fax: (301) 402-5585.
CD4+ immune escape and subsequent T-cell failure following chimpanzee immunization against hepatitis C virus†
Article first published online: 29 AUG 2006
Copyright © 2006 American Association for the Study of Liver Diseases
Volume 44, Issue 3, pages 736–745, September 2006
How to Cite
Puig, M., Mihalik, K., Tilton, J. C., Williams, O., Merchlinsky, M., Connors, M., Feinstone, S. M. and Major, M. E. (2006), CD4+ immune escape and subsequent T-cell failure following chimpanzee immunization against hepatitis C virus. Hepatology, 44: 736–745. doi: 10.1002/hep.21319
Potential conflict of interest: Nothing to report.
- Issue published online: 29 AUG 2006
- Article first published online: 29 AUG 2006
- Manuscript Accepted: 23 JUN 2006
- Manuscript Received: 21 DEC 2005
- National Vaccine Program Office
- National Cancer Institute. Grant Number: CA85883
Vol. 55, Issue 3, 983, Article first published online: 23 FEB 2012
Hepatitis C is a major cause of chronic liver disease, with 170 million individuals infected worldwide and no available vaccine. We analyzed the effects of an induced T-cell response in 3 chimpanzees, targeting nonstructural proteins in the absence of neutralizing antibodies. In all animals the specific T-cell response modified the outcome of infection, producing a 10- to 1,000-fold reduction in peak virus titers. The challenge of 2 immunized animals that had been previously exposed to hepatitis C virus resulted in subclinical infections. Immune responses in the third animal, naive prior to immunization, limited viral replication immediately, evidenced by a 30-fold reduction in virus titer by week 2, declining to a nonquantifiable level by week 6. After 10 weeks of immunological control, we observed a resurgence of virus, followed by progression to a persistent infection. Comparing virus evolution with T-cell recognition, we demonstrated that: (i) resurgence was concomitant with the emergence of new dominant viral populations bearing single amino acid changes in the NS3 and NS5A regions, (ii) these mutations resulted in a loss of CD4+ T-cell recognition, and (iii) subsequent to viral resurgence and immune escape a large fraction of NS3-specific T cells became impaired in their ability to secrete IFN-γ and proliferate. In contrast, NS3-specific responses were sustained in the recovered/immunized animals presenting with subclinical infections. In conclusion, viral escape from CD4+ T cells can result in the eventual failure of an induced T-cell response that initially controls infection. Vaccines that can induce strong T-cell responses prior to challenge will not necessarily prevent persistent HCV infection. (HEPATOLOGY 2006;44:736–745.)
Persistent infections caused by hepatitis C virus (HCV) occur in 70%-80% of the acutely infected population, most of whom will develop chronic hepatitis and be at risk for cirrhosis, end-stage liver disease, and/or hepatocellular carcinoma.1 Antiviral therapy at present is successful in about 50% of patients, but treatment carries significant side effects and is very costly.2, 3 At present, about 25,000 new HCV infections occur each year in the United States, making the development of a vaccine against this virus imperative.
Natural infection with HCV had previously been thought to afford no protective immunity from reinfection.4, 5 However, we and others have shown that following rechallenge, viremia and liver disease are significantly reduced in both intensity and duration6–8; this appears to be mediated through faster T-cell activation in peripheral blood and liver in chimpanzees8, 9 and possibly in humans.10
Despite developments with transgenic mouse systems using transplanted human hepatocytes,11 the only established animal model for HCV is the chimpanzee,12 which makes challenge vaccine experiments difficult and expensive. The development of HCV vaccines has also been hampered by the lack of an effective in vitro cell culture system. Recently, RNA from a genotype 2a strain of HCV was shown to replicate in cells and produce infectious virus.13 This development will aid future vaccine studies enormously. Also of importance is the natural variability of the virus, which results in the coexistence of quasispecies in infected individuals.14 It has been postulated that immune pressure on these variants brings about selection of escape mutants able to overcome host immune responses15, 16 and establish persistent infection.
