Subfamily: Parvovirinae

Genus: Erythroparvovirus


Distinguishing features

Members of this genus are primarily distinguished by their monophyletic position in the subfamily tree (Figure 6B. Parvoviridae) and by sequence identity criteria. However, because the alignment methods in use have changed since these guidelines were adopted, less than the standard 30% identity in NS1 protein sequence is permitted to accommodate viruses in the recognized rodent and ungulate erythroparvovirus species. Genomes are homotelomeric, ~5.5 kb, and are bracketed by terminal repeats (TRs) that end in long (~365 nt) hairpin telomeres. Several viruses, including the exemplar virus human parvovirus B19 (B19V), preferentially target human erythroid progenitor cells. 


See discussion under family description and Figures 1.Parvoviridae and 2. Parvoviridae therein The N-terminal VP1-specific region (VP1SR) of B19V differs from those encoded by most parvoviruses in being unusually long (227 amino acids) and by being positioned on the outside of infectious virions before entering cells. It shows relatively weak PLA2 enzymatic activity that is nevertheless essential for viral infectivity and is initially masked on the virion surface by a genus-specific N-terminal extension (Bönsch et al., 2010, Filippone et al., 2008). However, when virions bind to P antigen globoside, their first cell surface receptor, a structural rearrangement occurs that exposes the PLA2 domain and allows the displaced N-terminal extension to bind to a second (unknown) cell surface molecule, which is essential for infectious entry and appears to be cell-lineage specific (Leisi et al., 2013, Leisi et al., 2016). X-ray structures of VP2-only recombinant B19V virus-like particles (VLPs) (Kaufmann et al., 2004) and cryo-EM image reconstructions of DNA-containing B19V virions and empty particles from human sera (Kaufmann et al., 2008) show that a conserved glycine-rich VP peptide, which has been observed within the channel in virions from some other genera, in B19V lies between neighbouring fivefold-related VP chains, effectively positioning most of the extreme VP2 N-termini on the particle surface next to the cylinder. These models also indicate that the B19V fivefold channel itself is relatively short and appears constricted at the outer viral surface, gated by five symmetry-related threonines. However, it is hypothesized that three glycine residues (amino acids 136–138) immediately N-terminal to the threonines could provide the structural flexibility required for switching the channel from closed to open during virion morphogenesis. Overall, these studies indicate that, in B19V, the fivefold channels do not mediate post-assembly VP1 or VP2 N-terminal peptide extrusion, but are rather confined to a role in genome translocation. 

Genome organization and replication

B19V coding sequences were first cloned and sequenced in the early 1980s (Cotmore and Tattersall 1984, Shade et al., 1986) and for many years the same highly-conserved genotype (now called G1) was observed in western populations, but by 2002 two relatively rare variants had been reported, now called G2 and G3, which diverge in genome nucleotide sequence by ~10% (Nguyen et al., 1999, Hokynar et al., 2002, Nguyen et al., 2002, Servant et al., 2002). Previously it had been observed that following primary B19V infections, viral genomes commonly persist in solid tissues (Söderlund et al., 1997, Söderlund-Venermo et al., 2002, Hokynar et al., 2007), at least in part due to antibody mediated virus internalization by B-lymphocytes (Pyöriä et al., 2017). However, although G1 remains the predominant virus in circulation globally, both G1 and G2 forms could be found in solid tissue samples (in 25% and 11% of samples, respectively), with G1 occurring in tissues from all age groups whereas G2 was strictly confined to the tissues of subjects born before 1973 (Norja et al., 2006). This clearly suggested that G2 had been in circulation until the early 1970s, but had since been replaced by G1. Genomes retained in solid tissues were therefore dubbed the erythrovirus "bioportfolio", since they provide a permanent record of the viruses responsible for each individual's infectious history (Norja et al., 2006). In subsequent studies genotypes were assessed in the skeletal remains of World War II battle casualties from Finland, and found to be exclusively G2 (n=41) or G3 (n=2), indicating that G1 was likely absent in this area during the first half of the 20th century, while G2 was the major circulating virus. G3 appears to be a geographic variant that had previously been seen only in Ghana, Brazil and India, and both of the G3 tissue samples mentioned above were associated with human genotypic markers suggestive of non-European origins, likely reflecting the wider cultural diversity of Soviet armies (Toppinen et al., 2015). Where or when G1 arose and why it became pre-eminent remains uncertain, but to date there are no biological differences between viruses from the three genotypes and they all belong to the same serotype (Blümel et al., 2005, Ekman et al., 2007, Chen et al., 2009). 

