A mass spectrometry–based assay for mouse IgG N-glycan screening in biofluids

Abstract

N-Glycans represent an important post-translational modification of proteins that can serve as biomarkers of disease, injury, and inflammation. N-Glycosylation of the monoclonal antibody immunoglobulin G (IgG) impacts binding to receptors that initiate an immunological response. Herein, we describe the optimization of a high-throughput method for analyzing IgG glycosylation of multiple murine biofluid samples in a single analysis utilizing matrix-assisted laser desorption/ionization (MALDI) imaging mass spectrometry. Similar to an enzyme-linked immunosorbent assay (ELISA), our method begins by spotting a capture antibody into a well. However, glycosylation on the capture antibody causes a significant N-glycan background signal that interferes with the signal from IgG-derived glycans in serum samples. To eliminate endogenous capture antibody IgG glycans (i.e., chemical background), the capture antibody was deglycosylated using the enzyme PNGase F, purified using affinity chromatography, and analyzed using ELISAs to confirm there was no loss of binding affinity and selectivity. The performance of the deglycosylated capture antibody was then compared to that of the traditional glycosylated capture antibody using the MALDI IgG N-glycan screening assay. Background subtraction was performed for samples analyzed with both capture antibodies to compare signal intensities before and after background subtraction, which was previously used to correct for the chemical background produced by glycosylated capture antibodies. We show that the use of background subtraction is not necessary with the use of a deglycosylated capture antibody, and that using the deglycosylated capture antibody increases imaging mass spectrometry signal intensity, giving a more sensitive, accurate, and precise analysis of N-glycans present in murine biofluid samples.

Introduction

Antibody N-glycosylation is an important post-translational modification (PTM) found on asparagine 297 of the hinge region. This modification has marked impacts on the effector functions of antibodies such as complement activation through C1q binding and cellular activation/inhibition through binding to Fc γ receptors [1, 2]. These differences in glycan structures have been implicated in many autoimmune illnesses, diseases, cancers, and more recently, the development of primary graft dysfunction following lung transplantation [3–7]. Matrix-assisted laser desorption/ionization (MALDI) mass spectrometry has been utilized to assess the total IgG glycome successfully in many of these cases [8–14]. Therefore, it is evident that monitoring glycans can be useful in gaining insight into the antibody function and pathogenicity as well as studying their correlation with disease status.

Electrospray (ESI) coupled to mass spectrometry via direct infusion or using liquid chromatography is a powerful tool for the analysis of glycans [15], though it can be challenging to link N-glycan signatures to their carrier proteins. Previously, MALDI imaging mass spectrometry has been used to analyze glycan profiles of tissue, biofluid, and immune cell sample types [13, 14, 16–20]. For example, an array-based method has been previously established for the analysis of N-glycans from biofluids that enables detection of N-glycan profiles from target glycoproteins captured on a substrate [18]. This process utilizes a capture antibody, such as immunoglobulin G (IgG), adhered onto a microscope slide, which can then bind to IgG present in serum or supernatant sample that is subsequently added to the slide. Following sample capture, the slide is washed and then exposed to peptide-N-glycosidase F (PNGase F), which is an endoglycosidase that cleaves the core N-acetylglucosamine from the asparagine residue containing the glycan [21]. This releases the N-glycans from the antibodies and allows for their subsequent detection by MALDI imaging mass spectrometry. A confounding factor with this approach is the need to perform background subtractions, since the IgG capture antibody is glycosylated. Glycosylation is a natural process in the packaging and release of all proteins, including IgG. Therefore, the protein antibody used as the capturing factor for this analysis is also intracellularly deglycosylated by the PNGase F enzyme, meaning the N-glycans observed during MALDI imaging mass spectrometry are derived from both the capture antibody and the serum IgG antibodies. The capture antibodies thus emit a background signal that interferes with the sample signal, diminishing the sensitivity and accuracy of this method.

