Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 20.
Published in final edited form as: ACS Chem Biol. 2021 Dec 29;17(11):3059–3068. doi: 10.1021/acschembio.1c00932

In-depth Site-specific O-glycosylation Analysis of Glycoproteins and Endogenous Peptides in Cerebrospinal Fluid (CSF) from Healthy Individuals, Mild Cognitive Impairment (MCI), and Alzheimer’s Disease (AD) Patients

Zhengwei Chen 1,#, Danqing Wang 1,#, Qing Yu 2, Jillian Johnson 2, Richard Shipman 3, Xiaofang Zhong 2, Junfeng Huang 2, Qinying Yu 2, Henrik Zetterberg 4,5,6,7, Sanjay Asthana 8, Cynthia Carlsson 8, Ozioma Okonkwo 8, Lingjun Li 1,2,*
PMCID: PMC9240109  NIHMSID: NIHMS1771217  PMID: 34964596

Abstract

Site-specific O-glycoproteome mapping in complex biological system provides a molecular basis for understanding the structure-function relationships of glycoproteins and their roles in physiological and pathological processes. Previous O-glycoproteome analysis in cerebrospinal fluid (CSF) focused on sialylated glycoforms, while missing information of other glycosylation types. In order to achieve an unbiased O-glycosylation profiling, we developed an integrated strategy combining universal boronic acid enrichment, high-pH fractionation, and electron-transfer and higher-energy collision dissociation (EThcD) for enhanced intact O-glycopeptide analysis. This strategy has also been applied to analyze O-glycoproteome in CSF, resulting in the identification of 308 O-glycopeptides from 110 O-glycoproteins, covering both sialylated and non-sialylated glycoforms. To our knowledge, this is the largest dataset of O-glycoproteins and O-glycosites reported for CSF so far. We also developed a peptidomics workflow that utilized the EThcD and a three-step database searching strategy for comprehensive PTM analysis of endogenous peptides including both N-glycosylation and O-glycosylation and other common peptide PTMs. Interestingly, among the 1411 endogenous peptides identified, 89 of them were O-glycosylated and only one N-glycosylated peptide was found, indicating CSF endogenous peptides were predominantly O-glycosylated. O-glycoproteome and endogenous peptidome PTM analysis were also conducted in MCI and AD patients to provide a landscape of glycosylation pattern in different disease states. Our results showed a decreasing trend in fucosylation and increasing trend of endogenous peptide O-glycosylation, which may play an important role in AD progression.

Graphical Abstract

graphic file with name nihms-1771217-f0008.jpg

INTRODUCTION

Human cerebrospinal fluid (CSF) is predominantly produced in choroid plexuses, circulates within the ventricles of the brain, and surrounds the brain and spinal cord.1 The functional role of CSF includes mechanical protection of the central nervous system (CNS), homeostasis of the interstitial fluid in the brain, and regulation of neuronal functioning.2,3 Through direct contact with CNS, the CSF reflects the ongoing physiological or pathological state of CNS.4,5 Metabolites, peptides, proteins, enzymes, and hormones in CSF are involved in many biological processes, and changes in these compositions are viewed as a sign of pathological alterations in CNS. These biological compositional changes provide an opportunity to mine the CSF for biomarker discovery in neurological diseases.6

As one of the most complex protein post-translational modifications (PTMs), glycosylation serves as a key regulatory mechanism controlling protein folding, molecular trafficking, cell adhesion, receptor activation and signal transduction.7–10 Based on the amino acids that glycans attach to, glycosylation can be classified into two major categories: N-glycosylation and O-glycosylation. Biosynthesis of N-glycosylation is initiated by transferring a pre-assembled 14 monosaccharide complex glycan to asparagine residue (Asn) within the consensus motif (Asn-X-Ser/Thr, X≠P) followed by sequential addition or removal of certain monosaccharide in a well-defined process.11,12 In contrast, O-glycosylation synthesis involves the attachment of a single monosaccharide to the serine/threonine (Ser/Thr) residue of a polypeptide without any definable peptide consensus motif and subsequent attachment of numerous diverse monosaccharide residues. As a result, a higher degree of site occupancy, structural heterogeneity, and diversity have been observed in O-glycosylation.

O-glycosylation is further categorized into mucin and non-mucin types, according to the monosaccharide residue directly linked to polypeptides. The attached monosaccharide residue is N-acetylgalactosamine (GalNAc/HexNAc) in the mucin type, whereas the attached residue can be N-acetylglucosamine (GlcNAc/HexNAc), fucose (Fuc), galactose (Gal/Hex), mannose, glucose in the non-mucin type.11 Mucin type O-glycosylation on the cell surface and on secreted proteins has also been shown to modulate recognition, adhesion, communication between cells, and the cell’s surrounding environments.13 As a nutrient- and stress-responsive modification, non-mucin type O-GlcNAcylation is extensively involved in the spatiotemporal regulation of diverse cellular processes, including transcription, epigenetic modifications and cell signaling dynamics.14 Apart from modifying proteins, O-glycosylation can also happen on the endogenous peptides, including neurotransmitters and hormones. In fact, it has been well recognized that after initial peptidase cleavages, endogenous peptides can undergo further post-translational modifications such as amidation, acetylation, phosphorylation, sulfation, and N-/O-glycosylation. As an example, it has been reported that there was an extensive N-/O-glycosylation of gonadotropin, which is a glycoprotein polypeptide hormone.15 Recently, our group reported O-glycosylation on mouse insulin-1B and −2B chains and human insulin-B chain, as well as multiple O-glycoforms of signaling peptides.16 We also discovered 14 O-glycosylated neuropeptides in the crustacean nervous system.17 These findings highlight the crucial role of O-glycosylation on neurotransmitters and peptide hormones, and the necessity to characterize the glycosylation state of endogenous peptides.

