PNGase H + variant from Rudaea cellulosilytica with improved deglycosylation efficiency for rapid analysis of eukaryotic N‐glycans and hydrogen deuterium exchange mass spectrometry analysis of glycoproteins

Abstract

The analysis of glycoproteins and the comparison of protein N‐glycosylation from different eukaryotic origins require unbiased and robust analytical workflows. The structural and functional analysis of vertebrate protein N‐glycosylation currently depends extensively on bacterial peptide‐N4‐(N‐acetyl‐β‐glucosaminyl) asparagine amidases (PNGases), which are indispensable enzymatic tools in releasing asparagine‐linked oligosaccharides (N‐glycans) from glycoproteins. So far, only limited PNGase candidates are available for N‐glycans analysis, and particularly the analysis of plant and invertebrate N‐glycans is hampered by the lack of suitable PNGases. Furthermore, liquid chromatography–mass spectrometry (LC–MS) workflows, such as hydrogen deuterium exchange mass spectrometry (HDX‐MS), require a highly efficient enzymatic release of N‐glycans at low pH values to facilitate the comprehensive structural analysis of glycoproteins. Herein, we describe a previously unstudied superacidic bacterial N‐glycanase (PNGase H⁺) originating from the soil bacterium Rudaea cellulosilytica (Rc), which has significantly improved enzymatic properties compared to previously described PNGase H⁺ variants. Active and soluble recombinant PNGase Rc was expressed at a higher protein level (3.8‐fold) and with higher specific activity (~56% increase) compared to the currently used PNGase H⁺ variant from Dyella japonicum (Dj). Recombinant PNGase Rc was able to deglycosylate the glycoproteins horseradish peroxidase and bovine lactoferrin significantly faster than PNGase Dj (10 min vs. 6 h). The versatility of PNGase Rc was demonstrated by releasing N‐glycans from a diverse array of samples such as peach fruit, king trumpet mushroom, mouse serum, and the soil nematode Caenorhabditis elegans . The presence of only two disulfide bonds shown in the AlphaFold protein model (so far all other superacidic PNGases possess more disulfide bonds) could be corroborated by intact mass‐ and peptide mapping analysis and provides a possible explanation for the improved recombinant expression yield of PNGase Rc.

1. INTRODUCTION

N‐Glycosylation is a post‐translational protein modification that influences a myriad of cellular functions. The comparative analysis of protein N‐glycosylation from various sources has become increasingly important and has been reported to be a critical factor in the assessment of proteins in biotechnology, ¹ , ² , ³ food safety, ⁴ , ⁵ , ⁶ and structure/function relationship studies. ⁷ , ⁸ , ⁹ , ¹⁰ Generally, the analysis of N‐glycans requires the enzymatic release of the carbohydrate portion from glycoproteins using PNGases (Figure S1 [supporting information]). The release of vertebrate N‐glycans can be achieved by using recombinant PNGase F, which originates from the bacterium Elizabethkingia meningoseptica, ¹¹ and is currently widely applied for the analysis of N‐glycans from milk‐, egg‐, and blood serum glycoproteins. ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁷ , ¹⁸ PNGase F can release all types of vertebrate N‐glycans (high mannose‐, hybrid‐, and complex‐type N‐glycans), and pre‐denaturation of the glycoproteins (i.e., by sodium dodecyl sulfate [SDS] or urea) from the source material greatly enhances the efficiency of the N‐glycan release. ¹⁹ , ²⁰ However, PNGase F is not able to release N‐glycans which have an α1,3‐linked fucose on the innermost core GlcNAc (“core α1,3‐fucose”), and therefore this enzyme is insufficient for releasing N‐glycans from invertebrate and plant sources bearing this peculiar glycan modification. ²¹ , ²² A further limitation of PNGase F is the absence of activity in acidic environments as some liquid chromatography–mass spectrometry (LC–MS) workflows, such as hydrogen‐deuterium exchange mass spectrometry (HDX‐MS), require efficient deglycosylation at low pH values to enable comprehensive structural analysis of glycoproteins. ²³

