https://orcid.org/0000-0002-1941-1622
Abstract
Heparan sulfate 6-endosulfatases (SULFs) remove 6-O-sulfo groups from heparan sulfate polysaccharide chains. SULFs modify the functions of heparan sulfate and contribute to the development of cancers, organ development and endothelial inflammatory responses. However, direct measurement of the activity of SULFs from human and mouse plasma is not currently possible. Here, we report a liquid chromatography coupled with tandem mass spectrometry (LS-MS/MS) assay to measure the activity of SULFs. The method uses a structurally homogeneous heparan sulfate dodecasaccharide (12-mer) in which the glucuronic and iduronic acid residues are labeled with both 13C- and 2H-atoms. The 12-mers desulfated by the SULFs is subjected to degradation with heparin lyases to yield disaccharides, which is followed by LC–MS/MS. The amount of two specific disaccharides, ΔIIIS and ΔIVS, quantified by LC–MS/MS reports the activity of the SULFs with high sensitivity and specificity. This method allows for the determination of the activity from conditioned cell media and mouse plasma. Our findings offer an essential novel tool to delineate many roles of SULFs in biological processes.
Introduction
Heparan sulfates (HS) are sulfated linear polysaccharides widely distributed on the surface of mammalian cells as well as secreted into the extracellular matrix (Bishop et al. 2007; Lindahl and Li 2009). HS is composed of repeating disaccharide units of glucuronic acid (GlcA) linked to glucosamine or iduronic acid linked to glucosamine. The glucosamine residues carry sulfation at the N-, 3-OH and 6-OH position, and iduronic acid (IdoA) residues carry sulfation at the 2-OH position. GlcA residues carry sulfation at the 2-OH position although less abundantly. HSs are involved in a variety of physiological and pathophysiological functions including blood coagulation, angiogenesis, development, inflammation, viral infection, Alzheimer’s disease and cancer due to their ability to interact with a wide range of proteins including growth factors, proteases, protease inhibitors, cytokines, chemokines, morphogens and receptors (Bishop et al. 2007; Lindahl and Li 2009; Sarrazin et al. 2011; Xu and Esko 2014; Liu et al. 2016). The precise sulfation patterns in HS determine their biological functions (Gama et al. 2006).
Heparan sulfate 6-endosulfatases (SULFs) catalyze the hydrolysis of the 6-O-sulfo group of an N-sulfo glucosamine 6-O-sulfate (GlcNS6S) or an N-acetylated glucosamine 6-O-sulfate (GlcNAc6S) to yield corresponding desulfated saccharide units although GlcNAc6S residue is less reactive to SULFs. The action of SULFs occurs on the cell surface or extracellularly after the completion of the biosynthesis of HS glycans. Thereby, SULFs are viewed as the critical postsynthetic enzyme that edits the sulfation sequence of HS, which modulates their functions. Two isoforms of SULF, SULF-1 and SULF-2, are present in the human genome. Findings from available substrate specificity studies indicate that there is no detectable difference in substrate preference between the two isoforms (Ai et al. 2006).
SULFs are a major factor connecting the HS network to cancer progression and to various other diseases (Rosen and Lemjabbar-Alaoui 2010; Bret et al. 2011; Hammond et al. 2014). Some studies suggested that SULF-1 has tumor suppressing activity (Lai et al. 2008a) but it was noted that SULF-1 is lost in tumor cell lines much more frequently than in tumor tissues, which is explained by our observation that SULF-1 is supplied to tumors by cancer associated fibroblasts. Overall, SULF-1 expression is elevated in gastric (Junnila et al. 2010; Hur et al. 2012; Lee et al. 2016), colon (Vicente et al. 2015; Wei et al. 2017), urothelial (Lee et al. 2017) or pancreatic (Yunxiao Lyu et al. 2018) cancers. SULF-2 expression increases in lung cancer (Lemjabbar-Alaoui et al. 2010), glioma (Phillips et al. 2012), hepatocellular carcinoma (Lai et al. 2008b), and pancreatic ductal adenocarcinoma (Nawroth et al. 2007). Multiple clinical studies have associated high SULF-1 (Junnila et al. 2010; Hur et al. 2012; Vicente et al. 2015; Lee et al. 2016; Lee et al. 2017; Wei et al. 2017) and SULF-2 (Lai et al. 2008b; Tessema et al. 2009; Bret et al. 2011; Lui et al. 2012; Alhasan et al. 2016; Yang et al. 2020) with poor survival. In addition, a recent study associated SULF-2 with rheumatoid arthritis (Siegel et al. 2022), and another demonstrated that SULFs contribute to bone marrow hematopoiesis (Whitehead et al. 2024). SULF-2 upregulation may play a role in preventing lung injury and triggering lung repair as demonstrated in a mouse model (Yue 2017). In a sepsis mouse model, SULF-1 contributes to the postseptic suppression of pulmonary inflammation (Oshima et al. 2019).
