β-elimination of hyaluronate by red king crab hyaluronidase

Transcriptome analysis of red king crab hepatopancreas, cloning and sequencing of cDNA encoding the mature form of hyaluronidase

First de-novo transcriptome assembly allowed us to create a database consisting of 101,101 contigues covering a total transcriptome of the red king crab. The length of the transcripts varied from 200 bp to 20,595 bp, with an average length of 720 bp. Next, from the obtained transcripts, we created a local BLAST database and identified a potential gene for the red king crab hyaluronidase based on the similarity with the known arthropod hyaluronidase genes. The transcript containing the coding part of hyaluronidase consisted of 1628 bp with an open reading frame of 1119 bp encoding 372 amino acid protein. The primary amino acid sequence of the identified gene was closest to shrimp hyaluronidase Penaeus vannamei: (Score = 375.9 bits (964), Expect = 3E-125, Identities = 172/313 (54%), Positives = 224/313 (72%), Gaps = 3/313 (1%) (BLAST database).

To clone cDNA hyaluronidase, a total red king crab hepatopancreas (HPC) RNA preparation was obtained. Single-stranded DNA of all polyadenylated mRNAs was produced using oligo (dT)18-primer. After that, PCR was performed using primer oligonucleotides complementary to the hyaluronidase gene sequence identified in red king crab transcriptome analysis. The exception is the position of the amino acid Tyr193, in the codon of which there was a replacement: UAC (cloning)(rightarrow )UAU (transcriptome). In the future the resulting DNA sequence will be used to produce an expression system plasmid construct.

Search for homologues of the identified amino acid sequence of red king crab hyaluronidase

The SwissProt database contains the amino acid sequences of the following enzymes that can use HA as a substrate (date of reference to the database is 06.06.2021):

• EC 3.2.1.35 (catalyzes the cleavage of the ( beta )-(1(rightarrow )4) bond)—61 representatives, of which 2 are found in bacteriophages attacking the bacteria of Streptococcus genus (P15316 and Q54699), and 59 representatives are found in organisms of the kingdom Animalia (classes Insecta, Arachnida, Reptilia, Mammalia, as well as Gastropoda (genus Conus)).

• EC 3.2.1.36 (catalyzes the cleavage of the ( beta )-(1(rightarrow )3) bond)—1 representative, found in Hirudo nipponia (Korean blood-sucking leech) (X4Y2L4).

• EC 4.2.2.1 (catalyzes the cleavage of the ( beta )-(1(rightarrow )4) bond, ( beta )-elimination)—9 representatives (7 of them are found in bacteria and 2 in Homo sapiens), 3 of them (Q12794 (Homo sapiens), Q12891 (Homo sapiens) and P26831 (Clostridium perfringens)) also apply to EC 4.2.2.1 and EC 3.2.1.35.

The red king crab hyaluronidase has a length of 372 amino acid residues (Supplementary Fig. 2) corresponding to the molecular weight of 42.5 kDa, and is the first hyaluronidase described for the Anomura Infraorder. Among the orthologs detected by BLASTP in the SwissProt database there are hyaluronidases of the class 3.2.1.35 (including Q12794 (Homo sapiens), and Q12891 (Homo sapiens)). At the same time, the hyaluronidase of the red king crab is not included in any of the clades formed by known representatives of this class (Fig. 2).

Cleavage of HA at the (beta )-(1(rightarrow )4)-glycosidic bond in the presence of HPC homogenate

The enzymatic activity of the red king crab hypatopancreas homogenate was determined by using a colorimetric method based on the Morgan-Elson reaction^33,34, which leads to a chromogen formation (a color reaction product). N-acetyl-D-glucosamine heated under alkaline conditions forms furan derivatives that react with p-dimethylaminobenzaldehyde to form a red dye which is detected with spectrophotometry. Intense color is associated with high production of N-acetyl-D-glucosamine at the reduced end of the chain.

The commercial drug Lydazum was used as a standard control³⁵ for hyaluronidase activity measurement. The cleavage of HA with HPC homogenate was performed under conditions close to optimal (Supplementary Fig. 3). Remarkably the hyaluronidase of the HPC homogenate exceeds the commercial drug activity by order of magnitude when normalized to the drug’s total protein amount per reaction mixture (Fig. 3). Colored product detection indicates the cleavage of HA at the (beta )-1,4-glycosidic chain bond in the presence of the HPC homogenate enzyme. The colored product is not observed in the case of leech²⁵ or antarctic krill Euphausia superba hyaluronidases¹⁸ because they hydrolyze HA at (beta )-(1(rightarrow )3)-glycosidic bond to produce N-acetyl-D-glucosamine at the non-reduced end of the reaction products, which is not amenable to the Morgan-Elson reaction. The specific activity of the HPC homogenate was 2,100 units of hyaluronidase activity per 1 mg of the total protein (drug Lydazum had 87 units of hyaluronidase activity per 1 mg of the total protein). The titration curve of the HPC homogenate activity units almost precisely overlapped the commercial drug S. hyalurolyticus lyase (Sigma H1136) (Supplementary Fig. 4).

