Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Leucine-rich alpha-2-glycoprotein (LRG) has been repeatedly proposed as a potential plasma biomarker for myelodysplastic syndrome (MDS).

OBJECTIVE:

The goal of our work was to establish the total LRG plasma level and LRG posttranslational modifications (PTMs) as a suitable MDS biomarker.

METHODS:

The total plasma LRG concentration was determined with ELISA, whilst the LRG-specific PTMs and their locations, were established using mass spectrometry and public mass spectrometry data re-analysis. Homology modelling and sequence analysis were used to establish the potential impact of PTMs on LRG functions via their impact on the LRG structure.

RESULTS:

While the results showed that the total LRG plasma concentration is not a suitable MDS marker, alterations within two LRG sites correlated with MDS diagnosis ( $p=$ 0.0011). Sequence analysis and the homology model suggest the influence of PTMs within the two LRG sites on the function of this protein.

CONCLUSIONS:

We report the presence of LRG proteoforms that correlate with diagnosis in the plasma of MDS patients. The combination of mass spectrometry, re-analysis of publicly available data, and homology modelling, represents an approach that can be used for any protein to predict clinically relevant protein sites for biomarker research despite the character of the PTMs being unknown.

Keywords

Myelodysplastic syndrome MDS leucine-rich alpha-2-glycoprotein LRG proteomics

1. Introduction

The physiological role of leucine-rich alpha-2-glycoprotein (LRG) is not yet fully understood. Studies have demonstrated the possible involvement of LRG in innate immune response [1]; LRG was proposed a potential marker of granulocytic differentiation [2]. Alterations in LRG serum expression have been observed in several malignant diseases, including lung 3̧ and pancreatic [4] cancers. In our previous work focusing on the proteomics of myelodysplastic syndrome (MDS) [5, 6, 7, 8], LRG has been proposed as a potential biomarker candidate especially of advanced MDS subgroups [9]. Our results suggested that the degree of LRG plasma alterations could be related to the severity of the disease. However, the available 2D electrophoretic data cannot directly identify the source of the changes (total plasma concentration or posttranslational modification (PTM) specific alterations). The increase in LRG plasma level could be an acceptable explanation as up-regulated LRG serum expression has also been observed for other malignant diseases [3, 4]. Moreover, western blot analysis for the RAEB-2 MDS subgroup suggested an increase in LRG concentration [8]. On the other hand, mass spectrometry-based relative quantification of selected LRG peptides (for RAEB-1 and RAEB-2 MDS subgroups) strongly indicated the presence of PTMs [6, 8].

In order to shed more light on the existing LRG related knowledge, we used mass spectrometry in combination with ELISA to map LRG total plasma concentration and its PTMs for different MDS patient subgroups, together with healthy controls. In addition, re-analysis of published mass spectrometry plasma proteomic data was performed to target the potential PTMs and their LRG positions. In order to estimate the potential impact of PTMs on LRG function via modifications in the LRG structure, we performed homology modelling and sequence analysis.

2. Methods

2.1 Plasma samples

Blood samples were collected as described previously [5]. In total, there were 101 different plasma samples used in the study: 71 samples obtained from MDS patients and 30 samples from healthy controls. The diagnoses of MDS subgroups were established according to the WHO classification criteria [10]. Samples were obtained and analysed in accordance with the Ethical Committee regulations of the Institute of Hematology and Blood Transfusion in Prague and with the Helsinki Declaration. All individuals tested agreed to participate in the study on the basis of informed consent.

2.2 Mass spectrometry

Relative label-free peptide quantification on the basis of Selected Ion Monitoring was performed as described in detail previously [6] using an HCT ultra ion-trap mass spectrometer with nanoelectrospray ionization (Bruker Daltonics, Bremen, Germany) coupled to a nanoLC system (UltiMate 3000; Dionex, Sunnyvale, CA, USA). Plasma samples were processed using modified protocols of the acetonitrile precipitation of plasma proteins and trypsin digestion as described by Kay et al. [11]. All LRG peptides were selected according to the PeptideAtlas [12]. Results were validated on the basis of the MS/MS spectra and retention times [13].

2.3 ELISA

Two different ELISA kits were used to establish the LRG total plasma concentration: SEB934Hu (Cloud-Clone Corp., Houston, TX, USA) and ELH-LRG1 (RayBiotech, Norcross, GA, USA). Samples were prepared and processed according to the manufacturers’ instructions.

