Design of High Affinity Binders to Convex Protein Target Sites

While there has been progress in the de novo design of small globular miniproteins (50-65 residues) to bind to primarily concave regions of a target protein surface, computational design of minibinders to convex binding sites remains an outstanding challenge due to low level of overall shape complementarity. Here, we describe a general approach to generate computationally designed proteins which bind to convex target sites that employ geometrically matching concave scaffolds. We used this approach to design proteins binding to TGFβRII, CTLA-4 and PD-L1 which following experimental optimization have low nanomolar to picomolar affinities and potent biological activity. Co-crystal structures of the TGFβRII and CTLA-4 binders in complex with the receptors are in close agreement with the design models. Our approach provides a general route to generating very high affinity binders to convex protein target sites.

Inhibition of SARS-CoV-2 polymerase by nucleotide analogs from a single-molecule perspective

The absence of 'shovel-ready' anti-coronavirus drugs during vaccine development has exceedingly worsened the SARS-CoV-2 pandemic. Furthermore, new vaccine-resistant variants and coronavirus outbreaks may occur in the near future, and we must be ready to face this possibility. However, efficient antiviral drugs are still lacking to this day, due to our poor understanding of the mode of incorporation and mechanism of action of nucleotides analogs that target the coronavirus polymerase to impair its essential activity. Here, we characterize the impact of remdesivir (RDV, the only FDA-approved anti-coronavirus drug) and other nucleotide analogs (NAs) on RNA synthesis by the coronavirus polymerase using a high-throughput, single-molecule, magnetic-tweezers platform. We reveal that the location of the modification in the ribose or in the base dictates the catalytic pathway(s) used for its incorporation. We show that RDV incorporation does not terminate viral RNA synthesis, but leads the polymerase into backtrack as far as 30 nt, which may appear as termination in traditional ensemble assays. SARS-CoV-2 is able to evade the endogenously synthesized product of the viperin antiviral protein, ddhCTP, though the polymerase incorporates this NA well. This experimental paradigm is essential to the discovery and development of therapeutics targeting viral polymerases.

Abstract 6640: Stabilized recombinant trefoil factor 2 (TFF2-CTP) enhances anti-tumor activity of PD-1 blockade in mouse models of colorectal cancer

Despite remarkable responses to immune checkpoint blockade across multiple tumor types, the clinical benefit in colorectal cancer (CRC) is limited to microsatellite unstable tumors. PD-L1 expression is a negative prognostic marker in CRC but correlates with a better response to PD-1 blockade. Here, we investigated the role of PD-L1 in colorectal tumorigenesis and evaluated the utility of targeting myeloid-derived suppressor cells (MDSCs) in combination with PD-1 blockade in mouse models of CRC. We generated knockin mice that conditionally express the murine Pdl1 gene (R26-LSL-Pdl1-EGFP) and crossed them with LysM-Cre mice to overexpress PD-L1 specifically in the myeloid lineage. Mice received azoxymethane (AOM; 10 mg/kg i.p.) followed one week later with 2.5% dextran sodium sulfate (DSS) in the drinking water for 7 days. AOM/DSS-treated control mice formed tumors at 10 weeks and developed adenocarcinoma at 17 weeks post-AOM. LysM-Cre; R26-PD-L1 mice treated with AOM/DSS showed markedly enhanced colorectal tumorigenesis, with a significant increase in tumor number and size and enlarged spleen. In both groups, AOM/DSS treatment led to a significant expansion of myeloid cells, particularly CD11b+Gr-1+ MDSCs, in the colon and spleen, along with decreased NK cells compared with untreated mice. Notably, AOM/DSS-treated LysM-Cre; R26-PD-L1 mice showed decreased intratumoral CD8+ T cell infiltration compared to control tumors. Furthermore, there was a significant decrease in the percentage of Ki-67+ cells in intratumoral CD8+ T cells, indicating attenuated anti-tumor immunity. Trefoil factor 2 (TFF2), a secreted anti-inflammatory peptide, inhibits colon tumor growth by suppressing the expansion of CD11b+Gr-1+ MDSCs. TFF2 fused with two carboxyl-terminal peptide and three Flag motifs (TFF2-CTP-Flag) prolonged the circulation time in blood but retained bioactivity. We induced tumors in R26-PD-L1 mice with AOM/DSS, administered fusion recombinant TFF2-CTP-Flag (300 ug i.p.) and/or anti-PD-1 (RMP1-14; 200 ug i.p.) three times a week starting at 10 weeks and 14 weeks, respectively, and examined tumors at 18 weeks post-AOM. R26-PD-L1 mice treated with anti-PD-1 + TFF2-CTP showed a marked reduction in tumor growth while anti-PD-1 monotherapy failed to suppress growth. The combination of anti-PD-1 and TFF2-CTP significantly increased tumor-infiltrating CD8+ T cells and concomitantly decreased intratumoral regulatory T cells and CD11b+Gr-1+ myeloid cells. These early findings suggest that TFF2 augments the response rate of CRC to PD-1 blockade, possibly through suppressing MDSC expansion, supporting the potential of TFF2-CTP in combination I-O treatment for CRC. We are currently testing the efficacy of combined TFF2-CTP and anti-PD-1 therapy in the AOM/DSS model with PD-L1-overexpressing LysM-Cre; R26-PD-L1 mice.

