Screening Competition trNOESY Assay

The trNOESY competition assay requires a "marker" ligand, for which the binding site on the protein and preferably the binding constant are known. The crystal structures of BoNT/B ligand complexes show that doxorubicin (55) and 3'-sialyllactose (56) bind to Site-1 (Fig. 5). Site-1 is a common surface feature found in the structures of both TetC and BoNT, and preliminary results obtained from the crystal structure of TetC in complex with doxorubi-cin indicate that doxorubicin also binds to Site-1 on TetC (S. Swaminathan, personal communication, March 2002). Furthermore, the dissociation constant for the doxorubicin/TetC complex is known and is approx 10 ^M (57). Doxo-rubicin is thus an ideal marker ligand to use in this study.

Competition experiments are used to identify those compounds that bind to Site-1 or a different site by determining whether binding of doxorubicin to TetC is disrupted by a competing ligand binding to the same site. This is evidenced by an absence of trNOEs for doxorubicin and the presence of trNOEs for the other ligand. However, it must be stressed that these assays cannot identify the exact site of ligand binding, only whether the sites are different or the same (see Note 9 for additional caveats).

3.4.1. Method 1: Sequential Addition of Ligands

Sequential addition of each of the computationally predicted Site-2 ligands is used as a negative control to show whether any of these ligands will also

Table 1

Ligands Tested Positive by ESI-MS

for Noncovalent Complex Formation With TetC

Table 1

Ligands Tested Positive by ESI-MS

for Noncovalent Complex Formation With TetC

Predicted Site-1

Predicted Site-2

Doxorubicin3,6

Tyr-Glu-Trp

3'-Sialyllactose2

Lavendustin A

D-(+)-Cellotetraosea

Sar-RGDSP

Neohesperidin diHCl"

NF-GalPyr

Gly-Arg-Gly-Asp-Ser"

MP-biocytin

Hemorphin-5a

SGNYPIV

Etoposide phosphate

"Reported in ref. 57. fcBinds Site-1 in BoNT/B (55).

"Reported in ref. 57. fcBinds Site-1 in BoNT/B (55).

Table 2

Concentrations of Ligands and TetC and Observation of Binding Activity

Table 2

Concentrations of Ligands and TetC and Observation of Binding Activity

Ligand

[Lig] (||M)

[TetC] (|M)

[TetC]:[Lig]

Binding

Fig. 6

A

Doxorubicin

1081.4

54.3

1:20

Yes

B

Sar-RGDSP

1116.3

50.6

1:22

No

C

SQNYPIV

1050.6

47.4

1:22

Yes

D

Lavendustin A

993.3

45.0

1:22

Yes

E

NF-GalPyr

926.6

43.3

1:21

Maybe"

Fig. 8

A

Sialic acid

386.7

22.3

1:17

No

BC

Doxorubicin

365.6

21.5

1:17

Yes, nofc

C

3'-Sialyllactose

351.9

20.7

1:17

Yes

"Precipitation was observed, resulting in a decrease in the intensities of all crosspeaks. ^Doxorubicin binding was observed in Fig. 8B but not in Fig. 8C.

"Precipitation was observed, resulting in a decrease in the intensities of all crosspeaks. ^Doxorubicin binding was observed in Fig. 8B but not in Fig. 8C.

bind to Site-1 by displacement of doxorubicin. The results also show whether two or more ligands can bind simultaneously to TetC, thus identifying suitable pairs of ligands to link together in developing the synthetic high-affinity ligands (SHALs) as detection reagents for the Clostridium neurotoxins.

1. Prepare a TetC:doxorubicin complex with the concentrations approx [50 \M]/ [1000 \M] or a ratio of 1:20. Collect trNOESY data.

2. Add approx 1 mM each of the predicted Site-2 ligands sequentially (Table 2), collecting trNOESY data after each addition.

3. Analyze trNOESY data after each addition of ligand to determine whether crosspeaks belonging to doxorubicin or added ligand become weaker, positive, or disappear, indicating that doxorubicin has been displaced. The appearance of strong negative crosspeaks indicates that the ligand is able to bind simultaneously with doxorubicin and is thus a good candidate for linking to another ligand (Fig. 6).

3.4.2. Method 2: Evaluation of Ligands Binding Sequentially in Reverse Order to TetC

It is important to carry out the positive control, or the reverse experiment of method 1. Each ligand shown in Fig. 4 is first added to TetC in a 20-fold excess to determine whether binding could be observed without the interference possible owing to the presence of another ligand. Doxorubicin is next added to the mixture to determine whether it can displace the bound ligand being tested, and to confirm that the protein is active if no binding had occurred with ligand-1. If displacement occurs, then ligand-1 is identified as a Site-1 binder that exhibits a weaker binding affinity than the marker ligand, doxorubicin. Next, another ligand, ligand-3, is added to determine whether the ligand-1 or doxorubicin is displaced. If ligand-1 is displaced but not doxorubicin, then ligand-1 and -3 must compete for the same site, with ligand-3 having the higher binding affinity. A comprehensive analysis of the optimized short list of compounds using this strategy yielded the results shown in Fig. 7. Here is a step-by-step example of the method:

1. Add a 20-fold excess of sialic acid to TetC (Table 2) and acquire a 300-ms trNOESY spectrum (Fig. 8A). Sialic acid does not appear to bind.

