The development of an improved vaccine for controlling measles virus (MV) infections in the developing world will require an understanding of the immune mechanisms responsible for the clearance of this virus. To evaluate the role of humoral immunity in the containment of MV, rhesus monkeys were treated at the time of MV challenge with either anti-CD20 monoclonal antibody (MAb) infusion, to deplete B lymphocytes, or both anti-CD20 and anti-CD8 MAb, to deplete both B lymphocytes and CD8+ effector T lymphocytes. Although the MV-specific antibody response in CD20+ lymphocyte-depleted monkeys was delayed by >1 week, the kinetics of MV clearance did not differ from those for monkeys that received control MAb. Furthermore, unusual clinical sequelae of MV infection were not observed in these monkeys. In contrast, MV-infected rhesus monkeys depleted of both CD20+ and CD8+ lymphocytes had a prolonged duration of viremia and developed a desquamating skin rash. These findings indicate that humoral immunity plays a limited role in the control of MV replication in an MV-naive individual and suggest that new measles vaccination strategies should focus on the elicitation of cell-mediated immune responses, in addition to neutralizing antibodies, to facilitate rapid elimination of locally replicating virus.
Despite the availability of an effective live attenuated measles virus (MV) vaccine, measles remains a significant cause of infant morbidity and mortality in developing nations. The persistence of MV-induced disease in pediatric populations reflects some limitations of the current vaccine. The immunogenicity of the live attenuated vaccine in young children is limited by the immaturity of their immune systems and by the presence of circulating maternal antibodies . The ultimate elimination of MV from human populations will therefore likely require an improved vaccine. The development of new vaccination strategies for preventing MV infection can best be guided by an understanding of the immune mechanisms responsible for control and clearance of MV.
Cellular immunity has been implicated in the control of MV replication. Children with cellular immune deficiencies have particularly severe disease associated with MV infection [2–4], and the rate of recovery from MV infection in healthy individuals has been correlated with the strength of their cellular immune response . In fact, we have recently shown that CD8+ T lymphocytes play a major role in control of MV replication and clearance of viremia in rhesus monkeys .
Although antibody provides protection against MV infection in previously vaccinated individuals and newborn infants, its role in recovery from an established infection is not clear. Neutralizing antibody responses elicited by the currently available live attenuated vaccine have been correlated with protection fromMV-induced disease . Furthermore, a formalin-inactivated virus vaccine that induces the development of humoral immune responses has been shown to be part of an effective vaccination strategy when exposure to MV occurs within a few months of vaccination [7, 8]. Nevertheless, hypogammaglobulinemic children recover uneventfully from MV infection, suggesting that antibody responses may not be required for normal clearance of MV .
Rhesus monkeys provide a powerful animal model for studies of measles pathogenesis. MV replicates in rhesus monkeys and induces immunosuppression and clinical manifestations of disease, including a morbilliform rash . Therefore, the rhesus monkey is a useful model for studying the immune control of MV, as well as for testing novel measles vaccination strategies [10, 11]. In the present study, we investigated the role that humoral immunity plays in the containment of MV replication in rhesus monkeys, through depletion of CD20+ lymphocytes and subsequent challenge with MV. In addition, we depleted both CD20+ and CD8+ lymphocytes in monkeys, to evaluate the combined contribution of humoral and CD8+ lymphocyte-mediated immune responses in the recovery from MV infection.
Materials and Methods
Monkeys and virus. Twelve 2-year-old rhesus monkeys were shown to be MV naive by negative MV-specific ELISA and neutralizing antibody assays. Four of these monkeys received an intravenous infusion of a monoclonal mouse/human chimeric anti-human CD20 antibody (Rituxan; IDEC Pharmaceuticals), at a dose of 20 mg/kg, on days −3, 4, and 10 after MV infection. Four other rhesus monkeys were similarly infused, with both anti-CD20 monoclonal antibody (MAb) (20 mg/kg on days −3, 4, and 10) and a monoclonal mouse/human chimeric anti-CD8 antibody (cM-T807; Centocor; 5 mg/kg on days −3, 0, and 4). As controls, 2 rhesus monkeys were infused with a control anti-respiratory syncytial virus MAb (Synagis; MedImmune), at a dose of 5 mg/kg, on days −3, 0, and 4 after MV infection, and 2 other healthy rhesus monkeys were treated with the same control antibody, at a dose of 20 mg/kg, on days −3, 4, and 10 after infection.