The enzymatic activity of SULFs is regulated by extensive post-translational modification, including furin-catalyzed proteolytic cleavage, N-glycosylation and chondroitin/dermatan sulfate modification (Masri et al. 2022; Seffouh et al. 2023). While N-glycosylation is required for catalytic activity, modification by chondroitin/dermatan sulfate decreases the activity of SULFs (Masri et al. 2022, Seffouh et al. 2023). The unique contribution of post-translational modifications to the activity of SULFs underscores the importance of measuring enzymatic activity independent of the determination of the expression of the mRNA as well as the level of protein by Western analysis.
A reliable method to measure the activity of SULFs with high sensitivity and selectivity is currently lacking. Although HS substrates have been used for measuring the activity of SULFs, structural heterogeneity of the polysaccharide substrate leads to high background, making it difficult to measure activity of relatively low levels of SULFs (Dhoot et al. 2001). A chromogenic substrate, 4-MUS (methylumbelliferyl sulfate), was later used as a substrate. However, the use of 4-MUS dose not display the needed selectivity as 4-MUS reacts with other types of sulfatases in addition to SULFs (Benicky et al. 2023) (Pempe et al. 2012). Structurally homogeneous synthetic oligosaccharide substrates were recently introduced as substrates for SULFs. Octasaccharides isolated from depolymerized heparin were reportedly labeled with two fluorescent tags at both reducing and nonreducing ends to be used as substrates to measure the activity of SULFs (Przybylski et al. 2019). In another study, a synthetic octasaccharide that was prepared chemoenzymatically was used to determine the activity of SULFs (Benicky et al. 2023). Our research team reported the use of an 8-mer oligosaccharide, GlcNS6S-GlcA-GlcNS6S-IdoA2S-GlcNS6S-IdoA2S-GlcNS6S-GlcA-pNP, as a substrate for SULFs (Benicky et al. 2023). The desulfated 8-mers were detected by a HPLC analysis method with a UV detector, provided that a p-nitrophenyl group is present at the reducing end of the 8-mer with the absorbance at 310 nM (Benicky et al. 2023). The use of the 8-mer significantly increased selectivity for measuring the activity of SULFs when compared to 4-MUS. However, the sensitivity of this assay format is insufficient to determine the activity of SULFs in biological samples.
In this manuscript, we demonstrate a new method for measuring the activity of SULFs. The method uses a synthetically made 13C/2H-labeled HS dodecasaccharide (12-mer) substrate coupled with LC–MS/MS quantification. Instead of measuring the intact desulfated 12-mer products, the products are degraded into disaccharides using heparin lyases and subsequently subjected to disaccharide analysis using the LC–MS/MS method. The method shows good linearity between the desired disaccharide products and the concentration of recombinant SULF-2 enzyme in biological buffer and in mouse plasma. Furthermore, we demonstrated the capability of measuring the activity of SULFs from mouse plasma. Our new assay provides a sensitive and quantitative method to measure the activity of SULFs. Success will advance investigations of the physiological roles of SULFs in various biological processes.