Citation Format: Woosook Kim, Na Fu, Phaneendra Duddempudi, Zinaida Dubeykovskaya, Steven Almo, Chandan Guha, Seth Lederman, Timothy Wang. Stabilized recombinant trefoil factor 2 (TFF2-CTP) enhances anti-tumor activity of PD-1 blockade in mouse models of colorectal cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6640.

HVEM signaling promotes protective antibody-dependent cellular cytotoxicity (ADCC) vaccine responses to herpes simplex viruses

Herpes simplex virus (HSV) glycoprotein D (gD) not only is required for virus entry and cell-to-cell spread but also binds the host immunomodulatory molecule, HVEM, blocking interactions with its ligands. Natural infection primarily elicits neutralizing antibodies targeting gD, but subunit protein vaccines designed to induce this response have failed clinically. In contrast, preclinical studies demonstrate that an HSV-2 single-cycle strain deleted in gD, ΔgD-2, induces primarily non-neutralizing antibodies that activate Fcγ receptors (FcγRs) to mediate antibody-dependent cellular cytotoxicity (ADCC). These studies were designed to test the hypothesis that gD interferes with ADCC through engagement of HVEM. Immunization of Hvem-/- mice with ΔgD-2 resulted in significant reduction in HSV-specific IgG2 antibodies, the subclass associated with FcγR activation and ADCC, compared with wild-type controls. This translated into a parallel reduction in active and passive vaccine protection. A similar decrease in ADCC titers was observed in Hvem-/- mice vaccinated with an alternative HSV vaccine candidate (dl5-29) or an unrelated vesicular stomatitis virus-vectored vaccine. Unexpectedly, not only did passive transfer of immune serum from ΔgD-2-vaccinated Hvem-/- mice fail to protect wild-type mice but transfer of immune serum from ΔgD-2-vaccinated wild-type mice failed to protect Hvem-/- mice. Immune cells isolated from Hvem-/- mice were impaired in FcγR activation, and, conversely, addition of gD protein or anti-HVEM antibodies to in vitro murine or human FcγR activation assays inhibited the response. These findings uncover a previously unrecognized role for HVEM signaling in generating and mediating ADCC and an additional HSV immune evasion strategy.

