Mousepox Resulting from Use of Ectromelia Virus-Contaminated, Imported Mouse Serum

Neil S. Lipman,1* Scott Perkins,1 Hai Nguyen,1 Martin Pfeffer,2 and Hermann Meyer3

Abstract  Mousepox was identified in a single mouse-holding room in early 1999 after a group of 20 CAF1/Hsd mice were inoculated SC with a killed murine spindle cell tumor line, S1509A. The cell line had been used without complications multiple times and was determined to be free of viral contamination on the basis of results of mouse antibody production testing. Of the 20 mice inoculated, 12 mice died by postinoculation day 8. Severe lymphoid and hepatic necrosis was observed in select mice subjected to histologic examination. Ballooning degeneration of epithelial cells with intracytoplasmic eosinophilic inclusion bodies was observed in the skin overlying the inoculation site of the single mouse from which this tissue site was evaluated. Presence of ectromelia virus was confirmed by use of immunohistochemical and polymerase chain reaction analyses, and the virus was isolated after serum, pooled from 5 of the index cases, was inoculated into an immune-naive mouse. Investigation into the source of virus contamination included inoculating mice with aliquots of various S1509A freeze dates; chemically defined media and supplements, including fetal bovine serum; and two lots of pooled commercial mouse sera, after heat inactivation at 56ЊC for 30 minutes used as a medium supplement. One lot of pooled commercial mouse serum was identified as the source of ectromelia virus. This lot of serum was inadvertently used to feed S1509A cells that were subsequently inoculated into mice. We determined that the contaminated serum, which was purchased in late 1998, originated from China. The serum was imported into the United States as a batch of 43 L in early 1995. The serum was blended into a single lot and filtered (0.2 ␮m) before distribution to major suppliers throughout the country. The serum was sold or further processed to obtain a variety of serum-derived products. Because murine serum is generally sold in small aliquots (10 to 50 ml), we speculate that several thousand aliquots may have been derived from this batch of serum and, if inoculated into mice, would likely result in additional mousepox outbreaks.

Mousepox, a potentially devastating disease of murine research yoelii in mice, resulted in the destruction of more than 5,000 colonies, is caused by ectromelia virus (EV), an orthopoxvirus of mice housed in two of four buildings at the NMRI followed by mice. Epizootic EV infection may cause substantial morbidity formaldehyde gas decontamination. Few details were provided and mortality dependent on the strain of virus and the host’s in the report, or have been made available subsequently, as to genotype, age, sex, and immunologic status (1–3). Mousepox can the source of the contaminated serum, except that it was ob- be associated with high mortality and limited clinical signs; tained from a domestic mouse colony. generalized maculopapular rash with scabs and/or conjunctivi- Historically, mousepox has been reported to be enzootic in tis, occasionally associated with appendage loss; or as inappar- Europe, Japan, and China. Today it is less common and is infre- ent infection (2). Considerable effort is exerted to maintain quently reported in Europe and Japan. It is still reported as EV-free mouse colonies as a serious outcome may follow its in- common in China (2). advertent introduction. In early 1999, unexpected mortality was observed in mice in- Mousepox is rare in the United States. The first mousepox oculated with a murine-derived cell line at the Weill Medical epizootic was reported in the 1950s (4). Although there were College of Cornell University (WMC/CU). Evaluation of affected sporadic reports of outbreaks over the subsequent 2 decades, it mice revealed mousepox and subsequent epidemiologic evalua- wasn’t until a major outbreak occurred in the United States in tion traced the source of EV to imported pooled commercial the late 1970s and early 1980s at the National Institutes of murine serum. We describe the clinical features of the outbreak, Health (NIH) and numerous U.S. research institutions, that the the methods used to eradicate the virus, and the source and dis- full impact of a mousepox outbreak was recognized (5–14). The tribution of the contaminated serum. NIH outbreak led to the destruction of tens of thousands of mice at a cost of millions of dollars and lost research (4). After the Materials and Methods NIH outbreak, reports of mousepox were not published in the Animals: Hybrid (CAF1/Hsd) F1 mice, which subsequently United States until 1996, when the most recent mousepox out- developed mousepox, were purchased from a commercial vendor break was reported from the Naval Medical Research Institute (Harlan Sprague Dawley, Indianapolis, IN) reported to be free of (NMRI), Bethesda, Md (15). The outbreak, traced to pooled com- adventitious virus contamination, including Sendai, mouse mercial murine serum used as a carrier to passage Plasmodium hepatitis virus (MHV), pneumonia virus of mice (PVM), minute 1Research Animal Resource Center, Weill Medical College of Cornell University virus of mice (MVM), mouse parvovirus (MPV), transmissible and Memorial Sloan-Kettering Cancer Center, New York, NY; 2Institute for Medi- encephalomyelitis virus (GDVII), reovirus 3, mouse rotavirus cal Microbiology, Epidemic and Infectious Diseases, Ludwig-Maximilians Univer- sity, Munich, Germany; 3Institute for Microbiology, German Armed Forces Medical (EDIM), ectromelia, lymphocytic choriomeningitis (LCM), mouse Academy, Munich, Germany; *Corresponding Author. adenovirus (Mad), polyoma, mouse cytomegalovirus (MCMV), 426 Mousepox Associated with Contaminated Mouse Serum