2.4 Western blot

Western blot analysis was performed as described previously [14]. Monoclonal mouse anti-leucine-rich alpha-2-glycoprotein antibody (1: 500; Abcam ab57992) and rabbit anti-mouse IgG antibody conjugated with peroxidase (1: 80000; Sigma-Aldrich A9044) were used as a primary and secondary antibodies, respectively. The membranes were incubated for 60 minutes at 30 ${}^{\circ}$ C for both antibodies. A Spectra Multicolor Broad Range Protein Ladder (Thermo Scientific) was used to determine protein molecular weights.

2.5 LRG affinity isolation

LRG was isolated from blood and purified according to the protocol by Weivoda et al. [15].

2.6 Deglycosylation assay

An N-deglycosylation assay was performed according to Zielinska et al. [16] using PNGase F (G1549; Sigma-Aldrich, St. Louis, MO, USA).

Figure 1.

LRG plasma level quantified by ELISA. Box plot of LRG plasma level measured by monoclonal- (A) and polyclonal-based (B) ELISA kits in the control cohort (N) and MDS patient subgroups (RA-RARS, RCMD, RAEB-1, and RAEB-2).

2.7 Data re-analysis

In order to search for LRG posttranslational modifications, different published plasma mass spectrometry data sets (health and cancer samples, depleted and non-depleted plasma samples, using data-dependent and data-independent acquisition, etc.) were re- analysed [17, 18, 19, 20, 21]. The following software was used: AB Sciex MS Data Converter (Beta 1.3), DIA Umpire 2.1.3 [22], ProteoWizard 3.0 [23], Trans-Proteomic Pipeline 5.1.0 [24].

2.8 Sequence analysis

There were 119 mammalian reference sequences of the LRG protein (Supplement S1.1) downloaded from the NCBI database [25] and used for sequence analysis of orthologous proteins. Only the most usual splicing variant was considered for species with more than one splicing variant. For paralogue analysis, sequences designed as paralogues to LRG (ENST00000306390.7) protein in the Ensembl database [26] were used.

Sequences for both analyses were aligned with ClustalX [27] and sequence logos were created by the weblogo server [28]. Sequence analyses focused on the protein regions where PTMs were detected by mass spectrometry (E149 to K164 and V251 to R260).

2.9 Homology modelling

Homology modelling of human protein LRG (NP_443204) was performed with Modeller [29]. Alignment using crystal structures 2Z66_A [30] and 5A5C_A [31] as templates was prepared in ClustalX and adjusted manually. The best of the 25 models was chosen according to the Modeller objective function and stereochemical properties calculated with Procheck [32].

2.10 Statistical methods

For continuous data, we assessed statistical significance using tests and correlation for non-normally distributed data. Mann-Whitney U test was used to assess the differences in LRG concentrations. The results were analysed as follow: Spearman’s Rank-order correlation (for age and iron level), Mann-Whitney U test (for sex), and Kruskal-Wallis test (for diagnosis). ROC curve was used to estimate LRG diagnostic ability. Sample power calculations were performed (80% of power, $\alpha=$ 0.05) to demonstrate a difference of 20%. All statistical tests were two-sided, and the level of significance was set as 5% ( $p\leqslant$ 0.05). Statistical analysis was done in IBM SPSS statistics v23 and Mathworks MATLAB R2021a.

Table 1
Characterisation of LRG peptides for relative label-free quantification

Peptide sequence	Position	m/z	z	RT	PI	PI m/z
LQELHLSSNGLESLSPEFLRPVPQLR	94–119	987.2	3	31.1	b6	734.4
NALTGLPPGLFQASATLDTLVLK	126–148	1170.66	2	36.8	y17	1771
ENQLEVLEVSWLHGLK	149–164	947.51	2	33.0	y10	1181.7
ALGHLDLSGNR	165–175	576.81	2	20.0	y6	661.3
TLDLGENQLETLPPDLLR	192–209	1019.05	2	32.1	y6	710.4
DLLLPQPDLR	230–239	590.34	2	27.9	y6	725.4
YLFLNGNK	240–247	484.76	2	25.5	y3	318.2
VAAGAFQGLR	251–260	495.28	2	21.8	y8	819.45
DGFDISGNPWICDQNLSDLYR	292–312	1243.06	2	34.8	y13	1679.8

RT – retention time, PI – product ion.

