In animals, CS/DS glycosidic bond-cleaving enzymes are hydrolases that cleave the β-1,4-glycosidic bond between GalNAc and GlcA residues in CS chains via a hydrolysis mechanism to produce saturated oligosaccharide products (Kreil, 1995). In mammals, the so-called hyaluronidases (EC 3.2.1.35) have been reported to be the only hydrolases that cleave CS chains, and even some hyaluronidases do not degrade hyaluronan (HA) but only degrade CS (Kaneiwa et al., 2010). In fact, most animal-derived hyaluronidases and microorganism-derived CS/DS lyases, which we will introduce later, show both HA- and Chn/CS-degrading activities, which may be due to the very similar structural features of the two GAG polysaccharides. Both HA and CS chains have same types of β-glycosidic bonds in and between repeating disaccharide units consisting of GlcA and hexosamine residues, and the only difference in structure between them is that the acetylated hexosamine GlcNAc in HA is replaced by the GalNAc in the Chn skeleton of CS. In addition, like other glycoside hydrolases, hyaluronidases exhibit certain transglycosidase activities (Hoffman et al., 1956), which can be used for synthesizing of HA (Kobayashi et al., 2003), Chn (Kobayashi et al., 2003), CS (Fujikawa et al., 2005), their derivatives (Ochiai et al., 2007b; Kobayashi et al., 2003), and hybrids of HA-Chn and HA-CS (Ochiai et al., 2007a).

In human genome, six highly homologous genes have been found to encode hyaluronidase-like sequences including five functional hyaluronidases genes (HYAL1, HYAL2, HYAL3, HYAL4 and SPAM1) and a pseudogene HYALP1(also called HYAL6P) that is transcribed in humans but is not translated (Csoka et al., 2001; Stern and Jedrzejas, 2006). HYAL1, HYAL2, and HYAL3 are clustered in the chromosome 3p21.3 locus, whereas the HYAL4, SPAM1(encodes PH-20) and HYALP1 genes are found on chromosome 7q31.3 (Csoka et al., 2001; Stern, 2003). Based on the optimal pH, most of these hyaluronidases show their highest activity at an acidic pH (Lokeshwar et al., 2001; Sabeur et al., 1997). As mentioned above, these so-called hyaluronidases from humans also show a certain degree of CS-degrading activity. The sperm-specific enzyme PH-20 shows much higher activity against Chn than CSA and HA at pH 4.5, whereas prefers HA and CS to Chn at pH 4.0 (Honda et al., 2012). In contrast, plasma hyaluronidase HYAL1 prefers to degrade CS-A than HA at pH 4.5 but prefers HA than CS-A at pH 3.5 (Csoka et al., 2001; Honda et al., 2012; Yamada, 2015). Furthermore, C units in CS-C negatively affect the hyaluronidase activity of PH-20. More interestingly, HYAL4 has been shown to be a specific CS-degrading enzyme without any activity toward HA, which means that the name of hyaluronidase is a misnomer for this enzyme (Csoka et al., 2001; Stern, 2003; Jedrzejas and Stern, 2005; Kaneiwa et al., 2010). The HYAL1 and PH-20 with CS-degrading activity cannot cleave the galactosaminidic linkages in -GalNAc-IdoA- and -GalNAc-GlcA(2S)- sequences, which are often found in DS chains and in highly sulfated D unit-containing domains of CS chains, respectively. In contrast, HYAL4 could strongly cleave the galactosaminidic linkages in -GlcA(2S)-GalNAc(6S)-GlcA-GalNAc(4S or 6S)- (Kaneiwa et al., 2010), suggesting that HYAL4 plays an important role in the degradation of CS in mammal. However, the degradation mechanism of DS chains in animals remains to be further investigated. Enzymes HYAL1 and HYAL2 are the main HA-degrading enzymes in somatic tissues. HYAL1 is a lysosomal enzyme and by contrast HYAL2 binds to plasma membrane via a glycosylphosphatidylinositol anchor (Afify et al., 1993; Rai et al., 2001). The traditional HA degradation model is that high molecular weight HA is digested into low molecular weight HA oligosaccharides by extracellular HYAL2, then, the HA oligosaccharides are internalized by interaction with cell surface receptors, and the internalized HA oligosaccharides are further decomposed by lysosomal HYAL1, exoglycosidases β-glucuronidase and β-hexosaminidase (Hex) (Stern, 2003; Stern, 2004; Figure 2 and Table 2). As a dimeric enzyme, human Hex exists in two main isoforms HexA (αβ) and HexB (ββ), of which the α- and β-subunits are encoded by HexA and HexB genes, respectively (Chiricozzi et al., 2014). Interestingly, HexA can hydrolyze HA and CS chains from their non-reducing ends but HexB cannot (Thompson et al., 1973; Bearpark and Stirling, 1978). Like HA, CS can be internalized by interaction with cell surface receptors and degraded in the lysosome (Wood et al., 1973). Indeed, mice lost both HYAL1 and Hex activity show the accumulation of HA and CS (Gushulak et al., 2012).

