TMEM2 is a bona fide hyaluronidase possessing intrinsic catalytic activity
2023; Elsevier BV; Volume: 299; Issue: 9 Linguagem: Inglês
10.1016/j.jbc.2023.105120
ISSN1083-351X
AutoresTakuma Narita, Yuki Tobisawa, Andrey A. Bobkov, Michael R. Jackson, Chikara Οhyama, Fumitoshi Irie, Yu Yamaguchi,
Tópico(s)Carbohydrate Chemistry and Synthesis
ResumoTransmembrane protein 2 (TMEM2) was originally identified as a membrane-anchored protein of unknown function. We previously demonstrated that TMEM2 can degrade hyaluronan (HA). Furthermore, we showed that induced global knockout of Tmem2 in adult mice results in rapid accumulation of incompletely degraded HA in bodily fluids and organs, supporting the identity of TMEM2 as a cell surface hyaluronidase. In spite of these advances, no direct evidence has been presented to demonstrate the intrinsic hyaluronidase activity of TMEM2. Here, we directly establish the catalytic activity of TMEM2. The ectodomain of TMEM2 (TMEM2ECD) was expressed as a His-tagged soluble protein and purified by affinity and size-exclusion chromatography. Both human and mouse TMEM2ECD robustly degrade fluorescein-labeled HA into 5 to 10 kDa fragments. TMEM2ECD exhibits this HA-degrading activity irrespective of the species of TMEM2 origin and the position of epitope tag insertion. The HA-degrading activity of TMEM2ECD is more potent than that of HYAL2, a hyaluronidase which, like TMEM2, has been implicated in cell surface HA degradation. Finally, we show that TMEM2ECD can degrade not only fluorescein-labeled HA but also native high-molecular weight HA. In addition to these core findings, our study reveals hitherto unrecognized confounding factors, such as the quality of reagents and the choice of assay systems, that could lead to erroneous conclusions regarding the catalytic activity of TMEM2. In conclusion, our results demonstrate that TMEM2 is a legitimate functional hyaluronidase. Our findings also raise cautions regarding the choice of reagents and methods for performing degradation assays for hyaluronidases. Transmembrane protein 2 (TMEM2) was originally identified as a membrane-anchored protein of unknown function. We previously demonstrated that TMEM2 can degrade hyaluronan (HA). Furthermore, we showed that induced global knockout of Tmem2 in adult mice results in rapid accumulation of incompletely degraded HA in bodily fluids and organs, supporting the identity of TMEM2 as a cell surface hyaluronidase. In spite of these advances, no direct evidence has been presented to demonstrate the intrinsic hyaluronidase activity of TMEM2. Here, we directly establish the catalytic activity of TMEM2. The ectodomain of TMEM2 (TMEM2ECD) was expressed as a His-tagged soluble protein and purified by affinity and size-exclusion chromatography. Both human and mouse TMEM2ECD robustly degrade fluorescein-labeled HA into 5 to 10 kDa fragments. TMEM2ECD exhibits this HA-degrading activity irrespective of the species of TMEM2 origin and the position of epitope tag insertion. The HA-degrading activity of TMEM2ECD is more potent than that of HYAL2, a hyaluronidase which, like TMEM2, has been implicated in cell surface HA degradation. Finally, we show that TMEM2ECD can degrade not only fluorescein-labeled HA but also native high-molecular weight HA. In addition to these core findings, our study reveals hitherto unrecognized confounding factors, such as the quality of reagents and the choice of assay systems, that could lead to erroneous conclusions regarding the catalytic activity of TMEM2. In conclusion, our results demonstrate that TMEM2 is a legitimate functional hyaluronidase. Our findings also raise cautions regarding the choice of reagents and methods for performing degradation assays for hyaluronidases. Hyaluronan (HA) is an extremely large polysaccharide of the glycosaminoglycan family, representing one of the major components of extracellular matrices. Via its unique biochemical and biophysical properties, HA plays key roles in regulating the size, fluid pressure, and malleability of tissues. HA and small HA fragments also exert a range of effects on cell behavior and signaling (1Toole B.P. Hyaluronan: from extracellular glue to pericellular cue.Nat. Rev. Cancer. 2004; 4: 528-539Crossref PubMed Scopus (1746) Google Scholar). In vivo, HA is continuously turned over with a short half-life. It is believed that bulky HA polymers in the extracellular space are first cleaved into intermediate-sized fragments, which are subsequently internalized and degraded into oligosaccharides and monosaccharides in endosomes and lysosomes (2Stern R. Kogan G. Jedrzejas M.J. Soltes L. The many ways to cleave hyaluronan.Biotechnol. Adv. 2007; 25: 537-557Crossref PubMed Scopus (329) Google Scholar). Although multiple proteins in different cellular compartments have been implicated in these catabolic processes, the identity of the hyaluronidase that acts on the cell surface has been uncertain. Although HYAL2 has been shown to be anchored to the cell surface via a glycosylphosphatidylinositol linkage (3Rai S.K. Duh F.M. Vigdorovich V. Danilkovitch-Miagkova A. Lerman M.I. Miller A.D. Candidate tumor suppressor HYAL2 is a glycosylphosphatidylinositol (GPI)-anchored cell-surface receptor for jaagsiekte sheep retrovirus, the envelope protein of which mediates oncogenic transformation.Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 4443-4448Crossref PubMed Scopus (302) Google Scholar, 4Cherr G.N. Yudin A.I. Overstreet J.W. The dual functions of GPI-anchored PH-20: hyaluronidase and intracellular signaling.Matrix Biol. 2001; 20: 515-525Crossref PubMed Scopus (151) Google Scholar), its acidic pH optimum (5Afify A.M. Stern M. Guntenhoner M. Stern R. Purification and characterization of human serum hyaluronidase.Arch. Biochem. Biophys. 1993; 305: 434-441Crossref PubMed Scopus (100) Google Scholar, 6Lepperdinger G. Strobl B. Kreil G. HYAL2, a human gene expressed in many cells, encodes a lysosomal hyaluronidase with a novel type of specificity.J. Biol. Chem. 1998; 273: 22466-22470Abstract Full Text Full Text PDF PubMed Scopus (261) Google Scholar) and the intracellular localization of endogenous HYAL2 protein (7Chow G. Knudson C.B. Knudson W. Expression and cellular localization of human hyaluronidase-2 in articular chondrocytes and cultured cell lines.Osteoarthritis Cartilage. 2006; 14: 849-858Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar) suggest that it may not be a full-time cell surface hyaluronidase. In addition, in spite of the rapid turnover of extracellular HA during embryonic development, Hyal2−/− mice exhibit only mild embryonic phenotype (8Martin D.C. Atmuri V. Hemming R.J. Farley J. Mort J.S. Byers S. et al.A mouse model of human mucopolysaccharidosis IX exhibits osteoarthritis.Hum. Mol. Genet. 2008; 17: 1904-1915Crossref PubMed Scopus (80) Google Scholar, 9Jadin L. Wu X. Ding H. Frost G.I. Onclinx C. Triggs-Raine B. et al.Skeletal and hematological anomalies in HYAL2-deficient mice: a second type of mucopolysaccharidosis IX?.FASEB J. 2008; 22: 4316-4326Crossref PubMed Scopus (86) Google Scholar, 10Chowdhury B. Hemming R. Hombach-Klonisch S. Flamion B. Triggs-Raine B. Murine hyaluronidase 2 deficiency results in extracellular hyaluronan accumulation and severe cardiopulmonary dysfunction.J. Biol. Chem. 2013; 288: 520-528Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). In 2017, we reported that transmembrane protein 2 (TMEM2), a type II transmembrane protein of unknown function, is the sought-after cell surface hyaluronidase (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). Our subsequent studies have assembled multiple pieces of evidence supporting the physiological relevance of TMEM2 as a functional hyaluronidase. Specifically, we showed that: (i) tamoxifen-induced global knockout of Tmem2 in adult mice leads to rapid and pronounced accumulation of uncleaved and partially cleaved HA in bodily fluids and a variety of organs (12Tobisawa Y. Fujita N. Yamamoto H. Ohyama C. Irie F. Yamaguchi Y. The cell surface hyaluronidase TMEM2 is essential for systemic hyaluronan catabolism and turnover.J. Biol. Chem. 2021; 297101281Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar); (ii) Wnt1-Cre–driven Tmem2 conditional knockout results in severe defects in neural crest derivatives, accompanied by aberrant HA accumulation in those tissues (13Inubushi T. Nakanishi Y. Abe M. Takahata Y. Nishimura R. Kurosaka H. et al.The cell surface hyaluronidase TMEM2 plays an essential role in mouse neural crest cell development and survival.PLoS Genet. 2022; 18e1009765Crossref PubMed Scopus (5) Google Scholar); and (iii) knockdown of TMEM2 in human tumor cells greatly impairs their ability to degrade HA in a contact-dependent manner (14Irie F. Tobisawa Y. Murao A. Yamamoto H. Ohyama C. Yamaguchi Y. The cell surface hyaluronidase TMEM2 regulates cell adhesion and migration via degradation of hyaluronan at focal adhesion sites.J. Biol. Chem. 2021; 296100481Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar). Up until now, however, there has been no direct evidence demonstrating that the TMEM2 protein itself has intrinsic HA-degrading activity. The lack of such data appears to have caused some confusion and skepticism concerning whether TMEM2 is a functional hyaluronidase. In this paper, we present multiple pieces of evidence demonstrating that TMEM2 is a bona fide hyaluronidase that has intrinsic HA-degrading activity. This work also identifies technical factors that may confound the outcome of assays for TMEM2 enzyme activity. To determine whether TMEM2 has intrinsic hyaluronidase activity, we produced a recombinant soluble human TMEM2 protein representing its ectodomain (hereafter designated as TMEM2ECD). cDNA