Differentiation of variola, monkeypox, and vaccinia antisera by radioimmunoassay*

Bulletin of the World Health Organization. 55 (5): 613-623 (1977) Differentiation of variola, monkeypox, and vaccinia antisera by radioimmunoassay * HARRIET D. HUTCHINSON,1 DONALD W. ZIEGLER,2 DAVID E. WELLS,3 & JAMES H. NAKANO 4 Poxvirus antisera adsorbed with " homologous " and " heterologous " poxvirus-infected chorioallantoic membranes (CAM) were differentiated by solid-phase radioimmunoassay (RIA). Mixtures of the antiserum dilutions and infected CAM were added directly (without centrifugation) to poxvirus-infected CAM antigens affixed to wells ofmicrotitra- tion plates. The affixed antigens combined with unadsorbed antibodies, and the cross- reactive antigen-antibody complexes were removed by washing. The results showed that adsorption of an antiserum with variola-, vaccinia-, or uninfected-CAM antigen and subsequent reaction ofeach in RIA with monkeypox- and uninfected-CAM antigens allowed the identification of antivariola, antivaccinia, or antimonkeypox sera. The variola/vaccinia subgroup of the Orthopox- virus genus consists of a number of biologically and serologically related viruses. Among the members of this subgroup, variola, monkeypox, and vaccinia viruses are important because they cause infections in man (2). It is also important in diagnostic virology to be able to differentiate antibodies to each of the viruses. Members of the subgroup are characterized and categorized primarily by their biological properties. The different viruses show recognizable growth char- acteristics on chicken embryo chorioallantoic mem- branes (CAM), rabbit dermis, and in tissue culture. Recently, about 30 proteins each for variola, vaccinia, and monkeypox viruses (5) were enumer- ated by polyacrylamide gel electrophoresis. Most of these proteins are shared by the members of the subgroup; however, high resolution electrophoresis showed characteristic, unique proteins for each of these subgroup members. * Viral Immunology and Viral Exanthems Branches of the Virology Division, Center for Disease Control, Public Health Service, US Department of Health, Education, and Welfare, Atlanta, GA 30333, USA. Use of trade names in this article is for identification only and does not constitute endorsement by the Public Health Service or by the US Department of Health, Education, and Welfare. Research Microbiologist, Viral Immunology Branch. 'Chief, Viral Immunology Branch. Microbiologist, Viral Immunology Branch. 'Chief, Viral Exanthems Branch. Because of multiple specificities of antibodies di- rected towards component antigens, identifying anti- sera raised against the individual viruses has been difficult. Results of neutralization tests, haemaggluti- nation inhibition (HI) tests and immunodiffusion tests (4, 7, 9, 15, 22) show that antisera to the various subgroup members are strongly cross-reactive with heterologous viruses within the subgroup. In spite of the cross-reactivity, adsorption studies have allowed the unique antigens among the members of the poxvirus subgroup to be defined (1, 4, 9). Gispen & Brand-Saathof (9) and Esposito et al. (4) used selective adsorption to prepare specific antisera that permit identification of variola, vaccinia, or monkeypox viruses. Multiple adsorptions of potent antisera with heterologous viruses yielded sera that produced one or two immunodiffusion precipitin lines with homologous viruses. The precipitin lines obtained with adsorbed sera probably involve strain- specific antigens not shared by other related viruses of the subgroup. The antisera have been used to identify antigenically related viruses. Recently, Gis- pen et al. (10) described an immunofluorescent assay in which they differentiated vaccinia and monkeypox antisera after adsorption with homologous and he- terologous antigens. They did not, however, include antiserum to variola virus. An alternative procedure involved adsorbing pa- tients' sera with poxvirus antigens and subsequently characterizing residual antibodies with a sensitive 3646 - 613- H. D. HUTCHINSON ET AL. radioimmunoassay (RIA), which apparently allowed poxvirus antisera identification. Ziegler et al. (27) described an RIA for detecting antibodies to pox- viruses that was more sensitive than complement fixation (CF), HI, or plaque reduction neutraliza- tion (PRN). In this study we adsorbed antisera with homologous and heterologous antigens, and deter- mined the specific reactivity of the residual antibody activity by RIA. It was possible to differentiate antisera to closely related poxvirus strains by their adsorption patterns. MATERIALS AND METHODS Virus strains Poxvirus strains were obtained from the Viral Exanthems Branch, CDC, as CAM passage materi- als. Except for vaccinia, the viruses were originally isolated from crust materials from infected patients or infected nonhuman primates. The strains used are listed below. - Wyeth strain of vaccinia (Vac) was originally obtained from Wyeth Laboratories, Philadelphia, PA. - Harvey (Har), a variola (Var) major