Surface Plasmon Resonance (SPR) is an invaluable technique that generates information- rich data for a variety of biomolecular interactions. Researchers utilize SPR to understand biological pathways and to develop and characterize a range of potential therapeutics to treat disease and illnesses. These interactions include those occurring with and between the major classes of biological macromolecules. However, this article is focused solely on using SPR to study biomolecular interactions involving small molecules binding to proteins. To provide key information about these interactions, many researchers are relying on SPR, which is a widely used technique for the determination of kinetics and thermodynamics. “Small molecules” for purposes of this article generally refers to molecules with molecular weights of about 1,000 Da or less. There are certain standard approaches that researchers follow to carry out these types of experiments. Since a higher surface density is needed when there is a large molecular weight difference between the protein (target) and “small molecule” (analyte) whose interaction is being studied, direct coupling is usually the preferred means of attaching a target to the sensor chip. Of necessity, since many “small molecules” do not contain primary amines or thiol groups needed for direct coupling, the protein is usually coupled to a hydrogel surface (typically dextran) and then the small molecule analyte is injected over the surface at 5 or more concentrations and kinetics are determined. If testing is done at multiple temperatures, thermodynamics can also be determined. Since it is common for small molecules to show limited solubility in the aqueous running buffers used for SPR studies (eg. PBS, HEPES), analytes soluble in DMSO are typically diluted with running buffer to about 1-5% DMSO prior to SPR analysis. In some cases, detergents may also be used to help with solubility issues. Many of these experiments are carried out to aid in the development of small molecules as potential therapeutics for disease. We now highlight some recent articles that researchers have published using Reichert’s SPR systems for investigating low molecular weight molecules interactions with proteins.
Example 1: Target HIV – using a Reichert4SPR for screening
A 2019 article in ACS Infectious Diseases highlights how some of the unwanted functions of HIV Nef accessory protein, which plays a key role in HIV replication, can be blocked by small molecules1.
For the results presented in this article, researchers developed analogs to an earlier small molecule that showed promise in testing. They had more than 200 analogs to analyze and ranked them basedonbindingresultsobtainedusingaReichert4SPRsystem. RecombinantHIV-1Nefwas covalently bound to a dextran sensor chip as the target and binding interactions with the analogs were ranked in terms of affinity, residence time and binding response1. Of the 216 analogs screened, 45 exhibited no binding. The remainder of the responses were ranked based on the magnitude of the responses, their kinetics, including on- and off- rates, and activity from a separate assay1.
Figure 1: Overlaid sensorgrams show binding of analyte FC-8698 (black) containing the benzimidazole moiety (green section on structure on left) and binding of analyte FC-10580 (red) with the same backbone but without the key group. Using SPR, the KD for FC-8698 was determined to be 13 nM vs the KD of 9.8 μM for FC-10580. In both cases, binding is to recombinant HIV-1 Nef and curves were fit using a 1:1 binding model.
Example 2: Target Tuberculosis – using a 2-channel Reichert SPR for initial screeningA 2017 article in ChemMedChem describes how a SR7500DC (2-channel Reichert SPR instrument) has been used for the screening of potential antimycobacterial agents against Mycobacterium tuberculosis (Mtb)2. In this article, researchers explain that there is an urgency to their work because multi-drug resistance and the need for long term treatment are making it difficult to treat this disease and, they note that there are millions of new cases of Mtb infections each year worldwide (as of the year 2017 when they wrote their article)2. Previous work showed that the enzyme CYP121 is essential for Mtb growth making it a potential target for Mtb treatment.2 Screening in this study was carried out based on biophysical and microbiological methods. The initial screen was carried out using Surface Plasmon Resonance (SPR) with a Reichert 2-channel instrument and involved capturing biotinylated CYP121 on a streptavidin chip as the target and then analyzing binding to the target of 139 novel small molecule antimycobacterial compounds across six classes. Researchers built up small molecule compounds on six different scaffolds to probe a wide range of structures. Of the compounds screened, 32% showed a response using SPR. Seventeen of the compounds had higher responses than econazole, the positive control. When compared to the positive control econazole, forty-four of the compounds screened were found to be worthwhile for further testing (ie. to have responses R/Rpos of 0.5 or higher). Further testing with multiple techniques led the researchers to a promising target for further research, a compound they call I:47 which exhibited good antimycobacterial activity and bacterial selectivity with a favorable toxicity profile2.
