Crystal structure of N-terminal degron-truncated human glutamine synthetase

The crystal structure of N-terminal degron-truncated human glutamine synthetase was determined at 2.95 Å resolution, which revealed that the N-degron is not essential for decamer formation. The study also investigated the roles of N-degron in oligomerization and enzymatic activities through biochemical analyses.

Abstract

Glutamine synthetase (GS) is a decameric enzyme that plays a key role in nitrogen metabolism. Acetylation of the N-terminal degron (N-degron) of GS is essential for ubiquitylation and subsequent GS degradation. The full-length GS structure showed that the N-degron is buried inside the GS decamer and is inaccessible to the acetyltransferase. The structure of N-degron-truncated GS reported here reveals that the N-degron is not essential for GS decamer formation. It is also shown that the N-degron can be exposed to a solvent region through a series of conformational adjustments upon ligand binding. In summary, this study elucidated the dynamic movement of the N-degron and the possible effect of glutamine in enhancing the acetylation process.

1. Introduction  

Glutamine synthetase (GS; EC 6.3.1.2) is involved in many physiological processes by catalyzing the formation of glutamine through the condensation of ammonia and glutamate with the concomitant hydrolysis of ATP (Fig. 1 ▸ a). This conversion plays a role in a variety of cellular functions such as the detoxification of ammonia in the liver (Taylor & Curthoys, 2004 ▸), the protection of neurons against excitotoxicity and the regulation of pH in the kidneys. Additionally, GS plays a critical role in mTOR signaling, translation and autophagy to coordinate cell growth and proliferation (Nicklin et al., 2009 ▸). Given its importance in a plethora of cellular functions, it is perhaps not surprising that GS is well conserved among animals, plants and yeast. Generally, GSs can be grouped into three different classes, GS I, II and III, based on the species of origin and the multimeric arrangement. GS I and III enzymes are commonly observed in prokaryotes, including bacteria and archaea, where they tend to form dodecamers in solution (Eisenberg et al., 2000 ▸). GS II enzymes, on the other hand, are commonly found in eukaryotes, such as humans, and tend to form decamers in solution comprising two pentamers (Krajewski et al., 2008 ▸).

Figure 1.

Catalytic reaction of glutamine synthetase and domain characterization. (a) Catalytic reaction of glutamine synthetase. (b) Schematic representation of the domains of human glutamine synthetase.

Human GS belongs to a class of proteins that possess an N-terminal degron (N-degron), which acts as a degradation signal that triggers intracellular protein degradation mediated by the ubiquitin (Ub)–proteasome system (UPS; Varshavsky, 2019 ▸; Nguyen et al., 2016 ▸). The N-degron-containing region of GSs comprises an N-terminal meander (residues 1–24) followed by β-grasp (residues 25–112) and catalytic (residues 113–373) domains (Fig. 1 ▸ b). Previous structural studies have tended to focus on prokaryotic and plant GSs. To date, only two reports have described crystal structures of human GS complexes, one comprising full-length human GS bound to ADP and a phosphorylated form of the inhibitor methionine sulfoximine (MSO) at 2.6 Å resolution and the other comprising full-length human GS bound to ADP at 2.05 Å resolution (Krajewski et al., 2008 ▸). Recently, Nguyen and coworkers revealed that high concentrations of glutamine promote the acetylation of Lys11 and Lys14 in the N-degron of human GS. This glutamine-dependent lysine acetylation of the N-degron is catalyzed by P300/CBP acetyltransferase and triggers a series of events that includes E3 ubiquitin ligase (Cul4–Rbx–DDB1–CRBN complex)-dependent ubiquitination and subsequent GS degradation by proteasomes and valosin-containing protein (VCP)/p97 (Nguyen et al., 2016 ▸, 2017 ▸). These discoveries have provided a greater understanding of human GS regulation.