Though the CSF proteome and endogenous peptidome have been extensively characterized, there are only few reports on CSF O-glycosylation. One study reported 39 O-glycopeptides from 22 CSF O-glycoproteins,18 and another study identified 43 O-glycopeptides from 28 CSF O-glycoproteins.19 By utilizing an enrichment approach based on oxidation of sialic acid and hydrazide chemistry, both studies lost the information of sialic acid and characterized only a subset of O-glycoproteome. As a result, the current O-glycoproteome depth of human CSF is rather limited, hindering the design of studies to explore more disease-related O-glycosylation alterations. On the other hand, CSF is a valuable source for studying neurodegenerative diseases, with previous studies exploring O-glycosylation changes in CSF samples from Alzheimer’s disease (AD) patients. However, they generally targeted at some well-known glycoproteins such as amyloid precursor protein (APP) or apolipoprotein E (APOE).20,21 A comprehensive study at a system-wide scale used to explore the O-glycosylation changes in AD state is still needed. For endogenous peptidome study in CSF, one study identified 730 endogenous peptides, including 138 peptides with PTMs such as acetylation, amidation, phosphorylation, Gln to pyro-Glu conversion.22 However, none of the identified peptides were reported to be glycosylated. In another study with 563 endogenous peptides identified, the presence of glycan oxonium ions was observed in some spectra, and 28 O-glycopeptides were eventually identified with lower collision energy during fragmentation.23 Nonetheless, only two O-glycan compositions were found. It is highly possible that there are many more O-glycosylated endogenous peptides undiscovered. Therefore, an approach with a glycosylation-centered analysis workflow needs to be developed to systematically evaluate N-/O- glycosylation on endogenous peptides.

In the present study, we optimized the boronic acid-based enrichment strategy to efficiently enrich both sialylated and non-sialylated O-glycopeptides. High-pH (HpH) fractionation was performed to further improve O-glycoproteome coverage. With EThcD fragmentation, site-specific information was obtained for intact O-glycopeptide characterization. The optimized approach was then applied to O-glycoproteome analysis in CSF from healthy individuals, mild cognitive impairment (MCI) and AD patients. For endogenous peptide analysis, we developed a peptidomics workflow that combined CSF endogenous peptide extraction by 10kDa molecular weight cut-off (MWCO), EThcD fragmentation, and a three-step database searching strategy for comprehensive PTM analysis. This workflow was further adopted to study the endogenous CSF peptides in MCI and AD patients. Overall, we believe that the developed analytical strategies in this study are readily applicable for site-specific O-glycosylation analysis of both glycoproteins and endogenous peptides in other complex biological systems.

RESULTS AND DISCUSSION

Optimization of boronic acid enrichment strategy for O-glycoproteomic analysis.

To avoid the interference of highly abundant non-glycosylated peptides, enrichment is a key step to enhance glycopeptide detection. Common enrichment techniques include lectin affinity enrichment chromatography and hydrophilic interaction liquid chromatography (HILIC).24 However, each enrichment method possesses inherent limitations. Lectin only enriches specific type of glycans, while HILIC-based methods usually suffer from interference from many hydrophilic non-glycopeptides and is also complicated by the difference in hydrophilicity between O-glycopeptides and N-glycopeptides.25 Instead, boronic acid enrichment is a more universal method for enriching both N-glycopeptides and O-glycopeptides because the cis-diol groups on glycans can react with boronic acids to form reversible covalent bond. This can later be released under acidic conditions without any side effects.26,27 Despite its application to a large-scale N-glycoproteome study28 and a small-scale analysis of O-GlcNAcylation29, boronic acid enrichment has not yet been used for comprehensive O-glycoproteome analysis of complex biological samples.

Here, we used phenylboronic acid (PBA) solid phase extraction cartridge to extract the O-glycopeptides from complex tryptic digests. PNGase F was first used to remove the N-glycans to avoid interference from N-glycopeptides during enrichment and MS detection of O-glycopeptides. Starting from 200 μg tryptic peptides from PANC1 cells, a total of 213 intact O-glycopeptides were identified right after enrichment in three technical replicates and 100 O-glycopeptides could be identified in a single replicate on average (Figure 1a and 1c). To minimize the interference from co-enriched non-glycosylated peptides, we performed off-line HpH fractionation, which has shown high separation orthogonality to low-pH reversed-phase liquid chromatography coupled with MS analysis.30 229 intact O-glycopeptides were identified from seven HpH fractions in one replicate, representing nearly two times of the number of identifications in the direct analysis (Figure S1a, Supplemental Table S2). A closer look at the distribution of the number of O-glycopeptides among 7 fractions showed that O-glycopeptides were mainly in the first four fractions (Figure S1b). This could be explained by the hydrophilicity of O-glycopeptides that shifts the elution to an earlier time frame on a C18 column. To better utilize instrument time and run more replicates, the first fraction was combined with the last three fractions in another round of experiment. It is well-known that two or more technical replicates are needed to get the maximum coverage of peptides due to the randomness and stochastic sampling of data-dependent analysis (DDA) mode, which is more significant in glycoproteomic analyses.31–33 Therefore, we evaluated the glycoproteome coverage within the context of multiple replicates. Among all O-glycopeptides identified with four repetitive injections, two replicates only yielded less than 60% cumulative coverage, and at least three technical replicates were needed to reach a coverage of more than 90% of the total identified glycopeptides (Figure 1b). As shown in Figure 1c, prefractionation improved the glycoproteome coverage by four times compared to direct analysis with the same number of replicates. When 4 replicates were analyzed, the total identification number of O-glycopeptide could further reach up to 987, representing nearly a 5-fold increase.

Figure 1.

Figure 1.

Open in a new tab

Boronic acid enrichment optimization. a) Comparison of O-glycoproteome coverage between direct analysis and prefractionation with four fractions. (F: fractions); b) The effects of technical replicates on the overall coverage; c) Comparison of O-glycoproteome coverage between direct analysis, prefractionation with 3 technical replicates and prefractionation with 4 technical replicates.

Site-specific O-glycoproteome analysis in CSF.

Benefiting from the optimized O-glycopeptide enrichment strategy and site-specific information provided by EThcD fragmentation (as discussed in the Supporting Information), we identified 308 intact O-glycopeptides and 292 unique O-glycoforms from 181 O-glycosites and 110 O-glycoproteins in CSF from healthy individuals, which was a large increase compared to previous reports (Figure 2a). Overall, the majority of O-glycoproteins (72%) carried only a single O-glycosite, 25% carried 2-3 O-glycosites and a low percentage (5%) of glycoproteins were found to have more than 4 glycosites (Figure 2b). There were around 30% O-glycosites identified with more than 2 O-glycans, demonstrating the microheterogeneity of O-glycosylation (Figure 2c).

Figure 2.

Figure 2.

Open in a new tab

a) The O-glycoproteome mapping in CSF from healthy individuals; b) The distribution of the number of O-glycosites per O-glycoproteins; c) The distribution of the number of O-glycoforms per O-glycosite; and d) Number of O-glycoforms with 1, 2 or 3 sialic acid or fucose residues.

