| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Microbiology and the Norris Cotton Cancer Center, Dartmouth Medical School, Lebanon, NH 03756;
Department of Biology, Universitiy of California at San Diego, San Diego, CA 92093
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although three classes of murine retroviruses have been isolated from the LP-BM5 mixture, neither the ecotropic nor mink cell cytopathic focus inducing (MCF) replication-competent viruses, alone or in combination, are responsible for disease induction. Rather, they serve as helper viruses for the transmission of a replication defective virus, which is the proximal agent of MAIDS (2, 5, 6). Large deletions within the pol and env genes of this defective virus result in the production of a single gag protein (7). With the exception of 25 amino acids (AA) in the COOH terminus of p15 and 25 AA in the N terminus of p12, the gag polyprotein encoded by the defective virus shares significant homology with other murine retroviruses.
Several host genes confer resistance to MAIDS (8, 9). Among these, the MHC class I locus is of particular interest since the primary function of the class I molecules is presentation of foreign peptide Ags to CD8+ CTL. The observed linkage between resistance and class I genes suggests that virus-specific CTL may serve to protect mice from MAIDS. In general, susceptibility is associated with homozygosity for H-2 haplotypes b, s, and q, while resistance is associated with haplotypes a and d. Highly resistant A strain (H-2a) mice are rendered susceptible to MAIDS after chronic in vivo depletion of CD8+ T cells (10). In a recent report by Pavlovitch et al. (11), it was demonstrated that the offspring of MAIDS-infected mice were resistant to challenge with the LP-BM5 viral complex, and this protection was dependent on CD8+ T cells. Furthermore, it was demonstrated by Tang et al. (12) that perforin-dependent functions of CD8+ T cells contribute to MAIDS resistance. Finally, we have shown in our laboratory that genetically resistant BALB/cByJ and C57BLKs/J (H-2d) mice, but not the prototypic susceptible C57BL/6 and BALB.B congenic (H-2b) mice, generate gag-specific CTL directed against the ecotropic and defective components of LP-BM5. These gag-specific CTL appeared to be directed against an epitope located in the p30 viral protein, within a region where there is complete amino acid sequence homology between defective and ecotropic viruses (13).
In this report we show that these gag-specific CTL are directed against the SYNTGRFPPL peptide and that this peptide is generated from the ORF2 of both the defective and ecotropic retroviral gag. Our data suggest that during the course of natural virus infection, reading frames other than the primary ORF may be used to encode antigenic epitopes. The implications of alternative reading frame usage can be extended to suggest that other viruses, and perhaps other foreign genes, may express T cell epitopes from more than one reading frame.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
BALB/cByJ mice (5–6 wk old) were obtained from The Jackson Laboratories (Bar Harbor, ME) and maintained at Dartmouth-Hitchcock Medical Center (DHMC) Animal Facilities (Lebanon, NH).
Cell lines
HuTK- 143B (143B) and BHK-21 (cl.13) cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and cultured as specified. P815B cells, a generous gift from Dr. Jack Bennink (National Institute of Allergy and Infectious Diseases (NAIAD)/National Institutes of Health, Bethesda, MD), were maintained in RPMI 1640 containing L-glutamine, penicillin, streptomycin, and 5% FBS. SC.1 and SC.1/BM5 Eco fibroblast cells were maintained in RPMI 1640 containing 0.1 M nonessential amino acids, L-glutamine, penicillin, streptomycin, and 5% FBS. The 1710.5 clone, a MAIDS B cell lymphoma that expresses both defective and ecotropic gag, the generous gift of Dr. Herbert Morse (National Institutes of Health , Bethesda, MD), was maintained in RPMI 1640 containing 0.1 M nonessential amino acids, L-glutamine, penicillin, streptomycin, and 5% FBS.
P815B and 1710.5 MAIDS tumor fusions
The P2.A4 clone and other clones were derived from treatment of the P815B and 1710.5 cell lines with PEG 1500, under conditions promoting cell:cell fusions. Two-color staining and single cell sorting allowed for the cloning of the P2.A4 and other clones. All clones were maintained in RPMI 1640 containing L-glutamine, penicillin, streptomycin, and 5% FBS.
Recombinant viruses
Kd-Vac, which encodes the H-2Kd class I molecules, was a generous gift from Dr. Jack Bennink (NIAID/National Institutes of Health, Bethesda, MD). Additional recombinant vaccinia and Sindbis viruses were generated using established protocols (14, 15). Intermediate shuttle vectors, used to create the indicated recombinant viruses described below, were constructed using established procedures (16) or manufacturer-specified protocols. The pSC65 shuttle vector and the Western Reserve isolate of vaccinia virus were generously provided by Dr. Bernard Moss (NIAID/National Institutes of Health, Bethesda MD). The vectors for recombinant Sindbis virus generation, pH3'2J1 (shuttle vector) and pTE3'2J:CAT, were generously provided by Dr. Chang Hahn (University of Virginia, Charlottesville, VA). Plasmid maps will be provided on request.
(Recombinant) viral defective gag expression systems: vaccinia
viruses.
Inserts were cloned into the shuttle vector pSC65, and recombinant
viruses were generated following homologous recombination between pSC65
and the Western Reserve isolate of vaccinia virus. Construction of the
shuttle vectors used to create the indicated recombinant viruses is
detailed below. A plasmid containing a permuted clone of the
LP-BM5-defective (p1/27/A1) retrovirus (generous gift
from Dr. Sisir Chattopadhyay, NIAID/National Institutes of Health,
Bethesda MD) (7) was linearized with HindIII. This linear
DNA served as a template for PCR using oligonucleotides DS02 and DS03
as primers (see below). Amplified products were digested with
SalI and KpnI and subcloned into the polylinker
of pSC65. This vector will be referred to as pSC65.DG2 (defective gag).