All monkeys were inoculated intratracheally with 1×104 TCID50 of Bilthoven strain MV on day 0. Preinoculation excisional lymph node biopsies were performed on 4 control monkeys, and postinoculation lymph node biopsies (days 7 or 10 and 21 or 24 after infection) were performed on all monkeys. Punch-skin biopsies were performed as the rash appeared on all monkeys (days 10, 14, 17, and 21 after infection). For all biopsies and injection and inoculation procedures, monkeys were anesthetized with ketamine HCl. On day 42 after infection, all monkeys were killed, and complete necropsies were performed. The monkeys were maintained in accordance with the guidelines of the Committee on Animals for Harvard Medical School and the Guide for the Care and Use of Laboratory Animals. The animal protocol for the present study was reviewed and approved by institutional review boards at Harvard Medical School and New England Regional Primate Center.
Flow cytometry. Phenotypic characterization of lymphocytes in peripheral blood from the monkeys was performed by cell staining and flow-cytometric analysis. Monoclonal anti-CD20-fluorescein isothiocyanate (FITC) (B1; Beckman Coulter), anti-CD8-phosphatidylethanolamine (PE) (SK1; Becton Dickinson), anti-CD4-energy-coupled dye (19Thy5D7), and anti-CD3-allophycocyanin (APC) (FN18) were added to 100 µL of EDTAanticoagulated whole blood and incubated for 15 min, and 2 mL of lysis buffer (BD Biosciences) was added. Cells were then washed and resuspended in 2% formaldehyde. A separate tube of 100 µL of blood was stained with monoclonal anti-CD20-FITC (B1; Beckman Coulter) and anti-CD19-PE (J4.119; Beckman Coulter), for 15 min, and red blood cells were lysed as described above. Cells were then washed, and 500 µL of Cytofix/Cytoperm (BD PharMingen) was added and incubated for 45 min. Cells were washed twice with Perm/wash buffer (BD Phar-Mingen) and stained with monoclonal anti-CD79a-APC (HM47; BD PharMingen) for 15 min. Cells were washed again and resuspended in 2% formaldehyde. Data were acquired by use of a flow cytometer (FACSCalibur; Becton Dickinson) and analyzed with CellQuest software (Becton Dickinson Immunocytometry Systems). Absolute lymphocyte counts for blood specimens were obtained by use of a T540 Hematology Analyzer (Beckman Coulter).
Viremia. Serial dilutions of peripheral blood mononuclear cells (PBMCs) were cocultivated with B95-8 cells in triplicate, and wells were scored for the presence of syncytia at 72 h . Data are reported as the number of syncytia/100,000 PBMCs.
Antibody assays. Neutralizing antibody was measured in a plaque-reduction assay by use of the Chicago-1 strain of MV and Vero cells, as described elsewhere . MV-specific IgG was measured by use of the MV IgG Indirect Enzyme Immunoassay kit (Sigma), by substituting an alkaline phosphatase-conjugated rabbit anti-monkey IgG (Biomakor; Accurate Chemicals) for the secondary antibody. For all assays, serum samples were diluted 1:100, and secondary antibodies were diluted 1:2000, in 1% normal rabbit serum and 0.05% Tween 20 in PBS. Fast pNPP (Tab set-n2770; Sigma) was used as a substrate for the alkaline phosphatase-conjugated rabbit anti-monkey IgG, and plates were read by optical density. Samples were run in duplicate, and displayed values represent mean change in absorbance from preinfected serum obtained from the same monkey.
Quantitative reverse-transcriptase polymerase chain reaction (RT-PCR). A pBluescript SK plasmid containing theMVhemagglutinin gene (a gift from Alex Valsamakis, Johns Hopkins School of Medicine) was used to create an RNA standard, as described elsewhere . One-step RT-PCR was performed by use of the TaqMan One-Step RT-PCR kit (PE Applied Biosystems) andMVHgene-specific primers (5′-CAATCGAGCATCAGGTCAAGG-3′ and 5′-GTCCTCAGGCCCACTTCATC-3′) and labeled probe (5′-FAM-CGTGCTACACCACTCTTCAAAATCATCGG-TAMRA-3′) (PE Applied Biosystems), and the assay was performed by use of an ABI Prism 7700 (Applied Biosystems) under conditions reported elsewhere . For each sample, the cycle threshold value, defined as the minimum number of cycles necessary to exceed threshold values, was measured and applied to the standardization curve created from the RNA transcript dilution series. Sensitivity was 15 copies of MV RNA. Statistical comparison between control and experimental groups was performed by use of the Mann-Whitney U test.