Hantaan virus, lactate dehydrogenase elevating virus (LDV), K Table 1. Murine sentinel health-monitoring program virus, and mouse thymic virus (MTV), Mycoplasma sp., para- Panel A B C sites, and pathogenic bacteria. On arrival, mice were uncrated Frequency Bimonthly Semiannual Annual in an animal holding room housing mice for a single investiga- Agents MHV Bimonthly plus: Semiannual plus: Sendai Reovirus 3 Mad tive group. They were uncrated on an open bench and placed in MVM LCM Polyoma static isolator cages (Micro-barrier® standard height, Allentown GDVII Ectromelia Hantaan Caging Equipment Co. Inc., Allentown, NJ), using aqueous io- EDIM Cecal/duodenal protozoa MTV PVM Rectal specimen culture MCMV dophor (Wescodyne, AMSCO, Apex, NC)-disinfected forceps. MPV for Clostridium rodentium Clostridium piliforme They acclimated for eight days prior to use. In addition to the Mycoplasma Respiratory tract CAR bacillus pulmonis aerobic culture Encephalitozoon CAF1 mice, the room contained a unique transgenic (Tg) line, cuniculi C57BL/6J, and A/J mice. Skin scraping Examination of cecum and The animal holding room, housing EV-infected mice, was served Pooled fecal colon for helminths flotation by a single corridor and contained three shelf racks. At the time of Anal tape test the outbreak, the room held approximately 80 static isolator MHV = mouse hepatitis virus; MVM = minute virus of mice; GDVII = trans- cages. The corridor served a distinct section of the animal facil- missible encephalomyelitis virus; EDIM = mouse rotavirus; PVM = pneumo- ity in which there were 15 additional holding rooms housing nia virus of mice; MPV = mouse parvovirus; LCM = lymphocytic choriomeningitis virus; Mad = mouse adenovirus; MTV = mouse thymic virus; MCMV = predominantly mice, but also rats, guinea pigs, and frogs. Differ- mouse cytomegalovirus; and CAR = cilia-associated respiratory bacillus. ential pressures provided directional airflow so that air flowed from the corridor into the animal rooms. Mice in other holding later, an additional mouse from this group (C99-353) was found rooms within the facility were housed in either static or ventilated dead and was subjected to complete pathologic examination. (Maxi-Miser®, Thoren Caging Systems, Inc., Hazelton, PA) isola- Eight days after inoculation, 6 additional mice were found dead tor cages and were changed on an open bench or in a vertical without a detectable prodrome. At that time, the remaining live flow class-100 HEPA-filtered change station (Max-Miser® Mo- mice (C99-377-1 through C99-377-8) were euthanized (n = 8), blood bile Change Station, Thoren Caging Systems, Inc., Hazelton, was collected for serum antibody determination (n = 5), and the PA), using forceps. Cages contained autoclaved corncob bedding bodies were submitted for gross and histologic (n = 5) evalua- (Bed O’Cobs, The Anderson Co., Maumee, OH) and were changed tion. Histopathologic findings were consistent with EV infec- at least weekly. Mice were fed pelleted feed (Laboratory Rodent tion, and the room was quarantined. Additional testing was Diet 5001, PMI Feeds, Inc., St. Louis, MO) and were provided completed to confirm the presence of EV in tissue sections by tap water in water bottles ad libitum. use of indirect fluorescent antibody (IFA) testing (C99-353) and The animal holding room environment was controlled, main- polymerase chain reaction (PCR) analysis (C99-377-2 and C99- taining temperature at 22 Ϯ 1ЊC, relative humidity between 40 377-4). Subsequently, intensive evaluation was undertaken to and 60%, and provided a photoperiod of 12 hours’ light/12 hours’ identify the source(s) of the virus. dark. Ventilation in the holding rooms was 100% fresh air and After suspicion of mousepox was raised, cage-side examina- provided approximately 11 air changes/h. The facility is approved tions were performed on all mice housed within the room. Three by the Association for Assessment and Accreditation of Laboratory mice were identified with gross lesions suggestive of mousepox, in- Animal Care International (AAALAC), and components of this including: an unmanipulated CAF1/Hsd mouse with unilateral vestigation that involved animal manipulation were approved by conjunctivitis and bilateral ulcerative pinnae; a B6 mouse with the Institutional Animal Care and Use Committee. ulcerative skin lesions on its flank; and an A/J mouse with gen- Colony health status was monitored, using a comprehensive eralized pruritis and scruffy fur. Blood samples were obtained, sentinel mouse health evaluation program. Each mouse-holding complete necropsy was performed, and the presence of serum room contained two sentinel cages each housing three outbred antibodies against a variety of murine adventitious pathogens Swiss Webster (Tac:(SW)fBR) mice. Sentinel cages consisting of (panel A [Table 1] and EV) was determined. a shoebox and wire bar lid, but not a filter top, were each ex- Blood was collected for serologic evaluation (panel A and EV) posed to soiled bedding from at least 10 randomly selected from one mouse in each of 15 cages containing an irreplaceable colony cages, mixed 1:1 with fresh bedding, at weekly cage transgenic (Tg) line, S1509A live tumor-bearing mice inoculated change. Every two months, blood samples were collected from with a tumor batch distinct from that injected into the index cases, one mouse per sentinel cage and were evaluated for serologic and the room sentinels. Those mice were then quarantined in a evidence of exposure to a panel of adventitious viruses, M. BSL-2 facility. All remaining mice in the room were euthanized in pulmonis, parasites, and select bacteria. Table 1 lists the agents situ. Approximately 1 month later, the S1509A tumor-bearing mice in each test panel and the frequency of testing for each agent. and the room sentinels present in the room at the time of the out- The holding room in which mousepox was detected was free of break had blood samples drawn for serum antibody determination all agents in the sentinel test panels for at least two years pre- and the bodies were submitted for complete necropsy. ceding the outbreak. After the outbreak, the frequency of test- Specific-pathogen-free SW female mice (Tac:(SW)fBR) ob- ing for EV antibody in sentinels was increased to monthly for tained from a commercial vendor (Taconic, Germantown, NY) six months and bimonthly thereafter. were placed as contact sentinels, 1 mouse per cage, in each of History: In February 1999, 20 female CAF1/Hsd mice were the cages housing the Tg mice for 6 weeks. They were then re- inoculated SC in both flanks with approximately 106 S1509A moved, blood was drawn for EV antibody determination, and cells, a murine spindle cell tumor line, previously killed by re- mice were subsequently group housed and received dirty bed- peated freezing and thawing. On postinoculation day (PID) 6, ding from the cages housing Tg mice. Blood samples were ob- five mice were found dead without prior signs of illness. A day tained three, six, and 12 weeks after placement for EV antibody. 427 Vol 50, No 4 Comparative Medicine August 2000

Figure 1. Photomicrographs of sections of liver (A), spleen (B), Peyer’s patch in small intestine (C), and bone marrow (D) from CAF1 mice (index cases) inoculated SC with ectromelia virus-contaminated murine spindle cells. Notice severe necrosis; intracytoplasmic inclusion bodies (arrows) are observed in skin (E) overlying the inoculation site of a CAF1 mouse. Lung (F) with an interstitial cellular infiltrate and alveolar edema from a SW mouse inoculated IP with pooled serum from the index cases. H&E stain; 250 X. Complete necropsy was performed at the time of the last mousepox, sentinel mice, and all mice used for test article inocu- blood collection. lation. Representative tissues collected at necropsy included The holding room that housed EV-infected mice was decon- heart, lung, liver, spleen, kidney, adrenal gland, stomach, duode- taminated by thoroughly foaming all exposed surfaces, using a num, jejunum, ileum, cecum, colon, urinary bladder, urethra, quaternary ammonium disinfectant (Anlage 64, Quip Laborato- esophagus, trachea, ovary, uterus, submandibular and mesen- ries, Wilmington, DE). The holding room was left vacant for six teric lymph nodes, thymus, skin, and bone and bone marrow weeks. Subsequently, two open shoebox cages each housing (sternum and cervical vertebrae). Tissues were preserved for 72 three mice of the DBA/2 strain or SW stock were placed in the hours in buffered 10% formalin, processed by use of routine room. One month after placement, blood samples were obtained methods, cut into 5-␮m-thick sections, and stained with hema- for serum antibody determination (panel A and EV). toxylin and eosin according to routine methods. Tissues were Pathologic examinations: Complete necropsy was per- also frozen and saved from all cases evaluated subsequent to formed on the index cases, mice with clinical signs suggestive of the mousepox diagnosis. 428 Mousepox Associated with Contaminated Mouse Serum