2. Add a 20-fold excess of doxorubicin to the sialic acid/TetC solution. Acquire a 300-ms trNOESY spectrum (Fig. 8B) to confirm that the protein has not lost its activity or that the batch of protein used was bad.

3. Add a 20-fold excess of 3'-sialyllactose (another Site-1 binder) to the sialic acid/ TetC/doxorubicin solution. Acquire a 300-ms trNOESY spectrum. In this case, 3'-sialyllactose will displace doxorubicin, confirming that both ligands bind to Site-1, with 3'-sialyllactose having a higher binding affinity constant than doxorubicin.

3.4.3. Method 3: Addition of TetC to Different Combinations of Ligands

This method screens multiple ligands simultaneously for binding to TetC in a more high-throughput manner. Up to 10 ligands at a time can be rapidly screened, but since the set of ligands had been prescreened by computational and mass spectroscopy methods, we limited our combinations to three at a time, as shown in Fig. 8. The time lost by running a larger number of trNOESY experiments is easily regained by having to analyze much less complicated, and unambiguous, data sets. The experimental results also provided key

Fig. 6. Effect of doxorubicin binding in Site-1 on predicted Site-2 ligands binding to TetC: expanded regions of the 2D trNOESY spectra of TetC/ligand complexes at 200-ms mixing time and 20°C showing binding of (A) doxorubicin (Site-1 marker ligand). The arrows indicate one aliphatic and one aromatic crosspeak that belongs to doxorubicin, which remains bound throughout the sequential additions of (B) Sar-RGDSP, (C) SQNYPIV, (D) lavendustin A, and (E) NF-GalPyr. The appearance of new, negative crosspeaks in the panels is indicated by a box, and these crosspeaks indicate that Sar-RGRSP does not bind when doxorubicin is present, but that both SQNYPIV and lavendustin A do bind. The addition of NF-GalPyr (E) appears to partially displace lavendustin A, but not doxorubicin and SQNYPIV.

Fig. 6. Effect of doxorubicin binding in Site-1 on predicted Site-2 ligands binding to TetC: expanded regions of the 2D trNOESY spectra of TetC/ligand complexes at 200-ms mixing time and 20°C showing binding of (A) doxorubicin (Site-1 marker ligand). The arrows indicate one aliphatic and one aromatic crosspeak that belongs to doxorubicin, which remains bound throughout the sequential additions of (B) Sar-RGDSP, (C) SQNYPIV, (D) lavendustin A, and (E) NF-GalPyr. The appearance of new, negative crosspeaks in the panels is indicated by a box, and these crosspeaks indicate that Sar-RGRSP does not bind when doxorubicin is present, but that both SQNYPIV and lavendustin A do bind. The addition of NF-GalPyr (E) appears to partially displace lavendustin A, but not doxorubicin and SQNYPIV.

Site-3

Ser-GIn-Asn-Tyr-Pro-lle-Val

Site-2

3'-s!alyllactosc

Doxorubicin

Site-1

Tyr-Glu-Trp

Lavendustin A

HVIP-biocytin

Tyr-Glu-Trp

Eliminated

Eliminated

Fig. 7. Grouping of optimized set of ligands tested by trNOESY competition assay into three different binding sites on TetC. The lines indicate how a ligand from each site can be linked in unique ways to develop SHALs as detection reagents in nanosensors. Three ligands were eliminated because either binding was not detected or the ligand was insoluble in aqueous solutions.

No binding detected

Low Solubility in aqueous solutions

Naphthofluorescein di-(ß-D-galactopyranoside)

Fig. 7. Grouping of optimized set of ligands tested by trNOESY competition assay into three different binding sites on TetC. The lines indicate how a ligand from each site can be linked in unique ways to develop SHALs as detection reagents in nanosensors. Three ligands were eliminated because either binding was not detected or the ligand was insoluble in aqueous solutions.

information about the compatibility of the ligands under the same solvent conditions.

1. Negative NOE crosspeaks for small molecules in the absence of TetC were sometimes observed when the molecule contains protons attached to large ring systems, such as the protons attached to the four-ring system of doxorubicin. These protons exhibit less internal motion than those located in more flexible long carbon chains. However, although the sign of the trNOE crosspeaks corresponding to these aromatic resonances remained the same regardless of whether TetC was present or not, their intensities were always much stronger for the bound ligand. Thus, both the sign and intensity of the crosspeaks must be taken into consideration when distinguishing ligands that bind from those that do not.

