Abstract

T cell proliferation to human parvovirus B19 antigen was measured in 6 patients with recent B19 infection (1 with pneumonia and pleuritis), 1 patient with symptoms persisting >180 days after onset, 18 nonsymptomatic subjects with remote B19 immunity, and 12 B19-seronegative control subjects. Recombinantly expressed virus-like particles (VP1/2 capsids), a candidate B19 vaccine, were used as antigen. Virus-specific T helper cell proliferation was detectable in all the recently infected patients and in most (17/18) of the remotely infected subjects but not in the seronegative control subjects. The B19-specific T cell responses, in general, were most vigorous among the recently infected patients. However, such strong B19-specific proliferation was not confined within the acute phase, as 28% (5/18) of the remotely infected healthy individuals had B19-specific reactivity persisting at acute-phase levels, apparently for years or decades. These data indicate that B cells recognizing the VP1/2 capsids receive class II–restricted help from CD4+ T lymphocytes

Parvovirus B19 is a significant human pathogen that causes a wide spectrum of illnesses. A typical manifestation is fifth disease (erythema infectiosum), a benign febrile rash with or without arthropathy [1]. The joint symptoms occasionally persist for years and may mimic rheumatoid arthritis [2]. In patients with hemolytic disposition, B19 infection may cause transient aplastic crisis [3]. Among immune-compromised persons [4] and, rarely, among immunocompetent individuals [5, 6], B19 infection may remain chronic. B19 infection sometimes causes fetal hydrops and fetal death during pregnancy [7]

Because important morbidity accompanies B19 infection, recombinant vaccines for this virus are being developed [8]. For efficient B19 eradication, the vaccine should elicit neutralizing antibodies [4]. Empty capsids consisting of the 2 B19 structural proteins VP1 and VP2 have fulfilled this criterion in a murine model [8]. In the present study, we show for the first time how such a recombinant antigen stimulates human T helper lymphocytes that are essential for antibody class switch and affinity maturation [9]

Subjects and Methods

Study groupsTo study recently infected patients, we obtained peripheral blood mononuclear cells (PBMC) from 6 individuals (designated patients R1–R6) 30–55 days after the onset of sero-logically documented parvovirus B19 infection. All patients had to have detectable B19 IgM and an “acute pattern” of epitope-type specificity of VP2 IgG (see below). Patients included 1 man and 5 women who were 30–55 years old. Patients were symptomatic with a rash and arthralgia, and most of them also had a fever. None was pregnant. Patients R1–R5 recovered normally, and all but 1 (R4, who had a rash) were asymptomatic on the day of PBMC preparation. Patient R6 had particularly severe B19 infection and required hospitalization because of pneumonia and pleuritis, which were diagnosed 9 days after the onset of fever, rash, and arthralgia. Initial intravenous cefuroxime-levofloxacin treatment was ineffective, but a peroral course of prednisone (20 mg daily for 7 days, with a subsequently tapered dose) cured the pleuritis permanently. The prednisone treatment commenced 3 weeks before PBMC collection, at which time the patient was asymptomatic. Blood cultures were negative for bacteria, and acute B19 infection was considered to be the etiology of the pulmonary disease. Follow-up samples were obtained from patients R1 and R5, who were both asymptomatic, and from patient R6, who had recurrent severe arthralgia in knees and knuckles

We also studied a patient with persistent symptoms. The patient (designated L1) was a 30-year-old woman with recurrent rash (main complaint) and mild arthralgia for >6 months after B19 infection was serologically verified (as in recently infected patients). By the day of PBMC collection, all tests and examinations for rheumatoid arthritis, systemic lupus erythematosus, and common microbial etiologies for postinfectious arthropathy in Northern Europe (Salmonella, Campylobacter, Chlamydia and Streptococcus species and Sindbis virus) were negative, and a diagnosis of prolonged postinfectious arthropathy caused by B19 was made

