Skip to main content
SearchLoginLogin or Signup

Do I Need to Trypsin Digest Before Releasing IgG Glycans With PNGase-F?

Keywords: rate kinetics, deglycosylation, liquid chromatography-mass spectrometry, immunoglobulin G

Published onMar 13, 2023
Do I Need to Trypsin Digest Before Releasing IgG Glycans With PNGase-F?
·
history

You're viewing an older Release (#1) of this Pub.

  • This Release (#1) was created on Mar 13, 2023 ()
  • The latest Release (#2) was created on Mar 29, 2023 ().

ABSTRACT

Immunoglobulin G (IgG) is the main immunoglobulin in human serum, and its biological activity is modulated by glycosylation on its fragment crystallizable region. Glycosylation of IgGs has shown to be related to aging, disease progression, protein stability, and many other vital processes. A common approach to analyze IgG glycosylation involves the release of the N-glycans by PNGase F, which cleaves the linkage between the asparagine residue and the innermost N-acetylglucosamine (GlcNAc) of all N-glycans except those containing a 3-linked fucose attached to the core GlcNAc. The biological significance of these glycans necessitates the development of accurate methods for their characterization and quantification. Currently, researchers either perform PNGase F deglycosylation on intact or trypsin-digested IgGs. Those who perform PNGase F deglycosylation on trypsin-digested IgGs argue that proteolysis is needed to reduce steric hindrance, whereas the other group states that this step is not needed, and the proteolytic step only adds time. There is minimal experimental evidence supporting either assumption. The importance of obtaining complete glycan release for accurate quantitation led us to investigate the kinetics of this deglycosylation reaction for intact IgGs and IgG glycopeptides. Statistically significant differences in the rate of deglycosylation performed on intact IgGs and trypsin-digested IgGs were determined, and the rate of PNGase F deglycosylation on trypsin-digested IgGs was found to be 3- to 4-times faster than on intact IgG.

ADDRESS CORRESPONDENCE TO: Lily Birx, Alcami, 4620 Creekstone Drive, Durham, NC 27707. (Phone: 770-743-9366; E-mail: [email protected]).

Conflict of Interest Disclosures: The authors declare no conflict of interest.

Funding Sources: This work was supported in part by National Institutes of Health grants 2R42GM113666 and 2R44GM131533.

Keywords: rate kinetics, deglycosylation, liquid chromatography-mass spectrometry, immunoglobulin G

INTRODUCTION

Immunoglobulins (Igs) are soluble serum glycoproteins utilized in the immune system to identify and neutralize pathogens. Immunoglobulin G (IgG) is the most abundant antibody in the human body, made up of 2 heavy chains and 2 light chains. IgGs are involved in multiple humoral immune processes, including antigen neutralization, complement activation and complement-dependent cytotoxicity (CDC), and antibody-dependent cell-mediated cytotoxicity (ADCC).[1] There are 4 subclasses of IgGs (IgG1, IgG2, IgG3, and IgG4) that differ in their constant regions, particularly in their hinges and upper CH2 domains.[2] IgGs have a conserved N-glycosylation site at aspargine-297 (Asn-297), which is in the CH2 domain of the crystallizable fragment (Fc) part of the heavy chains. In human IgGs, the majority of the Fc glycans are complex biantennary structures with a high degree of heterogeneity due to the presence or absence of different terminal sugars. The glycans attached can affect protein stability, bioactivity, and immunogenicity.[3] The biological significance of these glycans makes the development of accurate methods to analyze them vital.