strain, ob- tained from Professor K. R. Dumbell, was iso- lated from a smallpox outbreak in Middlesex, England, 1944. - African smallpox (ASP), a variola major strain, was isolated from a patient in Sierra Leone in 1968 (CDC No. V68-I-258) (13). - Chimp 9 (Ch-9) virus, a whitepoxvirus strain (variola-like) was isolated from an apparently healthy chimpanzee in Zaire in 1971 (CDC No. V71-I-224) (16). - Utrecht monkeypox virus (MMP) was isolated from an orangutan in Rotterdam, Netherlands in 1965 during a poxlike epizootic outbreak (20). - Human monkeypox virus (HMP), a monkeypox virus strain, was isolated from a patient in Sierra Leone in 1970 (CDC No. V70-I-266) (13). To confirm the identity of the poxviruses used in this study, the Viral Exanthems Branch, CDC, tested coded samples of the poxvirus-infected CAM prepa- rations. Identifications were made on the basis of the pock characteristics on CAM as described by Na- kano (17). Antigen preparation Antigens for RIA and selective adsorption of sera were prepared by inoculating 12-day-old chicken CAM with 0.1-ml aliquots of dilutions (usually 10-2) of the virus preparations as described by Nakano & Bingham (18). Uninfected control CAM was pre- pared by inoculating the eggs with Mcllvaine's buffer (18). After 3 days of incubating the eggs at 35.5°C the membranes were harvested and frozen once at -700C. Membrane antigens were prepared by thaw- ing the membranes and homogenizing in a Sorvall omnimixer (three 1-min cycles) at 4°C. The suspen- sions were centrifuged for 10 min at 900 g. The supernatant was collected and stored in small ali- quots at -70°C. Each viral antigen preparation was titrated on CAM and the titre was reported as pock- forming units per millilitre (PFU/ml). Generally, antigen preparations containing more than 107 PFU/ml were satisfactory for RIA antigen or ad- sorption antigen. The protein concentrations of the RIA antigens were determined (14) and were ad- justed to 10 mg/ml before use. Antiserum preparations Hyperimmune antisera were prepared in albino rabbits by the Viral Exanthems Branch, CDC, as described by Esposito et al. (4). Each poxvirus anti- gen was prepared in primary rabbit kidney tissue culture cells. When the cytopathic effect was 100 %, infected cells in each 950-ml (32-oz) prescription bottle were scraped from the glass, centrifuged, and resuspended in 3 ml of Mcllvaine's buffer. The virus was dissociated from the cells by freezing and thaw- ing. All viral antigens had titres greater than 107 PFU/ml on CAM. Each rabbit was initially inocu- lated with a homogenate containing equal volumes of poxvirus antigen and Freund's complete adjuvant. 0.25 ml of this viral emulsion was inoculated into both front footpads and 0.5 ml into the right hindleg muscle. A booster inoculum containing only the poxvirus antigen was given 24 days later. 1 ml was given subcutaneously in the neck and 0.5 ml in the right leg muscle. The rabbits were exsanguinated 18 days after the booster inoculations. Antisera were prepared to Wyeth Vac virus (anti- Vac; CF titre > 512, HI titre = 320, PRN titre = 25 000), ASP virus (antiVar; CF titre = 512, HI titre = 160, PRN titre = 1800), Ch-9 virus (antiCh-9; CF titre = 1024, HI titre = 640, PRN titre = 3200), MMP virus (antiMMP; CF titre = 512, HI titre = 640, PRN titre = 9000), and HMP 614 DIFFERENTIATION OF POXVIRUS ANTISERA virus (antiHMP; CF titre = 512, HI titre = 320, PRN titre = 8000). Aliquots of the antisera were prepared and frozen at -70°C. Standardization ofpoxvirus antigens for indirect RIA Because optimum amounts of the reagents are necessary to achieve high sensitivity and reproduci- bility in RIA, the optimum concentration of each primary antigen and 125I-antirabbit globulin used in the RIA was determined as described by Lee et al. (12). Optimum amounts of each antigen, includ- ing uninfected CAM, were determined by reacting an antiserum diluted 1: 10 with various concentrations of antigen. Controls containing diluent instead of antiserum were included for each antigen concentra- tion. After a secondary reaction with the optimum concentration (11) of 1251-antirabbit globulin, the radioactivities (cpm) of the different components were determined. Binding ratios were determined for each antigen concentration by the following formula: (X - Y + Z)/Z where X = (cpm) antiserum against homologous antigen Y = (cpm) antiserum against uninfected antigen Z = (cpm) diluent control. The data were evaluated by plotting the binding ratios on the ordinate and the logarithm (log) of the protein concentration on the abscissa (Fig. 1). For each poxvirus antigen shown (Vac, Har, Ch-9, and UMP), the highest ratios were obtained with 15.6 ,tg of protein per well. Hence antigen preparations with this concentration of protein were used in the RIA procedures. Although the figure represents single lots of antigen preparations, results of evaluating subse- quent lots in an identical manner were comparable. Adsorption procedure Twofold (or threefold) antiserum dilutions pre- pared in phosphate buffered saline (PBS) containing 2000 fetal calf serum and 0.04 %Y sodium azide (PBS diluent) were transferred to separate tubes in 0.2-ml aliquots, and an equal volume of infected-CAM or uninfected-CAM adsorption antigen in PBS was added to each dilution. In preliminary experiments, PBS diluent was also added to a series of serum dilutions as controls. The serum titrations were mixed and held at 4°C for 16 to 18 h. Immediately before they were added to the antigen-containing microtitration plates, the serum dilutions were incubated for 30 min at 36°C. 7.0- 6.0 5.0- o 4.0 1 3.0- 2.0 1.0 0* ~~~ N ~~A 0.98 1.95 3.9 7.8 15.6 31.2 jig Protein per Well 62.5 125 Fig. 1. Poxvirus antigen titrations. Various concentra- tions of each antigen were reacted with a single (1 :10) dilution of ' homologous" antiserum. Ratios were calculated as indicated above. (A) Har antigen against antiVar serum; (B) Vac antigen against antiVac serum; (C) Ch-9 antigen against antiCh-9 serum; (D) MMP antigen against antiMMP serum. RIA procedure The indirect RIA procedure for poxviruses was previously described by Ziegler et al. (27). An opti- mal concentration of infected and uninfected CAM antigen materials was prepared in PBS and added to the wells of microtitration plates. The plates were air-dried overnight at room temperature, and fixed for 15 min with 10% buffered formaldehyde. Sequen- tial serum dilutions (0.025 ml aliquots) adsorbed with homologous or heterologous antigens were added to both homologous and heterologous CAM antigens and uninfected CAM antigen and allowed to incubate for 2 h at 36°C. After the plates were washed with PBS, an optimal concentration of 1251_ antirabbit globulin (11) was added to each well. The plates were incubated for 1 h at 360C and washed with PBS. Each sample was cut from the microtitration plate and was counted in a gamma scintillation spectro- meter. End-point dilution factors (titres) were deter- mined as described by Hutchinson & Ziegler (11). Titres were obtained by plotting the log of the specifically bound radioactivity [(cpm) antiserum against infected CAM-(cpm) antiserum against un- infected CAM] + (cpm) diluent control on the ordinate and the log of the reciprocal antiserum dilutions on the abscissa. The antiserum titration curves were interpolated or extrapolated to the reciprocal antiserum dilution at which the bound radioactivity equalled twice the diluent control. I 1- I I 615 H. D. HUTCHINSON ET AL. RESULTS Antiserum titrations with homologous antigens Viruses of the family Poxviridae contain mul- tiple common protein components. Thus, antibodies prepared against one of the viruses of the group have multiple specificities which cross-react with related viruses. Before examining the antigenic relationships among closely related poxviruses, we titrated each antiserum against its homologous antigen, compar- ing the results with those obtained when the anti- serum was titrated against heterologous antigens. Titration curves (log-log plots) of Vac-antiVac, Ch- 9-antiCh-9, MMP-antiMMP, and Har-antiVar are shown in Fig. 2. Except for the Har-antiVar curve, each titration curve was linear and essentially paral- lel to the others. The plot of Har-anti Var was s00o- 400- 300- 200- E 1500- lSoo A . D ' ____\3 2 xDiluent Control Diluent Controls 125 250 500 1000 2000 4000 8000 Reciprocol Antiserum Dilution Fig. 2. ' Homologous antiserum titrations. The hori- zontal lines represent control (cpm) x 2. (A) Har antigen against antiVar serum; (B) Vac antigen against antiVac serum; (C) Ch-9 antigen against antiCh-9 serum; (D) MMP antigen against antiMMP serum. slightly curvilinear; however, with repeated titrations the end-point titre of the antiserum was reproduc- ible. Antiserum titrations against both homologous and heterologous antigens consistently yielded curves of similar slopes. Selection ofantigen concentrations for adsorption Before attempting to adsorb antibodies from anti- sera with related antigens, we determined the quan- tity of homologous antigen required to remove virtually all antibodies from an antiserum. Each antiserum was adsorbed with 0.1, 1.0, or 10 mg of antigen per millilitre before it was titrated with homologous antigen. The titration curves obtained for antiCh-9 serum (Fig. 3) show that larger amounts of antibody against Ch-9 virus were removed as the quantity of homologous adsorbent antigen was in- 4000- 3000- 2000- 1500 900- 800- 700-6005000 400- ~~~@ .