Recurrent Respiratory Syncytial Virus Infection in a CD14-Deficient Patient

Abstract

Background: Recurrent respiratory syncytial virus (RSV) infection requiring hospitalization is rare and the underlying mechanism is unknown. We aimed to determine the role of CD14-mediated immunity in the pathogenesis of recurrent RSV infection.

Methods: We performed genotyping and longitudinal immunophenotyping of the first patient with a genetic CD14 deficiency who developed recurrent RSV infection. We analyzed gene expression profiles and interleukin (IL)-6 production by patient peripheral blood mononuclear cells in response to RSV pre- and post-fusion (F) protein. We generated CD14-deficient human nasal epithelial cells cultured at air-liquid interface (HNEC-ALI) of patient-derived cells and after CRISPR-based gene editing of control cells. We analyzed viral replication upon RSV infection.

Results: Sanger sequencing revealed a homozygous single-nucleotide deletion in CD14, resulting in absence of the CD14 protein in the index patient. In vitro, viral replication was similar in wild-type and CD14-/- HNEC-ALI. Loss of immune cell CD14 led to impaired cytokine and chemokine responses to RSV pre- and post-F protein, characterized by absence of IL-6 production.

Conclusions: We report an association of recurrent RSV bronchiolitis with a loss of CD14 function in immune cells. Lack of CD14 function led to defective immune responses to RSV pre- and post-F protein without a change in viral replication.

Keywords: CD14; Toll-like receptor; epithelium; monocyte; respiratory syncytial virus.

Figure 1.

Overview of the patient’s medical history. Timeline of critical events in the patient’s medical history, according to his age in months. CoVOC43, human coronavirus OC43; CRP, C-reactive protein; RSV, respiratory syncytial virus; RTI, respiratory tract infection; resp, respiratory.

Figure 2.

Phenotypic and genetic analysis of CD14 deficiency. (a) Respiratory syncytial virus (RSV) cycle threshold (Ct) value of the index patient during the first RSV infection and of 9 controls admitted with RSV at the pediatric intensive care unit in the same season. Mean + standard deviation is shown for the controls. (b) Genomic deoxyribonucleic acid sequence of CD14 comparing the patient (CD14^−/−) with his parents (both CD14^+/−) and sister (wild type). The band indicates the deletion of C in the index patient. In the family pedigree, circles and squares denote female and male family members, black dots represent a heterozygous mutation in CD14, and the black square represents the index patient. (c) Flow cytometry plot of the index patient. Cells were gated based on their forward and side scatter (SSC) properties, and single, nonapoptotic cells were selected. Monocytes were identified a CD14⁺/CD16⁻. (d) Plasma levels of soluble CD14 (sCD14) in the index patient compared to control infants (n = 6) and RSV infants (n = 10). Levels of lipopolysaccharide-binding protein (LBP) during and after the first RSV episode in the index patient, compared to control infants (n = 6) and RSV infants (n = 10). An unpaired t test was performed to compare expression between patient groups. Only significant values are shown. *, P = <.05; ****, P = <.0001. (^−/−), genetic knock out.

Figure 3.

Impaired innate immune response to Toll-like receptor (TLR) 1/2, 2/6, 4, and 5 stimulation in CD14 deficiency. (a) Gene expression by peripheral blood mononuclear cells (PBMCs) after 4 hours of stimulation with a panel of TLR ligands, quantified using NanoString Technology. Log₂ normalized counts are represented for the index patient (top) and a healthy control (bottom). Upregulated genes are depicted in gray, and downregulated genes are depicted in black. Each row represents a stimulus. Each column represents 1 gene. In Supplemental Figure S9, a large version of the heatmap can be found. (b–d) Expression of the top 20 response genes measured in a, after stimulation with lipopolysaccharide (LPS) (Escherichia coli [E. coli]) (10 ng/mL), R848 (1 µg/mL), and respiratory syncytial virus (RSV)-pre-fusion (pre-F) (10 µg/mL). Fold change compared with unstimulated is shown for a healthy control ([HC] black bars) and the index patient (gray bars). (e) Interleukin (IL)-6 protein expression by patient and control PBMCs (n = 4 donors) after TLR stimulation. Values show the mean + standard deviation of 2 independent experiments.

