albumin (BAPS). Virus infectivity titers were determined

Transcription

1 INFECTION AND IMMUNITY, Apr. 1979, p /79/4/71-6$2./ Vol. 24, No. 1 Delayed-Type Hypersensitivity Responses in Mice Infected with St. Louis Encephalitis Virus: Kinetics of the Response and Effects of Immunoregulatory Agents B. W. HUDSON,` K. WOLFF,' AND J. C. DEMARTINI2 Immunochemistry Branch, Vector-Borne Diseases Division, Center for Disease Control, Fort Collins, Colorado 8522,1 and Department of Pathology, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, Colorado Received for publication 3 January 1979 Labeled monocyte infiltration techniques have been used to study delayed-type hypersensitivity responses in mice immunied with St. Louis encephalitis virus. A delayed 24- to 48-h inflammatory response occurred 6 to 7 days after immuniation. This response can be potentiated by cyclophosphamide treatment, by BCG administration, or by splenectomy. Treatments known to selectivity inhibit T-cell function suppressed the response. The development of central nervous system lesions after St. Louis encephalitis (SLE) virus infection has long been associated with cellular infiltrates and inflammatory responses (18). Recent investigations have demonstrated that thymus-dependent cell (T-cell)-mediated responses play both protective and immunopathological roles in mice infected with a number of flaviviruses related to SLE (1, 9, 12, 15). In perhaps the best-studied model of virus disease in mice, lymphocytic choriomeningitis, the development of T-cell-mediated central nervous system disease is clearly associated with the development of peripheral footpad inflammatory responses (5). Peripheral delayed inflammatory responses in mice immunied against other viruses, however, are typically slight and difficult to evaluate quantitatively (4). In studies of cell-mediated response to viruses other than lymphocytic choriomeningitis and ectromelia (16), migration inhibition tests (12) or cytotoxic cell assays have usually been used (7). Direct T-cell-mediated cytotoxicity is relatively inefficient in controlling flavivirus replication in vitro (2, 7). It seems likely, therefore, that in vivo studies of the delayed inflammatory response might be useful in evaluating the immune response to flavivirus infection. We have investigated the use of labeled monocyte infiltration techniques (6, 1) for measurement of the delayed-type hypersensitivity (DTH) response in mice infected with SLE virus. MATERIALS AND METHODS Virus stock and control inoculum. SLE virus strain BeH was obtained from Robert Shope, Yale Arbovirus Research Unit, Yale University, New 71 Haven, Conn. Stock virus was prepared as a 1% (wt/ vol) suspension of infected suckling mouse brain in phosphate-buffered saline containing 5% bovine serum albumin (BAPS). Virus infectivity titers were determined by plaque formation on CER cells (13). The working stock virus preparation had a titer of 3.2 x 18 plaque-forming units (PFU) per ml. A suspension of normal suckling mouse brain (NSMB) was prepared by the same methods as those used for preparation of virus-infected suckling mouse brain. Mice. All mice used in this work were reared at our facility and were produced from stocks of National Institutes of Health general purpose albino mice. Mice were free of intercurrent infections with mouse pathogens with the exception of periodic problems caused by infection with Tyer's bacillus (Bacillus piliformis). All mice were weaned at 3 weeks of age, and the females were pooled, caged, and held for 4 to 6 weeks before being used in experiments. Mouse infection. Stock virus was diluted in M199 containing 5% fetal calf serum. Mice were infected by intraperitoneal or intravenous inoculation with SLE virus in.2 ml of diluent. Control mice received equivalent inocula of NSMB. DTH. DTH was assessed by measuring the inflammatory response after footpad challenge administered at intervals after infection. The 24- to 48-h dorsoventral increase in footpad thickness as measured by calibrated calipers was minimal (not exceeding.3 mm). Statistically valid measurements were obtained with experimental groups of 8 to 16 mice. These DTH measurements were supplemented by measurement of [3H]thymidine-labeled cellular infiltration adapted from techniques developed by Lefford (6) and Sabalovic et al. (1). At 24 h before footpad challenge, infected mice and control mice were inoculated intraperitoneally with 2 ici of [3H]thymidine. Each mouse was then challenged by inoculation of.5 ml of undiluted SLE-infected mouse brain suspension (1.6 x 17 PFU) into the left footpad. Each mouse was also inoculated in the right footpad with an equivalent