Immunohistochemical analysis: Paraffin-embedded sections small and large intestines were frozen at –70ЊC. If found dead or of spleen, liver, and skin from mouse C99-353 were stained for EV moribund prior to MAP test completion, complete necropsy was antigen by use of the avidin-biotin complex immunoperoxidase performed, tissues were frozen, and blood was collected and method with hyperimmune rabbit anti-vaccinia virus serum evaluated against panel C agents (Table 1). Liver and spleen col- as described (16). lected at necropsy from the mouse receiving pooled serum from Serologic testing: Enzyme-linked immunosorbent assays index cases were used for virus isolation. Liver, spleen, and lung (ELISA) and IFA testing were performed by WMC/CU’s Research from select mice receiving pooled serum from index cases and a Animal Diagnostic Laboratory. The ELISA was used to evaluate pooled sample consisting of 2 lots of commercially purchased mu- sera for antibody to MHV, Sendai, GDVII, PVM, MVM, LCM, ec- rine serum obtained from a single supplier were evaluated by tromelia, rotavirus (EDIM), reovirus 3, K, Mad, and polyoma vi- PCR analysis. All procedures were conducted within a class-II ruses, and M. pulmonis and cilia-associated respiratory (CAR) type-A biological safety cabinet, and animals were housed in a bacillus, using commercially prepared reagents (Charles River BSL-2 animal-holding facility. Laboratory [CRL], Wilmington, MA). The IFA test was used to Virus isolation: The spleen and liver from the mouse inocu- screen sera for antibody to MPV, MCMV, and MTV. Equivocal or lated with pooled serum from index cases were homogenized positive results were confirmed by submitting a duplicate sample separately, freeze-thawed twice, and sonicated once for 30 sec- to a commercial laboratory (CRL) for evaluation, using the ELISA onds at 80 Hz (Branson Sonifier Cell disruptor B15). After cen- and IFA test. Seropositive results for EV were also confirmed by trifugation (1,000 x g for 10 minutes), MA104 (African green use of a hemagglutination-inhibition (HAI) assay. Blood for sero- monkey kidney) cells were absorbed with the resulting superna- logic testing was collected from the retro-orbital plexus (survival) tant for one hour at 37ЊC and were incubated with minimal es- or by cardiac puncture (terminal) of anesthetized mice. sential medium containing 1% FBS. Cell line and biologics: Subsequent to the identification of Polymerase chain reaction: The DNA was prepared directly mousepox, investigative staff were queried extensively to deter- from 400 ␮l of supernatant of ground tissue specimens, from mine the nature of procedures to which EV-infected mice had mounted paraffin-fixed tissue sections, from 400 ␮l of infected tis- been subjected, the scope of the projects conducted in the ani- sue culture (fluid and cells), or from 400 ␮l of mouse serum as de- mal holding room for the preceding six months, and to identify scribed (17). The DNA of orthopoxvirus reference strains, cowpox the nature and source of all biologics that were inoculated into virus strain Brighton, vaccinia virus strain Western Reserve, the mice or used to prepare materials that were subsequently in- monkeypox virus strain Copenhagen, camelpox virus strain CP-1, oculated into mice. All cell lines and biologics used were subse- and mousepox virus strain MP-1 were used. To amplify ortho- quently tested for contamination by use of mouse inoculation. poxvirus specific DNA sequences, A27L-PCR was applied. The Murine serum purchased commercially and used by the laboratory A27L-PCR is directed at the gene coding for a 14-kDa fusion pro- was also subject to serologic evaluation, negative-staining electron tein, an immunogenic envelope protein of orthopoxviruses. Prim- microscopy, PCR analysis, and virus isolation. ers 14 kDa-1 and 14 kDa-2 (ATGGACGGAACTCTTTTCCC and Animal inoculation: Mouse antibody production (MAP) test- TAGCCAGAGATATCATAGCCGC) comprising positions 1–21 ing was performed by inoculating 6-week old female SW [Tac: and 271–292 of the gene, respectively, were used. The primers (SW)fBR] mice with the following test articles: each of four frozen generate an amplicon of 292 basepairs. Amplicons were ana- aliquots of the S1509A cell line representing different freeze dates; lyzed in 1% agarose gels. pooled mouse serum from the index cases, diluted 1:5 in Hanks’ A second PCR analysis was performed as recently described buffered saline solution (HBSS); a pooled sample comprising (18). The ATI-PCR assay was accomplished by selection of oli- S1509A cells and medium (RPMI 1640) supplemented with 10% gonucleotides corresponding to sequences that encode the so-called fetal bovine serum (FBS) collected from three 150-cm2 cell culture acidophilic, or A-type, inclusion protein of cowpox virus. The primer flasks maintained in the laboratory’s incubator at the time of the pair spans a genomic region that contains various deletions in outbreak, as well as an aliquot of the FBS batch in use at that comparison to corresponding parts of vaccinia, variola, monkeypox, time; and a pooled sample consisting of two lots of commercially mousepox, and camelpox viruses. To enhance resolution of frag- purchased murine serum obtained from a single supplier. ment size differences and to confirm specificity of the amplicons, 30 A 0.6-ml aliquot of each test article, except pooled mouse se- ␮l of the PCR products were incubated with 10 U of Bgl II restric- rum from the index cases, was inoculated intraperitoneally (IP; tion endonuclease. Fragmented PCR products were then separated 0.5 ml), intranasally (IN; 0.05 ml), and orally (PO; 0.05 ml) into in a 3% agarose gel. groups of three mice. All test articles, except the pooled mouse Electron microscopy: Both lots of commercial mouse serum serum from the index cases, were also diluted 1:10 with HBSS (0.2 ml) were fixed with 0.5 ml of 2.5% glutaraldehyde for five and inoculated as described previously into additional groups of minutes, then were mixed with 3% uranyl acetate for one minute. three mice. In addition, two groups of two mice were inoculated The mixture was placed on formvar carbon-coated grids and was by use of the same routes and volume, with RPMI 1640 medium allowed to air dry, then was examined by use of a JOEL 122-EX supplemented with10% FBS and HBSS to serve as controls. electron microscope. Because of limited sample size (0.5 ml), pooled mouse serum Epidemiologic investigation: After identification of a from the index cases was inoculated IP into a single mouse. single lot of commercial mouse serum as the source of EV con- Five days after test article administration, blood samples tamination, the supplier from which the serum was obtained were obtained from surviving mice and serum lactate dehydro- was contacted to determine the source of mice from which the genase activity was determined. Mice surviving to day 32 had serum was obtained, how the mice were maintained, how the blood samples drawn, then were euthanized, and complete serum was collected and processed, the size of the lot from necropsy was performed. Sections of liver, spleen, lung, and which the aliquot was purchased, and the subsequent distribu- 429 Vol 50, No 4 Comparative Medicine August 2000 tion of aliquots from that lot. The investigation indicated that Stocks of S1509A were maintained frozen (multiple freeze this lot of serum was distributed through multiple suppliers af- dates) and were periodically thawed and expanded for experi- ter its importation into the United States. Therefore, the im- mental use. Historically, S1509A cells had been passaged in vivo porter and all suppliers identified in the distribution network to maintain their biological behavior. Laboratory records were were contacted and questioned. insufficient to track the parental stock from which each freeze date was obtained. At the time of the outbreak, several cultures Results of S1509A cells were maintained for experimental use. It was Index cases: Six to eight days after SC inoculation with not possible to definitively identify the cultures that were used killed S1509A cells, 12 of 20 CAF1 mice were found dead with- to prepare the inoculum resulting in mousepox or to confirm out prior clinical signs of infection. Necropsy was performed on that one of the cultures maintained at the time of the outbreak nine mice. Necropsy findings from mouse C99-353 included red was used to prepare the inoculum. discoloration of the skin at the inoculation site, mottled appear- Simultaneously with the propagation of the S1509A cell line, ance of the liver, and reddening of the serosal surface of the the laboratory whose mice succumbed to mousepox was also duodenum, which contained blood-tinged contents. Histologi- culturing primary murine Langerhans cells from CAF1 mice. cally, necrosis was observed throughout the liver in haphazard The Langerhans cells were used only in in vitro assays. The pattern (Figure 1A); in the spleen, thymus, intestinal Peyer’s cells were grown in RPMI 1640 medium supplemented with patches, and mesenteric lymph nodes, with the spleen most se- commercially purchased mouse serum. Circumstantial evidence verely affected (Figures 1B and C); and in bone marrow from indicated that the batch of S1509A cells administered to the in- the sternum (Figure 1D). In the skin at the site of tumor cell dex cases was inadvertently cultured in medium supplemented inoculation, ballooning degeneration of epithelial cells, many of with mouse serum. The laboratory had purchased and partially which contained eosinophilic intracytoplasmic inclusion bodies, used two lots of mouse serum from a commercial supplier was apparent (Figure 1E). Mild mononuclear cell infiltrate was within three months preceding the outbreak. The mouse serum observed in the subjacent dermis and hypodermis. Ectromelia was heat inactivated at 56ЊC for 30 minutes prior to use. virus antigen was documented by use of immunohistochemical Animal inoculation: Clinical abnormalities were not de- analysis, and viral DNA was amplified by use of A27L-PCR analy- tected in animals inoculated with frozen S1509A stocks, the sis of the spleen, liver, and skin (pooled for PCR) specimens from pooled sample comprising S1509A cells in complete supplemented mouse C99-353. medium and FBS, or the control mice receiving complete supple- The eight additional animals euthanized on PID 8 had mul- mented medium or HBSS for the 32 days after inoculation. Se- tiple white spots < 1 mm in diameter in the liver (n = 8) and rum lactate dehydrogenase activity, measured on PID 5, was spleen (n = 7); peripheral lymph node enlargement (n = 7); dif- not increased, compared with that in controls and the animals fusely dark-red lungs (n = 1); and multifocal 1-mm dark-red remained free of antibody to the agents in test panel C. Lesions spots in the pulmonary parenchyma (n = 1). Microscopically, were not found at necropsy. necrosis of lymphoid tissues and liver was observed in all cases A summary of animal inoculation results for the mice receiv- as described in mouse C99-353. Gross skin lesions were not ob- ing pooled serum from index cases and a pooled sample consist- served nor were tissue sections examined from inoculation sites. ing of two lots of commercially purchased murine serum Serum antibody to any agents in panel A or EV was not identi- obtained from a single supplier are provided in Table 2. Mouse fied in blood collected at necropsy (n = 5). The A27L-PCR prod- C99-505 receiving pooled mouse serum from the index cases uct was amplified in the pooled liver and spleen specimen died 7 days after inoculation. At necropsy, the thorax was filled collected from both mice examined (C99-377-2 and C99- 377-4). with 2 ml of straw-colored fluid and the lungs were moist and The three colony mice with clinical signs suggestive of mouse- dark red. Acute interstitial pneumonia was observed on micro- pox did not have parenchymal lesions nor did they have serum scopic examination (Figure 1F). The interstitial infiltrate was antibody to EV. The contact sentinels placed with the Tg mice, composed principally of neutrophils and was accompanied by the S1509A live tumor-bearing mice, the room sentinels in place moderate interstitial and alveolar edema. Mild hepatic and at the time of the outbreak, and the DBA/2 and SW sentinel splenic necrosis also was evident. Viral DNA was amplified mice placed after room decontamination remained seronegative from the lung and the pooled liver and spleen specimens by use when tested for agents in panel A and EV. Subsequent testing of of A27L and ATI-PCR analyses. A virus was isolated from a the Tg mouse colony also has failed to detect EV antibody. pooled specimen of liver and spleen. Ninety-six hours after in- Cell line and biologics: For the 6 months preceding the out- oculation, small plaques were observed on MA 104 cells inocu- break, a single cell line, an A strain-derived murine spindle cell lated with the ground pooled liver and spleen. Subsequently, the tumor designated S1509A, was used in vivo by the laboratory im- plaques grew larger, resulting in a plaque-type cytopathogenic plicated in the outbreak. Depending on the experiment, S1509A effect. The isolate’s identity as mousepox virus was proved by was administered as live or dead cells. The index cases received an results of ATI-PCR analysis. A band of the same size was also inoculum of dead cells that had been prepared by freeze thawing amplified from DNA derived from the EV isolate and from for four cycles. Each cycle involved washing the cells in phosphate- pooled tissues (Figure 2, lanes 7 and 8). The Bgl II restriction buffered saline (PBS) and freezing at –70ЊC before centrifugation. enzyme cleavage patterns of the amplicons differed from those The S1509A cells were maintained in vitro by passaging in of other orthopoxviruses species (data not shown). complete medium (RPMI 1640) supplemented with 10% FBS. Four of six mice inoculated with pooled commercial murine The cell line had been MAP tested and was determined free of serum died (n = 3) or were sacrificed because of clinical signs adventitious agent contamination, including EV, at least three characterized by severe dyspnea and ruffled fur (n = 1). Two of times during the preceding five years at two laboratories. three mice inoculated with the undiluted test article died and 430 Mousepox Associated with Contaminated Mouse Serum