One way to mitigate this N-glycan chemical noise from the capture antibody is to compare and subtract the level of N-glycan signal in capture controls as compared to test samples. The N-glycan background signal level derived from the capture antibody can thus be approximated and subtracted from the N-glycan signal observed in the serum sample [10, 18]. However, background subtraction introduces variability and may result in stochastic error and a diminished sensitivity, particularly in samples with low concentrations. To address this technical limitation, we have modified here this array-based method of N-glycan detection by using PNGase F to remove glycosylation from the capture antibody as a means to reduce the background N-glycan signal generated by the release of glycans from the capture antibody. We propose this approach will improve the sensitivity, accuracy, and precision of this methodology. This protocol is also adapted for the first time to analyze rodent samples (i.e., all previous uses of this type of antibody panel have been performed using human samples, such as serum from liver cirrhosis and liver fibrosis patients [18, 19]). As disease and inflammation states are commonly studied in rodents before human trials, this adaptation significantly broadens the applicability of this mass spectrometry–based IgG N-glycan screening assay [22–24].

Experimental

Materials

Well slide modules (ProPlate Multi-Array Slide System, 24-well Tray Set, 3 Slide Modules, 5.4 mm × 6.8 mm wells) and nitrocellulose-coated glass microscope slides (PATH protein microarray slides) were purchased from Grace Bio-Laboratories (Bend, OR, USA). Bovine serum albumin (BSA), phosphate-buffered saline (PBS), α-cyano-4-hydroxycinnamic acid (CHCA), and trifluoroacetic acid were purchased from Sigma-Aldrich (St. Louis, MO, USA). HPLC grade water, HPLC grade acetonitrile, trifluoroacetic acid, Pierce sodium acetate buffer, and n-octyl-β-D-glucopyranoside were purchased from Fisher Chemical (Waltham, MA, USA). Peptide-N-glycosidase F (PNGase F) Prime was purchased from Bulldog Bio, Inc. (Portsmouth, NH, USA). Goat anti-mouse IgG-Fc fragment antibody was purchased from Bethyl Laboratories (Montgomery, TX, USA). All standards and secondary detection antibodies were purchased from Southern Biotech (Birmingham, AL, USA). Nunc Maxisorp plates were purchased from Fisher Scientific (Waltham, MA, USA). Hitrap Protein G columns were purchased from Cytiva (Marlborough, MA, USA). Normal mouse serum was purchased from Thermo Fisher Scientific (Waltham, MA, USA).

Antibody preparation

All antibody samples to be purified were diluted in a Pierce sodium acetate buffer and pumped through a Protein G HiTrap column at 1 mL/min. Antibody was eluted in 500 μL aliquots using a 0.1 M low pH glycine–HCl buffer (pH 2.7) and immediately neutralized with Tris–HCl buffer (pH 8). The eluted antibody was buffer exchanged into PBS and quantified via Agilent BioTek H1 microvolume plate reader. Aliquots of control and deglycosylated capture antibody were stored in −20 °C.

Deglycosylation of the capture antibody was performed using a goat anti-mouse Fc antibody solution prepared at 1 mg/mL, which was first buffer exchanged into PBS and then incubated with PNGase F for 24 h at 37° C. Following incubation, the antibody was purified using a HiTrap Protein G HP column as above.

Enzyme‑linked immunosorbent assay

The goat anti-mouse IgG-Fc fragment antibody was used as the capture antibody and coated onto the plate at a concentration of 1 μg/mL overnight at 4 °C on a rocker. Following this, plates were blocked with 0.5% nonfat milk for 1 h at room temperature while on an orbital shaker. Plates were washed with PBS with 0.1% Tween 20 and then incubated shaking with samples for 2 h at room temperature. After another wash, secondary antibodies were added and incubated shaking for 1 h at room temperature. After a final wash, 100 μL of 3,3′,5,5′-tetramethylbenzidine (TMB) was added to each well and incubated for 10 min before stopping the reaction with 100 μL of stop solution. All plates were read at 450 nm on Agilent’s Biotek Synergy H1 (Santa Clara, CA).

Antibody array preparation

The general antibody array workflow has been described previously [18, 19]. Briefly, a 24-well slide module was mounted to nitrocellulose-coated microscope slides. Control and deglycosylated capture antibodies were diluted in PBS and manually spotted in wells at 200 ng per 1.5 μL spot. Spots were left to adhere overnight at 4 °C in a humidity chamber made from a 15.2 cm × 10.2 cm × 3.2 cm Western blot incubation box lined with a Wypall L40 and two rolled KimWipes saturated with distilled water. Slides were then placed in an electronic dry cabinet (Ruggard, New York, NY, USA) set at 32% humidity to dry. Slides were then rinsed for 1 min with 200 μL 0.1% n-octyl-β-D-glucopyranoside in 1 × PBS (referred to as PBS-OGS) per well to remove any unbound protein from the slide.

