top of page




Pancreatic cancer has an extremely high mortality rate due to its aggressive metastatic nature. Resolving the underlying mechanisms will be crucial for treatment. After finding that overexpression of Interleukin-17B receptor (IL-17RB) strongly correlated with postoperative metastasis and inversely correlated with progression-free survival in pancreatic cancer patients (2013), we identified a newly derived monoclonal antibody (D9) against IL-17RB which blocked tumor metastasis and promoted survival in a mouse xenograft model (2015). We further determined the complex structure of IL-17RB bound by Fab of D9, which revealed an important loop region, clearly defined as the epitope for D9, in an unusual conformation (2019). This conformation presumably disrupts the interaction between ligand IL-17B and receptor IL-17RB, and prevents the receptor dimerization that is required for activation of the IL-17B/IL-17RB signaling. Humanization and affinity maturation of D9 resulted in an improved antibody 1B12 for potential therapeutic applications.


In addition to cancer-specific interleukin receptors, the change in glycosylation pattern can also be a distinct feature of cancer cells. Globo-series glycolipids, in particular the stage-specific embryonic antigen (SSEA4), have been found to be highly expressed on the surface of almost all cancer cells, including brain, lung, breast, esophagus, stomach, liver, bile duct, pancreas, colon, kidney, cervix, ovary and prostate. We have recently determined the crystal structures of a mouse IgG3 monoclonal antibody MC813 in complex with SSEA4, as well as a human IgG1 monoclonal antibody isolated from a cancer patient (unpublished), also in complex with SSEA4 (2019).  MC813 can suppress SSEA4+ brain tumor growth in vivo possibly through both complement-dependent cytotoxicity and antibody-dependent cellular cytotoxicity (ADCC). The structural features and the different binding modes provide basis for understanding the anti-cancer effects carried out by these antibodies.


A major challenge to end the pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is to develop a broadly protective vaccine that elicits long-term immunity. As the key immunogen, the viral surface spike (S) protein is frequently mutated, and conserved epitopes are shielded by glycans. Here, we revealed that S protein glycosylation has site-differential effects on viral infectivity. The S protein from lung epithelial cells contained more sialylated complex-type glycans to facilitate receptor binding, and glycosites N801 and N1194 were shown to be essential for S protein folding and viral infection. We found that S protein generated by lung epithelial cells has glycoforms associated with increased infectivity. The analysis of cell-specific glycoform distribution, sequence conservation, glycan shielding, and their mutual correlations, led to the design of SMG vaccine, in which essentially all glycan shields are removed. This made the conserved epitopes better exposed to the immune system for more effective B cell and T cell responses. Compared to the fully glycosylated S protein, immunization of S protein with N-glycans trimmed to the mono-GlcNAc-decorated state (SMG) elicited stronger immune responses and better protection for human angiotensin converting enzyme 2 (hACE2) transgenic mice against variants of concern (VOCs). In addition, a broadly neutralizing monoclonal antibody was identified from SMG immunized mice that could neutralize wild type (WT) SARS-CoV-2 and VOCs with sub-picomolar potency. As illustrated by this antibody, the conserved epitopes targeted by such antibodies could be used for next-generation vaccine development. Together, these results demonstrate that removal of glycan shields to better expose the conserved sequences has the potential to be an effective and simple approach for developing a broadly protective SARS-CoV-2 vaccine. In an effort to develop vaccines against the currently circulating and future SARS-CoV-2 variants, the impact of glycosylation on viral infection, protein integrity, immune response and vaccine design as demonstrated in this study should be considered together with other parameters.


While the annual mortality of seasonal influenza is about half a million death, the ongoing catastrophic COVID-19 pandemic has killed more than 400,000 globally within 5 months since late January 2020. The influenza virus hemagglutinin (HA) and coronavirus spike (S) protein mediate virus entry. HA and S proteins are heavily glycosylated, making them potential targets for carbohydrate binding agents such as lectins. Here, we show that the lectin FRIL, isolated from hyacinth beans (Lablab purpureus), has anti-influenza and anti-SARS-CoV-2 activity. FRIL can neutralize 11 representative human and avian influenza strains at low nanomolar concentrations, and intranasal administration of FRIL is protective against lethal H1N1 infection in mice. FRIL binds preferentially to complex-type N-glycans and neutralizes viruses that possess complex-type N-glycans on their envelopes. As a homotetramer, FRIL is capable of aggregating influenza particles through multivalent binding and trapping influenza virions in cytoplasmic late endosomes, preventing their nuclear entry. Remarkably, FRIL also effectively neutralizes SARS-CoV-2, preventing viral protein production and cytopathic effect in host cells. These findings suggest a potential application of FRIL for the prevention and/or treatment of influenza and COVID-19.