1) dG(1–187)-Vac: pSC65.DG2 was digested with KpnI and
NcoI, and ss overhangs were removed with T4 DNA polymerase.
The resulting shuttle vector, which retained nucleotides encoding amino
acids 1–187 of the gag polyprotein, was religated. 2) dG(208–250)-Vac
and dG(305–338)-Vac (see Fig. 1
): Linear
p1/27/A1 (see above) served as a template for PCR using
DS14 and DS16 (dG(208–250)) or DS15 and DS17 (dG(305–338)) as
primers. Amplified products were digested with SalI and
BglII and subcloned into the polylinker of pSC65.
|
Oligonucleotides
Oligonucleotides were designed based on published defective gag sequences (Def27 clone)(7). Lower case text indicates nucleotides that are not identical to the published sequences and, in certain cases (shown in italics), generate restriction endonuclease recognition sites or complementary termini. Translation initiation and termination codons (in both polarities) are shown in bold text.
Oligonucleotides synthesized by the University of Pennsylvania: DS02 (5'-GTTTgtcgacATATGGGACAGACCATAACC-3'); DS03 (5'-ACCCggtaccTCCCTAGTCACCTAAGG-3').
Oligonucleotides synthesized by the Molecular Genetics Center of Dartmouth Medical School: DS21 (5'-ctagccatggggCAATACTGGCCGTTTTCCTCCTCTGATCTAtag-3'); DS22 (5'-gatcctaTAGATCAGAGGAGGAAAACGGCCAGTATTGccccatgg-3'); DS23 (5'-ctagccatggggCCTTCCTTTTCTGAAGATCCAGGTAAATTGACCGCCTTAATTtag-3'); DS24(5'-gatcctaAATTAAGGCGGTCAATTTACCTGGATCTTCAGAAAAGGAAGGccccatgg-3'); DS25 (5'-ctagccatgGCTTTCCCACTCCGTTTGGGGGGTAATGGTCAGTTAtag-3'); DS26 (5'-gatcctaTAACTGACCATTACCCCCCAAACGGAGTGGGAAAGCcatgg-3'); DS27 (5'-ctagccatgGGGCTATATAACTGGAAAAATAACCCTTCCTTTTCtag-3'); DS28 (5'-gatcctaAGAAAAGGAAGGGTTATTATTTTTCCAGTTATATAGCCCcatgg-3').
Oligonucleotides synthesized by Research Genetics: DS31 (5'-TTTTTCCAGTACTAGTGATCAGAGA-3'); DS32 (5'-AAGGAAGGGTACTAGTTTTTCCAGTT-3').
Oligonucleotides synthesized by Integrated DNA Technologies, Inc.: SMM9 (5'-gggccctctagacatgGGTAAcGGTCAGTTACAATACTGGCCGTTTTCCTCCTCTGATCTATATggatccgggccc-3'); SMM10 (5'-gggcccggatccATATAGATCAGAGGAGGAAAACGGCCAGTATTGTAACTGACCgTTACCcattctagagggccc-3'); SMM11 (5'-gggccctctagataGGGTAATGGTCAGTTACAATACTGGCCGTTTTCCTCCTCTGATCTATATggatccgggccc-3'); SMM12 (5'-gggccctggatccATATAGATCAGAGGAGGAAAACGGCCAGTATTGTAACTGACCATTACCCtatctagagggccc-3').
Synthetic peptides
Synthetic peptides prepared at Dartmouth Medical School (Hanover, NH) were: dG:220–229 (QYWPFSSSDL); dG:238–246 (SFSEDPGKL); dG:216–226 (NGQLQYWPFSS); dG:225–235 (SSSDLYNWKNN); and ORF2 peptide VSYNTGRFPPLI.
Synthetic peptides prepared at National Institutes of Health, Bethesda, MD (generous gifts of Jon Yewdell) were: dG:220–230 (QYWPFSSSDLY), dG:220–231 (QYWPFSSSDLYN), and dG:219–230 (LQYWPFSSSDLY).
Peptides prepared by Research Genetics (Huntsville, AL) were: dG:208–225 (AFPLRLGGNGQLQYWPFS); DG:217–234 (GQLQYWPFSSSDLYNWKN); dG:226–243 (SSDLYNWKNNNPSFSEDP); dG:235–250 (NNPSFSEDPGKLTALI); dG:208–216 (AFPLRLGGN); dG:223–231 (PFSSSDLN); dG:229–236 (LYNWKNNN), dG:229–237 (LYNWKNNNP), and ORF2 peptides SYNTGRFLPPL and SYNTGRFPPLI. Peptides were usually synthesized with unmodified NH2- and COOH-terminal amino acids using fluorenylmethoxycarbonyl (FMOC)-based chemistry. One exception is dG:220–229, which was synthesized both plus and minus an acyl group at the N terminus. The peptide concentration range tested was from 1 fM to 10 mM.
