The Tandem Affinity Purification (TAP) method has been used extensively to isolate native complexes from cellular extract, primarily eukaryotic, for proteomics. Here, we present a TAP method protocol optimized for purification of native complexes for structural studies.
Affinity purification approaches have been successful in isolating native complexes for proteomic characterization. Structural heterogeneity and a degree of compositional heterogeneity of a complex do not usually impede progress in conducting such studies. In contrast, a complex intended for structural characterization should be purified in a state that is both compositionally and structurally homogeneous as well as at a higher concentration than required for proteomics. Recently, there have been significant advances in the application of electron microscopy for structure determination of large macromolecular complexes. This has heightened interest in approaches to purify native complexes of sufficient quality and quantity for structural determination by electron microscopy. The Tandem Affinity Purification (TAP) method has been optimized to extract and purify an 18-subunit, ~ 0.8 MDa ribonucleoprotein assembly from budding yeast (Saccharomyces cerevisiae) suitable for negative stain and electron cryo microscopy. Herein is detailed the modifications made to the TAP method, the rationale for making these changes, and the approaches taken to assay for a compositionally and structurally homogeneous complex.
Many major cellular processes are carried out by large protein and protein-RNA complexes1. A significant bottleneck to conducting biophysical and structural studies of such complexes is obtaining them of a suitable quality (i.e., homogeneity) and at an appropriate concentration. Isolating a complex from a native source has many advantages, including retaining relevant post-transcriptional and/or translational modifications of subunits and insuring proper complex assembly. However, large cellular complexes are often present in a cell at a low copy number and the purification must be highly efficient and occur under near physiological conditions to ensure complex integrity is maintained. Purifying a complex from a eukaryotic source is particularly challenging and can be financially prohibitive. Thus, strategies or methods that are efficient and yield a homogeneous complex are highly desired.
A strategy that has been successful in purifying native complexes from eukaryotic cells for their initial characterization is the Tandem Affinity Purification (TAP) method2,3. The TAP method was initially devised to purify a native protein from the budding yeast (S. cerevisiae) in complex with interacting factor(s)2. The TAP method utilized two tags, each tag fused in tandem to the same protein-coding gene sequence. The tags were selected so as to balance a need for tight and selective binding to an affinity resin with a desire to maintain near physiological solution conditions. This balance serves to preserve stable interaction(s) of the tagged protein with interacting factor(s) for post-purification characterization. The genomically incorporated TAP tag was placed at the end (C-terminal) of a protein-coding gene and consisted of a sequence coding for a Calmodulin Binding Peptide (CBP) followed by Protein A – an addition of just over 20 kDa to the tagged protein. CBP is short, 26 amino acids, and recognized by the ~ 17 kDa protein Calmodulin (CaM) in the presence of calcium with a KD on the order of 10-9 M 4. The Protein A tag is larger, consisting of two repeats of 58 residues with a short linker between the repeats. Each 58 amino acid repeat is recognized by immunoglobulin G (IgG) with a KD ~ 10-8 M 5. Between these two tags was incorporated a recognition site for TEV protease, an endopeptidase from the Tobacco Etch Virus6,7. As illustrated in Figure 1, in the first affinity step of the method the TAP tagged protein is bound to an IgG resin via the Protein A. The tagged protein is eluted by on column cleavage upon the addition of TEV protease, site-specifically cleaving between the two tags. This is a necessary step as the interaction of IgG and Protein A is very strong and can only be adequately perturbed under denaturing solution conditions. Lacking a Protein A tag, the protein is bound to CaM resin in the presence of calcium and eluted from this resin with addition of the metal ion chelator EGTA (ethylene glycol tetraacetic acid) (Figure 1).
Soon after the introduction of the TAP method, it was used in a large-scale study to generate a 'map' of complex interactions in S. cerevisiae8. Importantly, as a consequence of this effort an entire yeast-TAP Tagged Open Reading Frame (ORF) library as well as individual TAP tagged ORFs9 are available from a commercial source. Thus, one can obtain any yeast strain with a tagged protein for any yeast complex. The TAP method also spurred modifications or variations of the TAP tag, including: its use for purification of complexes from other eukaryotic as well as bacterial cells10,11; the design of a "split tag", wherein the Protein A and CBP are placed on different proteins12; and the tags changed, so as to accommodate the need of the investigator, such as sensitivity of the complex to Ca2+ or EGTA13.
