1.      Abbreviations

GAPDH                 :               Glyceraldehyde 3-phosphate dehydrogenase

PyGAPDH             :               Plasmodium yoeliiGAPDH

rPfGAPDH            :               Recombinant Plasmodium falciparumGAPDH

G-3-P                      :               D-Glyceraldehyde-3-phosphate

2-DE                       :               2-Dimensional Electrophoresis

PTM                       :               Post-Translational Modification

PBST                      :               Phosphate Buffer Saline Tween

PVDF                     :               Polyvinylidene fluoride 

2.      Introduction 

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH, EC 1.2.1.12) has emerged as a multifunctional protein with several moonlighting functions. In mammalian cells, involvement of GAPDH in RNA transport, DNA replication, vesicular transport, cytoskeletal reorganization, membrane fusion, apoptosis etc. is well documented[1]. The structural basis for such functional diversity has been attributed to multiple species of GAPDH arising due to multiple PTMs[2]. A few chemical modifications have been correlated to specific functions[1d,3] while majority of GAPDH species are yet to be characterized. Since each chemical modification leads to a new protein speciation, multiple PTMs in different combinations could create a vast number of species leading to complexity in cellular functions without any expansion of genome. For a complete understanding of such systems, knowledge about the structure of each protein species, its function and spatio-temporal distribution inside the cell will be needed. Recent strategies of applying the advanced proteomics technologies for protein separation and sequencing, to multiple molecular species of a single gene product is providing robust structural data laying down the foundation for understanding the cellular physiology [4].

In P. yoelii, GAPDH is associated with multiple organelles viz. cytosol, nuclei, cell membranes, cytoskeletal elements etc. and has a distinct organelle specific electrophoretic variant profile. 2DE western blots of P. yoeliisub-cellular fractions showed ≥20-25 different species of GAPDH [5] arising due to post translational modifications. This structural diversity is forming the molecular basis for the involvement of the parasite GAPDH in multiple non-glycolytic functions such as vesicular transport and biogenesis of the apical complex[6], likely involvement in merozoite invasion of red blood cells [7] and invasion of liver cells by sporozoites [8]etc. Thus, the observed functional and localization diversity does correlate with the underlying structural heterogeneity. For understanding the molecular basis of various cellular functions of this protein, it is essential that we determine the chemical structure of each molecular species. Initial attempts to excise the relevant spots from a 2-Dimensional gel, digest the protein with trypsin and sequence the peptides for PTM determination did not succeed largely due to low resolution of spots in 2D-gel and inadequate sensitivity of our mass spectrometer. To tide over these limitations, we took an alternative approach where the whole cell extract was fractionated in soluble and particulate fractions and proteins were analyzed on a 1D-SDS-PAGE. PyGAPDH containing protein bands were digested with trypsin and subjected to MS and MS/MS analysis. Results showed that twelve residues underwent modifications. The changes consisted of four phosphorylations (pS¹⁴⁴, pT¹⁴⁶, pS²⁰⁴ and pS²¹³), five ubiquitinations (uK⁷³, uK²¹⁸, uK²²², uK²³⁰ and uK³³⁶⁾, three acetylations (acK¹⁶³, acK²³⁰, acK³⁰¹), two methylations (mK²¹⁸ and mK²³⁰), one dimethylation (m2K²³⁰) and one nitrosylation (nC¹⁵⁷). 

3.      Materials and Methods 

3.1.  P. yoeliiculturing and whole cell extract preparation 

The lethal strain of P. yoelii17XL was grown in mice as described earlier[9]. The parasite pellet isolated from infected blood was suspended in lysis buffer (50 mMTris-HCl, pH 7.4, 2 mM EDTA, 25 mMNaCl, 1 mM PMSF and 1x protease inhibitor cocktail) and subjected to 3-4 cycles of freeze-thaw in liquid nitrogen. This was labeled as the Whole Cell Extract (WCE). Animal experiments involving mice were approved by the Institutional Animal Ethics Committee (IAEC) of the Tata Institute of Fundamental Research, which is constituted by the ‘Committee for the Purpose of Supervision and Experiments on Animals (CPCSEA)’, Government of India (Project approval no: TIFR/IAEC/2010-4 and TIFR/IAEC/2012-5). 

3.2.  Generation of Anti-rPfGAPDH Serum and Immunoprecipitation 

Anti-PfGAPDH sera were generated using purified recombinant PfGAPDH as described earlier[5].

For immuno-precipitation experiments, whole cell extract was centrifuged at 40,000xg and the supernatant was labeled as the ‘Soluble fraction’. The pellet was dissolved in 1% NP-40 buffer and centrifuged at 40,000xg. The supernatant was collected and labeled as the ‘Particulate fraction’. The protocol followed for immuno-precipitation was similar to that used earlier except that anti-rPfGAPDHIgGs were used in place of anti-rPfenoIgGs[9]. 

3.3.  Electrophoresis and western blotting 

Proteins were resolved on a 12% SDS-PAGE [10] and either stained with Coomassie Brilliant Blue R-250 or were transferred to a PVDF membrane as described earlier[5]. The blots were treated with mouse anti-rPfGAPDH serum (1:1000 dilution) followed by washing and incubation with HRP conjugated secondary antibody. The immunoblots were developed using di-anilinobenzene substrate. 