In contrast to PNGase F, a structurally unrelated enzyme isolated from almond seeds (PNGase A) allows the release of N‐glycans from a broad variety of plant and invertebrate glycoproteins. ²⁴ , ²⁵ The optimum pH for PNGase A is 5.0, and it shows some residual enzymatic activity even at lower pH values. We previously reported the use of PNGase A to study the conformation and interactions of glycoproteins using HDX‐MS. ²⁶ Although PNGase A releases a broad variety of N‐glycans from vertebrate, invertebrate, and plant sources, efficient N‐glycan release is achieved only from glycopeptides but not from native glycoproteins, which makes a proteolytic pretreatment of the source materials necessary. ²⁷ , ²⁸ Furthermore, PNGase A itself is a glycoprotein and can be self‐deglycosylated, and thereby the contamination of glycan samples by endogenous PNGase A glycan structures must be anticipated. ²⁹ So far, approximately 10 PNGase A homologues from plants and fungi have been described yet. ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ However, none of these variants could be successfully expressed in recombinant form in prokaryotic expression systems, thereby limiting their use in glyco‐analytical applications.

In recent years, we have worked on the discovery, functional improvement, and applications of a novel class of superacidic bacterial PNGases. ³⁶ , ³⁷ , ³⁸ , ³⁹ Although these enzyme variants could satisfy basic requirements for the release of N‐glycans, the recombinant expression levels and overall enzymatic activities could be improved to simplify application within glycoanalytics and LC–MS analysis of glycoproteins. The first reported recombinant variant (originating from Terriglobus roseus, in short, PNGase H⁺ or Tr) was sufficiently active to determine the overall substrate scope, and the basic biochemical parameters such as pH and temperature optima. Its use in glycoproteomic applications such as HDX‐MS suffered from low expression yields and purity in comparison to the PNGase variant isolated from almond seeds (PNGase A). ²⁶ , ⁴⁰ One advantage of PNGase Tr relative to PNGase A, however, was a high tolerance to strong reducing agents (e.g., TCEP). Such reducing agents are added during sample preparation during HDX‐MS analysis of disulfide‐bonded proteins. ⁴⁰ In a collaborative effort, we then reported the evaluation of 12 more putative PNGase H⁺ variants, of which 3 variants showed significantly higher deglycosylation activities compared to PNGase Tr. ³⁹ , ⁴⁰ The most active variant originating from Dyella japonica (in short, PNGase Dj) showed a ~ fourfold increase in deglycosylation activity of horseradish peroxidase (HRP) when compared to PNGase Tr, and outperformed PNGase A in deglycosylating Trastuzumab glycopeptides. ³⁶ Furthermore, the higher yields and homogeneity after protein purification allowed the immobilization of PNGase Dj on microfluidic chips for studying the glycosylated sema‐domain of the tyrosine‐protein kinase MET, ³⁶ or the use as an in‐situ deglycosylation agent for matrix‐assisted laser desorption‐mass spectrometry (MALDI)‐imaging of N‐glycans in soybean root nodules. ⁴¹

Herein, we describe and characterize a previously unstudied recombinant acidic PNGase variant from Rudaea cellulosilytica with improved enzymatic properties and expression yields compared to previously described superacidic PNGases. Our results indicate that PNGase Rc should be a highly suitable enzyme for use in LC/LC–MS workflows for the analysis of released N‐linked glycans from a broad array of complex glycoprotein samples as well as for the HDX‐MS technique that requires highly efficient deglycosylation at low pH to allow comprehensive analysis of glycoproteins.

2. METHODS

2.1. Materials

HRP was obtained from Duly Biotech Company (Nanjing, China). Bovine lactoferrin was purchased from Wako Pure Chemical Industries (Nanjing, China). Peaches (Prunus persica L. Batsch cvs Xia Hui 6) were obtained from the experimental garden facility of the Jiangsu Academy of Agricultural Sciences (Nanjing, China). King trumpet mushrooms (Pleurotus eryngii) were bought at a local supermarket. C57BL/6 mouse plasma was purchased from Shanghai Fantai Biotechnology Company (Shanghai, China). Caenorhabditis elegans var Bristol strain N2 was kindly provided by Dr. Di Chen from the Model Animal Research Center of Nanjing University. Ni‐NTA Sefinose Resin was purchased from BBI Life Sciences (Shanghai, China). All other standard chemicals and buffer reagents were of the highest grade available.

