Storage stability and HILIC-UHPLC-FLR analysis of immunoglobulin G N-glycome from saliva

Abstract

Immunoglobulin G (IgG) is the most abundant antibody in the blood and plays a critical role in host immune defense against infectious agents. Glycosylation is known to modulate the effector functions of IgG and is involved in disease development and progression. It is no surprise that the N-glycome of IgG from plasma has already been proposed as a biomarker for various physiological and pathological conditions. However, because saliva is easy to collect, it could be useful for exploring the functional role of salivary IgG N-glycosylation and its potential as a diagnostic biomarker. Therefore, in this study, we described a method for N-glycome analysis of IgG from saliva samples. Salivary IgG N-glycans were analyzed by ultra-high-performance liquid chromatography based on hydrophilic interactions with fluorescence detection (HILIC-UHPLC-FLR). In addition, we compared IgG N-glycan profiles from saliva with those from plasma, assessed the stability of salivary IgG N-glycan profiles under different storage conditions, and evaluated the effects of using a saliva preservation medium. This study provides an ultrasensitive UHPLC method for the analysis of total IgG N-glycosylation from saliva, gives insight into storage stability of salivary IgG, and highlights its (dis)advantages for further biomarker-related research.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s00216-023-04682-y.

Introduction

Immunoglobulin G (IgG) is the most abundant antibody in the bloodstream and is often referred to as the “workhorse” of systemic humoral immunity. It plays a central role in adaptive immunity but also in the dysregulation of immunity that can lead to chronic inflammatory diseases or cancer [1]. While its antigen-binding fragment (Fab) recognizes and binds antigens, the fragment crystallizable (Fc) region interacts with Fcγ receptors (FcγRs) on the surface of many immune cells, triggering distinct immune responses such as antigen neutralization, macrophage phagocytosis, antibody-dependent cell cytotoxicity (ADCC), and complement activation [2]. In the Fc region, IgG molecules contain a highly conserved N-glycosylation site at Asn-297. Fc N-glycosylation plays a critical role in the protein stability and effector functions of IgG. It affects the interaction with IgG Fc receptors and ligands and is crucial for maintaining the proinflammatory and anti-inflammatory effector functions of IgG [3, 4]. Analysis of the IgG glycome has shown that changes in monosaccharide composition are associated with several inflammatory and autoimmune diseases [5]. Therefore, it is being investigated as a potential biomarker for early detection, diagnosis, and prognosis of various pathologies [6–9], leading to more appropriate treatment and better patient outcomes. To this end, IgG glycosylation has been studied in various biological matrices, mainly plasma and serum. The protocol for the isolation of IgG from plasma/serum samples requires trained personnel and laboratory equipment for blood collection and serum/plasma separation, along with specific and costly storage and shipping conditions for infectious agents. In addition, blood collection is an invasive method. Therefore, it may not be an ideal solution for patients who have a negative attitude toward venipuncture due to pain or hygienic concerns.

The use of saliva for diagnostic purposes is increasing because it offers a few advantages over plasma or serum. First, saliva can be easily and noninvasively collected from donors without needing specialized equipment or training. Second, saliva samples are less infectious than blood and do not clot, so minimal manipulation is required. Some saliva tests are already available on the market, e.g., for detecting antibodies against severe acute respiratory syndrome coronavirus 2 [10]. In addition, saliva is not only used as a plasma/serum replacement in clinical diagnostics but is also being further investigated as a potential biomarker for a variety of local, systemic, and infectious diseases [11–13].

Previously, N-glycosylation of IgG from saliva has been studied using lectin assays [14]. More recently, using LC–MS analysis, Plomp et al. [15] analyzed the N-glycome of IgG from saliva and the corresponding IgG from plasma. One of the significant challenges in analyzing the IgG N-glycome from saliva is that the antibody concentrations are much lower than in plasma. While plasma contains about 12.5 mg/mL IgG, the concentration in unstimulated whole saliva is estimated to be about 0.014 mg/mL [16].

