Full text of DOI:10.1086/315871

1486
Synthetic Malaria Peptide Vaccine Elicits High Levels of Antibodies
in Vaccinees of Defined HLA Genotypes
Elizabeth H. Nardin,1 Giane A. Oliveira,1
J. Mauricio Calvo-Calle,1 Z. Rosa Castro,1
Ruth S. Nussenzweig,1 Barbara Schmeckpeper,2
B. Fenton Hall,4 Carter Diggs,6 Sacared Bodison,5
and Robert Edelman3
1Department of Medical and Molecular Parasitology, New York
2
University School of Medicine, New York; Immunogenetics
3
Laboratories, Johns Hopkins University, and Center for Vaccine
Development, University of Maryland School of Medicine, Baltimore,
4National Institute of Allergy and Infectious Diseases, National
5
Institutes of Health, Bethesda, and University Health Center,
6
University of Maryland, College Park; Malaria Vaccine Development
Program, USAID, Washington, DC
A multiple antigen peptide (MAP) malaria vaccine containing minimal Plasmodium falci-
parum circumsporozoite protein repeat epitopes was assessed for safety and immunogenicity
in volunteers of known class II genotypes. The MAP/alum/QS-21 vaccine formulation elicited
high levels of parasite-specific antibodies in 10 of 12 volunteers expressing DQB1*0603,
DRB1*0401, or DRB1*1101 class II molecules. In contrast, volunteers of other HLA geno-
types were low responders or nonresponders. A second study of 7 volunteers confirmed the
correlation of class II genotype and high responder phenotype. This is the first demonstration
in humans that a peptide vaccine containing minimal T and B cell epitopes composed of only
5 amino acids (N, A, V, D, and P) can elicit antibody titers comparable to multiple exposures
to irradiated P. falciparum–infected mosquitoes. Moreover, the high-responder phenotypes
were predicted by analysis of peptide/HLA interactions in vitro, thus facilitating the rational
design of epitope-based peptide vaccines for malaria, as well as for other pathogens.
The development of drug-resistant Plasmodium parasites and
These phase 1 and phase 2a trials demonstrated that anti–repeat
antibodies reactive with P. falciparum sporozoites were elicited
by the peptide vaccine in most of those vaccinated. Although
the antibody titers were relatively low, protection against chal-
lenge by the bite of P. falciparum–infected mosquitoes was ob-
tained in a small number of these volunteers, as evidenced by
an absence or significant delay in the development of a patent
malaria infection.
The limitations of this first-generation peptide-protein con-
jugate vaccine, which included carrier toxicity, low epitope den-
sity, possible epitopic suppression, and lack of parasite-derived
T cell epitopes, were addressed by the development of multiple
antigen peptides (MAPs) [5]. These synthetic constructs re-
placed the foreign protein carrier with a peptide core matrix
that uses the a and ␧ amino groups of lysine to construct a
branched scaffolding on which peptide epitopes can be syn-
thesized. MAPs containing repeat B cell epitopes of CS proteins
elicited high levels of antibodies and protection in mice chal-
lenged with rodent malaria sporozoites [5–7] and in Saimiri
monkeys challenged with sporozoites of the human parasite P.
vivax [8]. Moreover, recent studies that used MAPs containing
other sporozoite epitopes have demonstrated CD4ϩ T cell–
dependent protective immunity [9, 10] mediated by interferon
(IFN)–g, a potent inhibitor of the intracellular hepatic stages
of the parasite [11].
insecticide-resistant mosquito vectors has compromised tradi-
tional methods of malaria treatment and control, reemphasizing
the critical need for an effective vaccine. Monoclonal antibodies
(MAbs) directed against the repeat region of a major surface
antigen, the circumsporozoite (CS) protein, can block invasion
of the host cell and passively confer sterile immunity [1–3].
Clinical trials were carried out ∼10 years ago with a first-gen-
eration synthetic peptide malaria vaccine based on the tetramer
repeat sequence of the Plasmodium falciparum CS protein,
(NANP)3, conjugated to tetanus toxoid as a carrier protein [4].
Received 27 April 2000; revised 18 July 2000; electronically published 9
October 2000.
Presented in part: Gordon Malaria Conference, Oxford, UK, 1998; Gor-
don Parasitism Conference, Newport, RI, July 1999; 48th Annual Meeting
of the American Society of Tropical Medicine and Hygiene, Washington,
DC, December 1999 (abstract 381).
Informed consent was obtained from all volunteers, and the phase 1 trial
was approved by the institutional review board of the University of
Maryland.
Financial support: National Institutes of Health (NIH) grant AI-25085
(New York University); US Agency for International Development (New
York University); NIH research contract NO1-AI-45251 (University of
Maryland Center for Vaccine Development). HLA typing was performed
at Johns Hopkins Immunogenetics Lab as a subcontract from Georgetown
University NIH research contract NO1-AI-45242.
Reprints or correspondence: Dr. Elizabeth Nardin, New York University
School of Medicine, Dept. of Medical and Molecular Parasitology, 341 E.
25th St., New York, NY 10010 (Elizabeth.Nardin@med.nyu.edu).
An effective malaria vaccine should, therefore, contain a par-
asite-derived epitope to elicit CD4ϩ T cells that could function
both as a source of inhibitory lymphokines and of T helper
The Journal of Infectious Diseases 2000;182:1486–96
᭧ 2000 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2000/18205-0026$02.00
factors for antibody responses. We used a series of CD4ϩ
T

JID 2000;182 (November)
Phase 1 Trial of Malaria Peptide Vaccine
1487
Figure 1. (T1B)4 multiple antigen peptide (MAP) vaccine containing T and B cell epitopes of Plasmodium falciparum circumsporozoite (CS)
protein. A, Illustration of the P. falciparum CS protein showing the central repeat domain containing the T cell epitope, T1, located in the 50
repeat region and a B cell epitope, (NANP)3, located in the 30 repeat region. B, Diagram of the tetrabranched (T1B)4 MAP that contains the T1
epitope (DPNANPNV)2 synthesized in tandem with the B cell epitope (NANP)3 on each branch.
cell clones, derived from a volunteer immunized by irradiated
P. falciparum sporozoites, to identify a class II restricted T
helper epitope, designated T1, in the 50 repeat region of the CS
protein [12]. This epitope consists of alternating NANP NVDP
amino acid sequences that are conserved in all P. falciparum
strains.