2 72 HUDSON, WOLFF, AND DEMARTINI dosage of NSMB. At selected intervals after footpad challenge (usually 24 h), the footpad response of each group of mice was assessed by measuring the differential swelling of the footpads. Immediately after physical measurement, the feet were removed at the first joint and solubilied in 1 ml of NCS tissue solubilier (Amersham/Searle Corp.), and the suspension was added to 1 ml of scintillation cocktail. After scintillation counting, data uncorrected for quenching were used for calculating thymidine incorporation. Differential counts (counts per minute [left foot] - counts per minute [right foot]) of 2, to 2, cpm of 3H were generally measurable in infected mice responding to footpad challenge. Such differential counts show a large statistical variance because of individual differences in thymidine incorporation. The ratios of 3H counted in the challenged and the control feet of individual mice were not subject to such variation. Footpad cellular infiltration ratios (CIR) were therefore computed for each individual mouse, the CIR being equal to the ratio of counts per minute measured in the left and right footpads. Tests for statistical significance were based on the Student's t test adapted for use with paired observations, a =.5 (14). Histological examinations of the inflammatory response in the feet of additional groups of infected and control mice were made 24 h after footpad challenge. Both feet were fixed and demineralied in Bouin solution, embedded in paraffin, sectioned at 6 Am, and stained with hematoxylin and eosin. Humoral antibody determinations. Solid-phase radioimmunoassay determinations of SLE virus-specific immunoglobulin M (IgM) and IgG were performed by techniques adapted from those of Trent et al. (17). The antigen used was a sucrose-acetone-extracted preparation of SLE-BeH infected suckling mouse brain which was obtained from Nick Karabatsos, World Health Organiation Arbovirus Reference Laboratory, Center for Disease Control, Fort Collins, Colo. Each well of a 96-well plastic plate was coated with 32 hemagglutinating units of the SLE antigen. Triplicate wells were then incubated with a 1: 1 dilution of each normal and test mouse serum. After the plates were washed, the amount of mouse IgM and IgG bound to SLE antigen was assessed by using '25I-labeled anti-mouse IgM and IgG (goat antimouse IgM and goat anti-mouse IgG, Bionetics Laboratory Products, Kensington, Md.). Results are expressed as the difference in 12"I uptake by wells treated with test serums and control serums. RESULTS Preliminary investigations were made to determine the level of delayed 24-h footpad swelling and cellular infiltration during the 12 days after infection. The footpad response was maximal at 6 days postinfection. At this time, the course of inflammation and cellular infiltration was studied. Additionally, the character of the cellular infiltrate present at 24 h after challenge was examined histologically. Finally, the effects of various immunoregulatory treatments on the INFECT. IMMUN. humoral antibody and peripheral delayed inflammatory responses were studied. Delayed 24-h footpad swelling and cellular infiltration during the 12 days after infection. Mice were infected by intraperitoneal inoculation with 1, PFU of SLE virus. Control mice received equivalent inoculations of NSMB suspension. The level of the delayed 24- h inflammatory response was measured at intervals after infection (p.i.). At each interval the footpads of groups of eight mice were challenged. Differential footpad swelling and cellular infiltration were assessed 24 h later. Figure 1 presents the results obtained. Differential footpad swelling reached the highest levels on day 6 p.i. and decreased rapidly to negligible levels by day 1. Labeled CIR reached statistically significant levels on days 6 to 8, after which they declined rapidly, reaching normal control levels by day 1 Ṫime courses of inflammation and labeled cell infiltration. The time courses of inflammation and cellular infiltration were measured in mice infected as in the previous section and challenged 6 days p.i. The responses of groups of five immunied and five control mice were measured at intervals up to 72 h after footpad challenge. Figure 2 shows the development of the responses. All mice challenged with mouse brain suspensions exhibited a prominent nonspecific inflammatory response within min- 3r E 2 E (2 -J -J w In 4. 4 utes of footpad inoculation. Residual footpad swelling from this response was still measurable 2 to 4 h after inoculation. The swelling continued to decrease in control mice and was essentially undetectable 24 and 48 h after footpad challenge. A moderate inflammatory response was detect- I,- -3r 1.2 Z 1.1 J~~~~~~~~~L l W fl \ <U DAY AFTER INFECTION FIG. 1. Twenty-four-hour differential footpad swelling and tritium-labeled cellular infiltration measured during the 12 days after intraperitoneal infection with 1, PFU of SLE virus. Mean values + 1 standard error, n = 8. Symbols:, infected mice;, control mice. I I

