Robust and facile purification of full-length, untagged human Nedd4 as a recombinant protein from Escherichia coli

Abstract

Nedd4 is an E3 ubiquitin ligase that has received increased attention due to its role in the maintenance of proteostasis and in cellular stress responses. Investigation of Nedd4 enzymology has revealed a complex enzymatic mechanism that involves intermolecular interactions with upstream E2 conjugating enzymes and with substrates and intramolecular interactions that serve to regulate Nedd4 function. Thus, it is imperative that investigations of Nedd4 enzymology that employ recombinant enzyme be conducted with Nedd4 in its native, untagged form. We report herein an optimized, facile method for purification of recombinant human Nedd4 in its full-length form as a stable and active recombinant enzyme. Specifically, Nedd4 can be purified through a two-step purification which employs glutathione-S-transferase and hexahistidine sequences as orthogonal affinity tags. Proteolytic cleavage of Nedd4 was optimized to enable removal of the affinity tags with TEV protease, providing access to the untagged enzyme in yields of 2–3 mg/L. Additionally, investigation of Nedd4 storage conditions reveal that the enzyme is not stable through freeze-thaw cycles, and storage conditions should be carefully considered for preservation of enzyme stability. Finally, Nedd4 activity was validated through three activity assays which measure ubiquitin chain formation, Nedd4 autoubiquitination, and monoubiquitin consumption, respectively. Comparison of the method described herein with previously reported purification methods reveal that our optimized purification strategy enables access to Nedd4 in fewer chromatographic steps and eliminates reagents and materials that are potentially cost-prohibitive. This method, therefore, is more efficient and provides a more accessible route for purifying recombinant full-length Nedd4.

1. Introduction

Nedd4 serves as an E3 ubiquitin ligase, the third and final member of the ubiquitination signaling cascade [1]. Nedd4 works in conjunction with an E1 activating enzyme and E2 conjugating enzyme to covalently conjugate ubiquitin to itself or to substrate proteins by isopeptide bond formation at acceptor lysine residues. There are several classes of E3 ligases characterized by structural and mechanistic hallmarks. Nedd4 is part of the HECT (homologous to E6AP carboxyl terminus) ligase family and is a multidomain enzyme that contains an N-terminal C2 domain followed by four WW domains and a C-terminal catalytic HECT domain [2–4]. The mechanism of Nedd4 requires multiple steps, including a protein-protein interaction with its upstream E2 conjugating enzyme, transfer of ubiquitin from E2 to Nedd4 by transthioesterification, association with its substrate, and ligation of ubiquitin to the target by isopeptide bond formation. The mechanism of Nedd4 is further complicated by regulatory mechanisms in which its activity is modulated by intramolecular interactions between the HECT domain and upstream C2 domain or WW linker regions [5–16]

Interest in Nedd4 as a ubiquitin ligase has increased as the role of Nedd4 in maintenance of proteostasis and in response to cellular stressors (heat shock, protein aggregation-associated trafficking defects, regulation of proteinopathies, etc.) has become evident [17–25]. This is complemented by increased interest in ubiquitination more broadly as a system that provides novel drug targets and that can be exploited for inducible protein degradation. Due to the complexity of the Nedd4 mechanism and increased interest in Nedd4 enzymology, it is important to have access to a stable, highly pure, untagged form of the protein for in vitro analyses. To this end, we report herein optimization of the plasmid design, expression, and purification of full-length human Nedd4 from Escherichia coli. Specifically, we present a strategy that enables facile purification of Nedd4 as an untagged enzyme, a feature which is essential as the presence of a large fusion tag could interfere with the intra- or intermolecular interactions involved in the regulation and mechanism of Nedd4. The strategy described herein employs two affinity tags for enrichment and a proteolytic cleavage of affinity tags for robust, facile purification of Nedd4 in two chromatographic steps. The presented purification route improves upon previously reported purification methods by yielding Nedd4 in its full-length form as a stable, pure, and untagged recombinant enzyme in fewer chromatographic steps [26]. Further, activity of the recombinant enzyme was confirmed by various ubiquitination activity assays and is comparable with the reported activities of Nedd4 accessed through other purification methods.