Generation of bulk cytolytic effector cells
Six- to ten-week-old mice were infected i.v. with 107 pfu dG-Vac. At least 3 weeks postinfection, dG-Vac-primed mice were killed by cervical dislocation and splenic leukocytes recovered. An amount equal to 1 x 106 P815B cells were irradiated (8,000 rad), infected with the appropriate recombinant virus(es) (see below), and added to 4 to 5 x 107 primed leukocytes to obtain gag-specific CTL. Vaccinia virus-specific effectors were generated by adding vaccinia virus directly to responder cells at a multiplicity of infection (MOI) of 10:1. Peptide stimulation was achieved by the addition of synthetic peptides (1x PBS) at a final concentration of 10 µg/ml. Cells were cultured for 6 days in a humidified 5% CO2 atmosphere in 25-cm2 flasks containing 10 ml sensitization media (SM) (RPMI 1640, 100 µM NEAA, 1 mM sodium pyruvate, 50 µM 2-ME, and 10% FBS). Gag-specific CTL lines (1229C; 33:229) and CTL clones (D7) were established via biweekly stimulation and standard limiting dilution analysis, respectively. Established gag-specific CTL lines and clones were maintained in a humidified 5% CO2 atmosphere in 25-cm2 flasks containing 10 ml of SM media with IL-2 (9 U/ml-CETUS).
Infection of stimulator and target cells
An amount equal to 1 x 107 (vaccinia virus) or 5 x 107 (Sindbis virus) pfu were added to 1 x 106 cells in 500 µl of a balanced salt solution containing 0.1% BSA and allowed to adsorb for 1 h at 37°C with agitation. For stimulator cells, samples were washed with 10 ml SM before adding responding lymphocytes. For vaccinia target cells, samples were resuspended in 1 ml SM or P815B media, transferred to a 24-well, flat-bottom tissue culture plate, and incubated at 37°C for an additional 3 h before 51Cr labeling. Sindbis targets were plated in the same fashion but were allowed to express viral proteins for a minimum of 6 h before 51Cr labeling. Occasionally the 51Cr labeling was done simultaneously with the virus expression.
Chromium release assay
Target cells (1–2 x 106) were resuspended in 100 µl FBS and labeled with 200 µCi [51Cr]sodium chromate (New England Nuclear, Wilmington, DE) for 45 min at 37°C. After two washes with RPMI, 4 x 103 or 104 target cells were combined with serial dilutions of effector cells in 96-well V-bottom plates in a final volume of 200 µl. Following a 4-h (tumor targets) or 6-h (fibroblast targets) incubation at 37°C, 100 µl of cell-free supernatant was collected and counted. Percent specific lysis of tumor cells was determined using the formula ((a-b)/c) x 100, where a is experimental cpm released by target cells incubated with effector cells, b is cpm released by target cells incubated alone (spontaneous release), and c is cpm released by the freeze-thaw of target cells (approximately 80% of total cpm incorporated). The level of spontaneous release of target cells alone was never above 20% and <10% variation was seen between replicates.
RT-PCR
Total RNA was recovered using a CsTFA-based kit from Pharmacia (Piscataway, NJ), and 1 µg was resuspended in 20 µl of a buffered solution containing 100ng d(T)15 primer, 10 units RNase inhibitor, 1 mM each dATP, dCTP, dGTP, dTTP, and 40 units avian myeloblastosis virus reverse transcriptase. Samples were incubated for 60 min at 42°C and the reaction stopped with 100 mM EDTA. Following ethanol precipitation, each sample was resuspended in 100 µl of a buffered solution containing 200 µM each dATP, dCTP, dGTP, dTTP, 1 µM of each oligonucleotide primer, and 2.5U Taq polymerase. Each sample was then subjected to 30 cycles of amplification consisting of a 45 s denaturation at 94°C, a 60-s annealing at 50°C, and a 90-s chain elongation at 72°C.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It was previously demonstrated that a population of gag-specific CD8+ CTL generated from the MAIDS-resistant BALB/cByJ strain recognizes a Kd-presented epitope located within the gag protein of both the ecotropic and defective viruses of the LP-BM5 retroviral complex (13). BALB/cByJ mice primed with a recombinant vaccinia virus expressing the entire defective gag (def-gag) protein (dG-Vac), and restimulated with either a recombinant vaccinia virus expressing defective gag amino acids (AA) 1–342 or AA 188–532, induced CTL that lysed target cells infected with the entire def-gag protein (AA 1–532) with similar efficiencies (13). Although these data suggested that a major CTL epitope resides between AA 188 and 342, the possibility that each gag recombinant contained one of two dominant epitopes could not be excluded. To distinguish between these two possibilities, a panel of recombinant viruses expressing truncated versions of the def-gag polyprotein were generated and tested against polyclonal, gag-specific CTL.
As expected (Fig. 2A, Expt.
1), Kd-Vac-infected, human 143B target cells coinfected
with recombinant vectors expressing the entire defective gag (AA
1–532), gag AA 1–342, or gag AA 188–532 were lysed to similar levels
(83%, 68%, or 77%, respectively) when assayed with def-gag AA
1–342-stimulated effectors. Because secondary stimulation occurred in
the absence of AA 343–532, the carboxyl-terminal boundary for the CTL
epitope appeared to be AA residue 342. Likewise, target cells infected
with either recombinant vectors expressing the entire def-gag AA
1–532, or the overlapping recombinant vectors, def-gag AA 1–342 or AA
188–532, were lysed to similar levels (86%, 63%, or 74%,
respectively) when assayed with def-gag AA 188–532-stimulated
effectors, suggesting the amino-terminal boundary must be AA residue
188. Finally, although lysed by vaccinia virus-specific effectors (81%
lysis), indicating successful vaccinia infection of target cells,
gag-specific effectors failed to kill target cells infected with a
recombinant vector expressing def-gag AA 1–187. Collectively, these
data confirmed that a major CTL epitope(s) is located between AA
188–342 of the defective gag protein.