3 VOL. 24, 1979 DTH RESPONSE TO SLE VIRUS 73 2 I~~~~~~~~~~~~~B- E (I) LI~~~~~~~~~~a W - -I 21 o HOUR AFTER FOOTPAD CHALLENGE (Log) FIG. 2. Development of the footpad response during the 72 h after footpad challenge. (A) Differential footpad swelling. (B) Tritium-labeled CIR. Symbols:, mice infected with 1/KXJ PFU of SLE virus 6 days before footpad challenge;, control mice. Mean values ± standard error, n = 5. able in the SLE virus-infected mice. This response was highest 24 h after footpad challenge and, although significantly different from control values, did not exceed.25 mm. Labeled cellular infiltration in infected mice was not evident at 24 h after footpad challenge, was noticeable at 9 h after challenge, and was significantly different from that in control mice at 24 h (a <.1). Character of the cellular infiltrate. Histological examination of the footpads 24 h after challenge revealed a certain amount of inoculation trauma and accompanying polymorphonuclear cell infiltration in both feet of control animals. In the SLE virus-infected mice, however, cellular infiltration in the virus-challenged feet was three to four times that observed in the feet from control animals. The infiltrate contained a greater proportion of monocytes and small lymphocytes. In general, the cellular infiltrate in the virus-challenged feet of infected mice was composed of approximately 7% neutrophils, 2% macrophages, 1% small lymphocytes, and an occasional lymphoblast. Effects of immunoregulation on the peripheral 24-h delayed inflammatory response and associated humoral antibody response. The character of the cellular infiltrates found could be taken as evidence of antibody-mediated or DTH response in that neutrophils, lymphocytes, and monocytes were present (3). To evaluate the potential contributions of the cell-mediated and humoral immune responses to the footpad inflammatory response, we studied the effects of various treatments known to potentiate or suppress DTH. Peripheral DTH responses in mice can be enhanced by live BCG administration, by splenectomy, and by selective suppression of antibody production (8). Selective suppression of T- cell responses can be induced by administering phytohemagglutinin (PHA) followed by low doses of cyclophosphamide (11). An experiment was designed to evaluate the effects of various SLE infecting doses and immunoregulatory treatments on the development of 24-h cellular infiltration responses. At 18 days before SLE infection, a group of mice was inoculated intravenously with 17 live BCG (live M. bovis, Pasteur strain, TMC 111, obtained from the Trudeau Institute, Inc., Saranac Lake, N. Y.). A second group of mice was inoculated 1 days before SLE infection with intravenous PHA (Miles-Yeda Ltd., Miles Laboratories Inc., Elkhart, Ind.) (15 mg/kg) and 5 days before SLE with intravenous cyclophosphamide (Mead Johnson and Co., Evansville, Ind.) (4 mg/kg). A third group was inoculated intravenously with cyclophosphamide (2 mg/kg) 2 days before SLE infection. Two final groups were splenectomied and sham splenectomied 4 weeks before infection. At the time of SLE infection, all groups of mice were separated into five equal portions and inoculated intravenously with SLE (6.5 x 11, 6.5 x 13, 6.5 x 15, 6.5 x 17 PFU) and with a normal suckling mouse brain dilution equivalent to the highest SLE virus dose. Eight of the normal, PHA-cyclophosphamide-, BCG-, and cyclophosphamide-treated mice and five of the splenectomied and sham-splenectomied mice of each treatment group were challenged in the footpads at intervals for 1 days p.i. The CIR were measured 24 h later. No appreciable response was noted in any of the control groups inoculated with normal suckling mouse brain suspensions. All responding groups of SLE-infected mice, regardless of the infecting dose of SLE virus administered, yielded maximum or near maximum 24-h CIR at 5 to 6 days after infection. Normal mice, sham-splenectomied

4 74 HUDSON, WOLFF, AND DEMARTINI mice, mice treated with BCG, and mice given cyclophosphamide 2 days before SLE infection all showed a consistent SLE dose-related increase in cellular infiltration but were not significantly different from each other on day 6 p.i. Figure 3 shows the effects of SLE virus-infecting dose on the 24-h response to footpad challenge given on day 6 p.i. For simplicity of presentation, data for the normal SLE-infected mice have been excluded from the figure. Mice treated with PHA plus cyclophosphamide showed no evidence of significant labeled cell infiltration. A marked dose-dependent enhancement of the response was evident in splenectomied animals, with optimally enhanced responses being obtained with the dose of 6.5 x 15 PFU of SLE virus. Although neither BCG nor cyclophosphamide given