corresponding to Ser104–His1383 of human TMEM2, with a 6× His tag fused to its C terminus (Figs. 1A and S1), was transfected in FreeStyle 293-F cells, and TMEM2ECD was purified from culture supernatants (Fig. 1B). HA degradation assays with purified human TMEM2ECD were performed using fluorescein-labeled HA (hereafter designated as FA–HA) from Cosmo Bio (1200–1600 kDa) as a substrate, and the degradation of FA–HA was monitored by agarose gel electrophoresis. As shown in Figure 1C, FA–HA was completely degraded into 5 to 10 kDa fragments in 6 h, with the bulk of reaction occurring within 2 h. Analysis by size-exclusion chromatography yielded a similar result (Fig. 1D). These results directly demonstrate that the human TMEM2 ectodomain possesses intrinsic HA-degrading activity. Significantly, HA degradation by purified TMEM2ECD occurs much more rapidly than the HA degradation observed in cell-based assays using TMEM2-transfected 293T cells (∼72 h; (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar)), the system we originally employed to demonstrate the hyaluronidase activity of TMEM2. This large discrepancy in reaction rates suggests that the cell-based assays may be influenced by a component(s) that suppresses the activity of TMEM2 (see below for these data: Factors that may confound results obtained in TMEM2 hyaluronidase assays). To substantiate this initial finding, we performed four corroborating experiments using different recombinant TMEM2 proteins and assay systems. First, we purified mouse TMEM2ECD with the same structural configuration as human TMEM2ECD (Fig. 1B). As observed for human TMEM2ECD, mouse TMEM2ECD exhibits robust HA-degrading activity (Fig. 2A). The time course of FA–HA degradation by mouse TMEM2ECD is faster than that of human TMEM2ECD, the reaction being almost complete within 1 h. Second, we examined the possible effect of the position of the 6× His tag on the HA-degrading activity of TMEM2ECD. For this, we produced a second human TMEM2ECD in which a 6× His tag was fused to its N terminus. This N-terminally tagged human TMEM2ECD exhibits an HA-degrading activity that is nearly as robust as that of C-terminally fused TMEM2ECD (Fig. 2B). Third, we examined whether full-length TMEM2 associated with cell membranes exhibits HA-degrading activity. For this, the HA degradation assays were performed with the isolated membrane fraction of TMEM2-transfected 293T cells as the enzyme source. These assays demonstrate that both human and mouse full-length TMEM2 degrade HA (Fig. 2C). As seen in the experiments with TMEM2ECD (see Fig. 2A), mouse TMEM2 exhibits stronger activity than human TMEM2 in this assay system. Finally, the HA-degrading activities of both human and mouse full-length TMEM2 are further confirmed by in situ HA degradation assays using transfected 293 cells (Fig. 2D). Together, these results demonstrate that TMEM2 possesses intrinsic HA-degrading activity, independent of the species of TMEM2 origin, the position of epitope tag insertion, and whether or not it is anchored to cell membrane. To define the physiological importance of TMEM2 as a functional hyaluronidase, we performed two additional studies. First, we asked how the observed HA-degrading activity of TMEM2 compares with that of established hyaluronidases. To address this question, we chose to compare TMEM2 with HYAL2, which, like TMEM2, has been implicated in HA degradation on the cell surface (15Csoka A.B. Frost G.I. Stern R. The six hyaluronidase-like genes in the human and mouse genomes.Matrix Biol. 2001; 20: 499-508Crossref PubMed Scopus (505) Google Scholar, 16Triggs-Raine B. Natowicz M.R. Biology of hyaluronan: insights from genetic disorders of hyaluronan metabolism.World J. Biol. Chem. 2015; 6: 110-120Crossref PubMed Google Scholar). The recombinant human HYAL2 (R&D Systems) used for this analysis is appropriate for the side-by-side comparison—like our human TMEM2ECD, it is produced as a soluble protein with a C-terminal 6× His tag. Assays were performed at the pH optima for respective enzymes (pH 5.0 for HYAL2 (17Wu Z.L. Ertelt J.M. Assays for hyaluronidases and heparanase using nonreducing end fluorophore-labeled hyaluronan and heparan sulfate proteoglycan.Glycobiology. 