2. Our experiments were carried out on a Varian Inova 600-MHz spectrometer. Although the experiments can be carried out on a lower-field instrument, higher concentrations of both protein and ligands will be required because the sensitivity of the experiment decreases with decreasing magnetic field strength.

3. This step is only necessary if the protein concentration is high and/or the molecular weight of the protein is low. Either case would result in the detection of protein signals that may interfere with observation of the ligand signals. In our case, the TetC concentration is kept low (approx 50 ||M). Although TetC signals could be detected at higher concentrations (approx 72 ||M), they did not interfere with the detection of the ligand signals. In addition, the molecular weight of TetC is

4. Notes

Fig. 8. A 300-ms trNOESY spectra at 2°C of (A) sialic acid does not appear to bind to TetC, as evidenced by weak and positive trNOEs (boxed gray crosspeaks). (B) The addition of doxorubicin to a mixture of TetC and sialic acid results in doxorubicin binding, as evidenced by the strong negative trNOEs (circled black crosspeaks). (C) The addition of 3'-sialyllactose to a mixture of sialic acid, doxorubicin, and TetC shows that 3'-sialyllactose displaces doxorubicin from binding in Site-1. The circled positions of doxorubicin crosspeaks from (B) and the boxed gray sialic acid crosspeaks are shown in (C) for comparison.

Fig. 8. A 300-ms trNOESY spectra at 2°C of (A) sialic acid does not appear to bind to TetC, as evidenced by weak and positive trNOEs (boxed gray crosspeaks). (B) The addition of doxorubicin to a mixture of TetC and sialic acid results in doxorubicin binding, as evidenced by the strong negative trNOEs (circled black crosspeaks). (C) The addition of 3'-sialyllactose to a mixture of sialic acid, doxorubicin, and TetC shows that 3'-sialyllactose displaces doxorubicin from binding in Site-1. The circled positions of doxorubicin crosspeaks from (B) and the boxed gray sialic acid crosspeaks are shown in (C) for comparison.

sufficiently large (51 kDa) and it is known to dimerize, so that the protein signals are broadened out into the baseline.

4. If the ligands and proteins can be dissolved in deuterated solvents and buffers, then a simple presaturation pulse of the water signal can suffice to eliminate any residual water signal. However, if the samples need to be prepared in solvents or buffers that contain a high ratio of H2O/D2O, then it is necessary to implement one of these water suppression schemes.

5. Sialic acid did not bind TetC at 2, 10, 20, 30, or 37°C, as evidenced by the presence of weak positive crosspeaks in the spectra at all of these temperatures. However, a crystal structure of a TetC/sialic acid complex by Emsley et al. (58) shows that sialic acid binds to a site that is adjacent to Site-1. One possible explanation for the discrepancies between the NMR and X-ray results is that sialic acid binds TetC with <103 M or >106 M affinity in the temperature range that we tested. Another reason, however, may be owing to differences in binding between the crystal state and the solution state.

6. Whereas TetC can be purchased as a dry, lyophilized powder, other proteins may not be available dried or they cannot be dried without compromising their activity or risking their precipitation. In these cases, either the protein is dialyzed against deuterated buffer or the experiment is carried out in 90% H2O/10% D2O.

7. The minimum volume needed for the NMR experiment in a standard 5-mm tube is approx 500 |L. However, other NMR tubes and probes could be used instead to reduce the amount of sample needed in cases in which the protein is rare or has limited solubility. For example, 5-mm Shigemi tubes (200-300 |L) or a microprobe (2 |L) can be used to reduce the volumes. However, note that these are more expensive options and not always available in every NMR laboratory.

8. Ligand:protein ratios between 5 and 50 can be used, depending on the amount of protein available. The optimal molar ratio of TetC:doxorubicin was found to be between 1:15 and 1:25. These ratios provided good sensitivity for detection of trNOEs for a large range of structurally unrelated ligands

9. Several limitations of trNOESY competition assays need to be considered: the possibility that two or more ligands can bind the same site simultaneously cannot be completely ruled out based only on their lack of competition. Recently, Ma et al. (59) have proposed that a protein can preexist in ensembles of substrates, as a result of its conformational flexibility, and present a range of different binding site shapes to the incoming ligands, such that one site may recognize and bind multiple, diverse ligands. An assumption has been made that the ligands are binding specifically to only one site, which may not always be the case. For example, ^-acetylgalactosamine binds to two different sites in the crystal structure of the TetC complex (58). Similarly, if a predicted Site-2 ligand is binding with specificity to more than one site, a stronger competitor can displace it from one site, but not necessarily from the other sites. In such a case, both ligands would bind simultaneously to TetC. Nonspecific association of a ligand with TetC is also possible, especially at very high concentrations of ligand.

Was this article helpful?

0 0

Post a comment