Protective effect of pre-existing natural immunity in a nonhuman primate reinfection model of congenital cytomegalovirus infection

Abstract

Congenital cytomegalovirus (cCMV) is the leading infectious cause of neurologic defects in newborns with particularly severe sequelae in the setting of primary CMV infection in the first trimester of pregnancy. The majority of cCMV cases worldwide occur after non-primary infection in CMV-seropositive women; yet the extent to which pre-existing natural CMV-specific immunity protects against CMV reinfection or reactivation during pregnancy remains ill-defined. We previously reported on a novel nonhuman primate model of cCMV in rhesus macaques where 100% placental transmission and 83% fetal loss were seen in CD4+ T lymphocyte-depleted rhesus CMV (RhCMV)-seronegative dams after primary RhCMV infection. To investigate the protective effect of preconception maternal immunity, we performed reinfection studies in CD4+ T lymphocyte-depleted RhCMV-seropositive dams inoculated in late first / early second trimester gestation with RhCMV strains 180.92 (n = 2), or RhCMV UCD52 and FL-RhCMVΔRh13.1/SIVgag, a wild-type-like RhCMV clone with SIVgag inserted as an immunological marker, administered separately (n = 3). An early transient increase in circulating monocytes followed by boosting of the pre-existing RhCMV-specific CD8+ T lymphocyte and antibody response was observed in the reinfected dams but not in control CD4+ T lymphocyte-depleted dams. Emergence of SIV Gag-specific CD8+ T lymphocyte responses in macaques inoculated with the FL-RhCMVΔRh13.1/SIVgag virus confirmed reinfection. Placental transmission was detected in only one of five reinfected dams and there were no adverse fetal sequelae. Viral whole genome, short-read, deep sequencing analysis confirmed transmission of both reinfection RhCMV strains across the placenta with ~30% corresponding to FL-RhCMVΔRh13.1/SIVgag and ~70% to RhCMV UCD52, consistent with the mixed human CMV infections reported in infants with cCMV. Our data showing reduced placental transmission and absence of fetal loss after non-primary as opposed to primary infection in CD4+ T lymphocyte-depleted dams indicates that preconception maternal CMV-specific CD8+ T lymphocyte and/or humoral immunity can protect against cCMV infection.

Fig 1. Study design and kinetics of CD4+ T lymphocyte depletion in experimental groups.

(A) Schematic of study design of cCMV transmission in pregnant CMV-seropositive rhesus macaques. (B-C) Peripheral blood CD4+ T lymphocyte counts following anti-CD4 antibody administration in (B) CMV-seropositive Reinfection group (n = 5); and (C) CMV-seropositive Control group (n = 3).

Fig 2. RhCMV viral kinetics in blood and body fluids of individual CD4+ T lymphocyte depeleted CMV-seropositive macaques.

RhCMV in plasma (indicated in red), saliva (blue), urine (orange), and amniotic fluid (purple) in (A) five CMV-seropositive reinfected animals and (B) three CMV-seropositive control animals. Plasma and amniotic fluid are reported in mean RhCMV DNA copy number/mL of sample fluid; saliva and urine are reported as mean RhCMV DNA copy number/μg of input DNA. In CMV-seropositive controls, the equivalent post-infection time-points on the x-axis are aligned concurrently with the CMV-seropositive Reinfection group. The black vertical lines indicate time of anti-CD4 antibody (CD4R1) and RhCMV inoculation. Animals JP01, KK24, and KB91 were inoculated with RhCMV UCD52 and FL-RhCMVΔRh13.1/SIVgag; animals 274–05 and 292–09 were inoculated with RhCMV 180.92; animals 234–07, 309–09, and 222–02 remained without a reinfection. The horizontal stippled line indicates the baseline mean RhCMV DNA copy number/μg of input DNA in either saliva (blue) or urine (orange).

Fig 3. Early immunophenotypic changes following RhCMV reinfection in CMV-seropositive macaques.

Immunophenotyping of circulating peripheral blood mononuclear cells in acute RhCMV reinfection. Plots show the kinetics of different lymphocyte subsets in three CMV-seropositive reinfected macaques (JP01, KB91, and KK24). Paired non-parametric Wilcoxon Signed Rank test comparing baseline prereinfection values with values at time-point of maximal change in the first 7 days post reinfection was performed.

Fig 4. CMV-specific CD8+ T lymphocyte memory responses to RhCMV immediate early (IE) proteins and exogenous SIV Gag protein in CD4+ T lymphocyte depleted RhCMV reinfected macaques.

