Virus-like particles_ Designing an effective AIDS vaccine

Virus-like particles: Designing an e V ective AIDS vaccineKelly R. Young a,1, Sean P. McBurney a,b , Lukena U. Karkhanis a , Ted M. Ross a,b,¤aDepartment of Medicine, Division of Infectious Diseases, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, USAbCenter for Vaccine Research for Emerging Diseases and Biodefense, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, USAAccepted 5 May 2006AbstractViruses that infect eukaryotic organisms have the unique characteristic of self-assembling into particles. The mammalian immune sys-tem is highly attuned to recognizing and attacking these viral particles following infection. The use of particle-based immunogens, often delivered as live-attenuated viruses, has been an e V ective vaccination strategy for a variety of viruses. The development of an e V ective vac-cine against the human immunode W ciency virus (HIV) has proven to be a challenge, since HIV infects cells of the immune system causing severe immunode W ciency resulting in the syndrome known as AIDS. In addition, the ability of the virus to adapt to immune pressure and reside in an integrated form in host cells presents hurdles for vaccinologists to overcome. A particle-based vaccine strategy has promise for eliciting high titer, long-lived, immune responses to a diverse number of viral epitopes against di V erent HIV antigens. Live-attenuated viruses are e V ective at generating both cellular and humoral immune responses. However, while these vaccines stimulate immunity, chal-lenged animals rarely clear the viral infection and the degree of attenuation directly correlates with protection from disease. Further, a live-attenuated vaccine has the potential to revert to a pathogenic form. Alternatively, virus-like particles (VLPs) mimic the viral particle without causing an immunode W ciency disease. VLPs are self-assembling, non-replicating, non-pathogenic particles that are similar in size and conformation to intact virions. A variety of VLPs for lentiviruses are currently in preclinical and clinical trials. This review focuses on our current status of VLP-based AIDS vaccines, regarding issues of puri W cation and immune design for animal and clinical trials.© 2006 Elsevier Inc. All rights reserved.Keywords:AIDS; HIV; Virus-like particle; VLP; Pseudovirions; Vaccine1. Introduction1.1. Virus structure, pathogenesis, and host response The human immunode W ciency virus (HIV) 2 is a lentivirus and a member of the Retroviridae family. All retrovirusescontain a single-stranded RNA genome (100–120nm) that is converted to a double-stranded DNA form (provirus),which integrates into the host cell chromosomal DNA.Retroviruses are classi W ed as simple and complex retrovi-ruses. The simple viruses only contain the gag , pol , and env genes, whereas the complex retroviruses, which include*Corresponding author. Fax: +1 412 648 8455.E-mail addresses: kyoung@ (K.R. Young), rosst@ (T.M. Ross).1Present address: 824 Mary Ellen Jones Building, Carolina Vaccine Institute, University of North Carolina, Chapel Hill, NC 27599, USA.2Abbreviations used: HIV, human immunode W ciency virus; AIDS, acquired immune de W ciency syndrome; VLP, virus-like particle; SIV, simian immunode W-ciency virus; CRF, recombinant circulating forms; dsDNA, Double-stranded DNA; RT, reverse transcriptase; LTR, long terminal repeat; Gag, group associated antigen; Pol, polymerase; Env, envelope; SU, surface; TM, transmembrane; PR, protease; IN, integrase; pr55, Gag–Pol precursor polyprotein; MA, matrix; CA,capsid; NC, nucleocapsid; Vif, virion infectivity factor; Vpr, viral protein r; Nef, negative regulatory factor; Tat, trans-activator of transcription; Rev, regulator of virion protein expression; Vpu, viral protein u; vRNA, viral RNA; PIC, pre-integration complex; CTL, cytotoxic T lymphocyte; ADCC, antibody-dependent cel-lular cytotoxicity; APC, antigen presenting cell; MHC, major histocompatibility complex; CCR, chemokine co-receptor; NRE, negative regulatory element; i.v.,intravenous; i.m., intramuscular; i.d., intradermal; i.p., intraperitoneal; g.g., gene gun; p.v., post-vaccination; Vpx, viral protein x; SHIV, simian–human immuno-de W ciency virus; NIH, National Institute of Health; WHO, World Health Organization; SBBC, Sydney Blood Bank Cohort; LTNP, long-term non-progressor;LTS, long-term