before infection amplified the response at 6 days p.i., both treatments prolonged the response. This is shown in Fig. 4 which presents the development of the footpad cellular infiltration response from days 3 to 1 p.i. with 6.5 x 1i PFU of virus. It is apparent from these data that splenectomy, BCG treatment, and cyclophosphamide treatment either enhance or prolong the 24-h footpad response. Treatment with ct I- 1= -4 cx: -j w I FIG. 3. Effects of immunoregulatory treatments on group was used as a control and was inoculated the footpad response. Dose-response curves showing with equivalent dilutions of NSMB suspensions. the effect ct of intravenous SLE virus-infecting dose on Virus-infected mice were bled in groups of four 24-h labweled cellular infiltration. CIR were measured by footpad challenge on day 6 p.i. Mean values + on days 4, 6, 8, and 12. Diluent-moculated mice standardrd error. Symbols: A, splenectomied mice;, were bled on days, 6, and 12. Sera were col- (2 mg/kg) 2 days before infec- elected and SLE-specific igm and IgG antibody cyclophtosphamide tion;, live BCG 18 days before infection; A, phyto- content was measured with the solid-phase ra- (ca. 15 mg/kg) 1 days and cyclophos- dioimmunoassay technique. Figure 5 presents hemaggdlutinin phamid{e (4 mg/kg) 5 days before infection. the radioimmunoassay data obtained. Control 2D c:.-- -J -i -J INFECT. IMMUN. I DAY AFTER INFECTION FIG. 4. Development of 24-h footpad responses during the 1 days after intravenous infection ofmice with 6.5 x 15 PFU of SLE virus. The effect of immunoregulatory treatments on mean 24-h labeled CIR. Symbols: A, splenectomied mice; *, cyclophosphamide (2 mg/kg) 2 days before infection; U, live BCG 18 days before SLE infection;, phytohemagglutinin (ca. 15 mg/kg) 1 days and cyclophosphamide (4 mg/kg) 5 days before infection; L---L, normal untreated mice infected with SLE. PHA plus cyclophosphamide caused almost complete suppression of the 24-h response. The data presented in Fig. 3 and 4 suggest that the 24-h labeled cellular infiltration may be T-cell mediated. Thus, the treatments known to potentiate DTH also potentiate or prolong the 24-h cellular infiltration responses, and the treatment known to suppress T-cell responses causes virtual elimination of 24-h cellular infiltration. A remaining reservation was caused by the lack of available knowledge pertaining to the effects of the immunoregulatory treatments on antibody responses to SLE virus. These effects were therefore measured. Six groups of mice were prepared. These consisted of normal mice; mice pretreated with BCG, PHA plus cyclophosphamide, or cyclophosphamide alone; and splenec- 6.5X1' 6.5X13 6.5X15 6.5XlO' tomied and sham-splenectomied mice. Each DOSE (PFU) group was infected by intravenous inoculation of INFECTIVE, 6.5 x 14 PFU of SLE virus. A portion of each

5 VOL. 24, 1979 o x ~ 11I - IgM DAY AFTER INFECTION FIG. 5. Effect ofimmunosuppressive treatments on SLE-specific IgM and IgG responses of mice infected with SLE viruses. Solid-phase radioimmunoassay data are expressed as mean counts per minute of "2I above background, n = 4. Symbols:, PHA (ca. 15 mg/kg) 1 days and cyclophosphamide (4 mg/kg) 5 days before infection;, BCG (17) intravenously 18 days before infection;, cyclophosphamide (2 mg/kg) 2 days before infection; A, splenectomied mice; --EJ normal control mice infected with SLE virus. sera from each group were used as base line measures for computation of 125I uptake. Mean solid-phase radioimmunoassay data for normal and sham-splenectomied mice did not differ significantly (Student's t test) and have been pooled. These data indicate that the cellular infiltration measured by footpad challenge on day 6 was probably not antibody mediated. The PHAcyclophosphamide treatment, which suppressed the footpad response, left the antibody response essentially normal. Splenectomy, which enhanced the footpad response, resulted in suppression of humoral antibody 6 days after immuniation. In general we might postulate three possible relationships between antibody responses and CIR. The first, assuming that antibody contributes to the CIR, would lead one to expect a direct association between CIR and antibody levels on day 6 of the response. The second, assuming no relationship, would lead one to expect no association. The third, assuming that antibody responses may suppress the CIR, would lead to expectation of an inverse relationship. By using rank correlation techniques for 6- day CIR and IgG antibody levels in the five treatment groups, the latter hypothesis is barely tenable at the 5% significance level. DISCUSSION At 5 to 8 days p.i. with live SLE virus, mice exhibited a moderate 24-h delayed inflammatory response to footpad challenge. The cells infiltrating the challenged footpad were neutrophils, monocytes, and