Among different kinds of O-glycoforms, mucin-type core 1 (Galβ1→3GalNAc) is known to be the major component of O-glycans.34 In the present study, around 48.3% of O-glycoforms had core 1 glycoforms with a large portion of them sialylated (72.9%). This agreed well with the core 1 O-glycoform percentage in human serum (46.4%) as well as the dominant composition of sialylated core 1 glycoforms (~67%), because more than 80% of CSF proteins originated from the plasma filtrate and contained a similar glycosylation pattern.35,36 Interestingly, as another capping unit to elongate glycan branches, only 5.1% core 1 O-glycoforms were fucosylated and total fucosylated O-glycoforms only accounted for 29.1%, which was much lower than sialylated forms. As shown in Figure 2d, the number of O-glycoforms with 2 sialic acids far exceeded the number of O-glycoforms with 2 fucoses, while the number of O-glycoforms with 1 or 3 sialic acids/fucoses were similar.

Though there is no consensus motif for O-glycosylation, the role of adjacent proline residues is extensively studied, as it occurs frequently near the glycosylation site and may facilitate the process of protein glycosylation.37–39 Here, we performed a proline frequency analysis of the ±10 residues surrounding the 181 identified O-glycosites. Though such proline frequency analysis results may differ depending on the origins of the O-glycosite dataset and different tissues used, we did observe that the proline residue exhibited conserved sequence at the −1 and +3 positions in CSF (Figure 3a, c), which was consistent with previous global statistical analysis39 and reports in CSF.19 We also conducted the proline frequency analysis of the experimentally verified 435 O-glycosites from Uniprot (2017_12). The results revealed that the highest frequency of proline was from −4 to +7 positions, except for the +1 position right next to threonine (Figure 3b, d). The conserved proline frequency was most pronounced at −1 and +3 positions in accordance with our findings. This unique conserved sequence information may represent CSF-specific features that resulted from various factors including: the origins of the glycoproteins, different biosynthesis routes of these glycoproteins, glycotransferase enzyme activities, etc. Another CSF-specific feature was the higher frequency of glutamic acid (E) adjacent to the O-glycosite, compared with the results from global O-glycosite database, especially at +1 and +2 positions. A previous study showed that O-glycosylation was markedly reduced with glutamic acid residue substituted at positions −1 and +3, while glutamic acid replacement at +1 and +2 had no such effect.40 This may explain a higher frequency of glutamic acid at +1 and +2 positions, compared with other positions.

Figure 3.

Figure 3.

Open in a new tab

Weblogo probability plots and proline frequency analysis. a) Weblogo plot for 181 identified O-glycosites in the current study; b) Weblogo plot for 435 O-glycosites from Uniprot; c) Proline frequency analysis of 181 identified O-glycosites in the current study; d) Proline frequency analysis of 435 O-glycosites from Uniprot. ±10 amino acid residues surrounding the O-glycosite were analyzed.

O-glycosylation alteration in MCI and AD.

In addition to analyzing CSF from healthy individuals, we also applied this strategy to systematically analyze CSF samples from MCI and AD patients to obtain global profiles of O-glycosylation in these disease states. In total, 366 O-glycopeptides mapping to 197 O-glycosites and 128 O-glycoproteins were found in MCI CSF and 358 O-glycopeptides mapping to 214 O-glycosites and 136 O-glycoproteins were found in AD CSF (Supplemental Table S3). The Venn diagrams demonstrate the diversity and overlap of O-glycosylation at these three stages (Figure 4a). 22.8%, 20.9% and 22.1% of unique O-glycopeptides presented in each group, suggesting the necessity to obtain global profile of O-glycosylation landscape before performing further quantitative O-glycoproteomic study. The overlap of O-glycopeptides, O-glycosites and O-glycoproteins between MCI and AD were 27.3%, 30.2% and 39.0%, respectively, which was higher than the overlap between healthy vs. MCI and healthy vs. AD. These results indicated certain glycosylation pattern changes during disease progression. To highlight more specific features of O-glycosylation patterns at each stage, the percentages of the different disease states of glycoforms were compared (Figure 4b, Supplemental Table S3). Since the core 1 O-glycoform is the most prevalent glycoform, we compared the core 1 percentage across the three states along with their sialylated and fucosylated glycoforms. Most of the percentages were quite comparable among three stages with a slight increase in core 1 and sialylated glycoforms. However, we observed a trend in decreased global fucosylation during disease progression. This was quite interesting because we also found reduced fucosylation patterns in the N-glycosylation study of AD CSF.41 Future biological mechanistic investigations may be necessary to reveal the functional role of fucosylation in AD pathogenesis and progression.

Figure 4.

Figure 4.

Open in a new tab

O-glycoproteome comparison between control, MCI, and AD. a) Venn diagram analysis of total O-glycoproteome in control, MCI, and AD; b) Comparison of distribution of different types of O-glycoforms in control, MCI, and AD.

Selected examples of O-glycoproteins.

Among all identified O-glycoproteins, we selected some proteins of interest to discuss below. As structural components of lipoprotein particles, apolipoproteins play important roles in maintaining their structure and regulating their metabolism and enzyme activities. Studies have shown that apolipoproteins are often modified by O-glycosylation, including apolipoprotein E, A-I, A-II and C-III.42–44 In our study, four apolipoproteins, apolipoproteins D, A-I, E and J were found to be O-glycosylated (Table S4). Although N-glycosylation of apolipoprotein D (APOD) has been extensively studied and there are many potential O-glycosylation sites (8 serine residues and 10 threonine residues) that exist in its protein sequence, O-glycosylation of APOD has not been reported before. In our datasets, APOD was found to be O-glycosylated at Thr86. A total of three O-glycoforms were detected at this O-glycosite, and all of them were fucosylated. In contrast, none of the glycoforms were found in MCI and AD CSF (Figure 5a). Using lectin-based isolation method, Cubedo et al reported the O-glycosylation of serum apolipoprotein A-I, but the exact O-glycosite or O-glycoform information was lacking.45 In our study, one glycoform was detected in healthy and MCI CSF at Thr92 and an additional glycosite was found in AD CSF at Ser 76 (Figure 5b). For APOE, two O-glycosites at Thr36 and Thr212 were found in all three states and one additional glycosite at Ser 215 was found in MCI CSF. Over half of the glycoforms were sialylated and our results suggested a high degree of micro-heterogeneity of glycosylation at Thr 212 (Figure 5c). Despite that APOJ has been shown carrying several PTMs such as N-glycosylation, ubiquitination and phosphorylation, O-glycosylation has not been reported before.46–49 In our study, seven glycosites were found among three states and only Thr 105 and Ser 210 were glycosylated in all three states (Figure 5d). Compared to ApoE, glycosylation on ApoJ showed higher degree of macro-heterogeneity and the change in site occupancy may play an important role in AD progression.50

Figure 5.