Table 2

Relative LRG quantification results

Peptide sequence	Position	N	RA	RARS	RCMD	RAEB-1	RAEB-2
LQELHLSSNGLESLSPEFLRPVPQLR	94–119	1.0	1.3	1.5	1.5	1.7	2.4
NALTGLPPGLFQASATLDTLVLK	126–148	1.0	0.9	1.7	1.1	2.7	3.1
ENQLEVLEVSWLHGLK	149–164	1.0	3.1	1.4	4.6	33.1	50.0
ALGHLDLSGNR	165–175	1.0	1.4	1.9	1.4	2.1	3.9
TLDLGENQLETLPPDLLR	192–209	1.0	1.3	1.5	1.5	1.9	2.8
DLLLPQPDLR	230–239	1.0	1.4	1.9	1.6	3.0	4.3
YLFLNGNK	240–247	1.0	1.0	1.4	1.4	2.3	4.1
VAAGAFQGLR	251–260	1.0	7.0	1.6	5.3	16.7	30.8
DGFDISGNPWICDQNLSDLYR	292–312	1.0	1.7	1.6	1.8	3.3	4.9

Fold change vs control group (N) for RA, RARS, RCMD, RAEB-1, and RAEB-2 subgroups.

3. Results and discussion

The results indicated the decreased LRG concentration in MDS samples (Fig. 1A) with the lowest levels in the advanced MDS subgroups (mean/median values: 46.7/44.6, 47.5/47.7, 54.7/56.0, 35.8/38.8, and 38.5/37.7 $\mu$ g/ml for N, RA-RARS, RCMD, RAEB-1, and RAEB-2, respectively); the decrease was statistically significant for the advanced MDS subgroups ( $p=$ 0.0093, $p=$ 0.0158, and $p=$ 0.0016 for RAEB-1, RAEB-2, and advanced MDS samples, respectively). Interestingly, this observation was in contrast to previous results (obtained by MS, 2D SDS-PAGE, and western blot) indicating an increase in LRG plasma concentration. One of the possible explanations was that the ELISA kit was based on a monoclonal antibody (the presence of PTMs could hamper the binding of the antibody to its target LRG sequence). Indeed, the manufacturer confirmed using a monoclonal antibody as the capture antibody in the kit. We performed statistical analyses to search for potential relationships with cohort characteristics (risk/diagnosis, iron level for an MDS cohort together with age and sex for all donors). Interestingly, the results showed a correlation with diagnosis ( $p=$ 0.0011) independently of other parameters. Therefore, the results point out a potential MDS-related LRG proteoform (or a group of LRG proteoforms), unfortunately, the accurate LRG location of the PTM(s) is not known as the capture antibody was raised against the LRG sequence Val36–Glu197.

In order to quantify the total LRG plasma concentration, we consequently used a polyclonal antibody-based ELISA kit (ELH-LRG1). The results showed an increase in the LRG level in MDS samples (Fig. 1B) compared to the healthy controls (mean/median values: 22.6/21.9, 24.4/22.6, 32.4/30.5, 35.2/23.1, and 42.9/37.3 $\mu$ g/ml for N, RA-RARS, RCMD, RAEB-1, and RAEB-2, respectively), with the highest levels in the advanced MDS subgroups ( $p=$ 0.0172, and $p=$ 0.0084 for RAEB-2, and advanced MDS samples, respectively). These results were consistent with the previous observations obtained by MS, 2D SDS-PAGE, and western blot. Although the trend of increasing total LRG plasma level seems to follow the severity of MDS from the lowest levels in the control and low-risk MDS to advanced MDS samples, it does not seem to be a promising marker because of the large LRG level deviations among the samples. This assumption was confirmed by the ROC curve analysis – AUC did not exceed 0.5 for any of the groups studied (0.422, 0.297, 0.375, and 0.139 for RA-RARS, RCMD, RAEB-1, and RAEB-2, respectively). Thus, the effort (for a suitable MDS biomarker) should be focused on LRG proteoforms instead of its total plasma level.

To specify MDS-related LRG PTMs, we analysed plasma samples of different MDS subgroups by MS/MS. There were no specific LRG modifications found in the patient and control plasma samples; even using affinity isolated and purified LRG provided no clues. Western blot analysis of N-glycosylation also did not show any changes (Supplement S1.2). For this reason, we further performed relative label-free quantification of LRG peptides among all of the samples in order to map peptides that change their PTMs in MDS; peptides for this assay were selected using the PeptideAtlas [12] and previously collected spectra (Table 1). Two of the peptides were found to change markedly among the samples (251–260 VAAGAFQGLR, 149–164 ENQLEVLEVSWLHGLK), especially in advanced MDS subgroups (Table 2). Consequent statistical analysis showed a strong correlation with diagnosis for ENQLEVLEVSWLHGLK, indicating a relation of this peptide’s PTM(s) to MDS. There was no correlation with the iron level observed (as a possible source of oxidative PTMs). As these two LRG peptides emerged to be promising MDS related PTMs locations, we searched publicly available MS/MS data for possible LRG PTMs in order to find if these peptides are modified in other cancer-related or different control plasma or serum cohorts. ENQLEVLEVSWLHGLK was found in two datasets being modified (Trp to kynurenine); this PTM was not identified in our MDS dataset. Therefore, the results collected thus far showed the presence of unknown PTM(s) and two LRG sites whose alterations correlated with MDS diagnosis.