>>~~~~,:_> I. -N&B Diluent IControls 64 128 256 512 1024 2048 4096 8192 Reciprocal Antiserum Dilution Fig. 3. Adsorption of antiCh-9 serum with various concentrations of Ch-9 antigen. (A) No adsorption, or adsorption with each of the following: (B) 0.1 mg/ml; (C) 10 mg/ml; (D) 10 mg/ml of Ch-9 antigen. creased. The highest concentration (10 mg/ml) of adsorbent antigen removed virtually all of the anti- bodies. The end-point titres were <2.3 %, 5.9 %, and 16.2% of the unadsorbed titre values after adsorp- tion with 10, 1.0, and 0.1 mg of the antigen per millilitre, respectively. Like the antiCh-9 serum adsorptions, antiMMP, antiVac, and antiVar serum adsorptions with their homologous antigens reduced the antiviral antibo- dies in amounts proportional to the amount of adsorbent used. Adsorption with 10 mg of homolo- gous antigens per millilitre removed more than 95% of the antibody activity of each antiserum (Table 1). In subsequent experiments with antisera adsorbed with their homologous antigens (10 mg/ml) the antibody titres were consistently reduced by more than 95 %. Although we were unable to remove antibody activity completely by a single adsorption with 10 mg of the adsorbent antigens per millilitre, it was impractical to use higher concentrations. Adsorption of antisera with uninfected CAM Poxvirus antigens used in these investigations were crude suspensions prepared by homogenizing pox- virus-infected CAM. Thus, in addition to viral anti- gens, the preparations contained large amounts of host tissue that could react nonspecifically. The effect of adsorption of antisera with uninfected CAM homogenates was therefore examined. As shown in Table 2, the homologous titres of CAM-adsorbed antisera were 1.3-3.9 times higher than the titres obtained with the antisera without adsorption with CAM. The increased titres, after adsorption, were par- tially attributed to decreased binding of 1251-antirab- bit globulin to the uninfected-CAM controls (non- specific). However the major reason for increased 616 DIFFERENTIATION OF POXVIRUS ANTISERA Table 1. Adsorption of poxvirus antisera with various concentrations of homologous antigens Adsorbing antigen concentration Antigen against Unadsorbed 10 mg/ml 1.0 mg/ml 0.1 mg/mlantiserum Titre a Titre a % b Titre a % b Titrea % b UMP antiUMP 8 000 <250 <3.1 580 7.2 2000 25 Vac antiVac 1 680 <64 <3.8 88 5.2 560 33 Har antiVar 5500 <250 <4.5 285 5.2 1 300 24 Ch-9 antiCh-9 10800 <250 <2.3 640 5.9 1 750 16 a Titre is the end-point dilution factor. b Percentage of original (unadsorbed) antibody titre. titres was the enhanced binding of 1251-antirabbit globulin to the poxvirus-infected antigens (specific). Because this phenomenon was reproducible, adsorp- tion of antisera with uninfected CAM antigen was adopted as a part of the standard adsorption proce- dure. Homologous and heterologous antiserum adsorptions The antigenic relationships among five orthopox- viruses, Ch-9, Har, ASP, MMP, and Vac, were evaluated by RIA. Five antisera, antiVar, antiCh-9, antiVac, antiMMP, and antiHMP, were each ad- sorbed with six antigens, uninfected CAM, Ch-9, Har, ASP, MMP, and Vac viruses, and were tested by RIA for residual antibodies reactive with both homologous and heterologous antigens. The results of the adsorptions of antiVar, antiVac, and antiMMP sera are shown in Tables 3-5. Gen- erally, the CAM-adsorbed antisera had the highest Table 2. Antisera adsorbed with normal CAM antigen Unadsorbed CAM adsorbed a Antigen Antiserum Titreb Tteb FoldincreaseC Ch-9 antiCh-9 10 076 12 660 1.3 Har antiVar NTd 6 498 - Vac antiVac 853 3 292 3.9 MMP antiMMP 4 485 6 884 1.5 MMP antiHMP 2 202 4123 1.9 a I 0 mg of uninfected CAM antigen per millilitre used for adsorption. b Titre is the end-point dilution factor. c Fold increase in titre as compared to the unadsorbed titre. d NT=not tested. titres with homologous antigens. In some cases, how- ever, using this measurement as the sole criterion for differentiating the antisera could lead to equivocal or erroneous assignment of strain specificity. As an example, antiVar serum reacted with MMP antigen yielded a higher titre than it did when it was reacted with ASP antigen. This fact indicates the importance of considering the titres obtained after adsorption. The results of homologous and heterologous ad- sorption of antiVar serum are shown in Table 3. Adsorption of this antiserum with Ch-9, Har, or ASP antigen removed essentially all antibody activity to Ch-9, Har, ASP, and Vac antigen. However low residual antibody activity which was reactive with MMP antigen remained after this adsorption. Con- versely, adsorption of antiVar serum with MMP antigen yielded a serum with substantial antibody activity (8.0 °-18.8 %) to Ch-9, Har, and ASP anti- gens. The adsorbed antiserum had low antibody activity to MMP antigen (0.4%Y) and Vac antigen (2.8 %). The same antiserum (antiVar) adsorbed with Vac antigen resulted in a serum with antibody activity to Ch-9, Har, ASP, and MMP viruses, but none to Vac virus. Next, similar adsorptions of antiVac serum were examined (Table 4). After the serum had been adsorbed with Ch-9, Har, or ASP, most of the antibody activity to these antigens was removed whereas considerable antibody activity to MMP and Vac viruses remained (5.5%-1 1.4%). Adsorption of antiVac serum with MMP antigen yielded an anti- serum with antibody activity to Ch-9, Har, ASP, and Vac viruses (5.8%, 5.0%, 4.8%, and 3.7%, respec- tively) but completely removed the antibody activity to MMP antigen. After antiVac serum was adsorbed with homologous antigen, only very low residual 617 618 H. D. HUTCHINSON ET AL. Table 3. Homologous and heterologous antiserum titrations of antiVar serum