Figure 5.

Open in a new tab

Site-specific O-glycosylation analysis for a) apolipoprotein D, b) apolipoprotein A-I, c) apolipoprotein E, d) apolipoprotein J (clusterin) and e) amyloid-beta precursor protein (APP) in control, MCI, and AD.

APP is another protein of interest in AD research, because its aberrant proteolytic processing into amyloid β (Aβ) in the brain is an essential step during the pathogenesis of AD and O-glycosylation is considered to be related to Aβ formation.51 Here, we found Thr 22 was glycosylated at all three states with fucosylated glycoforms, and MCI state had one additional glycosite at Thr 667 (Figure 5e), which was one of the previously reported sites.52

In addition, we observed several interesting cases of O-glycosylation occurring at Thr residues of the N-glycosylation Asn-X-Thr consensus motif. These proteins are discussed in detail in the Supporting Information.

Strategy for Comprehensive PTM analysis of CSF endogenous peptides.

Glycosylation is not included as a possible dynamic modification in most peptidomic analyses, because it requires an additional glycan database to conduct the search, which requires much larger computational capacities. To overcome such limitation, we utilized EThcD to preserve the labile glycosylation and PTM-centric Byonic searching engine to enable site-specific glycoprofiling.53 We first tried to search the raw data directly by setting all the possible PTMs (glycosylation, phosphorylation, acetylation etc.) and then using the whole human protein and glycan databases. However, this caused computational issues. A second trial with a “focused” CSF protein database constructed based on the literature also did not work for this dataset, suggesting that the size of glycan database might be too large. Therefore, we developed a 3-step searching strategy (Figure S3) to facilitate the search; details can be found in the Methods section of Supporting Information. The key idea of this search method is that focused protein and glycan database are constructed by the first two rounds of searches. These databases will then be used to perform the final round search, where all the possible PTMs can be searched all at once.

For endogenous peptide separation, our lab previously developed a 10kDa MWCO-based protocol to achieve an optimal recovery rate.54,55 Here, we compare this 10kDa MWCO-based protocol with another 30kDa MWCO-based protocol reported in literature in terms of their performance on CSF samples.56 As shown in Figure S4 (Supplemental Table S5), 10kDa-based protocol outperformed 30kDa-based protocol in identification numbers of both total peptides and modified peptides. Therefore, we adapted the 10kDa-based protocol to separate CSF samples into peptide fraction and protein fraction (Figure S5).

CSF endogenous peptidome mapping in control, MCI, and AD.

With optimized analysis strategy, a total of 1411 endogenous peptides were identified in CSF samples from healthy individuals (Figure 6a), which was an over 2-fold increase compared to previous reports.22,23 Benefiting from EThcD fragmentation, labile PTMs including glycosylation and phosphorylation are well preserved (Figure S2). In total, 339 endogenous peptides with PTMs were identified, indicating that endogenous peptides in CSF went through extensive modifications. Among them, 89 and 34 peptides were O-glycosylated and phosphorylated, respectively. It is worth noting that this work is the first report to show extensive O-glycosylation on CSF endogenous peptides, and the number of O-glycosylated peptides exceeded the number of peptides with other PTMs. Overall, these O-glycopeptides were derivatized from 27 protein precursors, and 27 O-glycan compositions were identified. Interestingly, only 1 N-glycosylated peptide TNSTFVQALVEHVK from precursor protein prosaposin was found with N-glycosite at previously reported Asn215 residue.

Figure 6.

Figure 6.

Open in a new tab

Endogenous peptidome coverage in CSF from healthy individuals. a) The number of endogenous peptides and neuropeptides identified with/without PTMs; b) The number of endogenous peptides and neuropeptides with different PTMs.

As a subset of endogenous peptides, neuropeptides are of interest because of their involvement in various biological processes as inter-cellular signaling molecules.57 By referring to the human neuropeptide database NeuroPep,58 276 peptides were identified as neuropeptides. More than half of them (178, 64.5%) belonged to chromogranin/secretogranin family, followed by ProSAAS (28, 10.1%), NPY (18, 6.5%), VGF (17, 6.2%), opioid (11, 4.0%) and 7B2 (7, 2.5%) (Figure S6). 88 were modified, including acetylation, amidation, O-glycosylation, phosphorylation, Gln to pyro-Glu conversion. Among the 15 O-glycosylated neuropeptides, 14 originated from ProSAAS and 1 was from secretogranin-1. ProSAAS-derived neuropeptides, SAAS, PEN, and LEN, are among the most abundant peptides present in mouse hypothalamus,59–61 which have been implicated in the regulation of food intake and body weight.62,63 We detected two O-glycosylation sites (Thr53, Thr247) in the LEN and SAAS regions with 11 O-glycosylated LEN peptides and 3 O-glycosylated SAAS peptides (Supplemental Table S6). At site Thr53, two O-glycoforms with sialylated T-antigen were identified, whereas at site Thr247 three O-glycoforms with T-antigen in both sialylated and non-sialylated forms and one non-sialylated Tn-antigen glycoforms were found.

There was a decrease in the number of endogenous peptides in MCI and AD compared with healthy individuals. A closer look into the peptide modifications showed that the decrease mainly came from the unmodified peptides (Figure 7a, Supplemental Table S7). A more detailed analysis of PTM types revealed that there was a trend of increasing O-glycosylation, Gln->pyro-Glu, acetylation, phosphorylation, and a trend for slightly decreased oxidation (Figure 7b). These intriguing observations represent a global landscape of endogenous peptidome from control, MCI and AD, exhibiting distinct features at different states of AD progression. Further investigation is needed to explore the biological function, their potential roles of these endogenous peptides in AD progression, and how each PTM regulates their bioactivities.

Figure 7.

Figure 7.