The physiological role of LRG is not yet fully understood, however, some findings have shown its role in apoptosis [33, 34, 35, 36]. This involvement could represent a (patho)physiological link of LRG to MDS as apoptosis is one of the hallmarks of myelodysplastic syndrome [37]. Since we were not able to detect the identity of LRG PTMs, but we did detect their locations, we focused on homology modelling and sequence analysis in order to establish their potential impact on LRG functions through influencing the LRG structure.

To determine the position of sequences of interest within the protein tertiary structure, a homology model was constructed (Supplemental Fig. 1A and B). In the absence of crystal structure of any LRG protein, a homology model had to be made first. To do so, crystal structure 5A5C_A (LRRs containing a domain of synthetic construct of murine LRRTM2 protein) was used as a template for amino acids 38 to 306 of human LRG because it has a high homology with LRG (34%) because it covers a long part of the modelled protein. We were looking for a crystal structure containing disulfide bond at the positions representing C43 and C56 in LRG because such disulfide bond in LRG is predicted [38]. Further, we were looking for a template with a high homology to LRG and the resolution of crystal structure was the decisive factor among structures fulfilling the abovementioned criteria. Based on this, crystal structure 2Z66 was considered the best template for the N-terminal region (amino acids 6 to 46 in human LRG) of human LRG. We tried to improve the model further using crystal structure 3RFE [39] as a template for its C-terminal part which did not lead to improvement, so this structure is not considered in the presented final model. Evaluating the homology model, one must be aware that LRG is a glycoprotein and that saccharides may have an influence on protein secondary structure. The structure of saccharides bound to LRG is, to the best of our knowledge, unknown. For this reason, we could not incorporate them into the model.

The presented structure is, as far as we know, the first homology model of human LRG protein made by some other than “black-box” server approach; the mouse LRG structure was predicted by Wang et al. [40]. The model shows a molecule of a solenoid shape with a 10-stranded parallel $\beta$ -sheet on its concave (inner) site and $\beta$ -turn as dominant secondary structure elements on the convex site. LRG contains 8 leucine-rich repeats (LRRs), 7 N-terminal contains 24 amino acid residues and the C-terminal LRR contains 28 amino acid residues. The additional 4 amino acid residues are inserted from the first $\beta$ -strand of the C terminal cap region, where they form a loop. The C-terminal cap contains, apart from the mentioned $\beta$ -strand that is incorporated into the central $\beta$ -sheet, $\alpha$ -helical structure (3 ${}_{10}$ helix (D304 to N306) followed by $\alpha$ -helix (L307 to Q317)) on its convex site and a $\beta$ -hairpin on the concave site of the molecule. There are three $\beta$ -strands in the N-terminal cap region of protein. While those N-terminal ones form a $\beta$ -hairpin, the C-terminal one participates in the central $\beta$ -sheet. There are two disulfide bridges in LRG, one in the N-terminal cap (C43 with C56) and the other one in the C-terminal cap region (C303 with C329). We did not pay any attention to the glycosylation of the protein that may have an impact on its structure, because of the unknown structure of any of the 6 saccharides. Similarly, the geometry of 6 N- and C-terminal amino acid residues is in question, although we expect these residues will be disordered. Both amino acid sequences bearing PTM are located mainly on the convex site of the protein although the N-terminal sequences reach part of the preceding LxxLxLxxNxL sequence motif by which the LRR is defined.

Sequence analyses were performed to identify conserved amino acids in the regions, where PTMs were identified by mass spectrometry. Alternations of conserved amino acids are believed to have a bigger impact on protein structure and/or functions than other amino acids. The only amino acid residue that is conserved among orthologues and paralogues of LRG in the region E149 to K164 is N150 (Supplemental Fig. 2A and B) that is a part of the LRR structure motif. There are hydrophobic residues conserved at positions 152, 155, 160 and 163 (but for threonine in LRRC52 at this position) among paralogues and L152, L160 and L163 conserved among orthologues. There is a hydrophobic amino acid conserved at position 155 and tryptophan (but for 3 species out of 119 species bearing glycine and 2 species bearing serine) conserved at position 159 among orthologues. This hints that W159 is important for LRG and that it is more likely involved in its function. Therefore, W159 modification should influence LRG properties and functions substantially. Interestingly, this amino acid was found to be modified (Trp to kynurenine) in data re-analysis as mentioned above.