|
MHC class I allele-specific peptide motifs have been defined
through the comparison of various CTL minimal epitopes (17), amino acid
substitution studies (18), and microsequencing of acid-eluted peptides
(19). Thus, nonameric peptides presented by the H-2Kd
class I molecule preferentially contain a tyrosine (Y), or to a lesser
extent phenylalanine (F), "anchor" at position 2, and valine (V),
isoleucine (I), threonine (T), alanine (A), or leucine (L) at the
carboxyl terminus (20). To identify the location(s) of the dominant
epitope(s) for gag-specific CTL, the amino acid sequence of gag was
scanned for the presence of potential position 2 anchors (Y or F) (Fig. 1
, Materials and Methods). The Y/F residues of gag AA
188–342 were clustered so that all potential anchors could be
encompassed by two minigene constructs that encoded AA 208–250 or AA
305–338 of def-gag. The analogous Sindbis or vaccinia virus
recombinants that were constructed shared the following features: 1) at
the amino-terminus of each construct, the Y or F anchor (position 2)
was preceded by two amino acids: position 1 representing the natural
amino acid (for example, alanine at 208) and position -1 containing an
added methionine to initiate translation of the minigene product, and
2) any potential carboxyl-terminal anchor residue up to position 13
(relative to Y or F at position 2) was included. While the
carboxyl-terminal anchor is often located at the ninth residue for
Kd (relative to Y or F anchor at position two), there have
been reports of CTL epitopes longer than minimal size (21, 22).
The data of Figure 2A, Expt. 2 demonstrated that def-gag AA
208–250-infected P815B target cells were efficiently lysed when
assayed against either def-gag AA 1–342 or def-gag AA 188–532
restimulated, gag-specific CTL. That def-gag AA 305–338-infected
target cells were as efficiently lysed by vaccinia virus-specific
effectors as those infected with def-gag AA 208–250 clearly indicated
that these target cells were infected by and expressing each of these
recombinant vaccinia viruses. Reverse PCR was employed to verify
expression of the appropriately sized minigene message (Fig. 2B), and maintenance of an open reading frame was
confirmed by sequencing the viral construct (data not shown). These
data collectively suggested that the dominant epitope(s) for
gag-specific CTL resides between AA 208–250.
To facilitate the fine mapping of the CTL epitope(s), CTL lines (1229C
and 33:229) and clones (D7 and H9) were established. These CTL (lines
and clones) were gag specific, and did not cross-react with uninfected
P815B (H-2d) target cells. Consistent with the data for
bulk effectors, all of these cytolytic effectors, also of BALB/cByJ
origin, recognized an H-2Kd-restricted epitope(s) residing
between amino acids 208 and 250 of the defective gag polypeptide (data
not shown). The CTL lines expressed 
ß TCRs, the variable regions
of which were predominantly, if not exclusively, derived from the
Vß8 gene family. While both CTL lines and clones were
indistinguishable in terms of Ag specificity and phenotype
(CD3+, CD8+, CD4-,
ß+ TCR), the D7 clone and 33:229 line were utilized
most extensively because of their rapid growth in vitro.
ORF1 of defective and ecotropic gag does not contain an epitope recognized by gag-specific CTL
To define the minimal CTL epitope(s), two synthetic peptides
(QYWPFSSSDL and SFSEDPGKL) were generated that were located between AA
208 and 250 (Fig. 1, Materials and Methods), were of
the appropriate length, and incorporated the appropriate anchor
residues necessary for H-2Kd presentation (20). These
peptides failed to sensitize P815B target cells for recognition by
gag-specific CTL (bulk cultures and 1229C line; Table I, lines 4 and 11). Additional synthetic
peptides from defective gag AA 208–250, of "minimal" length (9–12
AA), and longer peptides (16- and 18-mers) that collectively spanned
amino acids 208–250, and overlapped by nine residues each, were tested
under a variety of experimental conditions but also failed to sensitize
P815B target cells for lysis by gag-specific CTL (1229C and 33:229
lines or the D7 clone) (Table I). To test the hypothesis that the
endogenous peptide epitope might be subject to post-translational
modification before presentation on the cell surface such that
unmodified, synthetic peptides could not be recognized, recombinant
Sindbis viruses were constructed that expressed def-gag AA 208–219, AA
220–229, AA 229–240, and AA 237–250. As shown in Table I, none of
the four recombinant viruses sensitized P815B target cells for lysis by
either gag-specific CTL lines (1229C and 33:229) or a clone
(D7).
|
), a
systematic, genetic approach was undertaken. To define the carboxyl
terminus of the defective gag CTL epitope, sequential truncations of
the def-gag AA 208–250 region were made, resulting in the generation
of four recombinant constructs encoding def-gag AA 208–240, AA
208–235, AA 208–230, and AA 208–225. The lack of gag-specific CTL
recognition of P815B targets infected with def-gag AA 208–225, and the
positive recognition of targets infected with def-gag AA 208–230 and
all larger constructs (Fig. 3), strongly
suggested that a critical carboxyl-terminal residue is located within
AA 226–230, inclusive.