2. Materials and methods

2.1. Materials

Competent E. coli strains were purchased from Agilent (BL21(DE3)-CodonPlus-RIL), Thermo Fisher Scientific (DH5α), or New England Biolabs (NEB5α). All chemicals were ordered from Sigma Aldrich or Alfa Aesar unless otherwise indicated and were used as received.

Plasmids:

The plasmid for recombinant expression of human Nedd4, pGEX-5X3-Nedd4, was a gift from Shiaw-Yih Lin via Addgene (Addgene plasmid #45043; http://n2t.net/addgene:45043; RRID:Addg ene_45043). As received, the plasmid contains the gene for human Nedd4 (NM_006154.2; NP_006145.2) with a N-terminal GST fusion and factor Xa protease cut site sequence. Plasmid identity was confirmed by sequencing upon receipt.

2.2. Methods

Site-directed mutagenesis:

The pGEX-5X3-Nedd4 plasmid was modified by two rounds of Q5 site-directed mutagenesis (New England Biolabs) using primers designed with the NEBaseChanger tool (New England Biolabs; https://nebasechanger.neb.com/). The primers were designed for insertion of a hexahistidine affinity tag at the N-terminus of the GST open reading frame and insertion of a TEV protease cut site sequence (amino acid sequence ENLYFQG) at the C-terminus of the GST fusion before the start of the Nedd4 gene. Forward and reverse primers, respectively, for the insertion of a hexahistidine sequence were as follows: 5′-CACCACCCACTCCCCTATACTAGGTATTG-3′ and 5′-ATGATGATGCATGAATACTGTTTCCTGTG-3’. Forward and reverse primers, respectively, for insertion of a TEV protease cut site were 5′-TTTTCAGGGCGCCCTTATGGCAACTTGC-3′ and 5′-TACAGGT TTTCGAATTCGGGGATCCCACG-3’. Use of these primers allows insertion of the hexahistidine sequence immediately after the start codon of the GST-Nedd4 open reading frame and of the TEV cleavage sequence between residues N230 and S231 of the GST-Nedd4 open reading frame (at C-terminus of GST, downstream of Factor Xa cut site; upstream of Nedd4 sequence). Mutagenesis was performed according to the NEB Q5 site-directed mutagenesis protocol and isolated mutants were confirmed by sequencing. Confirmed plasmids were transformed into DH5α competent E. coli for storage.

Protein expression:

The desired plasmid was transformed into BL21(DE3)-CodonPlus-RIL competent E. coli and plated on chloramphenicol/kanamycin double-selection LB agar plates. A starter culture (100 mL LB media) was inoculated with a streak from the transformation plate and grown overnight at 37 °C to saturation. Expression cultures were inoculated at a concentration of 10 mL saturated starter culture per liter expression media and OD₆₀₀ was monitored by UV. At OD₆₀₀ = 0.6, protein expression was induced by addition of isopropyl β-d-1-thiogalactopyranoside (IPTG) to a final concentration of 100 μM. Expression cultures were then moved to 21 °C and growth was continued for 18 h. Following expression, cell pellets were collected by centrifugation (Sorvall SLA-3000, 5,000 rpm (4,225 rcf), 20 min, 4 °C) and were used immediately for purification.

Protein purification:

All expression steps were performed with purification buffer (50 mM Tris, pH 7.4 with 250 mM NaCl) and any additives described herein. Buffers were used at 4 °C and were sterile filtered (0.22 μm nylon filter) prior to use. Purification steps were performed using the AKTA FPLC system (GE Life Sciences).