Our preclinical studies demonstrated that a P. falciparum
MAP containing the T1 epitope in combination with the
(NANP)3 B cell epitope, designated (T1B)4 MAP (figure 1),
elicited extraordinarily high levels of anti-sporozoite antibodies
in mice and Aotus monkeys [13, 14]. Immunofluorescence assay
(IFA) titers exceeding 1 ϫ 106 were obtained in some of the
murine and Aotus serum samples, a magnitude of antibody
response not previously reported for any recombinant or syn-
thetic CS construct. The (T1B)4 MAP also elicited a significant
anamnestic response when administered to P. falciparum spo-
rozoite-primed mice and monkeys [14, 15], suggesting that the
vaccine would significantly increase antibody levels, and con-
comitantly protection, in individuals living in areas where ma-
laria is endemic.
a DRB1*1101 restricted CD4ϩ T cell clone in a peptide com-
petition cellular assay [17]. In addition, low-affinity binding of
the T1 peptide to soluble DRB1*0401 molecules (IC50 80 mM)
was detected in a peptide competition radioimmunoassay,
which is consistent with the presence of a DRB1*0401 binding
motif in the T1 amino acid sequence ([18]; J. Hammer and F.
Sinigaglia, personal communication). The (NANP)3 B cell epi-
tope, which lacks the aliphatic or aromatic amino acid residues
that function as P1 anchors in the class II peptide binding
groove, did not bind to any of the class II molecules in these
in vitro assays [16–18].
The current Phase I trial was designed as a prospective study,
to determine whether the (T1B)4 MAP vaccine could elicit
immune responses in individuals expressing DQB1*0603-,
DRB1*1101-, or DRB1*0401-encoded molecules. Two cohorts
of volunteers of these selected class II genotypes or random
HLA haplotypes were immunized with an alum adsorbed
(T1B)4 MAP vaccine formulation, with or without QS-21 as
coadjuvant, to assess the safety and immunogenicity.
In the murine response to (T1B)4 MAP, high or low antibody
responses correlate with specific murine H-2 haplotypes [13].
The genetic restriction of the human response to the T1 or B
repeat peptides was defined by in vitro analysis of peptide HLA
interactions. The DQB1*0603 class II molecule was shown to
function as a restriction element for the presentation of the T1
peptide to CD4ϩ T cell clones derived from a P. falciparum
sporozoite–immunized volunteer [16]. Peptides containing the
T1 epitope also blocked presentation of a cognate peptide to
Subjects and Methods
Subjects. Cohort A consisted of 32 healthy male and female
volunteers (19–40 years old) recruited at the University of Mary-
land, College Park. Cohort B consisted of 7 volunteers recruited
from the Baltimore community. Informed consent was obtained
from all volunteers before admission into the study. The clinical
protocol was approved by the institutional review boards of the
University of Maryland at Baltimore and at College Park.

1488
Nardin et al.
JID 2000;182 (November)
Table 1.
Plasmodium falciparum (T1B)4 multiple antigen peptide
Medical history, physical examination, and routine standard lab-
(MAP) phase 1 cohort A study design.
oratory tests (complete blood count, serum chemistries, urinalysis,
immunoglobulin levels, and serology) were obtained to exclude
individuals with cardiovascular, hepatic, or renal functional ab-
normalities or a history of hepatitis or human immunodeficiency
virus or malaria infection. Class II genotypes (DRB1, DRB3,
DRB4, DRB5, and DQB1 genes) were determined with the stan-
dard reverse transcriptase–polymerase chain reaction sequence–
specific oligonucleotide probe typing scheme of the 11th and 12th
International Histocompatibility Workshops [19].
Class II genotypes
DRB1*0401,
MAP/alum,
QS-21,
No. of
volunteers
DRB1*1101,
DQB1*0603
Group
mga
mgb
Random
1
2
3
500
500
—
50
8
8
4
2
4
6
1000
1000
—
50
100
—
8
5
5
16
3
3
16
4/5
Total
4 ϩ 4
32
After each injection of vaccine, volunteers were observed for 60
min and examined for local and systemic side effects by a physician.
Clinical examinations carried out at 24 and 48 h and at 7, 14, and
28 days after immunization included an interval history of systemic
and local reactions, examination of the site of injection, and oral
temperature. The detailed clinical findings, including application
of immediate-type and delayed-type hypersensitivity skin tests, will
be reported elsewhere (J. Kublin and R. Edelman, unpublished
data).
Vaccine formulation. The (T1B)4 MAP was synthesized in a
stepwise fashion, as described elsewhere [5, 13], under Good Man-
ufacturing Practices (Peninsula Laboratories, San Carlos, CA). The
MAP contained equimolar amounts of the 16-mer T1 epitope,
(DPNANPNV)2, synthesized in tandem with the 12-mer B cell epi-
tope, (NANP)3, on each arm of a tetrabranched lysine core (figure
1). Adsorption to alum (aluminum hydroxide; Reheis, Berkeley
Heights, NJ) was carried out in acetate buffer, pH 6.2, at a final
concentration of 2.5 mg (T1B)4 MAP and 3.1 mg aluminum per
milliliter. As a coadjuvant, a purified saponin, QS-21 (Aquila Bio-
pharmaceuticals, Framingham, MA) [20, 21], was mixed with the
(T1B)4 MAP/alum immediately before subcutaneous (sc) injection
into the upper arm.
a
All volunteers were immunized subcutaneously on days 0 and 28. The third
dose of vaccine was administered on day 56 to groups 1–3 and on day 237 to
group 4/5. Immunization of group 5 volunteers was initiated 1 week after that
of group 4 volunteers.
b
The 500-mg and 1000-mg doses of MAP/alum contained 625 and 1250 mg of
aluminum, respectively. The QS-21 coadjuvant was mixed with the alum-adsorbed
(T1B)4 MAP at the time of injection.
(OD) greater than the day 0 mean DOD plus 2 SD. An indirect
immunofluorescence antibody assay (IFA) was carried out by use
of air-dried P. falciparum sporozoites, incubated with 2-fold dilu-
tions of serum samples, followed by fluorescein isothiocyanate–
labeled anti–human IgG or IgM (Kierkegaard & Perry, Gaithers-
burg, MD) diluted in PBS/0.4% Evans blue.
To determine the ability of the MAP-induced antibodies to react
with CS protein expressed on the surface of viable P. falciparum
sporozoites, CS precipitin (CSP) assays were carried out [22, 23].