Unlike the hyaluronidases from animals, CS/DS-lyases from microorganisms cleave the β-1,4-glycosidic linkage between hexosamine and hexuronic acid residues in HA or CS/DS chains through an elimination reaction to yield an unsaturated double bond between C-4 and C-5 on the uronic acid residue Δ^4,5hexuronate (ΔHexA) at the nonreducing end of the resulting oligosaccharide products. Conventionally, CD/DS lyases are named as chondroitinases (CSases), although some of them cannot digest CS, such as CSase B, which shows specific activity towards DS only. Moreover, CS/DS lyases can be classified into endolytic and exolytic types according to their substrate-degrading patterns, in which endolyases cleave CS/DS chains initially into larger oligosaccharides and finally to small disaccharides with a random cleavage pattern, whereas exolyases successively release disaccharides from the end of the sugar chains and do not produce any larger oligosaccharides in the process. The unsaturated double bond of oligosaccharides produced by lyases shows specific absorption of ultraviolet light at 232 nm, which is beneficial for the detection of CS/DS oligosaccharide products. In addition, CS/DS lyases have many advantages such as more diversity, better stability and activity and simpler preparation compared with hydrolases. Due to the outstanding features above, lyase has a wide range of applications in the preparation of oligosaccharides (Li et al., 2007; Mizumoto et al., 2013a), treatment of nerve damage (Janzadeh et al., 2017; Mondello et al., 2015; Sarveazad et al., 2017), and other CS structure-activity relationship studies.

The sulfation patterns of CS/DS chains play a key role in various functions of CS/DS. Technically, sulfatases with specific activity that allow them to selectively remove sulfate groups from CS/DS chains should be another important tool for the structural and functional studies of CS/DS. Based on the positions of sulfate groups in CS/DS chains, there are several types of specific sulfatases in animals and bacteria, such as N-acetylgalactosamine-4-O-sulfatase (Yamagata et al., 1968; Sugahara and Kojima, 1996; Wang et al., 2019b; Baum et al., 1959) and N-acetylgalactosamine-6-O-sulfatase (Yamagata et al., 1968; Sugahara and Kojima, 1996; Lim and Horwitz, 1981; Singh et al., 1976), which specifically hydrolyze sulfate groups on the C-4 and C-6 positions of GalNAc residues, respectively, and hexuronate-2-O-sulfatase, which specifically removes C-2 sulfate groups from saturated or unsaturated hexuronic acids derived from the digestion of CS/DS by GAG lyases (Sugahara and Kojima, 1996; Myette et al., 2003; Wang et al., 2019a; Tables 2, 3). Based on the sequence similarities, the GAG sulfatases were recently classified in the database SulfAtlas (http://abims.sb-roscoff.fr/sulfatlas/) (Barbeyron et al., 2016). Animal CS/DS sulfatases are lysosomal enzymes responsible for the degradation of endogenous CS/DS, and genetic defects of these enzymes result in the formation of several mucopolysaccharidoses (MPS) in humans, such as MPS II, MPS IVA and MPS VI (Khan et al., 2017). The well-studied animal CS/DS sulfatases are N-acetylgalactosamine-4-O sulfatase (also named Arylsulfatase B, ARSB) (Baum et al., 1959; Wicker et al., 1991), N-acetylgalactosamine-6-O-sulfatase (GALNS) (Lim and Horwitz, 1981; Bielicki and Hopwood, 1991; Tomatsu et al., 1991), iduronate-2-O-sulfatase (IDS) (Lim et al., 1974; Shaklee et al., 1985; Wilson et al., 1990) and Glucurono-2-sulfatase (Arylsulfatase K, ARSK) (Dhamale et al., 2017), which specifically remove the 4-O-sulfate group from sulfated GalNAc residues of CS/DS, 6-O-sulfate group from sulfated GalNAc residues of CS/DS and sulfated galactose of keratan sulfate, 2-O-sulfate groups of sulfate IdoA residues of DS and heparin (Hep), and 2-O-sulfate groups of GlcA residues of heparan sulfate (HS), respectively (Parenti et al., 1997; Tables 2, 3). All these animal CS/DS sulfatases belong to exosulfatases.