2.2. Plasmid construction, expression, and purification

The candidate gene Rc (from R. cellulosilytica DSM 22992, GenBank ID WP_169337280) with Kpn I and Xho I restriction sites was synthesized and ligated into the pET30a vector by Genscript (Nanjing, China). The constructed expression vector was transformed into Escherichia coli BL21 (DE3) competent cells and plated on LB agar supplemented with kanamycin. A single colony was transferred into a 2 L Erlenmeyer shaking flask containing 400 mL of lysogeny broth (LB) medium and shaken at 37°C until the culture density reached an absorbance of 0.5 at 600 nm. The final concentration of 1mM isopropyl 1‐thio‐β‐d‐galactopyranoside (IPTG) was used to induce the expression of recombinant Rc at a reduced temperature of 18°C. After an induction period of 20 h, cells were collected by centrifuging at 12 000g for 20 min. The resulting cell pellet was resuspended in 10 mL of lysis buffer (100mM, NaCl, 50mM Tris, 1% [v/v] Triton X‐100, 1mM PMSF, adjusted to pH 8.0) and sonicated for 30 min. The cell lysates were centrifuged for 20 min at 12 000g, and then the supernatant was applied onto a Ni‐NTA column (10 mL of bed volume). About 200 mL of washing buffer (consisting of 50mM NaCl, and 50mM Tris, adjusted to pH 8.0 with HCl) was used to wash off unspecifically retained proteins. A simple one‐step purification with 20 mL of elution buffer containing 500mM imidazole was directly used to elute target proteins. Samples were further analyzed using SDS‐PAGE, and the fractions including candidate protein were stored at 4°C for further experiments.

2.3. Site‐directed mutagenesis of PNGase Rc

PNGase Rc site‐directed mutagenesis was generated according to the QuickChange XL site‐directed mutagenesis protocol (Stratagene) using the former primer 5'‐ACCGGTACTCGTCAGTACGCCCGTACCGCTCGTATCTG‐3'and the reverse primer 5'‐CAGATACGAGCGGTACGGGCGTACTGACGAGTACCGGT‐3′. The mutated plasmid was verified using DNA sequencing and transformed into E. coli BL21 (DE3) competent cells for recombinant expression. The expression and purification was performed under the conditions mentioned earlier (Section 2.2). The activity test of PNGase Rc mutant was carried out using gel‐based deglycosylation assay. ³⁹

2.4. Substrate specificity of PNGase Rc

HRP and bovine lactoferrin were selected to evaluate substrate specificity of recombinant PNGase Rc. About 10 μL of each glycoprotein (20 μg/μL) was incubated with 190 μL of PNGase Rc reaction mixture (consisting of 10 μg of PNGase Rc and 300mM of citrate/sodium phosphate buffer [pH 2.0] in water). After the reaction mixtures were incubated for 1 h at 37°C, the samples were centrifuged for 5 min at 12 000g, and the supernatant was purified with solid‐phase extraction (Supelclean ENVI‐Carb SPE Tube, 3 mL, 250 mg, No 57088). The purified N‐glycans were analyzed using matrix‐assisted laser desorption/ionization time‐of‐flight mass spectrometry (MALDI‐TOF‐MS) in a positive ion mode using 2, 5‐dihydroxybenzoic acid (10 mg/mL in acetonitrile) as a matrix. The acquired data were processed using the Bruker Flex Analysis software (version 3.3).

2.5. Peptide mapping

Peptic digestion of reduced PNGase was performed by first diluting 5 μL of 200 pmol PNGase Rc 1:1 with 300mM phosphate buffer (pH 2.3), 6M guanidinium hydrochloride (GdnHCl), and 0.5M TCEP. The solution was incubated at 37°C for 30 min. About 90 μL of solvent A (0.23% formic acid) was added prior to injection. For non‐reduced samples, the 0.5M TCEP was not added. Digestion was performed on a home‐packed pepsin column (internal volume of 60 μL) coupled to the LC–MS system upstream of the C18 trap column (see below). For tryptic digestion of reduced Rc, 5 μL of 200 pmol of PNGase Rc was diluted in 10 μL of 6M GdnHCl, 50mM NH₄HCO₃ (pH 8.00), and the resulting mixture was incubated at 60°C for 30 min. The sample was reduced by adding 5 μL of 45mM dithiothreitol (DTT) and incubated at 60°C for 3 h. Reduced cysteines were alkylated by adding 5 μL of 100mM iodoacetamide (IAM) and incubated at room temperature for 30 min. About 7.7 μL of 650mM CaCl₂ was added to assist tryptic stability. The sample was diluted to 100 μL with 50mM NH₄HCO₃ (pH 8.00), and 5μM of trypsin was added in a w/w relationship of 1:20 and incubated overnight at 37°C. For non‐reduced samples, DTT was replaced with MilliQ water.