The availability of a bacterial plasmid carrying a full-length infectious B19V G1 genome (Zhi et al., 2004) has greatly facilitated laboratory analysis of its molecular biology. This homotelomeric genome is 5,596 nt, with long (383 nt) terminal repeats (TRs) that end in imperfectly palindromic hairpins of 365 nt (Figure 3B. Parvoviridae). The integrity of the hairpins is essential for viral infectivity, although why they are so complex remains unclear. However, it is known that the signal transducer and activator of transcription 5 (STAT5), which plays an important role in viral DNA replication, specifically interacts with the TRs (Ganaie et al., 2017). The genome has a single transcriptional promoter (P6), which gives rise to one full-length pre-mRNA, and two polyadenylation signals, one corresponding to the middle of the DNA (p(A)p) and the other (p(A)d) near its right end. The single pre-mRNA is alternatively spliced at one or two introns using a total of 2 donor and 4 acceptor sites, generating 12 viral mRNAs that encode the replication initiator protein (NS1), 2 structural proteins (VP1 and VP2) and two ancillary proteins (7.5 kDa and 11 kDa). Early in infection most transcripts are polyadenylated at the internal p(A)p site and encode NS1 or the 7.5 kDa protein, but after DNA replication becomes established, the p(A)p site is read through, allowing synthesis of full-length transcripts that encode the capsid and 11 kDa proteins (Guan et al., 2008). An intronic splice enhancer (ISE2) that contains a binding site for a cellular RNA binding protein (RBM38) lies immediately distal to the D2 donor. RBM38 expression during erythropoiesis makes it available to bind to ISE2, leading to enhanced recognition of the D2 splice site and high level expression of the 11 kDa protein (Ganaie et al., 2018).  This 11 kDa ancillary protein is known to be a potent inducer of apoptosis in erythroid progenitor cells (Chen et al., 2010b) and is essential for optimal viral DNA replication and virion release (Ganaie et al., 2018), whereas the function of the 7.5 kDa protein remains uncertain. 


B19V has an exceptionally narrow tissue tropism that in culture restricts its productive replication to a short time period following the differentiation of human bone marrow CD34+ stem cells into CD36+ erythroid progenitors (EPCs) (reviewed in detail in (Qiu et al., 2017)). It can also replicate productively, albeit much less efficiently, in a human megakaryoblastoid cell line, UT7/Epo-S1. The viability of both of these productive cell types depends upon access to erythropoietin (Epo), and Epo/Epo-receptor (Epo-R) signalling plays a critical role in promoting B19V infection via activation of Janus kinase 2 (Jak2) pathways. Jak2 further expands Epo-R phosphorylation and initiates a kinase cascade that activates STAT5A transcription and down-regulates signalling by mitogen-activated protein kinase (MEK/ERK), both of which lead to enhanced virus production. Culturing cells under hypoxic conditions (1% O2) to mimic the environment in human bone marrow, also significantly increases B19V DNA replication and progeny virus production (Pillet et al., 2004), although in EPCs this acts by regulating EpoR signalling rather than by the more common HIF-1α pathway (Luo and Qiu 2015). B19V infection of EPCs also induces a DNA damage response (DDR) with activation of all three phosphatidylinositol 3-kinase-related kinases (PI3KKs). The virus hijacks the induced ATR and the DNA-PKcs pathways to promote viral DNA amplification, inducing cell cycle arrest in late S phase that allows the cell's resources to be diverted for the replication of viral DNA (Luo and Qiu 2015, Zou et al., 2018). 

In children, B19V infection of EPCs commonly manifests as an immune complex exanthema called “fifth” disease, also known as erythema infectiosum or “slapped-cheek” syndrome, while in adults (especially women) polyarthralgia is common. In vulnerable populations a range of additional clinical disorders may occur. For example, EPC disfunction can cause persistent anaemia in immunosuppressed individuals, transient aplastic crisis in patients who require increased erythropoiesis (e.g. in sickle cell disease), or chronic pure red cell aplasia in congenitally immune-compromised patients. The virus can also cross the placenta, sometimes causing hydrops fetalis in developing 2nd trimester fetuses. Clinical observations suggest that B19V could also be implicated in hepatic or cardiovascular diseases such as myocarditis, certain autoimmune conditions and chronic fatigue syndrome, possibly by being taken into and perturbing non-productive cell types in these conditions, although how the virus induces such pathology requires further study (Qiu et al., 2017, Luo and Qiu 2015, Kerr 2016). 

Although little is known about the biology of other erythroparvoviruses, those that infect simian, pig-tailed or rhesus macaques all show a predilection for the bone marrow and can induce significant anemia in immunosuppressed animals (Brown and Young 1997, Green et al., 2000), suggesting that they may resemble B19V in their cell type specificities. 


See discussion under family description. The VP1 specific region of B19V is located on the outside of the virus and serves as a major antigenic target for neutralizing antibodies (Anderson et al., 1995). 

Species demarcation criteria

Viruses within a species are monophyletic and encode replication initiator proteins (called NS1 or Rep1, 68, or 78) that show >85% amino acid sequence identity.