity of infection of .1 pfu/cell, with a 50:50 mixture of RSV wild-type and mutant variant. After 3.5 days of incubation at 37°C in 5% carbon dioxide, a portion of culture supernatant was transferred to a fresh flask of HEp-2 cells. The remaining supernatant was stored at −80°C for subsequent neutralization and sequencing analysis. The virus mixture was passaged in this manner every 3.5 days for 10 weeks, or 20 passages total. The viral titers were determined using a plaque neutralization assay in Vero cells in the absence or presence of palivizumab or motavizumab.

RESULTS

In Vitro Selection and Genotype Mapping of RSV MARMs to Palivizumab and Motavizumab

Twelve palivizumab and 19 motavizumab MARMs were independently selected by multiple passages of RSV A in the presence of the respective mAbs in HEp-2 cells. The genes for the F and G proteins were amplified using RT-PCR and were subsequently sequenced. No amino acid changes from the prototypic RSV A Long sequence were detected in the G protein of all 31 clones. Sequence analysis of the F genes from the 12 palivizumab MARMs revealed 4 nonsynonymous substitutions, compared with the parental sequence (Table 1). These nucleotide changes respectively code for single amino acid substitutions from Lys (K) to Gln (Q), Met (M), Thr (T), or Asn (N) at amino acid position 272 in the highly conserved antigenic site A region. In contrast, sequence analysis of the F protein gene from the motavizumab MARMs demonstrated that all 19 isolates contained a single nucleotide substitution at position 814, corresponding to an amino acid change from K to Glu (E) also at amino acid position 272 (Table 1). No mutations in the F protein were observed when RSV was passaged in the presence of the antihuman CD2 control antibody Medi507.

Table 1.

Genotypic Changes Selected in the Monoclonal Antibody-Resistant Mutants (MARMs) and Phenotypic Susceptibility of these MARMs to Palivizumab and Motavizumab

  Nucleotide at Position
 
Amino Acid at Position 272 IC50, μg/mL (Fold Changes)
 
Reference 
Antibody Used for Selection Virus 814 815 816 Palivizumab Motavizumab 
None A Long 0.353±.06 (1.0±.2) 0.021±.017 (1.0±.8)  
Palivizumab MARM B1   1823±611 (5164±1731) 0.015±.009 (.7±.4) this study 
 MARM B2   >9000 (>25,000) 0.038±.025 (1.8±1.2) this study; Zhao et al, 2004 [21
 MARM B7   >9000 (>25,000) 0.057±.018 (2.7±.9) this study 
 MARM B9   >9000 (>25,000) 0.026±.031 (1.2±1.5) this study; Zhao et al, 2005 [30
Motavizumab MARM N3   >9000 (>25,000) 30.04±11.35 (1430±540) this study 
  Nucleotide at Position
 
Amino Acid at Position 272 IC50, μg/mL (Fold Changes)
 
Reference 
Antibody Used for Selection Virus 814 815 816 Palivizumab Motavizumab 
None A Long 0.353±.06 (1.0±.2) 0.021±.017 (1.0±.8)  
Palivizumab MARM B1   1823±611 (5164±1731) 0.015±.009 (.7±.4) this study 
 MARM B2   >9000 (>25,000) 0.038±.025 (1.8±1.2) this study; Zhao et al, 2004 [21
 MARM B7   >9000 (>25,000) 0.057±.018 (2.7±.9) this study 
 MARM B9   >9000 (>25,000) 0.026±.031 (1.2±1.5) this study; Zhao et al, 2005 [30
Motavizumab MARM N3   >9000 (>25,000) 30.04±11.35 (1430±540) this study 

NOTE. IC50, 50% inhibitory concentration.

The Phenotypic Susceptibility of MARMs to Palivizumab and Motavizumab

The representative MARM variants were used in a microneutralization assay to further evaluate the ability of palivizumab and motavizumab to neutralize these resistant viruses. The palivizumab-selected clones B1, B2, B7, and B9 containing mutations K272N, K272M, K272T, or K272Q were largely resistant to palivizumab neutralization, with increases in calculated IC50 values ranging from 5164-fold for clone B1 to >25,000-fold for B2, B7, and B9 MARMs, compared with the parental A Long strain (Table 1; Figure 1A). However, they were all susceptible to neutralization by motavizumab, with IC50 values within a 5-fold range, compared with wild-type RSV (Table 1; Figure 1B). In contrast, motavizumab MARM clone N3 containing the K272E mutation exhibited a dramatic reduction in motavizumab neutralization, with a calculated IC50 that was 1430-fold higher than the parental strain. This variant also exhibited complete resistance to palivizumab, with no detectable neutralization even at the highest palivizumab concentration tested (∼10 mg/mL).