Recent advances in both instrumentation and methodology have led to significant advances in the application of electron microscopy (EM) for structure determination, that have led to high, near atomic resolution images of macromolecular complexes14. The resolution obtainable of a complex by EM, however, remains contingent upon the quality of the complex under study. This study has utilized the TAP tag approach to purify from S. cerevisiae the U1 snRNP, an 18-subunit (~ 0.8 MDa) low copy number ribonucleoprotein complex that is part of the spliceosome15,16. A number of steps have been taken to purify this complex such that it is homogeneous and of an adequate concentration. Potential problems encountered at various stages of the purification are described and strategies taken to overcome challenges highlighted. By carefully assessing and optimizing steps in the purification, the U1 snRNP purified is of a quality and at a quantity suitable for negative stain and electron cryo microscopy (cryo-EM) studies. An optimized TAP method protocol for purification of native complexes for structural studies is described herein.
Note: The following protocol was devised for purification of a complex from 4 L of cell culture, approximately 40 g wet weight of cells. Once prepared, all buffers should be stored at 4 °C and used within a month of their preparation. Reducing agent and protease inhibitors are only added to buffers just prior to use.
1. Preparation of Whole Cell Extract for Tandem Affinity Purification
2. Column Purification step 1: IgG Chromatography
3. Column Purification Step 2: Calmodulin Affinity Chromatography
4. Post-Purification Analyses to Assess Complex Quality
5. Complex storage
Note: The presence of a detergent such as NP-40 and at a concentration above that of its critical micelle concentration may hinder or inhibit progress for some applications. The consequences of its removal from the protocol as well as replacement with other additives or co-solvents are discussed below. If no NP-40 is desired, an option is to use a commercial application for removal, for example Bio-beads.
A modified TAP method was used to purify from S. cerevisiae the U1 snRNP, an 18-subunit ribonucleoprotein complex. An initial TAP purification of the complex following the published protocol2,3 yielded a complex that appeared heterogeneous, migrating as three bands on a silver stained native polyacrylamide gel (Figure 2A). Multiple rounds of optimization of the TAP method, yielded a complex that migrated as primarily a single band on a native gel indicative of a more homogeneous assembly (Figure 2A). Using fluorescent dyes to stain for both protein and nucleic acid, it was established that the purified complex had both biopolymers present (Figure 2B).
The purified complex was also analyzed by SDS PAGE and Western blotting using a TAP tag antibody (Figure 3A). Consistent with the native gel result, the complex purified according to the published protocol exhibited proteolysis of the TAP tagged protein (Snu71). The amount of proteolysis, however, was significantly reduced with modifications of the TAP method. Nearly all of the seventeen proteins in the U1 snRNP complex were resolved by SDS PAGE and positively identified by MALDI mass spectrometry (Figure 3B). As well as native and SDS PAGE, complex quality was assessed at different stages of the purification by negative stain electron microscopy (Figure 4). The monodispersed particles observed in Figure 4A and B but absent from 4C are examples of good sample for structural study. This analysis method provided critical insight as to the impact of changes to the TAP method.
The TAP method calls for the inclusion of the non-ionic detergent NP-40 and at a concentration above that of its critical micelle concentration (CMC). The robustness of the methods for assessing the quality of the complex described in the protocol are well demonstrated by an experiment where the zwitterionic surfactant CHAPS, glycerol or no co-solvent were substituted for NP-40 in all buffers (Figure 5). By native PAGE and negative stain EM, the U1 snRNP appears most homogeneous and stable in NP-40 with decreasing stability from CHAPS to glycerol to no co-solvent added. It might be that the yeast U1 snRNP has a hydrophobic face and the presence of NP-40 prevents aggregation by coating this surface. Unfortunately, NP-40 cannot be removed by dialysis and concentrating the complex will then concentrate the NP-40. While the higher NP-40 did not appear to perturb the complex, it did negatively impact the freezing of the particle in vitreous ice for cryo-EM and large micelle-like aggregates were observed. Detergent absorbing beads were used successfully to remove the majority of the NP-40 from the U1 snRNP sample without appearing to impact the complex structurally.
Following the purification of the complex with the modifications detailed in the protocol, a complex suitable for both negative stain and cryo-EM was obtained, as evidenced by images of the lawn of particles and distinctive representative class averages (Figure 6).