In contrast, bacterial CS/DS sulfatases are essential for the biodegradation and utilization of CS/DS from animals, and a large number of potential sulfatase genes have been found in the genomes of various bacteria. However, only a few CS/DS sulfatases have been studied in detail. Three Δ^4,5HexA-2-O-sulfatases have been identified from Flavobacterium heparinum, Bacteroides thetaiotaomicron and Photobacterium sp. FC615, that can specifically remove 2-O-sulfate ester from a ΔHexA residue located at the nonreducing terminus of an unsaturated oligosaccharide (Sugahara and Kojima, 1996; Myette et al., 2003; Ulmer et al., 2014; Wang et al., 2019a). Two N-acetylgalactosamine-4-O-sulfatases from Proteus vulgaris and Photobacterium sp. FC615 specifically hydrolyze 4-O-sulfate groups on GalNAc residues at the reducing ends of CS/DS chains (Sugahara and Kojima, 1996; Wang et al., 2019b). An N-acetylgalactosamine-6-O-sulfatase from Proteus vulgaris has been shown to specifically act on 6-O-sulfates on GalNAc residues at the reducing termini of CS/DS oligosaccharides (Sugahara and Kojima, 1996), and another N-acetylgalactosamine-6-O-sulfatase from Bacteroides thetaiotaomicron can only attack the 6-O-sulfate group on monosaccharide GalNAc residues (Ulmer et al., 2014). Notably, most of the identified CS/DS sulfatases are exo-type enzymes, which only remove sulfate esters from the ends of CS/DS poly-/oligosaccharides and thus have very limited applications to structural and functional studies of CS/DS. Recently, two endo-type N-acetylgalactosamine-4-O-sulfatases were identified from Bacteroides thetaiotaomicron (Ulmer et al., 2014) and Vibrio sp. FC509 (Wang et al., 2015), which can effectively remove 4-O-sulfate groups from GalNAc residues within the CS/DS chains (Table 2 and Figures 2, 3). Compared with exosulfatases, endosulfatases should be more useful enzymatic tools for CS/DS studies but seem to be very rare in nature. Our recent study has shown that CS/DS sulfatases with similar specificities have common signature sequences and can cluster to form a single branch in the phylogenetic tree, although they descended from separate ancestral genes (Wang et al., 2019a). Based on this finding, a series of Δ^4,5HexA-2-O-sulfatases have been successfully predicted and verified from sequences in GenBank (Wang et al., 2019a), and we believe that some CS/DS endosulfatases could also be found by using this method. In fact, the existence of endosulfatases facilitates the catabolic metabolism of CS/DS by bacteria, in which the degradation of CS/DS chains by lyases can be significantly promoted via pre-desulfation by endosulfatases (Wang et al., 2015).

