The Pathophysiological Functions of Heparanases: From Evolution, Structural and Tissue-Specific Perspectives

Abstract

Heparanase 1 (HPSE1) is a unique endoglycosidase responsible for the enzymatic cleavage of heparan sulfate, thereby playing important functions in cancer processes. In contrast, the structurally related Heparanase 2 (HPSE2) lacks catalytic activity and appears to counteract HPSE1 activities. However, contradictory observations in various pathologies highlight the need for a better understanding of the respective contributions of both heparanases. In this review, we provide a comprehensive resource about the biology of HPSE1 and HPSE2 based on findings from different mouse models, with an emphasis on immune cells and their involvement in skin pathophysiology. In addition, we explore the evolutionary relationships between the two heparanases and describe the structure-function of HPSE2 using the advanced protein-prediction tool AlphaFold 3 (AF3). These approaches unveil new insights for deciphering the functional molecular determinants that distinguish HPSE1 from HPSE2.

Keywords: Heparanase; cancer; immune cell; molecular modeling; mouse model; skin inflammation.

FIGURE 1

(A) (Top) Schematic view of the HPSE1 architecture and maturation steps (prepro to the mature form from top to bottom) with the small (8 kDa), the linker (6 kDa) and large (50 kDa) subunit colored in blue, red and green, respectively. Triangles and stars denote the position of the N‐glycosylation sites and catalytic residues (Glu225 and Glu343), respectively. (Bottom) Overall views of the proHPSE1 (accession code 5la4) and mature HPSE1 (accession code 5e8m) crystal structures with the subunits colored as in A and viewed in a similar orientation looking down the active site cleft; the distribution of the electrostatic potential mapped on the molecular surface at −3 kT/e (red) to +3 kT/e (blue) is shown as inset. (B) (Top) Schematic architecture of the four human HPSE2 isoforms (HPSE2a, HPSE2b, HPSE2‐2 and HPSE2c from top to bottom) with the corresponding AF3 predicted structures viewed in a similar orientation and looking down the vestigial binding cleft, mapped by the electrostatic potential as inset. The position of the predicted N‐glycosylation sites (triangles colored in green for those conserved with HPSE1) and the vestigial residue pair (Glu262 and Gly381 as a pink star and circle) in HPSE2c are indicated. Differences in the overall architecture of the four HPSE2 isoforms due to the spliced segments within the (β/α)₈ domain and in the topology of the electropositive patches between HPSE1 and HPSE2 isoforms are evident. The three putative heparin‐binding motifs in HPSE2c are colored in blue and indicated by an asterisk. (C) (Left) Overlay of HPSE2c and proHPSE1 (colored as in B) showing the two surface regions that differ in HPSE2c colored in green. (Right) Close‐up view of the HPSE1 binding cleft in the structure of HPSE1‐heparin tetrasaccharide dp4 complex (accession code 5e9c) overlay on HPSE2c showing the steric clashes of the dp4 substrate with the HPSE2c helical lid.

FIGURE 2

Phylogenetic analysis of HPSE1 and HPSE2 sequences showing the early divergence between the two heparanases. Phylogenetic tree of aligned protein sequences of Hpse1 and Hpse2c orthologs. Protein sequences were aligned using MAFFT [126], clustered with CD‐HIT and curated using BMGE [127]. The tree was generated with IQ‐TREE [128] and displayed with iTOL [129]. The two groups are colored in cyan and yellow/orange with the position of human Hpse1 and Hpse2c indicated by an asterisk; the main organism classes are labeled.

FIGURE 3

Multiple sequence alignment of (A) Hpse1 and (B) Hpse2 performed using CLUSTAL O (1.2.4) multiple sequence alignment (https://www.uniprot.org/align) with the default parameters to identify conserved regions. All the protein sequences were obtained from the UniProt database. (C) Percentage of sequence identity between HPSE1 or HPSE2 from  Homo sapiens (Human), Heterocephalus glaber (Naked Mole‐Rat, NMR), Oryctolagus cuniculus (Rabbit), Mus musculus (Mouse), Rattus norvegicus (Rat), Pantherophis guttatus (Corn snake) and Danio rerio (Zebrafish). Identical residues are marked with an asterisk (*), strongly similar residues with a colon (:) and weakly similar residues with a period (.). HPSE2 in (B) and (C) refers to the longest human isoform HPSE2c.

FIGURE 4

Overall views of the AF3 predicted models of the HPSE1‐HPSE2c and HPSE1‐MBP complexes. (A) (Left) The two HPSE1 and HPSE2c molecules from the top‐ranked predictions of the human HPSE1‐HPSE2c complex are shown in two orientations rotated by 90° with a transparent surface and colored as in Figure 1 (colored by prediction confidence pLDDT score as inset). (Right) Overall views, oriented as in Figure 1, of the buried interfaces colored in orange at the molecular surface of HPSE2c (top) and HPSE1 (bottom) in the predicted complexes. Steric interactions of the helical lid of HPSE2c within the HPSE1 binding cleft with the two catalytic residues colored in pink are evident. The location of the human HPSE2c Pro140Arg variant at the tip of the helical lid, associated with the UFS rare disease, is highlighted in red and labeled; those of the two HPSE2c peptides reported to inhibit HPSE1 are highlighted in pink. (B) Overall views of the top‐ranked AF3 predictions for (left) pro (linker region colored in red) and (right) mature HPSE1 bound to the MBP C‐type lectin domain shown with a transparent surface with the HPSE1 molecule oriented and colored as in A. The catalytic cleft is indicated by an arrow with the two catalytic residues colored in pink. A similar binding mode highlighting a shape complementarity at the interface of the two proteins is observed.