3.4.   In-Gel Tryptic Digestion and LC-ESI-MS/MS Analysis 

Protein bands that corresponded to PyGAPDH positive in western blot were excised from a Coomassie stained gel and subjected to in-gel trypsin digestion as described earlier[9,11]. Extracted peptides were analyzed by LC-ESI-MS/MS using an Agilent 6520-Q-TOF. Details for the mass spectrometric analysis of peptides were as described earlier[9].Briefly, the extracted peptides were re-suspended into 3 μL 0.1% formic acid (Solvent A) of which 2.8 μL was applied to Agilent HPLC chip (G4240-62002) (injected at a rate of 40 μL/min). Mobile phases (A): 0.1% formic acid, (B): 90% acetonitrile, 0.1% formic acid. After sample injection, the column was washed by a gradient 3-12% of phase B for 3 min and peptides were eluted with linear gradient of varying slopes viz. 12-60% B from 3 to 23 min, 60-95% of B from 23 to 27 min. Q-TOF MS conditions were: drying gas 4L/Min, 300^oC; skimmer: 65 V; fragmentor: 175V; collision energy: slope 3.7 V, offset 2.5 V. The MS scan range was m/z =100-1700 and the scan rate was 5 spectra/sec. For MS/MS, scan range was m/z=100-1700 and scan rate was 3 spectra/sec. Active exclusion was set on for 0.5 min after 2 MS/MS spectra of a parent ion. For each MS, five most abundant precursor ions were sequenced. Preferred charge states were set to 2⁺, 3⁺ and 4⁺. 

3.5.  Data Analysis Using Mascot 

From all the MS data files, Mascot generic files (.mgf) were extracted using Agilent Mass hunter qualitative analysis software. All mass-spectrometric data were analyzed using a Matrix Science Mascot in house server[12]. MS/MS data were searched against the NCBInr database for P. falciparum and P.yoelii.Since a large fraction of P. yoeliithan P. falciparum proteomes are unannotated, most proteins are marked as hypothetical. For obtaining as many proteins annotated as possible, databases for both species of Plasmodium were used for search. However, hits obtained against a species different from the one used in the experiments were analyzed cautiously for sequence differences before reporting the PTMs. Parameters used for the search were: peptide mass tolerance in MS was set to 10 ppm and for MS/MS to 0.6 Da; peptide charges were set to 2⁺, 3⁺ and 4⁺; missed cleavage, 2; fixed modifications: carbamidomethyl (cysteine); variable modifications: oxidation of methionine and target modifications phosphorylation (Ser/Thr), phospho (Tyr), acetyl (Lys), methylation (Lys; mono, di and tri); nitrosylation on cysteine and Gly-Gly (Lys) for ubiquitination. All PTMs reported here were manually validated. This involved examining the ions detected in MS/MS spectra and if two consecutive ions differed in mass equivalent to the modified residue, it was treated as true positive. Although this approach is time consuming, it leads to greaterconfidence in identification and assignment of PTMs. Cases where a signature peak was missing, and if the site of modification could not be inferred from neighboring peaks, such PTMs were not reported. 

The mass spectrometric data have been deposited at the ProteomeXchange Consortium [13] via the PRIDE partner repository [14] with the dataset identifier PXD002313 and 10.6019/PXD002313. The desired pride XML files were obtained from Mascot .dat files using the PRIDE converter 2 tool [15]and inspected using the PRIDE Inspector tool [16] before uploading them. These PRIDE XML files were deposited to the repository along with the raw data files (Agilent .d files), peak lists (Mascot .mgf files) and the search results files (Mascot .dat files).

4.      Results

4.1.  PyGAPDH Variants With MW~51 kDa may Be Ubiquitinated

P. yoeliiWholeCell Extract (WCE) was subjected to centrifugation (40,000xg for 30 minutes). The supernatant was designated as the soluble fraction while the pellet containing nuclei, membrane vesicles and cytoskeletal elements was treated as the particulate fraction. All three fractions (WCE, Soluble and Particulate) were analyzed on a 12% SDS-PAGE and probed with anti-rPfGAPDHantisera in a western blot. As shown earlier[5], three major positive bands at ~27, 37 and 51 kDa were observed (Figure 1). 

Since molecular mass of PyGAPDH in its native state is 37 kDa, the lower mass band at ~27 kDa could arise as a result of controlled proteolysis. Observation of higher molecular mass species of PyGAPDH raised the possibility of post-translational modifications involving conjugation with multiple ubiquitin moieties or ubiquitin like modifiers (e.g. SUMO). To test the possibility of the higher molecular weight species of PyGAPDH in the soluble fraction of P. yoeliibeing ubiquitinated, an immuno-precipitation experiment was performed. Using purified fraction of IgGs derived from rPfGAPDH antisera, all variants of PyGAPDH present in the soluble and particulate (solubilized in 1% NP-40) fractions were pulled down (Figure 2A) and the proteins were run on a 12% SDS-PAGE. Blot of the gel (Western analysis) was probed using rabbit anti-ubiquitin antibody (Figure2B(ii) & (iii)).