Dilutions of serum samples collected from volunteers 20 days after
the final immunization were mixed with viable sporozoites and
incubated for 45 min at 37ЊC. The end-point titer was the last
serum dilution that could elicit a positive CSP reaction in 2 of 20
sporozoites, as determined by phase microscopy [22].
Study design. The cohort A study was an open-label trial to
assess the dose-dependent safety and immunogenicity of the (T1B)4
MAP adsorbed to alum, with or without QS-21 as coadjuvant, in
volunteers of known HLA class II genotypes (table 1). Four groups
of 8 individuals were immunized sc with either 500 or 1000 mg of
alum-adsorbed MAP, with or without 50 or 100 mg of QS-21 as
coadjuvant. Each group included 2–5 individuals of selected DR
or DQ genotypes (DQB1*0603, DRB1*0401, and DRB1*1101),
with the balance of volunteers of random class II haplotypes, ex-
clusive of these 3 genotypes. Groups 1–3 were immunized on days
0, 28, and 56, whereas group 4/5 was immunized on days 0, 28,
and 237. (Results for volunteers in groups 4 and 5 were pooled
because the same dose of vaccine and adjuvant was administered,
separated by 1 week as a safety precaution.) In the cohort B study,
7 volunteers of responder genotypes were immunized on days 0
and 28 with 500 mg of (TIB)4 MAP/alum and 50 mg of QS-21.
Immunogenicity. Serological responses were measured in serum
samples obtained at 2 weeks and 1 month after each immunization
and at sequential times after the third and final dose. Peptide-
specific antibodies were measured by ELISA by use of peroxidase-
labeled antibodies specific for the Fc of IgM or IgG (Cappel, West
Chester, PA). IgG subgroups IgG1–IgG4 were measured with spe-
cific MAbs (Southern Biotechnology, Birmingham, AL).
Cellular responses were measured with Ficoll-purified peripheral
blood leukocytes (PBLs) in a 6-day H-thymidine incorporation
3
proliferation assay. Cultures were incubated with medium or with
10-fold dilutions of MAP or a recombinant P. falciparum CS pro-
tein that lacked only the putative signal and anchor sequences [24].
Positive controls included the recall antigen tetanus toxoid (Wyeth-
Ayerst, Marietta, PA) or mitogens, phytohemagglutinin-P (Difco,
Detroit) and poke weed mitogen (Life Technologies, Grand Island,
NY). Negative controls included an unrelated MAP or a P. vivax
recombinant CS protein [25]. Interleukin (IL)–2 levels in culture
supernatants were measured with an IL-2–dependent T cell bio-
assay [12]. PBLs obtained from the volunteers before immunization
(day 0) did not proliferate or produce IL-2 in response to any of
the malarial antigens.
T cell lines (TCLs) were established from Cohort B PBL obtained
14 days after the second immunization. The PBL were cultured for
1 week with 4 mM (T1B)4 MAP, followed by the addition of fresh
medium containing IL-2 (100 U/mL) at 3–4 day intervals. After
∼3 weeks’ growth in vitro, the TCLs were assayed for proliferation,
in response to stimulation with 4 mM (T1B)4 MAP, using irradiated
autologous PBL as antigen-presenting cells. The amount of IL-2
in 24-h culture supernatants was measured by bioassay. IFN-g was
measured in 48-h culture supernatants by ELISA (R&D Systems,
Minneapolis).
ELISAs were carried out by use of 96-well plates coated with
(T1B)4 MAP, as described elsewhere [13, 14]. The end point was
defined as the dilution of serum samples giving an optical density
Statistical methods. The primary analysis compared the anti-
body response of individuals expressing responder genotypes with

JID 2000;182 (November)
Phase 1 Trial of Malaria Peptide Vaccine
1489
Figure 2. Correlation of high antibody responder phenotype with class II genotypes. IgG geometric mean titers were measured by multiple
antigen peptide (MAP) ELISA (solid bars) or indirect immunofluorescence assay with Plasmodium falciparum sporozoites (diagonal bars).
Serum samples were obtained after 3 immunizations with 500 mg MAP/alum and 50 mg QS-21 (A); 1000 mg MAP/alum and 50 mg QS-21 (B) ;
1000 mg MAP/alum and 100 mg QS-21 (C); or 500 mg MAP/alum without QS-21 (D). Underlined genotypes—DRB1*0401, DRB1*1101, and
DQB1*0603—encode HLA class II molecules that bind the malaria T1 epitope in vitro.
individuals of random genotypes. Additional comparisons were
made of the antibody response to different concentrations of MAP/
alum with or without QS-21 as coadjuvant. Between-group differ-
ences were assessed by 2-sided t tests after log transformation of
antibody titers. The correlation coefficient was calculated for
ELISA and IFA titers.
volunteers in group 3, administration of the final dose of vac-
cine to the group 4/5 volunteers was delayed. At 7 months after
the second dose, the volunteers tested negative by skin test, and
MAP-specific IgE antibodies were not detectable by ELISA (R.
Edelman, unpublished data). Subsequent administration of the
third dose of 1000 mg MAP/alum and 100 mg QS-21 at this
time did not elicit any adverse local or systemic reactions in
any of the volunteers.
Results
Eosinophilia (1500 eosinophils per microliter of blood) was
not induced by any of the vaccinations. No clinically significant
laboratory changes were noted in any of the immunized
volunteers.
Serologic responses of volunteers immunized with (T1B)4
MAP/alum with or without QS-21: correlation of high antibody
responses with specific class II genotypes. After 3 sc immu-
nizations, antibody levels in the serum samples of the vaccinees
were measured by ELISA, using MAP as antigen, or by IFA,
using P. falciparum sporozoites. High IgG titers were detected
in a number of the volunteers immunized with either 500 or
Safety and reactogenicity of the (T1B)4 MAP/alum with or
without QS-21. The vaccine and adjuvant formulations were
well tolerated at all doses and elicited minimal local and sys-
temic reactions in most volunteers. Local inflammation at the
vaccination site was more frequent in response to higher con-
centrations of the vaccine and adjuvants.
Two of 8 volunteers in group 3 developed local and systemic
immediate hypersensitivity reactions (urticaria) after the third
dose of vaccine. Neither volunteer developed signs of anaphy-
laxis, and the pruritic hives resolved after treatment with oral
antihistamines. Because of the development of urticaria in 2

1490
Nardin et al.
JID 2000;182 (November)
Table 2. Serological and cellular responses in cohort B volunteers.