Biochemical and Functional Characterization of the ICOSL TNFR2 Interaction

To evade the immune response, cancers hijack the anti-inflammatory function of T regulatory Cells (Tregs) through blockade of anti-tumor responses carried out by T cells (effector and cytotoxic). Tumor necrosis factor receptor 2 (TNFR2) positive Tregs are highly immunosuppressive and enriched in the tumor microenvironment. Thus, immunotherapies targeting TNFR2 with antagonistic or agonistic antibodies are promising therapeutics for decreasing or increasing Treg cell levels, respectively. TNFR2 is activated by engagement with the cytokine, TNF alpha (TNFa), which leads to cell survival and proliferation through activation of the NFkB pathway, and this is extensively studied in terms of Treg cell proliferation and survival. Using a high-throughput cell-cell based screening approach, the Almo lab identified ICOS ligand (ICOSL) as a novel interacting partner of TNFR2. ICOSL is a costimulatory molecule with wide tissue distribution and with the ability to bind type-I transmembrane proteins CD28, CTLA-4 and ICOSR. ICOSL interactions with these receptors have been implicated in both positive and negative regulation of T cell activity. While TNFa binding to TNFR2 can activate NFkB signaling, it is unclear if ICOSL binding serves a blocking or activating function. We hypothesize the ICOSL-TNFR2 interaction functions as an additional mechanism for activating/stimulating TNFR2+ Treg cells through processes such as survival, cytokine production and expansion.

Structural and Functional Investigations Into B7-1:NGFR

Although Nerve Growth Factor Receptor (NGFR, TNFSFR16) is traditionally thought of as a receptor that regulates the growth of neurons, it is also widely expressed throughout the immune system in response to inflammation where its function is poorly characterized. In addition to binding the four neurotrophins (NGF, BDNF, NT3, NT4), the Almo lab and others have recently discovered that NGFR can also act as a receptor for B7-1 (CD80). B7-1 is a major costimulatory molecule expressed on antigen presenting cells (APCs) such as dendritic cells, macrophages, and microglia in response to inflammation. B7-1 can bind to CD28 on naïve T-cells, which induces T-cell activation. It can then bind CTLA4 (upregulated on T cells after CD28 activation), which reduces activation. B7-1 also interacts with PD-L1. The structure and function of the B7-1-NGFR interaction is currently unknown. We have recently conducted epitope mapping experiments that suggest the NGFR binds to B7-1 on the same interface that CD28 and CTLA4 bind to B7-1. This interface appears to be independent of the PD-L1-B7-1 binding interface. NGFR epitope mapping data suggests that B7-1 binds to NGFR near the binding site for NGF, and similarly to other TNFRSF receptor-ligand interactions. Furthermore, our data shows that NGFR and CTLA4 can compete for binding to B7-1. In addition, we are investigating the role the B7-1:NGFR binding plays in neurons.

Development of high-throughput methods used to identify and characterize novel interactions within the human secretome with focus on the Ig and TNF receptor superfamilies

While it is estimated that roughly 15% of the human genome encodes secreted proteins, many remain “orphans” with no known ligand or are poorly characterized. Even for those with known ligands, new binding partners continue to be identified that fundamentally reshape our understanding of their role in specific biological processes and diseases (i.e. PD-L1:B7-1, ICOS-L:CD28, PD-L2:RGMb). This highlights the need for systematic screening efforts to identify such interactions, broadening our general knowledge while also providing potential new targets for therapeutic intervention. To this end we have developed a set of novel technologies using both cell microarray and high-throughput flow cytometry techniques to screen for novel extracellular protein: protein interactions. Initially we focused on screening for interactions within the Ig and TNF receptor superfamilies, two protein families that are the central underpinning of our immune response but for which much is still unknown. We have generated two expression-validated libraries for cell-based screening, a 366 member Type I expression library of Ig and TNF receptor proteins and a ~3800 member full-length human plasma membrane protein library encompassing much of the human membrane proteome. Through rounds of screening we identified novel interactions for B7-1 (CD80), ICOS-L, TrkA and TrkC. The identified interactions were validated biochemically and are being characterized for their biological significance. Additionally we are using these technologies to characterize the binding interfaces of both known and newly discovered protein complexes as well as to screen for potential off-target interactions of protein-based therapeutics.