|
Examination of the ORF1 AA sequence from the 208–230 region
suggested that no likely Kd-presented epitopes existed
beyond those already tested. However, analysis of the defective gag
nucleotide sequence revealed that a peptide (SYNTGRFPPLI), which
fulfilled the Kd-allele-specific peptide motif, was located
in an alternative reading frame (ORF2), with a potential initiating
methionine located two AA upstream (Fig. 4). To test the possibility that the CTL
epitope was located in this putative second reading frame, the 12-AA
peptide VSYNTGRFPPLI, containing the potential 10-mer or 11-mer epitope
plus the N-terminal valine located between the Met and position 1 of
the putative epitope, was synthesized and tested for its ability to
sensitize target cells for lysis. As shown in
Figure 5A, this 12-AA peptide
located in ORF2 (VSYNTGRFPPLI) sensitized P815B target cells for
CTL lysis by gag-specific polyclonal CTL. In the same experiment,
a representative peptide from ORF1 (208–230 region) fulfilling the
Kd-motif (QYWPFSSSDL) was not recognized, suggesting
that the lack of recognition observed by the previously tested ORF1
peptides was not the result of the experimental system (Fig. 5A). These results with polyclonal CTL were obtained
in 5/5 experiments, and similar results were obtained when the
VSYNTGRFPPLI peptide was tested using the gag-specific 33:229 CTL line
and the D7 CTL clone in 7/7 experiments (data not shown). To determine
the "minimal" peptide required to sensitize target cell lysis, the
C-terminal 11-mer and 10-mer related synthetic peptides, SYNTGRFPPLI
and SYNTGRFPPL, were similarly tested. Figure 5B (inset)
shows that the 10-mer, SYNTGRFPPL, sensitized target cell lysis at a
level comparable (57%) to that seen when targets infected with
recombinant Sindbis virus encoding the entire defective gag protein (AA
1–532) were tested (55%) against the D7 CTL clone. Because in
head-to-head experiments the synthetic 10-,11-, and 12-mer peptides
from ORF2 were recognized equally well (data not shown), these results
suggested that the total antigenic capacity of the defective gag
protein can be accounted for by the 10-mer, SYNTGRFPPL, located in
ORF2. Data from a peptide titration experiment confirmed that varying
the concentration of SYNTGRFPPL peptide resulted in a target cell
sensitization pattern similar to those published for "authentic"
(ORF1) antigenic peptides (Fig. 5B).
|
|
The ORF2-derived SYNTGRFPPL peptide accounts for the total CTL reactivity of the defective gag protein
Sindbis virus minigene constructs were produced in
which the defective gag coding sequence of either ORF1 or ORF2, within
the defined region of epitope expression, was expressed independently.
When the ATG encoding for ORF1 proteins was deleted and a stop codon
introduced, yielding a Sindbis virus encoding only alternative ORF
peptide(s), there was normal gag-specific CTL recognition of infected
P815B target cells (Fig. 6).
Alternatively, infection of P815B cells with a recombinant Sindbis
virus containing a mutation of the ATG for ORF2 resulted in a complete
lack of recognition by gag-specific CTL (Fig. 6). Nucleotide changes
were designed as conservative mutations; hence, the ORF1 "knockout"
construct (SIN:ORF1 KO) maintained an intact ORF2 initiation methionine
and coding sequence, while the ORF2 "knockout" construct (SIN:ORF2
KO) maintained an intact ORF1 initiation methionine and coding
sequence. RT-PCR was employed to demonstrate that P815B target cells
infected with these two recombinant Sindbis viruses produced relatively
equivalent message levels of the expected size (data not shown). Taken
together, these data confirmed the conclusions that the immunodominant
epitope residing within the defective gag protein is the ORF2-derived
SYNTGRFPPL peptide, and that the naturally occurring, initiation Met
(ATG) located at nucleotide 756 in ORF2 is required for expression of
SYNTGRFPPL, at least in the context of the minigene constructs used in
Figure 6.
|
To address the generation of the ORF2 peptide, SYNTGRFPPL, during
the course of normal retroviral gene expression, SC.1 fibroblast cells
infected with the MAIDS-associated BM5 ecotropic helper virus and thus
expressing gag (SC.1/BM5 Eco) were employed. Although readily
infectable by murine retroviruses, the SC.1 line is of feral mouse
origin, and SC.1/BM5 Eco cells had to be infected with vaccinia
constructs encoding the Kd-molecule
(Kd-Vac) for use as target cells. As mentioned above, the
gag gene nucleotide sequences of the defective and BM5 Eco
retroviruses are identical within the region containing the ORF2 CTL
epitope. This complete identity suggested that BM5 Eco-derived
SYNTGRFPPL should be recognized by gag-specific CTL, if this ORF2
peptide is generated. In duplicate experiments (where the efficiency of
Kd expression was sufficient to provide appropriate
positive control lysis, see below) Kd-Vac infection of
SC.1/BM5 Eco targets resulted in substantial levels of lysis by
gag-specific CTL. This level of CTL recognition was striking compared
with the essentially negligible levels of lysis of SC.1 target cells
expressing only Kd (no BM5 Eco virus) or 65-Vac infected
SC.1/BM5 Eco targets (not expressing Kd), as negative
controls (see representative experiment, Fig. 7) or other negative controls (legend to
Fig. 7). Based on the levels of lysis by the anti-gag CTL, the
amount of endogenously synthesized ORF2 peptide in the SC.1/BM5 Eco
(and Kd-Vac)-infected cells available to sensitize targets
for lysis was estimated to be approximately 40 to 50% compared with
that of the positive controls (i.e., SC.1 and SC.1/BM5 cells, both
Kd-infected and exogenously pulsed with ORF2
peptide-SYNTGRFPPL, Fig. 7). By comparison of the lysis of these
Kd-Vac-infected fibroblast targets and P815B targets
(endogenously expressing Kd), both pulsed with synthetic
SYNTGRFPPL as positive controls, it appeared that in this system lysis
of all types of SC.1 targets was limited by the number of
Kd molecules available (via Kd-Vac infection)
to present the ORF2 peptide. Apparently due to this constraint, three
additional experiments demonstrated a similar trend of recognition of
Kd-Vac-infected, BM5 Eco gag-expressing fibroblasts
(SC.1/BM5 Eco), although both the positive control and experimental
fibroblast target types had lower levels of lysis.