Figure 2: SPR was used to do an initial screen of 139 novel small molecule inhibitors and differentiate between binders and non-binders. In addition, the size of the response was compared to the response for the positive control, econazole (R(analyte)/R(econazole). The reddish line at 0.5 differentiates between analytes that were carried forward (ratio > 0.5) and the remainder that were dropped from further testing.
Example 3: Anti-Cancer Drug Candidates – comparison using a Reichert4SPRA 2017 Article in PLOS 1 includes SPR data obtained using a Reichert4SPR for the binding of small molecule anti-cancer drugs to human serum albumin (HSA)3. In this study, researchers looked at two potential anti-cancer drug candidates, NSC48693 (which is hydrophilic) and NSC290956 (which is hydrophobic) and determined where they bind to HSA using fluorescence quenching and molecular modeling. For information regarding the kinetics and thermodynamics of the small molecule/HSA interactions, researchers used a Reichert4SPR. HSA was amine coupled to a dextran sensor chip and then the two drug candidates were diluted and injected over the surface as analytes. Kinetic analysis resulted in the determination of the affinity for the HSA/NSC48693 interaction (KD = 13.8 μM) to be almost 10 times higher than the affinity for the HSA/NSC290956 interaction (KD = 116 μM).3 This information is expected to be particularly useful in later determination of the type of dosing that would be needed for clinical studies using each compound.
Figure 3: A and C (left) show the overlaid sensorgrams (with fit to 1:1 model in green) and Langmuir isotherm plot, respectively, for the HSA/NSC48693 interaction (KD = 13.8 μM). B and D (right) show the overlaid sensorgrams (with fit to 1:1 model in green) and Langmuir isotherm plot, respectively, for the HSA/ NSC290956 interaction (KD = 116 μM).
In a 2019 paper in ACS Chemical Biology researchers present results from their research into two known short linear motifs (SLiMs), PxIxIT and LxVP, which bind to Calcineurin (CN), an enzyme involved in a number of biological functions including such things as development, heart function, and immune response4. Although most cellular signaling involves protein/protein interactions, some also occurs between proteins and short linear motifs which are about 3-10 amino acids in size and are typically located in intrinsically disordered regions. Using SPR to investigate these interactions is not trivial because the motifs are small and have low binding affinities. For this project, researchers captured His-tagged CN on a NiNTA sensor chip and used a Reichert4SPR surface plasmon resonance (SPR) instrument to determine binding constants with peptides containing either PxIxIT or LxVP. Researchers employed Structure Based Shape Complementarity (SBSC) analysis which uses the shape of the preformed, rigid SLiM binding pocket to identify complementary amino acids for binding to expand the number of known SLiM sequences. Using this approach, they identified 119 CN regulators and substrates containing either PxIxIT or LxVP. Even though this type of interaction is usually weak, they were able to obtain kinetics. From SPR results, researchers concluded that LxVP SLiM binding to CN is on-rate-dependent4. Results for a typical compound they studied are shown here:
Example 4: The Binding of short linear motifs (SLiMs) to the Enzyme Calcineurin
Figure 4: Sensorgram overlays for an LxVP peptide, NFATc1, binding to captured CN. When fit to a 1:1 binding model, the on-rate was determined to be 1.97x104 M-1s-1, the off-rate 0.113 s1 and the KD 5.9 μM.