In a parallel experiment, the soluble fraction was also analyzed by Western using mouse anti-rPfGAPDH antibody. As expected, three major protein bands at MW~27, 37 and 51 kDa were observed (Figure 2B (i)). Certain minor bands present are likely to arise due to proteolysis. Higher molecular wt. species (MW~51 kDa) was present only in the soluble (cytosolic) fraction (Figure 2B). This is consistent with the earlier observations of electrophoretic variant profiles in 2DE[5]. In the anti-rPfGAPDH antibody pull down sample from the soluble (i.e. cytosolic) fraction, an intense band at ~51 kDa MW was observed that was positive for ubiquitin indicating it to be the ubiquitinated form of PyGAPDH. This sample also had ubiquitin positive band at ~37 kDa albeit of much lower intensity. Such a band could arise if the~61kDaubiquitinated form of PyGAPDH got proteolysed yielding a ~37 kDa form that still carried ubiquitin moieties (Figure 2B (ii)). The particulate fraction did not show any ubiquitinated form of PyGAPDH(Figure 2B (iii)).From the data presented here, we conclude that the higher molecular weight species (MW~51 kDa) of PyGAPDH observed in the soluble fraction and in 2DE of cytosol[5] arose due to ubiquitination of native PyGAPDH. The ~51 kDa band in the soluble fraction that is visualized by both antibodies (anti-PfGAPDH and anti-Ub) has an addition of mass of ~15-17 kDa to the native PyGAPDH. Conjugation of two molecules of ubiquitin (MW~8.5 kDa) to PyGAPDH can give rise to such species. 

4.2.  Detection and Sequence Coverage ofPyGAPDHas Analyzed By LC-ESI-Q-TOF-MS 

Identity of the three protein bands observed in Western blot analysis was further confirmed by the mass spectrometric analysis of peptides obtained from in-gel tryptic digestion of the three PyGAPDH species. Extracted peptides were separated on a reverse phase C-18 nano-chip and as the peptides eluted, MS and MS/MS spectra were acquired. The lists of matched m/z peptides for various fractions are presented in Table 1(A) to (D). 

Table 1: Analysis of Post Translational Modifications in peptides derived from tryptic digests of GAPDH positive bands as shown in Figure 1. Three bands from the soluble fraction (with MW ~ 27, 37 and 51 kDa) and 37kDa band of particulate fraction were individually digested with trypsin and peptides were sequenced using MS/MS. All peptides that were derived from GAPDH and had post-translational modifications are listed below.

Sequence coverage for soluble fraction GAPDH bands was in the range of 76-87%. Since in each case two independent samples were analyzed, in final tally the sequence coverage was 313 out of 337 residues (92.9%). Several peptide m/z matched by inclusion of certain PTMs defined as fixed and variable. For insoluble fraction, only ~37 kDa band was analyzed (Table 1(D)). The sequence coverage obtained was ~26% that largely covered the C-terminal half of the molecule. Generally high sequence coverage is obtained for the soluble proteins as compared to the membrane bound forms [17]. However, our expectation was to obtain much greater sequence coverage similar to soluble fraction (87% coverage; Table 1(B)). Membrane association of PyGAPDH is likely to be mediated through post-translational modifications involving membrane anchoring groups such as prenyl, palmitoyl or GlycosylPhosphatidylInositol (GPI) etc. or those that facilitate its binding with other membrane proteins. Lack of N-terminal peptides in ~37 kDa band from particulate fraction could arise because of post-translational modifications with membrane associating hydrophobic groups. Such regions may not be cleaved by trypsin or such peptides may not have eluted from C-18 chips that we used in our chromatographic separation. Recently, the possibility of N-terminal being palmitoylated to translocate GAPDH1 to cellular cortex in Toxoplasma gondii has been suggested [18]. Although the soluble fraction showed extensive coverage, certain stretches of sequence did not get covered. These consisted of¹⁴IGRLVFRSAQER²³, ¹⁹⁵GGKDWRAGR²⁰³ and ²⁶¹VAK²⁶³. Trypsin digestion of these segments will generate peptides that are too small in size and could have been missed detection.Thus, MS data presented in Table1 provided direct evidence that all the three different molecular mass species detected by anti-rPfGAPDH antibodies indeed contained GAPDH. 

4.3.  Identification of Post-Translational Modifications (PTMs)

Matched m/z in MS spectra led to identification of several peptides that have undergone modifications (marked in bold in Table 1).MS/MS spectra of all these peptides were manually verified and peptides that passed our acceptance criteria were selected. The peaks in MS/MS spectra were assigned to b and y ions and wherever possible, spectra for modified and unmodified forms of the peptide were compared to locate the modified residue and the PTMs. This approach is far superior and yields more reliable results as compared to most of the algorithms that automatically identify PTMs. Modified peptides identified with confidence in various fractions are listed in Table 2 along with the residue(s) (in bold) that have undergone the modification. The PTM search in PyGAPDH was set for phosphorylation of Ser, Thr and Tyr with the residue acquiring additional mass of 80 Da (∆m = 80 Da) or a neutral loss of 98 Da (∆m = -98 Da), acetylation (∆m = 42 Da), methylation (mono ∆m = 14 Da; dimethylation ∆m = 28 Da and trimethylation ∆m = 43 Da) and ubiquitination (∆m =114 Da) of Lys and nitrosylation (∆m = 29 Da) of Cys. Addition of 80 Da in mass also occurs on sulfation of tyrosine[19]. 