Figure 1.

Susceptibility of mutant variants to neutralization by palivizumab (A) and motavizumab (B). Microneutralization assays were performed in HEp-2 cells infected with respiratory syncytial virus (RSV) in the presence of serially diluted antibodies. Viral replication was measured using an F-protein–specific enzyme-linked immunosorbent assay (ELISA) represented by the A450 value. The neutralization titer is presented as 50% inhibitory concentration (IC50), the antibody concentration that gives a 50% reduction in the A450 value, compared with the control wells without antibody.

Figure 1.

Susceptibility of mutant variants to neutralization by palivizumab (A) and motavizumab (B). Microneutralization assays were performed in HEp-2 cells infected with respiratory syncytial virus (RSV) in the presence of serially diluted antibodies. Viral replication was measured using an F-protein–specific enzyme-linked immunosorbent assay (ELISA) represented by the A450 value. The neutralization titer is presented as 50% inhibitory concentration (IC50), the antibody concentration that gives a 50% reduction in the A450 value, compared with the control wells without antibody.

Genotypic Analyses of RSV F Gene Among Patients Who Failed Palivizumab or Motavizumab Prophylaxis

To investigate whether antibody-resistant viruses might also arise in human recipients of palivizumab or motavizumab, we evaluated 178 subjects (of >6600 enrolled) who were diagnosed with lower respiratory tract RSV disease after immunoprophylaxis with palivizumab or motavizumab in a Phase 3 clinical study [26]. These infants received either palivizumab or motavizumab at 15mg/kg monthly for 5 months during the 2004–2006 RSV seasons. In total, there were 190 RSV-positive nasal specimens available for nucleic acid recovery. The RSV F gene was amplified and genetic variation was examined in 157 evaluable nasal specimens, collected from 146 subjects. 64 of these samples were from motavizumab recipients and 93 were from palivizumab recipients (Table 2). Careful visual inspection of sequencing electrophoretograms was employed in the region of the F gene encoding antigenic site A, to detect potential mixtures of RSV subtypes. Sequences were compared to RSV isolates collected from treatment-naive patients. Of the 157 F gene sequences generated in this study, 8 samples contained nucleotide changes either from A to C or G at position 814, or from C to T at position 824 resulting in amino acid alterations from K to E or Q at position 272 and from Ser (S) to Phe (F) or Leu (L) at position 275 in the antigenic site A (Table 2). Three of these were derived from samples collected from motavizumab recipients, and 5 were from palivizumab recipients. In 3 patients, mutant virus was found mixed with wild type virus. The calculated amino acid mutation frequencies in antigenic site A were 5.4% for palivizumab recipients and 4.7% for motavizumab recipients (Table 2). Interestingly, MARMs for palivizumab and motavizumab were observed with higher frequency in the outpatient group compared to the hospitalized patients.

Table 2.

Sequence Changes in Antigenic Site A of Protein F Identified in Respiratory Syncytial Virus (RSV) Isolates Collected From Recipients of Antibody Immunoprophylaxis

Treatment Total No. of Isolates ID Nucleotide Change Amino Acid Change No. of Mutant Variants Frequency, Σn/N%(hospitalization: outpatients) 
Palivizumab 93 Clin PM-1 A814Ga K272Ea 5.4 (3.7 : 10.2) 
  Clin PM-2 A814Ca K272Qa  
  Clin PM-3b C824T S275F  
  Clin PM-4 C824T S275L  
Motavizumab 64 Clin MM-1b A814G K272E 4.7 (2.4 : 9.1) 
  Clin MM-2 C824Ta S275La  
Treatment Total No. of Isolates ID Nucleotide Change Amino Acid Change No. of Mutant Variants Frequency, Σn/N%(hospitalization: outpatients) 
Palivizumab 93 Clin PM-1 A814Ga K272Ea 5.4 (3.7 : 10.2) 
  Clin PM-2 A814Ca K272Qa  
  Clin PM-3b C824T S275F  
  Clin PM-4 C824T S275L  
Motavizumab 64 Clin MM-1b A814G K272E 4.7 (2.4 : 9.1) 
  Clin MM-2 C824Ta S275La  

NOTE.aA mixed (mutant and wild-type) sequence was detected at this position.

b

F gene sequences were identical in 2 mutant variants.