The other region bearing PTM (V251 to R260; Supplemental Fig. 2C and D) is considerably less conserved than the previous one. There is no homology among paralogues. Sequence analysis of orthologues revealed there are conserved hydrophobic amino acids at positions 251, 256 and 259. The absence of conserved amino acid residues does not allow us to predict which residues are important for protein structure and which PTM can be of importance. For this reason, we do not discuss this region further. N150 is the only amino acid residue conserved among orthologues and paralogues of LRG. This amino acid is a part of the LxxLxLxxNxL sequence motif of LRR repeats and participated in formation of the $\beta$ -sheet.

4. Conclusions

In conclusion, we report the presence of LRG proteoforms that correlate with diagnosis in the plasma of MDS patients. Using mass spectrometry, data re-analysis and other techniques we specified two posttranslationally modified LRG sites with alterations related to MDS. Despite the unknown character of these modifications, homology modelling and sequence analyses showed their influence on LRG functions. The results show the potential impact of LRG alterations in MDS biomarker research and specify the area for future focus. The nature of the PTMs can be difficult to reveal due to several reasons; PTMs can be too unstable to be analysed by the MS method used, there can be combinations of several PTMs at the protein site (sequence) so that the specific proteoforms (PTMs combinations) are too low in abundance to be identified, the natural individual variability of PTMs may hamper the identification of biomarker relevant PTM(s), etc. Using the approach that combines mass spectrometry, data re-analysis of publicly available data, and homology modelling can be used for any protein to predict clinically relevant protein sites for marker research even though the character of PTMs is unknown.

Author contributions

Conception: PM.

Interpretation or analysis of data: PM, ZS, KP, JS, ZG, PP, VI, JED.

Preparation of the manuscript: PM, ZS, KP, PP, VI, JED.

Revision for important intelectual content: PM, ZS, KP, PP, VI.

Supervision: PM.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.3233/CBM-210033.

sj-docx-1-cbm-10.3233_CBM-210033.docx - Supplemental material

Supplemental material, sj-docx-1-cbm-10.3233_CBM-210033.docx

Footnotes

Acknowledgments

This work was supported by the Czech Science Foundation (grant number 20-10845S) and by the Ministry of Health of the Czech Republic project for the conceptual development of the research organization (Institute of Hematology and Blood Transfusion, 00023736).

Conflict of interest

The authors declare no conflict of interest.

References

Shirai

Hirano

Ohkura

Ikeda

and Inoue

, Up-regulation of the expression of leucine-rich α2-glycoprotein in hepatocytes by the mediators of acute-phase response, Biochem. Biophys. Res. Commun 382 (2009), 776–779. doi: 10.1016/j.bbrc.2009.03.104.

O’Donnell

L.C.

Druhan

L.J.

and Avalos

B.R.

, Molecular characterization and expression analysis of leucine-rich α2-glycoprotein, a novel marker of granulocytic differentiation, J. Leukoc. Biol 72 (2002), 478–485.

Okano

Kondo

Kakisaka

Fujii

Yamada

Kato

Nishimura

Gemma

Kudoh

and Hirohashi

, Plasma proteomics of lung cancer by a linkage of multi-dimensional liquid chromatography and two-dimensional difference gel electrophoresis, Proteomics 6 (2006), 3938–3948. doi: 10.1002/pmic.200500883.

Kakisaka

Kondo

Okano

Fujii

Honda

Endo

Tsuchida

Aoki

Itoi

Moriyasu

Yamada

Kato

Nishimura

Todo

and Hirohashi

, Plasma proteomics of pancreatic cancer patients by multi-dimensional liquid chromatography and two-dimensional difference gel electrophoresis (2D-DIGE): Up-regulation of leucine-rich alpha-2-glycoprotein in pancreatic cancer, J. Chromatogr. B Anal. Technol. Biomed. Life Sci 852 (2007), 257–267. doi: 10.1016/j.jchromb.2007.01.029.

Májek

Reicheltová

Suttnar

Čermák

and Dyr

J.E.

, Plasma proteome changes associated with refractory cytopenia with multilineage dysplasia, Proteome Sci 9 (2011). doi: 10.1186/1477-5956-9-64.

Májek

Reicheltová

Suttnar

Čermák

and Dyr

J.E.

, Plasma protein alterations in the refractory anemia with excess blasts subtype 1 subgroup of myelodysplastic syndrome, Proteome Sci 10 (2012). doi: 10.1186/1477-5956-10-31.

Májek

Riedelová-Reicheltová

Suttnar

Pečánková

Čermák

and Dyr

J.E.