N-linked glycosylation follows a conserved consensus sequence of Asn-X-Serine/Threonine, where X is any amino acid except proline.[4] Peptide-N4-(N-acetyl-beta-glucosaminyl)asparagine amidase (PNGase F) is commonly used to remove N-linked glycans. PNGase F cleaves the glycosidic bond between the carbohydrate-holding Asn residue, converts the Asn residue to aspartic acid, and releases ammonia and the intact glycan while generating a carbohydrate-free peptide.[5],[6] PNGase F is not able to cleave N-linked glycans from glycoproteins when the innermost GlcNAc residue is linked to an α1-3 Fucose residue. This modification is most commonly found in plant and some insect glycoproteins.[7],[8]

An analysis of IgG glycosylation is typically done by releasing the N-glycans using PNGase-F. A complete release of all glycans is necessary for accurate quantitation. There is a current divide on the benefits of trypsin digestion prior to PNGase-F release. One camp argues that proteolysis is needed to reduce steric hindrance, while the other states that this step is not needed. There is minimal experimental evidence supporting either assumption. The importance of obtaining complete glycan release for accurate quantitation led us to investigate the kinetics of this deglycosylation reaction for intact IgGs and IgG glycopeptides and to determine the effect of trypsin digestion on the rate of glycan release from glycopeptides of IgGs. Studies have shown that trypsin digestions on glycoproteins have a possibility of being incomplete due to glycan steric hindrance,[9],[10] but until now, no study has investigated the effect of steric hindrance on deglycosylation kinetics.

To facilitate the analysis of IgG glycosylation, we investigated the kinetics of PNGase-F deglycosylation on both intact and trypsin-digested IgGs to determine the effect of trypsin digestion on the rate of glycan release from IgG-glycopeptides. It is hypothesized that the rate of PNGase-F deglycosylation will be faster on the sample treated with trypsin beforehand.

METHODS

The overall experimental workflow is shown in Figure 1.

Figure 1

Experimental workflow.

4Human serum, trypsin (tosyl phenyl alanyl chloromethyl ketone (TPCK) treated), dithiothreitol (DTT), iodoacetamide (IDA), ammonium bicarbonate, ammonium formate, and formic acid (liquid chromatography [LC]–mass spectrometry [MS] grade) were purchased from Sigma-Aldrich. Acetonitrile (ACN; LC-MS grade) was purchased from Thermo Fisher Scientific. Other reagents were analytical grade.

A Hi-Trap protein G column (Cytiva) was used to purify IgGs from human serum (HSPD IgGs). An initial solution containing intact HS-IgGs in 50 mM ammonium bicarbonate (pH 7.8) was spiked with [Glu1]-Fibrinopeptide B (quantitation standard) (GluFib) and trypsin-digested SILAC-labeled IgG1 (kinetic standard) (GlycoScientific). This solution was split in half; one half was digested with trypsin first, and the other was deglycosylated with PNGase F first.

For the trypsin-first sample, reduction and alkylation were performed with DTT and IDA. The sample was dried to volatilize the excess DTT and IDA remaining and then resuspended in 50 mM ammonium bicarbonate (pH 7.8). Next, trypsin was added to the sample with the ratio of 1 part trypsin to 20 parts sample, and digestion was carried out at 37°C overnight. The trypsin digest sample was dried as a means to desalt the sample. The dried tryptic digest was redissolved in 50 mM ammonium bicarbonate. The dissolved tryptic peptides/glycopeptides were then treated with 0.1 U PNGase F per μg of IgG. Aliquots of the digestion mixture were removed at various time points, quenched by lowering the pH to 3.5 with formic acid, and then immediately frozen in dry ice. The samples were dried and then resuspended in 68% ACN for analysis. The ratio of PNGase-F in units to substrate has been experimentally determined so that the deglycosylation reaction takes 1 hour, which is long enough to allow for sufficient time points to be sampled in a reasonable time period.