Phenotypic Characterization of Recombinant Viruses Containing Mutations in Antigenic Site A Identified in Clinical Isolates

Because the conditions for patient sample collection and storage in this clinical study were only appropriate for molecular genotypic analyses but not for recovery of proliferative RSV, none of the RSV-breakthrough isolates encoding an antigenic site A mutation grew in cell culture. To assess the effect of these antigenic site A mutations on the susceptibility to neutralization by motavizumab and palivizumab, recombinant RSV A2 (rA2) mutant variants rA2_K272Q, rA2_K272E, rA2_S275F, and rA2_S275L were generated and evaluated using a microneutralization assay. As shown in Table 3, the rA2_K272Q and rA2_K272E variants exhibited complete resistance to palivizumab at the highest concentration tested (∼10 mg/mL). However, the rA2_K272Q mutant was neutralized by motavizumab with a potency comparable to recombinant wild-type rA2 and native wild-type laboratory strain RSV A2, whereas the rA2_K272E mutant was >600-fold less susceptible to motavizumab, compared with wild-type A2 strains (Table 3). A similar pattern was also seen with the in vitro–selected MARMs containing K272Q and K272E mutations in A Long strain as described above (Table 1), indicating that rRSV recapitulates the same phenotype as the parental laboratory strain. The variants encoding mutations at position 275 (S275F and S275L), which had not been generated by in vitro selection in our hands, behaved similarly to the K272Q variant, in that rA2_S275F and rA2_S275L were neutralized by motavizumab with a potency comparable to wild-type RSV A2 but were entirely resistant to palivizumab to the highest concentration tested (∼10 mg/mL). Thus, only the K272E mutation conferred resistance to both palivizumab and motavizumab, whereas the other mutations in antigenic site A observed in MARMs or clinical breakthrough isolates exhibited resistance to palivizumab only.

Table 3.

Susceptibility of Recombinant Mutant Respiratory Syncytial Virus (RSV) Variants to Palivizumab and Motavizumab Measured by Microneutralization Assay

  IC50, μg/mL (Fold Changes)
 
Virus Amino Acid Changes Palivizumab Motavizumab 
A2 none 0.342±.091 (1.0±.27) 0.029±.008 (1.0±.28) 
rA2 none 0.388±.069 (1.0±.18) 0.029±.009 (1.0±.31) 
rA2_K272Q K272Q >9000 (>25000) 0.039±.01 (1.34±.34) 
rA2_K272E K272E >9000 (>25000) 18.14±10.26 (625±354) 
rA2_S275L S275L >9000 (>25000) 0.017±.003 (.59±.10) 
rA2_S275F S275F >9000 (>25000) 0.008±.001 (.3±.03) 
  IC50, μg/mL (Fold Changes)
 
Virus Amino Acid Changes Palivizumab Motavizumab 
A2 none 0.342±.091 (1.0±.27) 0.029±.008 (1.0±.28) 
rA2 none 0.388±.069 (1.0±.18) 0.029±.009 (1.0±.31) 
rA2_K272Q K272Q >9000 (>25000) 0.039±.01 (1.34±.34) 
rA2_K272E K272E >9000 (>25000) 18.14±10.26 (625±354) 
rA2_S275L S275L >9000 (>25000) 0.017±.003 (.59±.10) 
rA2_S275F S275F >9000 (>25000) 0.008±.001 (.3±.03) 

NOTE. IC50,50% inhibitory concentration.

Characterization of the Growth Kinetics and Fitness of mAb-Resistant Mutants

To determine whether the motavizumab and palivizumab MARMs had altered growth properties, in vitro selected motavizumab or palivizumab MARMs and recombinant resistant mutants were used as surrogates for naturally occurring RSV containing these mutations. To this end, the growth of parental RSV A Long and rRSV A2 and their derived mutants were compared. As shown in Figure 2A, the A Long MARMs K272E, K272Q, K272M, K272N, and K272T did not show enhanced growth kinetics in HEp-2 cell culture, compared with wild-type A Long strain in repeated studies, nor did the rRSV A2 S275L, S275F, K272E, or K272Q variants, compared with wild-type rRSV A2 (Figure 2B). Overall, the growth kinetics of wild-type A Long or rRSV A2 and their derived palivizumab or motavizumab MARMs were comparable except that recombinant rA2_S275L grew somewhat more slowly than did wild-type rRSV A2. There were also no substantial differences observed in the size or morphology of the MARM plaques, compared with those of wild-type virus (data not shown).