The N-degron of GS was thought to be involved in pentamer formation, as the N-degron is embedded in the inner core of the pentameric ring, as shown in the crystal structures of both human and dog GS (Krajewski et al., 2008 ▸; Supplementary Fig. S1). The embedded N-degron makes direct intra- and inter-subunit contacts with multiple subunits of pentameric GS and also participates in contacts with the other GS pentamer protomer. These contacts involve non­polar interactions and polar interactions such as hydrogen bonds, suggesting that the N-degron peptide contributes to GS pentamer formation and stabilization. Burial of the N-degron in the inner core of the pentameric ring prevents it from being acted upon by acetyltransferase and suggests that a structural change in the pentamer is required to facilitate the acetylation of the lysine of the N-degron by projection of the N-degron outside the pentamer cleft.

This study aims to investigate the dynamic features of an N-degron through structural analyses and to determine the effect of mutations on the overall structure and function of human GS. In an effort to characterize the effects of the N-degron on pentamer formation, we determined the crystal structure of N-degron-truncated GS at 2.95 Å resolution. The structure represents an apo form comprising a decamer consisting of head-to-head pentamers in the absence of ADP and the inhibitor MSO. Structural comparison with previously reported structures of nontruncated GS indicated that movement of the N-terminus could be achieved without disruption of decamer formation. Additionally, it was found that mutations of the N-degron did not reduce the in vitro enzymatic activity of GS in comparison to wild-type GS. Altogether, this study revealed that the N-degron was not essential for decamer formation or for GS enzymatic activity. Instead, the flexibility of the N-degron facilitated the degradation of GS upon stimulation by glutamine, a product of GS. The N-terminal polyhistidine tag of N-degron-truncated N-terminal His-tagged human GS was not observed in the electron-density map of the decameric protein.

2. Materials and methods  

2.1. Production and purification of N-terminal degron-truncated glutamine synthetase  

DNA encoding the full-length human glutamine synthetase (1–373) was cloned from cDNA (Invitrogen) and inserted into His-fusion vector pET-47b(+) (Novagen). The 22-residue truncated glutamine synthetase (GS^ΔN) was generated using the full-length wild-type human glutamine synthetase (GS^FL) plasmid as a template, and the integrity of the coding region was verified by DNA sequencing. The plasmids were transformed into Escherichia coli Rosetta2 (DE3) cells (Novagen), and the cells were grown in LB medium supplemented with 50 µg ml⁻¹ kanamycin and 35 µg ml⁻¹ chloramphenicol at 37°C until the OD_600 nm reached a value of 0.6. Protein expression was induced by the addition of isopropyl β-d-1-thiogalactopyranoside (IPTG) to a final concentration of 100 µM after the cells had been placed in an ice bath for 15 min. The culture was then incubated for 20 h at 20°C and the cells were harvested by centrifugation. The cells were then suspended in lysis buffer consisting of 20 mM Tris–HCl pH 8.0, 500 mM NaCl, 10 mM imidazole and disrupted by sonication in an ice bath. The soluble fraction was separated by ultracentrifugation and loaded onto an Ni–NTA agarose column (Qiagen). The column was washed with buffer consisting of 20 mM Tris–HCl pH 8.0, 100 mM NaCl, 20 mM imidazole and the target protein was eluted using the same buffer containing 250 mM imidazole in lieu of 20 mM imidazole. The column effluent was dialysed against 20 mM Tris–HCl pH 8.5, 50 mM NaCl, 1 mM β-mercaptoethanol at 4°C overnight. The dialysed protein was loaded onto a HiTrap Q anion-exchange column (GE Healthcare) and eluted using a gradient of 0–500 mM NaCl in a buffer consisting of 20 mM Tris–HCl pH 8.5, 1 mM β-ME. The eluted products were pooled and concentrated by centrifugation using an Amicon Ultra 30 000 molecular-weight cutoff filter (Merck Millipore). The concentrated product was finally loaded onto a Superdex 200 gel-filtration column (GE Healthcare) using buffer consisting of 10 mM HEPES–NaOH pH 7.5, 300 mM NaCl, 1 mM TCEP [tris(2-carboxyethyl)phosphine]. Peak fractions were collected and concentrated to 45 mg ml⁻¹ for GS^FL and 80 mg ml⁻¹ for GS^ΔN. SDS–PAGE of the protein samples gave one major band corresponding to 44 kDa for GS^FL and 42 kDa for GS^ΔN. Analysis of the sample using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS; Bruker Daltonics) confirmed that the target proteins were successfully purified without degradation. Samples were frozen in liquid nitrogen and stored at −80°C until use. Information pertaining to macromolecule production is summarized in Table 1 ▸.