Antigen reacted with Antiserum adsorbed Ch-9 Har ASP MMP Vac with - Titre a % b Titre a % b Titre a % b Titre a %b Titre a % b CAM 5510 6493 5062 5598 2914 Ch-9 9 0.2 38 0.6 34 0.7 108 1.9 <8 <0.3 Har 27 0.5 51 0.8 41 0.8 95 1.7 <8 <0.3 ASP 17 0.3 25 0.4 9 0.2 52 0.9 <8 <0.3 MMP 533 9.7 1222 18.8 403 8.0 25 0.4 83 2.8 Vac 392 7.1 494 7.6 178 3.5 257 4.6 <8 <0.3 a Titre is the end-point dilution factor. Underlined values indicate adsorptions or reactions with homologous antigens. b Percentage of CAM-adsorbed antibody activity remaining. Table 4. Homologous and heterologous antiserum titrations of antiVac serum Antigen reacted withAntiserum adsorbed Ch-9 Har ASP MMP Vac with _ _ _ _ _ _ _ _ _ _ _ Titrea %b Titre a% b b Titre % Titrea % b Titrea % b CAM 2081 2351 1444 2491 3400 Ch-9 <8 <0.4 <8 <0.3 <8 <0.6 225 9.0 386 11.4 Har 16 0.8 <8 <0.3 <8 <0.6 138 5.5 269 7.9 ASP 40 1.9 23 1.0 <8 <0.6 182 7.3 223 6.6 MMP 120 5.8 118 5.0 70 4.8 <8 <0.3 127 3.7 Vac 39 1.9 22 0.9 37 2.6 11 0.4 62 1.8 a Titre is the end-point dilution factor. Underlined values indicate adsorptions or reactions with homologous antigens. b Percentage of CAM-adsorbed antibody activity remaining. Table 5. Homologous and heterologous antiserum titrations of antiMMP serum Antigen reacted with Antiserum adsorbed Ch-9 Har ASP MMP Vac with Titre a % b Titre a % b Titre a % b Titre a % b Titre a%b CAM 1163 1111 884 3702 1807 Ch-9 <8 <0.7 <8 <0.7 <8 <0.9 1277 34.5 35 1.9 Har <8 <0.7 <8 <0.7 <8 <0.9 840 22.7 32 1.8 ASP <8 <0.7 <8 <0.7 <8 <0.9 550 14.9 26 <0.4 MMP <8 <0.7 <8 <0.7 <8 <0.9 49 1.3 <8 <0.4 Vac <8 <0.7 <8 <0.7 <8 <0.9 740 20.0 <8 <0.4 a Titre is the end-point dilution factor. Underlined values indicate adsorptions or reactions with homologous antigens. b Percentage of CAM-adsorbed antibody activity remaining. DIFFERENTIATION OF POXVIRUS ANTISERA Table 6. Homologous and heterologous antiserum titrations of antiCh-9 serum Antigen reacted with Antiserum adsorbed Ch-9 Har ASP MMP Vac with Titre a % b Titre a % b Titrea % b Titre a % b Titre a%b CAM 7911 7809 NT c 5809 4916 Ch-9 <8 <0.1 24 0.3 NT <8 <0.1 <8 <0.2 Har 15 0.2 32 0.4 NT 11 0.2 11 0.2 ASP <8 <0.1 35 0.4 NT <8 <0.1 <8 <0.2 MMP 1172 14.8 1050 13.4 NT <8 <0.1 95 1.9 Vac 401 5.1 398 5.1 NT 43 0.7 10 0.2 a Titre is the end-point dilution factor. Underlined values indicate adsorptions or reactions with homologous antigens. b Percentage of CAM-adsorbed antibody activity remaining. c NT=not tested. antibody activity remained to each of the antigens. The residual activity ranged from 0.9% to 2.6% of the CAM-adsorbed titres with no indication of specificity for any of the antigens. Residual activity was highest with Vac antigen, probably reflecting incomplete adsorption of the antibodies. Finally, homologous and heterologous adsorp- tions of antiMMP serum were studied (Table 5). When antiMMP serum was adsorbed with either Ch- 9, Har, or ASP antigens, all antibody activity to these three poxviruses was removed. Antibody activity to MMP virus remained high in the adsorbed serum (14.9% to 34.5 %), but very low residual activity to Vac antigen remained (<0.4-1.9%). Adsorption of antiMMP serum with MMP antigen completely removed antibody activity to each of the antigens except MMP. The low residual antibody activity (1.3%) to MMP antigen probably reflects incomplete adsorption and is of no consequence. Adsorption of antiMMP serum with Vac antigen produced an antiserum which was monoreactive with MMP anti- gen. The monoreactive antibody activity remaining after adsorption was 20.0% of the original antibody activity. Comparison of antiVar and antiCh-9 sera Ch-9-virus and variola-virus strains cannot be differentiated by conventional serological procedures and growth characteristics. Further, the RIA reactiv- ity patterns of antisera prepared against the two viruses and adsorbed with homologous and hetero- logous viruses are basically the same (Tables 3 and 6). However, when the antiVar serum was adsorbed with Ch-9, Har, or ASP antigen, it had low residual antibody titres of 108, 95, and 52, respec- tively, when titrated against MMP antigen. Unlike the adsorptions of antiVar, adsorption of antiCh-9 serum with Ch-9, Har, or ASP removed all antibo- dies which reacted with MMP virus in the RIA. This minor discrepancy between the reactions of the two antisera appears to reflect quantitative rather than qualitative differences. Such differences could be attributed to incomplete adsorption caused by differ- ences in the avidity of the two antisera. Comparison ofantiMMP and antiHMP sera Although the Utrecht strain of MMP virus was isolated from an orangutan in Rotterdam and HMP was isolated from a human patient in Sierra Leone, the two viruses are thought to be identical (13, 20). Several laboratories have used conventional tech- niques to investigate and confirm their identities. The two isolates have identical growth characteristics (13, 15) and cannot be differentiated by conventional serological procedures (22). Likewise, when adsorbed with homologous and heterologous viruses and tested by RLA, antiMMP and antiHMP sera had similar adsorption patterns (Tables 5 and 7). In fact, despite minor quantitative differences in the adsorption stu- dies, the two were qualitatively identical. Identification of a reaction scheme for simple separa- tion ofpoxviruses into groups The usefulness of a serological procedure depends upon its simplicity and sensitivity. The described RIA adsorption procedure is both easy and sensitive. 