2021; 31: 1435-1443Crossref PubMed Scopus (3) Google Scholar); pH 7.0 for TMEM2ECD (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar)). The results of the comparison demonstrate that while TMEM2ECD degrades FA–HA completely into 5 to 10 kDa fragments in 2 h, the reaction by HYAL2 does not appear to reach completion at the same time point (Fig. 3A). Moreover, the degradation products of HYAL2 action are much larger (ranging from 100 to 300 kDa) and more polydisperse than those produced by TMEM2ECD. These results indicate that the rate of cleavage is faster for TMEM2ECD than for HYAL2 and that TMEM2ECD cleaves HA polymers at many more sites than HYAL2. Assuming, for example, that the average molecular size of the fragments generated by TMEM2ECD is 10 kDa and that of HYAL2 is 200 kDa (estimates from the migration in agarose gels at 2 h), TMEM2ECD is estimated to cut at 20 times more sites along the HA during this time period. Thus, the enzymatic activity of TMEM2 appears significantly more potent than that of HYAL2, an observation that seems to be in line with the rapid and pronounced accumulation of HA in Tmem2 knockout than observed in Hyal2 knockout (12Tobisawa Y. Fujita N. Yamamoto H. Ohyama C. Irie F. Yamaguchi Y. The cell surface hyaluronidase TMEM2 is essential for systemic hyaluronan catabolism and turnover.J. Biol. Chem. 2021; 297101281Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar, 18Bourguignon V. Flamion B. Respective roles of hyaluronidases 1 and 2 in endogenous hyaluronan turnover.FASEB J. 2016; 30: 2108-2114Crossref PubMed Scopus (46) Google Scholar). These results indicate that the HA-degrading activity of TMEM2 is sufficiently robust for TMEM2 to be considered as a functional hyaluronidase. As a second additional means of establishing the physiological relevance of TMEM2, we examined the ability of TMEM2ECD to degrade unlabeled, high-molecular weight HA. This determination is necessary, because our results with FA–HAs from commercial sources other than Cosmo Bio have raised a note of caution regarding the use of fluorescence-labeled HA as a surrogate substrate for hyaluronidase assays (see the next section below for these data). Since agarose gel electrophoresis is not suitable for quantitative analysis of unlabeled HA, we instead used size-exclusion chromatography combined with the quantification of HA using the HA-binding protein assay (12Tobisawa Y. Fujita N. Yamamoto H. Ohyama C. Irie F. Yamaguchi Y. The cell surface hyaluronidase TMEM2 is essential for systemic hyaluronan catabolism and turnover.J. Biol. Chem. 2021; 297101281Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar). In addition, to enable reliable quantification of HA by this method, the degradation reaction was performed with 20-fold more HA than in the assays with FA–HA. This analysis demonstrates that unlabeled HA (1500–1750 kDa) was degraded by TMEM2ECD into 5 to 10 kDa fragment in a time-dependent manner (Fig. 3B). The longer time needed for complete degradation than that seen in assays with FA–HA most likely reflects the lower enzyme/substrate ratio used in this experiment. This result confirms the functional relevance of TMEM2 as an enzyme that degrades native high-molecular weight HA. Recently, two papers reported that they could not detect HA-degrading activity in TMEM2 (19Niu M. McGrath M. Sammon D. Gardner S. Morgan R.M. Di Maio A. et al.Structure of the transmembrane protein 2 (TMEM2) ectodomain and its apparent lack of hyaluronidase activity.Wellcome Open Res. 2023; 8: 76PubMed Google Scholar, 20Sato S. Miyazaki M. Fukuda S. Mizutani Y. Mizukami Y. Higashiyama S. et al.Human TMEM2 is not a catalytic hyaluronidase, but a regulator of hyaluronan metabolism via HYBID (KIAA1199/CEMIP) and HAS2 expression.J. Biol. Chem. 2023; 299104826Abstract Full Text Full Text PDF Scopus (2) Google Scholar). Niu et al. (19Niu M. McGrath M. Sammon D. Gardner S. Morgan R.M. Di Maio A. et al.Structure of the transmembrane protein 2 (TMEM2) ectodomain and its apparent lack of hyaluronidase activity.Wellcome Open Res. 2023; 8: 76PubMed Google Scholar) reported that recombinant soluble human TMEM2, similar to our TMEM2ECD, did not degrade FA–HA. While they employed an assay system similar to ours, a potentially key difference between their experiments and ours is the source of FA–HA products used as a substrate. We have consistently been using FA–HA from Cosmo Bio (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 12Tobisawa Y. Fujita N. Yamamoto H. Ohyama C. Irie F. Yamaguchi Y. The cell surface hyaluronidase TMEM2 is essential for systemic hyaluronan catabolism and turnover.J. Biol. Chem. 2021; 297101281Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar), while Niu et al. used FA–HA from Biosynth. Since these FA–HA products differ significantly in both HA size and the degree of substitution with fluorescein moieties (Table S1), we compared these FA–HA products side-by-side, along with a third FA–HA product from Creative PEGWorks. Surprisingly, agarose gel electrophoresis revealed that these three FA–HA products differ markedly in their quality (Fig. 4A). While the FA–HA from Cosmo Bio consists of a single high-molecular weight band that is completely degraded by TMEM2ECD, the other two FA–HA products contain no components that represent high-molecular weight HA. Instead, they contain multiple species of polydisperse smears and bands that are not affected by TMEM2ECD (Fig. 4A) or HYAL2 (Fig. S2). We have thus far been unable to determine the molecular