Alternatively, the other half of the initial solution containing intact HSPD IgGs was deglycosylated first by treating the sample with 0.1 U PNGase F per μg of IgG. Aliquots were removed at various time points, quenched by lowering the pH to 3.5 with formic acid, and then immediately frozen in dry ice. The aliquots were dried and then resuspended in 50 mM ammonium bicarbonate (pH 7.8) for trypsin digestion. Reduction and alkylation were performed with DTT and IDA on the solubilized intact, deglycosylated IgGs. The sample was dried to volatilize the excess DTT and IDA remaining and then was resuspended in 50 mM ammonium bicarbonate (pH 7.8). Next, trypsin was added to the sample with the ratio of 1 part trypsin to 20 parts sample, and digestion was carried out at 37°C overnight. The PNGase F first sample was dried as means to desalt the sample and then resuspended in 68% ACN for analysis.

LC-MS analysis was performed on a Waters Synapt-G2 with a HP1100 LC. Peptides were separated using a column 2.1 mm × 150 mm HALO Penta-HILIC column packed with 2.7 um diameter superficially porous particles that have a 90 Å pore diameter (Advanced Materials Technology) at 60°C. The mobile phases used in the separation were 0.1% formic acid, 50 mM ammonium formate in water (A), and 0.1% formic acid in acetonitrile (B). The glycopeptides were bound to the column in 68% B, and a linear gradient to 54% B over 20 minutes was initiated to elute the peptides. Skyline was used for data analysis.[11]

RESULTS AND CONCLUSIONS

The kinetics of PNGase F deglycosylation reaction were studied by its addition to 2 solutions, one containing tryptic-digested IgG glycopeptides purified human serum and the other solution containing intact IgG purified from human serum. The release of glycans by PNGase F was monitored in both solutions. This approach allowed a broad range of human serum IgG glycopeptides to be monitored simultaneously by LC-MS. The glycan structures of the glycopeptides observed in the human serum IgGs and their abbreviations are shown in Figure 2. The LC-MS chromatograms illustrate how the peak area of the glycopeptides decrease with increasing digestion times (Figure 3). The normalized peak area of each glycopeptide ([GP]) was calculated by dividing the peak area of each glycopeptide by the peak area of the internal standard (GluFib) at each time point.

Figure 2

Glycan structures analyzed. Each structure with a “G1” can have 2 possible linkages of Gal.

Figure 3

LC-MS chromatograms show the disappearance of glycopeptide from 0 minutes to 1 hour for deglycosylation performed on intact and tryptic-digested IgGs.

PNGase F releases N-linked glycans from the peptide backbone by hydrolyzing the amide group of the Asn sidechain (Equation 1). PNGase F, an enzyme, is neither consumed nor produced during the reaction, and thus its concentration stays constant. H2O as a solvent is in excess compared with other reagents and remains constant throughout the process. Consequently, this third-order reaction behaves as a pseudo first-order reaction (Equation 2) since 2 of the 3 reactants are not consumed in the reaction. The differential rate equation describing the decrease in glycopeptide concentration as a function of time for pseudo first-order kinetics is shown in Equation 3 and the integrated form in Equation 4. In Equations 3 and 4, k is the rate constant.[12]

Glycopeptide+H2O+PNGase F NGlycan+Peptide+PNGase FGlycopeptide + H_{2}O + PNGase\ F\ \rightarrow N - Glycan + Peptide + PNGase\ F (1)

Glycopeptide NGlycan+PeptideGlycopeptide\ \rightarrow N - Glycan + Peptide (2)

d[GP]dt=k[GP]\frac{- d\lbrack GP\rbrack}{dt} = k\lbrack GP\rbrack (3)

ln[GP]= kt+ln[GP0]\ln\lbrack GP\rbrack = \ - kt + ln\lbrack GP_{0}\rbrack (4)

Plotting the natural log of [GP] (ln [GP]) with respect to time yields a straight line, as seen in Figure 4 for 2 IgG1 glycopeptides. These ln[GP] versus time plots are referred to as kinetic profiles. The linearity observed confirms that the PNGase F deglycosylation reaction can be modeled as a pseudo first-order process and allows a rate constant for the deglycosylation of each glycopeptide to be determined by calculating the slope of ln [GP] versus time plots. A rate standard of trypsin-digested SILAC-labeled IgG1 was used to standardize the deglycosylation rate between the samples. The kinetic profile for 2 IgG1 glycopeptides obtained carrying the Fucosylated glycan with 1 galactose (A2G1F) and the Fucosylated glycan with 2 galactoses (A2G2F) on trypsin-digested IgG and on intact IgG was used to check the standardization as well.