We also studied 18 remotely infected subjects (4 men and 14 women; 21–51 years old), who were designated P1–P18. This group consisted of healthy staff members who were seropositive for parvovirus B19. In addition, we studied 12 healthy laboratory workers (5 men and 7 women; 22–46 years old) who were seronegative for parvovirus B19. The members of this nonimmune group were designated N1–N12

Antibody assaysWe measured serum B19 IgG in a 1:200 dilution by EIA, employing as antigen virus-like VP2 particles expressed and purified as described elsewhere [10]. B19 IgM was studied by using a commercial EIA (Biotrin). For confirmation or exclusion of recent B19 infection, all samples were further studied, as described elsewhere [10], for epitope-type specificity of VP2 IgG

Antigens for proliferation assaysVP1/2 capsids were expressed and purified by using Spodoptera frugiperda (Sf 9) cells infected with recombinant baculovirus, as described elsewhere [10], except for use of the p10 promoter with VP1 and ultracentrifugation in 28% CsCl gradients (at 100,000 g for 48 h). After dialysis against PBS, the capsids were sterile filtered by using 0.2-μm filters (Anotop 10 Plus; Whatman). Total protein concentration in the sterile capsid preparations was determined by bicinchoninic acid protein assay reagents (Pierce). Purity for B19 proteins was >90%, as determined by SDS–PAGE with silver staining (Novex SilverXpress Silver Staining Kit; Novel Experimental Technology) and densitometry (Gel Doc 2000 Gel Documentation systems with Quantity One Quantitation Software; Bio-Rad)

The capsid preparations contained ∼66% VP2 and 33% VP1, the ratios recommended for vaccine use [8]. Electron microscopy with negative staining showed native-like B19 capsids (data not shown). The level of endotoxin was 0.0035 EU/μg VP1/2 protein, as determined by the Limulus amebocyte lysate assay (QCL-1000; BioWhittaker), as used according to the manufacturer’s instructions. After titration, the VP1/2 capsids were used at 2.5 μg/mL

As control antigens, we used 5 μg/mL each of purified protein derivative of Mycobacterium tuberculosis (PPD; Statens Serum Institut) and of tetanus toxoid (TT; National Public Health Institute, Helsinki). Candida albicans antigen was prepared in house, was heat inactivated, and was used at 2.5 μg/mL

Isolation of PBMCBlood (8 mL) from staff members and from recently infected patients was collected into mononuclear cell separation tubes (Vacutainer CPT; Becton Dickinson). Cells were spun for 30 min at 1700 g were washed twice in PBS, and were prepared within 2 h (for staff members) or 8 h (for recently infected patients) of sampling

Proliferation assayIsolated PBMC were resuspended in complete RPMI 1640 containing 20 mM HEPES, 2 mMl-glutamine, streptomycin (100 μg/mL), penicillin (100 U/mL), 50 μM 2-mercaptoethanol, and 10% heat-inactivated human AB serum (Finnish Red Cross Blood Transfusion Service) containing B19 IgG. Manually counted PBMC (200,000 cells/well) and the antigens in triplicate were placed in 96-well U-bottom plates (Costar; Corning). Cells were cultured for 6 days (37°C, 5% CO2) and were pulsed for the last 16 h with 1 μCi of tritiated thymidine (specific activity 50 Ci/mmol; Nycomed Amersham). Thymidine incorporation was measured in a liquid scintillation counter (Microbeta; Wallac). Data are expressed both as Δ counts per minute (cpm) and as stimulation indices (SIs), respectively: Δ cpm=mean cpm (test antigen)-mean cpm (media); SI=mean cpm (test antigen)/mean cpm (media)

Depletion of CD4+or CD8+cellsCD4+ or CD8+ T cells were depleted from PBMC, using magnetic beads coated with CD4- and CD8-specific monoclonal antibodies (MAbs; Dynabeads M-450; Dynal), according to the manufacturer’s instructions. Depletion efficiency was confirmed by flow cytometry (FACScan; Becton Dickinson)

