Trypanosoma cruzi ¹³C-labeled O-Glycan standards for mass spectrometry

Abstract

Trypanosoma cruzi is a protozoan parasite that causes Chagas disease, a debilitating condition that affects over 10 million humans in the American continents. In addition to its traditional mode of human entry via the “kissing bug” in endemic areas, the infection can also be spread in non-endemic countries through blood transfusion, organ transplantation, eating food contaminated with the parasites, and from mother to fetus. Previous NMR-based studies established that the parasite expresses a variety of strain-specific and developmentally-regulated O-glycans that may contribute to virulence. In this report, we describe five synthetic O-glycan analytical standards and show their potential to enable a more facile analysis of native O-glycan isomers based on mass spectrometry.

Introduction

Trypanosoma cruzi is the cause of Chagas disease, the highest impact infectious disease of the Americas with 10–20 million humans affected (Tarleton 2016). Although T. cruzi infection is generally controlled by host immune responses, it is rarely cleared, resulting in a chronic infection, with the potential development of cardiac and/or gastrointestinal manifestations in approximately 30% of infected individuals. Chagas disease is plagued by poor diagnostics, inadequate treatment options and the absence of vaccines (Prata 2001).

Trypanosoma cruzi expresses a broad repertoire of protein and lipid-linked glycans implicated in differentiation, cell invasion, and regulation of the host immune response. Substantial glycosylation changes likely support success in the sequential environments confronted during its life cycle (Tyler and Engman 2001). O-linked glycans are particularly abundant on mucins and mucin-associated proteins expressed on the parasite surface. Two major groups of mucin-like GPI-anchored glycoproteins are inversely expressed on the proliferative insect dwelling stage and infective cell-derived trypomastigotes (Buscaglia et al. 2006). The dense mucin arrays may protect epimastigotes from proteases in the insect intestine (Mortara et al. 1992), support survival of metacyclic trypomastigotes and their penetration of the gastric mucosal epithelium after ingestion by the mammalian host (Hoft et al. 1996), and protect bloodstream forms from complement attack (Tomlinson and Raper 1998). Mucins have also been implicated in T-cell suppression possibly through binding L-selectin (Argibay et al. 2002; Alcaide and Fresno 2004). Antibodies (Abs) against mucin-type O-glycans inhibit host-cell invasion (Yoshida et al. 1989; Ruiz et al. 1993). Terminal α3 Gal-linkages of infective trypomastigotes induce Chagasic anti-αGal Abs that can reduce parasite burden by a complement-independent mechanism (Gazzinelli et al. 1999), and may affect binding to host galectins (Pineda et al. 2014). Galf is a novel and immunogenic sugar of many pathogens that is also found on T. cruzi O-glycans (Oppenheimer et al. 2011). Sialic acids are applied to O-glycans and other glycans by a family of novel transialidases (Freire-de-Lima et al. 2015) and have been implicated in reduced and increased infectivity, attachment to and invasion of host cells, protection from anti-αGal Abs, modulation of recognition by host cell galectins, and cleavage of C3 convertases by serum enhancing factor. Finally, T. cruzi interacts with host cell galectins in complex ways that implicate its O-linked β-galactosides in mediating the host-pathogen relationship (Benatar et al. 2015; Poncini et al. 2015; da Silva et al. 2017).

Based on in-depth NMR-based studies (Mendonça-Previato et al. 2013), the mucin-type O-glycans consist of 2–5 sugars and are particularly diverse, strain-specific, and differentiation dependent. They are linked via α-linked GlcNAc by a mechanism that is homologous to mucin-type O-glycan assembly in animals (Heise et al. 2009). Existing data indicate that, depending on the parasite strain, β-galacto-pyranose (Galp) can be linked in the Golgi to the 3-, 4- and/or 6-positions of the GlcNAc residue, and can be extended by other 2/3/4/6-linked βGal residues, before addition of terminal α2,3-linked sialic acid residues via cell surface transialidase. βGlcp or αGalp can be linked to the 4-position of the core αGlcNAc. A recent glycoproteomics analysis identified ~100 O-glycopeptides from ~50 different proteins across the life cycle (Alves et al. 2017).