LC–MS was performed by loading samples onto a UPLC‐MS system where desalting was performed via a C18 vanguard trap column with a 200 μL/min flow rate of solvent for 3 min. After desalting, peptides were separated on a C18 analytical column and were eluted off column through a gradient of solvent B (5% to 95% over 9 min). MS and MS/MS analysis was performed on a hybrid ESI‐Q‐TOF mass spectrometer (Synapt G2‐Si Milford, MA, USA) utilizing the fast DDA function and assessed via ProteinLynx Global server. The disulfide linked peptide was analyzed through a separate MS/MS experiment where its precursor ion's M + 6H⁶⁺ was selected for fragmentation through a ramping of the collision voltage (20–40 eV) over its given retention time.

2.6. Intact mass

Reduced and unreduced samples underwent the same sample preparation utilized in the pepsin digest peptide mapping experiments, bar on column digestion. 15 pmol of PNGase Rc in solvent A (0.23% formic acid) was injected onto a UPLC system where desalting was performed via a C4 vanguard trap column with a 200 μL/min flow rate of solvent for 3 min. The intact protein was then eluted off column through a gradient of solvent B (5% to 95% over 6 min), and mass analysis was performed on a hybrid ESI‐Q‐TOF mass spectrometer (Synapt G2‐Si). Data analysis was performed using Maxent1 function of the Masslynx software.

3. RESULTS AND DISCUSSION

Prior to this work, the PNGase H⁺ variant reported to have the highest recombinant expression levels in E. coli was PNGase Dj. By comparing AlphaFold ⁴² ‐derived protein structure models, we noticed that PNGase Dj had fewer cysteine residues compared with PNGase Tr (six versus eight cysteine residues), of which six were predicted to form three intramolecular disulfide bonds (Figure 1). Given that cytosolic expression in E. coli is not optimal for the expression of proteins containing disulfide bonds, ⁴³ we speculated that a lower number of cysteine residues present in the PNGase H⁺ protein sequence may be beneficial for the soluble expression of functional PNGases.

FIGURE 1.

Comparison of cysteine abundance in PNGase variants. Cysteines are shown in yellow (participating in disulfide bond formation) and green (not participating in disulfide bond formation). The protein models were generated and visualized using the AlphaFold toolset of the UCSF ChimeraX software suite (www.rbvi.ucsf.edu/chimerax) with the respective PNGase amino acid sequences obtained from GenBank (accession codes for Dj: WP_019464163.1, Tr: WP_014787206.1, Rc: WP_169337280.1, and A [from almonds]: XP_034202935) [Color figure can be viewed at wileyonlinelibrary.com]

A BLAST search of putative superacidic PNGases from bacterial genome databases identified a variant from R. cellulosilytica, which possessed only four cysteine residues in its protein sequence (Table S1 [supporting information]), and therefore it was chosen as a candidate for initial expression experiments and activity studies. Deglycosylation experiments showed that PNGase Rc was able to fully deglycosylate the glycoprotein substrates HRP and bovine lactoferrin within 10 min of incubation time (Figure 2). Notably, a variant of Rc where the aspartic acid residue at position 106 was substituted with alanine (RcD106A) was inactive (Figure 2), indicating the importance of this residue for the catalytic activity of Rc.

FIGURE 2.

PNGase Rc catalyzed N‐glycan release from horseradish peroxidase (HRP) and bovine lactoferrin. Time course analysis of the deglycosylation reaction of HRP using (A), an SDS‐PAGE shift assay and (B) and (C), ultra‐performance liquid chromatography (UPLC)‐ and matrix‐assisted laser desorption/ionization time‐of‐flight mass spectrometry (MALDI‐ToF‐MS)‐based analysis of the released N‐glycans. Panels (D), (E), and (F) used bovine lactoferrin as glycoprotein substrate. The red arrow marks the protein band of PNGase Rc in the assay mixtures. The substitution of the aspartic acid residue at position 106 with alanine (RcD106A) rendered this mutant variant inactive. N‐Glycans are depicted in the symbol nomenclature for glycans (SNFG) [Color figure can be viewed at wileyonlinelibrary.com]