Results
A decrease in the level of 6-O-sulfated HS in head and neck squamous cell carcinoma (HNSCC) tumor is associated with the overexpression of SULFs
Previously, two lines of evidence suggested that expression of SULFs is dysregulated in HNSCC tumors. First, the levels of SULFs mRNAs are elevated in HNSCC tumor tissue (Yang et al. 2020). Second, an increase in the SULF-2 protein was detected by histological staining using 8G1 antibody, a mouse anti-human SULF-2 monoclonal IgG (Flowers et al. 2016). To further prove the impact on the structure of HS from the overexpression of SULFs, we conducted disaccharide analysis for HS isolated from HNSCC tumors and from the surrounding healthy tissues (Fig. 1C). We discovered a significant decrease in the trisulfated disaccharide, ΔIS (ΔUA2S-GlcNS6S), in HS from tumor tissue (Fig. 1A). In the disaccharide analysis, we introduced internal 13C-labeled disaccharide calibrants, which substantially reduces data variability caused by individual differences in patient samples, leading to a high degree of accuracy. It should be noted that ΔIS disaccharide is within the highly sulfated domain of HS, which is the primary action site for SULFs (Pempe et al. 2012). A decrease in the level of ΔIS is consistent with previously reported elevations in expression of SULFs (Qiu et al. 2018). Simultaneously, we observed an increase in ΔIVS (ΔUA-GlcNS) disaccharide from HNSCC (P = 0.04) (Fig. 1D) and a decrease in the total amount of HS from tumor tissue (P = 0.04) (Fig. 1B). This suggests a significant remodeling of the HS epitopes in HNSCC tissue with a significant decrease in the content of trisulfated disaccharide (ΔIS). Our findings suggest that the expression of SULFs enzymes is increased in the progression of HNSCC, resulting in the observed changes in the structures and quantities of HS in the tumors. Our subsequent efforts were focused on developing a method for the measurement of the activity of SULFs.
Analysis of the HS from HNSCC tumor and healthy tissue. Panel a shows the decrease in the level of ΔIS disaccharide in the HS from HNSCC tumor and healthy tissue (P = 0.04, n = 4). Panel B shows the total amount of HS from HNSCC tumor and healthy tissue (P = 0.04, n = 4). Panel C shows the steps involved in the analysis of HS. Panel D shows the percentages of seven additional disaccharides of HS from HNSCC tumor and healthy tissue. Paired t-test was used to compare the tumor and adjacent healthy tissues of patients (n = 4).
Development of a method to measure the activity of SULFs using HS oligosaccharides
We next sought to improve our assay by employing structurally defined oligosaccharide substrates. One limiting factor when developing an assay was the need to quantify the substrate and desulfated products with a degree of sensitivity high enough to detect low-level SULFs activity in tissues. Although we previously demonstrated the use of a p-nitrophenyl group labeled 8-mer substrate in vitro (Benicky et al. 2023), the detection of oligosaccharides merely based on UV310 nm measurement did not provide adequate sensitivity in tissue. We attempted to employ an LC–MS method in the analysis with the hypothesis that mass spectrometry detection offers higher sensitivity toward the oligosaccharide substrate and products. However, the sensitivity of LC–MS for the intact 8-mer substrate was at the level of 100 ng, which was only marginally improved when compared to UV310 nm detection as reported previously (Benicky et al. 2023) (Fig. 2A). We therefore employed an alternative strategy, involving an LC–MS/MS based disaccharide analysis method, which was introduced to improve sensitivity (Li et al. 2015; Wang et al. 2020). Four disaccharide products are expected from the digested 8-mers: ΔIIIS and ΔIVS are from desulfated 8-mers after the treatment with SULF-2; whereas ΔIS and ΔIIS disaccharides are from the 8-mer substrate (Fig. 2C). The amount of ΔIIIS and ΔIVS from the disaccharide analysis reflects the quantity of desulfated 8-mers, thereby, permitting us to determine the activity of SULFs. The detection sensitivity for ΔIIIS and ΔIVS would be substantially higher than for the 8-mers. We successfully detected ΔIIIS disaccharide at levels as low as 10 pg, 10,000-time more sensitive than the 8-mer (Fig. 2B). Notably, an AMAC (3-aminoacridin-9-(10H)-one) group was coupled to the disaccharides prior to the LC–MS/MS analysis. AMAC-labeled disaccharides tend to have higher ionization efficiency. In addition, multiple reaction monitoring (MRM) techniques were employed to further improve sensitivity. ΔIVS disaccharide also was detected with high sensitivity by the LC–MS/MS method (Supplementary Fig. S1).