Antibody blocking assaysClass restriction of the T cell responses was further studied by using HLA class I–specific (IgG2a, clone W6/32; Dako) and class II–specific (HLA-DR) (IgG2a, clone L243; Becton Dickinson) MAbs, as described elsewhere [11], except that the antibodies were used (after extensive dialysis for removal of NaN3) at 1.5 μg/mL instead of 1.0 μg/mL. Percentage inhibition of proliferation was calculated by the formula 100×[1-(cpmantigen + MAb − cpmMAb)/(cpmantigen − cpm media)]

Statistical methodsThe Δ cpm and SI values were statistically evaluated by using the Mann-Whitney U test, and the distribution of responders having Δ cpm >5000 and SI >5 against antigen was studied by using Fisher’s exact test. P<.05 was considered significant

Results

Proliferation responses among remotely infected and seronegative subjectsThe B19-specific T cell responses of the remotely infected B19-seropositive subjects were, in general, much stronger (table 1) than the responses of the seronegative subjects, regardless of proliferation criteria (Δ cpm, SI, or responder having Δ cpm >5000 and SI >5; P<.0001). T cell reactivity was absent with proteins derived from uninfected Sf-9 cells (data not shown). With the control antigens TT, PPD, and Candida no statistically significant differences (by any criteria) were found between the B19-seropositive and -seronegative subjects, and the same held for background proliferation

Table 1

Peripheral blood mononuclear cell (PBMC) proliferation to VP1/2 antigen and to control antigens among recently, remotely, and persistently parvovirus B19–infected subjects, compared with proliferation responses among nonimmune subjects

Table 1

Peripheral blood mononuclear cell (PBMC) proliferation to VP1/2 antigen and to control antigens among recently, remotely, and persistently parvovirus B19–infected subjects, compared with proliferation responses among nonimmune subjects

Comparison of recently and remotely infected subjectsThe B19-specific responses of the recently infected patients were stronger than those of the remotely infected subjects, as assessed by Δ cpm (P=.04). However, statistical significance was not reached (P=.5) by SI calculation. With the control antigens, no statistically significant differences were found by Δ cpm, whereas by SI, the recently infected patients showed weaker (P=.02) TT-antigen reactivity (table 1). Among the recently infected patients, the highest B19-specific responses were seen with patient R1 (also highest TT- and PPD-specific reactivity) and patient R6 (most severe clinical picture)

Even though the recently infected patients generally had stronger B19 reactivity (by Δ cpm) than did the remotely infected subjects, some of the strongest responders (“top responders”) of the latter group had B19-specific responses comparable to the highest values seen among the recently infected patients. These top responders (subjects P1–P5) showed higher VP1/2-, TT-, and PPD-specific responses (P<.03 by Δ cpm and SI) than did the rest of the seropositive subjects (P6–P18), who, in turn, had comparable TT- and PPD-specific responses but stronger VP1/2-specific responses (P=.0003 by SI and P<.0001 by Δ cpm) than did the seronegative control subjects (table 1)

Comparison of initial and follow-up data from recent infectionAs shown in table 1, during follow-up, the B19-specific T cell response of patient R1, who had self-limiting infection, was well maintained and was comparable to the vigorous B19-specific T cell reactivity of patient L1, who had persistent rash and arthralgia after B19 infection. Patients R5, who had self-limiting infection, and R6, who had pneumonia, pleuritis, and relapsed arthralgia, both maintained virus-specific T cell reactivity during follow-up at acute-phase levels

Identification of the proliferating cellsTo characterize the proliferating cell population, we depleted the PBMC of either CD4+ or CD8+ T cells, using MAbs attached to magnetic beads. Although CD8 depletion had little effect on B19-specific T cell proliferation, CD4 depletion abrogated the responses among all the groups studied (figure 1A). Blocking experiments that used class II (HLA-DR)–specific MAbs reduced the T cell responses of the top responders by 74%–80% and those of the recently infected patients by 80%–89%. T cell proliferation of L1 was inhibited by 97%. Class I antibodies inhibited the T cell responses by only ⩽15% (figure 1B)