There is some evidence that antibodies (Abs) to surface proteins can neutralize HCV in vitro,13, 17, 18 and vaccine studies have been carried out in chimpanzees using the recombinant envelope glycoprotein E1E2.19–21 However, in natural infections, Abs to E1E2 occur more frequently in those with persistent viremia than in those with self-limiting infections.22, 23 Moreover, viral clearance following infection has been more closely associated with strong T-cell responses, often to nonstructural proteins, in humans24–26 and chimpanzees.9, 25, 27–30 Recently, vaccines targeting HCV core, E1E2, and NS3-NS5 resulted in attenuated infections in chimpanzees.31–33 However, in some studies it could not be clearly determined which component of the vaccine was controlling the virus replication, although E2-specific antibody responses were associated with sterilizing immunity in 1 animal.32 Together, these observations suggest that a vaccine-induced T-cell response would contribute significantly to control of virus replication.
To address this question, we undertook a sequential analysis of the effect of an enhanced or induced T-cell response on HCV disease in chimpanzees in the absence of Abs to surface antigens. We utilized a DNA prime/recombinant vaccinia virus (rVV)–boost strategy to study the effect of specific immune responses on the outcome of HCV challenge. Immunization protocols designed to stimulate specific T cells directed against HCV-NS3, NS5A, and NS5B epitopes were developed in order to study the role of T-cell immunity separate from the effects of neutralizing antibodies. We challenged the immunized animals with clonal HCV, which bypasses analytical complications associated with quasispecies variability of natural HCV inocula. The challenge virus had the same sequence as the antigens used for immunization. Animals were monitored for disease, infection, and immunologic parameters and were compared to control animals and to historical data on HCV-infected chimpanzees.
Materials and Methods
The housing, maintenance, and care of the chimpanzees (Pan troglodytes) in this study were in compliance with all NIH guidelines. The animals were: Ch1605, a naive, nonvaccinated control animal inoculated with 3.2 50% chimpanzee infectious doses (CID50) of cHCV that became infected, cleared the virus, and 10 months later was reinoculated (with 100 CID50 of cHCV)6; Ch1588, a recovered animal initially infected with HCV-H plasma (64 CID50)35 that 2 years after clearance was immunized with NS3-plasmid and rVV-NS3 and rechallenged (with 100 CID50 of cHCV); Ch1606, a recovered animal initially infected with cHCV (100 CID50) that 10 months after clearance received NS3-, NS5A-, and NS5B-based immunization (DNA and rVV) and was rechallenged (with 100 CID50 of cHCV); and Ch6394, a naive animal immunized and challenged in a manner identical to Ch1606.
Plasmids for DNA immunization expressing HCV-NS3, NS5A, or NS5B proteins under the control of the cytomegalovirus promoter were constructed following standard protocols. For rVV constructs the HCV-NS3, NS5A, and NS5B ORFs were cloned into the plasmid p7.5tk, which contained the VV constitutive 7.5k promoter flanked by the vaccinia thymidine kinase gene.36 The presence of an insert was verified by PCR and of protein expression by Western blot analysis of infected cells (data not shown). The virus was prepared by expansion in BSC-1 cells and spinner cultures of HeLa-S3 cells.
Immunization, Challenge, and Disease.
Immunization consisted of 3 intramuscular injections of DNA (1 mg DNA/dose) together with 0.5 mg of CpG oligonucleotides,37 at 0, 4, and 8 weeks, followed by a boost with 3 × 108 PFU of each rVV delivered subcutaneously at 16 weeks, together with 3 × 108 PFU of Tricom vector vT224 (a gift from Therion Biologics). The Tricom recombinant vaccinia virus expresses human costimulatory molecules B7.1, ICAM-1, and LFA-3 and was used to enhance immune responses to cells expressing HCV immunogens.
After challenge (9 weeks after the rVV booster dose for Ch1588 and 4 weeks after the rVV for both Ch1606 and Ch6394), the chimpanzees were monitored for the development of hepatitis by measurement of serum alanine amino transferase (ALT) and HCV RNA by nested reverse-transcription (RT)-PCR (detection limit of 40 RNA copies/mL) and real-time RT-PCR (detection limit of 200 RNA copies/mL) as previously described.38, 39 HCV-RNA titers during the primary infection of Ch1588 were monitored by quantitative RT-PCR as previously described.35
HCV Antigens and Ab Assays.