Example 5: Retinoic Acid – an important nutrientIn a 2019 article in Biochemistry, research on the binding of all trans Retinoic Acid (atRA), a metabolite of Vitamin A, to Cellular retinoic acid-binding protein 2 (CRABP2), an intracellular lipid- binding protein, is described5. CRABP2 delivers atRA to Retinoic Acid Receptors (RARs), which than leads to the activation of specific gene transcription. Using a Reichert4SPR, they captured His- tagged CRABP2 on a Ni-NTA sensor chip and injected the atRA over as analyte.5 All-trans retinoic acid (atRA) is a hydrophobic molecule and CRABP2 controls its movement between the nucleus and cytoplasm. They determined that the affinity for the CRAB2/atRNA interaction was in the nM range with a slow off rate. This information, combined with the hydrophobic nature of CRABP2, indicated to them that the atRA release from the CRABP2 binding pocket was probably mediated by external factors. They noted that the binding rate was determined by atRA’s rate of diffusion out of detergent, which had to be added to the solution due to atRA’s poor aqueous solubility5. A typical result is shown here:
Figure 5: Using SPR, a kinetic titration was run to determine kinetics for the binding of atRA to CRABP2. The on-rate was determined to be 6.92 X 105 ± 2.32 X 105 M-1s-1, the off-rate 4.01 X 10-3 ± 1.08 X 10-3 s-1 and the KD 5.94 ± 0.79 nM.
Example 6: Development of CK2 inhibitors for Cancer TreatmentIn a 2019 article in the Journal of Medicinal Chemistry, researchers used the protein kinase CK2 as target and looked for potential inhibitors6. The protein kinase CK2 is known to play an important role in pro-oncogenic pathways that are critical to cells including proliferation, differentiation, and survival. Previous research involving development of inhibitors for the protein kinase CK2 were targeted toward the ATP binding pocket, but it was found that the compounds developed could also act on other targets in the body with ATP binding pockets which could then cause unwanted side effects. Researchers in this study previously determined there was a more specific, potentially druggable site on the protein CK2. In their current work they sought to prove it to be a viable option by developing 2-Aminothiazole Derivatives that they could test. Researchers amine coupled Anti- GST to the surface of a dextran sensor chip and then captured GST-tagged CK2α over the surface as a target for the allosteric inhibitors they made6. Of the allosteric inhibitors they constructed and tested, they found one of the compounds, 2-hydroxy-4-((4-(naphthalen-2-yl)thiazol-2- yl)amino)benzoic acid to be the lead compound with a sub-micromolar IC50. Testing was carried out using a number of techniques including SPR6. Comparing IC50 (biochemical inhibition) results with SPR binding affinity constants obtained with a 2-channel Reichert SPR for a number of the inhibitors they tested, they found good agreement7.
Figure 6: Good correlation is seen between IC50 results obtained for a number of allosteric inhibitors binding to CK2 and KD results obtained for the same biomolecular interactions using SPR. Values are in μM.7
Example 7: Potential Anti-Influenza Agent
In a 2016 paper in Scientific Reports, researchers describe how using a "wedge" could destabilize matrix protein 1 (M1) of influenza A virus (IAV) to and keep it from self-associating8. They started out testing about 70,000 potential compounds using virtual screening. Then, they took the top ten hits and did in vitro testing and found that six of the compounds showed antiviral activity. However, five of the six also exhibited substantial cytotoxicity. One small molecule which they call PHE only showed minimal cytotoxicity and became their top candidate to carry into further testing8. PHE is a hydrophobic compound with a molecular weight 400 Da and was found to be straightforward to synthesize in just a few steps so it was carried forward into further "proof of concept" testing. Researchers used a 2-channel Reichert SPR to determine kinetics for M1 to itself and M1 to PHE. Biotinylated M1 was captured over a neutravidin surface. M1 was then injected over at various concentrations and affinity was determined (Figure 7). Next, they tested M1 binding to PHE and determined the affinity (Figure 7). The slow off rate for the M1/M1 interaction led researchers to conclude that self-association is an off-rate driven process8. The affinity decreases from about 50 pM for M1 to M1 binding to 870 nM for PHE binding to M1 which researchers stated is still fairly tight binding for a small molecule. Using molecular modeling, they found that PHE binds to multiple sites on M1 which they feel may give it a better chance of destabilizing M1. The method of inhibition is that the small molecule impedes M1 oligomerization.8
Figure 7: Top – left M1/M1 association curves – right – M1/M1 dissociation curves. Bottom–Left – Association and Bottom-Right – Dissociation curves for PHE binding to M1.