As a kind of highly heterogeneous polysaccharides, the exact structures of all the chains in CS/DS samples cannot be determined with current technology. Thus, the disaccharide composition is used as a basic parameter to characterize various CS/DS preparations used in basic studies and industrial production. Commercial CS/DS preparations are extracted from terrestrial and marine sources, such as the cartilages from bovine, porcine, chicken, shark and squid, and are wildly used in medicines and health products (Bao et al., 2005; Deepa et al., 2007a; Volpi, 2007, 2009; Valcarcel et al., 2017). However, the biological and pharmacological properties of these CS/DS preparations are seriously affected by the raw materials, manufacturing processes and many other factors impacting their production. Disaccharide analysis has been commonly used to evaluate the quality of CS/DS products. CS/DS lyases play a key role in disaccharide composition assays of CS/DS. In general, the disaccharide compositions of various forms of CS/DS with different sulfation patterns can be easily determined by digestion with the commercial CSase ABC followed by anion-exchange HPLC. However, digestion by CSase ABC causes the conversion of both GlcA and IdoA residues into unsaturated uronic acid and thus CS and DS in the test sample cannot be distinguished by this method. In this case, we can use the substrate specificity of the CSase AC and CSase B to investigate the disaccharide composition and proportions of CS and DS in samples.

For example, to determine the disaccharide composition of CS/DS extracted from shark liver, we used CSases that differed in their specificity (CSase ABC, mixture of CSase AC-I and CSase AC-II or CSase B) to digest the sample, and then the digests were labeled with 2-AB and analyzed by anion exchange HPLC on an amine-bound silica PA-03 column using a solvent system of 16 and 530 mM NaH₂PO₄ over a period of 1 h by fluorescent detection. Although the shark liver-derived CS/DS preparation contained highly heterogenous hybrid chains of CS-DS, the disaccharide composition and contents of CS and DS domains in the hybrid chains could be well determined by this method (Li et al., 2007). In summary, CSases with different substrate specificities play an important role in the disaccharide composition assays of CS/DS.

The various biological functions of CS/DS are thought to be due to their functional domains, some oligosaccharide sequences with specific structural features. For a specific target protein, the functional domains of CS/DS chains are usually not a single specific structure but some oligosaccharide domains with similar characteristics, such as a minimum size requirement and the enrichment of specific oversulfated disaccharide units (Trowbridge and Gallo, 2002; Trowbridge et al., 2002; Sugahara et al., 2003; Sugahara and Mikami, 2007; Li et al., 2010). Therefore, isolation of the functional oligosaccharide domains from CS/DS chains is key to not only structure-function relationship studies of CS/DS but also the preparation of functional oligosaccharides with specific activities. Compared with the harsh conditions of chemical and physical methods, the enzymatic method is mild and biocompatible for degrading CS/DS to prepare the functional domains. In general, functional oligosaccharides with specific activity can be obtained through the partial digestion of CS/DS chains with specific enzymes followed by a series of chromatographic separations, especially affinity chromatography. Various CS/DS glycosidic bond-cleaving enzymes, including hyaluronidases and lyases, have been used to prepare CS/DS oligosaccharides. As mentioned above, the CS/DS oligosaccharides produced by lyases bear an unsaturated 4,5-bond between C-4 and C-5 of ΔHexA at their nonreducing ends, which can be easily detected at 232 nm. Moreover, CS/DS lyases show more flexibility in terms of substrate specificity. Thus, CS/DS lyases, including CSase ABC, AC I and B, have been widely used to partially digest various forms of CS/DS for preparing functional oligosaccharides that specifically bind to certain proteins (Fukui et al., 2002; Bao et al., 2005; Kim et al., 2017; Li et al., 2007). In contrast, glycan arrays have the advantages of low dosage, high sensitivity, high throughput, and rapid analysis, which is suitable for the large-scale screening and investigation of potential biological functions of various glycans and their conjugates including CS/DS poly- and oligosaccharides (Fukui et al., 2002). By using this technique, E unit-rich polysaccharides and structure-defined tetrasaccharides have been shown to interact with TNF-α (Tully et al., 2006), bFGF (Miyachi et al., 2015), and midkine-derived and brain-derived neurotrophic factor (Gama et al., 2006) with high affinity. Moreover, this method was used to investigate the interaction between DS and its binding proteins (Yamaguchi et al., 2006). Thus, glycan arrays can be very a powerful tool for the identification of novel functions of CS/DS oligosaccharides with different structures derived from the digestion of CS/DS polysaccharides with various degrading enzymes.

