Background

Congenital CMV infection is a leading cause of birth defects with an annual direct cost of one billion dollars in the US alone, and represents a global unmet medical need (Bristow et al., 2011; Griffiths et al., 2013; Manicklal et al., 2013; Saint Louis, 2016). This would be preventable with an effective vaccine or therapeutic targeting women in their reproductive years. Human cytomegalovirus (HCMV) can infect a wide range of cell types, but a major barrier in the field is that extended passage of clinically derived HCMV strains in fibroblast cells leads to a loss of viral tropism for other cell types (Waldman et al., 1991; Sinzger et al., 1999). In the late 1960s, several laboratory-adapted strains of CMV were serially passaged in fibroblasts–including the HCMV AD169, Towne, and Davis strains, as well as the Smith strain of murine CMV–and became some of the first tools used to study the molecular biology of CMV (Plotkin et al., 1975). These laboratory-adapted strains–often created during unsuccessful attempts to generate a live attenuated vaccine–were found to have a number of mutations affecting (i) their ability to infect different cell types, (ii) rate of viral replication, and (iii) altering latency phenotypes (Albrecht and Weller, 1980; Yamane et al., 1983; Waldman et al., 1989; Kahl et al., 2000). Specifically, the HCMV open reading frames (ORFs) UL128, UL130, UL131, which comprise a viral glycoprotein entry complex, were found to accumulate mutations during passage in fibroblasts, leading to the loss of viral tropism for infection of epithelial cells, endothelial cells, macrophages, and dendritic cells (Sinzger et al., 1999, Hahn et al., 2004; Wang and Shenk, 2005; Adler et al., 2006). These lab-adapted HCMV strains were also shown to have lost several genes in the UL/b’ region of the viral genome, a region that confers immune evasion functions and replication dissemination in vivo (Cha et al., 1996). It is now known that sustained viral growth on fibroblast cultures removes the selection pressure to retain these sequences, resulting in genetic loss or rearrangement of sequences essential for replication and dissemination in other host cells and tissues. However, passage of HCMV clinical isolates (e.g., TB40/E and VR1814) in epithelial and endothelial cell settings maintains selection pressure to prevent loss of tropism for non-fibroblast cell types (Waldman et al., 1991; Hahn et al., 2002; Sinzger et al., 2008).

Differences between established laboratory-adapted strains of HCMV and clinical isolates of HCMV are important considerations when planning experiments, as the choice of virus strain used may influence results. Clinical isolates are much more similar to the viruses that replicate within patients, making them preferable for understanding clinical symptoms, as well as natural and drug-selected genetic variability of human CMV. These clinical isolates also maintain productive infection and latency phenotypes that best represent the wild-type virus population (Lee et al., 2015). Because clinical isolates spread through a cell-associated manner, yields from clinical isolates are significantly lower than those collected from laboratory-adapted virus strains, due in part to clinical strains being more limited to cell-associated spread. This contributes to time consuming and labor-intensive aspects of clinical virus stock preparation.

Here, we report a new, more efficient method for generating stocks of clinically derived HCMV isolates, represented by TB40/E IE2-EYEFP. This virus is genetically tagged with an EYEFP fusion to enable convenient monitoring of continuous infection in the bioreactor environment. Using a two-stage bioreactor system (Figure 1) with microcarrier beads and the advantage of the prolonged period of virus production characteristic of HCMV, we are able to maintain an end-stage infection and generate high titer virus stocks on primary adherent cell cultures that preserve genetic integrity of key viral tropism factors and the viral genome. We have used the TB40/E-IE2 EYFP tagged virus in development and characterization of the bioreactor infection. The fluorescent tag enables convenient monitoring of the virus in the bioreactor culture, by surveying aliquots of the infected samples with fluorescent microscopy.