(A) Paired IE-specific responses by CD107a expression and secretion of IFN-γ, IL-2, and TNF-α in four CMV-seropositive macaques reinfected with RhCMV UCD52 and FL-RhCMVΔRh13.1/SIVgag (n = 3) or RhCMV 180.92 (n = 1). Pre-reinfection responses were compared with responses at week 8–10 post RhCMV reinfection using paired t-test. (B) Polyfunctional SPICE analysis of IE-specific responses pre vs post RhCMV reinfection showing the proportion of four-functional, three-funtional, two-functional and single function responses. CD107a (blue arc), IFN-γ (red arc), IL-2 (orange arc), and TNF-α (green arc). Four-functional responses are displayed in white, three-functional responses in dark grey, two-functional response in light grey, and mono-functional responses in black. (C) Bar graph of the polyfunctional responses pre (grey) and post (black) RhCMV reinfection (n = 4) showing the frequency of memory CD8+ T lymphocytes responding to RhCMV IE peptides. The RhCMV IE-specific response was measured by intracellular cytokine staining after stimulation with RhCMV IE1 and / or IE2 peptide pools depending on the baseline immunodominant response in individual animals. Comparison of pre- and post reinfection Boolean responses were compared with the Wilcoxon rank sum test using SPICE v6 software.

Fig 5. RhCMV-specific antibody responses in CD4+ T lymphocyte depleted CMV-seropositive reinfected macaques.

(A) Kinetic data of anti-gB binding antibodies in graphs on individual animals showing log ED₅₀ of anti-gB binding titer (green line) superimposed on CD4+ lymphocytes/μL (red line). (B) Comparison of CMV-seropositive reinfected dams (n = 5) and CMV-seropositive controls (n = 3) by their anti-gB binding titer and fibroblast neutralization against RhCMV 180.92 at baseline preceding CD4+ lymphocyte depletion. (C) Difference from baseline value in anti-gB IgG ED₅₀ titers in the CMV-seropositive reinfection groups at each of the post-infection time-points compared with CMV-seropositive controls at equivalent post CD4-depletion time-points. (D) Difference from baseline value in neutralization titer in the CMV-seropositive reinfection groups at each of the post-infection time-points compared with CMV-seropositive controls at equivalent post CD4-depletion time-points. ED₅₀ = Effective Dose 50.

Fig 6. Evidence of reinfection in CD4+ T lymphocyte depleted FL-RhCMVΔRh13.1/SIVgag inoculated dams.

(A) RhCMV-specific (black line) and SIVgag-specific (grey line) real time PCR results in the saliva of one CMV-seropositive reinfected animal (KB91). All other animals were found negative for SIVgag DNA in saliva and urine. (B) Detection of SIV Gag-specific T lymphocyte responses measured longitudinally against a SIVmac239 Gag peptide pool in CMV-seropositive reinfected macaques (KB91, KK24, and JP01) inoculated with RhCMV UCD52 and FL-RhCMVΔRh13.1/SIVgag. Horizontal stippled line shows negative cut-off based on pre-reinfection values.

Fig 7. Protective effect of pre-existing immunity against congenital CMV transmission.

(A) Plasma RhCMV-specific PCR in CD4+ T lymphocyte depleted CMV-seronegative primary infected macaques (red; n = 6) compared to both CMV-seropositive reinfected (green; n = 5) and CMV-seropositive controls (grey; n = 3). Area Under the Curve (AUC) values of plasma RhCMV DNA between 0–99 days were compared between groups using the Man-Whitney test. P-values <0.05 denoted with a single * were considered significant. (B) Heatmap of RhCMV-specific DNA copy number in amniotic fluid in CMV-seronegative primary infected macaques (n = 6), CMV-seropositive reinfected (n = 5), and CMV-seropositive controls (n = 3). (C) Kaplan-Meir curve showing cCMV frequency based on RhCMV DNA detection in the amniotic fluid in CMV-seronegative primary infected macaques (n = 6), CMV-seropositive reinfected (n = 5), and CMV-seropositive controls (n = 3). (D) Kaplan-Meier curve showing fetal survival in CMV-seronegative primary infected macaques (n = 6), CMV-seropositive reinfected (n = 5), and CMV-seropositive controls (n = 3). Statistical comparisons by Log-rank (Mantel-Cox) test showing significance levels: * = <0.05 and ** = <0.01.