For making distinction between tyrosine phosphorylation or sulfation, more extensive experiments will be needed [20]. Here, we assumed phosphorylation as the modifying group. Figure 3 shows a few representative MS/MS spectra of the peptides in their native and modified forms.

Figure 3: Some representative MS/MS spectra of peptides in their modified and unmodified forms. Parent ion m/z and Retention Times (RT) are stated. (A) phosphorylation- unmodified (parent ion m/z = 476.2814³⁺; RT=22.6 minutes) and modified (parent ion m/z = 714.4139⁴⁺; RT=22.71 minutes); (B) acetylation- unmodified (parent ion m/z = 405.2241²⁺; RT=12.06 minutes) and modified (parent ion m/z = 681.3600³⁺; RT=25.00 minutes) and (C) ubiquitination- unmodified (parent ion m/z = 319.1928⁴⁺; RT=13.70 minutes) and modified (parent ion m/z = 347.7039⁴⁺; RT=13.31 minutes). Insets (blue) mark the peaks that account for unmodified and modified residue masses. 

In all twelve residues were identified that underwent modifications. Different modifications included four phosphorylations (pS¹⁴⁴, pS²⁰⁴, pS²¹³ and pT¹⁴⁶), two methylations (mK²¹⁸ and mK²³⁰) and a dimethylation (m2K²³⁰⁾, three acetylations (acK¹⁶³, acK²³⁰ and acK³⁰¹), one nitrosylation (nC¹⁵⁷) and five ubiquitinations (uK⁷³, uK²¹⁸, uK²²², uK²³⁰and uK³³⁶) (Table 2). Some PTMs were detected in more than one band (Table 3). 

Figure S1 has all the data and corresponding MS/MS spectra for the peptides listed in Table 2.Examination of missed cleavage pattern among modified lysine residues indicated that trypsin could cut at C-terminal end of mono-methylated lysine (e.g. mK²³⁰) but failed to cleave dimethylated residues. Trypsin cleavage at ubiquitinated lysineswas also observed (e.g. SALLNIIPASTGAAKAVGuK²²²). A non-tryptic peptide (-LLDLAIHITKH-) that showed ubiquitination at K³³⁶ was present in the particulate fraction. This peptide is a product of the C-terminal end of the protein. 

Peptide 1

Peptide 2

Peptide 3

Peptide 4

Peptide 5

 Peptide 6

Peptide 7

Peptide 8

Peptide 9

Peptide 10

Peptide 11

Peptide 12

Peptide 13

 Peptide 14

Peptide 15

Peptide 16

Peptide 17

Peptide 18

5.      Discussion 

It is increasingly being realized that post-translational modifications and combinations thereof play an important role in determining sub-cellular distribution and functions of a protein[21]. Mass spectrometry has become the tool of choice to obtain precise chemical structures of various protein species at single protein level[9,21,22]. The ultimate objective of defining precise chemical structure is to correlate it to cellular function [4a,4b,4d]. Work presented in this manuscript has used single protein analysis approach to obtain solid chemical data about the PTMs in PyGAPDH.

Western blot analysis using anti-rPfGAPDH antibodies showed the presence of PyGAPDH in three different sizes in the parasite cell extracts. These were further confirmed by MS and MS/MS analysis of peptides obtained by trypsin digestion of the proteins. Origin of ~51 kDa form present in cytosol was found to be due to ubiquitination of the native 37 kDa species of PyGAPDH. Extensive MS/MS sequencing of peptides derived from the three bands led to identification of several PTMs in PyGAPDH. At least twelve different residues in PyGAPDH were modified with five different kinds of chemical modifications. Most modifications mapped to the C-terminal domain of the protein.There were eleven modifications in the C-terminal half while N-terminal half had only five (Table 3). 

Residues 201-237 had most modifications. Functionally, N-terminal domain has the NAD⁺ binding site and the C-terminal domain forms a glyceraldehyde 3-phosphate binding site. A flexible S-loop (extending from residue 180 to 210) is believed to be the region that transmits structural changes induced by substrate binding to neighboring subunits (allosteric regulation) [18,23]. Examination of the pattern of PTMs in various regions of the molecule could provide some insightful information about the regulation of underlying physiological processes (Figure 4).