A Herpes Simplex Virus (HSV)-2 Single-Cycle Candidate Vaccine Deleted in Glycoprotein D Protects Male Mice From Lethal Skin Challenge With Clinical Isolates of HSV-1 and HSV-2

Herpes simplex virus (HSV) infections manifest as recurrent oral or genital mucosal lesions, meningoencephalitis, corneal blindness, and perinatal disease. Subunit vaccines have advanced into the clinic without success. None were tested preclinically in male mice. We compared a single-cycle candidate vaccine deleted in HSV-2 glycoprotein D (ΔgD-2) and subunit gD-2 or gD-1 protein vaccines in a male murine skin model. The ΔgD-2 provided complete protection against 10 times the lethal dose of HSV-1 or HSV-2 clinical isolates, and no latent virus was detected, whereas gD-1- and gD-2-adjuvanted proteins provided little or no protection. Protection correlated with Fc receptor activating but not neutralizing antibody titers.

Abstract B021: Single domain antibodies targeting CTLA-4 demonstrate a critical role for Fc in the antitumor response

Monoclonal antibodies targeting CTLA-4 are now a central part of the standard of care for metastatic melanoma, and have shown efficacy against diverse tumor types. However, the extent to which direct blockade of CTLA-4 contributes to the efficacy of these antibodies remains incompletely understood. Mice lacking FcγRIV fail to respond to αCTLA-4 therapy, and maintain robust tumor infiltrating Treg populations. These findings suggest that anti-CTLA-4 antibody therapy enhances antitumor responses through Treg depletion, though a requirement for FcγR in other aspects of the antitumor response has not been excluded, and the importance of CTLA-4 blockade alone has not be directly assessed. We have developed an alpaca-derived single domain antibody (H11) that binds CTLA-4 with high affinity, occluding the B7 binding motif as determined by x-ray crystallography. We are able to use this single domain antibody in 18F immuno-PET to produce a comprehensive image of CTLA-4 expression in a tumor-bearing live mouse, confirming penetration of H11 into the tumor microenvironment. Therapeutically, we find that H11 potently blocks CTLA-4 co-inhibition in vitro, but fails to elicit protective immunity in vivo. Therapeutic activity is restored when H11 is attached to the Fc region of murine IgG2a, but not when half-life alone is increased through H11 dimerization or conjugation to PEG, confirming a central role for Fc-dependent function in αCTLA-4 responses, as well as the minimal efficacy of CTLA-4 blockade alone. We further find that the efficacy of αCTLA-4 therapy can be enhanced through blockade of CD47, consistent with an important role for Treg depletion by myeloid cells in the activity of this therapy. These findings have implications for the development of combinations immune therapies.

Citation Format: Michael Dougan, Jessica Ingram, Olga Blomberg, Mohammad Rashidian, Edmund Keliher, Ralph Weissleder, Steven Almo, Hidde Ploegh. Single domain antibodies targeting CTLA-4 demonstrate a critical role for Fc in the antitumor response [abstract]. In: Proceedings of the Second CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; 2016 Sept 25-28; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2016;4(11 Suppl):Abstract nr B021.

A novel approach to generate LIGHT mutants with selectivity to binding partners (TECH2P.925)

LIGHT is a member of TNF superfamily, which participates in interactions with many TNF receptors, including membrane-anchored HVEM and LTβR, as well as solely soluble DcR3. Interaction of LIGHT with HVEM and LTβR directly triggers intracellular signaling, which regulates a wide range of immune response, such as T cell proliferation, thymocyte differentiation and lymphocyte trafficking and migration. LIGHT:DcR3 interaction tends to attenuate the immune response. Besides LIGHT, DcR3 also neutralizes two critical inflammatory molecules TL1A and FasL. The mechanisms that support the broad specificity of LIGHT need to be defined and the functions of LIGHT interacting with different receptors await to be dissected. We have solved the crystal structures of LIGHT:HVEM and LIGHT:DcR3 complexes, revealing the critical residues involved in the interactions of LIGHT with HVEM and DcR3. We have also developed a novel approach to generate LIGHT mutants with altered binding selectivities to different receptors, which allow the further dissection of the functions of LIGHT. This novel method creates LIGHT mutants by transplanting the loops responsible for receptor recognition from TL1A to the corresponding positions of LIGHT. The generated LIGHT chimeras still possesses the wild-type like binding abilities with DcR3 and LTβR, but dramatically compromise the binding affinities with HVEM. Applications of the LIGHT chimeras reveal cooperative signaling of LIGHT through HVEM and LTβR.