Figure 1. Overview of Tandem Affinity Purification (TAP) Method. (A) Schematic of the two step TAP method. Complex of interest (grey) includes a TAP tagged subunit (black; U1 snRNP subunit Snu71) genomically fused with a sequence comprised of a Calmodulin Binding Peptide (CBP, green), a recognition site for the site-specific protease TEV (Blue), and two IgG binding regions of Protein A (Pr A, Red). In the 1st step, cell lysate is incubated with an IgG resin and the complex is retained via its interaction with the Protein A sequence. The complex is released from the resin by TEV mediated cleavage. In the 2nd step, the complex is incubated with a Calmodulin resin, a Ca2+ dependent reaction. The complex is released via addition of the Ca2+ chelator EGTA. (B) Western blot following a TAP tagged protein through the purification steps illustrated in (A). Samples were taken of cell lysate, after the 1st step, and after the 2nd step of the TAP method. The reactive protein observed in cell lysate migrates faster after the 1st step, following TEV protease cleavage and thus loss of the Protein A sequence. Please click here to view a larger version of this figure.
Figure 2. Optimization of the TAP Method. (A) Silver stained native gel of TAP purified complexes. Native PAGE of: protein standard (lane 1); complex purified according to original TAP method3 (lane 2), three bands observed (arrowheads); three consecutive elutions from the same TAP purification following alterations to buffer, two bands observed (arrowheads) (lanes 3 – 5); further changes to the TAP method, yielding primarily a single complex band migrating at ~ 800 kDa (lane 6). (B) Analysis of composition of a purified complex by staining a single gel lane with both protein and RNA reactive stains. Native PAGE of: protein standard (lane 1) stained with a protein fluorescent stain; complex imaged with protein fluorescent stain (lane 2), two bands observed; same lane in lane 2, now imaged with a nucleic acid fluorescent stain (lane 3), two bands observed; and fluorescent signal overlay of stained complex in lanes 2 and 3 (lane 4). Please click here to view a larger version of this figure.
Figure 3. Optimization of TAP Method. (A) Western blot analysis of improvement in complex quality, following TAP tagged subunit Snu71 probed with α-TAP antibody. Several proteolyzed bands are visible initially (lane 1) but improvements reduce the amount of lower bands considerably (lanes 2 and 3). (B) A representative SDS gel stained with a fluorescent protein gel stain. Gel resolves 12 of the 17 protein subunits of the complex. Identity of protein in the bands was established by mass spectrometry. Please click here to view a larger version of this figure.
Figure 4. Sample Quality at Stages in the TAP Method Assessed by Negative Stain Electron Microscopy. Assessment by negative stain electron microscopy of complex after: (A) the 1st step of TAP; (B) the 2nd step of TAP; (C) complex destabilized. Scale bar, 125 nm. Please click here to view a larger version of this figure.
Figure 5. Impact on Sample Quality of Additives Present in TAP Method Buffers. Lanes 1 – 4 are the first four elutions of the final or 2nd step of TAP purification of complex purified in the presence of: (A) the non-ionic detergent nonyl phenoxypolyethoxylethanol (NP-40), analyzed by native PAGE (top) and negative stain EM (bottom); (B) the zwitterionic detergent 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), analyzed by native PAGE (top) and negative stain EM (bottom); (C) glycerol, analyzed by native PAGE (top) and negative stain EM (bottom); and (D) no additive, analyzed by native PAGE (top) and negative stain EM (bottom). Scale bar, 50 nm. Please click here to view a larger version of this figure.
Figure 6. Visualization of TAP Method Purified Complex. (A) Top, a representative image of a negative stain electron microscopy (EM) micrograph. Observed is a lawn of similarly shaped particles, two are boxed, purified by the optimized TAP method. Bottom, a selection of class averages taken of the particles that highlight particle distinguishing features. Scale bar, 50 nm. (B) Top, a representative image of a cryo-EM micrograph. Observed are regions of higher contrast representative of particles, see two such regions boxed. Bottom, class averages of selected particles frozen in vitreous ice. Scale bar, 50 nm. Please click here to view a larger version of this figure.