, Plasma proteome changes associated with refractory anemia and refractory anemia with ringed sideroblasts in patients with myelodysplastic syndrome, Proteome Sci 11 (2013). doi: 10.1186/1477-5956-11-14.

Majek

Riedelova-Reicheltova

Suttnar

Pecankova

Cermak

and Dyr

J.E.

, Proteome changes in the plasma of myelodysplastic syndrome patients with refractory anemia with excess blasts subtype 2, Dis. Markers 2014 (2014). doi: 10.1155/2014/178709.

Majek

Pecankova

Cermak

and Dyr

J.E.

, Plasma protein biomarker candidates for myelodysplastic syndrome subgroups, Biomed Res. Int 2015 (2015). doi: 10.1155/2015/209745.

10.

Vardiman

J.W.

Harris

N.L.

and Brunning

R.D.

, The World Health Organization (WHO) classification of the myeloid neoplasms, Blood 100 (2002), 2292–2302. doi: 10.1182/blood-2002-04-1199.

11.

Kay

Barton

Ratcliffe

Matharoo-Ball

Brown

Roberts

Teale

and Creaser

, Enrichment of low molecular weight serum proteins using acetonitrile precipitation for mass spectrometry based proteomic analysis, Rapid Commun. Mass Spectrom 22 (2008), 3255–3260. doi: 10.1002/rcm.3729.

12.

Desiere

Deutsch

King

E.W.

Nesvizhskii

N.L.

Mallick

A.I.

Eng

Chen

Eddes

Loevenich

and Aebersold

S.N.

, The PeptideAtlas project, Nucleic Acids Res 34 (2006), 655–658.

13.

Lange

Picotti

Domon

and Aebersold

, Selected reaction monitoring for quantitative proteomics: A tutorial, Mol. Syst. Biol 4 (2008). doi: 10.1038/msb.2008.61.

14.

Májek

Reicheltová

Suttnar

Malý

Oravec

Pečánková

and Dyr

J.E.

, Plasma proteome changes in cardiovascular disease patients: Novel isoforms of apolipoprotein A1, J. Transl. Med 9 (2011). doi: 10.1186/1479-5876-9-84.

15.

Weivoda

Andersen

J.D.

Skogen

Schlievert

P.M.

Fontana

Schacker

Tuite

Dubinsky

J.M.

and Jemmerson

, ELISA for human serum leucine-rich alpha-2-glycoprotein-1 employing cytochrome c as the capturing ligand, J. Immunol. Methods 336 (2008), 22–29. doi: 10.1016/j.jim.2008.03.004.

16.

Zielinska

D.F.

Gnad

Wiśniewski

J.R.

and Mann

, Precision Mapping of an In Vivo N-Glycoproteome Reveals Rigid Topological and Sequence Constraints, Cell 141 (2010), 897–907. doi: 10.1016/j.cell.2010.04.012.

17.

Zhao

Chang

Qin

Cao

Tian

Jiang

Zhu

Ying

and Qian

, Mining the human plasma proteome with three-dimensional strategies by high-resolution Quadrupole Orbitrap Mass Spectrometry, Anal. Chim. Acta 904 (2016), 65–75. doi: 10.1016/j.aca.2015.11.001.

18.

Geyer

P.E.

Kulak

N.A.

Pichler

Holdt

L.M.

Teupser

and Mann

, Plasma proteome profiling to assess human health and disease, Cell Syst 2 (2016), 185–195. doi: 10.1016/j.cels.2016.02.015.

19.

Liu

Buil

Collins

B.C.

Gillet

L.C.

Blum

L.C.

Cheng

L.-Y.

Vitek

Mouritsen

Lachance

Spector

T.D.

Dermitzakis

E.T.

and Aebersold

, Quantitative variability of 342 plasma proteins in a human twin population, Mol. Syst. Biol 11 (2015), 786–786. doi: 10.15252/msb.20145728.

20.

Bjelosevic

Pascovici

Ping

Karlaftis

Zaw

Song

Molloy

M.P.

Monagle

and Ignjatovic

, Quantitative Age-specific Variability of Plasma Proteins in Healthy Neonates, Children and Adults, Mol. Cell. Proteomics 16 (2017), 924–935. doi: 10.1074/mcp.M116.066720.

21.

Park

G.W.

Kim

J.Y.

Hwang

Lee

J.Y.

Ahn

Y.H.

Lee

H.K.

E.S.

Kim

K.H.

Jeong

H.K.

Yun

K.N.

Kim

Y.-S.

J.-H.

H.J.

Kim

J.H.

