THE JOURNAL OF INFECTIOUS DISEASES. VOL. 140, NO.4. OCTOBER 1979 © by The University of Chicago. 0022-1899/79/4004-0012$00.75
In Vivo Enhancement of Dengue Virus Infection in Rhesus Monkeys by Passively Transferred Antibody Scott B. Halstead
From the Department of Tropical Medicine and Medical Microbiology, John A. Burns School of Medicine, University of Hawaii, Honolulu, Hawaii
tibody at concentrations above the neutralization threshold [4, 5]. The mononuclear phagocyte has been identified as the cell that supports D2V replication. Infection results after immune complexes are internalized, a step that requires the presence of Fe receptors [5, 6]. Before this report, the enhancement of dengue infection in vivo by use of passively transferred antibody has not been described. Recently, our group reported that dengue infection-enhancing antibody at high titers can be found in the cord-blood sera from infants who are born to dengue-immune Southeast Asian women [7]. The present report describes an experiment in which a pool of cord-blood sera from these dengue-immune subjects was inoculated into susceptible rhesus monkeys. Control animals received nonimmune human cord-blood serum. Subsequently, both groups of animals were infected with D2V, and the ensuing viremia was quantitated. Reproducible enhancement of dengue viremia was detected in monkeys given small quantities of antibody to dengue virus. The implications of these findings for the pathogenetic mechanisms of human dengue hemorrhagic fever/ dengue shock syndrome (DHF/DSS) are also discussed.
In a previous study, 118 rhesus monkeys were inoculated sequentially with each of the four types of dengue virus. Titers of circulating virus were found to vary paradoxically with immune status [1]. In monkeys that were monotypically immune to type 1, 3, or 4 virus and were then challenged with dengue type 2 virus (D2V), virus circulated at significantly higher levels than in nonimmune animals that were infected with D2V by the same route and with the same dose. The mechanism of this immunological enhancement of dengue virus infection is not established, but it might be related to a similar in vitro phenomenon. When dengue viruses were added to cultures of peripheral-blood mononuclear leukocytes (PBL) obtained from dengue-immune monkeys, virus replicated to titers of 10-3-10-5 per million cells over the next two to six days of culture [2, 3]. Similar cultures prepared from nonimmune animals did not support growth of dengue virus at this level. Enhanced dengue virus infection in PBL cultures also can be produced by infection of cells and addition of anReceived for publication November 13, 1978, and in revised form March 26, 1979. This work was supported by grant no. ROI HD08693 from the National Institutes of Health. I thank Dr. Ralph Hale, Kapiolani Hospital, Honolulu, Hawaii, for providing cord-blood samples, and May Tom, Susan Hatch, Kay Larsen, and Carrie Uyehara for technical assistance. Please address requests for reprints to Dr. Scott B. Halstead, Department of Tropical Medicine and Medical Microbiology, 3675 Kilauea Avenue, Honolulu, Hawaii 96816.
Materials and Methods
Virus. D2V, strain no. 16681, isolated in 1964 from a Thai child with dengue shock syndrome
527
Downloaded from http://jid.oxfordjournals.org/ at Boston University on November 11, 2014
Five pairs of juvenile, dengue virus-susceptible rhesus monkeys were given normal or dengue-immune human cord-blood serum injected intravenously to a final dilution of 1:300. The pool of immune human cord-blood serum had a titer of antibody to dengue type 2 virus (D2V) of 1:140 in the plaque-reduction neutralization test and a titer of human monocyte infection enhancement of >1:2,000,000. Fifteen minutes after inoculation of serum, animals were infected with D2V (strain no. 16681). Daily titers of viremia were always higher in the animals that had received antiserum to D2V than in animals that had received normal cord-blood serum. Ratios of infection enhancement ranged from 2.7 to 51.4. The demonstration of antibody dependence of dengue virus infection in subhuman primates-a complex, outbred experimental host-supports the hypothesis that the severity of dengue in humans is regulated by antibody.