PBMC were isolated from total blood, washed, and stored as previously described.21 Peptide pools containing 30-45 overlapping peptides (18-mers overlapping 11 amino acids, NIH Reagent Program) or individual peptides were prepared in PBS. Prestimulation of the PBMCs and ELISpot analysis of cells producing interferfon gamma (IFN-γ) or interleukin 4 (IL-4) was performed as previously described,30 using peptides or peptide pools at 1 μg/mL of each individual peptide and Multiscreen-IP Opaque plates (Millipore) for the ELISpot test. Spot-forming units (SFUs) were counted using an ELISpot automated reader (ImmunoSpot, CTL Analyzers, Cleveland, OH). The number of specific SFUs was calculated by subtracting the mean number of duplicate control wells (irrelevant antigen) from the mean number of duplicate experimental wells. In some cases CD4+ and CD8+ T cells were depleted from the total T-cell population using antihuman CD4+ or CD8+ dynabeads (Dynal) and tested against wild-type and mutant peptides. The individual peptide sequences used for these depletion analyses are shown in the legend for Fig. 6.
T-Cell Proliferation Assays.
These assays were performed as previously described,21 using thawed PBMCs in 96-well plates (105 cells/well) in 200 μL of X-vivo15 media (Bio-Whittaker). Cells were stimulated with 1 μg/mL protein in 3 replicate cultures per stimulation condition.
Intrahepatic levels of IFN-γ, IL-4, interleukin 10 (IL-10), and the epsilon chain of CD3 (CD3-ϵ) mRNA were analyzed using real-time PCR. Relative mRNA quantification was normalized to an endogenous reference (human glyceraldehyde-3-phosphate dehydrogenase) and expressed relative to a calibrator (liver biopsy taken prior to HCV inoculation) as previously described.6
Thawed PBMCs were labeled with 5,6-carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes) according to the manufacturer's instructions. CFSE-labeled cells were seeded at 2 × 106 cells/well in 96-well plates (Nunc, Roskilde, Denmark) and cocultured for 6 days in X-vivo15 media (Bio-Whittaker) with 10 μg/mL of whole antigen. On day 7 cells were washed and restimulated for 6 hours with overlapping peptide pools (1 μg/mL each peptide), together with 1 μg/mL human anti-CD28 and anti-CD49d Abs. For intracellular detection of IFN-γ, cells were fixed and permeabilized with Cytofix/Cytoperm Plus (BD Pharmingen) prior to staining. Surface and intracellular staining was performed using PerCP-conjugated anti-CD4 and APC-conjugated anti-IFN-γ Abs (BD Pharmingen). Multiparameter flow cytometry was performed according to standard protocols.40 Profiles were gated on lymphocytes, and 60,000-250,000 total events were collected. Absolute frequencies of CD4+ and CD8+ IFN-γ-producing T cells were calculated using FlowJo software by initially gating on the CD4+/IFN-γ+ population and subsequently subtracting the background frequencies obtained on the unstimulated samples.
HCV-specific RT-PCR was performed on serum RNA and analyzed as previously described.38 Sequencing reactions on purified PCR product were carried out by SeqWright (Houston, TX) using HCV-specific primers, and the data were analyzed using DNAstar (Madison, WI).
Rechallenge Infection in Immunized Ch1588 and Ch1606.
Immunization of recovered animals prior to rechallenge was directed to the enhancement of memory T-cell responses. Rechallenge in both animals resulted in subclinical infections. Initial immunization in Ch1588 used only NS3 to induce specific T-cell responses. In this animal, infection was delayed for 5 weeks (Fig. 1B), in contrast to the immediate infection that occurred in an unvaccinated/recovered animal (Ch1605) upon rechallenge (Fig. 1A). Viremia in Ch1588 was shorter, 2 weeks versus 15 weeks in Ch1605 and 10-fold lower, quantifiable at only one time point, week 6 (104RNA copies/mL). The profile of HCV reinfection in Ch1605 was previously reported6 and found to be comparable to reinfection profiles of other recovered chimpanzees.6, 8
A second recovered animal, Ch1606, was immunized with a preparation including NS3, NS5A, and NS5B immunogens. Following challenge, viral replication was further controlled from that seen in Ch1588. Serum from Ch1606 was RNA positive only by highly sensitive nested PCR performed at week 5 postinoculation (p.i.; detection limit 40 RNA copies/mL). This transient viremia coincided with minimal elevation of ALT (Fig. 1C).
Primary Infection in Immunized Ch6394.