Orange Fluorescent Proteins: Filling the Spectral Gap

Fluorescent proteins (FPs) have become valuable tools for molecular biology, biochemistry, medicine, and cancer research. Starting from parent green fluorescent protein (GFP), most challenging task of the FPs studies was the development of FPs with longer excitation/emission wavelength. This pursuit was motivated by advantages of so-called red-shifted FPs, namely, lower background of cellular autofluorescence in microscopy, lower light scattering and reduced tissue absorbance of longer wavelengths for in vivo imaging. In addition to FPs with regular spectral properties, there are proteins of other types available, including FPs with a large Stokes shift and photoconvertible FPs. These special kinds of FPs have become useful in super-resolution microscopy, imaging of enzyme activities, protein-protein interactions, photolabeling, and in vivo imaging. According to their emission wavelength, red-shifted FPs could be divided in the following groups: 520-540 nm yellow FPs (YFPs), 540-570 nm orange FPs (OFPs), 570-620 nm red FPs (RFPs), and > 620 nm far-RFPs. Red shift of the excitation/emission bands of these FPs is predominantly achieved by extension of the conjugated system of the chromophore and its protonation/deprotonation. The variety of spectral properties of FPs (excitation and emission wavelength, quantum yield, brightness, photo- and pH- stability, photoconversion, large Stokes shift, etc) results from the different chromophore structures and its interactions with surrounding amino acid residues. In this work we focus on structural studies and molecular mechanisms of FPs with orange emission.

Unusual antigen recognition provides clue to pathogenicity of diabetogenic CD8 T cell (APP2P.101)

The defined length restriction as well as preference for amino acids at particular positions for each MHC allele is imposed by the physical as well as chemical nature of the peptide binding groove of the MHC molecule. Accordingly, the preferred peptide for the mouse MHC class I allele H-2Db is a 9mer with anchor residues N at P5 and M/I or L at position 9. However, in this study, we found that AI4, one of the most pathogenic CD8 T cell clones that participates in type 1 diabetes pathogenesis, recognizes an unusual 7mer peptide from the β cell specific protein insulin (InsA14-20) in the context of H-2Db. Further investigation showed that a majority of NOD mice had islet infiltrating T cells that recognized InsA14-20 or mimotopes derived from it, making the AI4-like T cells a very prevalent population. Crystallographic analysis revealed that InsA14-20 shares two solvent-exposed residues with all previously identified AI4 ligands which could explain the promiscuous nature of the AI4 T cell receptor. In addition, accessory anchoring residues were observed to compensate for the lack of the C-terminal anchor residue. In sum, the study demonstrates that less than ideal peptide-MHC interaction can still trigger T cell reactivity and contribute to disease pathogenesis, aided further by lack of thymic negative selection due to weak binding affinity.

Crystal structures of FasL:DcR3 and LIGHT:DcR3 complexes reveal the molecular basis for the broad specificity of DcR3 (CCR5P.255)

Decoy receptor 3 (DcR3) is a secreted TNF receptor discovered in human beings but not in mice. DcR3 is able to neutralize three TNF ligands, LIGHT, FasL and TL1A and further modulate the immune response. For example, neutralization of LIGHT by DcR3 can manipulate the co-stimulatory signal triggered by LIGHT:HVEM interaction or the cell death signal resulted from LIGHT:LTβR interaction. Upon ligation with FasL, DcR3 can block the cell apoptosis pathway commenced by FasL:Fas interaction. Besides, DcR3 binds to TL1A, which can interrupt its interaction with DR3 to boost T cell immunity or induce cell apoptosis in different cell types. Although the remarkable ability of DcR3 to neutralize three different ligands has already been realized, the mechanisms which support the broad specificity of DcR3 remain to be fully defined. Here I present the novel crystal structures of the FasL:DcR3 and LIGHT:DcR3 complexes. Along with the TL1A:DcR3 structure solved in our lab previously, the whole set of the structures of TNF:TNFR complexes engaged by DcR3 was solved. Our results show that the loops of the ligands are the major determinants for the recognition of DcR3. By appropriate conformational alterations of the flexible loops but not the rigid beta-strand sheets, the diverse residues on the loops of different ligands are able to capture the cognate residues on DcR3 by polar contacts and hydrophobic interactions, thus enable the different TNF ligands to bind to the same DcR3.