Our results allowed us to put forth a model for FRIL’s anti-influenza action as below: FRIL first binds and extracellularly cross-links virions, which results in either large aggregates rapidly cleared by the host immune system, or the FRIL-virus complex is endocytosed into host cells. The FRIL-bound virus is subsequently retained in the late endosome/lysosome and prevented from nuclear entry, until its ultimate degradation 24 hours post infection.


Although detailed mechanistic studies are still ongoing for SARS-CoV-2 infections, our preliminary data have showed strong evidence that FRIL is capable of neutralizing this virus according to immuno-fluorescence tracking of N and spike protein of SARS-CoV-2 (hCoV-19/Taiwan/04/2020 strain) in green and cyan colors on the right. Production of viral proteins (N and spike) starts at 4 hours post infection and peaks at 24 hours without FRIL (PBS groups), while cells of the FRIL groups are almost virus-free. Nuclei is counterstained with DAPI (blue).



Influenza viruses continue to cause annual epidemics and pose the threat of a deadly global pandemic. Vaccination is the most effective method to control and prevent the morbidity and mortality resulting from influenza virus infections in humans. However, influenza virus is constantly changing and the current flu vaccines are only effective against closely matched circulating strains. As a consequence, flu vaccine needs to be reformulated every year. Over the years we have pioneered an improved flu vaccine design using the monoglycosylated form of the major surface glycoprotein hemagglutinins (HAs). Understanding how glycans on HAs affect the immune response and knowledge of how broadly neutralizing antibodies are induced will pave the way for a cross-protective influenza vaccine that does not require frequent updates or annual immunizations. Trimming down the structurally non-essential glycans on HA may be the key for a better flu vaccine.


Recombinant monoglycosylated hemagglutinin (HAmg) with an intact protein structure from either seasonal or pandemic H1N1 can be used as a vaccine for cross-strain protection against various H1N1 viruses in circulation from 1933 to 2009 in mice and ferrets. In the HAmg vaccine, highly conserved sequences that were originally covered by glycans in the fully glycosylated HA (HAfg) are exposed and thus, are better engulfed by dendritic cells (DCs), stimulated better DC maturation, and induced more CD8+ memory T cells and IgG-secreting plasma cells. Single B-cell RT-PCR followed by sequence analysis revealed that the HAmg vaccine activated more diverse B-cell repertoires than the HAfg vaccine and produced antibodies with cross-strain binding ability (2014). The HAmg vaccine elicits cross-strain immune responses that may not only mitigate the current need for yearly reformulation of strain-specific inactivated vaccines, but also map a new direction for universal vaccine design. We further provided a simple and practical procedure to develop monoglycosylated influenza vaccine using the current egg-based process (2019).



Transpeptidase (TP) and transglycosylase (TG) on the bacterial cell surface are essential for cell wall synthesis, and many antibiotics have been developed to target the transpeptidase; however, the problem of antibiotic resistance has inevitably arisen and caused a major threat in treating bacterial infection. The transglycosylase has been considered to be another excellent target, but no antibiotics have been developed to target this enzyme.

We have successfully obtained functional full-length transglycosylases to develop a high-throughput screening assay for finding new hit compounds that can replace the known inhibitor moenomycin binding with TG (2008). Subsequently, we determined the crystal structure of E. coli full-length membrane-bound bifunctional pencillin-binding protein PBP1b in complex with the antibiotic moenomycin (2009). We have also solved the crystal structure of S. aureus membrane-bound monoglycosyltransferase in complex with a substrate lipid II analog (2012), which for the first time revealed the lipid-II-bound acceptor site of TG. The mechanistic characterization of the lipid II dimerization into lipid IV in the peptidoglycan synthesis was therefore clarified at the atomic level.The three-dimensional interaction network between moenomycin/lipid II and TG domain provides a foundation for structure-based drug design. Over the years there are mainly three classes of inhibitors, including substrate lipid II analogs, moenomycin (lipid IV) analogs, and novel inhibitors identified through screening methods with unknown binding sites. We also focused our efforts on the protein complex structure with novel inhibitors which revealed unexpected binding sites on TG, providing valuable information to guide antibiotic optimization for improved inhibitory potency (2019).


The figure above is a summary of potential inhibitors targeting the bacterial transglycosylase (GT51 domain). The lipid-II-analog-bound S. aureus MgtA structure (PDB 3vmt) with an additional moenomycin A modeled in the donor site (PDB 6ftb and 3hzs) was shown in the middle. Shown in the surroundings are selected moenomycin and lipid II analogs on the top and novel non-saccharide inhibitors at the bottom, with compound name and references labeled for each.

bottom of page