I was forced to look at the wikipedia entry for enzyme, which states:
Enzymes /ˈɛnzaɪmz/ are macromolecular biological catalysts that accelerate chemical reactions. [...] Most enzymes are proteins, although a few are catalytic RNA molecules. The latter are called ribozymes. [Enzymes' specificity comes from their unique three-dimensional structures.]
The definition encompasses a rather broad range of molecules (macromolecular biological catalysts), is more of a functional rather than a structural definition, and requires clarification of what is meant by macromolecular. Also, not everything the wikipedia says should be accepted as consensus, of course.
This is an interesting question. For one thing, it challenged my biased view of enzymes as proteins (even if I know of ribozymes). I don't usually think of glycoconjugates as enzymes, but there you go, you have presented such a case. However, this is the structure shown in pubchem:
I see a couple of amide bonds, but mostly I see sugar rings, consistently with the description of the molecule:
Alpha-L-Fucp-(1->3)-[alpha-D-Manp-(1->6)-[beta-D-Xylp-(1->2)]-beta-D-Manp-(1->4)-beta-D-GlcpNAc-(1->4)]-D-GlcpNAc is a branched amino hexasaccharide consisting of a GlcNAc residue at the reducing end with a Man-alpha(1->6)-[Xyl-beta(1->2)]-Man-beta(1->4)-GlcNAc moiety attached via a beta-(1->4)-linkage and a Fuc residue attached via an alpha-(1->3)-linkage.
And yet the wikipedia says that "Stem bromelain (SBM) (EC 184.108.40.206), is a proteolytic enzyme". That's because I also missed the fine print: the pubchem entry is for MUXF, carbohydrate moiety of bromelain (hat tip to Ivan). Bromelain itself is a glycoprotein. And I am still looking for an experimental 3D structure to go with the molecule. Heading down another rabbit hole, the best I could do for now is find a link to a theoretical structure (Ref. 1):
A polysaccharide, attached to a folded protein, is shown in the lower right of the image.
So you opened up your post with a false premise, that all enzymes are exclusively polypeptides. Some of them, such as bromelain, are glycoconjugated proteins. And some such as ribozymes lack a peptide component altogether.
Finally, as to why the hexasaccharide is presented in 2D rather than as protein enzymes usually are, as a folded 3D structure, pubchem explains (see under 3D structure):
Conformer generation is disallowed since too many atoms, too flexible
- Banerjee, N., Banerjee, A.. A three dimensional structural model of stem bromelain: a critical comparative analysis of glycosylation torwards functional structural stability. DOI: 10.5452/ma-c3x5z