There are two serine residues in the peptide 201-237. Both underwent phosphorylation. However, these phosphorylations were exclusive of each other i.e. in a given molecule only one of the two was phosphorylated. Further, a peptide containing pS²⁰⁴- did not show any modifications at K²¹⁸ or K²²². However, when S²¹³had phosphorylation (pS²¹³), one or more modifications at the two lysines were observed. Further, we observed that K²¹⁸ could either be methylated or ubiquitinated, but these modifications probably required phosphorylation of S²¹³. A serine residue in Toxoplasma gondii (S²⁰³) that is homologous to S²⁰⁴ofP. yoeliialso undergoes phosphorylation[24]. This modification has been implicated in the regulation enzyme activity presumably by interfering with oligomerization and allosteric activation[18].P. yoeliias well as T. gondii have two neighboring residues (S²¹³ and T²¹⁴) that can be phosphorylated. Interestingly S²¹³is phosphorylated in P. yoeliiwhile T²¹⁴ is phosphorylated in T. gondii[24]. Thus, there appears to be a high degree of conservation in PTMs among the related organisms. K²³⁰ is rather unusual in undergoing four different types of modifications i.e. mono or dimethylation / acetylation / ubiquitination.Although all three lysines in peptide 201-237 could undergo ubiquitination, modification occurred only at one of these lysines. In another peptide covering residues 141-163, phosphorylations of S¹⁴⁴ and T¹⁴⁶were observed. These were also exclusive of each other. It is likely that phosphorylation of either of these two residues could mediate the similar physiological function(s). Two mono-methylations at K²¹⁸and K²³⁰ were detected. These could occur in the same molecule or in different molecules individually. Both these lysines are conserved in all four-species compared here (Figure S2).

Figure S2: Sequence homology among P. falciparum, P. yoeliiand human GAPDH. Analysis of phosphoproteome of P. falciparumled toidentification of four sites. Modified residues detected in PyGAPDH are marked. S¹⁴⁴, S²⁰⁴, S²¹³ and T¹⁴⁶ that undergo phosphorylation in PyGAPDH, only S²¹³ is conserved in P. falciparum.

Dimethylation of K²³⁰ was also detected. In a recent study, methylated lysine proteome of blood stage P. falciparum was analyzed.However, in this proteome-wide lysine methylation analysis,trimethylation of K⁸⁰(⁷⁴VSVFAEKDPSQIPGW⁸⁸)[25]was the only modification reported. This residue in P. yoeliiGAPDH is replaced by R⁸⁰. All methylated residues detected in P. yoeliiare conserved in all four species of Plasmodia (Figure S2) suggesting that such methylations are likely to be present in Plasmodium falciparum GAPDH too. 

For understanding the functional significance of PTMs, it is essential to determine combinations of various PTMs that occur together and the spatio-temporal distribution of each distinct chemical species inside the cell. In the absence of such information, functional implications of PTMs will be difficult to establish. PTMs such as phosphorylations are one important mechanism by which the parasite controls the process of invasion and modification ofthe host cells[26]. Phosphorylation of four different residues viz. pS¹⁴⁴, pT¹⁴⁶, pS²⁰⁴ and pS²¹³ were observed in P. yoelii17XL GAPDH. In P. falciparum 3D7 GAPDH, four phosphorylation sites (pS⁷⁵, pS²¹³, pT²¹⁴ and pT²⁸⁰) have been identified[26]. Although three of these four residues (S²¹³, T²¹⁴ and T²⁸⁰) are conserved in both species, only phosphorylation of S²¹³ is observed in both species. Residues that are phosphorylated only in P. yoelii(pS¹⁴⁴, pT¹⁴⁶ and pS²⁰⁴) are not conserved in P. falciparum (K¹⁴⁴, L¹⁴⁶and C²⁰⁴). Such variation may imply species-specific physiological roles for different modifications. 

There are two Cys residues in the active site of PyGAPDH viz. C¹⁵³ and C¹⁵⁷. One of these was found to be nitrosylated. Nitrosylation of GAPDH has been reported in macrophages. In mammalian cells, cysteine residue at the catalytic site (C¹⁵²) undergoes nitrosylation that triggers the binding of GAPDH to Siah-1 (an E3 ubiquitin ligase) followed by nuclear translocation and apoptosis[1d]. This cascade of S-nitrosylated-GAPDH and Siah-1 may represent an important molecular mechanism of apoptotic cell death [1d,1e]. The catalytic Cys (C¹⁵³and C¹⁵⁷) are not only conserved in PyGAPDH, but one of these also undergoes nitrosylation (nC¹⁵⁷). This raises the possibility that nitrosylation of C¹⁵⁷may play a role in nuclear localization of PyGAPDH. 