Deglycosylation of PNGase Rc was significantly faster than the currently used PNGase Tr and Dj variants, which deglycosylated lactoferrin in an identical assay within 6 h and “overnight,” respectively. ³⁹ , ⁴⁴ A comparison of the overall amount of purified recombinant proteins showed that PNGase Rc yielded 3.8 times more protein than PNGase Dj (5.3 ± 0.5 mg of purified PNGase Rc and 1.4 ± 0.2 mg of PNGase Dj from 400 mL of expression culture, respectively, Figure S2 [supporting information]). In addition, PNGase Rc also showed a ~ 56% increased specific activity compared to PNGase Dj using a colorimetric HRP deglycosylation assay ⁴⁵ (122 ± 6 U/mg and 78.1 ± 3.7 U/mg, respectively, Figure S3 [supporting information]). The pH optimum of PNGase Rc was comparable to previously described superacidic PNGases, with the highest activities observed at pH values 2.0–2.5 and moderate activities recorded at pH values 3.0–3.5; however, only negligible deglycosylation activities were confirmed at higher pH values (Figure S4 [supporting information]).

We next characterized the primary structure of PNGase Rc using LC–MS (Figure 3). Mass analysis of the native intact (unreduced) PNGase Rc yielded a mass of 64199.0 Da. This experimental mass is in good agreement (within a mass accuracy of 17 ppm) with an expected theoretical mass of 64197.9 Da based on the cloned sequence and with the assumption that all four C residues of Rc participate in disulfide bonds. In support of this finding, the AlphaFold model of PNGase Rc indeed indicated two disulfide linkages in Rc, between C58‐C84 and C554–574 (Figure 1). To confirm the presence of the two disulfide bonds in Rc, we also performed mass analysis of intact Rc by treating it with the reducing agent TCEP, which yielded a mass of 64202.9 Da corresponding to the expected mass shift due to reduction of four cysteine residues (Figure S5 [supporting information]).

FIGURE 3.

Characterization of the primary structure of PNGase Rc. (A), Deconvoluted mass spectrum showing an experimental determined mass for the unreduced form of 64199.0 Da (theoretical mass 64197.9 Da, 17 ppm mass accuracy. (B), Deconvoluted mass spectrum showing an experimental determined mass for the reduced form was 64202.9 Da (theoretical mass 64202.0 Da, 14 ppm mass accuracy). (C), Map of identified peptides of reduced PNGase Rc after pepsin digestion and LC‐MS/MS analysis. Cysteines have been highlighted in red [Color figure can be viewed at wileyonlinelibrary.com]

The primary structure of Rc was further verified by proteolysis of unreduced and reduced Rc using either pepsin or trypsin and using LC–MS/MS to identify and map the resulting peptides to the expected sequence of Rc (Figures S6–S8 [supporting information]). The peptide mapping analysis using pepsin of reduced and unreduced Rc yielded peptides confirming 81% and 75% of the sequence, respectively, including the N‐ and C‐termini. The corresponding peptide maps using trypsin provided sequence coverages of 43% and 47%, respectively, including covering the N‐ and C‐termini. Thus, the data from intact mass analysis and peptide mapping collectively confirmed the sequence of Rc. In addition, peptide mapping of unreduced Rc using trypsin identified a disulfide‐linked peptide consisting of tryptic peptides 526–556 (α) and 564–589 (β) (Figure 4). The summed overall sequence coverage from all the approaches was 90.4%.

FIGURE 4.

Mapping of the disulfide bonds in PNGase Rc. (A), MS spectrum of a disulfide‐linked peptide, consisting of peptide 526‐556 (α) and peptide 564‐589 (β), resulting from a tryptic digest of unreduced PNGase Rc. (B), MS/MS spectrum of the identified disulphide‐linked peptide following targeted MS/MS using CID of the [M+6H]6+ precursor ion (m/z, 997.973, theoretical m/z, 997.963, 10 ppm mass accuracy). [Color figure can be viewed at wileyonlinelibrary.com]

Importantly, this peptide was absent in the corresponding reduced sample. Our experimental data thus directly confirmed a C554–C574 disulfide bond in Rc—and by inference thus also a disulfide bond between the two remaining C residues (C58‐C84) (Figure S9 [supporting information]).