(T1B)4 MAPc
rPfCS
Responder
genotype
ELISA
GMTa
CSP
Volunteer
titerb
Proliferation
IL-2
Proliferation
IL-2
1
2
3
4
5
6
7
DRB1*0401
DRB1*0401
DRB1*1101
DQB1*0603
DRB1*1101
DRB1*1101
DQB1*0603
8127
10,240
8127
10
10
20
Negative
20
20
5
3.7
3.9
3.8
38.6
36.9
23.8
1.0
90.5
237.9
6.1
24.3
16.9
7.4
3.3
9.7
39.9
27.8
5.8
!80
0.7
3.2
20,480
12,902
10,240
19.9
11.2
21.0
10.3
63.9
0.7
10.2
13.8
NOTE. CSP, circumsporozoite precipitin; GMT, geometric mean titer; IL-2, interleukin-2; MAP,
multiple antigen peptide; rPfCS, recombinant Plasmodium falciparum CS protein.
a
IgG ELISA GMT was determined in serum obtained 14 days after the second dose of vaccine
(day 42), using (T1B)4 MAP as antigen. The corresponding immunofluorescence assay titers against
air-dried P. falciparum sporozoites ranged from 1280 to 20,480, with a GMT of 4580 (data not shown).
b
The CSP titer represents the last dilution of day 42 serum samples that elicited positive CSP
reactions on 2 of 20 viable P. falciparum sporozoites, as detected by phase microscopy.
c
Proliferation shown as stimulation index (SI) obtained after incubation of peripheral blood leu-
kocytes with 4 mM (T1B)4 MAP or 25 mg/mL rPfCS. Control wells incubated with unrelated MAP,
recombinant P. vivax CS protein, or medium gave means of 951 ע 297, 2234 ע 1154, and 942 ע
365 cpm, respectively. IL-2 levels in 72-h culture supernatants are shown as SI of an IL-2–dependent
T cell line.
1000mg of (T1B)4 MAP/alum coadjuvanted with either 50 or
100 mg of QS-21 (figure 2A–2C). These high antibody responses
were dependent on the presence of QS-21 as coadjuvant. Only
1 of 8 vaccinees seroconverted after immunization with 3 doses
of 500 mg (T1B)4 MAP/alum without QS-21 (figure 2D).
The pattern of response in the volunteers immunized with
the (T1B)4 MAP/alum/QS-21 vaccine formulation was bimodal,
consisting of high responders and low or nonresponders. In all
groups, the high-responder phenotype correlated with specific
class II genotypes predicted by in vitro analyses of peptide/
HLA interactions (figure 2A–2C, underlined). Only the volun-
teers who expressed class II molecules encoded by DQB1*0603,
DRB1*0401, or DRB1*1101 developed high anti-peptide and
anti-sporozoite antibody titers. No correlation of antibody re-
sponses with class II DRB3, DRB4, or DRB5 genotypes was
observed in these volunteers.
Importantly, there was an excellent correlation between anti–
repeat antibodies measured by ELISA and reactivity with P.
falciparum sporozoites, as detected by IFA (r p 0.9). A geo-
metric mean titer (GMT) of 10,568 by ELISA and a corre-
sponding IFA GMT of 5511 were obtained in the 11 of 12
volunteers of high-responder genotypes. In contrast, for the 5
of 10 volunteers of random genotypes who seroconverted, the
corresponding GMTs were 206 when measured by ELISA and
80 by IFA. The difference in the IFA and ELISA antibody
titers in volunteers of responder versus random genotypes was
statistically significant (P ! .0001). No significant differencewas
noted in the titers of volunteers of responder genotypes im-
munized with different concentrations of MAP/alum (500 vs.
1000 mg; P p .1) with QS-21 coadjuvant (50 vs. 100 mg; P p
.5).
1:20 (table 2). CSP responses were similar in cohort A vol-
unteers of responder genotypes (data not shown). CSP titers
of this magnitude have not been reported for any previous CS
subunit vaccine. The IFA, ELISA, and CSP titers measured in
serum samples of the MAP vaccinees were comparable to those
obtained in volunteers who were protected after immunization
by exposure to the bites of large numbers of irradiated P. fal-
ciparum–infected mosquitoes [23, 26]. These titers in the MAP-
immunized volunteers were also orders of magnitude higher
than those obtained in a previous phase 1 trial of (NANP)3–
tetanus toxoid, the first generation malaria peptide-protein con-
jugate vaccine [4].
The antibody response to the malaria CS repeats appears to
be controlled by a single autosomal dominant gene. Individ-
uals expressing any 1 of the responder genes, DQB1*0603,
DRB1*0401, or DRB1*1101, were high responders (figure 2A–
2C). The presence of 2 responder alleles, as in volunteer 23
(DRB1*0401 and DQB1*0603) or volunteer 24 (DRB1*0401
and DRB1*1101), did not increase the antibody response above
those observed in the individuals expressing a single responder
gene. It is noteworthy that of the 2 volunteers who were
DRB1*0401 and DRB1*0701 heterozygous, 1 developed the
lowest antibody titer of all the responders (volunteer 21) and
the other (volunteer 28) did not seroconvert.
Kinetics and IgG subgroups of anti-MAP antibody response.
All the volunteers of responder genotypes developed a similar
pattern of IgG subclasses in their anti-MAP antibody re-
sponses. IgG1 and IgG3 predominated after the second dose
of (T1B)4 MAP/alum/QS-21 (figure 3A), which is consistent
with the known adjuvant effect of QS-21 [20, 21]. After the
third dose of vaccine, increased levels of MAP-specific anti-
bodies of IgG2 subtype were also observed, as well as low levels
of IgG4 in some volunteers (figure 3B). This pattern, along
with the development of IgE in the 2 volunteers (volunteers 17
It is of particular importance that the MAP-induced anti-
bodies also reacted at high levels with viable P. falciparum spo-
rozoites, producing CSP reactions at serum dilutions of 1:5 to

JID 2000;182 (November)
Phase 1 Trial of Malaria Peptide Vaccine
1491
Figure 3. IgG subclasses of anti–repeat antibodies in volunteers of responder genotypes. Serum samples were obtained after immunization
with 2 doses (A) or 3 doses (B) of MAP vaccine. Group 3 volunteers (volunteers 17, 20, 23, and 24) received 1000 mg (T1B)4 MAP/alum with
50 mg QS-21, and group 4/5 volunteers (volunteers 25–27 and 31) received the same amount of MAP/alum with 100 mg QS-21. Seropositive
volunteers of random genotypes developed low titer IgG1 and IgG3 antibodies (data not shown). Serum samples of seronegative volunteers did
not give detectable responses with any of the IgG subgroup–specific reagents.
and 24) who developed urticaria (data not shown), suggests
that a shift to Th2 type responses occurred with continued
peptide immunization.
was obtained in 4 of 5 volunteers of responder genotypes, with
reactivity to P. falciparum sporozoites increasing from low or
negative (figure 5, cross bars) to 5120–10,240 IFA titer after
this third dose of vaccine (figure 5, solid bars). The peak an-
amnestic responses in volunteers 25 and 27 were 5–8-fold higher
than titers obtained after the second dose of vaccine, whereas
in the remaining 2 volunteers (volunteers 26 and 31), titers
returned to the peak levels obtained after the second dose. The
ELISA titers directly correlated with IFA titers at all time points
studied (data not shown).