identity of these smears/bands. Since neither TMEM2ECD nor HYAL2 degrades these unidentified smears/bands, they may represent non-HA polysaccharides that contaminate the supposedly "pure" HA used for fluorescein labeling. Another possibility is that differences in fluorescein labeling condition and in the degree of substitution may have generated denatured or partially degraded HA species that are resistant to treatment with these hyaluronidases. In any event, these results emphasize the needs to re-scrutinize earlier data generated by use of these questionable FA–HA products. In a second contradictory paper, Sato et al. (20Sato S. Miyazaki M. Fukuda S. Mizutani Y. Mizukami Y. Higashiyama S. et al.Human TMEM2 is not a catalytic hyaluronidase, but a regulator of hyaluronan metabolism via HYBID (KIAA1199/CEMIP) and HAS2 expression.J. Biol. Chem. 2023; 299104826Abstract Full Text Full Text PDF Scopus (2) Google Scholar) reported that they could not detect HA-degrading activity of human TMEM2 using a cell-based assay with transfected 293T cells, although mouse TMEM2 exhibited HA-degrading activity in the same assay. These authors claim that human TMEM2 (in contrast to mouse TMEM2) is not a hyaluronidase. Since we noted earlier in the Results that cell-based HA-degrading assays might contain a factor(s) that suppresses the activity of TMEM2, we analyzed the HA-degrading activity of culture supernatants from human TMEM2ECD-transfected 293T cells. While little HA-degrading activity is detected in TMEM2ECD-containing culture supernatant, a noticeable HA-degrading activity emerges upon 30-fold concentration of the same culture supernatant using a 50 kDa cutoff spin column (Fig. 4B). Remarkably, 30-fold dilution of this concentrated supernatant with the regular reaction buffer results in the emergence of a highly robust activity. These results indicate that although strongly suppressed under the conditions used for cell-based HA-degrading assays, human TMEM2 does indeed possess HA-degrading activity. The suppression of TMEM2 activity must be mediated by an unidentified substance(s) that passes through a 50 kDa cutoff membrane and is then reduced to insignificant levels upon 30-fold dilution. At present, the identity and the origin of the substance(s) are unknown (see Discussion). Since our first report identifying TMEM2 as a novel cell surface hyaluronidase (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar), we have presented extensive data illustrating the functional significance of TMEM2 in systemic HA catabolism (12Tobisawa Y. Fujita N. Yamamoto H. Ohyama C. Irie F. Yamaguchi Y. The cell surface hyaluronidase TMEM2 is essential for systemic hyaluronan catabolism and turnover.J. Biol. Chem. 2021; 297101281Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar), cell adhesion and migration (14Irie F. Tobisawa Y. Murao A. Yamamoto H. Ohyama C. Yamaguchi Y. The cell surface hyaluronidase TMEM2 regulates cell adhesion and migration via degradation of hyaluronan at focal adhesion sites.J. Biol. Chem. 2021; 296100481Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar), and embryonic development (13Inubushi T. Nakanishi Y. Abe M. Takahata Y. Nishimura R. Kurosaka H. et al.The cell surface hyaluronidase TMEM2 plays an essential role in mouse neural crest cell development and survival.PLoS Genet. 2022; 18e1009765Crossref PubMed Scopus (5) Google Scholar). Yet, a lack of data directly demonstrating the catalytic activity of purified TMEM2 protein has left some uncertainty about the role TMEM2 as a hyaluronidase. This has led to speculation that TMEM2 may not itself be a hyaluronidase but rather may act as an accessory or regulatory molecule for a known or unknown HA degradation system (20Sato S. Miyazaki M. Fukuda S. Mizutani Y. Mizukami Y. Higashiyama S. et al.Human TMEM2 is not a catalytic hyaluronidase, but a regulator of hyaluronan metabolism via HYBID (KIAA1199/CEMIP) and HAS2 expression.J. Biol. Chem. 2023; 299104826Abstract Full Text Full Text PDF Scopus (2) Google Scholar). Here, we present data that convincingly demonstrate the catalytic activity of TMEM2. The key findings of this paper include (i) purified TMEM2ECD exhibits HA-degrading activity without requiring the participation of any other proteins; (ii) this activity is demonstrated in multiple recombinant TMEM2 proteins and assays; (iii) the activity of TMEM2ECD is more potent than that of HYAL2; and (iv) TMEM2ECD degrades native high-molecular weight HA. It should also be stressed that contrary to the recent claim that human TMEM2 lacks HA-degrading catalytic activity (20Sato S. Miyazaki M. Fukuda S. Mizutani Y. Mizukami Y. Higashiyama S. et al.Human TMEM2 is not a catalytic hyaluronidase, but a regulator of hyaluronan metabolism via HYBID (KIAA1199/CEMIP) and HAS2 expression.J. Biol. Chem. 2023; 299104826Abstract Full Text Full Text PDF