|
).
|
To determine whether the ORF2-encoded CTL determinant was expressed in vivo, we immunized MAIDS-resistant BALB/cByJ mice with the LP-BM5 virus complex and assessed for the ability of infection to prime for an anti-gag CTL response. Injection (i.p.) of approximately 1.8 x 105 pfu of the LP-BM5 complex primed efficiently for subsequent in vitro restimulation and induction of anti-gag, SYNTGRFPPL-specific CTL. Primed spleen cells harvested at short intervals (7–10 days) after in vivo infection demonstrated particularly high levels of SYNTGRFPPL-specific CTL induction. Two representative experiments (Tables II and III) clearly demonstrated that in the absence of recombinant vectors, in vivo LP-BM5 retrovirus infection led to the production of the immunogenic, ORF2-derived SYNTGRFPPL epitope, and this immunodominant gag-derived peptide was able to prime SYNTGRFPPL-specific precursor CTL. Furthermore, in vitro restimulation was accomplished only with responder spleen cells from LP-BM5 primed, not unimmunized, mice by addition of stimulator cells presenting either minimal or extended portions of recombinant Sindbis-encoded gag and by stimulator cells presenting minimal synthetic peptide. These experiments were done a total of five times with a similar pattern of results.
Collectively, these results clearly demonstrated that, during the course of a normal retroviral infection, there is expression of a major physiologic CTL epitope located in ORF2 of the gag gene of both the defective and ecotropic MAIDS retroviruses, namely SYNTGRFPPL.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). We have concluded
that the epitope recognized by gag-specific CTL does not reside in ORF1
because the ORF2-derived SYNTGRFPPL peptide readily sensitizes P815B
target cells for destruction by gag-specific polyclonal CTL, CTL lines,
and CTL clones, at relatively low peptide concentrations, with
half-maximal lysis for the CTL clones in the pM range (Fig. 5). In
contrast, many ORF1 (208–250)-encoded peptides, including ones
fulfilling the Kd allele-specific peptide motif, failed to
sensitize target cells for CTL lysis, even at high concentrations
(Table I and Fig. 5). Further evidence that the SYNTGRFPPL peptide is a
bona fide epitope, not cross-reactive with an epitope that resides in
the reading frame encoding the viral structural proteins, comes from
the use of Sindbis virus constructs in which the coding sequence of
either ORF1 or ORF2 was expressed independently. These data
demonstrated that anti-gag reactivity could be accounted for by the
expression of the ORF2-derived SYNTGRFPPL peptide in the absence of any
ORF1-derived peptides (Fig. 6). In addition, P815B targets cells
infected with a Sindbis or vaccinia viral recombinant encoding the
AKR623 gag protein (from the related prototypic ecotropic AKR623
retrovirus) were not recognized by the gag-specific 1229 line and
clonal D7 CTL in two experiments (data not shown). Because the AKR623
gag protein has 100% amino acid homology (23) to both defective and
ecotropic gag (Fig. 4), within the relevant ORF1 region (208–230; Fig. 3), any potential ORF1 epitope(s) would have been shared among these
viruses and conferred sensitivity to the anti-gag CTL. Examination
of ORF2 of the AKR623 gag gene, within the epitope-defined
region, revealed a similar, but not identical, peptide (SCSTGRFPPL)
might be encoded (relative to SYNTGRFPPL of defective and BM5 ecotropic
gag). Although the change to serine at position 3 might also be
important, the lack of CTL recognition of P815B target cells expressing
the AKR623 gag protein is most likely due to the cysteine rather than
tyrosine at anchor position 2, which is known to be important for
presentation by the Kd molecule.
Based on our data and recent reports by other labs, we propose that the
generation of antigenic peptides from ORFs other than the primary ORF
is an important phenomenon which to date has gone largely unexplored.
Hence, the usage of all reading frames to generate antigenic peptides
may substantially increase the number of possible epitopes. In the one
other, to our knowledge, study of a "naturally" occurring epitope,
Wang et al. (24) demonstrated that a human cancer T cell determinant
(MSLQRQFLR) was generated by the use of an alternate open reading frame
of the TRP-1 (gp75) gene. Using the TIL586
(tumor-infiltrating lymphocyte) line, this study demonstrated, via
granulocyte-macrophage (GM)-CSF release and cytolytic assays, that a
human tumor rejection Ag can be generated from a normal cellular gene
using an open reading frame other than that which encodes the
ORF1-derived gp75 protein. Relative to the possibility that the tumor
cell may represent a unique situation where certain genes are
overexpressed and/or show dysregulated expression, we have extended
this alternate reading frame theory to include a retrovirus-encoded CTL
determinant. In this report we demonstrate that during the course of a
normal retroviral infection, an antigenic peptide can be generated via
a novel, but as yet undefined, mechanism by which overlapping
ATG-defined open reading frames may result in the generation of
structural virion proteins from ORF1 and an antigenic peptide from
ORF2. In another report, it has been suggested that a non-ATG-defined
translational reading frame contains a "cryptic" translation
product that can also serve as a source of an antigenic peptide (25).