Zymography

Zymography of the HPC homogenate revealed the protein with hyaluronidase activity is roughly 40–50 kDa relative to the molecular marker used in electrophoresis (Fig. 4(a), full–length gel is presented in Supplementary Fig. 5(a)). In addition, heating of the HPC homogenate preparation in the sample buffer (with or without DTT) leads to an irreversible loss of hyaluronidase activity. This is not observed in the case of activity detection by zymography of hyaluronidase from actinobacterium Streptomyces hyalurolyticus—the enzyme retains its activity after heating in the buffer, but the presence of DTT in it prevents the formation of aggregates after heating (Fig. 4(b), full–length gel is presented in Supplementary Fig. 5(b)). According to the protein marker, the molecular weight of S. hyalurolyticus lyase is in the range of 15 to 25 kDa, which is close to S. koganeiensis lyase (21.6 kDa)^36,37. It should be noted that the molecular weight of S. hyalurolyticus lyase declared by the manufacturer is 91 kDa, which was determined by the results of gel filtration. Such a large difference can be explained by the oligomer formation of S. hyalurolyticus lyase under gel-filtration conditions.

Temperature dependent loss of the hyaluronidase activity may be associated with irreversible thermal denaturation of the hyaluronidase, or with the activation of proteolytic enzymes of the HPC homogenate at high temperatures during heating. In this case, it should be assumed that the presence of a detergent at a concentration of 0.2% (SDS) in the buffer did not radically affect the activity of some proteolytic enzyme of the HPC homogenate. Notably, hyaluronidase activity of the HPC homogenate in the substrate gel produced an extended band. This is probably due to the influence of post-translational modifications (glycosylation reduces the electrophoretic mobility of the protein). It was previously shown that the deglycosylation of the sperm membrane hyaluronidase (PH-20 protein) leads to an increase in its electrophoretic mobility in the gel³⁸. Mobility of recombinant bee hyaluronidase of Apis mellifera isolated from E. coli (i.e. without glycosylation) also differed in mobility from the native protein³⁹. Thus, the observed wide band of activity in the gel might be due to the glycosylation of hyaluronidase molecules in the homogenate, which leads to a spread in electrophoretic mobility.

Formation of a double bond during the β-elimination of HA in the presence of HPC homogenate

The formation of a double bond between C-4 and C-5 glucuronic acid atoms during the (beta )-elimination of HA makes it possible to detect the in situ reaction kinetics by measuring the UV absorption in reaction mixture (Fig. 5). As a positive control, the S. hyalurolyticus lyase was used causing an increase of a 232 nm UV absorption in the reaction mixture indicating the formation of a double bond during the cleavage of HA by the (beta )-elimination reaction. As a negative control, the preparation of testicular hyaluronidase (Lydazum) and leech head homogenate were used. As expected, both of these hyaluronidases did not lead to an increase in the UV absorption of the reaction mixture, despite their hydrolysis of HA, which is detected by other methods under the same conditions (Supplementary Fig. 6). Incubation of HPC homogenate in a phosphate buffer without HA in the mixture also did not lead to an increase in absorption at a wavelength of 232 nm. Immediately after the addition of the HPC homogenate to the HA solution, an increase in the absorption of UV radiation was observed, which indicates the cleavage of HA by the homogenate enzymes by the (beta )-elimination mechanism, as in the case of bacterial lyase from our positive control.

Analysis of HA hydrolysis products by (^1)H-NMR spectroscopy

NMR spectroscopy has previously been used to analyze reaction products and study the kinetics of the HA cleavage reaction^{19,29,40,41,42}. As we demonstrated earlier studying the HA of a cosmetic filler, NMR analysis of the reaction mixture can be successfully used to evaluate the kinetics of HA cleavage¹⁹. As shown, the cleavage of HA by the enzymes of the HPC homogenate leads to an increase in the signals of the protons of the acetyl group of N-acetyl-D-glucosamine in the HA: two peaks in the range 2.08–2.02 ppm. One of them corresponds to the proton signal of the N-acetyl-D-glucosamine acetyl group inside the HA chain (“inside protons”, peak 2.032 ppm), and the other one emanates from the N-acetyl-D-glucosamine acetyl group of the HA fragment ends during the cleavage process (“end protons”, 2.074 ppm).