Figure 1

A Antibody-mediated depletion of CD4+ and CD8+ cells, as determined by use of magnetic beads. Five subjects were assayed before and after depletion of T cell subsets. Typical percentages of CD3+ T cells in the unseparated peripheral blood mononuclear cells (PBMC) were 63% CD4+ and 34% CD8+. After CD4 depletion, there were 2% CD4+ and 87% CD8+, and, after CD8 depletion, there were 92% CD4+ and 3% CD8+. B Effect of HLA class I–specific (W6/32) and class II–specific (L243) monoclonal antibodies (MAbs) on PBMC proliferation of 5 subjects. In calculation of Δ counts per minute (cpm), wells with and wells without (controls) antigen contained the MAbs indicated. Subject groups: “P” indicates healthy B19-seropositive “remotely infected” patients (top responders are the strongest responders of the group); “R” indicates recently infected patients; and “L” indicates patient with symptoms persisting >6 months after confirmation of B19 infection

Figure 1

A Antibody-mediated depletion of CD4+ and CD8+ cells, as determined by use of magnetic beads. Five subjects were assayed before and after depletion of T cell subsets. Typical percentages of CD3+ T cells in the unseparated peripheral blood mononuclear cells (PBMC) were 63% CD4+ and 34% CD8+. After CD4 depletion, there were 2% CD4+ and 87% CD8+, and, after CD8 depletion, there were 92% CD4+ and 3% CD8+. B Effect of HLA class I–specific (W6/32) and class II–specific (L243) monoclonal antibodies (MAbs) on PBMC proliferation of 5 subjects. In calculation of Δ counts per minute (cpm), wells with and wells without (controls) antigen contained the MAbs indicated. Subject groups: “P” indicates healthy B19-seropositive “remotely infected” patients (top responders are the strongest responders of the group); “R” indicates recently infected patients; and “L” indicates patient with symptoms persisting >6 months after confirmation of B19 infection

Discussion

T cell reactivity with the B19 virus has been thought to be difficult to measure [4, 12]. Von Poblotzki et al. [11] were pio-neers in showing that the T cells of subjects with preexisting B19 immunity show HLA II–restricted responses against prokaryotically expressed B19 structural proteins. Murai et al. [13] observed that proliferation responses with a recombinant VP1 antigen were higher in patients with rheumatoid arthritis after B19 infection than they were in healthy control subjects

We studied patients with recent B19 infection and observed strong T helper cell reactivity with the VP1/2 capsid antigen among all the patients, including one treated with prednisone, which is known to inhibit the clonal expansion of effector T cells from naive T cells [14]

Our data show that strong B19-specific T cell activity is not confined to patients with arthropathy, because the follow-up patients who recovered normally showed vigorous B19-specific proliferation, as did our 2 patients with persistent or relapsing arthralgia, and we found top responders among healthy seropositive subjects. These top responders had B19-specific T cell activity comparable to the corresponding B19-specific responses among recently infected patients and comparable to the same subjects’ strong responses against established control antigens. Since the top responders had vigorous control antigen–specific reactivity, their strong VP1/2-specific reactivity is most readily explained by a good general ability to maintain T cell memory to recall antigens, and the same holds for R1, the recently infected patient. Vigorous B19-specific proliferation as a marker is clearly not sufficiently specific for diagnosis of recent or complicated B19 virus infection. However, as most of the recently infected patients showed relatively low proliferation with the control antigens, coexistence of high B19-specific and low TT- and PPD-specific proliferation could indicate recent or complicated B19 infection better than mere top responder status

All in all, we showed that B19-specific memory T cell proliferation is readily observable, using VP1/2 capsids. Whether such capsids induce in vivo antigen-specific T cells from naive T cells needs to be ascertained in further studies