Halstead
528
containing antibody to dengue virus and the growth in infected cultures to which normal serum had been added was expressed as fold-enhancement. Usually, replication data for days 2 and 3 were grouped. Infection of monkeys. Juvenile rhesus monkeys born in the United States were weighed and bled. An estimation of 36.4 ml of plasma/kg of body weight was used to calculate plasma volumes [13]. Animals were divided into three groups of two and one group of four animals. One animal in each pair received either undiluted pooled dengueimmune cord-blood serum or nonimmune cordblood serum, which was injected into the saphenous vein to effect a final dilution in plasma of 1:300. In some instances, cord-blood antibody was administered in divided doses. Five minutes after the inoculation of antibody, a blood sample was removed from the femoral vein for assay of infection-enhancing activity. At 10-15 min after inoculation of cord-blood serum, animals were injected sc with rv 1,000 pfu of D2V into the right forearm. A portion of the virus inoculum was frozen at - 70 C for assay. Monkeys were coded. Daily bleedings (from days 1 to 12) and assays for viremia were performed under code. Virus assay. Heparinized monkey plasmas were thawed from -70 C and diluted to 1:10 (in some instances, 1:100 or 1:1,000); 0.2 ml of each dilution was inoculated into three l-oz prescription bottles that contained LLC-MK2 monolayers. A single agar overlay was applied [10]. Mean and standard deviations of mean plaque counts were calculated. Viremia titers in paired experiments were analyzed by the statistical sign test. Results
Pooled cord-blood serum from seven dengueimmune human donors had an HAl titer of 1:80 to D2V. Titers of neutralizing antibody to the four dengue types were measured for six individual sera and for a pool that included a seventh sample. The entire seventh sample was added to the pool. As shown in table 1, five of six sera had significant titers of neutralizing antibody to two or more viruses. K-21 serum predominantly neutralized D2V. On four occasions the pool of cord-blood serum was assayed for enhancement of dengue infection in human PBL cultures. The fold-enhancement of
Downloaded from http://jid.oxfordjournals.org/ at Boston University on November 11, 2014
(DSS), was used in all studies [8]. This virus has received only limited passages in monkey kidney cells. Antibody donors. All cord-blood samples were from infants born at the Kapiolani Hospital, Honoulu, Hawaii. Mothers of these infants were dengue-immune women who were recent immigrants to Hawaii from Southeast Asia (Thailand, Vietnam, Laos, and the Philippine Islands). Cordblood samples obtained from four nonimmune Japanese, Chinese, and Korean infants were used as controls. Antibody measurement. HAl test. Cordblood sera were treated with kaolin and tested for antibodies with microtiter equipment (Cooke Engineering, Alexandria, Va.). Mouse brainadapted prototype dengue 1 (Hawaii), dengue 2 (TR 1752), dengue 3 (H-87), and dengue 4 (H-241) antigens were prepared by the sucroseacetone method [9]. Four antigen units were used. Antibody measurement. Plaque-reduction neutralization test. Neutralizing antibodies were measured in our standard assay that uses continuous rhesus kidney cells (LLC-MK2) [10] and the 50070 plaque-reduction end point [11]. Sera were tested against the following tissue culture-adapted Southeast Asian strains: dengue 1 (no. 16007), dengue 2 (no. 16681), dengue 3 (no. 16562), and dengue 4 (no. 4328-S). Infection enhancement in human monocytes. Mononuclear leukocytes from the blood of nonimmune donors were prepared by differential sedimentation on Ficoll-Hypaque [12]. PBL were washed three times and suspended in RPMI-1640 medium (Grand Island Biological Co., Grand Island, N.Y.) supplemented with 10070 heatinactivated fetal calf serum, glutamine, 20 mx HEPES (N-2-hydroxyethylpiperazine-N '-2-ethanesulfonic acid), 0.02% sodium bicarbonate, and antibiotics (200 IJg of streptomycin and 200 units of penicillin/ml). Cell suspensions were standardized to a concentration of 106 PBL/ml. In all experiments D2V was added to PBL cultures at a multiplicity of infection of 0.1. Then, dilutions of test antisera, normal serum, and a positive control serum were added to enough cells to permit studies of viral growth over a four-day period. Screwcapped glass vials containing 106 cells were incubated at 37 C for one to four days. Usually, one vial was harvested each day for virus assay. The ratio between the mean titers of growth in cultures
Antibody-Enhanced Dengue Infection
529
Table 1. Reciprocal titers of neutralizing antibody to four types of dengue virus in cord-blood sera from infants born to dengue-immune Southeast Asian women, as measured by the plaque-reduction neutralization test. Dengue virus type Serum no. 20 53 100 115 135 150 90
3
4
98 110 195 240 540 135 140
28 210 110 265 135 600 140
10 22 75 165 40 44 27
Experiment no. 1 2 3 4
Fold-enhancement of D2V titer at indicated dilution of antibody 10-2
10-3
10-4
10-1
10-6
10-7
4.8
7.9 8.2 37.7
8.2 12.6 9.6 6.3
6.5 2.0 4.8
4.5 1.3 1.9
1.8
7.1
NOTE. For these experiments, a 50070 plaque-reduction end point was used. • The pool was constituted with different amounts of the six sera and all of the serum in a seventh sample.