A robust yet sensitive workflow is essential for high-throughput analysis of IgG N-glycome from saliva. An important factor that can affect the accuracy of analysis is the storage stability of biomolecules, especially in longitudinal studies where sample collection and storage can take months or years. Studies on the stability of biomolecules such as proteins, lipids, and metabolites have shown that they are generally stable in frozen biofluids [17, 18]. Similarly, the N-glycosylation profile of IgG did not change dramatically when biological matrices (e.g., serum, plasma, dried blood spots) were stored under appropriate conditions [19–21]. As for saliva, available data suggest that different saliva analytes remained stable under different storage conditions [22–24]. However, there is no corresponding data on how the storage conditions of saliva may affect its IgG N-glycome.

In this work, we have described a method for analyzing the N-glycosylation profile of salivary IgG using bead-based immunoprecipitation followed by enzymatic deglycosylation and analysis of the labeled N-glycans by ultra-high-performance liquid chromatography based on hydrophilic interactions with fluorescence detection (HILIC-UHPLC-FLR). In addition, we performed a series of stability tests of IgG N-glycosylation in saliva from nine donors under different storage conditions.

Materials and methods

Collection of biofluids

In order to compare IgG N-glycosylation from plasma and saliva, blood and saliva samples were obtained from nine healthy donors, including four women and five men aged 26 to 51 years. Saliva and plasma samples were collected on the same day within a 2-h time window. Eight milliliters of peripheral blood was collected in an EDTA vial. After settling at room temperature for 1 h, blood samples were centrifuged at 2700 g for 10 min, and the clear upper fraction, which consisted of plasma, was collected. Saliva was collected by unstimulated drooling into a 5-mL Eppendorf tube for approximately 15 min and immediately put on ice. For storage stability experiments, saliva was collected from nine healthy volunteers: six women and three men aged 27 to 39 years. Saliva was collected by unstimulated drooling into 15-mL Falcon tubes for approximately 60 min. After the first 30 min, samples were put on ice, and donors were given a new 15-mL Falcon tube for the collection. After the desired volume was collected, saliva from each donor was pooled and put on ice. All donors were instructed to rinse their mouths prior to the collection and to abstain from eating 1 h before the collection, as well as during the collection of saliva. To increase the efficiency of IgG purification, all samples were subjected to an additional step: saliva and plasma samples were centrifuged at 10,000 g for 30 min and 12,000 g for 3 min, respectively. Supernatants were collected and used for the IgG purification.

All donors were familiarized with the aims of the study and samples were collected in accordance with Croatian regulations and after approval by the Genos Ethics Committee in accordance with the Declaration of Helsinki.

IgG purification from human saliva and plasma

For comparison of IgG N-glycans from saliva and plasma, IgG was captured from 5 mL of saliva and 30 μL of plasma diluted in 4 mL of phosphate-buffered saline (PBS). For purification, 20 μL of Protein G Agarose Fast Flow beads (Merck Millipore, Darmstadt, Germany) was added to 5-mL Eppendorf tubes containing either saliva or diluted plasma. For the storage stability experiment, isolation of IgG from saliva stored at different temperatures (− 80°C, − 20°C, 4, RT, 37°C, and 50°C) was performed at three time points: 0, 3, and 15 days after collection. Briefly, 20 μL of Protein G Agarose Fast Flow beads (Merck Millipore, Darmstadt, Germany) were added to 2-mL Eppendorf tubes containing either 0.5 mL of whole saliva or 0.5 mL of whole saliva with 0.5 mL of preservation medium (Biocomma Limited). To avoid possible batch effect between different 96-well plates, two rounds of sample randomization were performed: one before IgG isolation for all samples analyzed at a single time point, and another one before HILIC-UHPLC-FLR analysis of labeled N-glycans for all samples (all donors, all temperatures, all time points). Samples were incubated for 2 h at 800 rpm on the plate shaker to achieve IgG capturing. Samples were then briefly centrifuged, and the supernatants discarded. Sample volumes were brought to a total of 300 μL by the addition of PBS and transferred to the 96-well Orochem filter plate (10-μm pore size; Orochem Technologies, Naperville, IL). Using a vacuum manifold, IgG-bound beads were washed three times with 200 μL PBS and three times with 200 μL deionized water (Purelab Ultra, maintained at 18.2 MΩ). IgG release from the beads was achieved by incubation in 100 mM formic acid for 15 min at RT. IgGs were eluted from the filter plate by centrifugation for 2 min at 50 × g into a PCR plate containing 17 µL of 1 M ammonium bicarbonate. For N-glycan analysis, 50 µL of the samples was dried for 2 h at 37°C in a vacuum centrifuge.