Figure 4.

Airway epithelial derived CD14 does not inhibit respiratory syncytial virus (RSV) infection. (a) Schematic view of RSV infection of human nasal epithelial cells cultured at air liquid interface (HNEC-ALI). The HNEC-ALI was derived from healthy control (HC) children (n = 2), the index patient (IP), wild type ([WT] n = 4), and CD14 (n = 4). (b) Cultures were infected with RSV-A2-GFP. Cytokine production of soluble (s)CD14 at day 1, 2, and 3 postinfection in cultures of the index patient and controls is shown. No significant difference in cycle threshold (Ct) value was observed between the index patient and controls (P > .05 at both time points). (c) Plots show mRNA expression of CD14. We confirmed the absence of CD14 mRNA expression in CD14 knock out (CD14^−/−) cultures. (d) Cultures were infected with RSV-A2-GFP. Cytokine production of soluble CD14 (sCD14) at day 1, 2, and 3 postinfection in WT and CD14^−/− cultures is shown. No significant differences in cycle threshold (Ct) value were observed between WT and CD14^−/− cultures (P > .05 for all time points). Values show the mean + standard deviation of 4 donors. The experiment using patient-derived HNEC-ALI were performed twice. The experiments using HNEC-ALI with WT and CD14^−/− was performed twice, using 2 individual donors per experiment. All experiments were performed in duplicate. An unpaired t test was performed to compare expression (1) between the index patient and controls and (2) between WT and CD14^−/−. Only significant values are shown.

Figure 5.

CD14 expression on monocytes mediates the interleukin (IL)-6 response to respiratory syncytial virus fusion (RSV-F). (a) The IL-6 protein response after 4, 18, and 24 hours of stimulation by RSV-pre-F and post-F (10 µg/mL), by peripheral blood mononuclear cells of 2 healthy controls (HC) and the index patient (IP). One experiment, performed in duplicates, lines indicate mean per individual (+standard deviation). (b) The IL-6 protein response during 24 hours of stimulation by RSV-pre-F (5 µg/mL) by Percoll-isolated monocytes of 2 healthy controls, the index patient, and his father. One experiment, lines indicate mean per individual.

Figure 6.

Impaired interleukin (IL)-6 protein response to respiratory syncytial virus fusion (RSV-F) after blocking CD14 in healthy adult donor monocytes. The figure shows IL-6 protein production after stimulation of healthy donor monocytes with RSV pre-F (a) or post-F (b) at different concentrations, by healthy donor monocytes, after Percoll isolation, in the presence of CD14 blocking antibodies (aCD14), control antibody (aMHC-1), or no antibody. The lines indicate the mean (+standard deviation) of 3 independent experiments with a total of 7 donors.

Figure 7.

Normal pre-fusion (F)-specific antibody response in CD14 deficiency. (a) Seroconversion of respiratory syncytial virus (RSV)-specific antibodies after the first RSV infection in the index patient. Depicted are the concentrations immunoglobulin (Ig)G against Ga, Gb, post-F, N, and pre-F in arbitrary units/mL at different time points during and in between RSV infections. (b) Percentage pre-F-specific IgG bound after incubation with 1.5M NH4SCN (left graph) and 3M NH4SCN (right graph), at the time points indicated in a. (c) Conformation-specific RSV A and B neutralizing activity in sera obtained at 14, 22, and 30 months, in the absence and presence of excess post-F. Bars represent mean neutralizing activity normalized to International Units of Neutralization. As a reference, the mean neutralization of 15 adult controls is depicted. (d) B-cell specific binding to RSV-pre- and post-F, 8 months after the second RSV infection and 6 months after third RSV infection. Bars show assembled percentage of total IgG + B cells that bind pre-F, post-F, or both probes (dual binding) in the index patient. The interquartile range of healthy controls is depicted in gray. N.D., not determined due to too little available serum; RIVM, Rijksinstituut voor Volksgezondheid en Milieu (National Institute for Public Health and the Environment).