NOTE. The fold-enhancement is expressed as a ratio between the virus growth (in pfu) on days 2 and 3 in PBL cultures supplemented with antibody and the virus growth in cultures supplemented with nonimmune cord-blood serum.
dengue virus growth on days 2 and 3 is shown in table 2 and is expressed as a ratio between the virus growth in PBL cultures supplemented with serial 10-fold dilutions of antibody to D2V and the virus growth in companion cultures that had received similar dilutions of nonimmune cordblood serum. Between a dilution range of 1:100 and 1:100,000, the fold-enhancement of virus yield varied between 5 and 12. In one experiment,
a 1:1,000 dilution of immune serum produced a 38-fold enhancement of infection. Following the iv injection of immune or nonimmune cord-blood serum, 10 rhesus monkeys were infected with D2V in four separate experiments. The sex and weight of the monkeys and the size of inocula of D2V and cord-blood sera are summarized in table 3. In experiments no. 1 and 4, cord-blood sera were injected incrementally to
Table 3. Sex and weight of animals, volume of human cord-blood serum, and dose of virus used in experiments determining the ability of passively transferred antibody to enhance dengue 2 virus (D2V) infection in rhesus monkeys. Characteristics of monkeys Experiment no., monkey no.
Group"
Sex
Weight (kg)
Volume of serum
Dose of D2V (pfu)
It H-251 H-252
Antibody Control
M F
2.74 2.43
0.33 0.3
9,200 9,200
H-253 H-254
Antibody Control
F F
2.75 2.90
0.33 0.34
5,200 5,200
H-255 H-256
Antibody Control
M F
2.75 2.39
0.33 0.27
1,700 1,700
W-38 W-92 W-39 W-98
Antibody Control Antibody Control
M M M M
2.17 2.57 2.45 2.87
0.26 0.31 0.3 0.34
1,000 1,000 1,000 1,000
2
3
4t
• The monkeys labeled "antibody" received iv injection of dengue-immune human cord-blood serum, and monkeys labeled "control" received iv injection of nonimmune human cord-blood serum, before sc inoculation of D2V. t Cord-blood serum (0.03, 0.1, and 0.2 ml) inoculated at 5-min intervals. t Immune and nonimmune cord-blood sera inoculated at cumulative final dilutions of 1:3,000, 1:600, and 1:300 at 15-min intervals.
Downloaded from http://jid.oxfordjournals.org/ at Boston University on November 11, 2014
K-21 K-22 K-28 K-30 K-34 K-35 Pool"
2
Table 2. Assay of enhancement titer of pooled dengue-immune human cord-blood serum: foldenhancement of dengue 2 virus (D2V) growth in cultures of human peripheral-blood leukocytes (PBL), supplemented with serial lO-fold dilutions of cord-blood serum.
Halstead
530
Table 4. Assay of enhancement titer of pooled dengueimmune human cord-blood serum after incremental iv inoculation of pooled serum into monkey no. H-251: fold-enhancement of dengue 2 virus (D2V) growth by dilutions of monkey plasma.