Biochemical and structural studies reveal a canonical mode of molecular recognition between TIGIT and nectin-2 (IRM10P.735)

Besides antigen-specific signal from the interaction between T-cell receptor and peptide-MHC complex expressed on antigen presenting cells (APCs), the process of optimal T-cell activation and differentiation requires antigen-independent costimulatory signals. Costimulatory signals are provided by interaction between a number of activating as well as inhibitory receptors expressed on T cells with specific ligands expressed on APCs. The balance between activating and inhibitory signals critically controls the mammalian immune responses. Therefore, understanding the mechanisms and designing corrective therapeutics involving T-cell costimulation are imperative in both infectious as well as autoimmune diseases. TIGIT (T cell immunoglobulin and ITIM domain), a recently identified immune receptor expressed on T and NK cells, upon interaction with either of its two ligands, nectin-2 or PVR, that are over-expressed on certain tumors, inhibits activation of T and NK cells. Several biophysical studies, including X-ray crystallography suggest that TIGIT may exist in its monomeric as well as dimeric forms. Based on the structural and biochemical studies, we mapped the nectin-binding interface on TIGIT. Interestingly, structure-guided mutations that disrupt the interaction between nectin-2 and TIGIT also limit the homodimerization of TIGIT in solution. Our data provide structural insights into how TIGIT recognizes a cell-adhesion molecule nectin-2 to execute immunoregulatory function.

Exploration and blockade of the decoy strategy: crystal structure of the LIGHT:DcR3 complex and rational design of DcR3 antagonists (P1067)

LIGHT is a member of the tumor necrosis factor (TNF) family that initiates intracellular signaling via engagement of the two TNF receptors HVEM and Lymphotoxin beta Receptor (LTbetaR). In humans, LIGHT is neutralized by a soluble TNF receptor denoted as decoy receptor 3 (DcR3), which tightly binds LIGHT and inhibits its interactions with HVEM and LTbetaR, resulting in blockade of the corresponding signaling pathways. In addition to LIGHT, DcR3 also binds and neutralizes two other TNF ligands, FasL and TL1A, thus further modulating the immune response. Because of its remarkable ability to neutralize three different ligands, DcR3 is thought to play an important role in regulating the immune response, especially in cancer immunology as suggested by elevated expression levels and the associated poor prognosis in cancer patients. However, the mechanisms that support the broad specificity of DcR3 (i.e., recognition of three different TNF ligands) remain to be fully defined. We have determined the structure of the LIGHT:DcR3 complex, revealing the structural basis for the DcR3-mediated neutralization of LIGHT. Comparison with our previously reported structure of the TL1A:DcR3 complex affords new insights into DcR3 function and selectivity. Based on these structures, we designed a novel LIGHT mutant that specifically recognizes DcR3 and exhibits no interaction with HVEM; this reagent dissects the multiple biochemical functions of LIGHT and offers unique clinical opportunities.

Structure of Nectin-2 reveals determinants of homophilic and heterophilic interactions that control cell-cell adhesion and immune regulation (176.10)

Nectin comprises a family of four immunoglobulin-like molecules. Homophilic and heterophilic interactions among nectins are implicated in cell-cell adhesion, while their interactions with members of other protein families have diverse biological functions like host-pathogen interaction and immune modulation. In particular, nectin-2, found to be up-regulated on cancer cells is capable of interacting with two receptors, CD226 and TIGIT, expressed on T and NK cells. These interactions lead to the delivery of two opposing signals to both T and NK cells. This situation is reminiscent of the well-studied pathways in T cells, in which the coinhibitory receptor CTLA-4 binds the same ligand (B7) as the coactivating receptor CD28. In order to define the molecular and structural determinants underlying the homophilic and heterophilic recognitions of nectin-2, we examined the biochemical, biophysical and structural properties of human nectin-2. The structure of nectin-2 at 1.3 Å resolution reveals that the architecture of the nectin-2 homophilic dimer resembles other members of the immunoglobulin superfamily and defines the details responsible for recognition and selectivity. Of particular note, the close proximity of charged residues at the interface is a major determinant of binding affinity. Using these biochemical and structural data, we also characterized the heterophilic binding of nectin-2 with TIGIT, which is implicated in T cell and NK cell-mediated immune modulation.