Figure 2.

Growth kinetics of wild-type respiratory syncytial virus (RSV) A and mutant viruses. HEp-2 cells were inoculated at a multiplicity of infection of .1 pfu/cell with parental strain or mutant strains. The culture supernatant was sampled twice daily and titered for the presence of virus on HEp-2 cells using a 50% tissue culture infective dose (TCID50) assay. A, Growth kinetics of parental RSV A Long and mutant variants A Long K272E, A Long K272Q, A Long K272M, A Long K272N, and A Long K272T. B, Growth kinetics of parental recombinant RSVA2 (rA2) and mutant viruses rA2 K272E, rA2 K272Q, rA2 S275F, and rA2 S275L. Growth capacity is represented as a log value of TCID50 units per mL.

Figure 2.

Growth kinetics of wild-type respiratory syncytial virus (RSV) A and mutant viruses. HEp-2 cells were inoculated at a multiplicity of infection of .1 pfu/cell with parental strain or mutant strains. The culture supernatant was sampled twice daily and titered for the presence of virus on HEp-2 cells using a 50% tissue culture infective dose (TCID50) assay. A, Growth kinetics of parental RSV A Long and mutant variants A Long K272E, A Long K272Q, A Long K272M, A Long K272N, and A Long K272T. B, Growth kinetics of parental recombinant RSVA2 (rA2) and mutant viruses rA2 K272E, rA2 K272Q, rA2 S275F, and rA2 S275L. Growth capacity is represented as a log value of TCID50 units per mL.

Of interest, when wild-type RSV strains and the derived drug-resistant mutant variants were mixed in approximately equal number and passaged 20 times in HEp-2 cells, parental wild-type RSV strains appeared to have a competitive growth advantage. This was determined by a plaque focus forming assay performed on virus strains collected at different passage intervals in the absence or presence of palivizumab or motavizumab (Figure 3). This analysis revealed rapid disappearance of all resistant variants from the mixed population with similar dynamics; after only a few passages, almost all of the remaining RSV was susceptible to palivizumab or motavizumab neutralization. The proportion of mutant variants decreased from nearly 20%–50% at passage 1 to almost undetectable by passage 10 and remained low (<1%) up to passage 20 for all mutant variants tested except S275F, which reappeared at ∼10% of the total mixture. Nucleotide sequence analyses of viral RNA extracted at different passage numbers confirmed the conclusions of the plaque titration studies, such that the wild-type sequence became dominant very quickly and no variant sequence was detected after passage 7 in the mixed population (data not shown).

Figure 3.

Differential plaque neutralization focus forming assay determination of proportions of mutant variants during competitive replication with the A Long strain (A) or wild-type rA2 virus (B). Equal amounts of the parental respiratory syncytial virus (RSV) A and mutant variants were mixed and used to infect HEp-2 cells. A portion of virus-containing culture supernatant was passaged to new cells every 3.5 days. Remaining supernatants from each passage were stored at −80°C and tested by plaque neutralization focus forming assay in the absence or presence of palivizumab or motavizumab on Vero cells. The viral titer from the palivizumab containing plate represented palivizumab resistant virus that harbor the mutations K272Q, K272N, K272M, and K272T at antigenic site A of F protein. The viral titer from motavizumab containing plate represented mutant variant K272E that is resistant to motavizumab. The viral titer from the plain plate indicated both mutant and wild type virus. The percentage of mutant (resistant) virus of the total virus population (y-axis) was calculated by dividing the viral titer in the monoclonal antibody plate by that in the plate lacking monoclonal antibody.

Figure 3.