Time of sample (min) 15 30
45
Fold-enhancement of D2V titer at indicated dilution of serum 10- 4
10- 5
10-6
10- 7
14.5 6.3 6.3
18.8 14.1 13.0
9.4 6.3 4.1
In Vivo Enhancement of Dengue Virus Infection in Rhesus Monkeys by Passively Transferred Antibody Scott B. Halstead
From the Department of Tropical Medicine and Medical Microbiology, John A. Burns School of Medicine, University of Hawaii, Honolulu, Hawaii
tibody at concentrations above the neutralization threshold [4, 5]. The mononuclear phagocyte has been identified as the cell that supports D2V replication. Infection results after immune complexes are internalized, a step that requires the presence of Fe receptors [5, 6]. Before this report, the enhancement of dengue infection in vivo by use of passively transferred antibody has not been described. Recently, our group reported that dengue infection-enhancing antibody at high titers can be found in the cord-blood sera from infants who are born to dengue-immune Southeast Asian women [7]. The present report describes an experiment in which a pool of cord-blood sera from these dengue-immune subjects was inoculated into susceptible rhesus monkeys. Control animals received nonimmune human cord-blood serum. Subsequently, both groups of animals were infected with D2V, and the ensuing viremia was quantitated. Reproducible enhancement of dengue viremia was detected in monkeys given small quantities of antibody to dengue virus. The implications of these findings for the pathogenetic mechanisms of human dengue hemorrhagic fever/ dengue shock syndrome (DHF/DSS) are also discussed.
In a previous study, 118 rhesus monkeys were inoculated sequentially with each of the four types of dengue virus. Titers of circulating virus were found to vary paradoxically with immune status [1]. In monkeys that were monotypically immune to type 1, 3, or 4 virus and were then challenged with dengue type 2 virus (D2V), virus circulated at significantly higher levels than in nonimmune animals that were infected with D2V by the same route and with the same dose. The mechanism of this immunological enhancement of dengue virus infection is not established, but it might be related to a similar in vitro phenomenon. When dengue viruses were added to cultures of peripheral-blood mononuclear leukocytes (PBL) obtained from dengue-immune monkeys, virus replicated to titers of 10-3-10-5 per million cells over the next two to six days of culture [2, 3]. Similar cultures prepared from nonimmune animals did not support growth of dengue virus at this level. Enhanced dengue virus infection in PBL cultures also can be produced by infection of cells and addition of anReceived for publication November 13, 1978, and in revised form March 26, 1979. This work was supported by grant no. ROI HD08693 from the National Institutes of Health. I thank Dr. Ralph Hale, Kapiolani Hospital, Honolulu, Hawaii, for providing cord-blood samples, and May Tom, Susan Hatch, Kay Larsen, and Carrie Uyehara for technical assistance. Please address requests for reprints to Dr. Scott B. Halstead, Department of Tropical Medicine and Medical Microbiology, 3675 Kilauea Avenue, Honolulu, Hawaii 96816.
Materials and Methods
Virus. D2V, strain no. 16681, isolated in 1964 from a Thai child with dengue shock syndrome
527
Downloaded from http://jid.oxfordjournals.org/ at Boston University on November 11, 2014
Five pairs of juvenile, dengue virus-susceptible rhesus monkeys were given normal or dengue-immune human cord-blood serum injected intravenously to a final dilution of 1:300. The pool of immune human cord-blood serum had a titer of antibody to dengue type 2 virus (D2V) of 1:140 in the plaque-reduction neutralization test and a titer of human monocyte infection enhancement of >1:2,000,000. Fifteen minutes after inoculation of serum, animals were infected with D2V (strain no. 16681). Daily titers of viremia were always higher in the animals that had received antiserum to D2V than in animals that had received normal cord-blood serum. Ratios of infection enhancement ranged from 2.7 to 51.4. The demonstration of antibody dependence of dengue virus infection in subhuman primates-a complex, outbred experimental host-supports the hypothesis that the severity of dengue in humans is regulated by antibody.