The exception to these strong anamnestic responses was vol-
unteer 28 (DRB1*0401 and DRB1*0701). No significant IFA
(figure 5) or ELISA (data not shown) responses could be de-
tected after the third dose of vaccine. Volunteer 29, of random
haplotype, also remained seronegative after the final
immunization.
Cellular responses in volunteers immunized with MAP/alum/
QS-21. Immunization of a second group of 7 volunteers (co-
hort B), all of whom had responder genotypes, confirmed the
correlation of class II genotypes and high-responder pheno-
types. Six of 7 volunteers developed high levels of anti–repeat
and anti–sporozoite antibodies after 2 injections of the vaccine
(table 2). PBLs of these seropositive volunteers proliferated and
secreted IL-2 in response to in vitro challenge with the (T1B)4
MAP immunogen. Moreover, these immune cells recognized
the repeats in the context of a recombinant P. falciparum CS
protein, raising the expectation that MAP-sensitized T cells
would proliferate and release inhibitory lymphokines, as well
as helper factors, to maintain high levels of antibodies, after
exposure to P. falciparum sporozoites. No proliferation or lym-
The kinetics and persistence of the IgG anti-MAP antibody
response were also similar in all the volunteers, regardless of
the dose of MAP/alum or QS-21. As illustrated by the vol-
unteers in groups 2 and 3, peak IgG antibody responses were
observed 14 days after the third and final dose of vaccine in
responder, as well as random, HLA genotypes (figure 4A, 4B).
By ∼7 months after the final dose of vaccine, the high ELISA
titers gradually decreased to GMT 905 in group 2 and GMT
368 in group 3 volunteers (figure 4A, 4B), with a corresponding
IFA of 160 GMT in each group (data not shown).
Because the final immunization of the group 4/5 volunteers
was delayed due to the development of urticaria in 2 volunteers
in group 3, the effect of prolonging the interval between the
second and third doses of vaccine could be assayed. Of the 5
volunteers of responder genotype in group 4/5, 4 individuals
had IFA antibody titers 11000 after receiving the second dose
of vaccine (figure 5, diagonal bars) and similar levels of anti-
peptide antibodies by ELISA (data not shown). The volunteers
of random haplotypes did not have IgG antibodies detectable
by ELISA or IFA after the second dose of vaccine.
By ∼7 months after the second dose of vaccine, the IgG IFA
titers had declined to !1000 in all volunteers (figure 5, cross
bars). At this time, the volunteers were found to be negative
by skin test, and MAP-specific IgE antibodies were not de-
tectable by ELISA (R. Edelman, unpublished data). The third
dose of vaccine was administered to these volunteers without
any adverse reactions. A strong anamnestic antibody response

1492
Nardin et al.
JID 2000;182 (November)
IgG2 or IgG4, which is consistent with the pattern of IgG sub-
types observed in cohort A volunteers (figure 3).
Discussion
It has not been possible, for the most part, to define the basis
of high and low antibody responders in humans, although these
phenotypes have been noted in the response to other vaccines
[27–29]. Similarly, in individuals living in areas in which malaria
is endemic, antibody responses to P. falciparum CS repeats do
not correlate with class II haplotypes [30, 31]. In these naturally
infected individuals, HLA associations may be obscured by the
presence of multiple T cell epitopes in the CS protein and by
concomitant immune responses to the numerous sporozoite,
liver, and blood stage antigens produced during malaria
infections.
In the present phase 1 studies, we demonstrate that a syn-
thetic malaria MAP vaccine, containing only minimal epitopes
of the P. falciparum CS protein, elicits high antibody responses
in individuals of specific HLA genotypes. Our ability to dem-
onstrate the correlation of responder phenotype with class II
genotype was most likely due to the use of an immunogen of
limited antigenic complexity. The (T1B)4 MAP is composed of
only 5 different amino acids, N, A, P, V, and D, exclusive of
the nonimmunogenic lysine core. The original demonstration
of class II Ir gene control of murine antibody responses was
also dependent on the use of branched peptide polymers con-
taining only 3 or 4 amino acids, such as (T,G)AL and (H,G)AL
[32].
As found in the early murine studies of Ir genes, a single
autosomal dominant gene appears to control the high-re-
sponder phenotype in humans immunized with (T1B)4 MAP.
Volunteers expressing only 1 responder allele developed high
levels of antibody after immunization with 2 or 3 doses of
(T1B)4 MAP/alum/QS-21. Although individuals homozygous
for responder alleles were not available, the presence of 2 re-
sponder alleles in volunteers 23 (DRB1*0401 and DQB1*0603)
and 24 (DRB1*0401 and DRB1*1101) did not result in higher
levels of antibody, compared with individuals with 1 responder
allele (figure 2B).
Although the numbers of vaccinees are small, the results of
our study also suggest that additional genetic factors may play
a role in determining the magnitude of the anti–repeat antibody
response. In cohort A, 2 DRB1*0401-positive volunteers who
were heterozygous for DRB1* 0701, volunteers 21 and 28 (figure
2B, 2C), developed significantly lower antibody responses than
the other responder haplotypes. Similarly, in cohort B, a
DRB1*0701/DQB1*0603 heterozygote (volunteer 4) did not de-
velop detectable levels of antibody after immunization with 2
doses of vaccine (table 2). HLA haplotypes containing the
DRB1*0701 allele have been associated with poor antibody re-
sponses to hepatitis and measles vaccines [28, 29]. Analysis of
suppressor or regulatory T cell responses in the MAP-immunized
Figure 4. Kinetics and persistence of IgG anti–multiple antigen pep-
tide (MAP) antibody response. Antibody titers were measured in vol-
unteers of responder genotypes (closed symbols) or random haplotypes
(open symbols) in group 2 (A) and group 3 (B) . Days on which vol-
unteers were immunized are indicated (arrows).
phokine production could be detected with PBLs obtained from
the seronegative volunteer 4.
