Intrahost Dynamics of Human Cytomegalovirus Variants Acquired by Seronegative Glycoprotein B Vaccinees - PubMed (original) (raw)

Clinical Trial

. 2019 Feb 19;93(5):e01695-18.

doi: 10.1128/JVI.01695-18. Print 2019 Mar 1.

Affiliations

Clinical Trial

Intrahost Dynamics of Human Cytomegalovirus Variants Acquired by Seronegative Glycoprotein B Vaccinees

Cody S Nelson et al. J Virol. 2019.

Abstract

Human cytomegalovirus (HCMV) is the most common congenital infection worldwide and a frequent cause of hearing loss and debilitating neurologic disease in newborn infants. Thus, a vaccine to prevent HCMV-associated congenital disease is a public health priority. One potential strategy is vaccination of women of child bearing age to prevent maternal HCMV acquisition during pregnancy. The glycoprotein B (gB) plus MF59 adjuvant subunit vaccine is the most efficacious tested clinically to date, demonstrating 50% protection against primary HCMV infection in a phase 2 clinical trial. Yet, the impact of gB/MF59-elicited immune responses on the population of viruses acquired by trial participants has not been assessed. In this analysis, we employed quantitative PCR as well as multiple sequencing methodologies to interrogate the magnitude and genetic composition of HCMV populations infecting gB/MF59 vaccinees and placebo recipients. We identified several differences between the viral dynamics in acutely infected vaccinees and placebo recipients. First, viral load was reduced in the saliva of gB vaccinees, though not in whole blood, vaginal fluid, or urine. Additionally, we observed possible anatomic compartmentalization of gB variants in the majority of vaccinees compared to only a single placebo recipient. Finally, we observed reduced acquisition of genetically related gB1, gB2, and gB4 genotype "supergroup" HCMV variants among vaccine recipients, suggesting that the gB1 genotype vaccine construct may have elicited partial protection against HCMV viruses with antigenically similar gB sequences. These findings suggest that gB immunization had a measurable impact on viral intrahost population dynamics and support future analysis of a larger cohort.IMPORTANCE Though not a household name like Zika virus, human cytomegalovirus (HCMV) causes permanent neurologic disability in one newborn child every hour in the United States, which is more than that for Down syndrome, fetal alcohol syndrome, and neural tube defects combined. There are currently no established effective measures to prevent viral transmission to the infant following HCMV infection of a pregnant mother. However, the glycoprotein B (gB)/MF59 vaccine, which aims to prevent pregnant women from acquiring HCMV, is the most successful HCMV vaccine tested clinically to date. Here, we used viral DNA isolated from patients enrolled in a gB vaccine trial who acquired HCMV and identified several impacts that this vaccine had on the size, distribution, and composition of the in vivo viral population. These results have increased our understanding of why the gB/MF59 vaccine was partially efficacious, and such investigations will inform future rational design of a vaccine to prevent congenital HCMV.

Keywords: DNA sequencing; glycoprotein B; human cytomegalovirus; vaccine.

Copyright © 2019 American Society for Microbiology.

PubMed Disclaimer

Figures

FIG 1

FIG 1

Reduced peak saliva shedding, yet similar numbers of viral haplotypes and nucleotide diversity between HCMV-infected glycoprotein B vaccinees and placebo recipients. Peak plasma viral loads (A) as well as the peak magnitudes of virus shed in saliva (B), urine (C), and vaginal fluid (D) were compared between 11 gB vaccinees and 22 placebo recipients (samples not available from each compartment for each participant). Plasma (A), urine (C), and vaginal fluid (D) viral loads were not statistically different between HCMV-infected placebo recipients and gB/MF59 vaccinees, though there was reduced HCMV shedding in the saliva of vaccinees (B). Using SNAPP NGS data, the peak number of unique viral haplotypes (E) as well as peak nucleotide diversity (π) (F) were assessed for viral DNA amplified at the gB locus for 3 tissue culture virus isolates (TC virus), 13 placebo recipients, 5 gB/MF59 vaccinees, and 4 seropositive chronically HCMV-infected individuals (Sero+). (G) The magnitudes of nucleotide diversity resulting in synonymous (πS) versus nonsynonymous changes (πN) were compared. Horizontal bars indicate the median values for each group. *, P < 0.05 according to statistical tests: viral load, exact Wilcoxon rank sum test; haplotypes and π, Kruskal-Wallis test plus post hoc exact Wilcoxon rank sum test; πS versus πN, Wilcoxon signed-rank test.

FIG 2

FIG 2

High-magnitude viral shedding in vaginal fluid, yet similar numbers of unique viral variants and nucleotide diversity between anatomic compartments. Peak viral loads were compared between anatomic compartments for 11 gB/MF59 vaccinees, 22 placebo recipients, and 4 seropositive individuals (A), revealing high-magnitude HCMV shedding in vaginal fluid. Using SNAPP NGS data, the peak numbers of unique viral haplotypes (B) and peak nucleotide diversity (π) (C) were defined according to individuals for each anatomic compartment, resulting in data from 8 whole blood, 11 saliva, 10 urine, and 9 vaginal fluid samples from acutely infected gB vaccinees and placebo recipients as well as chronically HCMV-infected individuals. (G) The magnitudes of nucleotide diversity resulting in synonymous (πS) versus nonsynonymous changes (πN) were compared. Horizontal bars indicate the median values for each group. *, P < 0.05 according to statistical tests: viral load, Friedman test plus post hoc pairwise Wilcoxon signed-rank test; haplotypes and π, Kruskal-Wallis test plus post hoc exact Wilcoxon rank sum test; πS versus πN, Wilcoxon signed-rank test.