Differential plaque neutralization focus forming assay determination of proportions of mutant variants during competitive replication with the A Long strain (A) or wild-type rA2 virus (B). Equal amounts of the parental respiratory syncytial virus (RSV) A and mutant variants were mixed and used to infect HEp-2 cells. A portion of virus-containing culture supernatant was passaged to new cells every 3.5 days. Remaining supernatants from each passage were stored at −80°C and tested by plaque neutralization focus forming assay in the absence or presence of palivizumab or motavizumab on Vero cells. The viral titer from the palivizumab containing plate represented palivizumab resistant virus that harbor the mutations K272Q, K272N, K272M, and K272T at antigenic site A of F protein. The viral titer from motavizumab containing plate represented mutant variant K272E that is resistant to motavizumab. The viral titer from the plain plate indicated both mutant and wild type virus. The percentage of mutant (resistant) virus of the total virus population (y-axis) was calculated by dividing the viral titer in the monoclonal antibody plate by that in the plate lacking monoclonal antibody.

DISCUSSION

The quasispecies nature of RNA viruses allows rapid adaptation to changing selective pressure [31]. RSV is no different: monoclonal and polyclonal antibody resistant mutants of RSV have been readily selected in tissue culture [20–23, 32–36]. Previous reports have shown the K272Q, K272M, and N268I mutations can be selected after passaging RSV A2 virus in vitro in the presence of increasing concentrations of palivizumab [21–23]. In addition, for RSV and other viruses, selection of viral escape mutants has been described after active and passive immunization in animals, demonstrating that immunization can result in resistant virus in vivo [30, 37–39]. Although it was possible for RSV to develop resistance during palivizumab use, clinical failure of prophylaxis because of palivizumab-resistant virus has not been reported to date in postmarketing analysis [40]. In prophylaxed infants, it is possible that the mean concentrations of mAb in the lung were sufficient to neutralize RSV and prevent the emergence of escape mutant variants. However, earlier studies of potential resistant RSV in prophylaxed patients were limited in scope [17, 24, 40], and the assays used to detect resistant RSV were not highly sensitive. Recently, an independent study in which the RSV F protein was directly genotyped from nasal specimens identified a mutation at codon 272 (K272E or K272Q) in 2 of 16 RSV-positive samples collected from patients who failed palivizumab prophylaxis, suggesting the potential appearance of resistant RSV in some patients receiving palivizumab (G. Boivin, personal communication) [25].

In the present study, genotypic analysis of breakthrough RSV isolates collected from subjects receiving palivizumab or motavizumab during study MI-CP110 revealed a low frequency of amino acid changes in antigenic site A (∼5% in both treatment groups), leading to potential drug resistance. Of the 157 nasal wash specimens from breakthrough subjects that were examined, 8 contained RSV with amino acid changes within antigenic site A (K272Q, K272E, S275F, or S275L). Recombinant RSV with these antigenic site A mutations were completely resistant to palivizumab neutralization, as determined by an in vitro microneutralization assay. Conversely, all of these variants, except the K272E mutant, were sensitive to neutralization by motavizumab; the susceptibility of the K272E variant to motavizumab was significantly reduced. In this regard, detection of the palivizumab-resistant but motavizumab-susceptible S275L variant among RSV isolates collected from motavizumab-breakthrough subjects cannot be explained on the basis of the results of our in vitro selection of resistant RSV. However, it is possible that different neutralization phenotypes will be observed for MARMs selected in vivo rather than in vitro. We generated and tested >20 additional rRSVs encoding amino acid changes in the mature F protein outside antigenic site A, which were found only in breakthrough isolates but not in RSV isolates collected from treatment-naive subjects. None of these variants were resistant to palivizumab or motavizumab (data not shown), indicating that only changes in antigenic site A of the F protein have been demonstrated to confer resistance to palivizumab and motavizumab.

Very recently, analysis of cocrystals of the motavizumab antigen-binding fragment in complex with a 24-residue peptide spanning the antigenic site A of F protein suggested that mutations at positions 262, 268, 272, and 275 of the F protein could potentially disrupt the formation of the hydrogen bonds or ionic interactions with amino acids in motavizumab and affect antibody binding [41]. However, results from a cell based equilibrium binding assay to determine the binding activity of motavizumab and palivizumab to the F proteins in MARM-infected cells indicated that, although all the mutations described in this study at position 272 or 275 significantly decreased binding to palivizumab, only K272E reduced binding to motavizumab (data not shown). Thus, affinity-enhanced motavizumab may be more tolerant of sequence changes in antigenic site A than was predicted by crystallographic analysis.