Paik

Y.-K.

and Yoo

J.S.

, Integrated GlycoProteome Analyzer (I-GPA) for Automated Identification and Quantitation of Site-Specific N-Glycosylation, Sci. Rep 6 (2016), 21175. doi: 10.1038/srep21175.

22.

Tsou

C.-C.

Avtonomov

Larsen

Tucholska

Choi

Gingras

A.-C.

and Nesvizhskii

A.I.

, DIA-Umpire: Comprehensive computational framework for data-independent acquisition proteomics, Nat. Methods 12 (2015), 258–264. doi: 10.1038/nmeth.3255.

23.

Chambers

M.C.

Maclean

Burke

Amodei

Ruderman

D.L.

Neumann

Gatto

Fischer

Pratt

Egertson

Hoff

Kessner

Tasman

Shulman

Frewen

Baker

T.A.

Brusniak

M.-Y.

Paulse

Creasy

Flashner

Kani

Moulding

Seymour

S.L.

Nuwaysir

L.M.

Lefebvre

Kuhlmann

Roark

Rainer

Detlev

Hemenway

Huhmer

Langridge

Connolly

Chadick

Holly

Eckels

Deutsch

E.W.

Moritz

R.L.

Katz

J.E.

Agus

D.B.

MacCoss

Tabb

D.L.

and Mallick

, A cross-platform toolkit for mass spectrometry and proteomics, Nat. Biotechnol 30 (2012), 918–920. doi: 10.1038/nbt.2377.

24.

Deutsch

E.W.

Mendoza

Shteynberg

Slagel

Sun

and Moritz

R.L.

, Trans-Proteomic Pipeline, a standardized data processing pipeline for large-scale reproducible proteomics informatics, PROTEOMICS – Clin. Appl 9 (2015), 745–754. doi: 10.1002/prca.201400164.

25.

Sayers

E.W.

Barrett

Benson

D.A.

Bolton

Bryant

S.H.

Canese

Chetvernin

Church

D.M.

Dicuccio

Federhen

Feolo

Geer

L.Y.

Helmberg

Kapustin

Landsman

Lipman

D.J.

Madden

T.L.

Madej

Maglott

D.R.

Marchler-Bauer

Miller

Mizrachi

Ostell

Panchenko

Pruitt

K.D.

Schuler

G.D.

Sequeira

Sherry

S.T.

Shumway

Sirotkin

Slotta

Souvorov

Starchenko

Tatusova

T.A.

Wagner

Wang

John Wilbur

Yaschenko

and Ye

, Database resources of the National Center for Biotechnology Information, Nucleic Acids Res 38 (2010), D5–16. https://www.scopus.com/inward/record.uri?eid=2-s2.0-75549088687&partnerID=40&md5=f98616465e8b13c85b7095a8f1160876.

26.

Cunningham

Achuthan

Akanni

Allen

Amode

M.R.

Armean

I.M.

Bennett

Bhai

Billis

Boddu

Cummins

Davidson

Dodiya

K.J.

Gall

Girón

C.G.

Gil

Grego

Haggerty

Haskell

Hourlier

Izuogu

O.G.

Janacek

S.H.

Juettemann

Kay

Laird

M.R.

Lavidas

Liu

Loveland

J.E.

Marugán

J.C.

Maurel

McMahon

A.C.

Moore

Morales

Mudge

J.M.

Nuhn

Ogeh

Parker

Parton

Patricio

Abdul Salam

A.I.

Schmitt

B.M.

Schuilenburg

Sheppard

Sparrow

Stapleton

Szuba

Taylor

Threadgold

Thormann

Vullo

Walts

Winterbottom

Zadissa

Chakiachvili

Frankish

Hunt

S.E.

Kostadima

Langridge

Martin

F.J.

Muffato

Perry

Ruffier

Staines

D.M.

Trevanion

S.J.

Aken

B.L.

Yates

A.D.

Zerbino

D.R.

and Flicek

, Ensembl 2019, Nucleic Acids Res 47 (2019), D745–D751. doi: 10.1093/nar/gky1113.

27.

Thompson

J.D.

Gibson

T.J.

and Higgins

D.G.

, Multiple sequence alignment using ClustalW and ClustalX, Curr. Protoc. Bioinformatics Chapter 2 (2002). https://www.scopus.com/inward/record.uri?eid=2-s2.0-54049108449&partnerID=40&md5=c19930feaf659746d230a91741c5897e.

28.

Crooks

G.E.

Hon

Chandonia

J.-M.

and Brenner

S.E.

, WebLogo: A sequence logo generator, Genome Res 14 (2004), 1188–1190. doi: 10.1101/gr.849004.