The TAP method utilizes two tags that balance a need for tight and selective binding to an affinity resin with a desire to maintain near physiological solution conditions. This balance serves to preserve stable interaction(s) of the tagged protein with interacting factor(s) for post-purification characterization. In addition, individual TAP tagged ORFs are available from a commercial source, so that one can obtain any yeast strain with a tagged protein for any yeast complex. Preserving the integrity of a complex and the availability of a resource to test use of different tagged protein subunits in a complex for purification are two advantages for utilizing the TAP method for purifying a native complex. However, the original TAP method protocol was not devised to ensure that the complex purified is compositionally and structurally homogeneous. The TAP method has therefore been modified, as detailed above, to achieve this goal.
Various steps in TAP method were assessed to establish the optimal approach and conditions to purify a compositional and structural homogeneous complex. An early critical step is cell lysis. This step requires a balance between one's desire for a high yield and thus maximum lysis, while insuring that the complex obtained is of high quality, i.e., homogeneous. Different approaches to lysing yeast cells were explored, including using a bead-beater, high sheer fluid processor, and ball mill. Use of the less expensive coffee grinder was less efficient than the alternative methods (40 – 60% lysis), but the complex isolated and purified appeared monodisperse while several of these approaches appeared to either perturb particle integrity or cause a degree of aggregation. In the end, a significant number of cells were not lysed so as to ensure high quality complex was obtained. While 40 – 60% lysis is ideal for the yeast U1 snRNP, optimization is suggested when applying this protocol to a new biological complex.
It was critical to identify the steps in the method that could hasten purification to ensure that complex integrity was not compromised by cellular proteases, subunit dissociation as a result of dilution, and extended incubation at 18 °C during TEV protease cleavage. One simple step to achieve time efficient purification was to ensure that lipids did not carry over from the early lysate clarification, since lipids impede flow during gravity purification. Other steps included identifying an optimal balance of the amount resin and column size/dimensions to achieve efficient purification and a maximum flow rate. A particularly critical step was to reduce time of on column cleavage by TEV protease. The TEV protease used was purified in house to ensure it was protease/nuclease free and at a high concentration17. Significant amounts of TEV protease were used to obtain complete cleavage in under an hour.
In initial TAP purifications proteolysis was observed. Proteolysis during purification can be limited by working efficiently, adding a battery of inhibitors, and including in the purification high salt washes. A broad range of protease inhibitors supplied over the first half of the purification greatly improved the stability of the TAP tagged protein and complex. In addition, proteolysis was minimized by increasing the monovalent concentration of 300 to 500 mM during the cell lysis and IgG steps of purification. Even with the above modifications to the TAP method, successive eluted fractions off of the CBP affinity resin varied in their degree of compositional homogeneity. Thus, selective pooling of fractions from the second, CBP chromatography step, is warranted.
An aspect of TAP method that received particular focus was the importance of the non-ionic detergent NP-40 to complex purification and at a concentration above its CMC. The presence of NP-40 at a concentration of 0.08% or higher had an overall beneficial effect on complex integrity and the final yield. The apparent beneficial effect(s) of NP-40 may be due in part to reducing non-specific interactions between resin and the complex. While NP-40 appears to have a beneficial effect during the purification, it could be removed post-purification without causing complex aggregation.
Finally, critical to optimizing the TAP method for structural study were several complementary assays used to assess the integrity and homogeneity of the complex. These included light scattering, SDS and native PAGE probed using the TAP antibody by Western blot and/or stained with fluorescent dyes, as well as negative stain electron microscopy. The changes made to the TAP method detailed herein, yielded a homogeneous complex at sufficient quantities for electron microscopy. Additional changes may need to be made to the TAP method for other complexes, in particular a determination of the importance of NP-40.
In summary, the TAP purification method can be used to successfully purify macromolecular complexes from a eukaryotic source of suitable quality for structural study. We have established a suitable approach for purification of the S. cerevisiae U1 snRNP and detail the steps taken as well as the rationale for many of these changes. For other complexes, we suggest that the investigator similarly explore the steps in the protocol, which may yield adequate amounts and quality of complex. The steps that should be examined include cell lysis, salt and additive type and concentrations including detergent, and sensitivity of the complex to calcium and metal chelator EGTA. Further, it is critical for long-term studies that the complex can be frozen and thawed without perturbing structure. Importantly, one should have several alternative ways to assay the structure of the complex during and following purification. Native and SDS PAGE, negative stain EM, and light scattering have served us well in our investigation of how best to purify a homogeneous complex using the TAP method.