Expression culture pellets were resuspended in purification buffer (10 mL buffer per liter of expression culture pelleted) supplemented with protease inhibitor cocktail (1X, Bimake), 40 μM phenylmethylsulfonyl fluoride, and lysozyme (10 mg/mL). Cells were lysed using an EmulsiFlex-C5 homogenizer and cell debris was collected by ultracentrifugation (Beckman Coulter 70-Ti rotor, 40,000 rpm (~12,000 rcf), 4 °C, 45 min, < 0.02 Torr vacuum). Following lysis and centrifugation, the lysate supernatant was loaded onto Glutathione Agarose (Genesee Scientific) resin that was pre-equilibrated with purification buffer. Purification via FPLC was performed as follows: 10 column volume wash (purification buffer), 7 column volume elution (purification buffer with 20 mM reduced glutathione). Following elution, fractions containing protein (as indicated by UV signal at 280 nm) were pooled and dialyzed against 4 L dialysis buffer (50 mM Tris, pH 7.4 with 250 mM NaCl with 5% glycerol and 1 mM β-mercaptoethanol) for 16 h. His₆-TEV protease was added into the dialysis tubing with eluted protein. Following overnight dialysis, the dialysate was loaded onto a Ni²⁺-charged Chelating Sepharose Fast Flow (GE Life Sciences) that was pre-equilibrated with wash buffer (purification buffer plus 20 mM imidazole). The purification method was as follows: 100 mL wash followed by gradient elution (20–250 mM imidazole) over 250 mL and 100 mL elution at 250 mM imidazole. Eluted proteins were analyzed by SDS-PAGE and fractions containing desired purified protein were collected, concentrated and buffer exchanged into purification buffer. Protein concentration was determined by Bradford assay and proteins were stored in solution with 40% glycerol at −20 °C.

Immunoblotting ubiquitination activity assay:

In vitro ubiquitination activity assays were conducted as endpoint assays via immunoblot detection using recombinant human E1, E2, and E3 enzymes. Specifically, UBE1, Ubch5a, and Nedd4 were used as the E1, E2, and E3 enzymes, respectively. E1 (100 nM), E2 (1 μM) and E3 (5 μM) were incubated with ubiquitin (100 μM) in reaction buffer (100 mM Tris, 25 mM MgCl_2, 0.1% Tween, pH 8) and reaction was initiated by addition of ATP (2 mM). Reactions (30 μL final volume) were incubated for 1 h at 37 °C and quenched with the addition of SDS-PAGE loading buffer (4X, 10 μL). Samples were heated to 95 °C for 5–10 min and separated by SDS-PAGE. Gels were transferred to methanol-activated PVDF membrane for immunoblotting and ubiquitination activity was detected by blotting with anti-ubiquitin (1:2000, Abcam ab7780) and HRP-conjugated secondary antibody (1:1000, goat anti-rabbit, BioRad 1706515). Signal was detected with ECL reagents as described by the manufacturer (Genesee Scientific).

Colloidal Coomassie detection of autoubiquitination:

E1 (UBE1, 50 nM), E2 (Ubch5a, 1.5 μM) and E3 (Nedd4, 1 μM) were incubated with ubiquitin (50 μM) in reaction buffer (40 mM Tris, pH 7.5, 50 mM NaCl, 5 mM MgCl_2, 0.5 mM TCEP) and reaction was initiated by addition of ATP (2 mM). Reactions (30 μL final volume) were incubated for 0–15 min at 30 °C and quenched with the addition of SDS-PAGE loading buffer (4X, 10 μL). Samples were heated to 95 °C for 5–10 min, separated by SDS-PAGE, and detected by staining with colloidal Coomassie. Densitometry measurements were performed with ImageJ [27].