Immunization of a naive animal was designed to prime adaptive immunity against HCV infection using the same NS3, NS5A, and NS5B recombinants used in Ch1606. Upon challenge, Ch6394 was infected at week 1 p.i. However, in contrast to the viral kinetics seen in unimmunized animals (Fig. 2A-C), viral replication was immediately controlled and RNA titers decreased at week 2 p.i., reaching <200 copies/mL at weeks 6-7 p.i. (Fig. 2D). The animal did not exhibit elevation of ALT typical of primary infections in naive animals. This immediate control of virus replication resembled the profiles observed in unimmunized rechallenged animals (Fig. 1A), indicating the immune system of this naive animal had been primed by the vaccine such that rapid induction of specific T cells following infection was achieved. However, following this control of virus replication there was a resurgence of viremia at week 10 p.i., and the animal developed a persistent infection.
Enhancement of Specific Abs during Immunization and Challenge.
In the recovered animals anti-NS3 Ab levels increased by 10- to 25-fold during immunization (Fig. 3). In contrast, the naive/vaccinated chimpanzee (Ch6394) did not show substantial increase in Abs for any of the viral antigens until later time points following challenge. Abs to E1E2 were not detected in any of the animals (data not shown).
Stimulation of T-Cell Responses upon Immunization.
We analyzed the cellular responses induced following immunization and challenge in the 3 animals in this study (Fig. 4). T-cell immune responses in peripheral blood were tested by ELISpot analysis and proliferation assays. Intrahepatic T-cell activity was assessed by quantification of expression of cytokine mRNA. In all 3 animals, immunization stimulated antigen-specific T cells.
At or before the time of challenge (week 0), in all chimpanzees we detected high levels of IFN-γ-producing cells (Fig. 4A-C) or proliferative responses (Fig. 4D-F), S.I. = 10 for Ch1588 and Ch1606, S.I. = 5 for Ch6394. For clarity, responses at key points are shown for each animal. Note the change in scale for Fig. 4A-C. Much higher ELISpot responses were observed postchallenge in Ch1606 and Ch6394 than in Ch1588. In Ch6394, we also detected IL-4-producing cells specific for NS3, NS5A, and NS5B (data not shown).
Ch1588 and Ch1606 (the recovered animals) maintained strong immune responses in the first 5 weeks following challenge (Fig. 4A-B,D-E), with HCV RNA undetectable by nested RT-PCR (Fig. 1). The infections in Ch1588 and Ch1606 were resolved within 1-2 weeks, almost 10 times faster than in unimmunized, rechallenged animals (Fig. 1A). Ch6394 (the naïve, immunized animal) showed immediate control of the virus replication at week 1 (Fig. 2D). This was coincident with a decline in the levels of IFN-γ- and IL-4-producing cells in peripheral blood at week 2 p.i. and increased expression of intrahepatic cytokine mRNA for IFN-γ and CD3-ϵ (see below), suggesting a migration of T cells from the peripheral blood to the liver in response to infection. A similar profile was observed for proliferative responses in peripheral blood (Fig. 4F). These data suggest T-cell-mediated control of viral RNA up to week 10 p.i. However, control was lost and a persistent infection was established in Ch6394. Coincident with the resurgence of virus in Ch6394 (after week 10 p.i.), we again observed decreased T-cell responses in peripheral blood and significantly increased expression of IFN-γ and CD3-ϵ mRNA in the liver (Fig. 4C,J). An important observation in this animal was the loss of NS3-specific T-cell responses postchallenge, which, unlike the responses to NS5A or NS5B, did not recover after the first weeks of infection.
Intrahepatically, the recovered/immunized animals did not exhibit significant changes in expression of mRNA for Th1-type (IFN-γ) or Th2-type (IL-4, IL-10) cytokines. The unavailability of closer serial biopsies might have resulted in peaks of intrahepatic mRNA in these profiles being missed. Ch6394 presented intrahepatic mRNA patterns during the acute phase (Fig. 4J) similar to those described for the recovered/unimmunized animals, in that substantially elevated expression (>12-fold) of IFN-γ mRNA was observed at the first biopsy point (3 weeks p.i.). With the exception of this first time point, the increases in IFN-γ mRNA for Ch6394 coincided with increases above the baseline of CD3-ϵ mRNA.
Phenotype and Effector Function of HCV-Specific T Cells.