Table 1. Macromolecule-production information.

Source organism	Homo sapiens
Expression vector	pET-47b(+)
Expression host	E. coli Rosetta2 (DE3)
Complete amino-acid sequence of the construct produced†	MAHHHHHHSAALEVLFQGPGEKVQAMYIWIDGTGEGLRCKTRTLDSEPKCVEELPEWNFDGSSTLQSEGSNSDMYLVPAAMFRDPFRKDPNKLVLCEVFKYNRRPAETNLRHTCKRIMDMVSNQHPWFGMEQEYTLMGTDGHPFGWPSNGFPGPQGPYYCGVGADRAYGRDIVEAHYRACLYAGVKIAGTNAEVMPAQWEFQIGPCEGISMGDHLWVARFILHRVCEDFGVIATFDPKPIPGNWNGAGCHTNFSTKAMREENGLKYIEEAIEKLSKRHQYHIRAYDPKGGLDNARRLTGFHETSNINDFSAGVANRSASIRIPRTVGQEKKGYFEDRRPSANCDPFSVTEALIRTCLLNETGDEPFQYKN

^†The underlined sequence is the vector-derived sequence.

2.2. Crystallization  

Preliminary crystallization screening was performed by the vapor-diffusion method at both 4°C and 20°C using commercially available screening kits from Hampton Research and Qiagen. The first crystal of 0.5 mM GS^ΔN was grown in equilibrium against mother liquor consisting of 0.07 M sodium acetate pH 4.6, 5.6% PEG 4000, 30% glycerol (Qiagen) at 20°C. Optimized crystals were grown in equilibrium against a home-made mother liquor consisting of 0.07 M sodium citrate pH 5.1, 10% glycerol, 5.6% PEG 4000 at 20°C. Crystals were cryoprotected in the same mother liquor containing 30% glycerol. Information pertaining to the crystallization of N-terminal degron-truncated GS is summarized in Table 2 ▸.

Table 2. Crystallization.

Method	Hanging-drop vapor diffusion
Plate type	Hampton Research VDX 24-well plate
Temperature (K)	293
Protein concentration (mM)	0.5
Buffer composition of protein solution	10 mM HEPES–NaOH pH 7.5, 300 mM NaCl, 1 mM TCEP
Composition of reservoir solution	0.07 M sodium citrate pH 5.1, 10% glycerol, 5.6% PEG 4000
Volume and ratio of drop	1 µl:1 µl
Volume of reservoir (µl)	200

2.3. Data collection and processing  

All X-ray data were collected at SPring-8, Harima, Japan. During X-ray beam exposure, crystals were flash-cooled and maintained at 100 K using a nitrogen stream. Detailed statistics of the structure determination are shown in Tables 3 ▸ and 4 ▸. All diffraction data were indexed and merged using XDS (Kabsch, 2010 ▸). The phases of the GS^ΔN structure were solved by molecular replacement using Phaser-MR (McCoy et al., 2007 ▸) with the structure of GS^FL (PDB entry 2qc8; Krajewski et al., 2008 ▸) as a search model. The model structure was further refined using phenix.refine (Liebschner et al., 2019 ▸) and the CCP4 suite (Winn et al., 2011 ▸), and was manually adjusted using Coot (Emsley et al., 2010 ▸). The final refined structure was deposited in the Protein Data Bank (PDB) and was assigned the accession code 7evt.

Table 3. Data collection and processing.