Glycan release, labeling, and clean up

N-glycans were released, fluorescently labeled, and cleaned up as described previously [25], with modification by using procainamide hydrochloride (ProA) as a fluorescent labeling agent for N-glycans. Briefly, isolated IgG was denatured with sodium dodecyl sulfate (SDS) (Invitrogen, USA) and by incubation at 65°C. After enzymatic release with PNGase F (Promega, USA), N-glycans were fluorescently ProA labeled in a 2-step procedure. First, 25 µL of procainamide mixture containing 4.32 mg of procainamide hydrochloride (Sigma-Aldrich) in glacial acetic acid (Honeywell, Charlotte, NC, USA)/dimethyl sulfoxide (Sigma-Aldrich) (30:70) was added per sample, followed by incubation at 65°C for 1 h. Then, 25 µL of reducing agent solution consisting of 4.48 mg of 2-picoline borane (J&K Scientific, Beijing, China) in glacial acetic acid/dimethyl sulfoxide (30:70) was added per sample and incubated at 65°C for an additional 1.5 h. Samples were then cooled for 30 min at RT. Excess reagents were removed from the samples using hydrophilic interaction liquid chromatography–solid phase extraction (HILIC-SPE) as described in [25]. For the HILIC-SPE step, an AcroPrepadv 1 mL 0.2 μm wwPTFE plate (Pall) was used as the stationary phase. The labeled glycans were eluted with ultrapure water and stored at − 20°C until use.

HILIC-UHPLC-FLR analysis

Fluorescently labeled N-glycans were analyzed by HILIC-UHPLC-FLR on a Waters Acquity UPLC instrument (Waters, USA), which consisted of a quaternary solvent manager, a sample manager, and an FLR fluorescence detector set with excitation and emission wavelengths of 310 and 370 nm, respectively. Labeled N-glycans were separated on a Waters UPLC glycan bridged ethylene hybrid (BEH) amide chromatographic columns, 100 × 2.1 mm i.d., 1.7 µm BEH particles, with 100 mM ammonium formate, pH 4.4 as solvent A, and acetonitrile as solvent B. The separation was performed using a linear gradient of 25–38% solvent A at a flow rate of 0.40 mL/min in a 27-min analytical run. Samples were maintained at 10°C prior to injection, and the separation temperature was 60°C. After the system has been equilibrated to starting conditions and the FLR detector has been stabilized, an analysis was run. In each UHPLC analysis, samples are run along with blanks and internal UHPLC standards prepared in-house following internal standardized operational protocol (standard 1: ProA labeled glucose homopolymer (GHP) ladder (Dextran ladder) and standard 2: ProA labeled N-glycans from pooled IgG). Data processing was performed using an automated processing method with a traditional integration algorithm. Then, each chromatogram was manually corrected to maintain the same integration intervals for all samples. Before manual integration, Quality control (QC) of all chromatograms was performed. This included sufficient intensity of the signal for the integration and acceptable shape (without tailing, fronting, and peak splitting) of all peaks, absence of baseline drift, absence of significant shift of chromatogram peaks (≤ 0.05 glucose units) and dextran ladder (≤ 0.06 min), and absence of any unexpected peaks within the chromatogram. The chromatograms were all divided into 24 chromatographic glycan peaks (GP) in the same way as previously reported (Fig. S1) [26]. The amount of glycans in each peak was expressed as % of the total integrated area. From directly measured 24 glycan structures, six derived traits were calculated according to the following formulas: for agalactosylated G0 = GP1 + GP2 + GP4 + GP6, with one galactose G1 = GP7 + GP8 + GP9 + GP10 + GP11, with two galactoses G2 = GP12 + GP13 + GP14 + GP15, with a bisecting GlcNac B = GP3 + GP6 + GP10 + GP11 + GP13 + GP15 + GP19 + GP22 + GP24, with a core fucose F = GP1 + GP4 + GP6 + GP8 + GP9 + GP10 + GP11 + GP14 + GP15 + GP16 + GP18 + GP19 + GP23 + GP24 and sialylated glycans S = GP16 + GP17 + GP18 + GP19 + GP20 + GP21 + GP22 + GP23 + GP24; for each derived trait, the obtained sum was divided by the sum of all peaks and multiplied by 100.