MALDI-TOF ubiquitin activity assay:

Ubiquitination activity was quantified in a time-dependent manner by MALDI-TOF detection of monoubiquitin consumption according to a procedure adapted from De Cesare et al. [28] E1 (50 nM), E2 (250 nM), and E3 (500 nM) were combined in reaction buffer composed of 0.25 mg/mL BSA in 10 mM HEPES pH 8.5, 10 mM MgCl₂ and 1 mM ATP. Reactions were incubated at 37 °C and initiated with the addition of ubiquitin (10 μM). At desired time intervals, an aliquot (5 μL) was quenched with 10% TFA (1 μL). Following collection of all time points, samples were doped with DHAP matrix and 4 μM ¹⁵N,¹³C-ubiquitin (R&D Systems) as an internal standard (3:1:2 matrix:standard:sample) and spotted on an AnchorChip 384 BC plate (Bruker Daltonics). Samples were analyzed by MALDI-TOF (Bruker Autoflex Speed LRF MALDI-TOF System) using the following automated AutoXecute method: Reflector Positive mode with laser intensity at 80%, Laser Fuzzy Control switched off, and accumulation parameters set to 4000 satisfactory shots in 500 shot steps with movement parameters set to “Walk on Spot”. Spectra were accumulated by FlexControl software and processed using FlexAnalysis software. Monoubiquitin signal was normalized to signal from the heavy isotope derivative and plotted as normalized intensity versus time to determine time-dependent ubiquitination activity of the signaling cascade.

3. Results

Purification strategy:

The purification route described herein employs two chromatographic steps via orthogonal affinity tags to yield full-length human Nedd4 as a stable, active recombinant enzyme. Specifically, Nedd4 was first expressed as His₆-GST-Nedd4. The fusion protein is enriched by glutathione agarose purification followed by His₆-TEV protease cleavage and nickel affinity-based separation of untagged Nedd4 from His₆-TEV and His₆-GST (Fig. 1A).

Fig. 1.

(A) The expression and purification workflow presented herein for isolation of untagged Nedd4 begins with recombinant expression in E. coli followed by glutathione agarose purification, TEV protease cleavage of Nedd4 from affinity tags, and removal of affinity tags and TEV protease via nickel affinity chromatography. (B) Optimized plasmid design provided the desired open reading frame containing His₆-GST-Nedd4 sequence (with TEV protease cleavage site) under control of the tac promoter in the pGEX-5X3 expression construct. (C) Expression of Nedd4 (as the fusion protein His₆-GST-Nedd4) was screened in the BL21(DE3)-CodonPlus-RIL strain of E. coli. After induction at OD₆₀₀ = 0.6, aliquots of expression culture were collected through the expression time-course and analyzed by SDS-PAGE and Coomassie staining in a 10% acrylamide gel (left). Expression with 100 μM IPTG followed by growth at 21 °C for 18 h provided robust expression of the fusion protein, and expression was validated by immunoblotting detection (right) with anti-Nedd4 (Abcam ab27979, 1:1000 dilution). (D) Purification of Nedd4 from the fusion protein occurs through two chromatographic steps. First, His₆-GST-Nedd4 is enriched from crude lysate via glutathione agarose purification (eluted in ~20 mL volume). Following elution, His₆-GST-Nedd4 is collected and incubated with His₆-TEV protease during dialysis to allow cleavage of Nedd4 from the His₆-GST fusion. Finally, untagged Nedd4 is separated from His₆-GST and His₆-TEV via nickel affinity chromatography, affording untagged Nedd4 as a purified enzyme (eluted in ~40 mL volume). Protein purity was monitored via SDS-PAGE throughout the purification process with representative samples are shown above wherein an aliquot (11.25 μL) of protein sample was combined with 4X Laemmli loading buffer (3.75 μL), heated to 95 °C for denaturation, resolved by SDS-PAGE, and visualized by Coomassie staining.

Modification of pGEX-5X3-Nedd4 plasmid for Nedd4 expression in E. coli:

The pGEX-5X3-Nedd4 plasmid was modified by Q5 site directed mutagenesis for incorporation of a second affinity purification handle and an alternate protease cut site. Specifically, a hexahistidine affinity tag (His₆) was inserted by primer-guided mutagenesis at the N-terminus of the GST-Nedd4 open reading frame, and a TEV protease cut site sequence (ENLYFQG) was inserted between the affinity tag and Nedd4 sequence (Fig. 1B). Clones were confirmed for correct sequence by Sanger dsDNA sequencing.