Detection limits of 8-mer substrate and ΔIIIS disaccharide on LC–MS. panel a shows the ion chromatograms of 8-mer with different amounts from the LC–MS analysis. Below 100 ng (1 × 10−7 g), it became difficult to detect 8-mer. Panel B shows the ion chromatograms of ΔIIIS from the LC–MS/MS analysis. At the amount of 100 pg (1 × 10−10 g), both disaccharides were clearly detectable. The disaccharide was detectable even at 10 pg. the disaccharides were labeled with AMAC (3-aminoacridin-9-(10H)-one) prior to the MS analysis. Panel C shows the reactions involved in the use of disaccharide analysis for the analysis of the activity of SULF-2 using the 8-mer substrate. Arrows indicate the cleavage sites by heparin lyases for both 8-mer and desulfated 8-mer.
Using 8-mer substrate coupled with disaccharide analysis, we demonstrated the feasibility to determine the activity of SULF-2. The amount of both ΔIIIS and ΔIVS disaccharides released after the treatment with SULF-2 correlated linearly with the amount of recombinant SULF-2 protein (Supplementary Fig. S2). Although both ΔIIIS and ΔIVS disaccharides are released after the digestion of the desulfated 8-mers with heparin lyases (Supplementary Fig. S2), the amount of ΔIIIS is higher than ΔIVS, suggesting that SULF-2 displays greater reactivity towards the disaccharide structural domain of -IdoA2S-GlcNS6S- than towards -GlcA-GlcNS6S-. The desulfated -IdoA2S-GlcNS6S- product resulted in ΔIIIS disaccharide, whereas the desulfated -GlcA-GlcNS6S- product is ΔIVS (Fig. 2C). This observation is consistent with previous reports that SULF-2 displays the strongest reactivity towards the disaccharide domain of -IdoA2S-GlcNS6S- (Dhoot et al. 2001; Pempe et al. 2012).
Selection of oligosaccharide substrates for analyzing SULFs from biological samples
Superior detection sensitivity for disaccharides by the LC–MS/MS method prompted us to devise a disaccharide analysis-based method. Our primary concern was interference from endogenous HS polysaccharides in the biological samples because HS is present in plasma and mammalian tissues (Hippensteel et al. 2019). Endogenous HS releases ΔIIIS and ΔIVS disaccharides after digestion with heparin lyases, which are not necessarily related to the desulfated oligosaccharide products attributable to the action of SULF. To overcome the major limitation, we chose to use 13C/2H-labeled oligosaccharide substrates. The use of 13C/2H-labeled substrate not only eliminates interference from endogenous unlabeled HS but also maintains the capability for the LC–MS/MS quantitative analysis using 13C-labeled calibrants (Fig. 3). The analytical procedure includes three steps. First, we incubate a 13C/2H-labeled oligosaccharide substrate with SULFs to yield partially de-6-O-sulfated oligosaccharides (Fig. 3). Second, the partially desulfated oligosaccharide products are subjected to degradation by heparin lyases to yield disaccharides. Third, a LC–MS/MS based disaccharide analysis is performed to measure the levels of four disaccharides, including ΔIVS, ΔIIIS, ΔIIS and ΔIS, in the presence of 13C-labeled disaccharide calibrants. The analysis typically permits us to detect three isomeric disaccharides with different molecular weights. For example, the unlabeled ΔIIIS disaccharide has a molecular weight of 690.05, which is from endogenous HS present in the biological samples; 13C-labeled ΔIIIS disaccharide has a molecular wight of 696.05, which is from 13C-labeled disaccharide calibrant used as internal standard for quantitation; and 13C/2H-labeled ΔIIIS disaccharide has a molecular weight of 700.05, which is from the 13C/2H-labeled oligosaccharide substrate. Measuring the relative intensities of 13C/2H-labeled ΔIIIS and 13C-labeled ΔIIIS disaccharide calibrant allows us to quantify the disaccharides and to accurately determine the activity of SULFs in biological samples.