Structural and functional studies on PD-1: from crystals to mice (63.17)

Programmed death-1 (PD-1) inhibits T and B cell responses upon binding to its ligands, PD-L1 and PD-L2, and has crucial role in maintaining peripheral tolerance. We have recently reported the crystal structure of the complex between PD-1 and PD-L2. Performing structure-based mutagenesis studies, high-affinity, and PD-L2-specific PD-1 mutants were identified. High-resolution crystal structures of the high-affinity PD-1 mutant with PD-L1 and PD-L2, respectively, provide the atomic details of the tighter receptor-ligand binding. The crystal structure of the complex between the PD-L2-specific PD-1 mutant and PD-L2 provides the atomic details of the selective interaction of this mutant with PD-L2. Soluble Ig fusion proteins of the two PD-1 mutants were designed to modulate the PD-1 pathway for immunotherapy. Finally, these mutations were introduced into mice by generating “knock-in” mouse strains, which provide novel animal models to study the role of PD-1 in immunity. Current work is focused on analyzing the functional effects of these mutations in PD-1 mediated immune responses. In summary, our data demonstrate the value of structural studies in developing novel immunotherapy and animal models for in vivo studies.

The role of Programmed Death-1 (PD-1) in pulmonary M. tuberculosis infection: friend or foe? (40.10)

The programmed death-1 (PD-1) co-inhibitory receptor attenuates T and B cell responses and plays a crucial role in peripheral tolerance. Recently, PD-1 has been shown to inhibit T cell responses during chronic viral infections such as LCMV and HIV, and blockade of PD-1 restored anti-viral effector T cell responses. In this study, we examined the role of PD-1 in infection with Mycobacterium tuberculosis, a common co-infection with HIV. Surprisingly and paradoxically, PD-1 deficient mice showed dramatically reduced survival compared to wild-type mice. Moreover, lungs of the PD-1-/- mice showed uncontrolled bacterial proliferation, and focal necrotic areas with predominantly neutrophilic infiltrates, but a lower number of infiltrating T and B cells. Pro-inflammatory cytokines, such as TNF-α, IL-1, and especially IL-6 were significantly increased in the lung and sera of infected PD-1-/- mice, consistent with an excessive inflammation. Microarray analysis of the lungs infected with M. tuberculosis showed dramatic differences between PD-1-/- and control mice: 297 genes were differentially expressed, resulting in profoundly altered inflammatory responses, with implications in both innate and adaptive immunity. Our studies demonstrate that M. tuberculosis infection in the absence of PD-1 results in fatal inflammation with increased bacterial load; therefore we suggest that the PD-1 pathway is involved in controlling inflammatory responses after M. tuberculosis infection in the lungs.

Cellular localization of activated N-WASP using a conformation-sensitive antibody

The main regulators of Arp2/3 activity appear to be N-WASP and the other members of the Scar/WAVE family of proteins. We show here that after EGF stimulation, N-WASP is recruited to the nucleation zone of the dynamic leading edge compartment of carcinoma cells, with maximal recruitment of N-WASP within 1 min after EGF stimulation. The timing of N-WASP recruitment mirrors the timing of barbed-end formation at the leading edge. To determine the cellular activation of N-WASP after EGF stimulation, we made a conformation-sensitive antibody (CSA) against the CRIB domain of N-WASP that is predicted to recognize N-WASP in its open, active conformation, but not in its closed, inactive conformation. The ability of CSA to detect only active N-WASP was demonstrated by in vitro experiments using immunoprecipitation of active N-WASP from EGF-stimulated cells and Cdc42 activation of N-WASP activity. In cell staining experiments, N-WASP is maximally accessible to CSA 40 sec after EGF stimulation and this activated N-WASP is in the nucleation zone. These results indicate that active N-WASP is present at the leading edge of lamellipods, an unexpected finding given its reported involvement in filopod formation. This work establishes the feasibility of using antibodies directed against specific conformations or epitopes with changing accessibilities as a window on the status and localization of activity.