619 H. D. HUTCHINSON ET AL. Table 7. Homologous and heterologous antiserum titrations of antiHMP serum Antigen reacted with Antiserum adsorbed Ch-9 Har ASP MMP Vac with Titre a % b Titrea % b Titre a % b Titre a % b Titre a%b CAM NT c 811 734 1816 862 Ch-9 NT 13 1.6 9 1.2 285 15.7 60 7.0 Har NT 9 1.1 10 1.4 293 16.1 36 4.2 ASP NT <8 <1.0 <8 <1.1 126 6.9 20 2.3 MMP NT <8 <1.0 1 7 2.3 28 1.5 23 2.7 Vac NT <8 <1.0 <8 <1.1 156 8.6 <8 <0.9 a Titre is the end-point dilution factor. Underlined values indicate adsorptions or reactions with homologous antigens. b Percentage of CAM-adsorbed antibody activity remaining. c NT=not tested. Adsorptions of sera with antigens of each poxvirus group and subsequent testing of the adsorbed sera by RIA with each antigen allow identification of serum specimens. However the multiple steps make the utility of the method questionable. To determine whether a simple adsorption scheme could be estab- lished for identifying routine diagnostic specimens, we examined the results in Tables 3-7 for definitive steps for differentiating antisera. Identifying the specificity of a poxvirus antiserum depends upon evaluating the residual antibody activ- ity (percentage of original) remaining after adsorp- tion. Evaluation of the adsorption data revealed that the group identity of antisera could be determined by adsorbing individual aliquots of an unknown serum with uninfected CAM, Ch-9 (or Var), and Vac antigens and then allowing it to react with MMP and uninfected CAM antigens in RIA. When antiVar serum is adsorbed with Var antigen, a serum with very low or negligible antibody activity to MMP antigen is produced. When a second sample of the same serum is adsorbed with Vac antigen, residual antibody activity to MMP is likewise low. An antiVac serum had moderate residual antibody activity to MMP virus when adsorbed with Var antigen and low activity when adsorbed with Vac antigen. Conversely, an antiMMP or antiHMP se- rum had moderately high residual activity after adsorption with either Var or Vac antigens. Thus antisera prepared against either variola, vaccinia, or monkeypox viruses can be identified by a simple adsorption procedure. DISCUSSION The incidence of smallpox infection in man has rapidly diminished since the inception ofthe smallpox eradication programme in 1966. There have, how- ever, been sporadic smallpox cases in Somalia and Kenya during 1977 (24). Additionally, three docu- mented human monkeypox infections have been reported as recently as February and March 1977 (J. H. Nakano, WHO Collaborating Centre for Smallpox and Other Poxvirus Infections, un- published data). Human monkeypox virus is biologically indistin- guishable from viruses isolated from clinically ill monkeys (6, 13, 16) or from apparently healthy monkeys (15, 16). In addition, a virus that closely resembles variola (8, 16) has been isolated from a chimpanzee and a cynomolgus monkey. The pres- ence of these viruses in the animal population and the occurrence of sporadic human infections requires that smallpox surveillance be continued. Future sur- veillance of human and animal populations for smallpox and monkeypox infections in remote areas of the world will be made after obvious lesions have disappeared. Hence, virus isolation will be impos- sible, and only serum specimens will be collected. Therefore, a method for identifying poxvirus strain- specific antibodies will be necessary for successful retrospective studies. Serological surveys can be used only as a measure of the immunological status of a population to the Orthopoxvirus genus. Until recently, differentiation 620 DIFFERENTIATION OF POXVIRUS ANTISERA among closely related poxvirus antibodies has de- pended upon differences in the antibody titres to homologous and heterologous viruses (10, 22). Serological studies have shown repeatedly that the orthopoxviruses are antigenically related (1, 3, 7, 19, 23, 25, 26). Nevertheless, immunodiffusion proce- dures have revealed soluble antigens not shared by all of the orthopoxviruses (7, 19, 21). Specific anti- sera have also been prepared by adsorption with heterologous antigens (4, 9). These antisera were capable of differentiating vaccinia-, variola-, and monkeypox-related viruses. Members of the Orthopoxvirus genus possess anti- gens not found in other strains of the group (5). As a result, antisera raised against a specific strain would also have unique antibodies, as the results of Gispen & Brand-Saathof (9) and Esposito et al. (4) have shown. Their results suggest that selective adsorption of common antibodies from sera and subsequent identification of the residual antibodies