Mass spectrometry (MS) is an alternative method with the potential to comprehensively survey the O-glycome of small amounts of material, and MSⁿ approaches hyphenated with separation methods have great potential to characterize isomers with regard to anomericity, regiospecificity, and ring size and identity of the component sugars. However, chemically defined standards are needed to work out the methods. While certain T. cruzi O-glycans have been previously synthesized (Mendoza et al. 2010; Kashiwagi et al. 2012; Agustí et al. 2015; Schocker et al. 2016; Giorgi et al. 2017), the compounds are in limited supply and not generally available to the community. Here we characterize five chemo-enzymatically synthesized isomers that represent known T. cruzi mucin-type O-glycans and contain the main constituents known to be biosynthetically assembled in the parasite secretory pathway. It is anticipated that these standards will support new structure-function studies on the roles of glycans in parasite biology and virulence.

Results and discussion

The five glycans (Table I) were synthesized as alcohols, equivalent to their chemical form after release from glycoproteins using standard conditions of reductive β-elimination, by Omicron Biochemicals. To facilitate their characterization by MSⁿ, one of the βGal residues linked to the core αGlcNAc was uniformly-¹³C-labeled (U-¹³C-labeled). Each standard was analyzed by 1D and 2D NMR (Supplementary data, Fig. 1) and judged to be greater than 95% pure, confirming information supplied by the company (Table I). To enhance ionization and detection by mass spectrometry, the compounds were derivatized by permethylation and analyzed by Ion Trap MSⁿ fragmentation (Supplementary data, Fig. 2).

Table I.

Summary of O-glycan analytical standards

ID	Composition	SNFG/Oxford representation	Molecular formula	Molecular mass, Da (monoisotopic)	13C-Atom enrichment	Chemical Puritya
						HPLC	13C-NMR	1H-NMR
TRI-023	Galpα1,3[U-13C6]Galpβ1,4GlcNAcitol		13C6C14H37NO16	553.46	98%	95%	a	b
TET-023	Galpβ1,2Galpβ1,6([U-13C6]Galpβ1,4)GlcNAcitol		13C6C20H47NO21	715.60	99%	98%	a	b
TET-024	Galpβ1,2Galpβ1,6([U-13C6]Galpβ1,3)GlcNAcitol		13C6C20H47NO21	715.60	99%	>99%	a	b
TET-025	Galpβ1,2Galpβ1,6([U-13C6]Galfβ1,4)GlcNAcitol		13C6C20H47NO21	715.60	99%	>99%	a	a
TET-026	Galpβ1,3Galpβ1,6([U-13C6]Galfβ1,4)GlcNAcitol		13C6C20H47NO21	715.60	99%	>99%	a	a

^aProvided by the supplier: a = identity confirmed, no extraneous peaks detected; b = identify confirmed, no significant extraneous peaks detected.

To assess the potential for HPLC to resolve the isomers, a mixture of the five compounds was profiled on a standard C18 nLC platform typically used for proteomics studies. The base peak chromatogram showed near base-line separation in the range of 26–68 min (35–40% Mobile Phase B) (Fig. 1B). As expected, the trisaccharide eluted earliest, and the two Galf containing tetrasaccharides eluted substantially later than the two all-Galp tetrasaccharides. Within the Galf pair, baseline separation was achieved between isomers that differed only in the linkage of the non-reducing terminal βGalp, to the 2- or 3-position of the underlying β6Galp of the disaccharide branch. Within the other tetrasaccharide pair, near baseline separation was recorded for isomers that differed only in the linkage of the ¹³C-labeled lower arm βGalp to the 4- or 3-position of the core GlcNAc.

Fig. 1.

nLC-MS analysis of standard and T. cruzi O-glycans. (A)O-glycans were released from a preparation of T. cruzi mucins by reductive β-elimination, permethylated, and subject to nLC-MS on standard proteomics workstation utilizing a C18 column and a Thermo Velos Pro ITMS. Shown is the base-peak chromatogram of ions (500–2000 m/z) eluting from 12–98 min, with composition assignments based on exact monoisotopic mass ([M + Li]⁺¹ or [M + 2Li]⁺²). H, hexose; N, HexNAc, Sa, NeuNAc or sialic acid. (B) An equimolar mixture of the five standards were chromatographed in parallel. The structures are illustrated using a blend of the CFG and Oxford symbols, to highlight the linkage differences. (C) Extracted ion chromatogram of mono-lithiated ions from the data in panel A zooming from 20 to 70 min to show glycans with a composition of H2N1. All compositions were confirmed by MS², which differentiated linear from branched trisaccharides, as indicated. (D) Similar analysis to show glycans with an H3N1 composition. All compositions were confirmed by MS² (except for RT 57.97). (E) Similar analysis of the data in panel B for the standards. Vertical red dashed lines highlight positions of co-elution with the standards.