Sample capture and N-glycan release was first performed by blocking capture antibody spots with 200 μL of 5% BSA (prepared in PBS-OGS) for 1 h with gentle shaking. Wells were washed with 200 μL PBS (3 min × 2) and 200 μL double distilled water (1 min × 1) and dried in an electronic dry cabinet. Samples were diluted with PBS, added to wells (100 μL per well, saturating the capture antibody), and incubated at room temperature for 2 h in a humidity chamber with gentle shaking. Slides were then washed with 200 μL of PBS-OGS (1 min × 1), 200 μL PBS (3 min × 2), and 200 μL double distilled water (1 min × 1). The well module was removed, and an additional wash was performed by submerging the slide into 50 mL of water for 1 min to remove any residual salt. To cleave N-glycans from captured proteins, PNGase F Prime (0.1 μg/μL in HPLC grade water) was applied to the slides using a robotic sprayer (M5 TM-Sprayer, HTX Technologies, Chapel Hill, NC, USA) with the following spraying parameters: 15 passes at 45 °C, 10 psi, flow rate of 25 μL/min, and 1200 mm/min velocity. Slides were incubated overnight at 37 °C in humidity chambers made in cell culture dishes with WypAll L40 paper towels and two rolled KimWipes saturated with distilled water.

MALDI MS analysis

A CHCA MALDI matrix (7 mg/mL in 50% acetonitrile/0.1% trifluoroacetic acid) was applied to the slides using the same robotic sprayer mentioned above using the following spraying parameters: 2 passes at 80 °C, 10 psi, flow rate of 100 μL/min, and 1300 mm/min velocity. Slides were analyzed using a 7 T solariX XR Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer equipped with an Apollo II dual MALDI/ESI source and a dynamically harmonized ParaCell (Bruker Daltonics, Billerica, MA, USA). The MALDI source employs a Smartbeam II Nd:YAG laser system (2 kHz, 355 nm). Positive ion mode imaging mass spectrometry was performed using a pixel spacing of 300 μm, and data were collected from m/z 200 to 3750 using a 0.2447-s time-domain transient length, resulting in a resolving power of 8547 at m/z 1501.54. N-glycan m/z values are reported with mass defects to two decimal places due to the stringent wash procedure that ensures only N-glycans and matrix clusters are detected, with only N-glycans having mass defects ranging from 0.45 to 0.73. A random smartwalk pattern of 300 μm with 300 laser shots per pixel was used for sampling.

Data analysis

N-glycan localization and intensity were visualized using SCiLS Lab 2023c (Bruker Daltonics, Billerica, MA, USA), with data imported at a 0.98 ICR noise reduction threshold. An average mass spectrum was calculated for each sample spot in the array. Images were normalized to total ion current, and N-glycan peaks were assigned manually based on high-resolution accurate mass measurements. Spectra were recalibrated using Data Analysis 5.0 (Bruker Daltonics, Billerica, MA, USA) with an 8-point quadratic function based on CHCA matrix clusters. All ELISA analyses were conducted in GraphPad Prism. Background subtraction and signal quantification were performed using an in-house Python script (Python 3.11.3) that extracted the maximum intensity within ± 0.05 Da of the target m/z value for each pixel in the.imzML files. ROIs were determined by extracting the non-zero region of capture antibody signal at m/z 1647.59 for more precise quantification of N-glycan signal. Specifically, non-zero coordinates within each ROI were identified, and the bounding box encompassing these coordinates was used to extract the relevant region. The maximum peak intensity of m/z 1647.59 was used to perform quantification [25]. Error propagation was performed for all raw and background-subtracted intensities to calculate relative standard deviations (RSDs).