were necropsied on day 7 after inoculation, and the third was euthanized, a blood sample was obtained, and necropsy was performed on day 12. One of three animals receiving the diluted (1:5) test article was found dead on day 12. The remaining two animals from this group remained clinically normal until day 25 when one mouse was euthanized after development of a swollen, alopecic, and ulcerative muzzle; the remaining mouse

pos pos was clinically normal when euthanized on PID 32. Necropsy findings were similar among mice that were

Liver/spleen Liver/spleen euthanized or died on or before PID 12 (n = 4). At necropsy, the thorax of mice was fluid filled (n = 2) and the lungs were discol- ored and mottled (n = 4). The small intestinal contents were

ND Lung pos Liver/spleen pos ND ND ND stained reddish black (n = 3). Microscopically, severe acute in-

HAI pos terstitial pneumonia, similar to that described in the mouse receiving pooled serum from index cases, was apparent (n = 4). Severe autolytic change precluded interpretation of the intestinal lesions in two mice; however, diffuse areas of transmural necrosis with mixed inflammatory infiltrates were observed in the other two mice. Mild to moderate lymphoid necrosis was also present in the thymus (n = 3), spleen (n = 3), and liver (n = 1). Necrosis of the ovaries and bone marrow were observed in a single mouse dying on PID 12. Skin lesions, consisting of several

+++ necrosis (+ to ++) small ulcers with minimal inflammatory infiltrate were ob-

enteritis ovarian marrow,

enteritis + pos Liver/spleen neg served on the cranium of cagemates (n = 2) dying on PID 7. In- clusions were not observed in any tissues. Ectromelia virus antibody was detected by use of ELISA and IFA testing in serum from the mouse euthanized on PID 12. Viral DNA was amplified by use of A27L-PCR analysis from the lungs of the two animals examined; however, only in one of two cases was viral DNA amplified by PCR analysis from the pooled liver and

- Necrosis ++ Autolysis Thymic necrosis - Necrosis ++ Autolysis Thymic necrosis spleen specimen. There was no evidence of necrosis in these PCR-

EMH EMH Necrotizing - ELISA/IFA Lung pos ND

HAI = hemagglutination-inhibition test Misc = miscellaneous pos = positive

Necrosis + Necrosis + - - ND Lung pos Liver/spleen pos negative organs at necropsy.

Necrosis ++ Necrosis ++ Necrotizing bone Thymic, ND ND ND The muzzle lesion noted on the mouse euthanized on PID 25 was characterized by necrotic epidermis and severe epidermal and dermal infiltration, principally with neutrophils. Inclusion bodies Mouse (Swiss Webster) incoulation results Webster) Mouse (Swiss were not observed, and sections of skin from other body surfaces were not remarkable. Splenomegaly, resulting from lymphoid hy-

Table 2. Table perplasia, also was observed. The mouse euthanized on PID 32 was free of gross and histologic lesions. Ectromelia virus anti-

Pulmonary effusion/ Pulmonary effusion/ body was detected by use of ELISA, IFA, and HAI testing of sera from mice euthanized on PID 25 and 32. Commercial mouse serum: Mouse serum was purchased

1 1 on two occasions from a single supplier over a 2-month period prior to the mousepox outbreak. The serum samples were from two serum lots (lot A:1 X 10 ml; lot B: 3 X 10 ml). Both lots had

Presumptive bite lesions

1 been labeled “not to be used in humans or animals.” Viral DNA was amplified by use of A27L-PCR analysis from only lot A. Virus particles were not seen in either lot. Infective EV was not isolated by use of direct inoculation of both lots on the MA 104 cells. Both lots contained antibody to MHV, MVM, and M. pulmonis, and lot B also contained antibody to Sendai, GDVII, MPV, and EDIM. Neither lot had detectable antibody to EV by use of any of the test methods. The source of lot A was traced by contacting the importer and the three suppliers to which the lot had been distributed. The lot originated from a batch of > 43 L of serum imported from China in early 1995. The serum was obtained in China from mice acquired from the National Institute of Science or various government-owned research institutes (19). The mice

index cases interstitial pneumonia ++

murine serum on cranium interstitial pneumonia ++ (2 lots pooled) murine serum on cranium interstitial pneumonia ++ (2 lots pooled)

murine serum murine serum ulcers on hyperplasia HAI pos (2 lots pooled) (2 lots pooled) muzzle

murine serum murine serum (2 lots pooled) (2 lots pooled)

Test articleTest Preparation (S or D) Day Skin Lungs Liver Spleen GIT MiscAb Ectromelia PCR isolation Virus

Serum pooled 1:5 dilutionCommercial (D) 7 - Undiluted (D) 7 Pulmonary effusion/ Ulcers

Commercial Undiluted (D) 7 Ulcers

Commercial Undiluted (S) 12 - Interstitial pneumonia ++

Commercial 1:5 dilution (S) 25 Alopecia/ - - Lymphoid - - ELISA/IFA/ ND ND

Commercial 1:5 dilutionCommercial (D) 12 1:5 dilution - (S) 32 Interstitial pneumonia ++ ------ELISA/IFA/ ND ND

+ = mild ND = not done

++ = moderate (D) = died EMH = extramedulary hematopoeisis +++ = severe (S) = sacrificedwere - indirect fluorescent antibody test IFA transported to the collection location in China for one to 431 Vol 50, No 4 Comparative Medicine August 2000 two weeks, during which blood was collected from each mouse 12345678 9 two to three times, yielding between 0.2 and 0.5 ml of serum/ mouse. The batch of serum was tested for EV antibody at one of two institutions in China, using ELISA, as required by the Chinese Department of Agriculture (19). The serum was subsequently imported into the United States under permits is- sued by the United States Department of Agriculture (USDA) and the Centers for Disease Control. An aliquot from the serum batch was provided to the USDA for species verification. After importation, this batch of serum was sold to a distributor who blended and filtered (0.2 ␮m) the serum before distribution. The serum was then distributed to five suppliers between April 1995 and July 1996. One of these suppliers, which further distributed serum from this batch to six additional suppliers, as well as selling aliquots from the batch to their own clients, sup- plied serum to the supplier from which we obtained contaminated lot A. We were unable to determine dispensation of the serum received by the other suppliers.