Measurement of the activity of SULFs using 13C/2H-labeled oligosaccharide substrates. Scheme shows the reaction involved in the analysis of the activity of SULFs from biological samples. Here, the use of 13C/2H-labeled substrates yields 13C/2H-labeled disaccharides, which can be distinguished from unlabeled disaccharides from the endogenous HS as well as 13C-labeled disaccharide calibrants used for the LC–MS/MS analysis.
Chemoenzymatic synthesis of isotopically labeled oligosaccharide substrates for SULFs
We synthesized two HS oligosaccharide substrates, including 13C/2H-labeled 8-mer and 13C/2H-labeled 12-mer, for this study. Both oligosaccharide substrates carried 13C/2H-double labeled GlcA or IdoA2S residues at specific sites in the oligosaccharides. The 13C/2H-labeled GlcA residues were introduced by employing UDP-[13C/2H]GlcA using Pasteurella Multocida heparosan synthase 2 (pmHS2) (Fig. 4) (Xu et al. 2017). The subsequent sulfation and epimerization steps were carried out using sulfotransferases and C5-epimerase as described previously (Xu et al. 2014). Purity of the oligosaccharide substrates was confirmed by high-resolution anion-exchange HPLC (Fig. 4) and structures of the oligosaccharides were confirmed by LC–MS (Supplementary Figs. S3 and S4). The 13C/2H labeled IdoA2S saccharide units in both 8-mer and 12-mer substrates carry six [13C]carbon atoms and five deuterium atoms (Supplementary Figs. S3 and S4). Notably, C5-position of IdoA2S is occupied by a proton, instead of deuterium, which was introduced during the synthesis of an IdoA2S unit by C5-epimerase from a GlcA unit (Step D in Fig. 4), where D-5 of GlcA unit was replaced by H-5 after the epimerization reaction in H2O-based buffers (Sheng et al. 2012).
Synthetic schemes for the synthesis of 13C/2H-labeled 8-mer and 13C/2H-labeled 12-mer substrates. Top panel shows the synthetic schemes for the synthesis of 8-mer and 12-mer substrates. Bottom panel shows the chromatograms of 8-mer and 12-mer from high-resolution anion exchange HPLC. The molecular masses of labeled 8-mer and 12-mer were determined by high-resolution mass spectrometry as demonstrated in supplementary figs. S4 and S5.
Comparison of the reactivity of 8-mer and 12-mer substrates
We compared the reactivities of two 13C/2H-labeled oligosaccharide substrates to SULF-2 desulfation reaction with the expectation that the longer substrate would display stronger reactivity to SULFs enzyme. A linear increase in release of ΔIIIS from 12-mer substrate in response to the amount of SULF-2 at the concentration range from 5 to 16 ng/mL was observed, however, no release of ΔIIIS was observed from the 8-mer substrate at the tested concentrations of SULF-2 (Fig. 5). Our results suggest that the 12-mer substrate is a more reactive substrate for the analysis of SULF-2 activity. In fact, the use of 12-mer substate improved sensitivity to SULF-2 activity by at least 4-fold when compared to the use of 8-mer substrate.
12-mer substrate offers better sensitivity than 8-mer substrate for measuring the activity of SULF-2. The linearity curves of using both 12-mer and 8-mer for SULF-2 activity measurement. An excellent linear relationship between the amount of ΔIIIS and the concentration of SULF-2 was observed. Although both 12-mer and 8-mer after treatment with SULF-2 can release ΔIIIS disaccharide, the increase of ΔIIIS from 8-mer was very small under such low concentration of SULF-2. The workflow for using 12-mer or 8-mer substrate to detect SULF-2 activity is shown on the right.
To demonstrate our method can be used to detect the activity of SULFs in a biological matrix, we spiked different amounts of recombinant SULF-2 in mouse plasma samples. As expected, we observed a linear response in the range of concentrations of SULF-2 from 1 to 13 ng/mL (Supplementary Fig. S5), using the 13C/2H labeled 12-mer as the substrate. Our data demonstrated the feasibility of measuring the activity of SULFs in plasma with high sensitivity.