Statistical analysis

Data was analyzed and visualized using Python programming language (version 3.9.1) and Jupyter Notebooks [27]. Before statistical analysis, normalization by total area was performed to make measurements across samples comparable, and the measurements were log-transformed due to right skewness of their distributions. Differences between different conditions and different sources of glycomic measurements were analyzed using a general linear model while the Bonferroni method was used for a multiple-comparison correction.

Results

Method development

IgG glycome composition was analyzed in paired plasma and saliva samples from nine healthy donors. The composition of the total IgG N-glycome (combined Fc and Fab glycans) was determined by HILIC-UHPLC-FLR of ProA labeled N-glycans. Statistical analysis was performed for the directly measured glycan structures and the main summary features of IgG glycan composition (G0—glycans without galactose, G1—glycans with one galactose, G2—glycans with two galactoses, S—percentage of all glycans with sialic acid, F—percentage of core fucosylated glycans, and B—percentage of glycans with bisecting GlcNAc).

IgG from saliva exhibited a lower degree of monogalactosylation and a higher degree of agalactosylation compared with plasma (Table 1, Fig. 1). In addition, a higher abundance of sialylated IgG N-glycans was observed in plasma-derived IgG compared with saliva, specifically glycan structures GP23 (FA2G2S2) and GP24 (FA2BG2S2) (Table S1, Fig. S2). Interestingly, there was a statistically significant difference in the presence of the GP5 glycan structure between saliva and plasma (Table S1, Fig. S2). This high-mannose glycan structure was much more abundant in the N-glycome of salivary IgG than in plasma. Consistent changes in the content of core fucosylated and bisected N-glycan structures between salivary and plasma IgG were not detected.

Table 1.

Differences between IgG N-glycan composition isolated from saliva and plasma samples

Glycan trait	Effect	P-value	Padj-value*
G0	0.43	6.00E − 04	1.67E − 02
G1	− 0.14	4.00E − 04	1.14E − 02
G2	− 0.39	5.10E − 03	1.52E − 01
S	− 0.22	1.02E − 02	3.06E − 01
F	0.00	8.50E − 01	1
B	0.01	9.02E − 01	1

^*Bonferroni’s method

Bolded entries represent statistically significant results (p-value < 0.05)

Fig. 1.