survivors; PBMC, peripheral blood mononuclear cells; rVV, recombinant vaccinia virus; rBV, recombinant baculovirus; MID, monkey infec-tious doses; CMV-IE, cytomegalovirus immediate early; CT, cholera toxin; INF- , interferon gamma; IL, interleukin; ODN, oligonucleotide; ZDV, zidovudine.K.R. Young et al. / Methods 40 (2006) 98–11799lentiviruses, contain these genes plus regulatory and accessory genes that aid in viral replication (Fig.1). HIV-1, HIV-2, and the simian immunode W ciency virus (SIV)are primate lentiviruses [1,2] that induce an acquired immunode W ciency disease in an infected host [3,4]. HIV-1is genotypically divided into three distinct groups: major (M), outlier (O), and non-M non-O (N) with the majority of HIV-1 strains comprising the M group. Since its intro-duction into the human population, the M group has evolved into at least 10 distinct subtypes and 13 di V erent circulating recombinant forms (CRF) [5,6].The envelope of HIV-1 is composed of a lipid bilayer derived from the host cell membrane during the budding process and is embedded with multimeric glycoproteins.Each envelope glycoprotein (Env; 7–15 Env trimers per virion), is composed of a surface, globular domain (gp120,SU) and a transmembrane domain (gp41, TM). The matrix (MA, p17) protein lines the inner surface of the viral enve-lope and surrounds the capsid. The capsid (CA, p24) layer contains approximately 2000 molecules and encases the nucleocapsid (NC, p9) and the viral genome and associated viral proteins.The HIV-1 proviral genome consists of nine open read-ing frames that encode for 15 viral proteins [7] (Fig.1 and Table 1). The three major genes are (1) group-associated antigen (gag ) encoding structural core proteins, (2) poly-merase (pol ) encoding enzymatic proteins, and (3) enve-lope (env ) encoding the receptor binding protein (Table 1).The HIV-1 genome contains two regulatory genes (tat and rev ) and four accessory genes (nef , vif , vpu , and vpr ) that are required for e Y cient virion replication and matura-tion (Table 1). There are two long-terminal repeat (LTRs)which X ank both the 5Ј and 3Ј ends of the proviral DNA genome. The 5Ј LTR contains the HIV-1 promoter and enhancer sequences that regulate gene expression. For the function of each HIV-1 gene product, see [8]. Brie X y, Env binds to cell surface receptors (human CD4 and corecep-tors) resulting in viral fusion and entry into susceptible cells [9–12]. The matrix protein (MA or p17) is associated with the envelope lipid bilayer [13] and the capsid protein (CA or p24) constitutes the conical viral capsid that contains the RNA genome, which in turn is associatedwith the nucleocapsid proteins (NC or p7/p6). The reverse transcriptase (RT) converts the genomic RNA into provi-ral DNA; integrase (IN) integrates the proviral DNA into the host chromosome; and the protease (PR) is responsi-ble for proteolytic cleavage and activation of the Gag–Pol (p160) polypeptide. The process of reverse transcription,viral replication, and the role of the accessory viral proteins can be reviewed in [8].1.2. Disease course and immune response to HIV-1 in humansHIV-1 can be transmitted by sexual intercourse, blood products, contaminated needles from intravenous drug use, or from mother-to-child during the pre-natal period [41]. The timeline for disease progression from infection with HIV-1 to the development of AIDS varies between individuals. In general, the W rst symptoms of clinical AIDS become evident 8–15 years after infection. Many immunological and viral hallmarks are observed during the course of infection of HIV [42–46]. A transient peak in viremia occurs in the W rst 10 weeks after primary infec-tion with HIV-1. The activation of cytotoxic T cells (CTL) and anti-Env and anti-Gag antibodies result in containment of the initial viremia. The amount of virus in the blood (viral load) decreases to a set-point (steady state level of virus) and is prognostic for the course of infection and disease; the higher the set-point, the faster the onset of disease progression. Viral loads fewer than 103 copies of viral RNA per mm 3 of plasma generally are associated with a slower progression to AIDS. During this asymptomatic phase, low levels of viral replication occur and the number of CD4+ T cells slowly declines.However, the number of CD8+ T cells does not decrease even though