Example 8: Characterizing Inhibitors to Auxin Transport
Auxin is a plant growth regulator that has been widely studied. In a 2016 article in Plant Cell, researchers used a 2-channel Reichert SPR instrument to study binding interactions of several small molecule inhibitors with the protein Twisted Dwarf 1 (TWD1), which mediates the action of auxin transport inhibitors9. If a plant is grown on land, most of its development is dependent on the formation of auxin gradients called polar auxin transport or PAT. Synthetic auxin transport inhibitors, such as NPA (1-N-naphthylphthalamic acid), can inhibit the basal polar auxin flow required for plant development at low concentrations. Using a 2-channel Reichert SPR, researchers immobilized TWD1 to the surface of a dextran chip using both amine and thiol coupling and flowed over small molecules at various concentrations. UsingSPR, researchers determined that binding of small molecule inhibitors to the Twisted Dwarf protein was low affinity in vitro. This result was in agreement with NMR results9. SPR binding results for one of the small molecule inhibitors is shown here.
Figure 8: SPR binding curves for Triiodobenzoic acid (TIBA) fit to a 1:1 binding model.
In a 2019 article in Emerging Microbes and Infections, researchers used SPR to study the binding of Honokiol, a natural plant polyphenol which they determined to be a potential inhibitor of a common pathogen Staphylococcus aureus (S. aureus).10 Researchers used a Reichert4SPR to study this interaction. They immobilized α-Hemolysin (Hla) on a dextran chip and flowed over the Honokiol at various concentrations to determine the affinity. Researchers in this study confirmed that α-Hemolysin (Hla), an extracellular toxin secreted by most pathogenic S. aureus strains, bound to Honokiol with a modest binding affinity. Additional research using other techniques showed that Honokiol, a low molecular weight natural product, can inhibit inflammasome activation without affecting oligomerization.10 The SPR binding curves are shown here.
Example 9: Honokiol Inhibition of the pathogen Staphylococcus aureus (S. aureus)
Figure 9: SPR binding curves for Honokiol binding to Hemolysin. Data is fit to a 1:1 binding model. The association rate constant ka is 293 M−1s−1, the dissociation rate constant kd is 5.03 × 10−2 s−1 and the equilibrium dissociation constant KD is 172 μM.
Example 10: Small molecule zinc complexes binding to DNAIn a 2015 article in Dalton Transactions, researchers designed small molecule zinc complexes and determined how they fit into bulges and quadraplexes on DNA. Non-canonical nucleobases that are part of DNA bulges are common targets when designing small molecules for binding both organic and inorganic compounds. And Zn(II) complexes containing planar aromatic pendents with two fused rings bind to DNA Hairpin (T-bulge) more tightly than complexes with non-planar pendents.11 Both a better understanding of these types of interactions and the development of alternative ways to recognize unusual nucleic acid structures could lead to improved design of small molecules for future therapeutic applications. Researchers in this study used Reichert's 2-channel surface plasmon resonance (SPR) system to capture biotinylated DNA oligos on a planar streptavidin chip. They then flowed over the analytes at various concentrations to characterize the binding of Zn complexes to DNA T-bulge and T-loop structures.11 An example of a typical result is shown here:
Figure 10: Sensorgrams for Zn4Qone binding to T-bulge are shown. The global best fit is to a 1:1 binding model with a KD value of 36.4 μM.