Open in a new tab

Endogenous peptidome comparison among control, MCI, and AD. a) The number of endogenous peptides identified with/without PTMs; b) The comparison of the percentage of endogenous peptides with different PTMs identified in control, MCI, and AD.

CONCLUSIONS

In summary, we developed an effective workflow to simultaneously analyze site-specific O-glycoproteome and endogenous peptides with PTMs in CSF. For the O-glycoproteome study, an optimized boronic acid enrichment method enhanced site-specific O-glycopeptide analysis. In total, 308 intact O-glycopeptides from 182 O-glycosites and 110 O-glycoproteins were identified from healthy individuals, representing the largest dataset of site-specific O-glycoproteome study in CSF to date, including 154 novel O-glycosites reported for the first time. We also profiled the O-glycoproteome in MCI and AD patients. For endogenous peptide study, we developed a peptidomics workflow that enabled comprehensive PTMs analysis. From this analysis, a considerable number of CSF endogenous peptides were O-glycosylated in all three states, while only a few notable peptides were N-glycosylated. Although, an in-depth glycoproteomic biomarker discovery study on individual control, MCI and AD subjects is ideal for biomarker discovery, the relatively small size of CSF samples, instrument/labor time, and financial constraints limited these experiments in this study. Instead, the samples in each group are pooled to reduce biological variation and allow deeper profiling of low-abundance glycoproteins.64 Our findings shed light on the CSF O-glycoproteome landscape, dominant glycosylation differences, similarities during AD progression, and PTM focused peptidome mapping. Future investigations will be conducted using strategies developed here to explore potential O-glycosylated or endogenous proteolytic biomarker candidates in a more quantitative manner, such as using 12-plex DiLeu isobaric labeling to analyze individual subjects and pinpoint more specific changes of interesting glycoprotein(s)/glycosite(s)/glycoform(s).65

METHODS

Chemicals and materials.

See the Supporting Information for details.

CSF samples.

See the Supporting Information and Supplemental Table S1 for details. In short, 48 enrollees (16 cognitively normal individuals, 16 individuals with MCI and 16 individuals with AD dementia) in the Wisconsin Alzheimer’s Disease Research Center (ADRC) participated in this study. The University of Wisconsin Institutional Review Board approved all study procedures. CSF aliquots from 16 individuals of each group were pooled together for analysis.

Sample processing and protein digestion.

Details of PANC1 cell culture, protein extraction and digestion, and CSF sample processing were provided in the Supporting Information. Briefly, CSF sample was separated into a peptide fraction and a protein fraction using 10 kDa MWCO protocol. The peptide fraction was injected for LC-MS/MS analysis after desalting. The protein fraction was reduced, alkylated and digested with trypsin. N-glycans were removed by PNGase F before O-glycopeptide enrichment.

Boronic acid enrichment.

See the Supporting Information for details.

LC-MS/MS analysis.

Details of LC-MS/MS data acquisition were provided in the Supporting Information. Briefly, samples were analyzed on the Orbitrap Fusion™ Lumos™ Tribrid™ Mass Spectrometer (Thermo Fisher Scientific, San Jose, CA) coupled to a Dionex UPLC system. Data was acquired in data dependent acquisition (DDA) mode with EThcD fragmentation.

Data analysis.

All data were searched by Byonic (version 2.9.38, Protein Metrics Inc, San Carlos, CA) incorporated in Proteome Discoverer 2.1. Details of search parameters, databases and post-search processing were provided in the Supporting Information. Data are available at Mass Spectrometry Interactive Virtual Environment (https://massive.ucsd.edu) with deposit ID of MSV000087160.

Supplementary Material

Supplemental Info & Figures
Supplemental Table 6
Supplemental Table 7
Supplemental Table 4
Supplemental Table 3
Supplemental Table 5
Supplemental Table 1
Supplemental Table 2

Acknowledgements

This research was supported in part by the National Institutes of Health (NIH) grants RF1 AG052324, R21AG065728, U01CA231081, R01 DK071801, and P30AG062715. The Orbitrap instruments were purchased through the support of an NIH shared instrument grant (NIH-NCRR S10RR029531) and Office of the Vice Chancellor for Research and Graduate Education at the University of Wisconsin-Madison. L.L. acknowledges a Vilas Distinguished Achievement Professorship and Charles Melbourne Johnson Distinguished Chair Professorship with funding provided by the Wisconsin Alumni Research Foundation and University of Wisconsin-Madison School of Pharmacy. We thank M. Bern from Protein Metrics for providing 1-year access to Byonic software package.

Footnotes

Supporting Information Available: This material is available free of charge via the Internet: Results and discussion; Methods; Supplementary figures of boronic acid enrichment optimization, representative EThcD fragmentation spectra, three-step searching strategy, comparison between 10kDa and 30kDa MWCO-based protocols, workflow for O-glycoproteome and endogenous peptide analysis, and different families of neuropeptides in CSF.

Data Availability

Url: https://massive.ucsd.edu

Deposit ID: MSV000087160

Account name: MSV000087160_reviewer

Password: revieweronly

References:

  • (1).McComb JG Recent Research into the Nature of Cerebrospinal Fluid Formation and Absorption. J. Neurosurg 1983, 59, 369–383. [DOI] [PubMed] [Google Scholar]
  • (2).Stopa EG; Berzin TM; Kim S; Song P; Kuo-LeBlanc V; Rodriguez-Wolf M; Baird A; Johanson CE Human Choroid Plexus Growth Factors: What Are the Implications for CSF Dynamics in Alzheimer’s Disease? Exp. Neurol 2001, 167, 40–47. [DOI] [PubMed] [Google Scholar]
  • (3).Chodobski A; Wojcik BE; Loh YP; Dodd KA; Szmydynger-Chodobska J; Johanson CE; Demers DM; Chun ZG; Limthong NP Vasopressin Gene Expression in Rat Choroid Plexus; Springer, 1998; pp 59–65. [DOI] [PubMed] [Google Scholar]
  • (4).Zhang J Proteomics of Human Cerebrospinal Fluid – the Good, the Bad, and the Ugly. Proteom. - Clin. Appl 2007, 1 (8), 805–819. [DOI] [PubMed] [Google Scholar]
  • (5).Fonteh AN; Harrington RJ; Huhmer AF; Biringer RG; Riggins JN; Harrington MG Identification of Disease Markers in Human Cerebrospinal Fluid Using Lipidomic and Proteomic Methods. Dis. Markers 2006, 22, 39–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (6).Yuan X; Desiderio DM Proteomics Analysis of Human Cerebrospinal Fluid. J. Chromatogr. B 2005, 815, 179–189. [DOI] [PubMed] [Google Scholar]
  • (7).Ohtsubo K; Marth JD Glycosylation in Cellular Mechanisms of Health and Disease. Cell 2006, 126, 855–867. [DOI] [PubMed] [Google Scholar]
  • (8).Zhong X; Chen Z; Snovida S; Liu Y; Rogers JC; Li L Capillary Electrophoresis-Electrospray Ionization-Mass Spectrometry for Quantitative Analysis of Glycans Labeled with Multiplex Carbonyl-Reactive Tandem Mass Tags. Anal. Chem 2015, 87, 6527–6534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (9).Chen Z; Zhong X; Tie C; Chen B; Zhang X; Li L Development of a Hydrophilic Interaction Liquid Chromatography Coupled with Matrix-Assisted Laser Desorption/Ionization-Mass Spectrometric Imaging Platform for N-Glycan Relative Quantitation Using Stable-Isotope Labeled Hydrazide Reagents. Anal. Bioanal. Chem 2017, 409 (18), 4437–4447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (10).Chen Z; Huang J; Li L Recent Advances in Mass Spectrometry (MS)-Based Glycoproteomics in Complex Biological Samples. Trends Analyt Chem. 2018, 118, 880–892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (11).Varki A; Cummings R; Esko J; Freeze H; Hart G; Marth J Essentials of Glycobiology.; Cold Spring Harber Laboratory Press, New York; 1998. [PubMed] [Google Scholar]
  • (12).Chen Z; Glover MS; Li L Recent Advances in Ion Mobility–Mass Spectrometry for Improved Structural Characterization of Glycans and Glycoconjugates. Curr. Opin. Chem. Biol 2018, 42, 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (13).Tabak LA In Defense of the Oral Cavity: Structure, Biosynthesis, and Function of Salivary Mucins. Annu. Rev. Physiol 1995, 57, 547–564. [DOI] [PubMed] [Google Scholar]
  • (14).Tran DT; Hagen KGT Mucin-Type O-Glycosylation during Development. J. Biol. Chem 2013, 288, 6921–6929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (15).Yang X; Qian K Protein O-GlcNAcylation: Emerging Mechanisms and Functions. Nat. Rev. Mol. Cell Biol 2017, 18 (7), 452–465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (16).Thotakura NR; Blithe DL Glycoprotein Hormones: Glycobiology of Gonadotrophins, Thyrotrophin and Free α Subunit. Glycobiology 1995, 5, 3–10. [DOI] [PubMed] [Google Scholar]
  • (17).Yu Q; Canales A; Glover MS; Das R; Shi X; Liu Y; Keller MP; Attie AD; Li L Targeted Mass Spectrometry Approach Enabled Discovery of O-Glycosylated Insulin and Related Signaling Peptides in Mouse and Human Pancreatic Islets. Anal. Chem 2017, 89 (17), 9184–9191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (18).Cao Q; Yu Q; Liu Y; Chen Z; Li L Signature-Ion-Triggered Mass Spectrometry Approach Enabled Discovery of N- and O-Linked Glycosylated Neuropeptides in the Crustacean Nervous System. J. Proteome Res 2019, 19 (2), 634–643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (19).Nilsson J; Rüetschi U; Halim A; Hesse C; Carlsohn E; Brinkmalm G; Larson G Enrichment of Glycopeptides for Glycan Structure and Attachment Site Identification. Nat. Methods 2009, 6 (11), 809–811. [DOI] [PubMed] [Google Scholar]
  • (20).Halim A; Rüetschi U; Larson G; Nilsson J. LC-MS/MS Characterization of O-Glycosylation Sites and Glycan Structures of Human Cerebrospinal Fluid Glycoproteins. J. Proteome Res 2013, 12 (2), 573–584. [DOI] [PubMed] [Google Scholar]
  • (21).Flowers SA; Grant OC; Woods RJ; Rebeck GW O-Glycosylation on Cerebrospinal Fluid and Plasma Apolipoprotein E Differs in the Lipid-Binding Domain. Glycobiology 2019, 30 (2), 74–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (22).Nilsson J; Brinkmalm G; Ramadan S; Gilborne L; Noborn F; Blennow K; Wallin A; Svensson J; Abo-Riya MA; Huang X; Larson G Synthetic Standard Aided Quantification and Structural Characterization of Amyloid-Beta Glycopeptides Enriched from Cerebrospinal Fluid of Alzheimer’s Disease Patients. Sci. Rep 2019, 9 (1), 5522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (23).Hölttä M; Zetterberg H; Mirgorodskaya E; Mattsson N; Blennow K; Gobom J Peptidome Analysis of Cerebrospinal Fluid by LC-MALDI MS. PLoS One 2012, 7, e42555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (24).Zougman A; Pilch B; Podtelejnikov A; Kiehntopf M; Schnabel C; Kumar C; Mann M Integrated Analysis of the Cerebrospinal Fluid Peptidome and Proteome. J. Proteome Res 2007, 7, 386–399. [DOI] [PubMed] [Google Scholar]