the anti-HIV CTL activity slowly wanes [47].During the symptomatic phase, there is a sharp drop in the number of CD4+ T cells that is associated with a rise in viremia. A CD4+ T-cell count of less than or equal to 200cells/mm 3 of plasma de W nes the symptomatic phase known as AIDS [48]. The patient begins to develop opportunistic infections and eventually succumbs to an AIDS-related illness.1. Schematic representation of HIV-1 proviral genome. The proviral genome is X anked by two long-terminal repeats (LTRs) (white boxes). Transcrip-LTR. In the HIV-1 genome, there are nine open-reading frames, encoding 15 di products. The Gag–Pol precursor polypeptide is cleaved into Gag (CA, MA, NC, and p6) and Pol (PR, RT, and IN) gene products. The Env/Vpu mRNA ) that is processed to Env Vpr, and Nef) proteins are encoded by multiply-spliced Gag–Pol precursor mRNA.vprgp41100K.R. Young et al. / Methods 40 (2006) 98–1171.3. Anti-retroviral treatments for AIDSAt the end of 2005, »45 million people (»18 million women; »5 million new infections) were infected with HIV-1. The majority of these infected individuals lived in develop-ing countries (37 million) as estimated by the UNAIDS. The current treatment of highly active anti-retroviral ther-apy (HAART), which consists of an assortment of anti-ret-roviral drugs, reduces viral replication and the symptoms associated with the infection [49]. A combination of two or more anti-retroviral medications is generally more e V ective than using just one of these medications (monotherapy) for treating HIV infection [49]. HAART involves using an assortment of anti-retroviral drugs to reduce or prevent viral replication (usually inhibitors of HIV-1 PR and RT).The regimen usually consists of one protease inhibitor (mivudine) and one or more reverse transcriptase inhibitors (e.g. zidovudine or stavudine) and results in reduced levels of virus (<50 copies of viral RNA/mm3 blood) after 1 year of treatment in approximately 60–80% of patients [50]. The use of HAART has enhanced both the longevity and quality of life for infected individuals by con-trolling viral replication [50]. Some of the advantages of combining anti-retroviral drugs for the treatment of HIV are: (1) minimal incidence of HIV-related complications, (2) decrease in viral loads/induction of lower viral set points, (3) lessened severity and delayed onset of symptoms and (4) prolonged survival of infected individual [51,52]. Despite the e V ectiveness of HAART, several drawbacks accompany this treatment that limits its worldwide useTable 1Summary of HIV-1 proteins and potential use in VLP vaccinesProtein Function Advantageous characteristic VLPvaccine Disadvantageouscharacteristic VLP vaccinePrimaryimmunityelicitedLiteraturecitedCA, MA, NC Structural proteins encoded bygag Required for particle formation; elicitsstrong cell-mediated immune responseNone Cellular andhumoral[14–17]Viral core; encapsidation of viral RNAPR Enzyme encoded by pol Cleaves Gag–Pol and Gag polypeptideinto seven gene products May induce reversion orrecombination; resistance todrug therapyCellular[18,19]Cleaves Gag–Pol and Gag polypeptide into seven mature gene productsRT Enzyme encoded by pol Additional vaccine target May induce reversion orrecombination; replication iserror-prone, escape mutants Cellular[20,21]Converts genomic single stranded RNA to double stranded proviral DNAIN Enzyme encoded by pol Persistent expression of viral proteins;no boosting required (must have LTRs)Lifelong infection withvaccine strain; may inducereversion or recombinationCellular[22–24]Directs proviral integration into host chromosomeEnv Structural protein Binding/entry of VLPs into susceptiblecells; target of neutralizing antibodies May induce apoptosis ofbystander cellsHumoral andcellular[9,25–27]Virus binding and entry into susceptible cellsTat Regulatory protein Strong cell-mediated response seen earlyin infection; potentially transcribesvRNA Strong induction ofapoptosis, modulating ofexpression of many cellulargenesCellular andhumoral[28–30]Promoting and enhancing viral transcriptionRev Regulatory protein Required for nuclear export of Gag–Poland Env mRNAs to be translated intoPROTEINS None Little to none[30,31]Nuclear export of unspliced and singly spliced vRNAsNef Accessory protein Very strong cell-mediated response earlyin infection Down regulates CD4/MHCI; perturbs T-cell activationand infectivity of virusCellular[32–34]Down regulation of CD4 and MHC I; increases infectivityVpu Accessory protein Inhibits CD4-Env binding in ER andEnv degradation; enhances virusbudding Down regulates CD4; lowersexpression of bicistronicmRNALittle to none[35,36]Down regulation of CD4; enhances virus releaseVpr Accessory protein Additional vaccine target Induces cell cycle arrest atG2 phase; preventsincorporation of deleteriousdUTPs into virion Cellular[37–39]Nuclear localization of the PIC; cell cycle arrest (G2)Vif Accessory protein Additional vaccine target Enhances infectivity; inhibitscellular antiviral factors Little to none[36,40]Enhances infectivityK.R. Young et al. / Methods 40 (2006) 98–117101(particularly in developing nations). HAART does not pro-tect patients against initial infection nor does HAART clear viral infection. These drugs are highly toxic resulting in a high incidence of non-adherence by patients. In addi-tion, the e Y cacy of these drugs is reduced by interactions with other drugs and food or unfavorable pharmacokinet-ics. Anti-retroviral drugs have transportation and storage issues and are costly to produce. Viral resistance to these drugs often results in escape mutants and multi-drug resis-tance [53–57]. Therefore, one of the long-term goals of HIV/AIDS research is the development of a safe and e V ec-tive AIDS vaccine.1.4. Correlates of protectionThere are multiple immune e V ector mechanisms known to reduce viral replication in vitro, such as the binding of neutralizing antibodies to viral particles, cytolytic activity of CD8+ T cells, antibody-dependent cellular cytotoxicity (ADCC) of infected cells, and innate immune defenses. An ideal AIDS vaccine will most likely need to elicit both cross-reactive antibodies (at mucosal surfaces) and strong cell-mediated immune responses against multiple HIV anti-gens to protect individuals against viral challenge. Previous attempts to develop an e V ective AIDS vaccine have focused on eliciting these immune responses by administering either recombinant viral proteins, peptides, particles, or by expres-sion of viral proteins in vivo from viral vectors or DNA plasmids [58–65]. Recombinant monomeric forms of Env induced low and transient neutralizing antibodies [68–70] that do not elicit protective immunity in humans [66–69], do not neutralize primary strains of HIV [70–73], and do not elicit strong CTL responses [67,74]. In addition, these vaccines failed to protect rhesus macaques from SHIV challenge [75,76].The role of Env appears critical for the induction of protective immunity in recent AIDS vaccine candidates tested in non-human primates [77,78]. Inclusion of Env in a multi-component AIDS vaccine result in lower viral set points and higher CD4 counts following challenge com-pared to the same vaccines lacking Env. Envelope on the native virion most likely forms a trimer, although one model suggests that both functional and non-functional trimeric spikes are present on virions. Several vaccine strategies have incorporated an oligomeric/trimeric form of Env to elicit cross-reactive immunity that neutralizes viral infection [79–91]. Strategies using oligomeric Env gp140 or viral particles incorporating Env gp160 are favored because they may generate antibodies to the native structure of Env. The goal of trimer immunogen design is to present and preferentially induce conforma-tionally dependent antibodies that recognize epitopes present only on the native virion-associated spikes. Sev-eral of these trimeric Env immunogens do elicit slightly higher titers of neutralizing antibodies than monomeric Env gp120[82,90,92,93], however the breadth of neutraliza-tion is still somewhat limited. Often, these oligomeric Env proteins are produced by eliminating the natural cleavage site recognized by cellular proteases [89,94–96], which might in X uence the trimeric structure [84,85]. The lack of elicited high titer, broadly reactive neutralizing antibodies by these immunogens may be associated with the elicitation of primarily non-neutralizing antibodies [85,97,98], which may be because these uncleaved enve-lopes are in non-native forms or are processed through di V erent cellular pathways than cleaved forms of Env [84,85,99,100]. The presentation of Env on a viral particle appears to elicit broader cellular and humoral immunity, including neutralizing antibodies, than soluble Env molecules.In addition to an e V ective cellular response, a broadly cross-reactive humoral response will most likely be required to protect individuals against viral challenge. Anti-bodies that neutralize infection or induce antibody-depen-dent cytotoxicity (ADCC) may play preventive roles in the initial infection of HIV rather than in the control of chronic, established HIV infection. Passive antibody immu-nizations are an e V ective therapy for controlling infection in chimpanzees and humans infected with HIV-1 [101,102]. Although controversial, there is evidence that neutralizing antibodies may in X uence the level of chronic steady-state viremia [103]. This limited e V ect may be due to the rapid escape of virus from neutralizing antibodies observed in infected individuals [104,105].In the presence of pre-existing, vaccine-induced immu-nity, the correlates of protection may not be the same as those involved in preventing disease progression during natural infection with HIV-1. Research will need to further de W ne the mechanism of viral control and how pre-existing immunity may a V ect the outcome of HIV-1 infection. Nonetheless, an ideal HIV/AIDS vaccine will most likely need to elicit both cross-reactive, neutralizing antibodies and a strong cell-mediated immune response against multiple HIV antigens to protect individuals against viral challenge.2. Lentiviral particle vaccinesLentiviral particles (live virus or virus-like particles) spontaneously form in vivo when Gag gene products are produced and therefore, only the Gag proteins are neces-sary to produce a viral particle. However, natural viral par-ticles contain MA and NC, as well as viral (Env) and cellular proteins embedded in the viral membrane (Fig.2). In addition, several cellular and viral gene products are encapsidated into the viral particle, which assist during viral budding, uncoating, and replication. Viral particle immunogens can be produced as inert molecules in vivo from DNA plasmids, viral vectors in primate cells, and from yeast or insect cells, then puri W ed as immunogens. Therefore, a variety of mechanisms are available to pro-duce lentiviral particles, including live infection. In this review, only the use of lentiviral particles as vaccines will be discussed.102K.R. Young et al. / Methods 40 (2006) 98–1172.1. Live-attenuated lentiviral vaccines2.1.1. Overview of live-attenuated lentiviral vaccinesHistorically, the use of live-attenuated virus therapy has been used to control many viral infections such as measles, smallpox, mumps, and rubella [106,107]. These vaccines elicit both humoral and cell-mediated immune memory responses. Often, only one immunization is required to elicit protective immunity. More information on live-atten-uated vaccines, are reviewed in [108–117]. Live-attenuated vaccines elicit a robust, broad CTL response in conjunction with high levels of cross-reactive neutralizing antibodies [110]. Moreover, live-attenuated lentiviral vaccines persis-tently express viral antigens requiring fewer boosts, contain multiple viral antigens including the native Env conforma-tion(s), and are capable of infecting professional antigen presenting cells (APCs). However, research using live-atten-uated lentiviral vaccines has been limited to non-human primate studies and not human clinical trials due to: (1) potential reversion to a virulent form of the attenuated virus, (2) possible recombination of the vaccine strain with wild-type, pathogenic virus in an infected individual, (3) ability of the proviral genome to integrate into the host genome, (4) the dysregulation of the immune system by viral proteins, and (5) potential pathogenicity caused by the vaccine strain. In addition to the safety concerns, live-atten-uated lentiviral vaccines have an inverse relationship between attenuation and e Y cacy as observed in multiple rhesus macaque studies. Therefore, as the degree of attenu-ation increases the e Y cacy of the vaccine decreases [109,118,119].2.1.2. Live-attenuated SIV vaccinesThe SIV/monkey models have been used extensively to study the e Y cacy and safety of potential AIDS vaccines [118,120–125]. The T cell-line adapted molecular clone SIV1A11 was attenuated by the addition of a pre-mature stop codon in the vpr gene [126]. Even though rhesus macaques vaccinated with