Observations of modified lysine residues at C-terminus of tryptic peptides (e.g. mK²³⁰ and uK²²²) indicate that trypsin does cut at modified lysines. This was in contrast to earlier belief that such modified residues were not cleaved by trypsin. Cleavage at ubiquitinated lysines was also observed in the analysis of ubiquitome of MCF-7 breast cancer cells [27]. In all, five residues (K⁷³, K²¹⁸, K²²², K²³⁰ and K³³⁶) were detected that were ubiquitinated of which K²²² and K³³⁶ were present in ~37 kDa as well as ~51 kDa species (Table 3) while K²¹⁸and K²³⁰ were restricted to ~51 kDa species only. 51 kDa species could arise either by conjugation with a diubiquitin moiety or by bi-ubiquitination of the native 37 kDa form. Since tagging a protein for proteasomal degradation requires a K⁴⁸ or K¹¹ linked chain of>4-5 Ub subunits attached to a protein[28], it is unlikely that ubiquitinations observed here served as a signal for protein degradation. Di (or Bi) ubiquitinations of PyGAPDH are likely to have some other regulatory function(s) essential for maintaining the cellular homeostasis. Monoubiquitination has been shown to play a role in endocytic pathways and in some cases, single monoubiquitination was sufficient for internalization of the membrane proteins [29]. Our observation of ubiquitination in low MW forms of PyGAPDH (ubiquitination at K²²² and K³³⁶) would suggest their origin from 51 kDa form by limited proteolysis. There have been a few reports about ubiquitination of parasite proteins that are likely to be important in functions other than tagging for proteasomal degradation, e.g. actin [30], histone H2B [31] and enolase[9]. A recent study on P. falciparumubiquitome from erythrocytic stages led to identification of 73 different proteins [32] that included the three proteins mentioned above. 

6.      Conclusions 

Results presented here provide evidence for multiple structural modifications in Plasmodium spp. GAPDH that could easily account for several moonlighting functions that this protein may have [33]. The main objective for the identification of PTMs was to define precise chemical structure of each species and understand their functions. This task of exact correlation between structural variant and its function and/or sub-cellular localization remains yet to be accomplished. 

7.      Authors declare no conflict of interest 

8.      Author contribution 

GKJ conceived and designed the experiments. NJ, CB and SD performed the experiments and collected data. NJ, CB, SD and GKJ analyzed and interpreted the data. GKJ wrote the manuscript. All the authors have read the manuscript and agreed with its content.

 9.       Funding 

Financial support for the research work reported here was provided by a grant to GKJ from the Department of Atomic Energy, Government of India.

Graphical Abstract

Figure 1: (i) Coomassie stained SDS-PAGE of P. yoeliiwhole cell extract (WCE) and the two fractions i.e. soluble and particulate. Fractionation was done by centrifugation at 40,000g for 30 minutes. (ii) Western blot of three fractions probed using anti-rPfGAPDH antibodies. Note the presence of PyGAPDH in three different sizes with MW~27, 37 and 51 kDa. Absence of 51 kDa species in particulate fraction is quite evident.

Figure 2: Antibody pull-down assay to determine ubiquitination of PyGAPDH. (A) Fractionation scheme for the preparation of soluble and particulate fractions. (B) (i) Soluble fractions showed three major species of PyGAPDH at MW~27, 37 and 51 kDa. (ii) Probing a similar blot with anti-Ub antibody showed ~51 kDa band to be an ubiquitinated form of PyGAPDH (dotted box). *Represents the ~37 kDa form of PyGAPDH which showed a faint signal for ubiquitination. (iii) Blot showing absence of ubiquitination in PyGAPDH associated with particulate fraction of P. yoeliicell extract.

Figure 3: Some representative MS/MS spectra of peptides in their modified and unmodified forms. Parent ion m/z and Retention Times (RT) are stated. (A) phosphorylation- unmodified (parent ion m/z = 476.2814³⁺; RT=22.6 minutes) and modified (parent ion m/z = 714.4139⁴⁺; RT=22.71 minutes); (B) acetylation- unmodified (parent ion m/z = 405.2241²⁺; RT=12.06 minutes) and modified (parent ion m/z = 681.3600³⁺; RT=25.00 minutes) and (C) ubiquitination- unmodified (parent ion m/z = 319.1928⁴⁺; RT=13.70 minutes) and modified (parent ion m/z = 347.7039⁴⁺; RT=13.31 minutes). Insets (blue) mark the peaks that account for unmodified and modified residue masses.

Figure 4: Schematic representation of modifications in various peptides. (A)Peptide containing residues from 201-237 and (B) peptide containing residues 141-163 showing the PTMs that occur in combinations or in exclusion of each other in PyGAPDH.

Table 1A: MW~51 kDa species from soluble fraction (data from 2 independent samples): Protein Score: 1246; Sequence coverage: 76%.

Table 1B: MW~37 kDa species from soluble fraction (data from 2 independent samples): Protein Score: 2116; Sequence coverage: 87%.

Table 1C: MW~27 kDa species from soluble fraction (data from 2 independent samples): Protein Score: 918; Sequence coverage: 85%.

Table 1D: MW~37 kDa species from particulate fraction: Protein Score: 149; Sequence coverage: 26%.

Table 2: Post-translational modifications in PyGAPDH. List of validated PTMs with peptide sequence. Residues modified are marked in bold.

Table 3: List of residues in PyGAPDH that undergo post-translational modifications (PTMs). Some residues showed multiple modifications. Data are from Table 2.