Discussion Mousepox was identified in a single mouse holding room in early 1999 after a group of hybrid mice were inoculated with a killed murine cell line. The cell line had been used at our institution for over five years and was determined to be free of adventitious murine virus contamination by use of MAP testing three times previously. Pooled commercial mouse serum, derived from mice in China, was determined as the source of EV infection by use of PCR analysis and mouse inoculation studies. Circumstantial evidence supported the presumption that the pooled serum, which was intended to be used as a medium supplement to grow a cell line for in vitro assays, was inadvertently added to the culture from which cells were subsequently inoculated into mice which developed mousepox. Figure 2. Agarose gel electrophoresis of ATI-polymerase chain reac- The outbreak was restricted to a single holding room. This is tion (PCR) products amplified from either DNA or reference orthopoxvirus strains (lanes 2–6), ectromelia virus isolated from mouse in contrast to other reported outbreaks in which EV spread rap- C99-505 inoculated with pooled serum from the index cases (lane 7), idly through the facilities, necessitating large-scale depopula- and ectromelia virus-infected tissue from mouse C99-505 (lane 8). tion, formaldehyde gas or ␤-propriolactone fumigation, and Lanes:1 and 9, 1-kilobase pair (kbp) ladder DNA size marker (Gibco restocking with EV-free animals (3, 6, 15, 20). Although we de- BRL); 2, cowpox virus strain Brighton; 3, vaccinia virus strain West- ern Reserve; 4, monkeypox virus strain Copenhagen; 5, ectromelia cided to euthanize most of the mice maintained in the affected virus strain MP-1; 6, camelpox virus strain CP-1; 7, ectromelia virus holding room for biosecurity, there was no evidence of virus isolate C99-505; and 8, lysate of spleen and liver from C99-505. spread to animals other than those inoculated with the virus- room was based on risk assessment. The potential of environ- contaminated cell line. Neither the sentinel mice that were ex- mental contamination with poxvirus was considered low, whereas posed to soiled bedding at the time of the outbreak, nor mice use and containment of toxic formaldehyde gas posed considerable from the room and retained for subsequent evaluation, devel- concerns. The room was determined safe for housing laboratory oped clinical signs of infection, lesions, or serum antibodies con- mice after outbred SW mice and the sensitive DBA/2 strain were sistent with EV infection. maintained in the room for several weeks in open-top cages with- We attribute the sensitivity of the CAF1 hybrid mouse strain, out developing clinical disease or EV serum antibody. which sustained appreciable mortality when infected with EV, As reported (7, 12, 15, 24), serologic testing is of minimal ben- as the principal reason the outbreak was confined. Both hybrid efit in diagnosing mousepox in the face of an outbreak. The in- partners, BALB/c and A strains, are highly sensitive to EV and dex cases examined in our outbreak and the mice inoculated with are used as EV-sensitive mouse strains in experimental studies EV-contaminated mouse serum that died on or before PID 7 did of mousepox (21, 22). The sensitivity of these strains has been not develop detectable EV antibody. All three SW mice inoculated attributed, in part, to a reduced Th1 cytokine response, neces- with EV-contaminated serum, which survived to day 12, developed sary to generate a potent cellular immune response to the virus antibody to the virus. (22). Additionally, the static isolator cages in use would limit Although histologic changes observed in tissues collected spread of the virus, as documented by Bhatt and Jacoby (23), from the index cases were highly suggestive of mousepox, con- who successfully used this caging system to provide contain- firmatory testing by use of EM to detect viral particles in nega- ment when conducting experimental studies of mousepox. tively stained tissue homogenates, PCR to detect viral nucleic Our decision to forego the routine practice of formaldehyde acid, or IFA testing to detect viral antigen in tissue are recom- gas decontamination of the potentially contaminated holding 432 Mousepox Associated with Contaminated Mouse Serum