Determination of the activity of SULFs from cell culture and from mouse plasma
We first tested SULF-2 in serum-free conditioned media of wild-type SCC-35 cells and media from SULF-2-deficient cells prepared by Crispr/Cas9 genome engineering. The expression of SULF-2 in wild-type (WT) cells and knockout (KO) cells was assessed by Western analysis (Fig. 6A). Consistent with the result from Western analysis, the activity of SULF-2 in the KO media decreased approximately 13-fold (Fig. 6A). These data validated the capability to measure SULF-2 activity from cell cultures. In the KO cells, we still observed a low level background of SULFs activity, which is likely attributable to the presence of low level SULF-1 expression (Benicky et al. 2023).
Determination of the activity of SULF-2 from conditioned media and SULFs in mouse plasma. Panel a shows the image of western analysis of SULF-2 from the condition media of SCC-35 cells (WT) and SCC-35 cells (KO) that the expression of SULF-2 was silenced using Crispr/Cas9 gene editing technique. The media was concentrated 50-fold prior to western analysis. The measurement for the activity of SULF-2 from condition medium is also shown. A standard curve was generated using purified SULF-2 enzyme, allowing us to determine the amount of SULF-2 based on the amount of released ΔIIIS, as quantified by our LC–MS/MS MRM assay. Panel B shows the activity measurement of SULFs from fresh mouse plasma. The flow chart for the analysis is shown on the left. Freshly collected blood (250 μL) was drawn directly into ethylenediaminetetraacetic acid (EDTA)-containing tubes. The 13C/2H-labeled 12-mer substrate (0.05 mg/mL in 0.9% NaCl saline) was added to the blood within the EDTA tubes and mixed by gentle inversion. The tubes were then centrifuged to isolate plasma for analysis. To the samples, CaCl2 (5 mM) was added to each aliquot and were incubated for different time points. The reaction mixtures were then subjected to disaccharide analysis. P value was determined by two-tailed unpaired t-test.
Next, we measured SULFs activity from mouse plasma as shown in Fig. 6B. In this experiment, we incubated the 13C/2H-labeled 12-mer substrate with mouse plasma for different periods, including 0 h, 1 h and 24 h. The release of ΔIIIS disaccharide was 1.85 ± 0.14 (n = 6) ng/mL of plasma from 0-h sample and 2.07 ± 0.19 (n = 7) ng/mL of plasma from 1-h sample. Although the difference in values from two groups was small, it was detectable and statistically significant (P = 0.0372). The release of ΔIIIS from 24-h sample reached 10.29 ± 3.27 ng/mL (n = 7) (P < 0.0001), substantially higher than those from 1-h and 0-h. Our data demonstrated that the release of ΔIIIS disaccharide was detected in a timely fashion, confirming that our method successfully detected SULFs activity. Despite the success in detecting SULFs activity from freshly collected mouse plasma, we were unable to detect SULFs activity from mouse serum. Due to the fact that both SULF-1 and SULF-2 hydrolyze the 12-mer substrate, we were unable to determine the sulfatase activity is attributed to SULF-1 or SULF-2.
Discussion
SULFs catalyze the removal of 6-O-sulfo group from HS glycans to modify sulfation patterns of HS after glycans are biosynthesized in the Golgi apparatus. The action of SULFs modulate the functions of HS in the extracellular space and on the cell surface. A sensitive and specific method for the measurement of the activity of SULFs from biological samples was lacking. Although we previously reported that the use of a synthetic HS 8-mer substrate increases the selectivity for SULFs, the sensitivity of the method was merely sufficient to measure the activity in high-expressing cell lines (Benicky et al. 2023). Here, we report a new LC–MS/MS based method that assays SULFs activity with robust sensitivity and selectivity. We demonstrate that we can measure the activity from conditioned culture media as well as recombinant SULF-2 present in control plasma. More importantly, our method demonstrates for the first time the ability to detect SULFs activity from mammalian plasma. Two isomers of SULF, SULF-1 and SULF-2, are present in humans. Due to the shared substrate preference of SULF-1 and SULF-2, our assay cannot distinguish between two isoenzymes. While most of our method development and validation were performed using recombinant SULF-2, the shared substrate specificity suggests that the assay measures the combined activity of SULF-1 and SULF-2.