Expression of His₆-GST-Nedd4:

Expression of the desired protein was optimized from previously reported conditions through use of BL21(DE3)-CodonPlus-RIL, a codon enriched E. coli expression strain. As the original construct was cloned from amplified cDNA, it was not codon optimized for expression in bacterial hosts. Use of a codon enriched expression strain allowed for robust expression of His₆-GSTNedd4 as a soluble protein with minimal truncation products. Expression conditions ([IPTG] and expression temperature) were screened and optimized expression conditions were determined to be induction at OD₆₀₀ = 0.6 with 100 μM IPTG followed by 18 h of growth at 21 °C. Nedd4 identity in the optimized expression conditions was confirmed by immunoblotting with anti-Nedd4 (Abcam ab27979; Fig. 1C).

Glutathione agarose purification of His₆-GST-Nedd4:

Expression cultures were harvested by centrifugation and cell pellets were immediately used for purification. Expression from freshly transformed BL21(DE3)-CodonPlus-RIL cells and immediate purification provided a more robust sample of the recombinant protein with minimal degradation as compared to expression from a 40% glycerol stock or purification from a previously frozen cell pellet (data not shown). Purification was initiated by lysis of the expression cell pellet by high-pressure homogenization in the presence of protease inhibitors and lysozyme. Cell lysate was cleared by ultracentrifugation and the supernatant was subjected to Glutathione Agarose resin for affinity separation of His₆-GST-Nedd4. The purification method included 10 column volume wash followed by a 7 column volume elution with 20 mM glutathione in purification buffer. This method resulted in elution of His₆-GST-Nedd4 across 2 to 3 fractions, affording sample of the desired protein with moderate purity in a low volume (Fig. 1D).

Protease cleavage of the affinity tag fusion:

Protease cleavage of the fusion is desired to afford untagged, recombinant Nedd4 for in vitro experiments. Initially, protease cleavage of the fusion protein was conducted with Factor Xa, the protease for which the parent plasmid contained a cut site sequence. However, proteolytic cleavage of the fusion protein with Factor Xa resulted in undesirable degradation (data not shown). The plasmid was thus modified by site-directed mutagenesis for insertion of a TEV protease cut site sequence (ENLYFQG). Insertion of the TEV sequence allowed for specific cleavage of the fusion, affording His₆GST and Nedd4 from the initial fusion protein.

Protease cleavage was conducted during dialysis of His₆-GST-Nedd4 accessed from the first purification step. To this end, His₆-TEV protease (1 mg) was added with His₆-GST-Nedd4 in the dialysis tubing, and 1 mM β-mercaptoethanol (1 mM) was added to the dialysis buffer to improve TEV activity. During the course of dialysis and subsequent purification, slight degradation of Nedd4 was detected (data not shown). To further promote Nedd4 stability, glycerol (5%) was added to the dialysis buffer and subsequent purification buffers. Ultimately, optimization of the dialysis and protease cleavage step provided access to untagged Nedd4 from the fusion protein with high yield (Fig. 1D). Alternatively, if access to GST-Nedd4 as the full fusion protein is desired (e.g. for a GST pulldown experiment), protease cleavage can be excluded from the purification method by not adding His₆-TEV during dialysis, and the full fusion can be carried forward into the second purification step described below.

Nickel affinity chromatography for separation of untagged Nedd4 from His₆-TEV and His₆-GST:

To separate cleaved Nedd4 from the fusion tag and from the TEV protease, nickel affinity chromatography was employed. To this end, the dialysate was loaded onto an equilibrated, Ni²⁺-charged Chelating Sepharose resin. Inclusion of low concentrations of imidazole (20 mM) in the wash buffer prevented non-specific binding to the resin, and gradient elution with imidazole allowed for separation of Nedd4 from the His₆ tagged GST and TEV proteins. Typically, Nedd4 is separated very early in the elution step (between 20 and 50 mM imidazole) while His₆-TEV and His₆-GST are not eluted until 200–250 mM imidazole. In some cases, a trace amount of Nedd4 was eluted during the wash step, but the majority of the purified protein is collected over a small volume early in the elution gradient. We anticipate that reduction of the imidazole concentration in the wash buffer may further improve this method. Cumulatively, this step allows for high resolution in separation of the desired, purified protein from the fusion tag and protease (Fig. 1D).

