TL;DR
This paper investigates two glycomimetic compounds (1 and 2) as potential inhibitors of DC-SIGN, a key receptor in viral infections like HIV. While compound 2 initially appears much more potent, detailed biophysical studies reveal it unexpectedly clusters DC-SIGN tetramers, artificially inflating its measured activity. This highlights the critical need for thorough structural and biophysical characterization to avoid misleading results in drug design, ultimately favoring the simpler compound 1 as a better lead.
What Problem Does This Paper Address?
DC-SIGN is a crucial receptor on dendritic cells that pathogens like HIV exploit to infect host cells. Developing inhibitors for DC-SIGN is a promising strategy for preventing viral infections. Glycomimetics, which mimic natural sugars, are attractive candidates due to their stability and low toxicity. However, designing effective inhibitors is challenging because DC-SIGN naturally exists as tetramers, and pathogen ligands often present multiple binding sites (multivalency), leading to enhanced binding through avidity. This paper addresses the problem of accurately assessing the potency of potential DC-SIGN inhibitors, particularly when comparing compounds with seemingly different activities, and understanding how their presentation (monovalent vs. multivalent) affects their efficacy. Specifically, it investigates why one glycomimetic (compound 2) showed significantly higher potency in initial screens than a related, simpler compound (compound 1), and how this difference plays out when these compounds are presented in multivalent formats.
Methods
The researchers employed a multi-pronged approach to dissect the interaction between DC-SIGN and the glycomimetics:
- X-ray Crystallography: This technique was used to determine the precise atomic-level structure of DC-SIGN's carbohydrate-binding domain (CRD) in complex with both compounds (1 and 2). This allowed for direct visualization of how the molecules bind and to identify any differences in their binding modes.
- Isothermal Titration Microcalorimetry (ITC): ITC measures the heat released or absorbed during molecular binding events. This provided thermodynamic data (enthalpy, entropy, and free energy) on the interaction between DC-SIGN and the glycomimetics, offering insights into the forces driving the binding.
- Analytical Ultracentrifugation (AUC): AUC is used to determine the size and shape of molecules in solution by measuring their sedimentation velocity. This technique was crucial for detecting if the glycomimetics induced changes in the oligomeric state of DC-SIGN, such as aggregation or complex formation.
- Static and Dynamic Light Scattering (SLS and DLS): These techniques provide information about the size, molecular weight, and hydrodynamic radius of molecules in solution. They were used to further characterize the assembly of DC-SIGN induced by compound 2.
- HIV Trans-infection Inhibition Assays: These cellular assays were used to evaluate the functional efficacy of the glycomimetics, both in their monovalent and multivalent forms, in preventing HIV transmission.
Datasets
Not discussed in the paper. The study focuses on experimental data generated from purified proteins and synthesized compounds.
Key Insights and Ideas
- The "False Positive" Phenomenon: Compound 2, despite appearing to have a 10-fold higher inhibitory potency than compound 1 in initial screens, was found to be a "false positive" in a sense. Its high apparent potency was not due to superior intrinsic binding affinity but rather to an artifact of the experimental setup.
- DC-SIGN Clustering by Compound 2: The most striking finding is that compound 2, even in its monovalent form, can induce the clustering of DC-SIGN tetramers. This bridging of tetramers leads to the formation of larger complexes, which can artificially inflate its measured inhibitory potency in assays using soluble tetramers.
- Same Binding Mode, Different Behavior: Despite compound 2's clustering activity, X-ray crystallography revealed that both compounds bind to the DC-SIGN CRD in essentially the same way. This underscores that structural similarity doesn't always translate to similar functional behavior.
- Multivalency Levels the Playing Field: When both compounds were presented on multivalent scaffolds (dendrons and dendrimers), their inhibitory potencies became comparable. This suggests that the clustering advantage of compound 2 is lost when it's tethered, and the avidity effects of multivalency become dominant.
- Simplicity is Key for Lead Optimization: The study strongly advocates for prioritizing the simpler compound (1) for further drug development. Its synthetic accessibility and lack of problematic clustering behavior make it a more reliable and cost-effective starting point for optimization.
Key Results
- Potency Difference: Compound 2 showed an order of magnitude lower IC50 (higher potency) than compound 1 in SPR competition assays.
- Binding Mode: X-ray structures confirmed that both compounds bind to DC-SIGN CRD with the same mode, interacting with the Ca2+ ion and Val351.
- Clustering: AUC and DLS experiments demonstrated that compound 2, but not compound 1, induces the formation of larger DC-SIGN complexes (dimers of tetramers).
- Multivalent Efficacy: In HIV trans-infection assays, multivalent presentations of both compounds (dendrons and dendrimers) showed comparable efficacy, with the difference in monovalent potency largely disappearing.
- Optimal Lead: Compound 1 was identified as the optimal lead for further development due to its simpler structure and lack of clustering artifacts, despite its lower initial apparent potency.
Limitations and Future Work
- Dynamic Nature of Interactions: The authors acknowledge that the DC-SIGN/ligand system is dynamic, and the proposed supramolecular assembly model is a rough approximation. The low affinity of single mannose residues likely leads to various conformations and a mix of species in solution.
- Experimental Artifacts: The study highlights the potential for experimental artifacts (like clustering) to mislead drug discovery efforts. The "bridging" property of compound 2 is highly dependent on the assay format and the solution state of the protein.
- Future Work: The paper suggests that future optimization efforts should focus on compound 1, given its synthetic simplicity and more predictable behavior. Their recent improvements in this direction have been reported elsewhere, indicating ongoing research. The study also implicitly calls for more sophisticated assays that account for multivalent interactions and potential clustering effects early in the screening process.
Summary updated