Relative abundance of six IgG N-glycan derived traits in saliva and plasma. Lines are connecting corresponding results from plasma and saliva for each donor. G0, agalactosylation; G1, monogalactosylation; G2, digalactosylation; S, sialylation; F, core fucosylation; B, bisecting GlcNAc; P, plasma; S, saliva; Values, abundance (%) of derived trait in total IgG glycome

Storage stability of saliva IgG N-glycome

Prior to the storage stability experiment, we examined a series of different volumes of saliva to select the one that would provide adequate results in a cost-effective manner. Saliva from one subject was aliquoted into four different volumes—2 mL, 1 mL, 0.5 mL, and 0.2 mL. IgG was isolated from the saliva by bead-based immunoprecipitation as described above, followed by enzymatic deglycosylation and HILIC-UHPLC-FLR analysis of the labeled IgG N-glycans. The chromatographic profiles of N-glycans of IgG isolated from different initial volumes of saliva (2 mL, 1 mL, 0.5 mL, and 0.2 mL) were comparable (Fig. S3). We chose 0.5 mL as the minimum volume of saliva which provided satisfactory results considering interindividual variation of IgG concentration in saliva.

A series of storage conditions were investigated (Table 2) and compared with the reference condition at the sampling point (IgG isolated from saliva immediately after collection). The whole saliva was stored at six different temperatures (− 80°C, − 20°C, 4, RT, 37°C, and 50°C) for up to 15 days. Samples were stored in − 80°C/ − 20°C freezers because of their widespread use. Refrigerator storage (4°C) was considered for convenience, as was RT, which also resembles accidental sample storage. Storage at 37 and 50°C allowed investigation of conditions for accelerated storage stability [19]. In addition, a commercially available saliva preservation medium was tested for improvement of the stability of saliva analytes. For each condition, 0.5 mL of collected saliva was aliquoted into 2 mL tubes. Where indicated, the preservation medium was added in a 1:1 ratio. In the analysis, we only included data that passed the quality control (QC) (Table S2).

Table 2.

Storage stability conditions applied to saliva samples

	No preservation medium		Preservation medium
Temperature (°C)	Day 3	Day 15	Day 3	Day 15
− 80	-	+	-	+
− 20	+	+	+	+
4	+	+	+	+
RT	+	+	+	+
37	+	+	+	+
50	+	+	+	+

RT room temperature, + storage conditions that were investigated,—storage conditions that were not investigated

First, the samples with preservation medium were more stable than those without (Fig. 2, Fig. S4, Fig. S5, Table S2). We encountered loss of IgG in saliva stored at 37°C and − 20°C only in samples stored without preservation medium. While samples without preservation medium stored at − 80°C showed no changes in glycosylation-derived traits after 15 days, interindividual variation and glycosylation instability was observed already after 3 days at higher temperatures. In samples stored at RT and 50°C, there was a decrease in core fucosylation and monogalactosylation, while bisection was marginally increased (Table S3). Of individual glycosylation traits, there was a significant increase in bisected glycan structures with and without galactose (GP3 and GP13) and an increase in high-mannose structure (GP5) (Table S5). After 15 days, besides RT and 50°C, these differences were further observed in samples stored at 4 °C (Table S4, Table S6). The abundance of sialylated glycans remained stable. On the other hand, samples stored with preservation medium showed an overall stable N-glycosylation profile at lower temperatures (− 80°C, − 20°C, and 4°C), whereas higher temperatures, especially 37°C affected the composition of IgG N-glycans (Table S3, Table S4, Table S5, and Table S6). Again, the most prominent changes were observed in the abundance of core fucosylation. A decrease in core fucosylation was observed at RT, 37°C, and 50°C after a period of 15 days. A decrease was observed in corresponding core fucosylated glycan structures—GP4, GP8, GP9, and GP14 (Table S6). A temperature of 37°C affects the IgG glycosylation profile in saliva the most, as the relative abundance of core fucosylated glycans decreased to as low as 70% (compared to > 90% under reference conditions). Greater interindividual differences were also observed in galactosylation at 37°C, especially after 15 days, during which we observed a significant decrease in monogalactosylation (Fig. 2, Table S3, Table S4).

Fig. 2.