If purification of the fusion protein (without protease cleavage) is desired, then the method can be performed as described but without protease cleavage during dialysis. In this case, the fusion protein is eluted between 200 and 250 mM imidazole during the elution step of the nickel affinity purification step.

Yield of recombinant Nedd4 and storage conditions:

The purification method described typically affords 2–3 mg of active enzyme per liter of expression culture.

Several storage conditions were examined, including storage at 4 °C, at −20 °C with 40% glycerol, or at −80 °C. The enzyme showed partial degradation upon freeze-thaw cycles and thus is most favorably stored at 4 °C or −20 °C with glycerol (data not shown). The stability of Nedd4 is prolonged when stored at −20 °C with glycerol relative to storage at 4 °C.

Validation of His₆-GST-Nedd4 and Nedd4 activity:

Nedd4 activity was monitored using three previously established activity assays that employ different measures of E3 ligase activity (Fig. 2A) [7,29–31]. First, Nedd4 activity was measured via an immunoblotting-based end-point assay wherein ubiquitin chain formation is detected by anti-ubiquitin primary antibody, HRP-conjugated secondary antibody, and chemiluminescence signal (Fig. 2B). This assay demonstrated that ubiquitin chains were only formed in the presence of full-length Nedd4 with the complete ubiquitination cascade (E1, E2, Ub and ATP), demonstrating that Nedd4 purified by the method described above is active. It should be noted that there is some non-specific signal present in control conditions and the appearance of faint ubiquitination bands in the no-ATP control may be a result of residual ATP present in the commercially purified UBE1 sample. However, comparison of the control conditions to the full ubiquitination reaction reveal a significant, Nedd4-dependent increase in the immunoblotting signal. This result is indicative of Nedd4-dependent ubiquitin chain formation. Secondly, a colloidal Coomassie-based densitometry assay was used to measure Nedd4 autoubiquitination (Fig. 2C). In this assay, Nedd4 is incubated with E1 and E2 enzyme in the presence of ubiquitin and ATP. Following incubation, the reaction is quenched, and products are separated by SDS-PAGE and visualized with colloidal Coomassie staining. In this case, Nedd4 autoubiquitination is specifically measured by quantification of the colloidal Coomassie signal at molecular weights greater than that of full-length Nedd4 (> 104 kDa) as these bands are representative of Nedd4 with added ubiquitin chains. The activity of Nedd4, presented as percent unmodified Nedd4, measured during an abbreviated time-course was consistent with that of Nedd4 activities reported elsewhere [7]. Finally, a MALDI-TOF based activity assay [31] was used for quantitative measurement of Nedd4-dependent monoubiquitin consumption (Fig. 2D). In this method, Nedd4 activity is measured in a time-resolved manner by MALDI-TOF measurement of monoubiquitin signal relative to an internal standard. This assay also validates that Nedd4 is accessed as an active enzyme as judged by the decrease in monoubiquitin signal in the MALDI-TOF spectra relative to a constant amount of internal standard. Cumulatively, the assays employed to validate Nedd4 activity relied upon three measures of E3 ligase activity and demonstrate that Nedd4 is active as measured by 1) ubiquitin chain formation, 2) autoubiquitination, and 3) monoubiquitin consumption.

Fig. 2.

