Extinct type of human parvovirus B19 persists in tonsillar B cells

Abstract

Parvovirus B19 (B19V) DNA persists lifelong in human tissues, but the cell type harbouring it remains unclear. We here explore B19V DNA distribution in B, T and monocyte cell lineages of recently excised tonsillar tissues from 77 individuals with an age range of 2-69 years. We show that B19V DNA is most frequent and abundant among B cells, and within them we find a B19V genotype that vanished from circulation >40 years ago. Since re-infection or re-activation are unlikely with this virus type, this finding supports the maintenance of pathogen-specific humoral immune responses as a consequence of B-cell long-term survival rather than continuous replenishment of the memory pool. Moreover, we demonstrate the mechanism of B19V internalization to be antibody dependent in two B-cell lines as well as in ex vivo isolated tonsillar B cells. This study provides direct evidence for a cell type accountable for B19V DNA tissue persistence.

Figure 1. Viral DNA copies in tonsillar tissue.

B19V- and EBV-DNA copies were measured by qPCR and normalized to cell numbers with the human single-copy gene RNase P. Top panels: median copy numbers of B19V DNA (a) and EBV DNA (b) in tonsillar cell suspensions resulting from mechanical homogenization (n=33) (black) or after collagenase treatment (n=33) (light grey). The viral copy numbers between the two preparations were significantly different for B19V (P<0.001), while not significant for EBV (P>0.05). Bottom panels: median copy numbers of B19V DNA (n=33) (c) and EBV DNA (n=24) (d) in T-, B- and monocytic cells following homogenization (black) or collagenase digestion (light grey). The viral loads in B cells were statistically significant between preparations for B19V (P<0.001) while not significant for EBV (P>0.05). Error bars indicate 95% confidence intervals and were determined by normal approximation. Statistical significance was calculated with Mann–Whitney U asymptotic sig. (two-sided test).

Figure 2. Antibody-dependent uptake of VP2-VLPs in B cell lines.

The mean fluorescence intensity (MFI) of VP2-VLP-AF647 was measured by flow cytometry in Raji (left column) and GM12878 cells (right column) after incubation with total purified B19V-positive (red line) or -negative IgGs (blue line, squares) (a,b). Pooled sera from B19V-IgG⁺ individuals either heat inactivated (red solid line) or non-heat inactivated (red dotted line) or pooled sera from B19V-IgG^- individuals either heat inactivated (blue solid line, squares) or non-heat inactivated (blue dotted line) were used (c,d). * Represents statistical significance (P<0.001) between positive and respective negative controls at a given concentration. The effect on viral uptake was evaluated before (grey bars) and after (black bars) blocking with an anti-CD32 antibody (e,f). The difference was statistically significant (Raji P<0.001, GM12878 P=0.05). Error bars represent s.d. (of four replicates). Statistical significance was calculated using Student's t-test.

Figure 3. Antibody-dependent enhancement of VP2-VLPs in primary B cells and in a monocytic cell line.

The mean fluorescence intensity (MFI) of VP2-VLP-AF647 was measured by flow cytometry in tonsillar B cells from two seronegative individuals (a,b) as well as in U937 cells (c) after incubation with B19V-IgG⁺ (black solid line) or B19V-IgG^- (black solid line, squares) total purified preparations. Error bars represent standard deviation (of four replicates). *Represents statistical significance (P<0.001) between positive and negative IgGs at a given concentration as calculated by Student's t-test.

Figure 4. Association of VP2-VLPs with EEA1-positive endosomes.

Maximum intensity projections of serial confocal optical sections through Raji (a) and GM12878 cells (d) incubated overnight with VP2-VLP-AF647 (red) and treated with trypsin. White arrows represent enhanced cytoplasmic VLP localization in EEA1-positive endosomes (green). Differential interference contrast (DIC) merged with DAPI (blue) images are shown. Areas of localization (yellow) (b,e) and normalized fluorescent intensity profiles in zoomed areas (c,f) indicate VLP internalization. Scale bars, 2 μm.

Figure 5. Antibody-dependent uptake of native B19V in B cell lines.

ADE was tested using viremic plasma in Raji (left column) and GM12878 (right column) cells in the presence of total purified B19V-positive or -negative IgGs (a,b). The effect on viral uptake was evaluated before (grey bars) and after (black bars) blocking with an anti-CD32 antibody (c,d). Error bars represent s.d. (of three replicates). *Represents statistical significance (P<0.001) between positive and respective negative controls, as calculated by Student's t-test.