mended for rapid confirmation in suspect cases (2). To the authors’ served in only two and four of the seven mice inoculated with knowledge, cases of other naturally acquired orthopoxviruses in serum, respectively. In contrast, all five animals that died or laboratory mice have not been reported; however, results of a were euthanized on or before PID 12 developed interstitial recent study (25) indicated that laboratory mice are susceptible pneumonia, with or without accompanying pulmonary effusion. to cowpox infection when virus is administered via the com- Interestingly, the mouse receiving serum from the index cases bined intradermal and subcutaneous routes. Polymerase chain developed pulmonary effusion and interstitial pneumonia de- reaction analysis and tissue IFA testing were used to confirm spite not receiving serum via the intranasal route. Another mousepox in our mice before commencing depopulation. We mouse inoculated via the intranasal and intraperitoneal routes were unable to detect EV in the contaminated pooled commer- developed only interstitial pneumonia. Splenic or hepatic necro- cial mouse serum by use of ultrastructural examination or virus sis was not observed and EV DNA from these two sites in this isolation, reflecting the limited sensitivity of these techniques. animal was not amplified by use of PCR analysis, although EV In contrast, virus was detected using PCR, immunohistochemis- was detected in the lung. Intraperitoneal inoculation of EV try, and mouse inoculation. It should be noted that only a small would be expected to result preferentially in spleen and liver aliquot (75 ␮l) of virus-contaminated commercial pooled mouse infection (21). These findings suggest that the EV isolate associ- serum was available to attempt virus isolation. Isolation may ated with this outbreak may be more pneumotropic than has have been possible had a larger aliquot been available. Dick et been previously observed. Early investigators suggested that al. (15) were only able to isolate virus from a 500-␮l aliquot of EV, after serial lung passage, developed pneumotropic proper- pooled commercial mouse serum after repeated passaging, sug- ties (31). However, Fenner discounted this finding and inter- gesting that the concentration of virus in the sample was likely preted the apparent pneumotropism to be the result of intranasal small, as we also suspect in the serum associated with the out- administration of large quantities of virus (32). He concluded that break in our mice. Considering the difficulty, expense, and lim- severe pulmonary lesions developed in mice receiving large ited sensitivity of virus isolation, it is not recommended for amounts of virus, resulting in the animal’s death from localized routine screening of contaminated biological material for EV. pulmonary reaction before the virus spread to infect other organs. Although the quantity of virus in the contaminated serum Necrosis, in some cases severe, observed in the ovary, bone was suspect to be low, Karupiah has reported that as little as marrow, and small intestine of selected index cases and mice in- one infective viral particle can result in 100% mortality in BALB/c oculated with contaminated serum, also has been reported (15, 16). and A strain mice (22). Because the contaminated serum was in- Neither diffuse generalized exanthema nor conjunctivitis was ob- advertently used to supplement the growth of murine spindle served in the index cases or animals inoculated with the EV- cells, the virus may have been propagated to high titer in cul- contaminated test articles. Intracytoplasmic inclusions associated ture. Ectromelia virus has been documented to cause persistent with ballooning epithelial cell degeneration were observed in the infection of BALB/c mice-derived splenic dendritic cells and skin from the inoculation site where the virus-contaminated cells macrophages, producing active virus in culture for up to PID 60 were obtained from a single index case. Although we suspect simi- (26). Buller et al. documented EV’s ability to replicate in vivo in lar lesions were present in the remaining index cases, skin sections lymphoma cell lines and some hybridomas (27). were not collected from these sites. On PID 7 and 25 post EV- The resistance of EV to elimination by use of routine methods contaminated serum administration, three mice developed skin le- was clearly indicated in this outbreak. The virus-containing se- sions. Several small ulcers were observed on the skin overlying the rum was filtered (0.2 ␮m) by the distributor and heated (56ЊC cranium in two SW mice housed together and found dead on PID for 30 minutes) by the laboratory prior to its use. Ectromelia vi- 7. The lesions were suggestive of bite wounds, and epithelial cell rus is resistant to many disinfectants and heat, characteristics degeneration or inclusions were not observed. Additionally, the that enable it to persist in the environment. Although Marchal time elapsed after inoculation was likely too short for the lesions to described complete inactivation of EV in saline at 55ЊC for 30 be attributed to EV (32). The ulcerative muzzle lesion observed in minutes, Bhatt and Jacoby documented that the NIH-79 strain the SW mouse euthanized on PID 25 post EV-contaminated serum of EV was considerably more stable in FBS, compared with sa- administration was also devoid of features characteristic of mouse- line (28, 29). They also documented that EV persisted for 26 pox, although the lesion may have healed sufficiently so that inclu- weeks when maintained at 4ЊC in 100% FBS; that titer was higher sion bodies were no longer present (33). under several environmental conditions when EV was maintained There currently are no policies or procedures that might pre- in 100%, compared with 10% FBS in cell culture medium; and EV vent introduction of mousepox and other infective agents from was only inactivated when virus-laden mouse serum, diluted imported rodent sera. Chinese export requirements mandate 1:10 in saline, was heated at 60ЊC for 30 minutes (29). We sus- that serum be tested for EV antibody. We were unable to detect pect, on the basis of these findings and our observations, that EV antibody in the serum lot that contained virus based on EV maintained in undiluted mouse serum is better protected PCR analysis and animal inoculation. Although blood from indi- from thermal inactivation than is dilute sera. vidual animals used to prepare the serum lot may have had de- Gross and microscopic lesions observed in the index cases tectable antibody titer, the antibody would be diluted below the from this outbreak were similar to those described in sensitive detection limit of serologic assays when pooled with serum from mouse strains in other outbreaks and after experimentally in- other animals. Additionally, animals undergoing acute infection duced infection with EV (15, 21, 30). The disease hallmarks of lym- would have virus-positive, antibody-negative serum. Clearly the phoid and hepatic necrosis were observed in all index cases. Of colonies from which the serum was obtained were exposed to a note was the pathologic changes observed in mice inoculated variety of adventitious murine pathogens since antibody to a with either pooled serum from the index cases or virus-contami- variety of infective agents was detected. Although transmission nated commercial serum. Hepatic and splenic necrosis was ob- of murine-specific