To evaluate the deglycosylation activity of PNGase Rc, a variety of sample materials were subjected to an enzymatic N‐glycan release. UPLC analysis revealed that the N‐glycan profiles from king trumpet mushroom and mouse serum samples treated with PNGase Rc were very similar to the N‐glycan profiles obtained using PNGase F (Figures 5A and 5B), and showed the expected signature pattern for the N‐glycosylation of fungi (mostly high‐mannose type ⁴⁶ , ⁴⁷ ) and mammals (mostly complex type N‐glycans ⁴⁸ ). As anticipated, the N‐glycan profiles obtained from the peach fruit and C. elegans samples showed a significantly higher N‐glycan variety for PNGase Rc–treated samples (Figures 5C and 5D), which is caused by the inability of PNGase F to release N‐glycans bearing core α1,3‐fucose moieties, which are present in both plant N‐glycans and invertebrate N‐glycans. ²⁹ , ⁴⁹

FIGURE 5.

Ultra‐performance liquid chromatography (UPLC) analysis of N‐glycans released from (A), king trumpet mushroom; (B), mouse serum; (C), peach fruit; and (D), C. elegans released by PNGase Rc and F. The red triangles denote additional N‐glycan structures released by PNGase Rc [Color figure can be viewed at wileyonlinelibrary.com]

To further evaluate the effect of the acidic reaction conditions on acid‐labile carbohydrate moieties such as terminally linked sialic acid residues, sialylated glycoproteins such as human immunoglobulin G (IgG) or bovine fetuin were also incubated in a short time deglycosylation experiment (1 h enzymatic N‐glycan release) using PNGase Rc. UPLC and MALDI‐TOF analysis showed that the release efficiency of PNGase Rc is comparable to PNGase F, and that no loss of sialic acid residues occurred during the enzymatic release (Figure S10 [supporting information]).

4. CONCLUSIONS

Here we express and characterize a previously unstudied superacidic bacterial N‐glycanase (of the PNGase H⁺ family) that originates from the soil bacterium R. cellulosilytica (Rc). We show that the enzyme has significantly improved enzymatic properties for use in workflows for glycan and glycoprotein analysis, compared to previously described PNGase H⁺ variants. PNGase Rc can be expressed recombinantly and purified at significantly higher yield and concentrations, and also possesses a higher specific activity levels than the previously studied superacidic PNGases. We anticipate that the PNGase Rc will be of particular benefit to research within the fields of glycomics, glycoproteomics, and HDX‐MS. We note that PNGase Rc could, for instance, be readily used instead of PNGase A or PNGase Dj in our previously reported HDX‐MS workflows for analysis of glycoproteins. ³⁰ , ⁴⁰

PEER REVIEW

The peer review history for this article is available at https://publons.com/publon/10.1002/rcm.9376.

Supporting information

FIGURE S1 PNGases catalyze the release of N‐glycans from glycoproteins. The carbohydrate portion of the glycoprotein is depicted in the Symbol Nomenclature for Glycans¹⁵ (SNFG, in blue and green) and the ball‐and‐stick molecule style (in pink)

FIGURE S2 (A) Comparison of protein concentrations of purified PNGase Dj and Rc obtained from 400 mL fermentation culture. (B) The concentrations were measured using the Bradford method. Bovine serum albumin solutions at concentrations between 0 μg/mL and 30 μg/mL were used for protein quantification

FIGURE S3 Comparison of the specific activity of PNGase Dj and Rc. Typically, 20 μL of an activity test mixture contained 5 μL of 400 μM HRP, 2 μL of sodium phosphate citric acid buffer (500mM, pH 2.0), 11 μL of deionized water, and 2 μL of diluted enzyme solution. Samples were incubated at 37°C for 1.0 h. An equal volume of 2.5M TCA was added to stop the reaction and precipitate proteins. After centrifugation at 12 000g for 30 min, 10 μL of the supernatants containing free N‐glycans were transferred to a 384‐well microplate, then 5 μL of 4M NaOH and 10 μL of 1.7mM WST‐1 solution were added into the wells step by step. The mixture was incubated at 50°C for 1 h. Absorbance at 584 nm was measured using a microplate reader. The absorbance that the samples were incubated for 0 h was taken as blank control. The unit of enzyme activity (U) was defined as the amount of enzyme required to release 1 nmol N‐glycans from HRP at 37°C and pH 2.0 in 1 min