Acknowledgments

We thank Leena Kostamovaara for help with VP1/2 capsid production, Lea Hedman for carrying out the antibody assays, Leena Kaikkonen (Department of Virology, Haartman Institute and Helsinki University Central Hospital Diagnostic, University of Helsinki, Helsinki) for help with image processing, and John Brunstein (University of British Columbia, Vancouver) for densitometry. We are grateful to Helena Lanki and Ari Rantanen (Malmi Hospital, Helsinki) for clinical data

References

1
Anderson
MJ
Lewis
E
Kidd
IM
Hall
SM
Cohen
BJ
An outbreak of erythema infectiosum associated with human parvovirus infection
J Hyg
 , 
1984
, vol. 
93
 (pg. 
85
-
93
)
2
White
DG
Mortimer
PP
Blake
DR
Woolf
AD
Cohen
BJ
Bacon
BA
Human parvovirus arthropathy
Lancet
 , 
1985
, vol. 
1
 (pg. 
419
-
21
)
3
Saarinen
UM
Chorba
TL
Tattersall
P
, et al.  . 
Human parvovirus B19–induced epidemic acute red cell aplasia in patients with hereditary hemolytic anemia
Blood
 , 
1986
, vol. 
67
 (pg. 
1411
-
7
)
4
Kurtzman
GJ
Cohen
BJ
Field
AM
Oseas
R
Blaese
M
Young
NS
Immune response to B19 parvovirus and an antibody defect in persistent viral infection
J Clin Invest
 , 
1989
, vol. 
84
 (pg. 
1114
-
23
)
5
Faden
H
Gary
GW
Anderson
LJ
Chronic parvovirus infection in a presumably immunologically healthy woman
Clin Infect Dis
 , 
1992
, vol. 
15
 (pg. 
595
-
7
)
6
Söderlund
M
von Essen
R
Haapasaari
J
Kiistala
U
Kiviluoto
O
Hedman
K
Persistence of parvovirus B19 DNA in synovial membranes of young patients with and without chronic arthropathy
Lancet
 , 
1997
, vol. 
349
 (pg. 
1063
-
5
)
7
Anand
A
Gray
ES
Brown
T
Clewley
JP
Cohen
BJ
Human parvovirus infection in pregnancy and hydrops fetalis
N Engl J Med
 , 
1987
, vol. 
316
 (pg. 
183
-
7
)
8
Bansal
GP
Hatfield
JA
Dunn
FE
, et al.  . 
Candidate recombinant vaccine for human B19 parvovirus
J Infect Dis
 , 
1993
, vol. 
167
 (pg. 
1034
-
44
)
9
De Kruyff
RH
Rizzo
LV
Umetsu
DT
Induction of immunoglobulin synthesis by CD4+ T cell clones
Semin immunol
 , 
1993
, vol. 
5
 (pg. 
421
-
30
)
10
Kaikkonen
L
Lankinen
H
Harjunpää
I
, et al.  . 
Acute-phase–specific heptapeptide epitope for diagnosis of parvovirus B19 infection
J Clin Microbiol
 , 
1999
, vol. 
37
 (pg. 
3952
-
6
)
11
Von Poblotzki
A
Gerdes
C
Reischl
U
Wolf
H
Modrow
S
Lymphoproliferative responses after infection with human parvovirus B19
J Virol
 , 
1996
, vol. 
70
 (pg. 
7327
-
30
)
12
Collett
MS
Young
NS
Prospects for a human B19 parvovirus vaccine
Rev Med Virol
 , 
1994
, vol. 
4
 (pg. 
91
-
103
)
13
Murai
C
Munakata
Y
Takahashi
Y
, et al.  . 
Rheumatoid arthritis after human parvovirus B19 infection
Ann Rheum Dis
 , 
1999
, vol. 
58
 (pg. 
130
-
2
)
14
Brinkmann
V
Kristofic
C
Regulation by corticosteroids of Th1 and Th2 cytokine production in human CD4+ effector T cells generated from CD45RO and CD45RO+ subsets
 
J Immunol 1995; 155 3322–8
Informed consent was obtained from the recently infected patients and the staff members
Financial support: Helsinki University Central Hospital Research and Education Fund; Finnish Technology Advancement Fund