Rabies disease is incurable and cause for the highest human death toll among zoonotic infections [24, 25]. This is especially disturbing as the disease can be prevented with appropriate pre- and post-exposure prophylaxis [26–28].
Worldwide, rabid dogs are the major source of human infection [1]. The global strategic plan to end human deaths from dog-mediated rabies by 2030 was launched in 2015. Four organizations, the World Health Organization (WHO), the World Organization for Animal Health (OIE), the Food and Agriculture Organization of the United Nations (FAO) and the Global Alliance for Rabies Control (GARC), have joined forces as United Against Rabies to achieve this goal. Expanded dog vaccination in endemic countries is among the major measures aimed at elimination of dog-mediated human rabies [26].
Livestock can also be highly affected by canine rabies in endemic areas. For example, Turkey reports an incidence rate of 0.10 to 3.87 cases/100 000 cattle [29]. Substantially higher incidence has been reported in Tanzania with 12–25 cases/100 000 cattle annually in rural communities [6]. Annual livestock loss was estimated to accumulate to $12.3 million in Africa and Asia [3].
Dog vaccination coverage should be maintained continuously above the threshold of 70% over years to achieve the interruption of canine rabies transmission to humans. This challenge is beyond the capacity of most developing countries mainly due to the lack of resources [5, 10, 26]. A robust and transportable production process may improve local production of rabies vaccines, which in turn may facilitate access to a vaccine and enhance its availability. Such a scenario may help to increase the vaccination rate of canine populations in endemic and poor countries [5, 30].
In this work we studied the replication of PV rabies virus, initially adapted to BHK-21 cells, in avian CR and CR.pIX cell lines. These cell lines are derived from primary cells of a muscovy duck embryo and were used for the replication of different viruses [17, 19–21, 31]. They are adapted to proliferation in single-cell suspensions in chemically defined medium and amenable to batch cultivation or perfusion processes.
We demonstrate in this study that CR and CR.pIX cells are fully permissive for rabies virus. An infectious titer in the range of 107 to 108 FFU/ml was achieved with an MOI as low as 0.01. Similar results were obtained when the cells were grown in the specifically developed culture medium free of the recombinant growth factor LONG R3IGF, the sole protein component in the medium. This result is important because LONG R3IGF is one major cost factor in chemically defined media. Without this factor, costs of goods and medium complexity are further reduced.
The BHK-derived PV rabies virus strain replicated efficiently in single-cell suspensions of the anatine cell line apparently without dependence on virus adaptation. This property is beneficial in the light of a report where the same virus strain required at least three passages to achieve a high titer in BHK-21 cells that were adapted to proliferation in suspended aggregates in serum-free medium [32]. Although a comparision is difficult, rabies virus isolated from a rabid dog required seven passages in BHK-21 for adaptation to efficient replication [33]. The results obtained here with rabies virus are consistent with previous studies where CR and CR.pIX cells were fully permissive for influenza virus types (A and B) [34], poxviral vectors [18, 19], and a diverse spectrum of avian pathogens [20, 31] without adaptation.
The previous experiences also indicated that modified vaccinia Ankara replicates to higher titers in CR.pIX cells compared to parental CR cells [19]. The CR.pIX cell line was generated by stable transfection of established CR cells with an expression plasmid that encodes pIX [20], a minor structural protein of adenoviruses that appears to be involved in stabilizing adenvoviral capsids against thermal stress. The small adenoviral protein appears to have additional functions [35] and one model that explains augmented replication of viruses unrelated to adenoviruses suggests that pIX may constitutively elevate Hsp90 into an activated state in the avian cell lines [15]. As Hsp90 is involved in the replication of several viruses, including rabies virus [36], a potential connection between pIX and heat shock responses may also explain the increased yields for rabies virus that were observed in this study.
The Multiplicity of Infection (MOI) is another critical process parameter that often needs to be optimized for new virus production processes. Here, we observed that the MOI only slightly affected peak titers for rabies virus in the avian cell line. Interestingly, although titers were usually around 108 FFU/ml for all MOIs, the highest titer was obtained with an MOI of 0.001. This MOI is 100-fold lower compared to previous data generated within our group with BHK-21 or Vero cells as substrate [37, 38] which further demonstrates the high permissivity and productivity of CR.pIX cells for propagation of rabies virus.
Cell density at the time of infection is another critical process parameter. In this work, CR.pIX infection at 2x106 cells/ml resulted in the highest cell specific productivity compared to the other densities (1x and 4x106 cells/ml). At 2x106 cells/ml, the specific virus productivity was 2.7 or 1.2-fold higher compared to when the cells were infected at 106 or 4x106 cells/ml, respectively. These data confirm that the so-called “cell density effect” also impacts rabies virus production in CR.pIX cells. This effect describes an often observed decrease of the amount of virus produced in a variety of hosts at high cell densities [39–41]. However, high cell density processes using CR.pIX cells were successfully developed for MVA [18, 19]. In these studies, medium feeding and exchange strategies applied prior to or during the virus replication phase were demonstrated to restore a high specific productivity. We could show that perfusion during the virus production phase maintained high infectious titers also of rabies virus until the end of the culture and without the drop observed in batch cultures. As an additional advantage, multiple harvests could be collected (Table 1) and the potency of the experimental vaccine was enhanced 2-fold. Consequently, the expected total vaccine doses that could be produced using perfusion was increased by 9.6 fold (Table 1). This performance was even higher than our previous data obtained with BHK-21 cells grown in serum containing medium in perfusion mode in a 20-l bioreactor [37].
Table 1
Rabies virus production in CR.pIX suspension cells grown in a 7-L bioreactor using different culture modes
Culture mode | Time of culture (days) | Working volume (ml) | Harvest volume (ml) | Virus titer (FFU/ml) | GP level (µg/ml) | Activity (IU/ml) | Expected doses |
Batch | 8 | 2500 | 2500 | 2.3x107 | 13 ± 2 | 3.5 | 8750 |
Fed batch/perfusion | 13 | 4000* | 2000 | 1.9x107 | 2.8 ± 1 | 7 | 84000 |
| | | 2000 | 1.3x108 | 32 ± 12 |
| | | 2000 | 3.2x108 | 69.7 ± 5 |
| | | 2000 | 2.5x108 | 73.4 ± 4 |
| | | 4000 | 3.3x108 | 78.6 ± 10 |
*Working volume during the virus production phase |