Diffraction source	SPring-8 beamline BL44XU
Wavelength (Å)	0.900000
Temperature (K)	100
Detector	Rayonix MX300HE CCD
Crystal-to-detector distance (mm)	380
Rotation range per image (°)	1
Total rotation range (°)	1–360
Exposure time per image (s)	1
Space group	P21
a, b, c (Å)	117.92, 158.87, 118.91
α, β, γ (°)	90, 92.80, 90
Resolution range (Å)	50.00–2.95
Total No. of reflections	698766
No. of unique reflections	91909
Completeness (%)	99.7 (98.9)
Multiplicity	7.6 (7.6)
〈I/σ(I)〉	20.24 (3.09)
R r.i.m.	0.084 (0.828)
Overall B factor from Wilson plot (Å2)	83.45

Table 4. Structure solution and refinement.

Resolution range (Å)	44.43–2.95 (2.98–2.95)
Completeness (%)	99.7
σ Cutoff	F > 1.35σ(F)
No. of reflections, working set	87261 (2730)
No. of reflections, test set	4592 (144)
Final R cryst	0.219 (0.3648)
Final R free	0.250 (0.4129)
No. of non-H atoms	26419
Protein residues	3436
Ligands	0
Waters	0
R.m.s.d.s
Bond lengths (Å)	0.003
Angles (°)	0.566
Average B factor (Å2)	88.8
Ramachandran plot (%)
Favored	96.11
Allowed	3.89
Outliers	0.00
PDB code	7evt

2.4. Structural comparison  

The obtained structure was compared with those of mammalian GSs. Superposition of the structures was performed using LSQKAB (Kabsch, 1976 ▸). The comparison was performed with full-length human GS bound to ADP at 2.05 Å resolution (PDB entry 2ojw) or bound to both ADP and MSO (methionine sulfoximine) at 2.6 Å resolution (PDB entry 2qc8) and with the apo form of dog GS (PDB entry 2uu7) (Krajewski et al., 2008 ▸). Illustrations were prepared using PyMOL (Schrödinger).

2.5. Size-exclusion chromatography (SEC)  

Analytical gel filtration was performed with GS variants at 30 µM using a Superdex 200 (10/30) gel-filtration column in running buffer consisting of 10 mM HEPES pH 7.5, 300 mM NaCl, 1 mM TCEP at 4°C. To investigate the influence of glutamine on the oligomerization of GS, 6 mM glutamine was included in the running buffer. The elution profiles of the GS variants were compared with a standard elution profile. The molecular weights of the eluted peaks were calculated and compared with decameric standard molecular weights.

2.6. In vitro enzymatic assay  

Glutamine synthetase activity assays were performed at 37°C using coupled catalytic reactions that detected the reduction of β-NADH to β-NAD at 340 nm (Kingdon et al., 1968 ▸). The time-course enzyme assay was measured using an UV1900i temperature-controlled spectrophotometer (Shimadzu). The enzyme buffer (1000 µl) consisted of 30 mM imidazole–HCl pH 7.5 (Wako Chemical), 100 mM l-glutamic acid (Wako Chemical), 10 mM ATP (Sigma–Aldrich), 1 mM PEP (Sigma–Aldrich), 60 mM MgCl₂, 20 mM KCl, 40 mM NH₄Cl, 0.3 mM β-NADH (Sigma–Aldrich) and 40 U PK/LDH enzymes (Sigma–Aldrich). The measurement commenced with the addition of GS enzyme at a concentration of 50 nM. One unit of activity corresponds to the conversion of 1.0 mol of l-glutamate to l-glutamine in 1 min at pH 7.5 and 37°C.

3. Results and discussion  

3.1. Protein preparation and structural determination  

In an effort to investigate the potential conformational changes in GS, both N-terminally His-tagged full-length GS (hereafter designated GS^FL) and N-degron-truncated GS (GS^ΔN) were purified. Additionally, an acetylation mimic generated by the introduction of a double mutation into full-length GS comprising Lys11Gln and Lys14Gln (K11Q and K14Q; designated GS^QQ) was also purified for in vitro enzyme assay. The molecular weight of the purified proteins was verified by mass spectrometry and indicated that the proteins were purified without degradation. All of the GS proteins contained the 19-residue N-terminal extension MAHHHH­HHSAALEVLFQGP, which includes a His tag and an HRV3C protease recognition site.