The selective degradation of inactive domains is an ideal way to isolate functional oligosaccharides from CS/DS polysaccharides. Technically, this can be achieved by selecting enzymes with a certain substrate specificity. For example, in a previous study we found that shark skin/liver-derived CS/DS could strongly interact with PTN and HGF to promote neurite outgrowth and this activity could be abolished by treatment with the CSase B but not the CSase AC I. Based on this finding, a PTN- and HGF-binding hexasaccharide containing two iB units was eventually isolated from shark skin CS/DS through selective digestion with the CSase AC I followed by pleiotrophin affinity and anion exchange chromatographies (Li et al., 2007, 2010). In addition to the different selectivity of the CSase AC and B toward uronic acid epimers, some enzymes show high sensitivity to specific sulfation patterns of CS/DS chains, which is important for the preparation of oligosaccharides with specific structures, such as E unit-rich oligosaccharides, which play a key role in neuronal cell adhesion and outgrowth (Ueoka et al., 2000; Mikami et al., 2009; Nandini and Sugahara, 2006; Purushothaman et al., 2007), cancer cell metastasis (Li et al., 2008; Mizumoto et al., 2012, 2013b) and virus infections (Bergefall et al., 2005; Kato et al., 2010), and D unit-rich oligosaccharides, which significantly promote hippocampal neurite outgrowth through interacting with various growth factors and other proteins (Nandini and Sugahara, 2006; Clement et al., 1998; Shida et al., 2019). Testicular hyaluronidase can efficiently digest non-sulfated and low sulfated domains but not highly sulfated domains in CS chains, which makes it a good choice for the selective isolation of highly sulfated domains from CS/DS chains, such as the preparation of D unit-rich oligosaccharides from CS-D (Nadanaka and Sugahara, 1997) and E unit-rich oligosaccharides from CS-E (Kinoshita et al., 2001), respectively. Furthermore, our studies have shown that HCLase like CSase AC I cannot cleave the β–1,4–linkage between GalNAc and D unit (Han et al., 2014), and in contrast, HCLase Er cannot efficiently act on the β–1,4–linkage between E unit and GlcA (Peng et al., 2018), which makes these enzymes more specific tools for the selective isolation of D unit-rich domains and E unit-rich domains in CS/DS chains. Figure 4 shows a schematic diagram of the preparation of various structurally determined hexasaccharides by CS/DS degradation of enzymes with specific activities. In addition, various CS/DS sulfatases combined with sulfotransferases can be potential tools to further edit the sulfation pattern of the obtained oligosaccharides for functional evaluation (Li et al., 2017; Shioiri et al., 2016; Wang et al., 2015; Wang et al., 2019b). Recently, Li et al. developed a method to synthesize CS oligosaccharides using multiple glycosyltransferases and sulfotransferases, and synthesized 15 homogeneous CS oligosaccharide by this method (Li et al., 2017). However, the de novo synthesis of CS/DS oligosaccharides is very cumbersome, heavy workload, time-consuming and costly. By contrast, it can be a relatively simple, efficient and low-cost choice to prepare basic oligosaccharide structures from CS/DS polysaccharides by treatment with various degrading enzymes and further modify these basic structures with specific synthetases to prepare structure-defined oligosaccharides (Cai et al., 2012; Zhang et al., 2019).

In brief, CS/DS-degrading enzymes, including hydrolases, lyases and sulfatases, are very useful tools for preparing functional oligosaccharides with specific structures from CS/DS polysaccharides. With the identification of an increasing number of enzymes with novel substrate specificity, the directional isolation and preparation of oligosaccharides with specific structures from various CS/DS forms will be achieved, which may represent a more feasible and low-cost way to prepare CS/DS oligosaccharides with specific bioactivity compared with synthetic methods.