could pro- vide a means of identifying antisera of unknown specificity. In fact, Gispen et al. (10) recently showed that adsorbed antisera to monkeypox virus and vac- cinia virus can be differentiated; antisera to variola virus were not included in the study, however. The relationships among antisera prepared against poxviruses revealed by RIA indicate that adsorption of antisera with heterologous antigens did not uni- formly produce monospecific reactivity; however adsorption with an antigen removed antibodies reactive with that antigen (i.e., homologous). The cross-reactions with heterologous viruses do not preclude the presence of monoreactive antigens because complete adsorption of cross-reactive antigens was not attempted. This study confirmed that (a) there is an antigen group on variola virus that is common to vaccinia virus but not to monkeypox virus (9), and (b) monkeypox virus contains a unique antigen group (9). Residual antibody activities were observed which seemed to reflect two antigen groups that have not been reported previously (4, 9). Variola and monkey- pox viruses have an antigenic group in common that is not found in vaccinia; and vaccinia and monkey- pox viruses have antigen(s) in common that variola virus does not have. The adsorption patterns of antibody produced to chimp-9 virus and African smallpox virus were nearly identical, as were those produced to monkey monkeypox and human monkeypox viruses. Thus these adsorption studies reconfirm that the following pairs of virus isolates are closely similar if not identical: chimp-9 virus and variola viruses; and isolates of monkeypox viruses obtained from mon- keys and those obtained from humans. The results of this study using hyperimmune rabbit sera show that the antigenic relationship among monkeypox, variola, and vaccinia allows a relatively simple adsorption scheme to be established. To determine the specificity of the antibodies, adsorbed antisera must be reacted only with crude monkeypox virus-infected and uninfected chorioallantoic mem- branes. The use of this adsorption scheme for differentiating the poxvirus specificity of diagnostic sera is currently being investigated. ACKNOWLEDGMENTS The authors express their appreciation to Mrs Denise R. Brown, Mr Brian P. Holloway, and Mrs Donna Miller for excellent technical assistance in this study. RESUME DIFFRENCIATION PAR LA MfTHODE RADIO-IMMUNOLOGIQUE DES IMMUNSiRUMS ANTI-VARIOLE, ANTI-MONKEYPOX ET ANTI-VACCINE L'identification des anticorps A l'egard des divers virus du sous-groupe variole/vaccine du genre Orthopoxvirus est malais6e en raison d'importantes r6actions croisees. Cette identification est d'un interet considerable du point de vue du diagnostic virologique, mais la mise au point de methodes permettant une differenciation indu- bitable des anticorps a l'egard de souches specifiques de poxvirus s'est rev6lee difficile. Dans le present travail, on a evalu6 une m6thode de diff6renciation des immun- serums a l'egard des virus de la variole, de la vaccine, ainsi que des virus du monkeypox de l'homme et du singe. L'adsorption des immunserums a l'egard de cha- cune de ces souches virales par des antigenes de pox- virus specifiques et la caracterisation ulterieure des anti- 621 622 H. D. HUTCHINSON ET AL. corps residuels par dosage radio-immunologique (RIA) ont fourni un moyen d'identifier les immunserums contre les poxvirus. Des dilutions de ces serums anti- poxvirus ont et6 adsorbees avec des preparations d'anti- genes provenant de membranes chorio-allantoldiennes (MCA) infectees de poxvirus # homologues > ou a hete- rologues >; ces dilutions ont ensuite ete ajout6es directe- ment (sans centrifugation) a des antigenes de MCA infectees de poxvirus, fixes A des cupules de plaques pour microtitrage. Les anticorps non adsorbes restants etaient libres de reagir avec l'antigene fixe A la phase solide, alors que les anticorps A reactivite crois6e, adsorbes par la MCA infectee de virus, n'etaient pas libres de le faire. En raison de la communaute antigenique etendue chez les virus de ce groupe, les adsorptions ne produisaient pas invariablement des immunserums ayant une reac- tivite a monospecifique# A l'egard de l'antigene homo- logue. Les adsorptions ont confirm6 la presence de deux groupes antigeniques anterieurement decrits: 1) le virus de la variole et le virus de la vaccine qui possedent un groupe antigenique commun; et 2) le virus monkeypox qui a un groupe antigenique propre. D'apres ces recher- ches, il semble en outre qu'il existe deux autres relations antigeniques, ainsi: 1) les virus de la variole et du monkey- pox ont un groupe antigenique commun que ne possede pas le virus de la vaccine; et 2) les virus de la vaccine et du monkeypox ont un groupe antigenique commun que ne possede pas le virus variolique. Les r6sultats de l'adsorption des anticorps A l'6gard du virus chimp-9 et du virus de la variole (variole africaine) etaient presque identiques; il en etait de meme dans le cas des anticorps produits A l'egard des virus de monkey- pox simien et du monkeypox