Detection of Cache Valley Virus in Biologics Manufactured in CHO Cells

October 1, 2008

Avoid manufacturing failures by effective viral inactivation.

Chinese hamster ovary (CHO) cell substrates are widely used for manufacturing biologics, such as recombinant proteins and monoclonal antibodies, used as human therapeutic and diagnostic agents. These cells are free of infectious endogenous viruses and have proven to be relatively resistant to contamination by adventitious viruses derived from the environment, manufacturing staff, and raw materials used in the processes.1,2 The few viral contaminants that have been detected with some frequency include the mouse parvovirus (mouse minute virus), the respiratory/enteric virus reovirus, and the Bunyavirus Cache Valley virus (CVV).3,4 We describe the detection of CVV in biologics manufactured in CHO cells over a 10-year period.



From 1996 through 2006, we performed viral safety testing of a variety of biologics manufactured in CHO cells in both bioreactor and large-scale monolayer culture processes. A major component of such testing, the in vitro virus screen, is required by the US Food and Drug Administration for each manufactured lot of a biologic.5,6 The assay consists of inoculating the biologic bulk harvest samples (taken before purification) onto monolayer cultures of three or more detector cell lines. Cache Valley virus causes a striking and characteristic response in this assay. It produces an unusually rapid lytic infection of the detector cells, but does not produce hemadsorption or hemagglutination of red blood cells. At high viral loads (>106 tissue culture infectious doses/mL), cytopathic effects are observed in the detector cells within 48–72 hours of inoculation. Total cell lysis rapidly follows initial observation of cytopathic effects, typically within 24 hours.

At least three separate encounters with CVV were documented at BioReliance. A biologic was found to be contaminated in the year 2000. Cache Valley contamination of a cluster of five samples from one client was detected in the year 2003. In 2004, a separate cluster of four samples from one client was contaminated. The viral infection in each case manifested itself as the manufacturing process progressed (before collecting samples for viral safety testing). The manifestations included the absence of cell sheets in the monolayer processes, and perturbations in parameters (oxygen and base consumption) indicative of death of the cells in the bioreactor processes.

Figure 1. Transmission electron micrograph of Vero cells infected with the year-2000 Cache Valley virus isolate (V, a particle apparently budding from the plasma or vesicle membrane; S, external spikes on a viral particle). 99,800x magnification. Reproduced from reference 14, with permission.

Transmission electron microscopic evaluations of cells infected with the year-2000 isolate, the five year-2003 isolates, and two of the year-2004 isolates were performed during the course of the contaminant identification process. Average viral particle size, structure, and location in the infected cells were recorded. Viral particles were typically observed in vesicles in the cell or were observed extracellularly. The electron micrograph of the year-2000 isolate displayed particles with characteristic external spikes, including one particle apparently budding from a plasma or vesicle membrane (V in Figure 1). The particle sizes for the year-2000 isolate were in the 80–100 nm range. Viral particle sizes for the isolates from the years 2003 and 2004 ranged from 70–90 nm (Table 1). A micrograph of one of the year-2003 isolates displays numerous 74–86 nm particles located extracellularly (Figure 2).

Table 1. Transmission electron microscopic findings for Cache Valley virus isolated from biologicals

In each case, the contaminant was ultimately identified as CVV by means of reverse transcriptase–polymerase chain reaction (RT–PCR) assays using bunyavirus-specific primer sets targeting the S- or M-segment G1 envelope glycoprotein genes, followed by nucleotide sequencing of the resulting amplicons. Species level identity was obtained by comparing the sequences obtained for the amplicons to published viral sequence data (GenBank) using basic local alignment search tool (BLAST) searches. For the year-2000 isolate, 100% sequence identity was obtained for a 198 base-pair amplicon sequence and sequences posted for CVV (GenBank accession numbers X73465.1, DQ315775.1). For one of the isolates obtained in 2003, a 98% sequence identity was obtained for a 493 base-pair amplicon sequence and sequences posted for CVV (GenBank accession numbers AF186242, AF231119, AF231113). A concensus region of 167 base pairs was obtained between the four isolates detected in 2004. There was 99% sequence identity between the determined consensus sequence and sequences posted for CVV (GenBank accession numbers X73465, DQ315775). For each BLAST search, the other close matches were all members of the Bunyamwera serotype.

Figure 2. Transmission electron micrograph of Vero cells infected with a year-2003 Cache Valley virus isolate (arrow points to one of a cluster of viral particles). 90,700x magnification.


Cache Valley virus is a mosquito-borne arbovirus that is a member of the family Bunyaviridae, genus Bunyavirus, serogroup Bunyamwera. The virus is widely distributed throughout North America.7 Neutralizing antibodies to the virus have been detected in deer, sheep, horses, cattle, and in humans.8–11 Cache Valley virus has been reported to cause disease in humans on at least two occasions. A case of encephalitis and multiorgan failure leading to death of a human was reported in 1997 to be attributed to CVV infection.12 More recently, a case of acute meningitis in a patient was attributed to CVV.13 The CVV isolated from the encephalitis patient in 1997 was found by thin-section electron microscopic analysis of infected Vero cells to contain enveloped viral particles of 55 to 84 nm (average 70 nm) size in intracellular vacuoles of smooth cytoplasmic membranes.12 In the case of the meningitis patient in 2006, evaluation of negative-stained concentrated cell culture supernatants revealed viral particle sizes ranging from 60–80 nm.15,13 Although the 80–100 nm particle size observed for the year-2000 isolate in the present study was in the nominal range (80–120 nm) expected for a bunyavirus,7 the isolates from 2003 and 2004 displayed particle sizes in the mid-70 to mid-80 nm range, substantially smaller than expected but similar to those noted in the human cases.12,13 This initially confounded the identification of the viral contaminant for these episodes, because it led us incorrectly in the direction of alphaviruses and togaviruses and appeared to rule out the bunyaviruses. The smaller viral particle sizes observed for CVV in many of these cases do not appear to be an artifact of the method used to size the particles (i.e., thin–section versus negative staining of viral pellets).