FIGURE S4 Determination of the pH optima of PNGase Rc (A) using a gel‐based deglycosylation assay to observe the molecular weight shift of the HRP glycoprotein band upon treatment with the PNGase enzymes. (B) Control experiment using the inactive PNGase mutant variant RcD106A for the deglycosylation experiment. The red arrow shows the protein band of the PNGase Rc variants in the assay mixtures. The pH optimum was tested in 20 μL assay mixtures consisting of 5 μL of 1 μg/μL purified PNGase Rc solution, 1 μL of 10 μg/μL HRP, 6 μL of sodium phosphate–citrate buffer (with different pH values; the final concentration of the buffer was 500mM, final pH was 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, respectively) and 8 μL of deionized water. The inactive mutant PNGase RcD106A was used as a negative control. All samples were incubated at 37°C for 1 h, and then a sodium hydroxide aqueous solution was added to the mixture for adjusting the pH to a neutral value. Sample aliquots were further incubated with 5 × Laemmli buffer at 95°C for 10 min, separated by SDS/PAGE and visualized by Coomassie Brilliant Blue R‐250 staining. (C) Wild‐type PNGase Rc and RcD106A (aspartic acid residue at position 106 was substituted with alanine) amino acid sequence alignment

FIGURE S5 (A) Deconvoluted mass spectrum showing an experimental determined mass for the unreduced form of 64 199.0 Da (theoretical mass 64 197.9, 17 ppm mass accuracy). (B) Deconvoluted mass spectrum showing an experimental determined mass for the reduced form was 64 202.9 Da (theoretical mass 64 202.0, 14 ppm mass accuracy)

FIGURE S6 Map of identified peptides of unreduced PNGase Rc following trypsin digestion and

LC–MS/MS analysis. Cysteines have been highlighted in red

FIGURE S7 Map of identified peptides of reduced PNGase Rc following trypsin digestion and

LC–MS/MS analysis. Cysteines have been highlighted in red

FIGURE S8 Map of identified peptides of unreduced PNGase Rc following pepsin digestion and LC–MS/MS analysis. Cysteines have been highlighted in red

FIGURE S9 Isolation MS/MS of the disulfide linked peptide (tryptic DDA data set) (A,B) chromatograms and (C) MS/MS fragmentation spectra of the disulfide linked peptide. MS/MS fragmentation setup: acquisition time, 60.0 s; trap collision energy ramp start, 20.0 eV; trap collision energy ramp end, 40.0 eV. It is located toward PNGase Rc's C termini MH⁶⁺ precursor ion (997.963 m/z) obtained by isolated MS/MS. Peptide 526–556 is indicated by the α, and peptide 564–589 is indicated by β

FIGURE S10 Comparison of N‐glycan release from human IgG and bovine fetuin using PNGase F and PNGase Rc. The samples were prepared as follows: 10 μL of human IgG or bovine fetuin glycoprotein (concentration ~10 μg/μL) was mixed with either 190 μL of PNGase F reaction mixture containing ~6 U of PNGase F, 50 μL of 6M urea, 25.5 μL of deionized water, 23 μL of 500mM sodium phosphate solution (pH 7.5), 12.5 μL of sodium dodecyl sulfate solution (2% w/v sodium dodecyl sulfate in 1M β‐mercaptoethanol), 19 μL of Triton‐X100 (10% w/v) or 190 μL of PNGase Rc reaction mixture containing ~6 U of PNGase Rc and 60 μL of sodium phosphate–citrate buffer (1M, pH 2.0). The N‐glycan release was carried out at 37°C for 1 h, and subsequently, the released glycans were purified with SPE columns, labeled with 2‐aminobenzamide and detected using ultra‐performance liquid chromatography (UPLC) (see method “Enzymatic Release: An Analysis of N‐Glycans”). About 1 μL of a 1:10 dilution was used for MALDI‐TOF MS analysis in a positive ion mode using 2, 5‐dihydroxybenzoic acid (10 mg/mL in acetonitrile) as matrix. The acquired data were processed using the Bruker Flex Analysis software (version 3.3), and the mass peaks derived from the mass spectra were evaluated using GlycoWorkbench (version 1.1)

Guo R‐R, Zhang T‐C, Lambert TOT, et al. PNGase H + variant from Rudaea cellulosilytica with improved deglycosylation efficiency for rapid analysis of eukaryotic N‐glycans and hydrogen deuterium exchange mass spectrometry analysis of glycoproteins. Rapid Commun Mass Spectrom. 2022;36(21):e9376. doi: 10.1002/rcm.9376

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.