The crucial innovation in our method is to use high purity 13C/2H-labeled 12-mer substrate and the implementation of disaccharide analysis using the LC–MS/MS method. Our method takes advantage of high sensitivity of the disaccharide analysis using LC–MS/MS, correlating the amount of two 13C/2H-labeled disaccharides, including ΔIIIS and ΔIVS, with SULFs activity. The use of 13C/2H-labeled 12-mer substrate eliminates the interference from endogenous HS during the disaccharide analysis. Furthermore, because 13C/2H-labeled ΔIIIS and ΔIVS disaccharides do not exist in biological samples, background signals are nearly zero in the control mouse plasma (Suppl Fig. S5). The low background signals contribute substantially to the increase in the sensitivity of our method.
We have not yet employed our method for the analysis of solid tissue samples due to a concern for stability of SULFs activity when preparing such samples. Furthermore, while the activity of SULFs can be detected in plasma, the unexpected failure to detect activity from serum potentially offers novel insight into SULFs activity when blood clotting occurs. While serum and plasma are both cell-free blood preparations, serum is isolated from whole blood after it is allowed to clot whereas plasma is isolated after being placed in EDTA containing tubes to inhibit clotting. The failure to detect SULFs activity in serum using this novel assay may be caused by blood clots acting as SULFs sink. Further investigation is needed to understand this unexpected observation in serum and to adapt and optimize this method to quantify activity from other tissues. Given that SULFs are known to function extracellularly, it is presumed that the active form of SULFs is present and should be detectable in plasma and potentially other biological fluids. The capability of measuring SULFs activity from plasma is clearly a major step forward for studying the biology of SULFs.
The synthesis of 13C/2H-labeled 12-mer with high purity is achieved using a chemoenzymatic method incorporating the use of UDP-[13C/2H]GlcA. Both the synthesis of the 12-mer and the sugar donor, UDP-[13C/2H]GlcA, are completed cost effectively. In the current study, we successfully synthesized 5 mg of 12-mer, which is adequate for the analysis of ten thousand samples. Once the method is accepted by the research community, we anticipate increasing the synthesis of the 12-mer by 10- to 100-fold without any difficulties. Consequently, the availability of the substrate will expand our ability to analyze samples from experimental animals and patient samples. This new tool will undoubtedly accelerate the investigation into this important family of enzymes.
CRediT author statement
Zhangjie Wang (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Methodology [equal]), Julius Benicky (Methodology [equal]), Pritha Mukherjee (Data curation [equal], Methodology [equal]), Justin Laing (Data curation [equal], Methodology [equal]), Yongmei Xu (Data curation [equal], Formal analysis [equal], Methodology [equal]), Vijayakanth Pagadala (Data curation [equal], Investigation [equal]), Shuangni Wu (Data curation [equal], Formal analysis [equal]), Joseph A Hippensteel (Data curation [equal], Funding acquisition [equal], Investigation [equal], Supervision [equal]), Jian Liu (Conceptualization [equal], Resources [lead], Supervision [lead]).
Author contributions
ZW designed the project, synthesized the oligosaccharide substrates and carried out the LC–MS/MS analysis. JB, PM and RG provided recombinant SULF-2 enzymes and condition media expressing SULF-2. J Laing and JAH completed the activity of measurement of SULF in mouse plasma. YX and VP prepared the intermediates for the synthesis of oligosaccharides. SW participated in the disaccharide analysis for measuring the activity of SULF-2. ZW and Jian L wrote the manuscript. RG and JAH also participated in writing the manuscript. All authors reviewed the manuscript.
Funding
This work was supported in part by National Institutes of Health (R42GM140693, R44GM142304, K08HL159353, and R03AG074056).
Conflict of interest statement
ZW, VP and SW are employees of Glycan Therapeutics; Jian L and YX are founders for Glycan Therapeutics. ZW, VP, Jian L and YX have equity of the company. Other authors declare no competing interest.
Data availability
The data underlying this article are available in the article and in the online Supplementary Information.