Scopus (2) Google Scholar), our results clearly demonstrate that while somewhat less active than mouse TMEM2, human TMEM2 possesses intrinsic HA-degrading activity that is more potent than that of human HYAL2. In addition to these core findings, our study has revealed factors that may confound the results of TMEM2 hyaluronidase assays. One such factor is the use of FA–HA as a substrate. As revealed in this study, the quality of FA–HA is vastly different among products from different vendors. From the migration patterns in agarose gel, the FA–HA products from sources other than Cosmo Bio consist of multiple fluorescein-labeled bands and smears that do not appear to be intact HA. Those poorly characterized molecular species may represent degraded or denatured HA or even molecules totally unrelated to HA. Since the vendors of FA–HA do not provide detailed information regarding the source and purification of HA species used for labeling, it is difficult to speculate about the cause(s) for these differences in FA–HA quality. However, it is known that bacteria used for large-scale fermentation for HA production, such as Streptococci, can produce not only HA but also other nonsulfated glycosaminoglycans (21Radhouani H. Correia S. Gonçalves C. Reis R.L. Oliveira J.M. Glycosaminoglycans.in: Oliveira J.M. Radhouani H. Reis R.L. Polysaccharides of Microbial Origin: Biomedical Applications. Springer Nature, Cham, Swizerland2020: 167-184Google Scholar, 22Datta P. Linahardt R.J. Sharfstein S.T. Industrial production of glycosaminoglycans.in: Encyclopedia of Microbiology. 4th ed. Elsevier, Amsterdam2017: 681-690Google Scholar). Also, a high degree of substitution with fluorescein labels, as seen in the FA–HA from Biosynth, could affect the recognizability and/or cleavability by TMEM2, as the fluorescent labels can significantly affect the chemical properties of labeled molecules (23Yin L. Wang W. Wang S. Zhang F. Zhang S. Tao N. How does fluorescent labeling affect the binding kinetics of proteins with intact cells?.Biosens. Bioelectron. 2015; 66: 412-416Crossref PubMed Scopus (47) Google Scholar, 24Liu Y. Xiao M. Zhao J. Zhang X. Hu X. Goff H.D. et al.Fluorescent labeling affected the structural/conformational properties of arabinoxylans.Carbohydr. Polym. 2021; 265118064Crossref Scopus (11) Google Scholar). Although we cannot rule out the possibility that the problematic nature of the Biosynth FA–HA is specific to the lot we obtained, our results nonetheless emphasize the importance of a careful quality check of purchased FA–HA products, as well as the need to re-examine earlier data generated via use of these FA–HA products. Another potentially confounding factor uncovered in this study is the inhibitory effect of unknown components present in cell-based HA degradation assays. This is most likely the reason why our earlier experiments required a much longer time course for complete HA degradation (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar) than we observe in assays with purified TMEM2ECD. The failure of Sato et al. (20Sato S. Miyazaki M. Fukuda S. Mizutani Y. Mizukami Y. Higashiyama S. et al.Human TMEM2 is not a catalytic hyaluronidase, but a regulator of hyaluronan metabolism via HYBID (KIAA1199/CEMIP) and HAS2 expression.J. Biol. Chem. 2023; 299104826Abstract Full Text Full Text PDF Scopus (2) Google Scholar) to detect the catalytic activity of human TMEM2 might also have been due to this issue. The identity of the inhibitory substance(s) and its origin, whether it is a component(s) of culture media or a cell-derived factor, is currently unknown. Standard components of cell culture media, such as vitamins and sodium pyruvate, have been shown to inhibit the activity of various enzymes (25Rao Y. Gammon S.T. Sutton M.N. Zacharias N.M. Bhattacharya P. Piwnica-Worms D. Excess exogenous pyruvate inhibits lactate dehydrogenase activity in live cells in an MCT1-dependent manner.J. Biol. Chem. 2021; 297100775Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar, 26Zhao Y.J. He W.M. Zhao Z.Y. Li W.H. Wang Q.W. Hou Y.N. et al.Acidic pH irreversibly activates the signaling enzyme SARM1.FEBS J. 2021; 288: 6783-6794Crossref PubMed Scopus (9) Google Scholar, 27Radons J. Heller B. Burkle A. Hartmann B. Rodriguez M.L. Kroncke K.D. et al.Nitric oxide toxicity in islet cells involves poly(ADP-ribose) polymerase activation and concomitant NAD+ depletion.Biochem. Biophys. Res. Commun. 