3.2. The N-degron is not essential for decamer formation  

Although the N-degron of GS is embedded in the core of the pentameric ring with direct inter- and intra-subunit interactions (Krajewski et al., 2008 ▸), the N-degron may possess the ability to move out from the core and become exposed to the solvent. We set out to determine the hydrodynamic character of GS^ΔN using a Superdex 200 column and compared the results with those for GS^FL and other mutant GSs. Our analytical size-exclusion chromatography (SEC) revealed a GS^FL peak corresponding to a decamer with a small shoulder on the higher molecular-weight side, indicating that a small fraction exists as an icosamer (Fig. 2 ▸). The high protein concentration (30 µM) in this experiment probably induced decamer–decamer association to produce the icosamer. Intriguingly, GS^ΔN remained as a decamer without dissociation, suggesting that the N-degron may not be essential for decamer formation or for association of the pentameric ring. Next, to generate an acetylation mimic, two acetylated lysine residues, Lys11 and Lys14, of the N-degron were replaced with glutamines to produce the K11Q and K14Q double mutation to neutralize the positive charges of the lysine side chains. In our SEC analysis, the acetylation mimic GS^QQ also forms a decamer.

Figure 2.

All GS enzymes, including the mutants, tended to exist in a decameric form in solution. A small portion corresponds to a dimer of the decamer, as represented by the shoulder peak observed at the 10 ml eluted fraction position. Gel-Filtration Standard (Bio-Rad): peak 1, thyroglobulin (670 kDa); peak 2, γ-globulin (158 kDa); peak 3, ovalbumin (44 kDa); peak 4, myoglobin (17 kDa); peak 5, vitamin B₁₂ (1.3 kDa).

Previously, a high concentration of glutamine was reported to be one of the priming factors for acetylation of the N-degron of GS (Nguyen et al., 2016 ▸). These results suggested that glutamine might play a role as an allosteric inducer to dissociate decameric GS into the monomeric form to expose the N-degron for acetylation. Our SEC experiment, however, showed that the presence of a high concentration (6 mM) of glutamine does not affect decamer formation by GS (Fig. 2 ▸). Although it is possible that glutamine might influence the stability of the pentamer and facilitate release of the N-degron from the inner core of the pentameric ring, GS retains the pentameric ring and forms the decamer. These observations indicate that the presence of the N-degron may not be essential for formation of the pentamer or decamer, and that unknown global and/or local structural changes may be responsible for release of the N-degron from the inside of the pentamer ring.

3.3. Overall structure of N-degron-truncated GS  

The crystal structure of apo-form GS^ΔN was determined at 2.95 Å resolution (PDB entry 7evt, Table 4 ▸). In the crystal GS^ΔN forms a pentamer, which is associated with another pentamer in a head-to-head configuration to form a decamer (the ten GS molecules are referred to as molecules A–J; Figs. 3 ▸ a and 3 ▸ b). Electron-density maps were well defined for residues 23–371, except for residues 70–75 in molecule D, 298–319 and 333–338 in molecule E and 300–307 in all molecules. The vector-derived 19-residue N-terminal tag was not visible in any of the ten subunits. This poor electron density suggests that the N-terminal His-tag regions were disordered in the current structures. The current GS^ΔN structure consists of a β-grasp domain comprising two α-helices and five β-strands, followed by a catalytic domain comprising eight α-helices and eight β-strands (Fig. 3 ▸ c and Supplementary Fig. S1). The GS^ΔN pentamer forms via inter-subunit interactions including inter-subunit β-sheet formation between the β4 strand from the β-grasp domain of one protomer and the β7 strand from the catalytic domain of the neighboring protomer. The antiparallel β4–β7 association resulted in the formation of 13-stranded β-sheets defining the pentameric geometry (Fig. 3 ▸ c). The GS^ΔN decamer comprises two pentamers stacked on top of one another through interactions between the β6–α3 loops from the top pentamer and the nearest β6–α5 loops from the bottom pentamer (Supplementary Fig. S2d ).

Figure 3.