To examine whether the loss of control of the infection by the naïve/vaccinated chimpanzee was a result of imparity in either CD4+ or CD8+ T-cell frequency and/or their effector function, we analyzed PBMCs by flow cytometry (Fig. 5). We compared the absolute frequencies of HCV-specific IFN-γ-producing (effector) CD4+ and CD8+ T cells in the total T-cell population (Fig. 5, bars) and their ability to proliferate (Fig. 5, hatched area). The percentage of IFN-γ-positive cells that had proliferation potential are indicated as the hatched areas in Fig. 5 and were calculated as follows: [(IFN-γ+CFSE−)/(IFN-γ+)] cells × 100. For some data points in Fig. 5, as many as 100% of the cells on the graphs are shown as proliferating; however, note that these numbers represent only those HCV-specific cells producing IFN-γ. The data are not meant to imply that 100% of all T cells were proliferating or producing IFN-γ at any time point.
Ch1606 showed an increasing frequency of IFN-γ-producing T cells over time, with similar profiles for CD8+ and CD4+ T cells. Upon clearance, the number of proliferating effector T cells increased 8- to 14-fold, which was maintained during the following 5-6 months. Ch1588 showed a pattern similar to that of Ch1606, with strong NS3-specific responses, high frequency of proliferating effector T cells during the acute phase, and maintenance of specific responses (Fig. 5). In Ch6394 specific effector CD4+ T-cell responses were detected following immunization (week 0) but, in contrast to Ch1606 and Ch1588, the frequency of those cells did not increase when the animal became viremic. In addition, in Ch6394 we observed a decreased capacity of the effector T cells to proliferate, especially CD4+ T cells, early in the acute phase (week 4 p.i.) and later, when the animal became persistently infected (after week 10 p.i.).
Sequence Mutations and Immune Escape.
Sequence analysis of the NS3, NS5A, and NS5B regions of viral RNA isolated from Ch6394 at weeks 8 and 13 p.i. revealed several changes. These dates represent resurgence of viremia after significant control to less than 200 RNA copies/mL (Fig. 2D). We found 7 nucleotide mutations, 5 of which were synonymous. Two mutations (nt4171 in NS3 and nt7210 in NS5A) led to amino acid substitutions (N1277T and R2290Q, respectively). Neither mutation has been associated with previously described T-cell epitopes.
We performed IFN-γ-specific Elispot assays on PBMCs from Ch6394-isolated pre- and postvaccination and up to 1 year postchallenge using individual overlapping wild-type (WT) and mutant peptides that spanned these amino acid residues (Fig. 6A). As expected, T cells collected prevaccination (week −36) did not recognize any peptide. T cells collected postvaccination recognized both the WT peptides from NS3 and NS5A but did not react against the peptides carrying the mutations at amino acids 1277 (NS3-mut) and 2290 (NS5A-mut). After challenge, the T cells were still capable of recognizing the WT peptides although these responses declined considerably as the persistent infection progressed. There was no recognition of the mutant peptides even after 53 weeks of viral infection, suggesting no new T-cell responses were generated against the mutant sequences.
To determine whether this immune escape was CD4+ or CD8+ T-cell specific, we performed depletion experiments using magnetic beads specific for these cell types. Depleted samples were more than 90% pure for CD4+ or CD8+ T-cell populations as assessed by flow cytometry (data not shown). Fig. 6B shows IFN-γ-producing cells from 7 weeks p.i. stimulated with WT and mutant peptides. Neither mutant peptide was recognized by these T cells, confirming our previous data suggesting immune escape. We saw a significant reduction in signal against the WT peptides using CD4+-depleted populations, whereas good responses were maintained using the CD8+-depleted cells, suggesting these escape mutations were in CD4 epitopes.
In this article we report the effect of induced cellular immune responses in the absence of neutralizing Abs on the control of HCV replication following challenge. As the chimpanzee is the only available model for HCV infection and because of the costs associated with chimpanzee research, this study was limited to 3 animals. However, this laboratory has extensive experience with the clonal HCV inoculum used as the challenge in this study for both primary infections and for reinfections in recovered animals.6, 41 Because this inoculum results in highly reproducible and predictable infections in chimpanzees, it is possible to draw important and informative conclusions from the present study of 3 chimpanzees. In unimmunized, recovered animals viremia occurred 1 week postchallenge. Memory T-cell responses generated by primary infections immediately controlled HCV replication, but viremia lasted for several months with mild ALT elevations.6 In contrast, boosting the immune response by vaccination of the recovered animals resulted in delayed subclinical infections in both animals, with no biochemical evidence of hepatitis. Vaccine-induced control of HCV infection was linked with early activation of specific T cells that produced IFN-γ during the immediate weeks after challenge, followed by increased proliferation of T cells, which peaked during the viremic phase.