Figure 4

The plot of ln [GP] versus time for the IgG1 glycopeptide carrying the fucosylated glycan with 1 galactose (A2G1F) and the fucosylated glycan with 2 galactoses (A2G2F). This data was obtained following the addition of PNGase F at a concentration of 0.1 U PNGase F/μg IgG. The kinetic profile for A2G1Fand A2G2F on trypsin-digested IgG and on intact IgG is included in the graph.

The effects of trypsin digestion on the kinetics of PNGase F deglycosylation were investigated. The slopes of the lines (ln [GP] versus time), which are proportional to the rate constant of the PNGase F release, were calculated for each species, and these values are presented in Table 1. For each glycoform, the rate of PNGase F deglycosylation on trypsin-digested IgG glycopeptides is 3- to 4-times faster than on intact IgG. The faster rate of deglycosylation on trypsin-digested IgGs is presumably because of the reduced steric hindrance of the glycopeptides compared to the intact IgGs. The trypsin digest results in glycans that are more accessible to PNGase F, allowing for a faster deglycosylation compared to the glycans being buried in the intact structure. An additional factor that may account for some of the difference in kinetics centers on the increase diffusion rate of the smaller glycopeptide compared to the glycoprotein. The rate that the enzyme “encounters” the substrate is dependent on how fast the 2 species travel through solution, ie, diffuse. The calculated diffusion coefficient for a 2.5-kDa peptide is 4-times larger than that calculated for a 150-kDa protein at 20oC (2 × 10−10 m2/s versus 5 × 10−10 m2/s, assuming a globular shape for both species), which would account for the difference in deglycosylation.[13] While differences in steric hindrance and diffusion rates both offer potential explanations to the increased deglycosylation rate observed for glycopeptides over intact IgGs, there may be other reasons that account for this difference. Furthermore, the experimental data presented here cannot differentiate between these explanations.

Table 1

Average rate of deglycosylation on intact human serum IgGs and trypsin-digested human serum IgG glycopeptides

Rate (k) (min-1) (average +/- standard deviation)