lymphocytes. Measurement of X~~~~~~~ DTH RESPONSE TO SLE VIRUS 75 the kinetics of the cellular infiltrate by tritiumlabeled cell techniques, originally developed to IgG - quantitate monocyte infiltration (6, 1), revealed a time course characteristic of delayed responses. Immunomodifications such as splenectomy, cyclophosphamide treatment, and BCG administration, which are known to potentiate T-cell-mediated peripheral DTH responses (16), either potentiated or prolonged the delayed 24- h response in mice. Treatment with PHA plus cyclophosphamide, a procedure developed to suppress splenic T-helper-cell responses (11), suppressed the footpad inflammatory response. The degree of inflammation appeared to be independent of the antibody response. PHA-cyclophosphamide treatment, which suppressed the footpad response, did not alter the antibody response. Either splenectomy or cyclophosphamide treatment given 2 days before SLE infection enhanced the footpad responses but reduced the level of antibody produced at 6 to 8 days. The footpad response after SLE infection, therefore, meets most of the criteria of the typical transitory peripheral DTH responses shown to occur after stimulation of mice with nonreplicating antigens (3, 8). The T-cell-mediated response, assessed by cytotoxicity assays or by splenocyte migration indexes, has been shown to play both protective and immunopathological roles in mice infected with a number of related flaviviruses (1, 4, 5, 12). Differentiation of the relative contributions of cell-mediated and humoral factors to immunity has been difficult, however, since the assessment of protection has usually depended on intracerebral inoculation of the virus into immunied animals which exhibit both cellular and humoral immunity. Since both peripheral delayed inflammatory responses and humoral antibody levels can be easily manipulated by established immunoregulatory techniques, it should be possible to assess the differential role of humoral and cell-mediated immunity in the development of resistance to SLE virus infection. LITERATURE CITED 1. Bhatt, P. M., and R.. Jacoby Genetic resistance to lethal flavivirus encephalitis. II. Effect of immunosuppression. J. Infect. Dis. 134: Catanaro, P. J., W. E. Brandt, N. R. Hogrefe, and P. K. Russell Detection of dengue cell-surface antigens by peroxidase-labeled antibodies and immune cytolysis. Infect. Immun. 1: Crowle, A. J Delayed hypersensitivity in the mouse. Adv. Immunol. 2: Feinstone, S. M., E. H. Beachey, and M. W. Rytel Induction of delayed hypersensitivity to influena and mumps virus in mice. J. Immunol. 13: Hotchin, J., and L. Benson The pathogenesis of lymphocytic choriomeningitis in mice: the effects of

6 76 HUDSON, WOLFF, AND DEMARTINI different inoculation routes and the footpad response. J. Immunol. 91: Lefford, M. J The measurement of tuberculin hypersensitivity in rats. Int. Arch. Allergy 47: McFarland, H. F In vitro studies of cell-mediated immunity in an acute viral infection. J. Immunol. 113: Mackaness, G. B., P. H. Lagrange, T. E. Miller, and T. Ishibashi The formation of active T-cells. Excerpts Med. Int. Congr. Ser. 325: Rook, G. A. W., and H. E. Webb Antilymphocyte serum and tissue culture used to investigate role of cellmediated response in viral encephalitis in mice. Br. Med. J. 4: Sabolovic, D., M. C. Beugnot, F. Dumont, and M. Bujadoux A new method to measure the specific cellular component of a delayed hypersensitivity response in the ear of the mouse. Eur. J. Immunol. 2: Schware, G Drug-induced immunological unresponsiveness: selective inhibition of T-cell 'helper function' by cyclophosphamide in mice pretreated with phytohaemagglutinin. Clin. Exp. Immunol. 27: Semenov, B. F., V. V. Khoinski, and V. V. Vargin. INFECT. IMMUN The damaging effect of cellular immunity in flavivirus infection of mice. Med. Biol. 53: Smith, A. L., G. H. Tignor, K. Mifune, and T. Motohashi Isolation and assay of rabies serogroup viruses in CER cells. Intervirology 89: Stearman, R. L Statistical concepts in microbiology. Bacteriol. Rev. 19: Thind, I. J., and W. H. Price Recovery from primary infection with Langat E5 virus in normal and cyclophosphamide-treated mice: relative roles of virus multiplication, interferon, antibody and serum protective factor. Am. J. Epidemiol. 89: Tosolini, F. A., and C. A. Mims Effect of murine strain and viral strain on the pathogenesis of lymphocyte choriomeningitis infection and a study of footpad responses. J. Infect. Dis. 123: Trent, D. W., C. L. Harvey, A. Qureshi, and D. Le- Stourgeon Solid-phase radioimmunoassay for antibodies to flavivirus structural and nonstructural proteins. Infect. Immun. 13: Webster, L. T., and A. D. Clow The limited neurotropic character of the encephalitis virus (St. Louis type) in susceptible mice. J. Exp. Med. 63: Downloaded from on August 23, 218 by guest