Results

Deglycosylated capture antibody performance

The IgG capture antibody was deglycosylated using PNGase F in order to reduce the chemical background in the resulting MS-based N-glycan assay. However, glycosylation is an important structure on antibodies that contributes to stability and binding efficacy [1, 2, 26]. In order to ensure that the capture antibody retains full binding function following deglycosylation, a sandwich ELISA was performed that analyzed binding to mouse total IgG standard at varying capture concentrations (Fig. 1). The deglycosylated capture antibody bound IgG with similar efficiency compared to the control capture antibody when plated at various concentrations. Binding affinities to antibody subclasses were then assessed using the same type of sandwich ELISA (Fig. 2b–f). The deglycosylated capture antibody bound all IgG subclasses at varying concentrations with similar affinities to those of control fully glycosylated capture. ELISA assays were also performed using normal mouse serum to ensure similar binding kinetics to the IgG standards (Fig. 3). As before, the deglycosylated capture antibody displayed similar binding to all IgG subclasses in serum samples.

Fig. 1.

Capture antibody performance was assessed using ELISA. Plates were coated with a starting concentration of 1 μg/mL and serial dilutions down the column of control or deglycosylated antibody. The binding of each to 100 ng/mL of mouse IgG standard was assessed via detection with an anti-mouse IgG-HRP antibody. Data show absorbance of the plate at 450 nm (n = 3)

Fig. 2.

Deglycosylation does not impact binding efficacy. Performance of glycosylated and de-glycosylated capture antibody was assessed using a standard IgG ELISA. Performance of IgG standards was compared using an ELISA plate coated with 1 μg/mL of either the control or deglycosylated capture antibody and incubated with serial dilutions of standards. Binding of capture to IgG total and subtype standards was assessed via detection with a anti-mouse IgG-HRP, b anti-mouse IgG1-HRP, c anti-mouse IgG2b-HRP, d anti-mouse IgG2c-HRP, and e anti-mouse IgG3-HRP. Data shows absorbance of the plate at 450 nm (n = 3)

Fig. 3.

Capture antibody performance of total IgG and subtypes in mouse serum was compared using an ELISA plate coated with 1 μg/mL of either the control or deglycosylated capture antibody. a IgG total binding was assessed by incubation with serum diluted 1:30,000 and serial dilutions and was detected via anti-IgG-HRP. IgG subtype binding was assessed with serum diluted 1:10,000 and serial dilutions and was detected with b anti-mouse IgG1-HRP, c anti-mouse IgG2b-HRP, d anti-mouse IgG2c-HRP, and e anti-mouse IgG3-HRP. All data show the absorbance at 450 nm (n = 3)

Glycan-glycan interactions are able to mediate IgG interactions with Fc γ receptors on cells [26]. We wanted to ensure deglycosylation of the capture antibody would not cause differential binding due to the absence of possible glycan-glycan interactions when binding to IgG. Using a mouse IgG2b mAb that was deglycosylated in a similar fashion, the ability of the control and deglycosylated capture antibody to bind mAb with and without glycans was compared (Fig. 4). No difference in binding capacity was observed between the control and deglycosylated capture. Overall, the ELISA data show that deglycosylation of the capture antibody does not alter binding potential to IgG, regardless of IgG glycan repertoire.

Fig. 4.

Binding of the control capture antibody to mouse monoclonal antibody with (mAb) and without (degmAb) glycosylation was assessed using an ELISA plate coated with 1 μg/mL of either the a control or b deglycosylated capture antibody. Bound IgG in both was assessed via anti-mouse IgG-HRP. Data show absorbance at 450 nm (n = 3)

Mass spectrometry–based N‑glycan assay

The workflow for an imaging mass spectrometry–based antibody panel has been previously described [18, 19]. Briefly, this involves spotting of a capture antibody onto a slide that can then selectively bind to the corresponding antibody in a biofluid sample. The N-glycans from the sample are then enzymatically cleaved from the sample antibody and, after application of a MALDI matrix, are analyzed utilizing imaging mass spectrometry (Fig. 5). In order to perform relative quantification of glycan levels between biofluid samples, background subtraction has previously been performed to remove N-glycan signal arising from the capture antibody. By instead using a deglycosylated capture antibody with this workflow, the need for background subtraction is removed. This allows for more accurate determination of glycan levels in samples and eliminates the unintended loss of sample signal that typically accompanies background subtraction.