agents is of concern, the potential for intro- 433 Vol 50, No 4 Comparative Medicine August 2000 ducing a zoonotic or exotic pathogen should also be considered. place, could hasten transmission of EV and expand the number The importer indicated that mice used as serum donors, for the of institutions affected by an outbreak, if mice infected with EV batch of serum from which our lot was acquired, were obtained were transferred unwittingly between institutions. from several mouse colonies in China and brought together, then kept for approximately a week during which blood was collected multiple times. Depending on the caging and husbandry systems Acknowledgments in place, re-use of blood collection equipment without adequate We thank Robert Jacoby for immunohistochemical staining and sanitization, and the techniques used, it is possible that EV could evaluation of ectromelia virus-containing tissue sections, Pravin Bhatt have been transmitted to immune-naive mice during this period. and Abigail Smith for their expert advice, and Jo Ann Henry and Karen Jackson for technical assistance. Subsequently, blood may have been collected from a large number of mice during the viremic stages of EV infection. It is interesting to estimate the volume of EV-contaminated serum, and therefore the number of mice, necessary to contami- References 1. Briody, B. A. 1955. Mousepox (ectromelia) in the United States. nate a 43-L serum batch so that a typical aliquot administered to Proc. Anim. Care Panel 6:1–8 a mouse would transmit EV-causing infection. On the basis of in- 2. Fenner, F. 1994. Mousepox (ectromelia). In A.D.M.E. Osterhaus formation provided by the importer, a 43-L batch of serum (ed.), Virus infections of rodents and lagomorphs, p. 5–25. Elsevier would require blood collection from approximately 123,000 Science B.V., Amsterdam, The Netherlands. mice. Blood contains 10,000 EV infective particles/ml during the 3. Fenner, F. 1981 Mousepox (infectious ectromelia): Past, present, and future. Lab. Anim. Sci. 31:553–559. peak of the viremic phase of infection, and a single infective par- 4. Held, J. R. 1981. Summary (ectromelia (mousepox) in the United ticle is reported to be sufficient to induce disease (22, 33). There- States. Lab. Anim. Sci. 630–632. fore, as few as 246 of 123,000 mice, or < 0.2 %, would need to be 5. Wagner, J. E., and R. A. Daynes. 1981. Observations of an viremic to contaminate a 43-L serum batch if the estimated se- outbreak of mousepox in laboratory mice in 1979 at the Uni- versity of Utah Medical Center, USA. Lab. Anim. Sci. 31: rum dose per mouse was 0.05 ml, the infective particles re- 565–569. mained in the serum after coagulation, and the virus was evenly 6. Whitney, R. A., Jr., J. D. Small, and A. E. New. 1981. Mouse- distributed in the batch. pox - National Institutes of Health experiences. Lab. Anim. Sci. The USDA’s current testing program on imported murine 31:570–573. sera is limited to verification that the sera are of murine origin 7. Manning, P. J., and C. S. Frisk. 1981. Clinical, pathologic, and serologic features of an epizootic of mousepox in Minnesota. Lab. and not derived from livestock. The USDA has authority over Anim. Sci. 31:574–577. importation of EV and laboratory mice, and could require more 8. Smith, E. K., P. M. Hartroft, and M. H. Greider. 1981. Pre- stringent testing, if they determined it necessary (14). We be- liminary report of an outbreak of ectromelia (mousepox) at Wash- lieve it to be in the best interest of U.S. biomedical research pro- ington University in St. Louis. Lab. Anim. Sci. 31:585–589. 9. La Regina, M. C. and R.E. Doyle. 1981. Mousepox at St. Louis grams for the USDA to require testing of imported murine University-Preliminary report. Lab. Anim. Sci. 31:583–584. serum by PCR analysis or animal inoculation. Professional or- 10. Dixon, L. W. 1981. Control of mousepox epizootics in St. Louis ganizations should actively lobby the USDA to change their re- and Chicago. Lab. Anim. Sci. 31:578–589. quirements for rodent serum importation. Alternatively, do- 11. Werner, R. M., A. M. Allen, J. D. Small, J. D., et al. 1981. Clini- mestic colonies should be considered preferentially as a source for cal manifestations of mousepox in an experimental animal holding room. Lab. Anim. Sci. 31:590–594. pooled murine sera; however, the NMRI outbreak was report- 12. Allen, A. M., G. L. Clarke, J. R. Ganaway, et al. 1981. Pathol- edly associated with domestic sera (15). ogy and diagnosis of mousepox. Lab. Anim. Sci. 31:599–608. The identification of the serum source, the size of the lot from 13. Wallace, G. D., R. M. Werner, P. L. Golway, et al. 1981. Epi- which it originated, and its distribution throughout the United zootiology of an outbreak of mousepox at the National Institutes of Health. Lab. Anim. Sci. 31:609–615. States by many suppliers raises considerable concern about the 14. Small, J. D., and A. E. New. 1981. Prevention and control of potential risk to domestic mouse colonies. Although the quan- mousepox. Lab. Anim. Sci. 31:616–629. tity of serum imported annually into the United States is un- 15. Dick, E. J., Jr., C. L. Kittell, H. Meyer. et al. 1996. Mouse- known, the importer we contacted estimated the quantity to be pox outbreak in a laboratory mouse colony. Lab. Anim. Sci. 46: between 500 and 1,000 L. Because mouse serum is generally 602–611. 16. Jacoby, R. O., and P. N. Bhatt. 1987. Mousepox in inbred mice sold to the end user in small aliquots (approx. 10 ml) or may be innately resistant or susceptible to lethal infection with ectrome- processed to obtain a variety of serum-derived products, the risk lia virus. II. Pathogenesis. Lab. Anim. Sci. 37:16–22. of distribution and use of EV-contaminated material is high. 17. Meyer, H., M. Pfeffer, and H. J. Rziha. 1994. Sequence alter- Therefore, it is not surprising that there have been two addi- ations within and downstream of the A-type inclusion protein genes allow a differentiation of orthopoxvirus species by poly- tional mousepox outbreaks in the United States during the past merase chain reaction. J. Gen. Virol. 73:1975–1981. year (34). Investigative staff, veterinarians, and animal re- 18. Neubauer, H., M. Pfeffer, and H. Meyer. 1997. Specific detec- source personnel should be aware of this potential and vigi- tion of mousepox virus by polymerase chain reaction. Lab. Anim. lantly monitor their colonies. An important consideration, 3:201–205. which was not an issue during the last major U.S. mousepox 19. Importer. Personal communication. 20. Bhatt, P. N., W. G. Downs, S. M. Buckley, et al. 1981. Mouse- outbreak two decades ago, is widespread development and use pox epizootic in an experimental and a barrier mouse colony at of gene-targeted mice. Thousands of unique transgenic and Yale University. Lab Anim Sci. 31:560–564. knock-out mouse lines, many of which are maintained at a 21. Bhatt, P. N., and R. O. Jacoby. 1987. Mousepox in inbred mice single institution and may be susceptible to fatal EV infection, innately resistant or susceptible to lethal infection with ectromelia virus. I. Clinical responses. Lab. Anim. Sci. 37:11–15. could be permanently lost. Also exchange of these unique mouse 22. Karupiah, G. 1998. Type 1 and type 2 cytokines in antiviral de- lines between institutions around the globe, which is common- fense. Vet. Immunol. Immunopathol. 63:105–109. 434 Mousepox Associated with Contaminated Mouse Serum