These results confirmed the importance of T-cell responses in controlling HCV infections and justified the immunization of a naive chimpanzee using the NS3, NS5A, and NS5B recombinants. Immunization led to priming of the T cells, and although challenge resulted in early viremia, replication was immediately controlled, as shown by the almost 100-fold decreased titer between weeks 1 and 3 p.i. In primary infections of humans26 and naive/unvaccinated chimpanzees,28, 41 the virus replicates to high levels during the first 1-2 weeks of infection (with an average doubling time of 0.5 days in chimpanzees41). Hepatitis, as measured by ALT elevations, coincides in all naive animals with immune responses at about week 10, leading to control of viral replication. This suggests that the immune responses induced in Ch6394 to NS3, NS5A, and NS5B resulted in memory T cells that could be recalled upon challenge with infectious virus and which led to control of replication similar to that observed in recovered animals.
Despite this early control, Ch6394 became persistently infected. Sequence analysis identified mutations in the NS3 and NS5A regions at weeks 8 and 13 p.i. (following a period when the viremia appeared to be controlled). We showed that these amino acid changes resulted in immune escape that was CD4+ specific. It is probable that these mutations arose under pressure from the early immune response induced in this animal following challenge. CD8+ T cells are potent mediators of viral clearance, and most viral escape mutations identified have been in CD8+ epitopes. Although CD4+ T cells are not required for the induction of CD8+ T cells during acute infection, CD4+ T-cell help is indispensable to sustain antiviral CD8+ T-cell activity during the control of chronic infections42, 43 and has been shown to be necessary for viral clearance.29 More data is pointing to the role of CD4+ T cells in the selection of viral escape mutants.44 NS3-specific CD4+ escape mutants have been identified in patients. The variants attenuate or fail to stimulate T-cell proliferation,45 which is consistent with our observation of T-cell proliferation in Ch6394 after the emergence of mutations in NS3 and NS5A.
Strong NS3-specific CD4+ T-cell responses have been associated with self-limited HCV infection in humans.46, 47 The NS3-specific immune response was more robust and sustained in the recovered chimpanzees, partly because of the priming by the primary infection. In contrast, Ch6394 lost the potency of the NS3-specific response during the acute phase. We showed the amino acid changes we observed led to a loss of T-cell recognition, suggesting immune escape. We propose that the mutations in CD4+ epitopes caused the reduced stimulation of T-cell proliferation seen during the later acute phase of HCV infection in Ch6394. This immune escape may also have led to down-regulation of important lineages of T cells involved in viral clearance.
The degree of cell activation is likely to be an important parameter in determining whether CD8+ T cells are deleted or become functionally defective, and both mechanisms have been described in viral infections.48–50 Our results have demonstrated that despite the enhancement of specific T cells with effector functions by immunization, differences in the specificity of the cellular immune response, in the frequency of the sustained response, and in the ability of the specific T cells to proliferate were observed between the recovered and the naive, immunized animals. Folgori et al.33 observed persistence of virus in 1 of 5 vaccinated naive chimpanzees using a T-cell vaccine approach. However, in that study the persistence was associated with a weak T-cell response following vaccination and immune escape at the CD8+ T-cell level. In our study, despite the induction of a broad multispecific T-cell response in Ch6394, immune escape from CD4+ T cells still occurred. This resulted in the eventual failure of the induced response to clear infection, leading to a persistence of the virus. This has important consequences for future T-cell vaccines. In conclusion, attention needs to be given to the type of T cells induced, in particular, epitope specificity and maintenance of CD4+ T-helper responses.
We thank Estella Jones and Ray Olsen for the expert care and handling of the chimpanzees and the samples within the FDA facility.
- 35MargolisHS, AlterMJ, LiangTJ, DienstagJL, eds. Protective antibodies in immune globulins prepared from anti-hepatitis C virus (HCV)-positive plasma units: update of a chimpanzee study. In: Viral Hepatitis and Liver Disease. London: International Medical Press, 2002.
- 36Generation of recombinant vaccinia viruses. In: AusubelFM, BrentR, KingstonetRE, eds. Current Protocols in Molecular Biology. New York: Green Publishing Associates/Wiley Interscience, 1991: 16.17.1–16.17.16., .
- 40ColiganJ, KruisbeekA, MarguliesD, SheevacE, StroberWJ, eds. Current Protocols in Immunology. New York: Green Publishing Associates/Wiley Interscience, 1995: 5.3.1–5.3.23., , . In: