the Peter Kim Lab
 
Screen Shot 2019-09-03 at 2.59.08 PM.png
 

 

Vaccines are one of the most profound ACCOMPLISHMENTS OF BIOMEDICAL SCIENCE.

Smallpox has been eradicated.  Three infectious agents (polio, measles, rubella) have been targeted for elimination by the World Health Organization and another seven have been at least 90% eradicated in the U.S. largely due to the efficacy of vaccines.  But it has not been possible to create vaccines for some of the most important infectious diseases.  We aim to create vaccines for such difficult targets and to devise new strategies to enable vaccine creation.  Our vaccine targets include HIV-1, pandemic influenza and Ebola. 

Our vaccine research is primarily focused on antibody-mediated immune responses.  As such, our work involves investigation of antigen-antibody interactions using biochemical, biophysical and structural biological methods.  We also interrogate human and murine naïve and antigen-stimulated B cell populations using fluorescence activated cell sorting (FACS) and single-cell sequencing methods.

We are also interested in the mechanism of viral membrane fusion.  Our focus is on the Class I membrane-fusion proteins, which includes those from HIV-1, influenza, Ebola and respiratory syncytial virus.  These proteins fold as a precursor that is then proteolytically processed, yielding a native form of the membrane-fusion protein that is trapped in a thermodynamically metastable state.  The energy stored in the metastable protein is used to overcome the activation barrier for membrane fusion – no external source of energy (e.g., ATP hydrolysis) is utilized.  In addition to providing mechanistic insights, studies of inhibitors of the membrane fusion process help to inform drug and vaccine development efforts.

We are also interested in establishing approaches to enable small-molecule drug discovery toward targets that have proven refractory (i.e., “undruggable” targets).  We are investigating the HIV-1 gp41 hydrophobic pocket and the cancer immunotherapy target PD-1.

 

 

Creating high-resolution, epitope-focused vaccines

Over recent decades, many monoclonal antibodies (mAbs) with potent and broad-spectrum neutralizing activity against certain viruses have been isolated and characterized.  Several of these broadly neutralizing mAbs (bnAbs) are being evaluated for their ability to prevent and/or treat infections in humans.  There are examples of bnAbs against many pathogens including HIV-1 and influenza.

The discovery of bnAbs reignited major efforts to create vaccines against important pathogens that have evaded conventional vaccine approaches, including HIV-1 and pandemic influenza.  Nonetheless, to date, it has not been possible to create a vaccine that elicits a bnAb-like response.  This is because bnAbs are rare – typically, the antibody responses to vaccines or infections are directed toward strain-specific and/or non‑neutralizing epitopes. 

Various “immunofocusing” strategies have been employed in attempts to elicit a polyclonal antibody response toward a specific epitope.  However, these approaches are generally “low-resolution” – that is, they focus the immune system to a region of the protein much larger than an antibody epitope.  This can lead to strain-specific antibodies that bind outside the epitope of interest

We recently introduced a method for creating epitope-focused vaccines called PMD (for protect, modify, deprotect).  The steps for PMD are: (i) protect an epitope on a protein by binding a bnAb, (ii) modify the bnAb­–antigen complex chemically to render solvent-exposed surfaces non-immunogenic, and (iii) deprotect the epitope by dissociating the bnAb–antigen complex. This produces an antigen in which the only unmodified region is the epitope of the bnAb (Fig. 1).  In our proof-of-concept study, using influenza hemagglutinin (HA), we demonstrated that PMD is capable of low-resolution epitope focusing. 

Importantly, in theory, PMD is capable of creating vaccines that focus an antibody response to the unprecedented level of individual amino acid residues – in an epitope specific manner.  In addition, PMD is readily generalizable – the key starting material is a mAb against the infectious agent.  We are working to enable creation of vaccines that elicit an antibody response against any given bnAb epitope, and only that epitope.

Figure 1. Schematic of the PMD strategy. First, the epitope is protected by combining the mAb (blue hashed) with the antigen (orange). Then the surfaces of the complex are modified to render them non-immunogenic (shown as darker shading). Finally the epitope is deprotected by removal of the mAb.

Figure 1. Schematic of the PMD strategy. First, the epitope is protected by combining the mAb (blue hashed) with the antigen (orange). Then the surfaces of the complex are modified to render them non-immunogenic (shown as darker shading). Finally the epitope is deprotected by removal of the mAb.

 

publication highlightS

Payton A. Weidenbacher and Peter S. Kim Proc. Natl. Acad. Sci. USA (2019) 116: 9947-9952. Protect, Modify, Deprotect (PMD): A strategy for creating vaccines to elicit antibodies targeting a specific epitope. (PDF)


 

Viral Membrane Fusion

Enveloped viruses are characterized by a lipid bilayer that surrounds the nucleocapsid. In order to infect a cell, the membrane surrounding the virus must fuse with the membrane surrounding the host cell (Fig. 2). This membrane-fusion process is mediated by virally encoded, transmembrane-anchored glycoproteins. We are interested in understanding the mechanism of viral membrane fusion.

Figure 2. Membrane fusion of an enveloped virus and its target cell.

Figure 2. Membrane fusion of an enveloped virus and its target cell.

 
 

publication highlights

Debra M. Eckert and Peter S. Kim Ann. Rev. Biochem. (2001) 70: 777-810. Mechanisms of Viral Membrane Fusion and Its Inhibition. (PDF)

Chavela M. Carr, Charu Chaudhry and Peter S. Kim Proc. Natl. Acad. Sci. USA (1997) 94: 14306-14313. Influenza Hemagglutinin is Spring-Loaded by a Metastable Native Conformation. (PDF)

Chavela M. Carr and Peter S. Kim Cell (1993) 73: 823-832. A Spring-Loaded Mechanism for the Conformational Change of Influenza Hemagglutinin. (PDF)


HIV-1

Figure 3. Schematic of membrane fusion mediated by the HIV-1 gp120/gp41 Envelope glycoprotein

Figure 3. Schematic of membrane fusion mediated by the HIV-1 gp120/gp41 Envelope glycoprotein

The mature HIV envelope glycoprotein consists of two parts:  a surface protein (SU) called gp120 and a transmembrane protein (TM) called gp41.  On the virus surface, these glycoproteins exist as a trimer of gp120/gp41 heterodimers (Fig. 3).

Interactions between gp120 on the virus and CD4 receptors on target cells mediate attachment of HIV to CD4+ T cells.  This binding event induces a conformational change in gp120 that facilitates additional interactions between gp120 with co-receptors (CCR5 or CXCR4), leading to a dramatic “spring-loaded” conformational change in gp41 (Fig. 3).  As a result, gp41 adopts a pre-hairpin intermediate conformation, in which the protein is associated with two membranes simultaneously: the host cell membrane via the “fusion peptide”, and the viral membrane via the transmembrane domain (Fig. 3). Interactions between the N-heptad repeat (NHR) and C-heptad repeat (CHR) regions (orange and blue, respectively, in Fig. 3) lead to the formation of a “trimer-of hairpins,” or six-helix bundle, which brings the two membranes together.

Structures of the native gp120/gp41 envelope glycoprotein and the final trimer-of-hairpins structure of gp41 are available. However, structural information for the pre-hairpin intermediate is lacking. We are interested in understanding what the pre-hairpin intermediate looks like, what stabilizes it, and what triggers the transition to complete membrane fusion.

publication highlights

David C. Chan and Peter S. Kim Cell (1998) 93: 681-684. HIV Entry and Its Inhibition. (PDF)

 

Inhibitors of HIV membrane fusion

Peptides that bind to the NHR region of the pre-hairpin intermediate (PHI) disrupt fusion by preventing formation of the trimer-of-hairpins. Such inhibitors include the C-peptides, derived from the CHR region (Fig. 4). Importantly, the FDA-approved HIV drug enfuvirtide (Fuzeon™) is a C-peptide.

A prominent pocket on the surface of the three-stranded coiled coil formed by the gp41 NHR region is a potential target for drugs that inhibit HIV infection by preventing formation of the trimer-of-hairpins. Using mirror-image phage display, we identified cyclic, D-peptide inhibitors of HIV infection that bind to this gp41 pocket (Fig. 4). These studies validate the pocket as a drug target.

Despite evidence suggesting that the gp41 pocket is an attractive drug target, small molecule ligands that bind with high affinity and high specificity have not been reported in the literature. The pocket therefore serves as an interesting model system for “undruggable” targets. We want to understand why it is difficult to identify small molecule ligands that bind to the gp41 pocket. We are also interested in ways to enhance the ability to discover such ligands.

 
Figure 4. The pre-hairpin intermediate (PHI). The N-heptad repeat (NHR) and C-heptad repeat (CHR) regions are indicated. Different inhibitors work by binding to various regions of the PHI, thereby preventing formation of the trimer-of-hairpins. (review: Eckert & Kim [2001] Ann. Rev. Biochem.)

Figure 4. The pre-hairpin intermediate (PHI). The N-heptad repeat (NHR) and C-heptad repeat (CHR) regions are indicated. Different inhibitors work by binding to various regions of the PHI, thereby preventing formation of the trimer-of-hairpins. (review: Eckert & Kim [2001] Ann. Rev. Biochem.)

In a manner complementary to inhibitors that bind to the NHR region, it is possible to inhibit fusion with molecules that bind to the CHR region of the PHI. For HIV, we accomplished this with a designed protein called 5-Helix, in which five of the six helices that make up the trimer-of-hairpins are connected with short peptide linkers (Fig. 4). Since the third CHR helix in the trimer-of-hairpins is missing, 5-Helix binds to the endogenous CHR region of gp41 and is a potent (nanomolar IC50) inhibitor of HIV infection.

publication highlights

Michael J. Root, Michael S. Kay and Peter S. Kim Science (2001) 291: 884-888. Protein Design of an HIV-1 Entry Inhibitor. (PDF)

Debra M. Eckert, Vladimir N. Malashkevich, Lily H. Hong, Peter A. Carr and Peter S. Kim Cell (1999) 99: 103-115. Inhibiting HIV-1 Entry: Discovery of D-Peptide Inhibitors that Target the gp41 Coiled-Coil Pocket. (PDF)

David C. Chan, Christine T. Chutkowski and Peter S. Kim Proc. Natl. Acad. Sci. USA (1998) 95: 15613-15617. Evidence that a Prominent Cavity in the Coiled Coil of HIV type 1 gp41 is an Attractive Drug Target. (PDF)


A common fusion mechanism

Figure 5. The trimer-of-hairpins is a common feature of diverse membrane-fusion proteins

Figure 5. The trimer-of-hairpins is a common feature of diverse membrane-fusion proteins

It appears that the membrane-fusion mechanism used by HIV-1 (Fig 3) is also used by many other enveloped viruses, such as influenza, respiratory syncytial virus (RSV), and Ebola. The membrane-fusion proteins for these viruses have been characterized as belonging to “Class I”. In some cases, these viruses have no apparent phylogenetic relationship. Three observations strongly support the notion that a common fusion mechanism is utilized by Class I viral membrane-fusion proteins: (i) a common trimer-of-hairpins structure, (ii) identification of C-peptide inhibitors for several of these viruses, and (iii) native protein structures that suggest a common spring-loaded feature.

The trimer-of-hairpins motif in these viral proteins consists of a characteristic core made up by a three-stranded coiled coil, presumably in place to present the fusion peptide at its tip (Fig 5).  At the base of the coiled coil, there is a loop structure that folds back to form a hairpin.  The structures of the polypeptide chains that pack outside of the coiled coil to form the hairpin are variable (Fig 5).

C-peptide inhibitors (see above) containing peptide sequences corresponding to the structures that pack against the central coiled coil of the trimer-of-hairpins (in HIV-1, the CHR region; see Fig 3) have been identified for several of these viruses.  Beyond HIV-1, such inhibitors have been identified (by other labs) for Ebola virus, avian sarcoma and leukosis virus type A (ASLV-A), human T-cell leukemia virus type 1 (HTLV-1), Nipah virus, Middle East respiratory syndrome coronavirus (MERS-CoV), severe acute respiratory syndrome coronavirus (SARS) and others.  The finding that C-peptide-derived peptides can inhibit membrane fusion in many viruses provides evidence that the pre-hairpin intermediate is a common feature in the mechanism used by Class I viral membrane-fusion proteins.

Figure 6. The trimer-of-hairpins formed by Ebola GP2

Figure 6. The trimer-of-hairpins formed by Ebola GP2

Through the efforts of multiple labs, native pre-fusion structures for several of the membrane-fusion proteins that form trimer-of-hairpins have been determined.  These include influenza HA1/HA2, parainfluenza F protein, Ebola GP1/GP2, respiratory syncytial virus (RSV) F glycoprotein, and HIV-1 gp120/gp41.  Strikingly, across these different cases, the residues that form the trimer-of-hairpins are in very different conformations in the native state. 

These findings provide evidence that a “spring-loaded” native state, first identified for influenza HA1/HA2, is a general feature shared by the Class I membrane-fusion proteins.

We are interested in understanding details about this common mechanism of membrane fusion.  In addition to HIV-1, we are studying the mechanism of Ebola virus invasion of cells. The Ebola GP2 protein forms a trimer-of-hairpins (Fig 6) and there is additional evidence that membrane fusion mediated by Ebola GP1/GP2 proceeds through a pre-hairpin intermediate.  

publication highlights

Debra M. Eckert and Peter S. Kim Proc. Natl. Acad. Sci. USA (2001) 98: 11187-11192. Design of Potent Inhibitors of HIV-1 Entry from the gp41 N-peptide Region. (PDF)


The gp41 pre-hairpin intermediate as a potential vaccine target

Figure 7. An HIV vaccine approach based on eliciting antibodies that bind to the pre-hairpin intermediate (Eckert et al. [1999] Cell).

Figure 7. An HIV vaccine approach based on eliciting antibodies that bind to the pre-hairpin intermediate (Eckert et al. [1999] Cell).

Figure 8. D5, an HIV-neutralizing antibody (Miller et al. [2005] PNAS), binds to the gp41 pocket (Luftig et al. [2006] Nature Struct. Mol. Biol.)

Figure 8. D5, an HIV-neutralizing antibody (Miller et al. [2005] PNAS), binds to the gp41 pocket (Luftig et al. [2006] Nature Struct. Mol. Biol.)

Earlier, we proposed that the pre-hairpin intermediate (PHI) was an attractive target for vaccine development. The specific goal is to elicit antibodies that bind to the PHI and thereby prevent HIV infection by inhibiting formation of the trimer-of-hairpins (Fig. 7). The PHI is an attractive potential target for vaccine development because it has been validated as a therapeutic target in humans with the anti-HIV drug, enfuvirtide (see above). In addition, the NHR region of the PHI (i.e., the target of enfuvirtide) is highly conserved among different HIV-1 strains, so antibody-escape mutants are predicted to be less frequent than for other regions of the gp120/gp41 envelope protein.

Because the PHI is transient, eliciting an immune response requires engineering of stable “mimetics” of the PHI to serve as immunogens. Using such PHI mimetics, we and others have elicited polyclonal antibody responses against the NHR region of the PHI. The resultant antisera inhibit HIV infection in cell culture, although the neutralization responses are weak. Monoclonal antibodies (mAbs) that bind to the gp41 NHR region and inhibit HIV infection in cell culture have been isolated by us and others. Crystal structures for a few of these mAbs have been determined and, in each case, the mAb binds to the prominent gp41 pocket (Fig. 8). As with the polyclonal antibody responses, the neutralization potencies of these mAbs are generally weak. We are pursuing different strategies to understand the immune response to PHI mimetics studied previously, and to create PHI mimetics that will elicit more strongly neutralizing antisera.

publication highlights

Elisabetta Bianchi, Joseph G. Joyce, Michael D. Miller, Adam C. Finnefrock, Xiaoping Liang, Marco Finotto, Paolo Ingallinella, Philip McKenna, Michael Citron, Elizabeth Ottinger, Robert W. Hepler, Renee Hrin, Deborah Nahas, Chengwei Wu, David Montefiori, John W. Shiver, Antonella Pessi, and Peter S. Kim Proc. Natl. Acad. Sci. USA (2010) 107: 10655-10660. Vaccination with peptide mimetics of the gp41 prehairpin fusion intermediate yields neutralizing antisera against HIV-1 isolates. (PDF)

Michael D. Miller, Romas Geleziunas, Elisabetta Bianchi, Simon Lennard, Renee Hrin, Hangchun Zhang, Meiqing Lu, Zhiqiang An, Paolo Ingallinella, Marco Finotto, Marco Mattu, Adam C. Finnefrock, David Bramhill, James Cook, Debra M. Eckert, Richard Hampton, Mayuri Patel, Stephen Jarantow, Joseph Joyce, Gennaro Ciliberto, Ricarrdo Cortese, Ping Lu, William Strohl, William Schleif, Michael McElhaugh, Steven Lane, Christopher Lloyd, David Lowe, Jane Osbourn, Tristan Vaughan, Emilio Emini, Gaetano Barbato, Peter S. Kim, Daria J. Hazuda, John W. Shiver, and Antonello Pessi Proc. Natl. Acad. Sci. USA (2005) 102: 14759-14764. A human monoclonal antibody neutralizes diverse HIV-1 isolates by binding a critical gp41 epitope. (PDF)


Cancer Immunotherapy

Recently, immune-checkpoint inhibitors, such as anti-PD-1 or anti-CTLA-4 monoclonal antibodies (mAbs), have shown dramatic effects in some cancer patients.  These drugs work by enhancing the endogenous anti-tumor activity of T-cells.  Unfortunately, inhibition of PD-1 and CTLA-4 can result in serious side effects that are exacerbated by the long half-lives of mAbs.  We are interested in discovering small molecules that target immune-checkpoint proteins, that would offer safety advantages resulting from their much shorter half-lives as compared to mAbs and possibly also offer efficacy advantages resulting from increased penetration and distribution within the tumor microenvironment.