The contamination of biologics manufacturing processes with CVV has been attributed to the large volumes of non-gamma-irradiated fetal bovine serum (FBS) used as a cell growth medium component. Although the FBS is tested for viral contaminants before use, the lots of serum involved are large, and potential viral contaminants are likely to be nonhomogenously distributed among the individual bottles comprising each lot. Modeling of CVV growth in CHO cells under the conditions used in biologics production has suggested that very low levels of the CVV in the FBS used during cell expansion can cause the manufacturing failures observed.16 The rapid lytic infections resulting from introduction of CVV into CHO-cell processes typically manifest themselves through aberrations in process parameters that are monitored during the production runs. Therefore, the presence of the virus typically leads to premature termination of the manufacturing processes, and the investigatory testing that is required leads rapidly to identification of CVV as the root cause of the failed production runs.

Because CVV is relatively large (80–120 nm) and contains a lipid envelope, it is highly susceptible to inactivation by a variety of physical (gamma-irradiation) and chemical (detergent, solvent) means.7 The viral inactivation and removal processes that must be included in the downstream processing of biologics derived from CHO-cells should readily inactivate or remove any low level CVV infections escaping detection in the viral screening assays performed on the unprocessed bulk samples.5,6

Fortunately, the risk of introducing CVV to patients through contaminated biologics should be very low because of: 1) the rarity of the contamination events; 2) the facility with which the infections are detected during production through in process monitoring and in vitro virus screening; 3) the fact that known contaminated lots of product are discarded; and 4) the relatively high susceptibility of any undetected CVV to physical and chemical inactivation strategies used during product purification.

At the time when this article was written, Raymond W. Nims, PhD, was a scientific director, Sandra K. Dusing, PhD, was a scientific director, and Wang-Ting Hsieh, PhD, was a senior scientist, all at BioReliance, Rockville, MD. Archie Lovatt, PhD, was a scientific director, G. Gordon Reid, PhD, was a scientist, David Onions, PhD, FRSE, was the chief scientific officer, and Euan W. Milne was a senior scientist, all at BioReliance UK, Glasgow, Scotland, UK. Corresponding author Nims currently is at Amgen, Longmont, CO, 303.401.2354,


1. Hojman F, et al. Biological and molecular characterization of an endogenous retrovirus present in CHO/HBs—A Chinese hamster cell line. Dev Biol Stand. 1989;70:195–202.

2. Dinowitz M, et al. Recent studies on retrovirus-like particles in Chinese hamster ovary cells. Dev Biol Stand. 1992;76:210–207.

3. Garnick RL. Experience with viral contamination in cell culture. Dev Biol Stand. 1996;88:49–6.

4. Nims RW. Detection of adventitious viruses in biologicals—A rare occurrence. Dev Biol. 2006;123:153–164.

5. US FDA. Points to consider in the characterization of cell lines used to produce biologicals. FDA/CBER, 1993.

6. US FDA. Points to consider in the manufacture and testing of monoclonal antibody products for human use. FDA/CBER, 1997.

7. Gonzalez-Scarano F, Nathanson N. Bunyaviridae. In: Fields BN, Knipe DM, Howley PM, editors. Fields Virology, 3rd edition, Philadelphia, PA: Lippincott-Raven; 1996. p.1473–1504.

8. Blackmore CG, Grimstad PR. Cache Valley and Potosi viruses (Bunyaviridae) in white-tailed deer (Odocoileus virginianus): experimental infections and antibody prevalence in natural populations. Am J Trop Med Hyg. 1998;59:704–709.

9. Sahu SP, et al. Serologic survey of cattle in the northeastern and north central United States, Virginia, Alaska, and Hawaii for antibodies to Cache Valley and antigenically related viruses. Am J Trop Med Hyg. 2002;67:119–122.

10. McLean RG, Calisher CH, Parham GL. Isolation of Cache Valley virus and detection of antibody for selected arboviruses in Michigan horses in 1980. Am J Vet Res 1987;48:1039–1041.

11. Buescher EL, et al. Cache Valley virus in the Del Mar Va Peninsula. I. Virologic and serologic evidence of infection. Am J Trop Med Hyg. 1970;19:493–502.

12. Sexton DJ, et al. Life-threatening Cache Valley virus infection. N Engl J Med. 1997;336:547–550.

13. Campbell GL, et al. Second human case of Cache Valley virus disease. Emerg Infect Dis. 2006;12:854–856.

14. Nims R, et al. Adventitious agents: Concerns and testing for biopharmaceuticals. In: Rathore AS, Sofer G, editors. Process validation in manufacturing of biopharmaceuticals: guidelines, current practices, and industrial case studies. Boca Raton, FL: CRC Press/Francis Informa; 2005. p. 143–67.

15. Powell JW. Personal communication.

16. Onions D, unpublished data.