SIV1A11 were not protected from SIV superinfection, disease progression was delayed [127], indicating live-attenuated viral infection could elicit protective immunity. One of the earliest attempts to attenu-ate a lentivirus focused on abrogating Nef function in SIV macC8 by deleting 12 base-pairs in the region overlap-ping nef and the 5Ј LTR [128]. Interestingly, 17 weeks post-vaccination, the vaccine strain had reverted to a pathogenic form and the monkeys developed AIDS-like disease. Virus isolated from the peripheral blood mononuclear cells (PBMC) of these vaccinated monkeys contained a func-tional Nef protein and upon further analysis, the deleted gene sequences were restored. Similar results were seen with a live-attenuated SIV mac239 vaccine containing a single base-pair mutation that introduced a pre-mature stop at amino acid 93 in Nef [129]. These results not only showed the importance of Nef in SIV infection, which has been corrob-orated in HIV-1 infected patients, but also demonstrated that future SIV-based live vaccines would require additional attenuation to further debilitate the virus.Two attenuated SIV strains, SIVmac293 nef (SIV nef) containing a complete deletion of the nef gene, and SIV-mac293 3 (SIV 3), containing a combination of three deletions including the nef and vpr genes, as well as the negative regulatory element (NRE) of the LTR were con-structed [130,131]. These deletions resulted in low levels of virus replication and did not lead to development of dis-ease. Rhesus macaques were vaccinated (intravenously, i.v.) with SIV nef or SIV 3 and the majority of animals devel-oped persistent infection with the vaccine strain and had long lasting anti-Env and anti-Gag antibodies. When chal-lenged with wild-type, pathogenic SIV mac251 (10 animal infectious doses), macaques with persistent immune responses had no signs of simian AIDS compared to naïve monkeys. Initial safety and e Y cacy studies with SIV 3 vac-cine appeared promising in adult macaques [132,133], how-ever, further analysis revealed that this vaccine caused an AIDS-like illness in adult macaques and subsequently death in neonatal macaques at high doses [134,135]. There-fore, it is conceivable that a live-attenuated HIV-1 vaccine based on similar attenuation strategies may result in AIDS in vaccinated patients.Another interesting aspect of attenuated lentiviruses is that the degree of attenuation inversely correlates with ability to elicit e V ective immune responses. Johnson et al.K.R. Young et al. / Methods 40 (2006) 98–117103demonstrated that SIV strains with varying degrees of attenuation di V ered in their ability to elicit high titer immune responses in rhesus macaques [119]. Female mon-keys were vaccinated intravenously with one of the follow-ing attenuated SIV mac239 strains: SIV 3, SIVmac293 3X (SIV 3X), or SIVmac293 4 (SIV 4). SIV 4 contained the same three deletions as SIV 3X with the addition of a com-plete deletion of the vpr gene. Rhesus macaques vaccinated with any of these three vaccines and then challenged with pathogenic SIV mac251 had low levels of viremia and normal CD4+ cell counts. Thirty-three percent of macaques vacci-nated with SIV 3X and 50% of macaques vaccinated with SIV 4 were also superinfected with the challenge virus. After challenge, vaccinated monkeys had lower viral loads than challenged naïve animals. The most likely reason for the ability of the vaccinated monkeys to survive challenge was the induction of an early CTL response directed at epi-topes in the Gag protein and not neutralizing antibodies against Env.Live-attenuated SIV vaccines also induce cross-protec-tion against SHIV challenge [76,136,137]. Rhesus macaques vaccinated with SIV 3 or SIVmac239 were resistant to challenge with the highly pathogenic SHIV89.6P or pathogenic SIV smE660[137–139]. Monkeys immunized with SIV 3 and challenged intravenously with SHIV89.6P (37 months p.v.) had normal CD4+ T-cell counts and were free of AIDS-like disease even though all the monkeys had X uctuating viremia (300–10,000 copies of RNA/ml of plasma). Other monkeys vaccinated with SIV 3 and challenged with SIV smE660 had lower levels of viremia compared to naïve animals. However, the levels of viremia in SHIV89.6P-challenged monkeys were signi W-cantly lower than the viremia in monkeys vaccinated with SIV mac239 and signi W cantly less compared