One caveat of this report was that the cryptic translation product,
JAL8-encoded epitope (SVVEFSSL), was only partially derived from the
cDNA insert (-tubulin gene). Rather, this CTL
neodeterminant spanned the junction of the vector (including the entire
5' linker nucleotide sequence) and cDNA insert of ORF3 of a construct
encoding the -tubulin gene. The inclusion of vector and
linker sequences in the epitope suggested that JAL8 is not a
physiologically derived epitope. Indeed, in a follow-up report by
Malarkannan et al. (26), it was demonstrated that the SLVELTSL (SEL8)
peptide, derived from the primary open reading frame of the murine
adenosine phosphoribosyl transferase (APRT)
gene, shared five of eight residues with the previously identified JAL8
peptide, stimulated the same T cell population, and thus probably
represents the naturally occurring epitope.
In this report we clearly demonstrate that the entire antigenic epitope
is contained within the retroviral cDNA sequence insert, suggesting
that this is an authentic epitope, and that the use of viral vectors in
our experimental system did not contribute sequence information to the
generation of the SYNTGRFPPL peptide. Three lines of evidence support
the conclusion that during the course of a normal retroviral infection
(i.e., one in which the SYNTGRFPPL peptide is not generated in the
context of an expression vector) Kd- and gag-positive
targets express sufficient levels of SYNTGRFPPL to be lysed by
gag-specific CTL. First, SC.1 fibroblast cells expressing the BM5
ecotropic gag protein, with 100% amino acid homology to the def-gag
protein within the region of the CTL epitope, are recognized and lysed
by gag-specific CTL when the Kd molecule is delivered via
vaccinia virus expression (Fig. 7). Second, expression of the
SYNTGRFPPL epitope in 1710.5 (H-2b) MAIDS lymphoma cells
was assessed by the approach of fusion with P815B tumor cells to
provide the required Kd restriction element and was
demonstrated by the recognition of two resulting clones (including
P2.A4, Fig. 8) by gag-specific CTL. Third, spleen cells from
MAIDS-resistant BALB/cByJ mice primed in vivo with the LP-BM5
retroviral complex, and restimulated in vitro with recombinant Sindbis
or vaccinia virus-expressing def-gag protein or SYNTGRFPPL peptide,
generated CTL specific for defective gag and, more specifically,
ORF2-derived SYNTGRFPPL (Tables II and III).
The thesis here that, in addition to the primary ORF, alternate reading frames can be used to generate antigenic peptides may rely on the documented high level of TCR sensitivity. This sensitivity, in turn apparently due to the serial-engagement model of TCR-mediated activation (27, 28) allows for the recognition of "rare" peptide/MHC complexes, perhaps present at low to very low frequencies, relative to epitopes derived from the ORF1-encoded structural gene products. Although many studies have shown that CTL with TCR of average affinity require 100 to 300 epitopes per target cell for lysis to be triggered, two independent studies have shown that T cell activation can be achieved by fewer than 10 complexes per target cell (29), or with as few as 3 to 5 peptide/class I complexes per target cell (30). Additionally, the recent publication of Sykulev et al. (31) suggested that as few as 1 to 3 peptide-MHC complexes on the surface of a target cell is sufficient to trigger activation of a T cell response. Because of this low threshold, due to the extremely high sensitivity of TCR-peptide/class I recognition, it is not surprising to us that peptides generated from alternative reading frames are sufficient to trigger a specific CTL response, even if the efficiency of translation is perhaps several logs less than that of the ORF1 proteins. Similarly, in a recent review by Yewdell et al. (32) it is proposed that a significant source of self and viral peptides is defective ribosomal products (DRiPs), which represent prematurely terminated polypeptides and mis-folded polypeptides produced as byproducts during the translation of mRNAs, although this concept was applied only to the consideration of DRiPs encoding primary ORF-derived T cell epitopes.
In keeping with the concept of alternative ORF-derived T cell epitopes, by use of a model CTL epitope, Bullock and Eisenlohr (33) engineered a series of frameshift mutations into the influenza A/PR/8/34 nucleoprotein gene. These data confirmed that three defined NP epitopes were expressed and recognized by CTL when frameshift mutations (+1 or -1) were engineered to place the NP50–57, NP147–155, and NP366–374 epitopes into alternative reading frames. These authors demonstrated initiation codon readthrough (termed scanthrough, where the scanning ribosome does not use the conventional initiation codon, rather it initiates translation from a downstream AUG) as one mechanism of CTL epitope production. Similarly, translation of overlapping reading frames from a single viral mRNA has been described (34, 35, 36), although these alternative open reading frame products have not been tested for any immunologic relevance.
In our system, the ORF2-derived SYNTGRFPPL epitope of def-gag could be
generated from two possible initiation methionines, one residing 157
nucleotides downstream of the ORF1 initiation site for the retroviral
gag proteins and/or a second ORF2 initiation methionine located 646
nucleotides downstream of the ORF1 initiation site. It is important to
note that both putative ORF2 initiation codons are "strong"
(CACUAGAUGGTA) or "very strong"
(GGGGTAAUGGTC) initiation codons, respectively, as compared
with the Kozak consensus sequences for initiation in higher eukaryotes
(GCCGCC(A/G)CCAUGG). By this measure an initiator codon is
"strong" if there is a G at position +4, and considered "very
strong" if in addition to the G at +4 there is a purine at position
-3 (37). If translation initiates at an AUG in ORF2, it seems likely
that the first initiating methionine (nt 157) might have been used in
our larger def-gag constructs to produce coding information
for SYNTGRFPPL, whereas the second initiating methionine (nt 646) was
probably used to generate the SYNTGRFPPL peptide from all of the
smaller viral constructs. Although we have shown in the context of
extended minigene constructs that the generation of the SYNTGRFPPL
peptide is dependent on the use of an alternative ORF2 Met-initiation
codon (Fig. 6), additional experiments will be necessary to determine
the exact mechanism used to generate the SYNTGRFPPL peptide from full
length BM5 ecotropic and defective gag genes. It is possible either
that the SYNTGRFPPL peptide is generated through a ribosomal frameshift
or that two separate transcripts are generated from the gag
gene. Additionally, these and other possibilities are not mutually
exclusive.