  • (25).Levery SB; Steentoft C; Halim A; Narimatsu Y; Clausen H; Vakhrushev SY Advances in Mass Spectrometry Driven O-Glycoproteomics. Biochim. Biophys. Acta - Gen. Subj 2015, 1850 (1), 33–42. [DOI] [PubMed] [Google Scholar]
  • (26).Kong S; Zhang Q; Yang L; Huang Y; Liu M; Yan G; Zhao H; Wu M; Zhang X; Yang P; Cao W Effective Enrichment Strategy Using Boronic Acid-Functionalized Mesoporous Graphene-Silica Composites for Intact N- and O-Linked Glycopeptide Analysis in Human Serum. Anal. Chem 2021, 93, 6682–6691. [DOI] [PubMed] [Google Scholar]
  • (27).Sparbier K; Koch S; Kessler I; Wenzel T; Kostrzewa M Selective Isolation of Glycoproteins and Glycopeptides for MALDI-TOF MS Detection Supported by Magnetic Particles. J. Biomol. Tech 2005, 16, 407. [PMC free article] [PubMed] [Google Scholar]
  • (28).Xu Y; Wu Z; Zhang L; Lu H; Yang P; Webley PA; Zhao D Highly Specific Enrichment of Glycopeptides Using Boronic Acid-Functionalized Mesoporous Silica. Anal. Chem 2008, 81, 503–508. [DOI] [PubMed] [Google Scholar]
  • (29).Chen W; Smeekens JM; Wu R A Universal Chemical Enrichment Method for Mapping the Yeast N-Glycoproteome by Mass Spectrometry (MS). Mol. Cell Proteomics 2014, 13, 1563–1572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (30).Wang X; Yuan Z-F; Fan J; Karch KR; Ball LE; Denu JM; Garcia BA A Novel Quantitative Mass Spectrometry Platform for Determining Protein O-GlcNAcylation Dynamics. Mol. Cell Proteomics 2016, 15, 2462–2475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (31).Batth TS; Francavilla C; Olsen JV Off-Line High-PH Reversed-Phase Fractionation for in-Depth Phosphoproteomics. J. Proteome Res 2014, 13, 6176–6186. [DOI] [PubMed] [Google Scholar]
  • (32).Liu H; Sadygov RG; Yates JR A Model for Random Sampling and Estimation of Relative Protein Abundance in Shotgun Proteomics. Anal. Chem 2004, 76, 4193–4201. [DOI] [PubMed] [Google Scholar]
  • (33).Huang J; Dong J; Shi X; Chen Z; Cui Y; Liu X; Ye M; Li L Dual-Functional Titanium(IV) Immobilized Metal Affinity Chromatography Approach for Enabling Large-Scale Profiling of Protein Mannose-6-Phosphate Glycosylation and Revealing Its Predominant Substrates. Anal. Chem 2019, 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (34).Chen Z; Yu Q; Hao L; Liu F; Johnson J; Tian Z; Kao WJ; Xu W; Li L Site-Specific Characterization and Quantitation of N -Glycopeptides in PKM2 Knockout Breast Cancer Cells Using DiLeu Isobaric Tags Enabled by Electron-Transfer/Higher-Energy Collision Dissociation (EThcD). Analyst 2018, 143 (11), 2508–2519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (35).Mitoma J; Petryniak B; Hiraoka N; Yeh J-C; Lowe JB; Fukuda M Extended Core 1 and Core 2 Branched O-Glycans Differentially Modulate Sialyl Lewis X-Type L-Selectin Ligand Activity. J. Biol. Chem 2003, 278, 9953–9961. [DOI] [PubMed] [Google Scholar]
  • (36).Qin H; Cheng K; Zhu J; Mao J; Wang F; Dong M; Chen R; Guo Z; Liang X; Ye M; Zou H Proteomics Analysis of O-GalNAc Glycosylation in Human Serum by an Integrated Strategy. Anal. Chem 2017, 89 (3), 1469–1476. [DOI] [PubMed] [Google Scholar]
  • (37).Darula Z; Sherman J; Medzihradszky KF How to Dig Deeper? Improved Enrichment Methods for Mucin Core-1 Type Glycopeptides. Mol. Cell Proteomics 2012, 11, O111. 016774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (38).Elhammer AP; Poorman RA; Brown E; Maggiora LL; Hoogerheide J; Kezdy F The Specificity of UDP-GalNAc: Polypeptide N-Acetylgalactosaminyltransferase as Inferred from a Database of in Vivo Substrates and from the in Vitro Glycosylation of Proteins and Peptides. J. Biol. Chem 1993, 268, 10029–10038. [PubMed] [Google Scholar]
  • (39).O’Connell B; Hagen F; Tabak L The Influence of Flanking Sequence on the O-Glycosylation of Threonine in Vitro. J. Biol. Chem 1992, 267, 25010–25018. [PubMed] [Google Scholar]
  • (40).Wilson I; Gavel Y; Heijne GV Amino Acid Distributions around O-Linked Glycosylation Sites. Biochem. J 1991, 275, 529–534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (41).Nehrke K; Hagen FK; Tabak LA Charge Distribution of Flanking Amino Acids Influences O-Glycan Acquisition in Vivo. J. Biol. Chem 1996, 271, 7061–7065. [DOI] [PubMed] [Google Scholar]
  • (42).Chen Z; Yu Q; Yu Q; Johnson J; Shipman R; Zhong X; Huang J; Asthana S; Carlsson C; Okonkwo O; Li L In-Depth Site-Specific Analysis of N-Glycoproteome in Human Cerebrospinal Fluid and Glycosylation Landscape Changes in Alzheimer’s Disease. Mol. Cell Proteomics 2021, 20, 100081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (43).Remaley A; Wong A; Schumacher U; Meng M; Brewer H; Hoeg J O-Linked Glycosylation Modifies the Association of Apolipoprotein A-II to High Density Lipoproteins. J. Biol. Chem 1993, 268, 6785–6790. [PubMed] [Google Scholar]
  • (44).Beisiegel U; Weber W; Havinga JR; Ihrke G; Hui DY; Wernette-Hammond ME; Turck CW; Innerarity TL; Mahley RW Apolipoprotein E-Binding Proteins Isolated from Dog and Human Liver. Arterioscler. Thromb. Vasc. Biol 1988, 8, 288–297. [DOI] [PubMed] [Google Scholar]
  • (45).Wada Y; Kadoya M; Okamoto N Mass Spectrometry of Apolipoprotein C-III, a Simple Analytical Method for Mucin-Type O-Glycosylation and Its Application to an Autosomal Recessive Cutis Laxa Type-2 (ARCL2) Patient. Glycobiology 2012, 22, 1140–1144. [DOI] [PubMed] [Google Scholar]
  • (46).Cubedo J; Padró T; Badimon L Glycoproteome of Human Apolipoprotein AI: N-and O-Glycosylated Forms Are Increased in Patients with Acute Myocardial Infarction. Transl. Res 2014, 164, 209–222. [DOI] [PubMed] [Google Scholar]
  • (47).Woody SK; Zhao L Clusterin (APOJ) in Alzheimer’s Disease: An Old Molecule with a New Role; InTech, 2016. [Google Scholar]