         1.(a) Campanale N, Nickel C, Daubenberger CA, Wehlan DA, Gorman JJ, et al.(2003) Identification and characterization of heme-interacting proteins in the malaria parasite, Plasmodium falciparum. J BiolChem 278: 27354-27361.
  (b) Sirover MA (1999) New insights into an old protein: the functional diversity of mammalian glyceraldehyde-3-phosphate dehydrogenase. BiochimBiophysActa 1432: 159-184.
  (c) Sirover MA (2011) On the functional diversity of glyceraldehyde-3-phosphate dehydrogenase: biochemical mechanisms and regulatory control.BiochimBiophys Acta 1810: 741-751.
  (d) Hara MR, Agrawal N, Kim SF, Cascio MB, Fujimuro M, et al. (2005) S-nitrosylated GAPDH initiates apoptotic cell death by nuclear translocation following Siah1 binding. Nat Cell Biol 7: 665-674.
   (e) Tristan C, Shahani N, Sedlak TW, Sawa A (2011) The diverse functions of GAPDH: Views from different subcellular compartments. Cell Signal 23: 317-323.
2.   Thiede B, Koehler CJ, Strozynski M, Treumann A, Stein R, et al. (2013) High resolution quantitative proteomics of HeLa cells protein species using stable isotope labeling with amino acids in cell culture(SILAC), two-dimensional gel electrophoresis(2DE) and nano-liquid chromatography coupled to an LTQ-OrbitrapMass spectrometer. Mol Cell Proteomics 12:529-538.
3.   (a) Tisdale EJ, Azizi F, Artalejo CR(2009) Rab2 utilizes glyceraldehyde-3-phosphate dehydrogenase and protein kinase C{iota} to associate with microtubules and to recruit dynein. J BiolChem 284: 5876-5884.
    (b) Tisdale EJ, Kelly C, Artalejo CR (2004) Glyceraldehyde-3-phosphate dehydrogenase interacts with Rab2 and plays an essential role in endoplasmic reticulum to Golgi transport exclusive of its glycolytic activity. Journal of Biological Chemistry 279: 54046-54052.
4.  (a) Jungblut PR(2013) Back to the future--the value of single protein species investigations. Proteomics13 :3103-3105.
(b) Jungblut PR (2008) The speciation of the proteome. Chem Cent J 2: 16.
(c) Jungblut PR(2014) The proteomics quantification dilemma. J Proteomics107: 98-102.
(d) JungblutPR (2016) Towards deciphering proteomes via the proteoform, protein speciation, moonlighting and protein code concepts. J Proteomics 134: 1-4.
5.  Sangolgi P, Balaji, Dutta S, Jindal N, Jarori GK (2016) Cloning, expression, purification and characterization of Plasmodium spp. glyceraldehyde-3-phosphate dehydrogenase. Protein ExpresPurif 117: 17-25.
6. (a) Daubenberger CA, Poltl-Frank F, Jiang G, LippCerta J,certa U, et al.(2000) Identification and recombinant expression of glyceraldehyde-3-phosphate dehydrogenase of Plasmodiumfalciparum. Gene 246: 255-64.
(b) Daubenberger CA, Tisdale EJ, Curcic M, Diaz D, Silvie O, et al.(2003)The N'-terminal domain of glyceraldehyde-3-phosphate dehydrogenase of the apicomplexan Plasmodiumfalciparummediates GTPase Rab2-dependent recruitment to membranes. BiolChem 384: 1227-1237.
7.  Pal-Bhowmick I, Andersen J, Srinivasan P, Narum DL, Bosch J, et al. (2012) Binding of aldolase and glyceraldehyde-3-phosphate dehydrogenase to the cytoplasmic tails of Plasmodium falciparummerozoiteduffy binding-like and reticulocyte homology ligands.MBio, 3.
8. Cha S, KimJ (2016) Identification of GAPDH on the surface of Plasmodium sporozoites as a new candidate for targeting malaria liver invasion. Journal of Experimental Medicine, 213 (10): 2099-2112.
9.  Shevade S, Jindal N, Dutta S, Jarori GK (2013) Food vacuole associated enolase in plasmodiumundergoes multiple post-translational modifications: evidence for atypical ubiquitination. PLoS One 8: edn72687.
10.  Laemmli UK,(1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685.
11.  (a) ShevchenkoA,Jensen ON, Podtelejnikov AV, Sagliocco F, Wilm M, et al.(1996) Linking genome and proteome by mass spectrometry: large-scale identification of yeast proteins from two dimensional gels. Proc Natl AcadSci USA 93: 14440-14445
(b) GharahdaghiF,Weinberg CR(1999) Mass spectrometric identification of proteins from silver-stained polyacrylamide gel: a method for the removal of silver ions to enhance sensitivity. Electrophoresis, 20 :601-605.
12.  Perkins DN, Pappin DJ, Creasy DM, Cottrell JS (1999) Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20:3551-3567.
13.  Vizcaino JA, Deutsch EW, Wang R, Czardas A, Reisinger F, et al. (2014) Proteome change provides globally coordinated proteomics data submission and dissemination. Nat Biotechnology 32: 223-226.