Sequencing of the isolated functional oligosaccharides with special structures is essential for structure-function studies of CS/DS. Various methods, such as NMR and MS, have been used to determine the structures of CS/DS oligosaccharides. Ly et al. reported successful sequencing of bikunin CS chains with up to 43 saccharide units with FT-ICR-MS/MS (Ly et al., 2011). It is still difficult to distinguish different units with the same sulfation degree, such as monosulfated A and C units or disulfated D and E units. Meanwhile, the distinction of hexuronate epimers and the loss of sulfate groups during analysis are also major challenges. In addition, instrument requirements formethods using NMR and MS are more sophisticated, and NMR, in particular, requires a much greater amount of sample. Sugahara lab has developed a method to analyze and monitor CS/DS oligosaccharides with highly sensitively via the fluorescent labelling of reducing ends with 2-AB followed by anion exchange HPLC, and the detection limit for 2-AB-oligosaccharides is as low as 1 pmol (Kinoshita and Sugahara, 1999). However, further study showed that the 2-AB labelling completely inhibited the cleavage of the β-1,4- linkage in the tetrasaccharide at the reducing end of the CS/DS chain by CSase ABC (Kinoshita and Sugahara, 1999). Whereas, in most cases, some exolyases, such as CSase AC II and HCDLase, can effectively degrade 2-AB-labelled CS oligosaccharides to release the 2-AB-labelled reducing-end disaccharides. By taking advantage of these characteristics of enzymes, CS/DS oligosaccharides that are shorter than decasaccharides can be easily sequenced by enzymatic methods (Bao et al., 2005; Deepa et al., 2007b). Recently, we exploited a novel exolyase HCDLase combined with CSase ABC to sequence a complex octasaccharide (ΔC-A-D-C) (Wang et al., 2017). In this method, the disaccharide composition of the octasaccharide was determined by digestion with the CSase ABC followed by 2-AB labelling and HPLC assay. Then, the octasaccharide was labeled with 2-AB and partially digested with CSase ABC to prepare the 2-AB-labelled reducing-end hexasaccharides. The reducing-end C unit can be directly determined through the digestion of the 2-AB-labelled octasaccharide with HCDLase followed by HPLC-fluorescence detection. To determine the first and second disaccharide units from the nonreducing terminus, the 2-AB-labelled octasaccharide and reducing-end hexasaccharide prepared as described above were individually digested by the CSase ABC and analyzed by HPLC after relabelling with 2-AB. This strategy is theoretically feasible for sequencing longer oligosaccharides but is not easy to achieve due to the rapid increase in operating steps with the increase of saccharide chains. Moreover, Shioiri et al. developed an enzymatic method for the sequencing of a synthesized CS dodecasaccharide (C-C-O-A-O-O) by using a strategy involving dual-fluorescence labelling and dual-digestion (Shioiri et al., 2016). This provides a possibility of sequencing longer CS/DS oligosaccharides. However, the inability of testicular hyaluronidase to cleave DS and highly sulfated CS limits the application of this method, and thus, it is necessary to find alternative enzymes with better features.