To examine the utility of the compounds for the analysis of T. cruzi O-glycans, an extract from epimastigotes was subjected to standard conditions of reductive β-elimination, permethylated, and chromatographed in parallel. All glycans that eluted prior to 90 min (separation range for the synthetic standards) could be assigned as ranging from a neutral disaccharide to an anionic tetrasaccharide and a neutral pentasaccharide, based on m/z-based composition analysis (annotated in Fig. 1A and summarized in Supplementary data, Table I). Other minor glycan species that eluted after 90 min included additional tetramer isomers, pentasaccharides and hexasaccharides (Supplementary data, Table I). Zooming in on the trisaccharide region revealed six isomers with a composition of H2N1 (Fig. 1C), none of which co-eluted with the trisaccharide standard (Fig. 1E), as expected because this structure is expressed at a different stage of the life cycle. MS² (not shown) indicated that two of the isomers were linear while the other four were branched. Zooming in on the tetrasaccharide region revealed eight isomers with a composition of H3N1, and each was branched based on diagnostic ions in MS² (not shown). Only the isomer eluting at 43.6 min is a potential match to a known standard, which has been described in the CL-Brener strain (Mendonça-Previato et al. 2013).

To investigate their potential to yield fragment ions for the analysis of unknown glycans, the standards were directly infused into a hybrid Ion Trap-Orbitrap mass spectrometer (Orbitrap Elite) using nanospray ionization in positive ion mode in the presence of LiOAc. The standards containing a β4-linked lower arm Gal yielded a unique cross-ring fragment ion in MS² that was not observed in the standard containing a β3-linked lower arm Gal (Supplementary data, Fig. 2A, C), whereas the latter yielded a characteristic Z-ion product in MS^2-4 experiments (Supplementary data, Fig. 2B). The standard containing a βGal 2-linked to the upper arm β6-linked Gal was differentiated by a unique fragment ion (m/z 95) in MS⁵ that was not observed in the isomer containing a non-reducing terminal β3Gal (Supplementary data, Fig. 2F), which was characterized by greater neutral losses of water and higher abundance of certain cross-ring cleavage products. Isomers that vary only in the pyranose or furanose configuration of the lower arm β3Gal were more difficult to differentiate by this method, though a characteristic fragment ion (m/z 128) was relatively more abundant for the latter isomer at the MS⁵ level (Supplementary data, Fig. 2G).

To further investigate the identity of the T. cruzi H3N1 isomer co-eluting at 43.6 min with Galpβ1,2-Galpβ1,6-[Galpβ1,4]-GlcNAcitol (TET-023), we analyzed a mixture of the T. cruzi sample and TET-023. As shown in Supplementary data 2, Fig. 3, the ¹³C-isomers co-elute precisely with the minor contaminating ¹²C-isomers present in the samples, indicating the utility of the ¹³C-isomer as an internal standard. MSⁿ analysis of the T. cruzi H3N1 (¹²C-labeled) ion in the mixture up to MS⁵ yielded fragment ions that were indistinguishable from that of the standard, indicating that it is identical to TET-023 (Supplementary data, Fig. 4).

These data establish the value for the glycan standards to differentiate, at the level of HPLC retention time and MSⁿ, many different isomers of T. cruzi O-glycans. They are expected to be highly useful for optimizing separation and assignment platforms not only for T. cruzi but for O-glycans from all species containing related isomers. The standards have been made possible by an NIH Common Fund grant, and aliquots are freely available to the scientific community. Additional information and requests for standards can be obtained at the website (https://ast.uga.edu/t-cruzi-o-glycans/) or by directly contacting the corresponding author.

Materials and methods

See the Supplementary data.

Funding

This work was supported in part by an National Institutes of Health Glycoscience Common Fund grant R21AI123161 to C.M.W., L.W. and Rick Tarleton (UGA), and other grants from the National Institutes of Health (R01GM111939 to L.W., P41GM103490 to L.W., P01GM107012 to L.W., and R01GM84383 to C.M.W.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Conflict of interest statement

None declared.