Immunohistochemistry. Skin and lymph node biopsy specimens were fixed in 10% buffered neutral formalin. The formalin-fixed tissues were embedded in paraffin, sectioned at 5 mm, and immunostained with either anti-CD20 MAb (clone L26; Dako) or anti-CD8 MAb (1A5; Vector Laboratories). Tissues stained for CD8 were preheated in an electric pressure cooker for 15 min with Trilogy solution (Cell Marque). All tissues were immersed in 3% hydrogen peroxide, to quench endogenous peroxidase, and then were treated with serum-free protein block (DAKO) for 10 min before staining with anti-CD20 MAb (1:350 dilution for 30 min) or anti-CD8 MAb (1: 50 dilution for 60 min). CD20 staining was followed by the addition of biotinylated goat anti-mouse Ig for 30 min and ABC Elite (Vector) for 30 min. CD8 staining was followed by the addition of EnVision+-labeled polymer (DAKO). In both cases, antigen was visualized by use of diaminobenzidine as the chromogen. For each primary antibody, a negative control of mouse IgG1 was used at the same concentration. Sections were counterstained with hematoxylin, dehydrated, and mounted.
Effect of MAb infusions on the targeted subsets of lymphocytes in peripheral blood from the monkeys. CD8+ T lymphocytes were quantitated by use of a PE-conjugated MAb that was able to bind to CD8 in the presence of cM-T807, as reported elsewhere . CD20+ B lymphocytes were quantitated by use of an FITC-conjugated anti-CD20 MAb. Flow-cytometric analysis using anti-CD20, anti-CD19, and anti-CD79a all provided similar results, as reported elsewhere . When CD8+ lymphocytes were undetectable, >95% of the remaining lymphocytes were CD20+ or CD4+ lymphocytes. Moreover, when CD20+ lymphocytes were undetectable, >95% of the remaining lymphocytes were CD3+ lymphocytes.
CD20+ lymphocytes were undetectable in peripheral blood from the 4 monkeys treated with anti-CD20 MAb alone, for at least 35 days after infection (figure 1B). In the 4 monkeys treated with the anti-CD20 and anti-CD8 MAbs, CD20+ lymphocytes remained undetectable in peripheral blood for at least 21 days after infection (figure 1C). CD20+ lymphocytes were subsequently detected in peripheral blood from half of the monkeys in both groups of monkeys by day 42 after infection (figure 1B and 1C). CD8+ T lymphocytes remained undetectable on day 14 after infection in peripheral blood from the 4 monkeys treated with the anti-CD8 MAb and were detectable by day 21 in peripheral blood from all monkeys (figure 1F).
A marked depletion of CD20+ lymphocytes was seen in lymph node samples obtained on day 7 from monkeys treated with the anti-CD20 MAb (figure 2B and 2C). The majority of the anti-CD20 staining in the lymph nodes from these anti-CD20-treated monkeys was noncellular and reticular in pattern (figure 2C), indicating that a portion of the infused anti-CD20 MAb is captured in the follicular dendritic cell network, as described elsewhere . The majority of CD8+ lymphocytes were depleted from lymph nodes from monkeys treated with the anti-CD8 MAb, on day 7 after infection (figure 2F). CD20+ and CD8+ lymphocyte numbers in peripheral blood or inguinal lymph nodes from the monkeys were not altered by administration of control MAb (figures 1A and 1D and 2A and 2D).
Delay and suppression of MV-specific humoral responses in monkeys treated with the anti-CD20 MAb. MV-specific IgG was detectable by ELISA in serum from all control monkeys on day 14 after infection, with peak titers reached by day 28 (figure 3A). Only 1 of 4 CD20+ lymphocyte-depleted monkeys developed a high-titer MV-specific IgG response (change in absorbance >0.6), and this response was delayed until day 42 after infection (figure 3B). Moreover, 2 of the CD20+ and CD8+ lymphocyte-depleted monkeys failed to develop detectable anti-MV IgG responses, as determined by a change in absorbance >0.4 between the serum samples. The remaining 2 monkeys developed delayed low-titer antibody responses (figure 3C). On day 14 after infection, significant (>100 mIU) MV-specific neutralizing antibody titers were seen in serum from all control monkeys (figure 3D), but not in serum from the CD20+ lymphocyte-depleted monkeys (figure 3E and 3F). On day 28 after infection, 2 of the 4 CD20+ lymphocyte-depleted monkeys and 3 of the 4 CD20+ and CD8+ lymphocyte-depleted monkeys had neutralizing antibody titers >100 mIU. Therefore, depletion of CD20+ lymphocyte was successful in delaying the development of MV-specific humoral responses during acute measles infection and suppressed the magnitude of the response in some monkeys.