Overall structure of N-degron-truncated human GS (GS^ΔN). (a) Top view of the decamer of GS^ΔN. The dodecamer is formed by the stacking of two pentamers. Each pentamer has five active sites. One of the active-site pockets is indicated by a red circle. (b) Side view of the stacking of pentamers of N-­degron-truncated human GS to form a decamer. (c) Molecule A (green) of GS^ΔN is superimposed onto those of CfGS (pink), GS^ADP (cyan) and GS^ADP+MSO (orange). Each monomer comprises the β-grasp domain followed by the catalytic domain. The secondary-structure elements are labeled. For visualization, all β-strands are colored blue and the side chain of the conserved Glu24 is displayed. (d) A dimer formed by molecules I (green and blue) and J (purple and yellow) of GS^ΔN. The active site is located in the funnel-shaped pocket (red dotted lines) formed by the β-grasp domain associated with the highly curved β-sheet from the catalytic domain of the adjacent subunit.

3.4. Structural comparison  

The current decameric structure of GS^ΔN was compared with previously reported mammalian GS structures such as human GS bound to ADP and MSO (GS^ADP+MSO; PDB entry 2qc8), human GS bound to ADP (GS^ADP; PDB entry 2ojw) and apo-form dog GS (CfGS; PDB entry 2uu7) (Krajewski et al., 2008 ▸). In agreement with previous size-exclusion chromatographic analyses, the crystal structure of GS^ΔN displayed high similarity to the structures of GS^ADP+MSO, GS^ADP and CfGS. Since both GS^ΔN and CfGS are apo forms, the GS^ΔN structure displays higher similarity to CfGS, with a small root-mean-square deviation (r.m.s.d) of 0.94 Å, than to GS^ADP+MSO or GS^ADP, with r.m.s.d.s of 1.33 and 1.92 Å, respectively (Supplementary Figs. S3a , S3b and S3c ). The pentameric geometry of GS^ΔN is essentially the same as that of the full-length GS pentamers, although the GS^ΔN pentamer displays a larger central hole than that of the full-length GS pentamer due to the absence of the N-terminal α-helix formed by the N-degron (Supplementary Figs. S3d , S3e and S3f ).

Since each GS pentamer has five active sites, the stacking of two pentamers to generate the GS decamer yields a total of ten active sites (Fig. 3 ▸ a). The interface forming the active site is built by a β–β interaction between the N-terminal β-grasp domain of one subunit and the highly curved β-sheet of the C-terminal catalytic domain in the adjacent subunit so as to form a funnel-shaped pocket (Krajewski et al., 2008 ▸; Fig. 3 ▸ d). The overall structure of the active-site pocket of GS^ΔN resembles those of GS^ADP+MSO, GS^ADP and CfGS, suggesting that no structural transition of the active site could be induced by the N-degron.

Two lysine residues, Lys11 and Lys14, are well conserved in mammalian GS N-degrons and are subject to acetylation. Upon N-degron binding to the pentamer hole, these lysine residues play a key role by participating in polar interactions. The positively charged Lys11 forms a salt bridge with the negatively charged Asp174 (in the α3 helix of the neighboring molecule) and a hydrogen bond to the main chains of Asp231 and Ser6 (in the α4 helix and η1′ helix of the same molecule, respectively) (Figs. 4 ▸ a and 4 ▸ b). Another key residue, Lys14, was extended into a pocket comprised of the negatively charged side chains of Glu174, Glu177 and Asp321 to electrostatically stabilize N-degron binding. In GS^ADP, a glycerol molecule is bound to the pocket and forms a hydrogen bond to Lys14, which could contribute towards glycerol-mediated interactions to stabilize N-degron binding (Fig. 4 ▸ b; Krajewski et al., 2008 ▸). In our GS^ΔN structure, this pocket is empty and structural changes such as a shift in helices were not observed (Fig. 4 ▸ c). Thus, the helix packing of the pocket providing N-degron binding is sufficiently structurally rigid to maintain the structure without any dynamic structural perturbation by N-degron binding.

Figure 4.