In this work, the process of HA cleavage by S. hyalurolyticus lyase, HPC homogenate, and leech head homogenate was studied. In the case of HPC homogenate and bacterial lyase, two peaks were observed at 2.032 and 2.074 ppm (Fig. 6a). The HA hydrolysis by leech head homogenate results in the appearance of these two peaks roughly from 2.032 and 2.059 ppm. As to the 2.059 ppm peak, it is probably due to the location of N-acetyl-D-glucosamine at the non-reduced end of the HA chains in the case of hydrolysis by leech hyaluronidase, which cleaves the HA by the (beta )-(1(rightarrow )3)-glycosidic bond.

The peak area “end protons” shows the total concentration of all formed chains. Thus, using the NMR spectra of the last time points of the experiments, the average number of dimers in the HA chains formed during hydrolysis was calculated. For the reaction mixture with HPC homogenate, the average number of dimers is close to the mixture with S. hyalurolyticus lyase (2.62 and 2.59, respectively), for leech head homogenate it is 3.95. In addition, the proton signal ratio of the “end protons” to “inside protons” in reaction mixtures with HPC homogenate and bacterial lyase practically does not differ during the entire time of observation of the course of reactions (Supplementary Fig. 7). This once again demonstrates the similarity of the reaction mechanism of the studied hyaluronidase of the eukaryotic organism to the bacterial one.

It is known that the chemical shift of the H-4 proton at the double bond of unsaturated oligoglucuronates is in the down-field region⁴³. As expected, the (^1)H-NMR spectroscopy analysis of the reaction mixture of HA hydrolysis in the presence of leech head homogenate enzymes did not show signals in this region (Figure 6b). In the case of HA cleavage by S. hyalurolyticus lyase and HPC homogenate, a proton signal in the down-field region is detected in the spectrum (5.865 ppm). This confirms that HPC hyaluronidase as well as bacterial lyase cleaves HA by the mechanism of (beta )-elimination with the formation of a double bond.

Ion-exchange chromatography of the HA after treatment by HPC homogenate

We carried out ion-exchange chromatography of the reaction mixture for HA cleavage using S. hyalurolyticus lyase, HPC homogenate, and leech head homogenate (samples from the end point of the experiment of Supplementary Fig. 7). The absorption at 210 nm detected the cleavage products in all three reaction mixtures (Fig. 7). The absorption at 232 nm was used to detect products containing double bonds in (Delta ^{4,5})-unsaturated glucuronic acid residue of oligosaccharides. As expected, absorption at 232 nm was observed in reaction mixtures with S. hyalurolyticus lyase and HPC homogenate, and was not observed in the case of reaction products after treatment with leech head homogenate. At the same time, the results showed that the products of HA degradation obtained using red king crab hyaluronidase are almost identical to the products after HA treatment with bacterial lyase detected both at 210 nm and at 232 nm. As is known, the final product of HA cleavage by S. hyalurolyticus lyase is the unsaturated tetrasaccharide^23,28,29. Thus, the final product of HA cleavage by the HPC homogenate is also the unsaturated tetrasaccharide. The difference between the reaction products of these two enzymes is the presence of an additional peak in the case of lyase in the region of 26–28 min of elution. Small absorption peaks at 210 nm to the left and right of the main one (marked with an asterisk) are low molecular weight components of the substrate solution (Supplementary Fig. 8). In the case of leech hyaluronidase, the final product of hydrolysis is tetrasaccharide^25,26,27. However, on the chromatogram, the peak corresponding to it, as well as other peaks, is slightly shifted relative to the rest of the chromatogram profiles (Fig. 7(c)). This is probably due to a change in the mobility of oligosaccharides with N-acetyl-D-glucosamine at the non-reduced end and D-glucuronic acid at the reduced end (due to hydrolysis by leech hyaluronidase at the (beta )-1,3-glycosidic bond HA). The results of ion-exchange chromatography showed that the products of HA degradation obtained using red king crab hyaluronidase are almost identical to the products after HA treatment with bacterial lyase. Unsaturated tetrasaccharide is the final product of the cleavage reaction by red king crab hyaluronidase.

Substrate specificity: cleavage chondroitin sulfate

Many hyaluronidases show broad specificity and are able to cleave chondroitin sulfate (CS) (Table 1). Using the method of zymography, we showed that the hyaluronidase of the red king crab is able to cleave chondroitin sulfate (Supplementary Fig. 9). However, compared to HA, CS cleavage is much slower, which was shown using the Morgan-Elson reaction (Supplementary Fig. 10) and in situ-monitoring of the double bond formation during the (beta )-elimination reaction of CS (Supplementary Fig. 11). The same insignificant level of CS cleavage was shown by NMR analysis of the reaction mixture containing HPC homogenate (Supplementary Fig. 12). Thus, it can be argued that red king crab hyaluronidase shows slight CSase activity in comparison with HA degradation.