Fetal inheritance of chromosomally integrated human herpesvirus 6 predisposes the mother to pre-eclampsia

Abstract

Pre-eclampsia (typically characterized by new-onset hypertension and proteinuria in the second half of pregnancy) represents a major determinant of the global burden of disease^1,2. Its pathophysiology involves placental dysfunction, but the mechanism is unclear. Viral infection can cause organ dysfunction, but its role in placentally related disorders of human pregnancy is unknown³. We addressed this using RNA sequencing metagenomics^4-6 of placental samples from normal and complicated pregnancies. Here, we show that human herpesvirus 6 (HHV-6, A or B) RNA was detected in 6.1% of cases of pre-eclampsia and 2.2% of other pregnancies. Fetal genotyping demonstrated that 70% of samples with HHV-6 RNA in the placenta exhibited inherited, chromosomally integrated HHV-6 (iciHHV-6). We genotyped 467 pre-eclampsia cases and 3,854 controls and found an excess of iciHHV-6 in the cases (odds ratio of 2.8, 95% confidence intervals of 1.4-5.6, P = 0.008). We validated this finding by comparing iciHHV-6 in a further 740 cases with controls from large-scale population studies (odds ratio of 2.5, 95% confidence intervals of 1.4-4.4, P = 0.0013). We conclude that iciHHV-6 results in the transcription of viral RNA in the human placenta and predisposes the mother to pre-eclampsia.

Extended Data Fig. 1. Characteristics of the study groups in the Pregnancy Outcome Prediction (POP) study.

Data are expressed as median (IQR) or n (%) as appropriate. The overall rate of preeclampsia in these participants was 6.4%. For fields where there is no category labelled “missing”, data were 100% complete. Maternal age was defined as age at recruitment. All other maternal characteristics were defined by self-report at the 20 weeks questionnaire, from examination of the clinical case record, or linkage to the hospital’s electronic databases. Socio-economic status was quantified using the Index of Multiple Deprivation (IMD) 2007, which is based on census data from the area of the mother’s postcode. Stillbirths (n=8) and spontaneous preterm deliveries (n=100) were included in the analysis, while miscarriages (n=7) and terminations of pregnancy (n=11) were excluded. Abbreviations: non-cases denote patients without preeclampsia; FTE denotes full time education; BMI denotes body mass index; DM denotes diabetes mellitus.

Extended Data Fig. 2. Placental RNA-seq reads mapped to HHV-6A and HHV-6B genomes.

Placental HHV-6 positive samples identified by RNA-seq. Reads were identified by Kraken as aligning to the HHV-6 genomes and mapped with BWA to the HHV-6A or HHV-6B reference genomes (Supplementary Methods); note that the total number of reads/sample recognized by the two software is not always identical. CON denotes a healthy pregnancy without FGR or preeclampsia (see Methods); FGR denotes fetal growth restriction; PE denotes a patient with preeclampsia; HHV-6A denotes human herpesvirus 6, variant A; HHV-6B denotes human herpesvirus 6, variant B; DRL: direct repeat left; DRR: direct repeat right. Repetitive regions are in Italic. HHV-6A and HHV-6B genomes have been described by Dominguez G et al.

Extended Data Fig. 3. HHV-6 detection in fetal and parental samples.