Acetylation of Lys11 and Lys14 of the N-degron exposed to the solvent. (a) In ADP- and MSO-bound GS^MSO (PDB entry 2qc8), Lys11 binds to the main chain of Asp231 from the same molecule and the side chain of Asp174 from the adjacent molecule. Additionally, Asp231 also interacts with Tyr15 of the N-degron from the adjacent molecule, thereby holding the N-degron in the core of the pentamer ring. (b) In ADP-bound GS^ADP (PDB entry 2ojw), a glycerol molecule occupies the space between the N-degron and the catalytic domain. In addition to Asp231 and Asp174, Lys11 also interacts with the main chain of Ser6. Another residue, Lys14, binds to the glycerol molecule. Acetylation of Lys11 and Lys14 will lead to disruption of these interactions and may expose the N-degron to the solvent. (c) In our apo-form GS^ΔN, deletion of the N-degron did not affect the geometry of Lys11- or Lys14-interacting residues. (d) Relative in vitro enzyme activity of wild-type GS (FL), N-degron-deleted GS (ΔN) and K11Q/K14Q double-mutant GS (QQ). Both mutants of GS^ΔN and GS^QQ displayed higher relative activities compared with GS^FL, indicating that the N-degron is not essential for the observed in vitro enzyme activity.

3.5. Enzymatic activity  

We have shown that deletion of the N-degron induces no significant changes in the overall structure of the GS decamer or pentamer as determined by our hydrodynamic and structural data. These data suggest that the N-degron may not have any dynamic effects on enzymatic activity, although acetylation of the conserved lysine residues of the N-degron triggers ubiquitin-dependent degradation of this enzyme in cells by proteasomes (Nguyen et al., 2016 ▸, 2017 ▸). In an effort to clarify a possible role of the N-degron in terms of the enzymatic function of GS, the enzymatic performances of GS^ΔN and GS^QQ were estimated using an in vitro enzymatic assay and compared with that of GS^FL (Fig. 4 ▸ d). Interestingly, GS^ΔN exhibited glutamine synthase activity comparable to or slightly higher than that of GS^FL. Moreover, GS^QQ, the acetylation-mimic form with glutamine replacement of Lys11 and Lys14, exhibited an unexpectedly high activity in comparison to wild-type GS. The mechanisms that are responsible for the enhanced activity observed in our assay are unknown at present. However, we speculate that deletion of the N-degron may contribute to the dynamics of enzyme action by increasing the structural flexibility of the GS decamer/pentamer. The acetylation-mimic N-degron of GS^QQ lacks the positive charges necessary to interact with the negatively charged binding pocket of GS and should be exposed to the solvent region. It is possible that the exposed N-degron is mobile in the solvent region and could induce conformational flexibility of the core body of the GS decamer/pentamer forming the catalytic site. Intriguingly, this speculation is supported by the fact that a portion of GS^QQ eluted as an icosamer in our gel-filtration analysis, indicating potential conformational differences between GS^QQ and GS^WT. In the GS structure, the N-degron directly connects to the β1 strand (residues 26–33), which participates in maintaining the catalytic pocket. It is probable that exposure of the N-degron increases its dynamic mobility. The increased mobility should affect the stability and geometry of the β1 strand, which may contribute to the activity enhancement of GS^QQ and GS^ΔN.

In conclusion, we presented a crystal structure of N-degron-truncated human GS, which revealed that the N-degron is not essential for the maintenance of decamer formation as previously thought. We provide biochemical analyses of the oligomerization using gel filtration, and in vitro enzyme activity indicated that GS^ΔN exists as an enzymatically active decamer formed by the stacking of two pentamers. These findings suggest that acetylation and subsequent ubiquitylation of the N-degron of GS do not diminish its activity. The complete inhibition of GS activity in cells could be accomplished by the removal of GS molecules by ubiquitylation-mediated protein degradation at proteasomes.

Supplementary Material

PDB reference: N-terminal degron-truncated human glutamine synthetase, 7evt

Funding Statement

This work was funded by Takeda Science Foundation; Japan Agency for Medical Research and Development grant JP17gm1010008 to Toshio Hakoshima; Ministry of Education, Culture, Sports, Science and Technology.