(A) Validation of Nedd4 activity was confirmed through three previously established activity assays that measure independent products of Nedd4 activity including ubiquitin chain formation, Nedd4 autoubiquitination, and monoubiquitin consumption, respectively. For all assays, specific reaction conditions are provided in the Methods section. (B) Immunoblotting detection via anti-ubiquitin primary antibody was used in an end-point assay to measure total ubiquitin chain formation in the presence of the ubiquitin signaling cascade with UBE1 (100 nM), Ubch5a (1 μM), and Nedd4 (5 μM) as the E1, E2 and E3 enzymes, respectively. Reactions were initiated with addition of ubiquitin (100 μM) and incubated for 1 h at 37 °C then quenched with addition of SDS loading buffer. Samples were separated by SDS-PAGE and transferred to PVDF membrane for immunoblotting detection with anti-ubiquitin primary (1:2000), HRP-conjugated secondary, and ECL detection. Ubiquitin chains were formed only in the presence of all three members of the signaling cascade. (C) Time-dependent Nedd4 autoubiquitination assays were conducted with UBE1 (50 nM), Ubch5a (1.5 μM), Nedd4 (1 μM) and ubiquitin (50 μM) in the presence of ATP (2 mM). Reactions were incubated at 30 °C for the indicated time and were subsequently quenched with SDS loading buffer, separated via SDS-PAGE, and detected with colloidal Coomassie staining. The degree of Nedd4 autoubiquitination was determined by measuring the signal at MWs greater than that of unmodified Nedd4 (> ~104 kDa). Signal intensity was measured through densitometry analysis and quantified using ImageJ. Percent unmodified Nedd4 is reported as average ± s.e.m. of duplicate experiments. (D) Quantitative MALDI-TOF detection of monoubiquitin enabled quantification of Nedd4-dependent ubiquitin consumption over time by comparison to an internal standard. Ubiquitination reactions were initiated with the addition of ubiquitin (10 μM) to a reaction containing UBE1 (50 nM), Ubch5a (250 nM), and Nedd4 (500 nM) in buffer with ATP (1 mM). Aliquots (5 μL) were taken at 5 min intervals and quenched with 10% TFA (2 μL). Quenched aliquots were mixed with DHAP matrix and ¹⁵N,¹³C-ubiquitin (4 μM) in 2:3:1 ratio and spotted on an AnchorChip plate (Bruker Daltonics) for MALDI-TOF analysis. Quantitation was performed by calculating the ratio of ubiquitin: ¹⁵N,¹³C-ubiquitin in each timepoint, and data was normalized to time = 0 min. Data is shown as average ± s.e.m. of triplicate experiments.

4. Discussion

The purification route described above enables robust purification of full-length, recombinant Nedd4 as a stable and active enzyme. Nedd4 is a complex, multidomain enzyme that is dependent upon protein-protein interactions for its activity and is subject to complex intramolecular regulatory mechanisms. Thus, it is imperative that it can be purified as a stable, properly folded, untagged enzyme to prevent interference by affinity tags in its inherent activity.

The method presented allows purification of Nedd4 as an untagged enzyme through a two-step purification strategy. Specifically, we present an optimized method which employs orthogonal affinity tags coupled with high specificity proteolytic cleavage. In the method described, Nedd4 is recombinantly expressed via an expression construct that was engineered to contain the Nedd4 gene in fusion with a glutathione-S-transferase and hexahistidine affinity tag sequences. Induction of protein expression from codon-enriched competent E. coli provide Nedd4 in fusion as His₆-GST-Nedd4 with a TEV protease cut site between the N-terminus affinity tags and the Nedd4 protein. Nedd4 can be purified as a stable, untagged recombinant enzyme via sequential purification steps including 1) enrichment via glutathione agarose purification, 2) cleavage from affinity tags via incubation with His₆-TEV during overnight dialysis, and 3) separation from affinity tags (His_6-GST) and His₆-TEV via nickel affinity purification.