No effect of the 35-day depletion of CD20+ lymphocytes alone on the clinical course of MV infection in monkeys. The control monkeys and the CD20+ lymphocyte-depleted monkeys developed a clinically indistinguishable maculopapular abdominal skin rash between days 10 and 14 after infection (figure 4A and 4B). These rashes resolved by day 21 after infection. The histologic appearances of the abdominal skin biopsy specimens obtained on day 14 or 17 from the control monkeys (n = 2) and the CD20+ lymphocyte-depleted monkeys (n = 1) were similar. They were characterized by minimal epidermal hyperplasia and mild multifocal perivascular edema, with a predominantly mononuclear inflammatory cell infiltrate in the dermis (figure 4D and 4E).
However, the CD20+ and CD8+ lymphocyte-depleted monkeys developed a desquamating rash between days 10 and 14 after infection that remained evident until day 24 after infection (figure 4C). The rash on these monkeys was more severe and extensive than that observed on the control monkeys and the CD20+ lymphocyte-depleted monkeys, covering the abdomen, chest, neck, and face. The skin biopsy specimens of the abdominal desquamating rash from the CD20+ and CD8+ lymphocyte-depleted monkeys (n = 4) obtained on days 14 and 17 showed much more severe pathological abnormalities. The biopsy specimens showed multifocal parakeratosis and occasional subcorneal pustules and marked epidermal hyperplasia with intercellular edema, as well as diffuse edema and mixed inflammatory cell infiltrates in the superficial dermis (figure 4F).
Similar extent and duration of MV replication in the control monkeys and the CD20+ lymphocyte-depleted monkeys. Infectious virus, measured by cocultivation techniques, was undetectable in both the control monkeys and the CD20+ lymphocyte-depleted monkeys by day 14 after infection (figure 5A and 5B). Moreover,MV RNA was undetectable by quantitativeMVspecific RT-PCR in PBMCs from all control monkeys and CD20+ lymphocyte-depleted monkeys by day 17 (figure 5D and 5E). CD20+ lymphocyte-depleted monkeys did not have higher peaks of infectious virus or viral RNA in PBMCs than did the control monkeys. Moreover, no MV was detected in the cerebrospinal fluid (CSF) from control monkeys or CD20+ lymphocyte-depleted monkeys (data not shown).
In contrast, MV replication in the CD20+ and CD8+ lymphocyte-depleted monkeys was greater and more prolonged than that in the control monkeys. Infectious virus remained detectable in the PBMCs from these lymphocyte-depleted monkeys until day 21 after infection (figure 5C), and MV RNA was detectable in these PBMCs until day 28 after infection (figure 5F), even when these measures were corrected for white blood cell counts (data not shown). Compared with those in control monkeys, MV RNA levels on day 24 after infection remained significantly different in the CD20+ and CD8+ lymphocyte-depleted monkeys (P = .0286) but were not significantly different in the CD20+ lymphocyte-depleted monkeys (P = .49). Finally, on day 14 after infection, MV was detected in CSF from 3 of the 4 CD20+ and CD8+ lymphocyte-depleted monkeys by cocultivation techniques and in all 4 monkeys by RT-PCR (data not shown). Therefore, although depletion of CD20+ lymphocytes alone did not result in increased MV replication in monkeys during primary infection, depletion of CD20+ and CD8+ lymphocytes was associated with increased MV replication.
In the present study, we sought to determine the role that antibody responses play in the clinical outcome and viral clearance kinetics of MV infection in the rhesus monkey model. The depletion of CD20+ lymphocytes and the resulting delay in the generation of MV-specific antibody responses did not change the virologic or clinical consequences of MV infection in rhesus monkeys. However, monkeys depleted of both CD20+ and CD8+ lymphocytes and then infected with MV developed prolonged measles viremia, with detectable virus in the CSF and a more severe skin rash. These findings suggest that MVspecific antibody responses play a minor role in the control of MV replication in nonimmunized monkeys. In addition, the present study has confirmed the importance of CD8+ lymphocytes in the control and clearance of MV.
The depletion of CD20+ lymphocytes in the monkeys resulted in a 1-week delay in their development ofMV-specific antibody. This delay, however, did not change the course of the MV infection. Although virus became undetectable in the blood of all the CD20+ lymphocyte-depleted monkeys by day 17, a significant neutralizing anti-MV antibody response (>100 mIU) was not detected until day 21 after infection. These findings suggest a temporal discordance between viral clearance and the development of a functional MV-specific antibody. These findings are consistent with the possibility that antibody does not play a role in containment of MV. It is, however, also possible that the very-low-titer MV-specific antibody responses detected before day 21 in the CD20+ lymphocyte-depleted monkeys were sufficient to facilitate viral clearance.