Synchrotron protein footprinting: a technique to investigate protein-protein interactions

Traditional approaches for macromolecular structure elucidation, including NMR, crystallography and cryo-EM have made significant progress in defining the structures of protein-protein complexes. A substantial number of macromolecular structures, however, have not been examined with atomic detail due to sample size and heterogeneity, or resolution limitations of the technique; therefore, the general applicability of each method is greatly reduced. Synchrotron footprinting attempts to bridge the gap in these methods by monitoring changes in accessible surface areas of discrete macromolecular moieties. As evidenced by our previous studies on RNA folding and DNA-protein interactions, the three-dimensional structure is probed by examining the reactions of these moieties with hydroxyl radicals generated by synchrotron X-rays. Here we report the application of synchrotron footprinting to the investigation of protein- protein interactions, as the novel technique has been utilized to successfully map the contact sites of gelsolin segment-1 in the gelsolin segment 1/actin complex. Footprinting results demonstrate that phenylalanine 104, located on the actin binding helix of gelsolin segment 1, is protected from hydroxyl radical modification in the presence of actin. This change in reactivity results from the specific protection of gelsolin segment-1, consistent with the substantial decrease in solvent accessibility of F104 upon actin binding, as calculated from the crystal structural of the gelsolin segment 1/actin complex. The results presented here establish synchrotron footprinting as a broadly applicable method to probe structural features of macromolecular complexes that are not amenable to conventional approaches.

Molecular recognition of human CD1b antigen complexes: evidence for a common pattern of interaction with alpha beta TCRs

Ag-specific T cell recognition is mediated through direct interaction of clonotypic TCRs with complexes formed between Ag-presenting molecules and their bound ligands. Although characterized in substantial detail for class I and class II MHC encoded molecules, the molecular interactions responsible for TCR recognition of the CD1 lipid and glycolipid Ag-presenting molecules are not yet well understood. Using a panel of epitope-specific Abs and site-specific mutants of the CD1b molecule, we showed that TCR interactions occur on the membrane distal aspects of the CD1b molecule over the alpha1 and alpha2 domain helices. The location of residues on CD1b important for this interaction suggested that TCRs bind in a diagonal orientation relative to the longitudinal axes of the alpha helices. The data point to a model in which TCR interaction extends over the opening of the putative Ag-binding groove, making multiple direct contacts with both alpha helices and bound Ag. Although reminiscent of TCR interaction with MHC class I, our data also pointed to significant differences between the TCR interactions with CD1 and MHC encoded Ag-presenting molecules, indicating that Ag receptor binding must be modified to accommodate the unique molecular structure of the CD1b molecule and the unusual Ags it presents.

An atomic model of fimbrin binding to F-actin and its implications for filament crosslinking and regulation

Using a new procedure that combines electron-density correlation with biochemical information, we have fitted the crystal structure of the N-terminal actin-binding domain of human T-fimbrin to helical reconstructions of fimbrin-decorated actin filaments. The map locates the N-terminal calcium-binding domain and identifies actin-binding site residues on the two calponin-homology domains of fimbrin. Based on this map, we propose a model of a fimbrin crosslink in an actin bundle and its regulation by calcium.

The immunoglobulin E-allergen interaction: a target for therapy of type I allergic diseases

The interaction of immunoglobulin E and otherwise harmless antigens (allergens) leads in sensitized individuals through effector cell activation to the immediate induction of a cascade of inflammatory reactions, the hallmark of type I allergy. Recently, the molecular and structural characterization of allergens, specific IgE antibodies and their epitopes has made rapid progress. Here we discuss active and passive strategies for therapy of type I allergy, which are based on interfering with the IgE-allergen interaction.