Glycopeptide

Trypsin-digested IgG

Intact IgG

IgG1 A2G0

0.0133 +/- 0.0047

0.0036 +/- 0.0008

IgG1 A2G1

0.0161 +/- 0.0034

0.0082 +/- 0.0013

IgG1 A2G2

0.0139 +/- 0.0054

0.0047 +/- 0.0040

IgG1 A2G0N

0.0297 +/- 0.0042

0.0077 +/- 0.0048

IgG1 A2G1N

0.0263 +/- 0.0047

0.0160 +/- 0.0114

IgG1 A2G2N

0.0293 +/- 0.0104

0.0142 +/- 0.0020

IgG1 A2G0F

0.0171+/- 0.0026

0.0068 +/- 0.0028

IgG1 A2G1F

0.0169 +/- 0.0040

0.0070 +/- 0.0036

IgG1 A2G2F

0.0150 +/- 0.0051

0.0066 +/- 0.0027

IgG1 A2G2F1S1

0.0071 +/- 0.0018

0.0045 +/- 0.0032

IgG2/3 A2G2

0.0196 +/- 0.0045

0.0043 +/- 0.0023

IgG2/3 A2G0N

0.0165 +/- 0.0072

0.0058 +/- 0.0056

IgG2/3 A2G1N

0.0150 +/- 0.0053

0.0060 +/- 0.0024

IgG2/3 A2G2N

0.0143 +/- 0.0027

0.0046 +/- 0.0036

IgG2/3 A2G0F

0.0158 +/- 0.0043

0.0048 +/- 0.0042

IgG2/3 A2G1F

0.0181 +/- 0.0056

0.0072 +/- 0.0055

IgG2/3 A2G2F

0.0184 +/- 0.0039

0.0064 +/- 0.0017

IgG2/3 A2G2F1S1

0.0109 +/- 0.0055

0.0023 +/- 0.0019

IgG4 A2G1

0.0178 +/- 0.0046

0.0052 +/- 0.0018

IgG4 A2G2

0.0170 +/- 0.0030

0.0046 +/- 0.0004

IgG4 A2G1N

0.0157 +/- 0.0069

0.0086 +/- 0.0029

IgG4 A2G0F

0.0215 +/- 0.0086

0.0062 +/- 0.0041

IgG4 A2G1F

0.0179 +/- 0.0051

0.0068 +/- 0.0037

IgG4 A2G1NF

0.0101 +/- 0.0042

0.0036 +/- 0.0015

The half-life (t1/2) (Equation 5) of the glycopeptides deglycosylation performed on both intact IgGs and trypsin-digested IgGs was calculated using this adjusted rate, shown in Table 2. To enable multiple time points to be acquired in a reasonable period, a 100-times-lower concentration of PNGase F was used compared to standard conditions for deglycosylation; 0.1 U PNGase F per μg glycoprotein was used, whereas the standard amount of PNGase F typically used in a laboratory setting is 10 U PNGase F per μg glycoprotein. To estimate values under normal conditions, the rate of deglycosylation for each glycoform was multiplied by 100 for the half-life calculation. For each glycoform, the half-life trypsin-digested IgG glycopeptides were calculated to be an average of 3-times faster than on intact IgG. The time until deglycosylation, ie, the time it takes to completely remove a glycan, can be calculated in multiples of t1/2, and after 7 iterations, 0.01% of the glycopeptide would remain; thus, this would be our time until full deglycosylation, shown in Table 2.

t1/2= ln2k 0.693kt_{1/2} = \ \frac{\ln 2}{k}\ \approx \frac{0.693}{k} (5)

Table 2

Half-life and time until full deglycosylation on intact human serum IgGs and trypsin-digested human serum IgG glycopeptides

Half-life (min)

Time until Full deglycosylation (min)