23. Bhatt, P. N., and R. O. Jacoby. 1983. An inexpensive containment 28. Marchal, J. 1930. Infectious ectromelia: A hitherto undescribed laboratory for mousepox research [abstr]. Lab. Anim. Sci. 33:495. virus disease of mice. J. Pathol. Bacteriol. 33:713–728. 24. Collins, M. J., Jr., R. L. Peter, and J. C. Parker. 1981. Sero- 29. Bhatt, P. N., and R. O. Jacoby. 1987. Stability of ectromelia logical detection of ectromelia virus antibody. Lab Anim Sci. virus strain NIH-79 under various laboratory conditions. Lab. 31:595–598. Anim. Sci. 37:33–35. 25. Bennett, M., A. J. Crouch, M. Begon, et al. 1997. Cowpox in 30. Wallace, G. D., and R. M. L. Buller. 1985. Kinetics of ectrome- British voles and mice. J. Comp. Pathol. 116:35–44. lia virus (mousepox) transmission and clinical response in C57BL/ 26. Spohr De Faundez, I., Gierynska, M. G. Niemialtowski, et 6J, BALB/cByJ and AKR/J inbred mice. Lab. Anim. Sci. 35:41–46. al. 1995. Ectromelia virus establishes a persistent infection in 31. Kikuth, W., and R. Gonnert. 1940. Erzeugung von Ektromelie spleen dendritic cells and macrophages of BALB/c mice following durch Provokation. Arch. Ges. Virusforsch. 1:295–312. the acute disease. Adv. Exp. Med. Biol. 2:257–261. 32. Fenner, F. 1949. Mouse-pox (infectious ectromelia of mice): A re- 27. Buller, R. M. L., A. C. Weinblatt, A. W. Hamburger, et al. view. J. Immunol. 63:341–373. 1987. Observation on the replication of ectromelia virus in mouse- 33. Fenner, F. 1948. The clinical features and pathogenesis of mouse- derived cell lines: Implications for epidemiology of mousepox. Lab. pox (infectious ectromelia of mice) J. Pathol. Bacteriol. 60:529–552. Anim. Sci. 37:28–32. 34. Lipman, N. S. Personal communication.

435