Fig. 5.

The imaging mass spectrometry N-glycan assay is performed by first a adhering capture antibodies to a microscope slide and then blocking with BSA. A biofluid sample is then added to the spotted wells and glycoproteins in the sample are bound to the capture antibodies. b N-glycans are enzymatically cleaved from the peptide backbone utilizing PNGase F. A homogeneous layer of a MALDI matrix is then applied to the slide. c Finally, the panel is analyzed by MALDI imaging mass spectrometry to produce an average mass spectrum for each well in the array. Individual m/z peaks (i.e., N-glycans) can be visualized to determine the abundance of each N-glycan. Putative glycan structures shown are assigned based on literature [13, 18].

Ion images can be generated for each individual glycan detected, allowing for a visual representation of the relative intensity in the sample. For example, the glycan detected at m/z 1809.64 is present in all wells (i.e., those containing serum, IgG standard, and PBS control) with the control capture antibody (Fig. 6a). This glycan is tentatively identified as Hex5dHex1HexNAc4 based on an accurate mass measurement (0.049 ppm). However, in wells that have the deglycosylated capture antibody, signal is only observed in wells containing serum and IgG, and not in the PBS control. This indicates that removal of N-glycans from the capture antibody was successful and suggests that this glycan is present on both the control capture antibody and in the serum samples. This deglycosylation of the capture antibody allows for improved identification of N-glycans from samples. Deglycosylation of the capture antibody allows for the similar identification of other glycans that are absent in PBS control samples. For example, the glycans detected at m/z 1647.59 (tentatively identified as Hex4dHex1HexNAc4, 3.4 ppm) (Fig. S1a) and m/z 1663.59 (tentatively identified as Hex5HexNAc4, 3.1 ppm) (Fig. S1b) are both present in both the control capture antibody and serum samples. Conversely, the glycan detected at m/z 1704.61 is uniformly distributed across all wells with control capture antibody (Fig. 6b). This glycan is tentatively identified as Hex4HexNAc5 based on an accurate mass measurement (1.0 ppm). However, this glycan is not detected in wells that have the deglycosylated capture antibody, indicating that this glycan is only present on the capture antibody and not found in serum or IgG standard. The use of the deglycosylated capture antibody removes uncertainty of the presence of this glycan in the sample. Other glycans that are effectively removed from the capture antibody and are not present in serum and IgG samples include m/z 1339.48 (tentatively identified as Hex3HexNAc4) (Fig. S2a), m/z 1501.54 (tentatively identified as Hex4HexNAc4) (Fig. S2b), m/z 1850.67 (tentatively identified as Hex4dHex1HexNAc5) (Fig. S2c), m/z 1866.66 (tentatively identified as Hex5HexNAc5) (Fig. S2d), and m/z 2012.72 (tentatively identified as Hex5dHex1NAc5) (Fig. S2e).

Fig. 6.

Imaging mass spectrometry analysis of normal mouse serum, IgG standard, deglycosylated IgG standard, and PBS control is performed at 300 μm pixel size using either control or deglycosylated capture antibodies. Numerous glycans are detected, including a m/z 1809.64, which is observed in all wells with the control capture antibody, but is only observed in serum and IgG wells with the deglycosylated capture antibody; b m/z 1704.61, which is observed in all wells with the control capture antibody and is absent in all wells with the deglycosylated capture antibody; and c m/z 1485.53, which is observed across all wells with control capture antibody, but with minimal signal in the control, and observed at increased abundance in serum and IgG wells with deglycosylated capture antibody. Putative glycan structures shown are assigned based on literature.[13, 18] Structures were created with the use of GlycoWorkbench [27]