In summary, we suggest that at least two separate reading frames of the BM5 ecotropic and defective gag mRNA can be used to generate separate proteins/peptides. One of these ORFs contains the retroviral gag proteins, while the second reading frame generates an antigenic peptide recognized by gag-specific CTL. The results of these experiments suggest that it may be necessary to consider the use of reading frames other than the primary reading frame when trying to identify antigenic peptides, and that the use of peptides generated from alternate reading frames may add to the total pool of processed peptides available for recognition by T cells. Alternate reading frame-derived peptides could also prove to have a serious impact on various aspects of the immune system, including the functional T (and perhaps B) cell repertoires. For example, such "self" peptides may play a role in such processes as positive and negative thymic selection, functioning as antagonists or agonists for either the expansion of certain TCR or conversely the formation of "holes" in the TCR repertoire. Alternative ORF-derived peptide epitopes might also be important in the generation of immune responses to foreign agents, as in this study, as well as in autoimmune processes. By consideration of these and other potential settings, it becomes clear that the existence of even a small amount of a given peptide, regardless of its point of origin, may result in profound effects on both the development and function of the immune system.
|
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
3 Address correspondence and reprint requests to Dr. William R. Green, Department of Microbiology, Dartmouth Medical School, Lebanon, NH 03756.
4 Abbreviations used in this paper: MAIDS, murine acquired immunodeficiency syndrome; ORF2, open reading frame 2; dG-Vac, defective gag recombinant vaccinia virus; def-gag, defective gag; AA, amino acids; SC.1/BM5 Eco, SC.1 fibroblast cells infected with the MAIDS-associated BM5 ecotropic helper virus; SM, sensitization media (RPMI 1640, 100 µM NEAA, 1 mM sodium pyruvate, 50 µM 2-ME, and 10% FBS); Y, tyrosine; F, phenylalanine; KO, knockout, NP, nuclear protein.
Received for publication September 23, 1996. Accepted for publication September 15, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. A. Green, T. Okazaki, T. Honjo, W. J. Cook, and W. R. Green The Programmed Death-1 and Interleukin-10 Pathways Play a Down-Modulatory Role in LP-BM5 Retrovirus-Induced Murine Immunodeficiency Syndrome J. Virol., March 1, 2008; 82(5): 2456 - 2469. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Ho and W. R. Green Alternative Translational Products and Cryptic T Cell Epitopes: Expecting the Unexpected J. Immunol., December 15, 2006; 177(12): 8283 - 8289. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. B. Zook, M. T. Howard, G. Sinnathamby, J. F. Atkins, and L. C. Eisenlohr Epitopes Derived by Incidental Translational Frameshifting Give Rise to a Protective CTL Response. J. Immunol., June 1, 2006; 176(11): 6928 - 6934. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Ho and W. R. Green Cytolytic CD8+ T Cells Directed against a Cryptic Epitope Derived from a Retroviral Alternative Reading Frame Confer Disease Protection J. Immunol., February 15, 2006; 176(4): 2470 - 2475. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Schirmbeck, P. Riedl, N. Fissolo, F. A. Lemonnier, A. Bertoletti, and J. Reimann Translation from Cryptic Reading Frames of DNA Vaccines Generates an Extended Repertoire of Immunogenic, MHC Class I-Restricted Epitopes J. Immunol., April 15, 2005; 174(8): 4647 - 4656. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gaur and W. R. Green Role of a Cytotoxic-T-Lymphocyte Epitope-Defined, Alternative gag Open Reading Frame in the Pathogenesis of a Murine Retrovirus-Induced Immunodeficiency Syndrome J. Virol., April 1, 2005; 79(7): 4308 - 4315. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Bain, P. Parroche, J. P. Lavergne, B. Duverger, C. Vieux, V. Dubois, F. Komurian-Pradel, C. Trepo, L. Gebuhrer, G. Paranhos-Baccala, et al. Memory T-Cell-Mediated Immune Responses Specific to an Alternative Core Protein in Hepatitis C Virus Infection J. Virol., October 1, 2004; 78(19): 10460 - 10469. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Cardinaud, A. Moris, M. Fevrier, P.-S. Rohrlich, L. Weiss, P. Langlade-Demoyen, F. A. Lemonnier, O. Schwartz, and A. Habel Identification of Cryptic MHC I-restricted Epitopes Encoded by HIV-1 Alternative Reading Frames J. Exp. Med., April 19, 2004; 199(8): 1053 - 1063. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. UMEMURA, H. NISHIMURA, T. YAJIMA, W. WAJJWALK, T. MATSUGUCHI, M. TAKAHASHI, Y. NISHIYAMA, M. MAKINO, Y. NAGAI, and Y. YOSHIKAI Overexpression of interleukin-15 prevents the development of murine retrovirus-induced acquired immunodeficiency syndrome FASEB J, November 1, 2002; 16(13): 1755 - 1763. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Saulquin, E. Scotet, L. Trautmann, M.-A. Peyrat, F. Halary, M. Bonneville, and E. Houssaint +1 Frameshifting as a Novel Mechanism to Generate a Cryptic Cytotoxic T Lymphocyte Epitope Derived from Human Interleukin 10 J. Exp. Med., February 4, 2002; 195(3): 353 - 358. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