29.

Šali

and Blundell

T.L.

, Comparative protein modelling by satisfaction of spatial restraints, J. Mol. Biol 234 (1993), 779–815. doi: 10.1006/jmbi.1993.1626.

30.

Kim

H.M.

Park

B.S.

Kim

J.-I.

Kim

S.E.

Lee

S.C.

Enkhbayar

Matsushima

Lee

Yoo

O.J.

and Lee

J.-O.

, Crystal Structure of the TLR4-MD-2 Complex with Bound Endotoxin Antagonist Eritoran, Cell 130 (2007), 906–917. doi: 10.1016/j.cell.2007.08.002.

31.

Paatero

Rosti

V Shkumatov

Sele

Brunello

Kysenius

Singha

Jokinen

Huttunen

and Kajander

, Crystal Structure of an Engineered LRRTM2 Synaptic Adhesion Molecule and a Model for Neurexin Binding, Biochemistry 55 (2016), 914–926. doi: 10.1021/acs.biochem.5b00971.

32.

Laskowski

R.A.

MacArthur

M.W.

and Thornton

J.M.

, Chapter 21.4 PROCHECK: validation of protein-structure coordinates, in: Crystallogr. Biol. Macromol. Second Online Ed., 2012.

33.

Cummings

Walder

Treeful

and Jemmerson

, Serum leucine-rich alpha-2-glycoprotein-1 binds cytochrome c and inhibits antibody detection of this apoptotic marker in enzyme-linked immunosorbent assay, Apoptosis 11 (2006), 1121–1129. doi: 10.1007/s10495-006-8159-3.

34.

Codina

Vanasse

Kelekar

Vezys

and Jemmerson

, Cytochrome c-induced lymphocyte death from the outside in: Inhibition by serum leucine-rich alpha-2-glycoprotein-1, Apoptosis 15 (2010), 139–152. doi: 10.1007/s10495-009-0412-0.

35.

Xiao

and Zhu

, Leucine-rich alpha-2-glycoprotein1gene interferes with regulation of apoptosis in leukemia KASUMI-1 cells, Med. Sci. Monit 24 (2018), 8348–8356. doi: 10.12659/MSM.911249.

36.

Takemoto

Serada

Fujimoto

Honda

Ohkawara

Takahashi

Nomura

Inohara

and Naka

, Leucine-rich α-2-glycoprotein promotes TGFβ1-mediated growth suppression in the Lewis lung carcinoma cell lines, Oncotarget 6 (2015), 11009–11022. doi: 10.18632/oncotarget.3557.

37.

Greenberg

P.L.

, Apoptosis and its role in the myelodysplastic syndromes: Implications for disease natural history and treatment, Leuk. Res 22 (1998), 1123–1136. doi: 10.1016/S0145-2126(98)00112-X.

38.

Bateman

Martin

M.J.

O’Donovan

Magrane

Alpi

Antunes

Bely

Bingley

Bonilla

Britto

Yeh

L.-S.

and Zhang

, UniProt: The universal protein knowledgebase, Nucleic Acids Res 45 (2017), D158–D169. doi: 10.1093/nar/gkw1099.

39.

McEwan

P.A.

Yang

Carr

K.H.

Zheng

and Emsley

, Quaternary organization of GPIb-IX complex and insights into Bernard-Soulier syndrome revealed by the structures of GPIβ and a GPIβ/GPIX chimera, Blood 118 (2011), 5292–5301. doi: 10.1182/blood-2011-05-356253.

40.

Wang

Abraham

McKenzie

J.A.G.

Jeffs

Swire

Tripathi

V.B.

Luhmann

U.F.O.

Lange

C.A.K.

Zhai

Arthur

H.M.

Moss

S.E.

and Greenwood

, LRG1 promotes angiogenesis by modulating endothelial TGF-β signalling, Nature 499 (2013), 306–311. doi: 10.1038/nature12345.

Mass spectrometry,data re-analysis,and homology modelling predict posttranslational modifications of leucine-rich alpha-2-glycoprotein as a marker of myelodysplastic syndrome

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Methods

2.1 Plasma samples

2.2 Mass spectrometry

2.3 ELISA

2.4 Western blot

2.5 LRG affinity isolation

2.6 Deglycosylation assay

2.8 Sequence analysis

2.9 Homology modelling

2.10 Statistical methods

Table 1 Characterisation of LRG peptides for relative label-free quantification

4. Conclusions

Author contributions

Supplementary data

sj-docx-1-cbm-10.3233_CBM-210033.docx - Supplemental material

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Characterisation of LRG peptides for relative label-free quantification