to previous studies with SIV mac251[109,132,133]. Interestingly, mon-keys vaccinated with SHIV were resistant to SIV mac239 challenge [113].Overall, live-attenuated SIV vaccines demonstrate that a broadly cross-reactive, long-lived immunity can be elicited in non-human primates after vaccination. However, the variability in attenuation/reversion to a pathogenic pheno-type is inversely correlated with the e Y cacy of the vaccine. The use of live-attenuated HIV strains in humans may result in similar complications and outcomes.2.1.3. Live-attenuated SIV–HIV (SHIV) chimeric vaccinesChimeric SIV–HIV (SHIV) viruses generally are recom-binant SIVs containing the HIV-1 envelope glycoprotein. The use of SHIV o V ers the opportunity to study infection and to elicit immunity to infection in monkeys with a enve-lope from HIV [140–144]. The prototypic SHIV, NM-3rN, encodes the env, tat, rev, vpu, and vpr gene sequences of HIV-1NL4-3 and the long-terminal repeats, gag, pol, vif, vpx, and nef genes from SIV mac239[115,145,146]. Attenuation of a SHIV strain (usually achieved by deletion of nef, vpr, or both genes) usually results in a non-pathogenic strain for monkeys [109,147]. In contrast, live-attenuated SIV or HIV vaccines often cause disease in primates [134,135,148]. Live-attenuated SHIV vaccines induce strong and long-lasting cell-mediated and humoral immune responses in monkeys [147,149,150]. In addition, SHIV vaccines allow for the unique opportunity to measure immunity directed against HIV-1 in the context of a live virus infection. Numerous live-attenuated SHIV vaccines have been constructed [140,141,147,149–152] including SHIV-dn (deleted nef), SHIV-drn (deleted vpr and nef), SHIV-dxrn (deleted vpx, vpr, and nef), SHIV-NI (deleted nef, plus interferon- expression), and live-attenuated SHIV-4 (deleted vpu and nef) and SHIV ppc (deleted vpu). SHIV-dn, SHIV-drn, SHIV-dxrn, and SHIV-NI vaccines are derived from the non-pathogenic parental strain, SHIV NM-3rN[147]. Macaques vaccinated with SHIV-dn and SHIV-drn had transient vire-mia of the vaccine strain, whereas no virus replication was detected in monkeys inoculated with SHIV-dxrn. Overall, SHIV vaccinated macaques remained disease-free before challenge. F ifty percent of the macaques immunized with SHIV-dn had neutralizing antibodies and CTLs speci W c for SIV Gag and HIV-1 Env proteins. In addition, most ani-mals demonstrated elevated natural killer cell activity. Following intravenous challenge with SHIV NM-3rN, no signs of integrated genome were found in the plasma, PBMCs, or inguinal lymph nodes two years after vaccination. Macaques, immunized with SHIV-dn, were completely pro-tected from homologous challenge, whereas animals vacci-nated with SHIV-drn (2 of 4) and SHIV-dxrn (4 of 4) were infected with the challenge strain (detected by PCR). Mon-keys vaccinated with SHIV-NI expressed all the genes of SHIV NM-3rN, as well as rhesus INF- . All vaccinated mon-keys showed increases in cytokine production without a reduction in viral replication following challenge. Macaques that were PCR positive (SHIV NM-3rN) main-tained the presence of much lower viral loads that were sporadic and delayed compared to naïve animals. Results from this study also support the observation of an inverse relationship between attenuation and e Y cacy of live-atten-uated virus vaccines. SHIV-dn was then re-evaluated for the ability to protect non-human primates with the highly pathogenic SHIV89.6P[153]. Levels of SHIV89.6P were three to W vefold lower in vaccinated animals compared to naïve macaques without a decline in the number of CD4+ T cells that correlated with protection to SHIV89.6P. Interestingly, the vaccinated animals did not develop an immunode W-ciency disease even though they were infected with the challenge virus.Recent studies using live-attenuated SHIV-4 (deleted vpu and nef) indicated that these live-attenuated viruses admin-istered individually did cause a delayed onset to disease, but AIDS-like symptom did develop [141,154]. Monkeys vacci-nated with SHIV ppc (deleted vpu) did elicit long-lasting pro-tective immunity against challenge against pathogenic SHIV KU[155,156]. The success of the vpu deleted SHIV was most likely due to productive infection compared to the vpu/nef deleted SHIV-4. Combining these viruses into one vaccine regimen reduced viral load and disease。