The authors have nothing to disclose.
The authors are grateful for the support and advice of Nikolaus Grigorieff. We thank Anna Loveland, Axel Brilot, Chen Xu, and Mike Rigney for helpful discussions and EM guidance. This work was funded by the National Science Foundation, Award No. 1157892. The Brandeis EM facility is supported by National Institutes of Health grant P01 GM62580.
S. cerevisiae TAP tagged strain | Open Biosystems | YSC1177 | This is the primary yeast strain used to develop the TAP protocol. Its background is S288C: ATCC 201388: MATa, his3Δ1, leu2Δ0, met15Δ0, ura3Δ0, SNU71::TAP::HIS3MX6 |
Coffee grinder | Mr. Coffee | IDS77 | Used for cell lysis |
Hemocytometer, Bright Line | Hausser Scientific | 3120 | Used to assess cell lysis |
JA 9.100 centrifuge rotor | Beckman Coulter, Inc. | Used to harvest the yeast cells | |
JA 20 fixed-angle centrifuge rotor | Beckman Coulter, Inc. | Used to clear the cell extract of non-soluble cellular material | |
Ti 60 fixed-angle centrifuge rotor | Beckman Coulter, Inc. | Used to further clear the soluble cell extract | |
Thermomixer | Eppendorf | R5355 | Temperature controlled shaker |
Novex gel system | Thermo Fisher Scientific | ||
IgG resin | GE Healthcare | 17-0969-01 | Sepharose 6 fast flow |
Calmodulin resin | Agilent Technologies, Inc. | 214303 | Affinity resin |
Protease inhibitor cocktail, mini tablets | Sigma Aldrich | 589297 | Mini cOmplete ultra EDTA-free tablets |
Protease inhibitor cocktail, large tablets | Sigma Aldrich | 5892953 | cOmplete ultra EDTA-free tablets |
Phenylmethanesulfonyl fluoride (PMSF) | Dissolved in isopropanol | ||
2 mL Bio-spin column | Bio-Rad Laboratories, Inc. | 7326008 | Used to pack and wash the Calmodulin resin |
10 mL poly-prep column | Bio-Rad Laboratories, Inc. | 7311550 | Used to pack and wash the IgG resin |
Precast native PAGE Bis-Tris gels | Life Technologies | BN1002 | Novex NativePAGE Bis-Tris 4-16% precast polyacrylamide gels |
NativeMark protein standard | Thermo Fisher Scientific | LC0725 | Unstained protein standard used for native PAGE. Load 7.5 μL for a silver stained gel and 5 μL for a SYPRO Ruby stained gel |
Precast SDS PAGE Bis-Tris gels | Life Technologies | NP0321 | Novex Nu-PAGE Bis-Tris 4-12% precast polyacrylamide gels |
PageRuler protein standard | Thermo Fisher Scientific | 26614 | Unstained protein standard used for Western blotting |
SDS running buffer | Life Technologies | NP0001 | 1x NuPAGE MOPS SDS Buffer |
TAP antibody | Thermo Fisher Scientific | CAB1001 | Primary antibody against CBP tag |
Secondary antibody | Thermo Fisher Scientific | 31341 | Goat anti-rabbit alkaline phosphatase conjugated |
BCIP/NBT | Thermo Fisher Scientific | 34042 | 5-bromo-4-chloro-3-indolyl-phosphate/nitro blue tetrazolium |
Dialaysis units | Thermo Fisher Scientific | 88401 | Slide-A-Lyzer mini dialysis units |
Centrifugal filter units, 100kDa MWCO | EMD Millipore | UFC5100008 | Amicon Ultra-0.5 Centrifugal Filter Unit with Ultracel-100 membrane |
Detergent absorbing beads | Bio-Rad Laboratories, Inc. | 1523920 | Bio-bead SM-2 absorbants |
SYBR Green II | Thermo Fisher Scientific | S-7564 | Flourescent dye for nucleic acid staining, when detecting with SYPRO Ruby present, use excitation wavelength of 488 nm and emission wavelength of 532 nm |
SYPRO Ruby | Molecular Probes | S-12000 | Flourescent dye for protein staining, when detecting with SYBR Green II present, use excitation wavelength of 457 nm and emission wavelength of 670 nm |
Copper grids | Electron Microscopy Sciences | G400-CP |