The abnormal expression of CS/DS has been shown to be closely related to the occurrence and development of various diseases, such as glial scar formation after brain injury, tumor metastasis, skeletal disorder and viral infection, indicating that the treatment with CS/DS-degrading enzymes should affect the progression of the related diseases and these enzymes might be used as therapeutic agent for the related diseases. In the research of a spinal cord injury (SCI) model, Lemons et al. expounded that CSPGs increased in the lesion and inhibited the growth of axons, that is, inhibited the recovery of the function of the lesion (Lemons et al., 1999; Mondello et al., 2015). Injecting of CSase ABC induces abnormal axon growth or enhance axon regeneration in zebrafish (Bernhardt and Schachner, 2000; Becker and Becker, 2002), adult rats (Bradbury et al., 2002; Janzadeh et al., 2017; Moon et al., 2001), mice (Li et al., 2013), and cats (Mondello et al., 2015). Moreover, CSase ABC combined with other operations such as human adipose derived stem cells (Sarveazad et al., 2017) or low level laser therapy (Janzadeh et al., 2017) can promote the treatment of SCI. Additionally, CSase ABC is used to treat some diseases related to nerve damage, such as glaucoma (Tribble et al., 2018), lumber intervertebral disc (Hoogendoorn et al., 2007; Sugimura et al., 1996; Lü et al., 1997), and to delay the progress of Parkinson’s disease (Kauhausen et al., 2015) and Alzheimer’s disease (Howell et al., 2015). Overall, these studies suggest that CS/DS-degrading enzymes, in particular CSase ABC with broad substrate spectrum, are very promising therapeutic agents for the treatment of nerve injury-related diseases. Most recently, studies have also found that 4-O-sulfated CS GAG chains are increased significantly at the injury site after SCI (Wang et al., 2008) and optic nerve injury (Pearson et al., 2018). ARSB, a mammalian 4-O-sulfatase, treatment improves locomotor function recovery after SCI (Yoo et al., 2013) and improves regeneration after optic nerve injury (Pearson et al., 2018). Comparing with the exo-sulfatases, the newly discovered 4-O-endosulfatases may show better effect in nerve injury repair, which remains to be investigated.

The abnormal expression of CS/DS or CS/DSPGs in cells and tissues is closely related to many tumorigenic processes including cell growth and survival, adhesion, migaration and invasion (Theocharis et al., 2010; Iozzo and Sanderson, 2011; Li et al., 2008). A series of studies have shown that CSases have potential application value in anti-tumor. For example, CSase ABC and CSase AC could significantly inhibited the growth of tumor while streptomyces hyaluronidase, and β-glucuronidase could not (Takeuchi, 1972), CSase AC and CSase B can inhibits the invasion and proliferation of melanoma (Denholm et al., 2001), CSase ABC can assist temozolomide in the treatment of glioblastoma (Jaime-Ramirez et al., 2017), the adhere ability of squamous tongue carcinoma SCC-9 LN-1 cells can be reduced by treatment with CSaseABC (Kawahara et al., 2014), the metastasis of Lewis lung carcinoma LM66-H11 cells can be effectively inhibited by treatment with CSase ABC (Li et al., 2008), and so on (Sullivan et al., 2018). Moreover, Link et al. found that treatment with CSase ABC enhanced integration of both immature and mature self-assembled articular cartilage to native tissue, indicating that Case ABC has a potential therapeutic target for the integration of neocartilage (Link et al., 2020). In fact, some hyaluronidases, such as ovine testicles hyaluronidase (Vitrase^®), bovine testicular hyaluronidase (Hydase^TM) and recombinant human hyaluronidase PH20 (ENHANZE^®), have been clinically used in ophthalmic surgery (Sarvela et al., 1994) and in cosmetic dermatosurgery for the treatment of complications caused by filler injection (Hirsch et al., 2007).

Additionally, the lack of sulfatase leads to the long-term accumulation of highly sulfated oligosaccharides in the lysosome, which cause lysosomal storage disorders MPSs. MPSII (Hunter, OMIM 309900), MPS IVA (Morquio A, OMIM 253000) and MPS VI (Maroteaux-Lamy, OMIM 253200) resulted from the deficiencies of IDS, GALNS and ARSB, respectively (Bondeson et al., 1995; Matalon et al., 1974; Dorfman et al., 1976). Enzyme replacement therapy is the standard treatment option for MPSs, which can start treatment immediately and improve prognosis. Recombinant human IDS (idursulfase and idursulfase beta), GALNS (elosulfase alfa) and ARSB (galsulfase) are clinically used to treat the corresponding MPSs (Whiteman and Kimura, 2017; Sawamoto et al., 2020; Harmatz and Shediac, 2017). Hematopoietic stem cell transplantation is also available for MPSs treatment (Coppa et al., 1995; Mullen et al., 2000). Moreover, gene therapy should be another potential choice for MPSs (Sawamoto et al., 2018; Ponder and Haskins, 2007).

However, the clinical application of CS/DS-degrading enzymes still faces many problems such as immunogenicity, instability and low activity in vivo, which need further study to solve.