The kinetics of the cellular responses paralleled the humoral
responses and increased with each dose of vaccine (figure 6). Low
but positive proliferative responses were detected in 2 of 7 vol-
unteers after the first immunization, and all the seropositive vol-
unteers had detectable cellular responses after a second dose of
vaccine. Moreover, CD4ϩ TCLs could be derived by short-term
peptide expansion of the volunteers’ PBLs, with the exception
of seronegative volunteer 4 (table 3). These MAP-specific TCLs
secreted both IL-2 and IFN-g after peptide stimulation, sug-
gesting a Th1/Th0 response. A Th1-type T helper cell response
in the cohort B volunteers was also suggested by the predomi-
nance of IgG1 and IgG3 anti–repeat antibodies, with little or no

JID 2000;182 (November)
Phase 1 Trial of Malaria Peptide Vaccine
1493
Figure 5. Anamnestic responses in multiple antigen peptide–immunized volunteers. Immunofluorescence assay (IFA) titers of 6 volunteers in
group 4/5 were measured 2 weeks after the second immunization (diagonal bars), ∼7 months later on the day of the third immunization
(cross bars), and 2 weeks after the third immunization (solid bars). Two seronegative individuals of random HLA haplotypes, volunteers 30 and
32, did not receive the third dose of vaccine.
vaccinees may help define other HLA-associated factors that
regulate antibody responses to P. falciparum CS repeats.
The cohort A study also emphasizes the importance of ad-
juvant formulation, especially for peptide immunogens that
lack the nonspecific costimulatory properties of vaccines based
on intact parasites, recombinant vectors, or complex carrier
proteins. In the absence of the coadjuvant QS-21, minimal or
no ELISA or IFA responses were detected in group 1 volun-
teers, even though 4 of these vaccinees were of responder ge-
notype (figure 2D). Moreover, peptide-specific T cell responses
could not be detected in PBLs of these 4 individuals (data not
shown). In contrast, peptide-specific proliferation and lym-
phokine production was detected in PBLs of all volunteers who
developed high antibody titers after immunization with the
MAP/alum/QS-21 vaccine formulation (table 2). These findings
suggest that one role of the QS-21 coadjuvant is to stimulate
the production of costimulatory molecules, such as IL-12 [33],
to expand the peptide-specific T helper cells to levels sufficient
for optimal antibody production, thus facilitating detection by
in vitro assays. These findings illustrate that even a potent an-
tigen tested in a high -responder genotype may be mistakenly
identified as nonimmunogenic in the absence of the correct
adjuvant formulation.
levels of antibodies reactive with P. falciparum sporozoites. The
magnitude of this response was comparable with that obtained
in human volunteers immunized with large numbers of P. fal-
ciparum sporozoites, administered by multiple exposures to the
bites of irradiated infected mosquitoes [23, 26], which represents
the current “gold standard” for pre-erythrocytic malaria vac-
cine development.
Although it is currently unknown what immune mechanisms
are required for sterile immunity in humans, high antibody
titers alone may be sufficient for protection. The passive trans-
fer of polyclonal anti–repeat antibodies derived from MAP-
immunized rodents can protect naive recipients against spo-
rozoite challenge [6, 7]. Anti-sporozoite immune serum samples
can elicit a precipitin reaction on the surface of the parasite,
termed the CSP reaction, which immobilizes the sporozoite [34],
thereby inhibiting invasion of the hepatocytes that is dependent
on parasite motility [35]. In the current study, positive CSP
Table 3. T cell lines derived from multiple antigen peptide (MAP)–
immunized volunteers.
Volunteer
Responder genotype
Proliferation
IL-2
IFN-g
1
2
3
4
5
6
DRB1*0401
DRB1*0401
DRB1*1101
DQB1*0603
DRB1*1101
DRB1*1101
18.7
52.0
14.7
1.0
23.4
55.9
29.0
79.7
28.2
0.8
41.2
35.5
25.2
57.2
32.6
0
642.4
77.2
Sporozoites circulate in the blood for only a brief period of
time before invading the host hepatocyte, and this limited win-
dow for immunological attack necessitates the presence of high
levels of antibody. The current phase 1 studies demonstrate for
the first time that a synthetic peptide vaccine containing min-
imal CS epitopes, when presented in the correct adjuvant for-
mulation to individuals of appropriate genotype, can elicit high
NOTE. Proliferation of T cell lines stimulated with 4 mM (T1B)4 MAP is
shown as a stimulation index (SI). Interleukin (IL)–2 in culture supernatants was
assayed by bioassay, and results were expressed as SI. The levels of interferon
(IFN)–g (pg/mL) in culture supernatants were measured by ELISA.

1494
Nardin et al.
JID 2000;182 (November)
Figure 6. Kinetics of multiple antigen peptide (MAP)–specific proliferative responses. Proliferation after stimulation with (T1B)4 MAP (4 mM)
measured in peripheral blood leukocytes of cohort B volunteers on day 0 (open bars), 2 weeks after the first immunization (diagonal bars), or 2
weeks after the second immunization (solid bars). Results are shown as D counts per minute (cpm; mean cpm in quadruplicate wells with antigen
minus mean cpm in quadruplicate wells with medium).
reactions correlated with high anti-peptide and anti-sporozoite
titers measured by ELISA and IFA, respectively (table 2).
Alternatively, parasite-specific CD4ϩ T cells could mediate
protection through the release of IFN-g, a potent inhibitor of
the intracellular liver stages of the parasite [11]. In the murine
model, CD4ϩ T cell– and IFN-g–dependent protective immu-
nity against sporozoite-induced malaria can be elicited by MAP
immunization [9, 10]. In the current studies, CD4ϩ T cell lines
derived from PBLs of the MAP-immunized volunteers were
shown to produce IFN-g after MAP stimulation in vitro (table
3). Whether repeat-specific antibodies and/or CD4ϩ T cells are
sufficient to protect against P. falciparum sporozoite challenge
can be determined only by phase 2 studies. However, recent
clinical trials of a hybrid CS recombinant protein, termed
RTS,S, have clearly demonstrated that CS subunit vaccines can
elicit protective immunity that correlates with high levels of
anti-CS antibody and CD4ϩ T cells in the absence of detectable
CD8ϩ T cells [36, 37].