humain. Ainsi, ces etudes confirment une fois de plus la grande similitude, sinon l'identite, du virus chimp-9 et du virus de la variole, d'une part, ainsi que des virus monkeypox isol6s des singes et de l'homme, d'aut're part. La methode d'adsorption a fourni un moyen pratique de differencier les immunserums A l'egard des virus de la variole, du monkeypox et de la vaccine. II n'est pas necessaire d'adsorber les echantillons de s6rums avec chacun des adsorbants viraux et de d6terminer les titres residuels contre chacun des virus du sous-groupe. Les immunserums ont et6 d'abord adsorbes par de la MCA non infectee, du virus de la vaccine ou de la variole; puis on les a fait reagir avec de la MCA non infectee et de la MCA infectee de virus monkeypox en RIA. Les immunserums A l'egard de chacun de ces virus ont presente une activite anticorps residuelle caract6ristique a l'egard du virus monkeypox apres adsorption. REFERENCES 1. BAXBY, D. A comparison of the antigens present on the surface of virus released artificially from chick cells infected with vaccinia virus, and cowpox virus and its white pock mutant. J. Hyg (Lond.), 70: 353- 365 (1972). 2. BAXBY, D. Identification and interrelationships of the variola/vaccinia subgroup of poxviruses. Progr. med. Virol., 19: 215-246 (1975). 3. DOWNIE, A. W. & MACDONALD, A. A study of the poxviruses by complement fixation and inhibition of complement fixation methods. J. Pathol. Bacteriol., 62: 389-401 (1950). 4. EsposiTo, J. J. ET AL. Serologic relatedness of mon- keypox, variola, and vaccinia viruses. J. med. Virol., 1: 35-47 (1977). 5. Esposrro, J. J. ET AL. The virion and soluble antigen proteins of variola, monkeypox, and vaccinia viruses. J. med. Virol., 1: 95-110 (1977). 6. FOSTER, S. 0. ET AL. Human monkeypox. Bull. World Health Organ., 46: 569-576 (1972). 7. GISPEN, R. Analysis of poxvirus antigens by means of double diffusion. A method for direct serological differentiation of cowpox. J. Immunol., 74: 134-141 (1955). 8. GIsPEN, R. & BRAND-SAATHOF, B." White "poxvirus strains from monkeys. Bull. World Health Organ., 46: 585-592 (1972). 9. GISPEN, R. & BRAND-SAATHOF, B. Three specific antigens produced in vaccinia, variola, and monkey- pox infections. J. inf. Dis., 129: 289-295 (1974). 10. GISPEN, R. ET AL. Monkeypox-specific antibodies in human and simian sera from Central Africa. Bull. World Health Organ., 53(4): 355-360 (1976). 11. HUTCHINSON, H. D. & ZIEGLER, D. W. Criteria for preparing, evaluating and standardizing iodinated globulins for radioimmunoassay procedures. Appl. Microbiol., 28: 935-942 (1974). 12. LEE, T.-K. ET AL. Comparison of rabies humoral antibody titers in rabbits and man by indirect radioimmunoassay, rapid fluorescent focus inhibition technique and indirect fluorescent antibody. J. clin. Microbiol., 5(3): 320-325 (1977). 13. LoURIE, B. ET AL. Human infection with monkeypox virus: Laboratory investigation of six cases in West Africa. Bull. WorldHealth Organ., 46: 633-639 (1972). 14. LowRy, 0. H. ET AL. Protein measurement with the Folin phenol reagent. J. biol. Chem., 193: 265-275 (1951). 15. MARENNIKOVA, S. S. ET AL. Isolation and-properties of the causal agent of a new variola-like disease (monkeypox) in man. Bull. World Health Organ., 46: 599-611 (1972). DIFFERENTIATION OF POXVIRUS ANTISERA 623 16. MARENNIKOVA, S. S. ET AL. Poxviruses isolated from clinically ill and asymptomatically infected monkeys and a chimpanzee. Bull. World Health Organ., 46: 613-620 (1972). 17. NAKANO, J. H. Comparative diagnosis of poxvirus diseases. In: Kurstak, E., ed., Comparative diagnosis of viral diseases, New York, Academic Press, 1977, Chapter 8. 18. NAKANO, J. H. & BINGHAM, P. G. Smallpox, vacci- nia, and human infections with monkeypox viruses. In: Lennette, E. H. et al., ed., Manual of clinical microbiology, 2nd ed. Washington, American Society for Microbiology, 1974, pp. 782-794. 19. NICoLI, J. & JOLIBOIS, C. Antigenes solubles des " poxvirus ". I. Etude des antigenes solubles vario- lique et vaccinal. Annales de l'Institut Pasteur (Paris), 107: 374-383 (1964). 20. PETERS, J. C. Eine " monkey-pox" enzootic im affenhaus des tiergartens " Blijdorp ". Die Kleintier- Praxis, 11(3): 65-70 (1966). 21. RONDLE, C. J. M. & DUMBELL, K. R. Antigens of cowpox virus. J. Hyg. (Lond.), 60: 41-49 (1962). 22. RONDLE, C. J. M. & SAYEED, K. A. R. Studies on monkeypox virus. Bull. WorldHealth Organ., 46: 577- 583 (1972). 23. SMADEL, J. E. ET AL. Nucleoprotein antigen of vaccinia virus. I. A new antigen for elementary bodies of vaccinia. Arch. Path., 34: 275-285 (1942). 24. Smallpox Surveillance Weekly Epidemiol. Rec. (WHO), 52(9): 85-88 (1977). 25. WILCOX, W. C. & COHEN, G. H. The poxvirus antigens. Curr. Topics Microbiol. Immunol., 47: 1-19 (1969). 26. WOODROOFE, G. M. & FENNER, F. Serological relationships within the poxvirus group: An antigen common to all members of the group. Virology, 16: 334-341 (1962). 27. ZIEGLER, D. W. ET AL. Detection by radioimmuno- assay of antibodies in human smallpox patients and vaccinees. J. clin. Microbiol., 1: 311-317 (1975).