HHV-6A and HHV-6B representative signals in cord (A) and parental (B) DNA samples detected using a multiplex qPCR approach. These analyses were performed in 5,061 and 86 samples, respectively, and each sample was analyzed in triplicate. qPCR amplification curves for the HHV-6A and HHV-6B 9 U67/68 genes are represented in green and red, respectively; RNase P curves are in blue and confirmed presence of DNA in the wells. ciHHV-6 corresponds to a high HHV-6 DNA signal in the sample measured by qPCR, i.e. within 4 cycles of the RNase P signal. HHV-6 non-integrated corresponds to a HHV-6 DNA signal in the sample detected at more than 4 Ct higher compared to the RNase P signal. Negative samples lack HHV-6A or HHV-6B DNA signal. C) RT-qPCR amplification plot of placental RNA samples showing detection of the HHV-6 U100 gene. Eight representative samples are shown, two with viral transcript amplification (total n=48 samples, each analyzed in triplicate). Five negative controls (samples without reverse transcriptase enzyme in the RT reaction) lacked U100 amplification (not shown). U100 curves are in black and RNase P curves are in blue. Rn (normalized reporter value) represents the fluorescence of the reporter dye normalized to the signal of the passive reference dye for a given reaction. The ΔRn is the Rn value of an experimental reaction minus the Rn value of the baseline signal generated by the instrument. This parameter indicates the magnitude of the fluorescent signal generated in the qPCR assay. ciHHV-6 denotes chromosomally integrated human herpesvirus 6; HHV-6A denotes human herpesvirus 6, variant A; HHV-6B denotes human herpesvirus 6, variant B; RNase P denotes the human positive control gene RPPH1.

Extended Data Fig. 4. Identification of informative HHV-6B SNPs.

DNA-seq reads of 2 randomly selected samples were compared to the HHV-6B reference genome. *Informative SNP sites, i.e. SNPs present in just one of the two analyzed samples (gapped vertical red lines). **SNPs present in both analyzed samples, i.e. sites concordantly different from the reference genome, were considered not informative (continuous vertical red lines). Throughout the 162kb HHV-6B genome 187 SNPs were classified as informative.

Figure 1. SNP analysis of HHV-6B genome sequenced in fetal and parental samples.

A) Comparison of HHV-6B genomes sequenced in different samples revealed 187 informative SNPs, represented as black or white lines in the “barcode” graph if concordant or discordant to the HHV-6B reference genome (GCF_000846365.1), respectively. B) Comparison of placental and parental SNPs based on RNA-seq and DNA-seq reads aligning to the HHV-6B gene U100, which codes for glycoprotein Q (gQ). The genes of the HHV-6B reference genome are in light green in the upper part of the panel, and the U100 gene is indicated in dark green; repeat regions, including the two large direct repeat regions on both termini, are indicated in amber. The RNA-seq coverage of the U100 gene (highlighted in light-red and enlarged) is shown for one placental sample (Infant 2 in Table 1 and Extended Data Figure 2) and represented by the surface area of the black peaks. The DNA-seq reads obtained from the corresponding paternal sample are shown in the bottom part of the panel. Red vertical lines indicate positions where both the infant RNA and the paternal DNA concordantly differ from the HHV-6B reference genome.

Figure 2. Fetal inheritance of ciHHV-6 and the risk of preeclampsia.

A) Proportion of iciHHV-6 positive samples in preeclampsia cases (black bars) and non-cases (white bars) in the following datasets: placental RNA samples within the POP study analyzed by RNA-seq (“RNA-seq”, n=279); all samples analyzed by cord DNA genotyping, excluding those in the first group, i.e. with RNA-seq data available (“Other”, n=4,042); the combined study population, i.e. all the genotyped cord DNA samples from both studies (“All”, n=4,321). n represents the number of patients analyzed in each group. For the 3 datasets “RNA-seq”, “Other” and “All” the P values for the association between fetal iciHHV-6 and the risk of preeclampsia were 0.14, 0.022 and 0.008, respectively, using the recommended Fisher-Boschloo unconditional exact test (2-sided); and 0.17, 0.026 and 0.008, respectively, using the more widely used Fisher’s exact test (2-sided). B) Odds ratios and 95% confidence intervals (CI) for preeclampsia by the presence of iciHHV-6 calculated in each dataset and in the pooled study population. C) Proportion of iciHHV-6 positive samples in the following datasets: large-scale population studies (n>1,000) of healthy/control patients currently available in the literature (white symbols; see Methods); pooled values of the healthy/control populations (dataset “Total”, grey symbol); GOPEC samples (dataset “GOPEC”, black symbol). Symbols represent the proportion of iciHHV-6 positive samples with 95% CI; n represents the number of patients involved in each study; the location where the studies were conducted is in parenthesis.