Due to variation in antibody glycosylation and sites, some N-glycans are inherently less concentrated on the capture antibody, which decreases the background signal observed during imaging mass spectrometry. This can allow for better visualization of sample signal directly from the image in these cases; however, the use of the deglycosylated antibody can still improve detection. For example, the glycan detected at m/z 1485.53, tentatively identified as Hex3dHex1HexNAc4 based on an accurate mass measurement (2.6 ppm), is detected in all wells with control capture antibody but is very lowly abundant in the PBS control (Fig. 6c). This suggests that there is a small amount of this glycan present on the capture antibody. This residual signal in the PBS control is absent when using the deglycosylated capture antibody. Interestingly, Hex3dHex1HexNAc4 glycan signal from serum and IgG standard is increased when using the deglycosylated capture antibody. This increase may be due to improved binding of IgG to the capture antibody, which may arise from reduced glycan-glycan interactions due to capture antibody deglycosylation, which results in less steric hindrance [1, 2]. A few N-glycans were not present on the capture antibody and only detected in serum samples. For example, the glycan detected at m/z 1282.45 (tentatively identified as Hex3dHex1HexNAc3, 0.71 ppm) is detected in serum and IgG wells with the control capture antibody and in serum wells with the deglycosylated capture antibody (Fig. S3a). This glycan is absent from the PBS controls with both the control and deglycosylated capture antibodies, suggesting it is not present on the capture antibody. Similarly, the glycan detected at m/z 2138.73 (tentatively identified as Hex5dHex1HexNAc4NeuGc1, 6.3 ppm) is only present in 1:100 serum wells with the deglycosylated capture antibody (Fig. S3b). This is most likely due to the glycan not being present on the capture antibody and the serum 1:100 samples having the highest concentration compared to other samples tested.

Performing this N-glycan assay with control capture antibodies requires the use of background subtraction to account for glycan signal originating from the capture antibody and enable accurate relative quantification of N-glycans. However, this method is imprecise and can result in the removal of sample signal, diminishing the limit of detection (LOD) of the assay. The use of a deglycosylated capture antibody alleviates the need for background subtraction and allows for direct measurement of sample N-glycans. The LOD is improved by a factor of 500, and the limit of quantitation (LOQ) is improved by a factor of 15 with the use of the deglycosylated capture antibody compared to the control capture antibody (Fig. S4, Tables S1, S2 and S3). N-glycan signal was still observed in the 1:25,000 serum dilution wells with the deglycosylated capture antibody, a three-fold improvement in sensitivity compared to the control capture antibody (data not shown). Table 1 shows the raw and background subtracted ion intensities of all glycans detected from the 1:100 mouse serum sample. In general, higher raw signal intensities are observed with the control capture antibody compared to the deglycosylated capture antibody, which indicates most glycans are present on the capture antibody to some degree. The relative standard deviations (RSDs) of the raw intensities are similar between the control and deglycosylated capture antibodies and are indicative of the inherent spot-to-spot and shot-to-shot variations within MALDI analysis due to uneven matrix crystallization and inconsistent laser fluences, respectively [28, 29]. Glycans highlighted in yellow in Table 1 are not observed with the deglycosylated capture antibody, indicating they are only present on the capture antibody and are not found in the serum. Background subtraction was then performed for the control capture antibody wells by subtracting the average intensities of the control PBS wells from the average intensities of the 1:100 serum dilution wells. As background glycan signal is absent or extremely minimal with the deglycosylated capture antibody, background subtraction has very little effect on the reported ion intensities in these wells. On average, 3.5-fold higher glycan signal intensities following background subtraction are observed with deglycosylated capture antibodies compared to the control capture antibodies. This demonstrates superior detection sensitivity and accuracy with the deglycosylated capture antibody workflow. Also, lower RSDs are observed with the deglycosylated capture antibody versus the control capture antibody after background subtraction. On average, background subtraction when using a control capture antibody results in an average standard deviation of ~ 130% across the glycans detected, with many glycans showing RSDs above 100% and as high as 594%. Many glycan signals are reported as negative intensities following background subtraction, highlighting the potential inaccuracy of this approach. Conversely, the use of the deglycosylated capture antibody resulted in an average standard deviation of ~ 23% across the glycans detected. This indicates improved precision with the deglycosylated capture antibody workflow. An exception to this trend is the glycan detected at m/z 1282.45, tentatively identified as Hex3dHex1HexNAc3, which is present only in the sample and not on the capture antibody. There is no difference between the raw and background subtracted intensities in the deglycosylated capture antibody samples since the background is zero.

Table 1.