Escape mutants and their fitness clearly have important implications for the capacity of mutant variants to spread in nature. Of interest, whereas the parental and variant rRSVs all replicated at about the same rates individually, when grown as mixtures, the parental strains had a definite growth advantage in vitro, compared with the variants, and the variants soon diminished to the limits of detection. This pattern has been observed previously with other viruses, including vesicular stomatitis virus and measles virus [42, 43]. In the present study, all of the palivizumab- and motavizumab-resistant variants were less fit than the parental viruses, suggesting that these variants are less likely to disseminate in the community because of a growth disadvantage in the absence of motavizumab and palivizumab selective pressure. Finally, although it is not established that the RSV disease in those few RSV- breakthrough patients was caused by strains containing amino acid changes in the F protein antigenic site A, a review of the clinical findings and time course for recovery for those patients were as expected (unpublished data), suggesting that these mutations did not increase virulence or worsen the patient's clinical outcome.

Funding

This work was supported by MedImmune.

We thank G. Boivin, for helpful discussion and sharing information, and J. Hong and J. Schickli, for kindly providing the pRSVC4G cDNA construct.

References

1.
Collins
PL
McIntosh
K
Channock
RM
Respiratory syncytial virus
Fields virology
 , 
1996
3rd ed
Philadelphia, PA
Lippincott-Raven
(pg. 
1313
-
96
)
2.
Johnson
PR
Jr
Olmsted
RA
Prince
GA
, et al.  . 
Antigenic relatedness between glycoproteins of human respiratory syncytial virus subgroups A and B: evaluation of the contributions of F and G glycoproteins to immunity
J Virol
 , 
1987
, vol. 
61
 (pg. 
3163
-
6
)
3.
Brandt
CD
Kim
HW
Arrobio
JO
, et al.  . 
Epidemiology of respiratory syncytial virus infection in Washington, D.C. 3. Composite analysis of eleven consecutive yearly epidemics
Am J Epidemiol
 , 
1973
, vol. 
98
 (pg. 
355
-
64
)
4.
Glezen
P
Denny
FW
Epidemiology of acute lower respiratory disease in children
N Engl J Med
 , 
1973
, vol. 
288
 (pg. 
498
-
505
)
5.
Hall
CB
Respiratory syncytial virus
Textbook of pediatric infectious diseases
 , 
1998
4th ed
Philadelphia, PA
WB Saunders
(pg. 
2084
-
110
)
6.
Groothuis
JR
Gutierrez
KM
Lauer
BA
Respiratory syncytial virus infection in children with bronchopulmonary dysplasia
Pediatrics
 , 
1988
, vol. 
82
 (pg. 
199
-
203
)
7.
Groothuis
JR
Simoes
EA
Levin
MJ
, et al.  . 
Prophylactic administration of respiratory syncytial virus immune globulin to high-risk infants and young children. The Respiratory Syncytial Virus Immune Globulin Study Group
N Engl J Med
 , 
1993
, vol. 
329
 (pg. 
1524
-
30
)
8.
Hall
CB
Kopelman
AE
Douglas
RG
Jr
Geiman
JM
Meagher
MP
Neonatal respiratory syncytial virus infection
N Engl J Med
 , 
1979
, vol. 
300
 (pg. 
393
-
6
)
9.
Heilman
CA
From the National Institute of Allergy and Infectious Diseases and the World Health Organization. Respiratory syncytial and parainfluenza viruses
J Infect Dis
 , 
1990
, vol. 
161
 (pg. 
402
-
6
)
10.
Law
BJ
De Carvalho
V
Respiratory syncytial virus infections in hospitalized Canadian children: regional differences in patient populations and management practices. The Pediatric Investigators Collaborative Network on Infections in Canada
Pediatr Infect Dis J
 , 
1993
, vol. 
12
 (pg. 
659
-
63
)
11.
MacDonald
NE
Hall
CB
Suffin
SC
Alexson
C
Harris
PJ
Manning
JA
Respiratory syncytial viral infection in infants with congenital heart disease
N Engl J Med
 , 
1982
, vol. 
307
 (pg. 
397
-
400
)
12.
Raes
M
Peeters
V
Alliet
P
, et al.  . 
Peripheral blood T and B lymphocyte subpopulations in infants with acute respiratory syncytial virus brochiolitis
Pediatr Allergy Immunol
 , 
1997
, vol. 
8
 (pg. 
97
-
102
)
13.
Hiatt
PW
Grace
SC
Kozinetz
CA
, et al.  . 
Effects of viral lower respiratory tract infection on lung function in infants with cystic fibrosis
Pediatrics
 , 
1999
, vol. 
103
 (pg. 
619
-
26
)
14.
Sigurs
N
Bjarnason
R
Sigurbergsson
F
Kjellman
B
Respiratory syncytial virus bronchiolitis in infancy is an important risk factor for asthma and allergy at age 7
Am J Respir Crit Care Med
 , 
2000
, vol. 
161
 (pg. 
1501
-
7
)
15.
Feltes
TF
Cabalka
AK
Meissner
HC
, et al.  . 
Palivizumab prophylaxis reduces hospitalization due to respiratory syncytial virus in young children with hemodynamically significant congenital heart disease