Halstead
528
containing antibody to dengue virus and the growth in infected cultures to which normal serum had been added was expressed as fold-enhancement. Usually, replication data for days 2 and 3 were grouped. Infection of monkeys. Juvenile rhesus monkeys born in the United States were weighed and bled. An estimation of 36.4 ml of plasma/kg of body weight was used to calculate plasma volumes [13]. Animals were divided into three groups of two and one group of four animals. One animal in each pair received either undiluted pooled dengueimmune cord-blood serum or nonimmune cordblood serum, which was injected into the saphenous vein to effect a final dilution in plasma of 1:300. In some instances, cord-blood antibody was administered in divided doses. Five minutes after the inoculation of antibody, a blood sample was removed from the femoral vein for assay of infection-enhancing activity. At 10-15 min after inoculation of cord-blood serum, animals were injected sc with rv 1,000 pfu of D2V into the right forearm. A portion of the virus inoculum was frozen at - 70 C for assay. Monkeys were coded. Daily bleedings (from days 1 to 12) and assays for viremia were performed under code. Virus assay. Heparinized monkey plasmas were thawed from -70 C and diluted to 1:10 (in some instances, 1:100 or 1:1,000); 0.2 ml of each dilution was inoculated into three l-oz prescription bottles that contained LLC-MK2 monolayers. A single agar overlay was applied [10]. Mean and standard deviations of mean plaque counts were calculated. Viremia titers in paired experiments were analyzed by the statistical sign test. Results
Pooled cord-blood serum from seven dengueimmune human donors had an HAl titer of 1:80 to D2V. Titers of neutralizing antibody to the four dengue types were measured for six individual sera and for a pool that included a seventh sample. The entire seventh sample was added to the pool. As shown in table 1, five of six sera had significant titers of neutralizing antibody to two or more viruses. K-21 serum predominantly neutralized D2V. On four occasions the pool of cord-blood serum was assayed for enhancement of dengue infection in human PBL cultures. The fold-enhancement of
Downloaded from http://jid.oxfordjournals.org/ at Boston University on November 11, 2014
(DSS), was used in all studies [8]. This virus has received only limited passages in monkey kidney cells. Antibody donors. All cord-blood samples were from infants born at the Kapiolani Hospital, Honoulu, Hawaii. Mothers of these infants were dengue-immune women who were recent immigrants to Hawaii from Southeast Asia (Thailand, Vietnam, Laos, and the Philippine Islands). Cordblood samples obtained from four nonimmune Japanese, Chinese, and Korean infants were used as controls. Antibody measurement. HAl test. Cordblood sera were treated with kaolin and tested for antibodies with microtiter equipment (Cooke Engineering, Alexandria, Va.). Mouse brainadapted prototype dengue 1 (Hawaii), dengue 2 (TR 1752), dengue 3 (H-87), and dengue 4 (H-241) antigens were prepared by the sucroseacetone method [9]. Four antigen units were used. Antibody measurement. Plaque-reduction neutralization test. Neutralizing antibodies were measured in our standard assay that uses continuous rhesus kidney cells (LLC-MK2) [10] and the 50070 plaque-reduction end point [11]. Sera were tested against the following tissue culture-adapted Southeast Asian strains: dengue 1 (no. 16007), dengue 2 (no. 16681), dengue 3 (no. 16562), and dengue 4 (no. 4328-S). Infection enhancement in human monocytes. Mononuclear leukocytes from the blood of nonimmune donors were prepared by differential sedimentation on Ficoll-Hypaque [12]. PBL were washed three times and suspended in RPMI-1640 medium (Grand Island Biological Co., Grand Island, N.Y.) supplemented with 10070 heatinactivated fetal calf serum, glutamine, 20 mx HEPES (N-2-hydroxyethylpiperazine-N '-2-ethanesulfonic acid), 0.02% sodium bicarbonate, and antibiotics (200 IJg of streptomycin and 200 units of penicillin/ml). Cell suspensions were standardized to a concentration of 106 PBL/ml. In all experiments D2V was added to PBL cultures at a multiplicity of infection of 0.1. Then, dilutions of test antisera, normal serum, and a positive control serum were added to enough cells to permit studies of viral growth over a four-day period. Screwcapped glass vials containing 106 cells were incubated at 37 C for one to four days. Usually, one vial was harvested each day for virus assay. The ratio between the mean titers of growth in cultures
Antibody-Enhanced Dengue Infection
529
Table 1. Reciprocal titers of neutralizing antibody to four types of dengue virus in cord-blood sera from infants born to dengue-immune Southeast Asian women, as measured by the plaque-reduction neutralization test. Dengue virus type Serum no. 20 53 100 115 135 150 90
3
4
98 110 195 240 540 135 140
28 210 110 265 135 600 140
10 22 75 165 40 44 27
Experiment no. 1 2 3 4
Fold-enhancement of D2V titer at indicated dilution of antibody 10-2
10-3
10-4
10-1
10-6
10-7
4.8
7.9 8.2 37.7
8.2 12.6 9.6 6.3
6.5 2.0 4.8
4.5 1.3 1.9
1.8
7.1
NOTE. For these experiments, a 50070 plaque-reduction end point was used. • The pool was constituted with different amounts of the six sera and all of the serum in a seventh sample.