Averaged N-glycan signal intensities for wells containing either control or deglycosylated capture antibodies following exposure to 1:100 serum. Raw ion intensities were taken directly from the MS images and background-subtracted ion intensities were calculated using an in-house Python script, where the average intensities of the control PBS wells were subtracted from the average intensities of the sample wells. All measurements reflect the average of n = 3 samples

		Raw Intensities		Background Subtracted Intensities
Glycan ID	m/z	Normal Capture	Deglycosylated Capture	Normal Capture	Deglycosylated Capture
Hex3dHex1HexNAc3	1282.45	9.80×104 ± 13%	5.40×104 ± 30%	9.80×104 ± 13%	5.40×104 ± 30%
Hex3HexNAc4	1339.48	1.26×106 ± 8%	0	−2.65×105 ± 145%	0
Hex3dHex1HexNAc4	1485.53	2.99×106 ± 10%	3.28×106 ± 9%	2.92×106 ± 11%	3.28×106 ± 9%
Hex4HexNAc4	1501.54	3.73×106 ± 10%	2.78×104 ± 17%	−9.34×105 ± 110%	2.78×104 ± 17%
Hex4dHex1HexNAc4	1647.59	6.06×106 ± 7%	5.59×106 ± 7%	3.79×106 ± 20%	4.26×106 ± 9%
Hex5HexNAc4	1663.59	5.62×106 ± 15%	7.81×105 ± 27%	−2.47×105 ± 594%	7.81×105 ± 27%
Hex4HexNAc5	1704.61	2.38×106 ± 7%	0	−9.32×105 ± 92 %	0
Hex5dHex1HexNAc4	1809.64	3.66×106 ± 9%	1.50×106 ± 15%	8.76×105 ± 94%	1.50×106 ± 15%
Hex4dHex1HexNAc5	1850.67	7.40×105 ± 6%	0	−3.67×105 ± 110%	0
Hex5HexNAc5	1866.66	2.01×106 ± 8%	0	8.72×105 ± 95%	0
Hex5dHex1HexNAc5	2012.72	1.11×106 ± 5%	0	−4.41×105 ± 112%	0
Hex5dHex1HexNAc4NeuGc1	2138.73	4.22×104 ± 173%	6.19×104 ± 57%	4.22×104 ± 173%	6.19×104 ± 57%

only present in sample

only present on capture

successfully removed from capture

The normal serum mouse IgG concentration range is 5.1–8.6 mg/mL [30]. Changes in total serum IgG level are seen in rodent models with immune diseases such as systemic lupus nephritis and Sjögren’s syndrome, which are associated with increased circulating levels of IgG [31–35]. Similarly, diseases such as immunodeficiency and cancer are often associated with decreased circulating IgG levels [36–38]. However, these overall systemic changes would likely not impact the performance of this assay, as it is based on saturation of the capture antibody and as little as 1 μg of IgG is needed for saturation. In the unlikely event that saturation could not be achieved for this assay due to low circulating levels of IgG, total serum IgG concentrations and load equivalent IgG concentrations could be quantified to facilitate comparisons.

Conclusions

Here, we present a novel murine fluid phase protocol for IgG glycan assessment. This workflow leverages MALDI imaging mass spectrometry to profile N-glycans following their ELISA-like capture and subsequent release via enzymatic cleavage. The capture antibody here is deglycosylated via PNGase treatment to remove background signal. This deglycosylation results in no change in binding affinity or efficiency of the capture antibody to IgG, either from purified IgG standard or from IgG in mouse serum. The use of a deglycosylated capture antibody in the MALDI imaging mass spectrometry panel allows for better visualization of glycan signal intensity differences between samples. This eliminates the need for background subtraction and results in improved sensitivity, accuracy, and precision. Future incorporation of an internal standard could further improve accuracy and precision. The deglycosylation of the capture antibody provides an improved tool to assess glycosylation of proteins from different sample types. To our knowledge, this is the first MALDI-based glycan array in murine biological samples, which will improve murine glycan analysis in many fluid samples. Furthermore, the array layout enables the analysis of samples from multiple subjects and/or time points, which can be adapted for high-throughput analyses. We believe that this will provide a sensitive and robust approach to analyze glycans in various settings, including preclinical models.

Supplementary Material

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00216-025-05994-x.

Data availability

Data are available upon reasonable request.