J Pediatr
 , 
2003
, vol. 
143
 (pg. 
532
-
40
)
16.
Palivizumab, a humanized respiratory syncytial virus monoclonal antibody, reduces hospitalization from respiratory syncytial virus infection in high-risk infants.The IMpact-RSV Study Group
Pediatrics
 , 
1998
, vol. 
102
 (pg. 
531
-
7
)
17.
Johnson
S
Oliver
C
Prince
GA
, et al.  . 
Development of a humanized monoclonal antibody (MEDI-493) with potent in vitro and in vivo activity against respiratory syncytial virus
J Infect Dis
 , 
1997
, vol. 
176
 (pg. 
1215
-
24
)
18.
Wu
H
Pfarr
DS
Johnson
S
, et al.  . 
Development of motavizumab, an ultra-potent antibody for the prevention of respiratory syncytial virus infection in the upper and lower respiratory tract
J Mol Biol
 , 
2007
, vol. 
368
 (pg. 
652
-
65
)
19.
Wu
H
Pfarr
DS
Tang
Y
, et al.  . 
Ultra-potent antibodies against respiratory syncytial virus: effects of binding kinetics and binding valence on viral neutralization
J Mol Biol
 , 
2005
, vol. 
350
 (pg. 
126
-
44
)
20.
Beeler
JA
van Wyke Coelingh
K
Neutralization epitopes of the F glycoprotein of respiratory syncytial virus: effect of mutation upon fusion function
J Virol
 , 
1989
, vol. 
63
 (pg. 
2941
-
50
)
21.
Zhao
X
Chen
FP
Megaw
AG
Sullender
WM
Variable resistance to palivizumab in cotton rats by respiratory syncytial virus mutants
J Infect Dis
 , 
2004
, vol. 
190
 (pg. 
1941
-
6
)
22.
Zhao
X
Chen
FP
Sullender
WM
Respiratory syncytial virus escape mutant derived in vitro resists palivizumab prophylaxis in cotton rats
Virology
 , 
2004
, vol. 
318
 (pg. 
608
-
12
)
23.
Zhao
X
Liu
E
Chen
FP
Sullender
WM
In vitro and in vivo fitness of respiratory syncytial virus monoclonal antibody escape mutants
J Virol
 , 
2006
, vol. 
80
 (pg. 
11651
-
7
)
24.
DeVincenzo
JP
Hall
CB
Kimberlin
DW
, et al.  . 
Surveillance of clinical isolates of respiratory syncytial virus for palivizumab (Synagis)-resistant mutants
J Infect Dis
 , 
2004
, vol. 
190
 (pg. 
975
-
8
)
25.
Boivin
G
Caouette
G
Frenette
L
Carbonneau
J
Ouakki
M
De Serres
G
Human respiratory syncytial virus and other viral infections in infants receiving palivizumab
J Clin Virol
 , 
2008
, vol. 
42
 (pg. 
52
-
7
)
26.
Carbonell-Estrany
X
Simoes
EA
Dagan
R
, et al.  . 
Motavizumab for prophylaxis of respiratory syncytial virus in high-risk children: a noninferiority trial
Pediatrics
 , 
2010
, vol. 
125
 (pg. 
e35
-
51
)
27.
Buchholz
UJ
Finke
S
Conzelmann
KK
Generation of bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissue culture, and the human RSV leader region acts as a functional BRSV genome promoter
J Virol
 , 
1999
, vol. 
73
 (pg. 
251
-
9
)
28.
Jin
H
Clarke
D
Zhou
HZ
, et al.  . 
Recombinant human respiratory syncytial virus (RSV) from cDNA and construction of subgroup A and B chimeric RSV
Virology
 , 
1998
, vol. 
251
 (pg. 
206
-
14
)
29.
Lennette
EH
Schmidt
NJ
Diagnostic procedures for viral, rickettsial and chlamydial infections
 , 
1979
5th ed
Washington, DC
American Public Health Association Publications
30.
Zhao
X
Sullender
WM
In vivo selection of respiratory syncytial viruses resistant to palivizumab
J Virol
 , 
2005
, vol. 
79
 (pg. 
3962
-
8
)
31.
Domingo
E
Holland
JJ
Morse
SS
Mutation rates and rapid evolution of RNA viruses
The evolutionary biology of viruses
 , 
1994
New York, NY
Raven Press
(pg. 
161
-
84
)
32.
Crowe
JE
Jr
Gilmour
PS
Murphy
BR
, et al.  . 
Isolation of a second recombinant human respiratory syncytial virus monoclonal antibody fragment (Fab RSVF2-5) that exhibits therapeutic efficacy in vivo
J Infect Dis
 , 
1998
, vol. 
177
 (pg. 
1073
-
6
)
33.
Garcia-Barreno
B
Portela
A
Delgado
T
Lopez
JA
Melero
JA
Frame shift mutations as a novel mechanism for the generation of neutralization resistant mutants of human respiratory syncytial virus
EMBO J
 , 
1990
, vol. 
9
 (pg. 
4181
-
7
)
34.
Lopez
JA
Penas
C
Garcia-Barreno
B
Melero
JA
Portela
A
Location of a highly conserved neutralizing epitope in the F glycoprotein of human respiratory syncytial virus
J Virol
 , 
1990
, vol. 
64
 (pg. 
927
-
30
)
35.
Sullender
WM
Edwards
KG
Mutations of respiratory syncytial virus attachment glycoprotein G associated with resistance to neutralization by primate polyclonal antibodies
Virology
 , 
1999
, vol. 
264
 (pg. 
230
-
6
)
36.
Walsh
EE
Falsey
AR
Sullender
WM
Monoclonal antibody neutralization escape mutants of respiratory syncytial virus with unique alterations in the attachment (G) protein
J Gen Virol
 , 
1998
, vol. 
79
 