1994; 199: 1270-1277Crossref PubMed Scopus (202) Google Scholar, 28Uchigata Y. Yamamoto H. Kawamura A. Okamoto H. Protection by superoxide dismutase, catalase, and poly(ADP-ribose) synthetase inhibitors against alloxan- and streptozotocin-induced islet DNA strand breaks and against the inhibition of proinsulin synthesis.J. Biol. Chem. 1982; 257: 6084-6088Abstract Full Text PDF PubMed Google Scholar, 29Denu J.M. Vitamin B3 and sirtuin function.Trends Biochem. Sci. 2005; 30: 479-483Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar, 30Ahn H. Lee G.S. Riboflavin, vitamin B2, attenuates NLRP3, NLRC4, AIM2, and non-canonical inflammasomes by the inhibition of caspase-1 activity.Sci. Rep. 2020; 1019091Crossref Scopus (28) Google Scholar, 31Sisecioglu M. Cankaya M. Ozdemir H. Effects of some vitamins on lactoperoxidase enzyme activity.Int. J. Vitam. Nutr. Res. 2009; 79: 188-194Crossref PubMed Scopus (6) Google Scholar, 32Nishino T. Tsushima K. Interaction of milk xanthine oxidase with folic acid. Inhibition of milk xanthine oxidase by folic acid and separation of the enzyme into two fractions on Sepharose 4B/folate gel.J. Biol. Chem. 1986; 261: 11242-11246Abstract Full Text PDF PubMed Google Scholar, 33Hwang S.Y. Siow Y.L. Au-Yeung K.K. House J. O K. Folic acid supplementation inhibits NADPH oxidase-mediated superoxide anion production in the kidney.Am. J. Physiol. Renal Physiol. 2011; 300: F189-198Crossref PubMed Scopus (70) Google Scholar, 34Ali F. Manzoor U. Bhattacharya R. Bansal A.K. Chandrashekharaiah K.S. Singh L.R. et al.Brain metabolite, myo-inositol, inhibits catalase activity: a mechanism of the distortion of the antioxidant defense system in Alzheimer's disease.ACS Omega. 2022; 7: 12690-12700Crossref PubMed Scopus (6) Google Scholar). Inorganic anions in culture media, serum added to media, and lactate produced by cells could all reduce the effective concentration of calcium (35Bazydlo L.A. Needham M. Harris N.S. Calcium, magnesium, and phosphate.Lab. Med. 2014; 45: e44-e50Crossref Google Scholar), a necessary factor for TMEM2 activity (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). The possibility that 293T cells secrete a biological TMEM2 inhibitor(s) cannot be ruled out at present. Lastly, although not specifically investigated in this study, it should be noted that, when FA–HA is used as a substrate for hyaluronidase assays, agarose gel electrophoresis is preferable to size-exclusion chromatography for the analysis of the degradation products. Agarose gel electrophoresis allows the analysis of HA fragments over the entire molecular weight range that needs to be covered. In contrast, size-exclusion chromatography using a resin with a small pore size, such as Sephacryl S-200, poses a risk of missing partial HA degradation in the high-molecular weight range. In conclusion, our results clearly demonstrate that TMEM2 is a bona fide hyaluronidase that possesses an intrinsic and physiologically relevant activity. While the mechanistic details of TMEM2-mediated HA degradation on the cell surface remain to be determined, these new findings are important in resolving uncertainties that have surrounded the role of TMEM2 as a hyaluronidase. This resolution should facilitate research on this intriguing molecule that is thought to play important roles in embryonic development (13Inubushi T. Nakanishi Y. Abe M. Takahata Y. Nishimura R. Kurosaka H. et al.The cell surface hyaluronidase TMEM2 plays an essential role in mouse neural crest cell development and survival.PLoS Genet. 2022; 18e1009765Crossref PubMed Scopus (5) Google Scholar, 36Smith K.A. Lagendijk A.K. Courtney A.D. Chen H. Paterson S. Hogan B.M. et al.Transmembrane protein 2 (Tmem2) is required to regionally restrict atrioventricular canal boundary and endocardial cushion development.Development. 2011; 138: 4193-4198Crossref PubMed Scopus (46) Google Scholar, 37Totong R. Schell T. Lescroart F. Ryckebusch L. Lin Y.F. Zygmunt T. et al.The novel transmembrane protein Tmem2 is essential for coordination of myocardial and endocardial morphogenesis.Development. 2011; 138: 4199-4205Crossref PubMed Scopus (51) Google Scholar, 38De Angelis J.E. Lagendijk A.K. Chen H. Tromp A. Bower N.I. Tunny K.A. et al.Tmem2 regulates embryonic Vegf signaling by controlling hyaluronic acid turnover.Dev. Cell. 2017; 40: 123-136Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar), tumor progression (39Lee H. Goodarzi H. Tavazoie S.F. Alarcon C.R. TMEM2 is a SOX4-regulated gene that mediates metastatic migration and invasion in breast cancer.Cancer Res. 2016; 76: 4994-5005Crossref PubMed Scopus (60) Google Scholar, 40Kudo Y. Sato N. Adachi Y. Amaike T. Koga A. Kohi S. et al.Overexpression of transmembrane protein 2 (TMEM2), a novel hyaluronidase, predicts poor prognosis in pancreatic ductal adenocarcinoma.Pancreatology. 