The population frequency of the DQB1*0603, DRB1*0401,
and DRB1*1101 responder genotypes suggests that the (T1B)4
MAP/alum/QS-21vaccine formulation would be highly im-
munogenic in ∼20%–35% of individuals, depending on ethnic
background [19]. However, the design of effective vaccines for
malaria, as well as other pathogens, requires the ability to elicit
immune responses in individuals expressing a broad range of
HLA molecules. To address this limitation of peptide vaccines,
we have used peptide/HLA analyses to identify a CS-derived
T cell epitope that can bind to multiple DR and DQ molecules
in vitro [16]. The incorporation of this epitope into a CS repeat–
based MAP provided T cell help for anti–repeat antibody re-
sponses in 8 of 8 strains of mice [16]. A synthetic peptide con-
taining this malaria universal T cell epitope in combination
with the P. falciparum CS repeats [38] recently has been shown
to elicit cellular and humoral immune responses in volunteers
expressing a broad range of HLA class II genotypes [39].
In contrast to more complex delivery systems such as viral
particles or bacterial vectors, synthetic peptides containing min-
imal epitopes provide a means to define precisely the parasite-
specific immune mechanisms regulating humoral responses,
thus facilitating the rational design of highly immunogenic ma-
laria vaccines. The empirical approach used in this study to
define peptide class II restrictions—that is, human T cell clones
or peptide competition assays—has been advanced by the re-
cent development of more sophisticated, computer-driven al-
gorithms that can predict T cell epitopes capable of interacting
with a broad range of HLA class II molecules [40]. The in vitro
and in vivo correlates defined in the current study raise the
hope that these new methods of predicting highly immunogenic
T cell epitopes in known protective antigens, as well as in the
large number of new proteins that will be identified by genome
sequencing projects, will facilitate the rational design of epitope-
based vaccines for malaria, as well as other pathogens.
Acknowledgments
We thank the volunteers for their participation in the study. We
gratefully acknowledge the technical assistance of Rita Altzsuler and
the clinical assistance of Helen Secrest and JoAnna Becker. We would

JID 2000;182 (November)
Phase 1 Trial of Malaria Peptide Vaccine
1495
an alum-adsorbed synthetic multiple-antigen peptide based on B- and T-
cell epitopes of the Plasmodium falciparum CS protein: possible vaccine
application. Vaccine 1994;12:1012–7.
like to acknowledge Eveline Tierney, National Institutes of Health, for
assistance with regulatory affairs and Lorraine Soisson, US Agency
for International Development, for help in program coordination. We
thank Francesco Sinigaglia and Victor Nussenzweig for critical review
of the manuscript. Alum adsorption and bottling of the (T1B)4 multiple
antigen peptide vaccine were carried out under the supervision of Ken-
neth Eckels at the Walter Reed Army Institute of Research Pilot Vac-
cine Production facility. We thank Oscar Kashala, Aquila Biophar-
maceuticals, for providing QS-21. Plasmodium falciparum (NF54)–
infected mosquitoes were generously provided by the Naval Medical
Research Institute.
16. Calvo-Calle JM, Hammer J, Sinigaglia F, Clavijo P, Moya-Castro ZR, Nardin
EH. Binding of malaria T cell epitopes to DR and DQ molecules in vitro
correlates with immunogenicity in vivo: identification of a universal T cell
epitope in the P. falciparum CS protein. J Immunol 1997;159:1362–73.
17. Sinigaglia F, Guttinger M, Romagnoli P, Takacs B. Malaria antigens and
MHC restriction. Immunol Lett 1990;25:265–70.
18. Hammer J, Bono E, Gallazzi F, Belunis C, Nagy Z, Sinigaglia F. Precise
prediction of major histocompatibility complex class II peptide interaction
based on peptide side chain scanning. J Exp Med 1994;180:2353–8.
19. Kimura A, Sasazuki T. Eleventh international histocompatibility workshop
reference protocol for the HLA DNA-typing technique. In: Tsuji K, Ai-
zawa M, Sasazuki T, eds. HLA 1991 proceedings of the eleventh inter-
national histocompatibility workshop and conference (Yokohama, Japan).
Vol 1. Oxford: Oxford University Press, 1992:397–419.
20. Kitamura K, Livingston PO, Fortunato SR, et al. Serological response pat-
terns of melanoma patients immunized with a GM2 ganglioside conjugate
vaccine. Proc Natl Acad Sci U S A 1995;92:2805–9.
21. Helling F, Zhang S, Shang A, et al. GM2-KLH conjugate vaccine: increased
immunogenicity in melanoma patients after administration with immu-
nological adjuvant QS-21. Cancer Res 1995;55:2783–8.
References
1. Potocnjak P, Yoshida N, Nussenzweig RS, Nussenzweig V. Monovalent frag-
ments (Fab) of monoclonal antibodies to a sporozoite surface antigen (Pb44)
protect mice against malarial infection. J Exp Med 1980;151:1504–13.
2. Nardin EH, Nussenzweig V, Nussenzweig RS, et al. Circumsporozoite pro-
teins of human malaria parasites Plasmodium falciparum and Plasmodium
vivax. J Exp Med 1982;156:20–30.
3. Ak M, Bower J, Hoffman S, et al. Monoclonal antibodies of three different
immunoglobulin G isotypes produced by immunization with a synthetic
peptide or native protein protect mice against challenge with Plasmodium
yoelii sporozoites. Infect Immun 1993;61:2493–7.
22. Vanderberg J, Nussenzweig R, Most H. Protective immunity produced by
the injection of X-irradiated sporozoites of Plasmodium berghei. V. In vitro
effects of immune serum on sporozoites. Mil Med 1969;134:1183–90.
23. Clyde DF. Immunity to falciparum and vivax malaria induced by irradiated
sporozoites: a review of the University of Maryland studies, 1971–75. Bull
World Health Organ 1990;68(Suppl):9–12.
4. Herrington DA, Clyde DF, Losonsky G, et al. Safety and immunogenicity
in man of a synthetic peptide malaria vaccine against Plasmodium falci-
parum sporozoites. Nature 1987;328:257–9.
24. Takacs BJ, Girard M-F. Preparation of clinical grade proteins produced by
recombinant DNA technologies. J Immunol Methods 1991;143:231–40.