  • (48).Burkey B; Harmony J Intracellular Processing of Apolipoprotein J Precursor to the Mature Heterodimer. J. Lipid Res 1991, 32, 1039–1048. [PubMed] [Google Scholar]
  • (49).Liang H-C; Russell C; Mitra V; Chung R; Hye A; Bazenet C; Lovestone S; Pike I; Ward M Glycosylation of Human Plasma Clusterin Yields a Novel Candidate Biomarker of Alzheimer’s Disease. J. Proteome Res 2015, 14, 5063–5076. [DOI] [PubMed] [Google Scholar]
  • (50).Kapron JT; Hilliard GM; Lakins JN; Tenniswood MP; West KA; Carr SA; Crabb JW Identification and Characterization of Glycosylation Sites in Human Serum Clusterin. Protein Sci. 1997, 6, 2120–2133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (51).Zhang Q; Ma C; Chin L-S; Li L Integrative Glycoproteomics Reveals Protein N-Glycosylation Aberrations and Glycoproteomic Network Alterations in Alzheimer’s Disease. Sci. Adv 2020, 6 (40), eabc5802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (52).Shi J; Ku X; Zou X; Hou J; Yan W; Zhang Y Comprehensive Analysis of O-Glycosylation of Amyloid Precursor Protein (APP) Using Targeted and Multi-Fragmentation MS Strategy. Biochim. Biophys. Acta - Gen. Subj 2021, 1865 (10), 129954. [DOI] [PubMed] [Google Scholar]
  • (53).Halim A; Brinkmalm G; Rüetschi U; Westman-Brinkmalm A; Portelius E; Zetterberg H; Blennow K; Larson G; Nilsson J Site-Specific Characterization of Threonine, Serine, and Tyrosine Glycosylations of Amyloid Precursor Protein/Amyloid β-Peptides in Human Cerebrospinal Fluid. Proc. Natl. Acad. Sci. U. S. A 2011, 108 (29), 11848–11853. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (54).Bern M; Kil YJ; Becker C Byonic: Advanced Peptide and Protein Identification Software. Curr Protoc Bioinform 2012, 40 (1), 13.20.1–13.20.14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (55).Cunningham R; Wang J; Wellner D; Li L Investigation and Reduction of Sub-microgram Peptide Loss Using Molecular Weight Cut-off Fractionation Prior to Mass Spectrometric Analysis. J. Mass Spectrom 2012, 47, 1327–1332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (56).Wang J; Cunningham R; Zetterberg H; Asthana S; Carlsson C; Okonkwo O; Li L Label-free Quantitative Comparison of Cerebrospinal Fluid Glycoproteins and Endogenous Peptides in Subjects with Alzheimer’s Disease, Mild Cognitive Impairment, and Healthy Individuals. Proteom. - Clin. Appl 2016, 10 (12), 1225–1241. [DOI] [PubMed] [Google Scholar]
  • (57).Secher A; Kelstrup CD; Conde-Frieboes KW; Pyke C; Raun K; Wulff BS; Olsen JV Analytic Framework for Peptidomics Applied to Large-Scale Neuropeptide Identification. Nat. Commun 2016, 7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (58).Sossin WS; Fisher JM; Scheller RH Cellular and Molecular Biology of Neuropeptide Processing and Packaging. Neuron 1989, 2, 1407–1417. [DOI] [PubMed] [Google Scholar]
  • (59).Wang Y; Wang M; Yin S; Jang R; Wang J; Xue Z; Xu T NeuroPep: A Comprehensive Resource of Neuropeptides. Database 2015, 2015 (0), bav038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (60).Fricker LD; McKinzie AA; Sun J; Curran E; Qian Y; Yan L; Patterson SD; Courchesne PL; Richards B; Levin N Identification and Characterization of ProSAAS, a Granin-like Neuroendocrine Peptide Precursor That Inhibits Prohormone Processing. J. Neurosci 2000, 20, 639–648. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (61).Fricker LD Neuropeptides and Other Bioactive Peptides: From Discovery to Function; Morgan & Claypool Life Sciences, 2012; Vol. 1, pp 1–122. [Google Scholar]
  • (62).Mzhavia N; Berman Y; Che F-Y; Fricker LD; Devi LA ProSAAS Processing in Mouse Brain and Pituitary. J. Biol. Chem 2001, 276, 6207–6213. [DOI] [PubMed] [Google Scholar]
  • (63).Gomes I; Bobeck EN; Margolis EB; Gupta A; Sierra S; Fakira AK; Fujita W; Müller TD; Müller A; Tschöp MH Identification of GPR83 as the Receptor for the Neuroendocrine Peptide PEN. Sci. Signal 2016, 9, ra43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (64).Gomes I; Aryal DK; Wardman JH; Gupta A; Gagnidze K; Rodriguiz RM; Kumar S; Wetsel WC; Pintar JE; Fricker LD GPR171 Is a Hypothalamic G Protein-Coupled Receptor for BigLEN, a Neuropeptide Involved in Feeding. Proc. Natl. Acad. Sci. U. S. A 2013, 110, 16211–16216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (65).Kendziorski C; Irizarry RA; Chen K-S; Haag JD; Gould MN On the Utility of Pooling Biological Samples in Microarray Experiments. Proc. Natl. Acad. Sci. U.S.A 2005, 102 (12), 4252–4257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (66).Frost DC; Greer T; Li L High-Resolution Enabled 12-Plex DiLeu Isobaric Tags for Quantitative Proteomics. Anal. Chem 2015, 87 (3), 1646–1654. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Info & Figures
NIHMS1771217-supplement-Supplemental_Info___Figures.pdf (712.3KB, pdf)
Supplemental Table 6
NIHMS1771217-supplement-Supplemental_Table_6.xlsx (9KB, xlsx)
Supplemental Table 7
NIHMS1771217-supplement-Supplemental_Table_7.xlsx (8.6MB, xlsx)
Supplemental Table 4
NIHMS1771217-supplement-Supplemental_Table_4.xlsx (12.2KB, xlsx)
Supplemental Table 3
NIHMS1771217-supplement-Supplemental_Table_3.xlsx (539.7KB, xlsx)
Supplemental Table 5
NIHMS1771217-supplement-Supplemental_Table_5.xlsx (5.4MB, xlsx)
Supplemental Table 1
NIHMS1771217-supplement-Supplemental_Table_1.xlsx (8.3KB, xlsx)
Supplemental Table 2
NIHMS1771217-supplement-Supplemental_Table_2.xlsx (434.2KB, xlsx)

Data Availability Statement

Url: https://massive.ucsd.edu

Deposit ID: MSV000087160

Account name: MSV000087160_reviewer

Password: revieweronly

RESOURCES