2020; 20: 1479-1485Crossref PubMed Scopus (12) Google Scholar), and aging (41Schinzel R.T. Higuchi-Sanabria R. Shalem O. Moehle E.A. Webster B.M. Joe L. et al.The hyaluronidase, TMEM2, promotes ER homeostasis and longevity independent of the UPR(ER).Cell. 2019; 179: 1306-1318.e18Abstract Full Text Full Text PDF PubMed Scopus (75) Google Scholar). For this study, three FA–HA products from different manufacturers were used as substrates for HA degradation assays, namely: CSR-FAHA-H2 from Cosmo Bio; YH4531 from Biosynth; and HA-804 from Creative PEGWorks. Unlabeled HA (1500 ∼ 1750 kDa) was obtained from Sigma (63357), while FA–HA molecular weight standards (CSR-FAHA-H2, CSR-FAHA-L2, and CSR-FAHA-U2) were from Cosmo Bio. Recombinant soluble human HYAL2 protein was purchased from R&D Systems (11012-GH-020). Expression plasmid for human TMEM2ECD fused with a C-terminal 6× His tag was constructed by inserting the cDNA into pSecTag2B, in which the original Myc/His tag was inactivated by inserting a stop codon. For the production of mouse TMEM2ECD, the construct used for the expression of a tag-less TMEM2 ectodomain (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar) was modified by inserting a 6× His tag at its C terminus. TMEM2ECD proteins were produced using the FreeStyle 293-F system (ThermoFisher, R79007), as described previously (42Fifield A.L. Hanavan P.D. Faigel D.O. Sergienko E. Bobkov A. Meurice N. et al.Molecular inhibitor of QSOX1 suppresses tumor growth in vivo.Mol. Cancer Ther. 2020; 19: 112-122Crossref PubMed Scopus (13) Google Scholar). Culture supernatants were first applied to HisPur Cobalt Resin (ThermoFisher, R79007), and bound TMEM2ECD proteins were eluted with PBS containing 0.5 M imidazole and 10% glycerol. TMEM2ECD proteins were further purified by size-exclusion chromatography on HiLoad 26/600 Superdex 200 (Cytiva, 28989336). Concentration of purified TMEM2ECD proteins was determined by measuring absorbance at 280 nm. For assays with FA–HA as substrates, 2.5 μg of purified TMEM2ECD was incubated at 37 °C with 0.25 μg FA–HA in 500 μl of 50 mM Hepes buffer (pH 7.0) containing 2 mM CaCl2 ("reaction buffer") at 37 °C. For assays with unlabeled HA as substrates, 10 μg of purified human TMEM2ECD was incubated with 5 μg HA in the reaction buffer at 37 °C. In the comparative analysis of N- and C-terminally 6× His-tagged TMEM2ECD proteins and the analysis of TMEM2 inhibition, 293T cells were transfected with respective TMEM2ECD constructs and cultured in Opti-MEM serum-free medium for 3 days. After a 30-fold concentration using Vivaspin 6 (50 kDa cutoff, Cytiva 28932294), the concentrated culture supernatants were used as enzyme sources in HA degradation assays. In the case of the former analysis, the input of N- and C-terminally tagged TMEM2ECD proteins was normalized based on the densitometric analysis of immunoblots with anti-TMEM2 polyclonal antibody (Invitrogen PA5-85901) using ChemiDoc MP Imaging System (Bio-Rad). For assays with full-length TMEM2 in the membrane fraction, C-terminally FLAG-tagged TMEM2 were transfected into 293T cells, from which the P2 membrane fraction was prepared as described (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar) and used as enzyme source. In all these experiments, the enzyme reaction was terminated by adding formamide (final 10 M). In experiments using FA–HAs as substrate, degradation of HA was analyzed by agarose gel electrophoresis in 1% SeaKem GTG (Lonza 50070) or size-exclusion chromatography on Sephacryl S-300HR (Cytiva 17059910) (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 12Tobisawa Y. Fujita N. Yamamoto H. Ohyama C. Irie F. Yamaguchi Y. The cell surface hyaluronidase TMEM2 is essential for systemic hyaluronan catabolism and turnover.J. Biol. Chem. 2021; 297101281Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar). Degradation of unlabeled HA was analyzed by size-exclusion chromatography on Sepharose CL-2B (Cytiva 17014001), combined with the quantification of HA using the Hyaluronic Acid LT Assay (Fujifilm Wako Chemicals). The ability of TMEM2-transfected cells to degrade substrate-bound HA was analyzed by in situ HA degradation assays, as described (11Yamamoto H. Tobisawa Y. Inubushi T. Irie F. Ohyama C. Yamaguchi Y. A mammalian homolog of the zebrafish transmembrane protein 2 (TMEM2) is the long-sought-after cell surface hyaluronidase.J. Biol. Chem. 2017; 292: 7304-7313Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 14Irie F. Tobisawa Y. Murao A. Yamamoto H. Ohyama C. Yamaguchi Y. The cell surface hyaluronidase TMEM2 regulates cell adhesion and migration via degradation of hyaluronan at focal adhesion sites.J. Biol. Chem. 2021; 296100481Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar). All data are contained within the article and supporting information. This article contains supporting information. The authors declare that they have no conflicts of interest with the contents of this article. We thank Dr William Stallcup for reading the manuscript. T. N., Y. T., A. B., and M. J. investigation; T. N. and F. I. data curation; M. J. and Y. Y. supervision; C. O. and Y. Y. funding acquisition; F. I. methodology; F. I. and Y. Y. writing–reviewing and editing; Y. Y. conceptualization; Y. Y. writing–original draft. This work was supported by the Mizutani Foundation Research Grant.
Referência(s)