Previous in vitro and clinical studies have suggested a significant role for humoral immunity in the control of MV replication. Treatment of MV-infected cells with polyclonal antibody and MAbs specific for MV proteins altered the expression of viral antigens and down-regulated MV replication [15–18]. In MV-infected individuals, the evolution of MV-specific antibody-dependent cellular cytotoxicity correlated temporally with clearance of cell-associated virus . In studies of exposure to MV in MV-naive individuals, administration of MV immunoglobulin within 5 or 6 days of exposure prevented the clinical manifestations of MV infection. If the immunoglobulin was administered later in the incubation period, the manifestations of measles were altered [20, 21]. However, in most of these studies, the role of cellular immunity in clearance of MV was not addressed.
The results of the present study are consistent with other clinical observations concerning containment of MV. Although prolonged viral shedding in Zambian children was seen in association with HIV infection, it did not correlate withMV-specific functional antibody titers . The results of the present study of monkeys also support the early clinical observation that measles is uncomplicated in hypogammaglobulinemic children but particularly severe in children with deficiencies in cell-mediated immunity .
The MV rash is a manifestation of the cellular immune response of the infected individual . Therefore, the quality of the rash is likely related to the status of that individual's cellular immune response. In a study of Zambian children hospitalized with measles, children whose growth was stunted or who were infected with HIV were likely to have a desquamating rash . It is interesting that the rash seen on the monkeys depleted of both CD8+ and CD20+ lymphocytes was desquamating and diffusely distributed, whereas the rash seen on control monkeys and CD20+ lymphocyte-depleted monkeys was morbilliform and localized to the abdomen. It has been previously shown that both CD4+ and CD8+ lymphocytes are present in the skin rash . Since the development of the measles rash is dependent both on MV replication in the skin and on the MV-specific cellular immune response, the severe rash seen on the CD8+ and CD20+ lymphocyte-depleted monkeys suggests that CD4+ T lymphocytes are important mediators of this clinical manifestation of disease. In our previous study of CD8+ lymphocyte-depleted monkeys that were inoculated with the same dose and stock of virus as in the present study, the rash was more extensive and prolonged, compared with that in control monkeys, but was not desquamating . This finding may indicate that MV-specific antibody contributes to modulation of MV replication or the immune response to MV in the epithelia. Alternatively, the differences in the character of the rashes in the groups of monkeys in these 2 studies could be explained by the younger ages of the monkeys in the present study.
Although our findings support a limited role for humoral immunity in the control of replicating virus in MV-naive hosts, antibody can certainly mediate protection against MV infection. The formalin-inactivatedMVvaccine, a vaccine that is replication incompetent and therefore unable to prime CD8+ T lymphocyte immune responses , was successful in preventing infection before antibody titers waned . The development of a neutralizing antibody response after immunization with the currently available live attenuated vaccine has been correlated with protection against infection , and newborn infants are protected from infection by passively transferred neutralizing antibody. In addition, the ability of the live attenuated vaccine to induce immunity in young children is blocked by circulating maternal antibody, presumably because this passively transferred antibody neutralizes the vaccine virus before sufficient replication and immune priming occurs . Such a discordance between the role of preexisting antibody in protection against a viral infection and the clearance of virus in an established infection is apparent with other viruses, including simian immunodeficiency virus [27–31] and hepatitis B virus [32, 33].
Although there is no absolute requirement for vaccine-elicited T cell immunity in protection from infection, cell-mediated immunity is a viable mechanism for clearance of replicating virus. Since vaccine-elicited antibody titers wane, cell-mediated immunity may prove to be an important means of abolishing early MV replication on exposure to MV. The ideal MV vaccine should therefore elicit both neutralizing antibody and cell-mediated immunity, to ensure protection or rapid clearance of locally replicating virus.
We are grateful to V. Petkova, A. Valsamakis, C. H. Pan, R. Khunkhun, M. Stremlau, and C. Lord, for assistance, reagents, and advice.
Figures and Tables
- excretory function
- drug clearance
- monoclonal antibodies
- antibody formation
- cd20 antigens
- developing countries
- macaca mulatta
- measles vaccine
- measles virus
- immunity, humoral
- immune response, cell-mediated
- neutralizing antibodies
- infusion procedures