25. Herrington DA, Nardin EH, Losonsky G, et al. Safety and immunogenicity
of a recombinant sporozoite malaria vaccine against Plasmodium vivax.
Am J Trop Med Hyg 1991;45:695–701.
26. Herrington D, Davis J, Nardin E, et al. Successful immunization of humans
with irradiated malaria sporozoites: humoral and cellular responses of the
protected individuals. Am J Trop Med Hyg 1991;45:539–47.
5. Tam JP, Clavijo P, Lu Y-A, Nussenzweig V, Nussenzweig RS, Zavala F.
Incorporation of T and B epitopes of the circumsporozoite protein in a
chemically defined synthetic vaccine against malaria. J Exp Med 1990;
171:299–306.
6. Wang R, Charoenvit Y, Corradin G, et al. Induction of protective polyclonal
antibodies by immunization with a Plasmodium yoelii circumsporozoite
protein multiple antigen peptide vaccine. J Immunol 1995;154:2784–93.
7. Reed RC, Louis-Wileman V, Cosmai EV, et al. Multiple antigen constructs
(MACs): induction of sterile immunity against sporozoite stage of rodent
malaria parasites, Plasmodium berghei and Plasmodium yoelii. Vaccine
1997;15:482–8.
27. Weissman JY, Tsuchiyose MM, Tong MJ, Co R, Chin K, Ettenger RB. Lack
of response to recombinant hepatitis B vaccine in nonresponders to the
plasma vaccine. JAMA 1988;260:1734–8.
28. McDermott AB, Zuckerman JN, Sabin CA, Marsh SGE, Madrigal JA. Con-
tribution of human leukocyte antigens to the antibody response to hepa-
titis B vaccination. Tissue Antigens 1997;50:8–14.
29. Poland GA. Immunogenetic mechanisms of antibody response to measles
vaccine: the role of HLA genes. Vaccine 1999;17:1719–25.
30. Graves PM, Bhatia K, Burkot TR, Prasad M, Wirtz RA, Beckers P. Asso-
ciation between HLA type and antibody response to malaria sporozoite
and gametocyte epitopes is not evident in immune Papua New Guineans.
Clin Exp Immunol 1989;78:418–23.
31. Riley EM, Olerup O, Bennett S, et al. MHC and malaria: the relationship
between HLA class II alleles and immune responses to Plasmodium fal-
ciparum. Int Immunol 1992;4:1055–63.
8. Collins WE, Sullivan JS, Morris CL, et al. Protective immunity induced in
squirrel monkeys with a multiple antigen construct against the circumspo-
rozoite protein of Plasmodium vivax. Am J Trop Med Hyg 1997;56:200–10.
9. Migliorini P, Betschart B, Corradin G. Malaria vaccine: immunization of
mice with a synthetic T helper epitope alone leads to protective immunity.
Eur J Immunol 1993;23:582–5.
10. Wang R, Charoenvit Y, Corradin G, de la Vega P, Franke E, Hoffman SL.
Protection against malaria by Plasmodium yoelii sporozoite surfaceprotein
2 linear peptide induction of CD4ϩ T cell– and IFN-g–dependent elimi-
nation of infected hepatocytes. J Immunol 1996;157:4061–7.
11. Ferreira A, Schofield L, Enea V, et al. Inhibition of development of exoery-
throcytic forms of malaria parasites by gamma interferon. Science
1986;232:881–4.
32. McDevitt H, Sella M. Genetic control of the antibody response. II. Further
analysis of the specificity of determinant specific control and genetic anal-
ysis of the response to (H,G)-A-L in CBA and C57 mice. J Exp Med
1967;126:969–78.
12. Nardin EH, Herrington DA, Davis J, et al. Conserved repetitive epitope
recognized by CD4ϩ clones from a malaria-immunized volunteer. Science
1989;246:1603–6.
33. Miller MA, Skeen MJ, Ziegler HK. A synthetic peptide administered with
IL-12 elicits immunity to Listeria monocytogenes. J Immunol 1997;159:
3675–9.
13. Munesinghe DY, Clavijo P, Calvo Calle M, Nussenzweig RS, Nardin E.
Immunogenicity of multiple antigen peptides (MAP) containing T and B
cell epitopes of the repeat region of the P. falciparum circumsporozoite
protein. Eur J Immunol 1991;21:3015–20.
14. Moreno A, Rodriguez R, Oliveira GA, et al. Preclinical evaluation of a
synthetic Plasmodium falciparum MAP malaria vaccine in Aotus monkeys
and mice. Vaccine 1999;18:89–99.
34. Stewart MJ, Nawrot RJ, Schulman S, Vanderberg JP. Plasmodium berghei
sporozoite invasion is blocked in vitro by sporozoite-immobilizing anti-
bodies. Infect Immun 1986;51:859–64.
35. Sultan AA, Thathy V, Frevert U, et al. TRAP is necessary for gliding motility
and infectivity of Plasmodium sporozoites. Cell 1997;90:511–22.
15. de Oliveira G, Clavijo P, Nussenzweig R, Nardin EH. Immunogenicity of

1496
Nardin et al.
JID 2000;182 (November)
36. Stoute JA, Slaoui M, Heppner DG, et al. A preliminary evaluation of a
recombinant circumsporozoite protein vaccine against Plasmodium falci-
parum malaria. N Engl J Med 1997;336:86–91.
37. Lalvani A, Moris P, Voss G, et al. Potent induction of focused Th1-type
cellular and humoral immune responses by RTS,S/SBAS2, a recom-
binant Plasmodium falciparum malaria vaccine. J Infect Dis 1999;
180:1656–64.
chemoselective ligation of repeat B-cell epitopes and a universal T-cell
epitope of CS protein. Vaccine 1998;16:590–9.
39. Nardin EH, Calvo-Calle JM, Oliveira GA, et al. A totally synthetic poly-
oxime malaria vaccine containing Plasmodium falciparum B cell and uni-
versal T cell epitopes elicits immune responses in volunteers of diverse
HLA types. J Immunol 2000 (in press).
40. Sturniolo T, Bono E, Ding J, et al. Generation of tissue-specific and pro-
miscuous HLA ligand databases using DNA microarrays and virtual HLA
class II matrices. Nat Biotechnol 1999;15:555–61.
38. Nardin EH, Calvo-Calle JM, Oliveira GA, et al. Plasmodium falciparum
polyoximes: highly immunogenic synthetic vaccines constructed by