Relative abundance of six N-glycan derived traits of IgG from saliva stored at different temperatures for three and fifteen days. Condition 1 represents saliva stored without preservation medium and condition 2 represents saliva stored with preservation medium. R represents the reference condition. Value, abundance (%) of derived trait in total IgG glycome. Median glycan values for each timepoint are bolded. RT, room temperature

Discussion

Here, we report a comprehensive analysis of total IgG N-glycome from human saliva and its stability in different storage conditions. IgG was purified from the plasma and saliva of nine healthy individuals and analyzed by HILIC-UHPLC-FLR. We found that the N-glycan profile of IgG from saliva had a lower degree of monogalactosylation compared to plasma-derived IgG. Regarding sialylation, a significantly lower abundance of structures FA2G2S2 and FA2BG2S2 was observed in saliva-derived IgG compared with IgG from plasma. The abundance of structures with core fucose and/or bisecting GlcNAc was similar between IgG derived from plasma and saliva. Published data on site-specific glycoprofiling of IgG from saliva provided similar results, showing that IgG1 (the most abundant of all IgG subclasses [1]) and IgG4 had slightly lower levels of galactosylation and sialylation compared with IgG from plasma [15]. Decreased galactosylation and sialylation are recognized features of proinflammatory IgG [4]. In addition, salivary IgG has been observed to have a higher abundance of oligomannose N-glycan structures, another feature associated with inflammation. Glycans with exposed mannose are recognized and bound by mannose-binding lectin (MBL), initiating the lectin pathway of complement activation [28]. Also, IgG containing high-mannose Fc glycans has previously been shown to be cleared more rapidly than other glycan forms [29]. This may, in a sense, represent a safety switch leading to a short-term proinflammatory effector function of salivary IgG when it encounters pathogens, without triggering a chronic inflammatory state.

The stability of biomolecules is an important factor affecting their efficacy as clinical biomarkers. Although several studies have shown that different salivary analytes remain stable under different storage conditions [22–24], the effects of saliva storage conditions on the stability of N-glycosylation of IgG have not yet been investigated. Using saliva samples from nine donors, we investigated the short-term stability and accelerated storage stability of salivary IgG N-glycosylation and the effects of the saliva preservation medium. As for short-term stability, we did not observe any changes in the IgG N-glycosylation features after 15 days in samples stored frozen (− 80°C and − 20°C) with preservation medium. These conditions are common in glycomics, as samples are stored at − 20°C or sometimes at − 80°C between collection and N-glycosylation analysis, depending on the biochemical properties of the sample. However, we encountered an unexplained loss of IgG in whole saliva stored at − 20°C without preservation medium. This was evident after both three and 15 days, even though samples were stored in separate set of tubes for each timepoint and randomized prior to isolation. Therefore, it is unlikely that bias or systemic error is responsible for the loss. As for 4°C, IgG glycosylation was stable in saliva with preservation medium, whereas instability was observed in saliva without it. Therefore, saliva samples may even be briefly stored refrigerated prior to analysis of IgG N-glycosylation. Storage for 15 days at RT reduced core fucosylation and to some extent galactosylation, while bisection and sialylation remained stable. The effects were more pronounced in saliva without preservation medium, where only half of the samples passed the QC. These results demonstrate higher stability of salivary IgG N-glycans when preservation medium is used for storage.