Glycopeptide

Trypsin-digested IgG

Intact IgG

Trypsin-digested IgG

Intact IgG

IgG1 A2G0

0.5

1.6

3.2

11.2

IgG1 A2G1

0.4

0.8

2.7

5.6

IgG1 A2G2

0.7

1.7

4.7

11.8

IgG1 A2G0N

0.2

1.1

1.5

7.7

IgG1 A2G1N

0.3

0.6

1.9

3.9

IgG1 A2G2N

0.2

0.5

1.6

3.6

IgG1 A2G0F

0.4

0.9

2.4

6.4

IgG1 A2G1F

0.4

0.9

2.8

6.2

IgG1 A2G2F

0.5

0.9

3.3

6.6

IgG1 A2G2F1S1

1.0

1.6

7.2

11.2

IgG2/3 A2G2

0.3

1.9

2.4

13.4

IgG2/3 A2G0N

0.4

1.5

2.8

10.3

IgG2/3 A2G1N

0.4

1.3

2.8

9.1

IgG2/3 A2G2N

0.8

1.3

5.6

9.1

IgG2/3 A2G0F

0.4

1.2

3.0

8.6

IgG2/3 A2G1F

0.4

0.9

2.6

6.0

IgG2/3 A2G2F

0.4

1.5

2.4

10.3

IgG2/3 A2G2F1S1

0.6

6.33

4.2

43.9

IgG4 A2G1

0.3

0.7

2.2

5.4

IgG4 A2G2

0.48

1.2

3.3

8.3

IgG4 A2G1N

0.4

0.8

3.0

5.4

IgG4 A2G0F

0.3

1.1

2.0

7.5

IgG4 A2G1F

0.3

1.1

2.3

7.4

IgG4 A2G1NF

0.6

1.3

4.2

9.1

A Mann–Whitney U test was used to investigate if the differences in rates among deglycosylation performed on tryptic-digested IgGs and intact IgGs are statistically significant. The Mann–Whitney U test is a nonparametric test of the null hypothesis in which 2 samples have equal averages versus the alternative hypothesis in which the sample means from the 2 samples are not equal. The Mann–Whitney U test was selected over the 2-sample t test, as the Mann–Whitney U test does not require the assumption of normal distributions or a specific sample size.[14] The results from this Mann–Whitney U test are listed in Table 3 for each comparison category. The smaller the P value, the more likely one can reject the null hypothesis that the difference between the 2 groups is a result of random sampling. Therefore, small P values lead one to conclude that the populations are distinct. For example, a P value of 0.05 indicates a 5% risk of finding that a difference exists between 2 populations when there is no actual difference.

Table 3

Results of MannWhitney test 1: calculating the significance of differences in rates among deglycosylation performed on tryptic-digested IgGs and intact IgGs

Comparison

P

Result

k (IgG1 - trypsin first) vs. k (IgG1 - PNGase F first)

<0.001

k (IgG1 - PNGase F first) < k (IgG1 - trypsin first)

k (IgG2/3 - trypsin first) vs. k (IgG2/3 - PNGase F first)

<0.001

k (IgG2/3 - PNGase F first) < k (IgG2/3 - trypsin first)

k (IgG4 - trypsin first) vs. k (IgG4 - PNGase F first)

<0.001

k (IgG4 - PNGase F first) < k (IgG4 - trypsin first)

The results from the rate of deglycosylation performed on every glycoform attached to the intact IgG1 was compared to the rate of deglycosylation performed on every glycoform attached to the IgG1 tryptic glycopeptides. The P value of <0.001 obtained when the rate of deglycosylation of intact IgGs is compared with the rate of deglycosylation performed on tryptic-digested IgG glycopeptides suggests that the observed differences in rate constants between these 2 groups have a <0.1% chance of arising from random sampling, and thus the results are statistically significant.

These results show that complete deglycosylation performed on intact IgGs is expected when digestions are allowed to proceed overnight, or longer than 43.9 minutes; however, shorter times may result in incomplete glycan release. The difference in the rates observed for different glycans suggests errors in quantitation will result, as the rate of deglycosylation for some glycoforms is faster than others. For example, on intact IgGs, it takes 6.4 minutes to fully deglycosylate IgG1 A2G0F, whereas it takes 11.2 minutes to fully deglycosylate IgG1 A2G2F1S1. The same comparison on tryptic-digested IgGs found that it takes 2.4 minutes and 7.2 minutes, respectively. This data suggests that researchers who try to complete a rapid PNGase F glycan release, even on trypsin-digested IgGs and while utilizing a normal amount of PNGase F, may encounter issues with complete deglycosylation, as some glycoforms have not reached full deglycosylation in that time frame.

CONCLUSION

Significant differences in deglycosylation rate constants were observed between IgGs treated with trypsin and intact IgGs. The rate of PNGase F deglycosylation on trypsin-digested IgG glycopeptides is 3- to 4-times faster than on intact IgG. The time until full deglycosylation was also found to be 3- to 4-times faster on trypsin-digested IgGs than on intact IgG. The use of a normal amount of PNGase F would digest intact IgG, just at a slower rate. The differences in deglycosylation rate performed on trypsin-digested IgGs and intact IgGs suggest that deglycosylation should be performed on IgGs digested with trypsin to ensure complete deglycosylation in the fastest time to avoid quantitation errors because of incomplete deglycosylation.

Comments
0
comment
No comments here
Why not start the discussion?