Additionally, the activity of Nedd4 purified through the described method was confirmed via three measures of ubiquitin ligase activity: 1) ubiquitin chain formation detected by end-point immunoblotting assays, 2) Nedd4 autoubiquitination via time-dependent, endpoint gel-shift assays, and 3) consumption of monoubiquitin as measured by quantitative MALDI-TOF analysis. We demonstrated, through these three orthogonal measures of Nedd4 E3 ligase activity, that the recombinant enzyme is active when purified with our described purification strategy. To determine if the specific activity of our enzyme is comparable to previously reported activities, we compared Nedd4 activity as measured through the densitometry-based Nedd4 autoubiquitination assay (Fig. 2C). In previous reports of Nedd4 activity using this autoubiquitination assay, Nedd4 was shown to have 71% [32] or 77% ± 1⁷ remaining unmodified Nedd4 after a 30 min reaction time. In another instance, there was 66% ± 3 and 49% ± 9 unmodified Nedd4 remaining after 10 and 30 min, respectively [7]. Herein, we report that 75% ± 7 Nedd4 remains unmodified after 15 min in the same assay conditions, demonstrating that the activity of our ligase is comparable with recombinant Nedd4 activity previously reported.

We sought access to highly pure and stable full-length Nedd4 for the purpose of conducting enzymology studies of Nedd4 activity and biophysical analyses of Nedd4 conformation. While there are several studies that investigate Nedd4 structure and conformation, some of these focus on the interactions of specific Nedd4 subunits as individual recombinant proteins (i.e. isolated C2, WW, or HECT domains) [6,13] or accessed full-length Nedd4 through immunoprecipitation from mammalian tissue culture [15].

There are also several studies which report methods for accessing recombinant Nedd4 in its full length form [7,32,33]. The initial citation of the pGEX-5X3-Nedd4 plasmid, available through the Addgene repository, that we used as the starting point for our plasmid design reported a purification method in which Nedd4 cDNA was cloned into the desired expression vector. The protein was subsequently expressed in BL21(DE3) cells and was purified in one reported step using Glutathione Sepharose enrichment according to manufacturer-recommended conditions (see protocol for GE Lifesciences product #17075601 for details) [33]. We found that expression of the pGEX-5X3-Nedd4 construct in BL21(DE3) strain provided little full-length protein and produced numerous truncation products (data not shown), likely due to differences in codon bias that prevent efficient expression of a non-codon optimized human gene in the E. coli host. After switching to a codon enriched strain, BL21(DE3)-CodonPlus-RIL, expression improved and truncation products decreased, but purification in a single chromatographic step did not afford sufficiently pure enzyme.

There are additional studies [7,32] which report methods for accessing full-length recombinant Nedd4 using the following general strategy: 1) Nedd4 is expressed as a GST-fusion protein in a codon-enriched E. coli strain; 2) Following expression, Nedd4 is isolated from lysed cells via Glutathione Sepharose (GE Lifesciences) enrichment; 3) Eluted GST-Nedd4 is incubated with PreScission protease (GE Lifesciences) during dialysis to cleave the GST tag from Nedd4; 4) Nedd4 is separated from the cleaved GST tag via a second round of Glutathione Sepharose chromatography; 5) Nedd4 is purified in an additional step using Size Exclusion Chromatography (SEC). In comparing our optimized purification method with this general method, we found several advantages to the approach we presented herein. Firstly, our purification method enables access to Nedd4 in two chromatographic steps instead of three, eliminating a costly SEC purification step (in terms of both time and chromatography materials). Additionally, the steps used in the purification presented herein could be modified to enable batch purification rather than FPLC-based purification. Finally, the engineering of the expression plasmid to allow TEV protease-based cleavage of the affinity tags from the His₆ GST-Nedd4 prevents undesired proteolytic degradation and employs a readily available and easily accessible protease [34,35].

Cumulatively, the method presented herein improves upon previously established purification strategies through optimization of the expression construct, expression conditions, chromatographic methods, and proteolytic cleavage. These optimizations provide a more efficient and less expensive approach for purifying recombinant, full-length Nedd4 as an untagged, active enzyme. As further investigations of Nedd4 structure, activity, regulation, and specificity are conducted, it is important to ensure that recombinant Nedd4 used in in vitro and structural investigations reflects its native conformation and activity. Therefore, we envision that the described route can improve the means by which recombinant Nedd4 is accessed in further biochemical and biophysical investigations.