Pt 3
(pg. 
479
-
87
)
37.
Carman
WF
Zanetti
AR
Karayiannis
P
, et al.  . 
Vaccine-induced escape mutant of hepatitis B virus
Lancet
 , 
1990
, vol. 
336
 (pg. 
325
-
9
)
38.
Nara
PL
Smit
L
Dunlop
N
, et al.  . 
Emergence of viruses resistant to neutralization by V3-specific antibodies in experimental human immunodeficiency virus type 1 IIIB infection of chimpanzees
J Virol
 , 
1990
, vol. 
64
 (pg. 
3779
-
91
)
39.
Protzer-Knolle
U
Naumann
U
Bartenschlager
R
, et al.  . 
Hepatitis B virus with antigenically altered hepatitis B surface antigen is selected by high-dose hepatitis B immune globulin after liver transplantation
Hepatology
 , 
1998
, vol. 
27
 (pg. 
254
-
63
)
40.
Groothuis
JR
Nishida
H
Prevention of respiratory syncytial virus infections in high-risk infants by monoclonal antibody (palivizumab)
Pediatr Int
 , 
2002
, vol. 
44
 (pg. 
235
-
41
)
41.
McLellan
JS
Chen
M
Kim
A
Yang
Y
Graham
BS
Kwong
PD
Structural basis of respiratory syncytial virus neutralization by motavizumab
Nat Struct Mol Biol
 , 
2010
, vol. 
17
 (pg. 
248
-
50
)
42.
Carsillo
T
Zhang
X
Vasconcelos
D
Niewiesk
S
Oglesbee
M
A single codon in the nucleocapsid protein C terminus contributes to in vitro and in vivo fitness of Edmonston measles virus
J Virol
 , 
2006
, vol. 
80
 (pg. 
2904
-
12
)
43.
Novella
IS
Ball
LA
Wertz
GW
Fitness analyses of vesicular stomatitis strains with rearranged genomes reveal replicative disadvantages
J Virol
 , 
2004
, vol. 
78
 (pg. 
9837
-
41
)

Author notes

Potential conflicts of interest: All authors are employees of MedImmune.
Presented in part: 2010 Pediatric Academic Societies Annual Meeting, Vancouver, BC, Canada, 1–4 May 2010 (Abstract 2869.563).