14.  Vizcaino JA, Cote RG, Czardas A, Dianes JA, Fabregat A, et.al. (2013)PeTheProteomicsIdentifications (PRIDE) database and associated tools: status in 2013. Nucleic Acids Res 41:1063-1069.
15.  Cote RG, Griss J, Dianes JA, Wang R, Wright JC, et al. (2012) The ProteomicsIdentification (PRIDE) Converter 2 framework: an improved suite of tools to facilitate data submission to the PRIDE database and the Proteome change consortium. Mol Cell Proteomics 11: 1682-1689.
16.  Wang R, Fabregat A, Rios D, Ovelleiro D, Foster JM,et al.(2012)pride Inspector: a tool to visualize and validate MS proteomics data. Nat Biotechnol 30: 135-137.
17.  (a)Wu C C, MacCoss MJ, HowellKE,Yates JR  (2003) A method for the comprehensive proteomic analysis of membrane proteins. Nature Biotechnology 21: 532-538.
(b) WuCC and Yates JR (2003) The application of mass spectrometry to membrane proteomics. Nature Biotechnology 21: 262-267.
18.  Dubey R, Staker BL, Foe IT, Bogyo M, Myler PJ, et al. (2017) Membrane skeletal association and post-translational allosteric regulation of Toxoplasma gondii GAPDH1. MolMicrobiol103: 618-634.
19.   Monigatti F, Hekking B, Steen H (2006) Protein sulfation analysis--A primer. BiochimBiophysActa1764 : 1904-1913.
20.  Itkonen O, Helin J, Saarinen J, Kalkkinen N, Ivanov KI,et.al. (2008) Mass spectrometric detection of tyrosine sulfation in human pancreatic trypsinogens, but not in tumor-associated trypsinogen. FEBS J 275: 289-301.
21.  Jeffery CJ (2016) Protein species and moonlighting proteins: Very small changes in a protein's covalent structure can change its biochemical function. J Proteomics 134: 19-24.
22.   Locke, Bian, Harris (2009) Post-translational modifications of connexin26 revealed by mass spectrometry. Biochemical Journal 424: 385-398.
23. Biesecker, Harris, Thierry, Walker, Wonacott, et al. (1977) Sequence and Structure of D-Glyceraldehyde 3-Phosphate Dehydrogenase from Bacillus-Stearothermophilus. Nature 266: 328-333.
24.  Treeck M, SandersJL, EliasJE, Boothroyd JC (2011) The PhosphoproteomesofPlasmodiumfalciparumand Toxoplasma gondii Reveal Unusual Adaptations Within and Beyond the Parasites' Boundaries. Cell Host & Microbe 10: 410-419.
25.  Kaur I, Zeeshan M, Saini E, Kaushik, Mohammad A,et al.(2016) Widespread occurrence of lysine methylation in Plasmodiumfalciparumproteins at asexual blood stages. Sci Rep-Uk6.
26.  Treeck M, Sanders J L, Elias JE, Boothroyd JC (2011) The phosphoproteomes of Plasmodiumfalciparumand Toxoplasma gondii reveal unusual adaptations within and beyond the parasites' boundaries. Cell Host Microbe 10: 410-419.
27.  Denis NJ, Vasilescu J, Lambert JP, Smith JC, Figeys D (2007) Tryptic digestion of ubiquitin standards reveals an improved strategy for identifying ubiquitinated proteins by mass spectrometry. Proteomics 7: 868-874.
28.  Pickart CM andFushmanD(2004) Polyubiquitin chains: polymeric protein signals. CurrOpinChemBiol 8: 610-616.
29.   (a)Haglund K, Di Fiore PP, Dikic I (2003) Distinct monoubiquitin signals in receptor endocytosis. Trends BiochemSci 28: 598-603.
(b) HickeL(2001) Protein regulation by monoubiquitin. Nat Rev Mol Cell Biol 2: 195-201.
(c) Mukhopadhyay Dand Riezman H (2007) Proteasome-independent functions of ubiquitin in endocytosis and signaling. Science 315: 201-205.
(d) Saksena S, Sun J, Chu T, Emr SD (2007) S DESCRTing proteins in the endocytic pathway. Trends BiochemSci 32: 561-573.
30.  Field S J, Pinde JC, Clough B, Dluzewski AR, Wilson RJ,et al. (1993) Actin in the merozoite of the malaria parasite, Plasmodiumfalciparum. Cell Motil Cytoskeleton 25:43-48.
31. (a)Trelle MB,Salcedo-Amaya,Cohen AM,Stunnenberg HG,Jensen ON (2009) Global histone analysis by mass spectrometry reveals a high content of acetylated lysine residues in the malaria parasite Plasmodiumfalciparum. J Proteome Res 8:3439-3450.
(b) Picchi, Zulkievicz, Krieger, Zanchin (2017) Post-translational Modifications of Trypanosoma cruzi Canonical and Variant Histones. Journal of Proteome Research 16: 1167-1179.
32.   Ponts N, Saraf A, Chung DW, Harris A, Prudhomme J, et al (2011) Unraveling the ubiquinone of the human malaria parasite. J BiolChem, 286: 40320-40330.
33.   Sriram G, and Martinez (2005) Single-gene disorders: what role could be moonlighting enzymes play? Am J Hum Genet 76: 911-924.