Garman lab: Interconversion of lysosomal enzyme specificities

The research is about two related enzymes, GAL and NAGAL. They are found in the same location in the human body (in the lysosome, an acidic organelle responsible for breaking down molecules the cell no longer needs), their primary sequence is 50% identical, they catalyze the same reaction (hydrolysis of α-glycosidic bonds), and they have very similar active sites. However, they differ in substrate specificity (one cleaves the bond with galactose, and the other with N-acetyl glucosamine. In terms of structure, the backbone conformation is quite similar (compare and ). The active sites differ in only two residues, and the substrates bind in almost the same way (compare the active site of and ).

The researchers asked the following question: Is it possible to turn one into the other (in terms of reaction catalyzed)? Their hypothesis was that a simple swap of the two amino acids in the active site that are different would accomplish an interconverson of specificities.

To test this, they made variants called GAL(SA) and NAGAL(EL), in which one active site has the amino acids of the other active site and vice versa (by swapping the two residues that are different). The data obtained by enzyme kinetics supported their hypothesis; the preference for Galactose vs N-acetyl galactosamine is swapped (not shown here, but the data is in their paper). Crystal structures show how GAL(SA) is able to bind to either or . Comparing the structures of the NAGAL: N-acetyl galactosamine complex and the GAL(SA): N-acetyl galactosamine complex shows that .

You can explore the structural data further by going through the figures below and clicking on the buttons. If during your exploration, you get lost or some figures behave strangely, press first to reset the 3D browser and then try again.

The initial shows the sugar N-acetyl galactosamine.

Figure 1

Panel B (show )

Panel C (show )

Use these buttons to switch back and forth between the two enzymes or to animate the switching

If you click on pop-up on the bottom of the 3D browser window, maximize the pop-up window and turn on stereoview (right click on the model, select Style>Stereographic>...), the active site will really pop.

Figure X (bonus figure)

For the animation in Figure 1, the carbon alpha atoms of the shown active site residues were superimposed (RMSD = 0.3 Å). The following views of the active site differences shows a superposition of the six common carbon atoms (RMSD = 0.02 Å) in the bound sugar. It becomes obvious that the sugar is bound in a slightly different orientation with respect to the overall protein structure.

(use the buttons above to compare with α-NAGAL)

(use the buttons above to compare with α-NAGAL)

Figure Y (bonus figure)

The shape of the active site is often complementary to the molecule it binds to (lock-and-key concept). This figure shows the contacts between bound molecules and active site residues. Contacts are shown as the overlap of Van der Waals spheres around atoms. Slight overlap is shown in yellow while larger overlap is shown in red. When two functional groups form a hydrogen bond, atoms come closer than they would if they interact via Van der Waals interactions, so you expect to see red overlap for hydrogen bonds. Here are the contacts for and . These are observed structures, so the contacts seen explain why they bind to their ligands.

In contrast, here are the contacts for two hypothetical models, galactose bound to the α-NAGAL active site and N-acetyl galactosamine bound to the α-GAL active site. There are less contacts in the and severe clashes in the (clashes involve the acetyl group of the ligand and residues Glu 203 and Leu 206 of the active site). In fact, experiments show that α-NAGAL does bind galactose (though much more weakly than N-acetyl galactosamine) while α-GAL does not bind N-acetyl galactosamine.

Figure 2: Structure of GAL(SA)

GAL(SA) is derived from GAL by replacing actives site residues glutamate 203 with serine and leucine 206 with alanine. Having these smaller amino acids in the active site increases the substrate binding cavity, and makes the active site of αGAL(SA) very similar to that of αNAGAL. With these substitutions, the catalytic activity of GAL(SA) is more similar to NAGAL than to GAL (the data is not shown here, but can be found in the research paper).

: in complex with GalNAc

Crystal structure of α-GAL(SA) bound to GalNAc. α-GAL(SA) active site residues are shown in yellow and the product, GalNAc, is shown in gray. The blue mesh around GalNAc represents its electron density.

Use the buttons to hide the model of GalNAc in the figure. This is what a crystallographer would interpret to figure out what is bound in the active site.

: in complex with Gal

: Superposition with alpha-NAGAL bound to GalNAc

Superposition of structures: GAL(SA) bound to N-acetyl glucosamine, in yellow, and NAGAL bound to N-acetyl glucosamine, in dark blue. The sugar bound to GAL(SA) is shown in light brown and the one bound to NAGAL is shown in light blue.