NOTE. The fold-enhancement is expressed as a ratio between the virus growth (in pfu) on days 2 and 3 in PBL cultures supplemented with antibody and the virus growth in cultures supplemented with nonimmune cord-blood serum.
dengue virus growth on days 2 and 3 is shown in table 2 and is expressed as a ratio between the virus growth in PBL cultures supplemented with serial 10-fold dilutions of antibody to D2V and the virus growth in companion cultures that had received similar dilutions of nonimmune cordblood serum. Between a dilution range of 1:100 and 1:100,000, the fold-enhancement of virus yield varied between 5 and 12. In one experiment,
a 1:1,000 dilution of immune serum produced a 38-fold enhancement of infection. Following the iv injection of immune or nonimmune cord-blood serum, 10 rhesus monkeys were infected with D2V in four separate experiments. The sex and weight of the monkeys and the size of inocula of D2V and cord-blood sera are summarized in table 3. In experiments no. 1 and 4, cord-blood sera were injected incrementally to
Table 3. Sex and weight of animals, volume of human cord-blood serum, and dose of virus used in experiments determining the ability of passively transferred antibody to enhance dengue 2 virus (D2V) infection in rhesus monkeys. Characteristics of monkeys Experiment no., monkey no.
Group"
Sex
Weight (kg)
Volume of serum
Dose of D2V (pfu)
It H-251 H-252
Antibody Control
M F
2.74 2.43
0.33 0.3
9,200 9,200
H-253 H-254
Antibody Control
F F
2.75 2.90
0.33 0.34
5,200 5,200
H-255 H-256
Antibody Control
M F
2.75 2.39
0.33 0.27
1,700 1,700
W-38 W-92 W-39 W-98
Antibody Control Antibody Control
M M M M
2.17 2.57 2.45 2.87
0.26 0.31 0.3 0.34
1,000 1,000 1,000 1,000
2
3
4t
• The monkeys labeled "antibody" received iv injection of dengue-immune human cord-blood serum, and monkeys labeled "control" received iv injection of nonimmune human cord-blood serum, before sc inoculation of D2V. t Cord-blood serum (0.03, 0.1, and 0.2 ml) inoculated at 5-min intervals. t Immune and nonimmune cord-blood sera inoculated at cumulative final dilutions of 1:3,000, 1:600, and 1:300 at 15-min intervals.
Downloaded from http://jid.oxfordjournals.org/ at Boston University on November 11, 2014
K-21 K-22 K-28 K-30 K-34 K-35 Pool"
2
Table 2. Assay of enhancement titer of pooled dengue-immune human cord-blood serum: foldenhancement of dengue 2 virus (D2V) growth in cultures of human peripheral-blood leukocytes (PBL), supplemented with serial lO-fold dilutions of cord-blood serum.
Halstead
530
Table 4. Assay of enhancement titer of pooled dengueimmune human cord-blood serum after incremental iv inoculation of pooled serum into monkey no. H-251: fold-enhancement of dengue 2 virus (D2V) growth by dilutions of monkey plasma.
Time of sample (min) 15 30
45
Fold-enhancement of D2V titer at indicated dilution of serum 10- 4
10- 5
10-6
10- 7
14.5 6.3 6.3
18.8 14.1 13.0
9.4 6.3 4.1