The effects of long-term storage on IgG N-glycosylation were investigated in an accelerated storage study. Samples were exposed to elevated temperatures to simulate the long-term storage of frozen samples. This approach was adopted from the storage stability study of N-glycosylation of serum/plasma IgG [19]. More specifically, samples were stored at 37 and 50°C for 15 days and the results were compared with the reference condition (IgG isolated from saliva immediately after collection). The use of the saliva preservation medium provided a more reliable and stable IgG N-glycosylation profile than whole saliva without it. This was particularly evident at 37°C, as we did not obtain IgG from saliva without medium after either three or 15 days. A logical explanation for the loss of IgG could be the fact that this is the optimal temperature for most human enzymes, including proteases present in saliva, which could degrade immunoglobulins in addition to other proteins. This would prevent the capture of IgG via bead-based immunoprecipitation and make analysis of its N-glycans impossible. On the other hand, the preservation medium usually contains detergents and/or protease inhibitors that could prevent this. Indeed, IgG from only one saliva sample stored with preservation at 37°C did not pass the QC. However, significant changes in glycosylation were observed after storage at 37°C for 15 days. The most notable change was the decrease in levels of core fucosylation. Interestingly, the enzyme responsible for the cleavage of core fucose, α-L-fucosidase (FUCA1), is highly expressed in healthy salivary glands [30] and whole saliva [31], so its presence in saliva and the fact that the biocatalytic activity of human enzymes shows the greatest effects at 37°C may contribute to the substantial decrease in core fucosylation. A decrease was also observed at 50°C after 15 days, but not to the same extent as at 37°C. This could be due to chemical hydrolysis, which would be expected to increase with temperature, rather than enzymatic activity, as 50°C could potentially induce enzyme degradation. Furthermore, galactosylation has been shown to be unstable under accelerated storage conditions. Similar to core fucosylation, the most profound changes were observed at 37°C. The presence of β-galactosidase in whole saliva [31], an enzyme that cleaves β-1–4-linked galactose, may contribute to decreased monogalactosylation at 37°C, whereas hydrolysis of glycosidic linkages at higher temperatures, e.g., 50°C, may be responsible for higher instability and variation in galactosylation of salivary IgG.

These results suggest that saliva samples in preservation medium could be stored frozen (− 80°C, − 20°C) or at 4°C for weeks without noticeable effect on subsequent analysis of IgG Fc N-glycosylation. Storage of saliva at higher temperatures or for extensive period of time is likely to alter the composition of salivary IgG N-glycosylation. Therefore, appropriate storage and/or timely isolation of IgG from saliva will preserve the uniformity of its N-glycome.

In conclusion, our study provides a novel UHPLC method for analyzing total IgG N-glycans in saliva and a better understanding of the stability of salivary IgG N-glycosylation during sample storage. We have shown that saliva is a suitable starting material for the analysis of IgG N-glycans and can be stored under appropriate conditions for an extended period of time. Furthermore, our results suggest that saliva may not be a reliable substitute for plasma because of the significant differences between the N-glycome profile of salivary and plasma IgG. Nevertheless, alterations in IgG N-glycome in saliva and plasma could go in the same direction in different diseases. Also, saliva collection has advantages over plasma collection, mainly because it is painless and does not require specially trained personnel or instruments. Therefore, N-glycosylation of IgG in saliva represents a potential for future research on its functional role as a novel biomarker and/or therapeutic target in local immunity.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Conceptualization, B. R. and I. G.; methodology, B. R., I. G., and G. L.; software, I. G.; validation, B. R. and I. G.; formal analysis, I. G.; investigation, B. R. and I. G.; resources, I. G. and G. L.; data curation, B. R. and I. G.; writing—original draft preparation, B. R. and I. G.; writing—review and editing, B. R., I. G., and G.L.; visualization, B. R. and I. G.; supervision, I. G. and G. L.; project administration, I. G.; funding acquisition, I. G. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the project “GLYCARD: Glycosylation in Cardiovascular Diseases” (UIP-2019–04-5692), funded by the Croatian Science Foundation.

Declarations

Ethics approval

All donors (volunteers from Genos Ltd.) were familiarized with the aims of the study and samples were collected in accordance with Croatian regulations and after approval by the Genos Ethics Committee in accordance with the Declaration of Helsinki.

Conflict of interest

G. L. declares that he is the founder and owner of Genos Glycoscience Ltd., which offers a commercial service of glycomic analysis and has several patents in this field. I. G. is an employee of Genos Glycoscience Ltd. The other author declares no conflict of interest.