FIG 3

FIG 3

Large numbers of low-frequency viral variants detected at gB locus in both primary HCMV-infected and chronically infected individuals. The relative frequency of each unique gB haplotype identified by SNAPP is displayed by individual patient and time point of sample collection. Tissue culture viruses (A) exhibited reduced population complexity by comparison. In primary HCMV-infected placebo recipients (C) and gB vaccinees (D), as well as chronically HCMV-infected women (B), there were typically one or more high-frequency haplotypes representing the dominant viral variants within the population, which were accompanied by haplotypes at very low frequency representing minor viral variants (<1% of viral haplotype prevalence). All haplotypes displayed exceeded the 0.44% threshold of PCR and sequencing error established for the SNAPP method (see Materials and Methods for detail).

FIG 4

FIG 4

Evidence of viral genetic compartmentalization at gB locus in 3 of 4 gB vaccinees. (A) Table indicating the results of 6 distinct tests of genetic compartmentalization performed on the pool of unique gB haplotypes identified per patient, including Wright’s measure of population subdivision (FST), the nearest-neighbor statistic (Snn), the Slatkin-Maddison test (SM), the Simmonds association index (AI), and correlation coefficients based on distance between sequences (r) or number of phylogenetic tree branches (_r_b). For each test, >1,000 permutations were simulated. Significant test results suggesting genetic compartmentalization are shown in gray with bold text. Values for FST, Snn, SM, r, and _r_b represent uncorrected P values, with P < 0.05 considered significant. An AI of <0.3 was considered a significant result. The presence of three or more positive tests per patient was considered strong evidence for genetic compartmentalization, indicated in green. UP, underpowered (fewer than 5 haplotypes were present in each compartment, making FST and Snn error prone). (B to D) Networks of unique viral haplotypes by individual patient, with 1 patient lacking tissue compartmentalization (B) and 2 patients demonstrating strong evidence of viral genetic compartmentalization (D). Samples are organized chronologically from left to right, with blood shown in red, saliva in blue, urine in yellow, and vaginal fluid in purple. The size of each node reflects the relative prevalence of each haplotype. Light blue lines connect identical viral variants between time points and compartments, green lines connect variants with a synonymous mutation, and red lines those with a nonsynonymous mutation.

FIG 5

FIG 5

Possible protection against gB1/2/4 genotype supergroup viruses in gB/MF59 vaccinees. (A) Unrooted phylogenetic tree in a polar layout constructed using full gB open reading frame consensus sequences for each sequenced sample. Note, a consensus sequence was not defined for all samples from each subject, so this tree depicts multiple samples representing multiple time points/compartments for each trial participant. Samples with discrepant genotypes from Sanger sequencing not excluded. Clades representing gB genotypes are highlighted in different colors: gB1, purple; gB2, yellow; gB3, blue; gB4, red; gB5, green. (B) Numbers of distinct gB vaccinees and placebo recipients who acquired viral variants belonging to two supergroups of genetically similar gB genotypes (gB1/2/4 and gB 3/5) as defined by Sanger sequencing of the cleavage site (highlighted in green) and NGS of the full gB protein (highlighted in blue). (C) Force-of-infection modeling closely predicts observed gB/MF59 vaccine trial efficacy (D). Force-of-infection model iterations indicate that gB vaccinees are universally protected against acquisition of gB1/2/4 variants (blue line; consistent with full gB NGS data) or 50% protected (green line; consistent with Sanger sequencing data). Model makes assumption that (i) viruses belonging to the gB1/2/4 genotype supergroup represent 52% of the circulating virus pool and (ii) the HCMV force of infection is 5.7 per 100 person-years. (Panel D adapted from The New England Journal of Medicine [5].)

References

    1. Dollard SC, Grosse SD, Ross DS. 2007. New estimates of the prevalence of neurological and sensory sequelae and mortality associated with congenital cytomegalovirus infection. Rev Med Virol 17:355–363. doi: 10.1002/rmv.544. -DOI -PubMed
    1. Manicklal S, Emery VC, Lazzarotto T, Boppana SB, Gupta RK. 2013. The “silent” global burden of congenital cytomegalovirus. Clin Microbiol Rev 26:86–102. doi: 10.1128/CMR.00062-12. -DOI -PMC -PubMed
    1. Ramanan P, Razonable RR. 2013. Cytomegalovirus infections in solid organ transplantation: a review. Infect Chemother 45:260–271. doi: 10.3947/ic.2013.45.3.260. -DOI -PMC -PubMed
    1. Stratton KR, Durch JS, Lawrence RS (ed). 2000. Vaccines for the 21st century: a tool for decisionmaking. National Academies Press, Washington, DC. -PubMed
    1. Pass RF, Zhang C, Evans A, Simpson T, Andrews W, Huang ML, Corey L, Hill J, Davis E, Flanigan C, Cloud G. 2009. Vaccine prevention of maternal cytomegalovirus infection. N Engl J Med 360:1191–1199. doi: 10.1056/NEJMoa0804749. -DOI -PMC -PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources