G4 Resolvase1 binds to G-quadruplex (G4) structures with the tightest reported affinity for a G4-binding protein and represents the majority of the G4-DNA unwinding activity in HeLa cells. We describe a novel protocol that harnesses the affinity and ATP-dependent unwinding activity of G4-Resolvase1 to specifically purify catalytically active recombinant G4R1.
Higher-order nucleic acid structures called G-quadruplexes (G4s, G4 structures) can form in guanine-rich regions of both DNA and RNA and are highly thermally stable. There are >375,000 putative G4-forming sequences in the human genome, and they are enriched in promoter regions, untranslated regions (UTRs), and within the telomeric repeat. Due to the potential for these structures to affect cellular processes, such as replication and transcription, the cell has evolved enzymes to manage them. One such enzyme is G4 Resolvase 1 (G4R1), which was biochemically co-characterized by our laboratory and Nagamine et al. and found to bind extremely tightly to both G4-DNA and G4-RNA (Kd in the low-pM range). G4R1 is the source of the majority of G4-resolving activity in HeLa cell lysates and has since been implicated to play a role in telomere metabolism, lymph development, gene transcription, hematopoiesis, and immune surveillance. The ability to efficiently express and purify catalytically active G4R1 is of importance for laboratories interested in gaining further insight into the kinetic interaction of G4 structures and G4-resolving enzymes. Here, we describe a detailed method for the purification of recombinant G4R1 (rG4R1). The described procedure incorporates the traditional affinity-based purification of a C-terminal histidine-tagged enzyme expressed in human codon-optimized bacteria with the utilization of the ability of rG4R1 to bind and unwind G4-DNA to purify highly active enzyme in an ATP-dependent elution step. The protocol also includes a quality-control step where the enzymatic activity of rG4R1 is measured by examining the ability of the purified enzyme to unwind G4-DNA. A method is also described that allows for the quantification of purified rG4R1. Alternative adaptations of this protocol are discussed.
G4 structures are highly stable nucleic acid secondary structures that form within guanine-rich regions of DNA and RNA. G4 structures are stabilized via Hoogsteen-bonding interactions and coordinate bonding within the central cavity with monovalent cations (i.e. K+ and Na+) that significantly contribute to the remarkable thermal stability of G4 structures1,2. Early bioinformatics studies suggested that the human genome contains >375,000 “potential G4-forming motifs”3,4. More recent study estimates suggest that the number of G4 motifs is higher by a factor of 2-55, while another study predicts 716,310 distinct potential G4-forming sequences in the human genome6. G4-forming sequences are evolutionarily conserved and not randomly dispersed in the genome. G4 motifs are enriched in gene coding regions, and upwards of 40% of all gene promoters contain G4 motifs7. Interestingly, the degree of enrichment of G4 motifs in a gene has been demonstrated to suggest the function of the gene. For example, proto-oncogenes and genes involved in development have significantly greater enrichment of G4 structures than tumor suppressor genes8,9.
With high thermal stabilities, a nearly ubiquitous presence throughout the genome, and the potential to significantly affect major cellular processes, it is unsurprising to find that the cell has evolved enzymes to manage these structures. One such enzyme is G4 Resolvase1 (G4R1; also called RHAU and DHX36), which we characterized as the source of the majority of tetramolecular G4-DNA resolving activity in human (HeLa) cells10. Since then, it has been shown that G4R1 tightly binds and catalytically unwinds tetramolecular and unimolecular G4-DNA and G4-RNA with the tightest reported KDs for a G4-binding protein11,12,13. Additionally, the G4-resolving activity of G4R1 has been implicated in a wide range of biochemical and cellular processes, including telomere/telomerase biology11,14,15,16, transcription and splicing17,18,19,20, development21, hematopoiesis21, and immune regulation22,23. With a preponderance of G4 sequences specifically situated throughout the genome and the diverse cellular processes that G4R1 has recently been implicated to be involved with, the ability to express and efficiently purify highly active rG4R1 will be of the utmost importance for elucidating the biochemical mechanisms and behaviors of this protein.
Here, we demonstrate a novel expression and purification scheme (Figure 1) that takes advantage of the ATP-dependent, G4-resolving activity of rG4R1 to efficiently isolate active enzyme. This scheme could be adapted to purify other ATP-dependent nucleic acid enzymes for which the product of the enzymatic reaction is no longer a substrate for binding, as is the case for G4R1.
1. Preparation of G4-DNA Structures to be Used for the Purification of rG4R1 (Formation of Biotinylated G4-DNA G-Quadruplex)
2. Preparation of G4-DNA Structures to be Used in An Enzymatic Activity Assay of rG4R1 (Formation of TAMRA-labeled G4-DNA)
3. Transform (DE3) PlysS Competent Cells with pTriEx4-DHX36 (Plasmid Encoding Human G4R1) and Grow/Induce Large Bacterial Cultures
4. Purification of Human rG4R1
5. Quality Control Enzymatic Activity Assay of Purified rG4R1
6. Pooling of Highly Active rG4R1 Preps, Aliquoting, and Storage
NOTE: This step requires 2 people. The number and enzymatic requirements of the downstream assays that the purified rG4R1 will be used in will determine how many preparations are needed prior to aliquoting. A typical large preparation consists of the pooling of 8 highly active preparations (thus using a total of 16 induced 500 mL bacterial cultures), but this is an arbitrary number and is laboratory-specific.
7. Quantification of Purified rG4R1 Concentration
This protocol (Figure 1) routinely yields nearly pure, catalytically active rG4R1. As a measure of enzymatic activity, it is typically observed that 50% of 0.2 pmol of TAMRA-labeled tetramolecular G4-DNA is converted into monomers within the 0.2 – 0.013 µL range of rG4R1, as assayed by the G4 activity assay outlined above (Figure 2). Coomassie staining of purified rG4R1 indicates a single band at the expected 120 kDa size, with a minor contaminating band at ~75 kDa, which may represent truncated rG4R1 that has maintained its ability to bind G4-DNA beads and to elute in an ATP-dependent manner (Figure 3). The band of interest is quantified against a protein standard curve, and this protocol typically obtains 200 µL of 20-100 nM purified enzyme per 1 L of bacteria culture.
Figure 1: Schematic of a Two-step Purification of G4R1. A 6xHis-tagged rG4R1 is bulk-purified from E. coli lysates by first binding to and eluting from Co2+-conjugated beads. A binding step to G4-DNA-conjugated magnetic beads follows. Finally, an ATP-dependent elution step is necessary to obtain relatively pure and enzymatically active rG4R1. Please click here to view a larger version of this figure.
Figure 2: Quality-control G4-DNA Unwinding Assay. Lanes 1 – 6: A constant concentration of TAMRA-labeled tetramolecular G4-DNA was incubated at 37 °C for 30 min in the presence of 4x serial dilutions of purified rG4R1 representing 3.9 µL, 0.83 µL, 0.2 µL, 0.05 µL, 0.013 µL, and 0.003 µL, respectively. Lane 7: tetramolecular G4-DNA in the absence of rG4R1. Lane 8: tetramolecular G4-DNA boiled to reduce the G4 structure into monomers in the absence of rG4R1. Please click here to view a larger version of this figure.
Figure 3: Quantification of Purified rG4R1 Concentration. Broad-range MW protein markers were loaded in the following quantities: 500, 250, 125, 62.5, 31.3, and 15.7 ng in Lanes 1 – 6, respectively, and were used to generate a standard curve of protein concentrations. Lane 7 represents 35 µL of rG4R1. This particular gel was quantified, as part of a triplicate set of gels, resulting in an average protein concentration of 62 ± 22 nM standard deviation (SD; N = 9). Please click here to view a larger version of this figure.
This protocol represents a highly efficient expression, purification, and quantification scheme for the isolation of the DHX36 gene product, G4-Resolvase1 (G4R1, also called RHAU and DHX36) (Figure 1). This protocol utilizes two purification steps: His-tag affinity purification on cobalt affinity beads and enzymatic purification on G4-DNA-conjugated beads. The latter step is unique in that it takes advantage of the tight affinity, high specificity, and catalytic unwinding activity that rG4R1 has for G4 structures. The tight affinity for G4 structures (Kd in the low-pM range) allows for a high salt wash (near the solubility limit of NaCl) prior to elution from the G4 beads, greatly reducing the presence of non-specific proteins. The catalytic activity of G4R1 on G4 structures allows for the specific elution of active rG4R1 upon the addition of ATP and MgCl2. This two-step purification scheme consistently produces highly pure, catalytically active rG4R113 in quantities suitable for most biochemical analyses.
Several key steps will ensure a good yield of highly pure, catalytically active rG4R1. The first is to ensure the proper induction conditions of His-rG4R1 in the bacterial expression strain. We have found that the best yield of recombinant protein occurs when cells are grown to an OD600 of no more than 0.4-0.6 just prior to induction. Growing the cells beyond this point may result in an overall loss of protein recovery, possibly due to the incorporation of rG4R1 into inclusion bodies. Second, we obtained a higher concentration of purified protein by “serially-binding” the lysates to the cobalt affinity beads. For example, we bound the lysate from 0.5 L of induced cultures to the cobalt affinity beads and then performed a second binding of another lysate from an additional 0.5 L of culture to the same cobalt affinity beads. This step ensures a more concentrated preparation of protein by increasing the number of rG4R1 molecules bound to a given volume of beads, thus utilizing the full capacity of the cobalt affinity beads. Thirdly, the high salt wash after binding the cobalt affinity elutions to the G4 beads ensures that nearly all non-specific protein binding to the beads is removed. Fourthly, the ATP/MgCl2 elution step allows for rG4R1 bound to the G4 beads to catalytically unwind the tetramolecular structure into single strands, causing rG4R1 to be released from the beads. We cannot completely rule out the possibility that ATP elutes rG4R1 in a competitive rather than a catalytic manner; however, this is less likely to be the case, since we have previously shown that a non-hydrolyzable ATP analogue is not sufficient for competitive binding13,18. The affinity of rG4R1 for the unwound, single-stranded DNA is an order of magnitude less than the starting tetramolecular G4-DNA, and thus rG4R1 should not re-bind to the beads. In order to reduce this possibility, however, this step should be done at 37 °C, and the elution volume should be separated from the beads as quickly as possible. The elution step is repeated twice to ensure maximum recovery. If downstream applications require the protein to be free of DNA contaminants, we recommend an additional cleanup step in which the purified preparation is re-bound to streptavidin beads in order to remove any biotinylated DNAs, if present.
We have found that rG4R1 is susceptible to degradation if the proper conditions are not maintained throughout the protocol. In order to maintain the integrity and activity of the enzyme, we employ the following critical safeguards. Protease inhibitors are kept present throughout the purification procedure. The protocol is performed at 4 °C, unless otherwise noted. The protein is purified in the presence of lactalbumin and β-mercaptoethanol. The protocol is performed in a timely fashion (in 1 day for the purification). Additionally, we have found that multiple freeze-thaw cycles negatively impact the activity of the protein, so we aliquot the protein into “one-time use,” 7 µL aliquots following purification and store them at -80 °C.
Although the presence of lactalbumin in the preparation is required to maintain the integrity and activity of the protein, as mentioned above, we have found that this may also impede downstream applications. Other potential interfering molecules that are present in the purified rG4R1 preparation include ATP, β-mercaptoethanol, and singled-stranded DNA. For example, we have found this protein preparation to be incompatible with BIACORE analysis due to the high background signal from the buffer components. Also, the presence of lactalbumin in the protein preparation precludes the use of standard Bradford and BCA protein quantitation assays. However, we have developed an alternative gel-based quantitation method to circumvent this limitation.
This purification procedure, which harnesses the enzymatic activity of G4R1 as a means to specifically purify it, makes this method distinct from other methods. For example, other groups have expressed FLAG-tagged rG4R1 in human cells15,25 or GST-tagged rG4R1 in insect cells26 and purified it by FLAG- or GST-affinity chromatography, respectively. These methods have the advantage of being done in a eukaryotic expression system compared to a bacterial expression system. Estimated Kd values of the resulting purified GST-G4R1 for G4 structures were found to be an order of magnitude higher14 than our reported Kd values12,13. We attribute this discrepancy in Kd values to differences associated with a bulkier GST-tag versus a 6xHis-tag, differences in the purities obtained from these two purification schemes, and differences in the extent and type of post-translational modifications acquired in a bacterial versus an insect expression system. Our approach has a distinct advantage over the aforementioned alternatives because the purification of this protein directly hinges on its enzymatic activity. Therefore, we primarily obtain an enzyme that has been folded and modified in such a way that maintains its enzymatic properties. Other affinity-tag and/or size-exclusion techniques are unable to separate active enzyme from enzymatically dead enzyme. It would be useful for future protocol development to combine the strengths of other groups’ purification protocols15,25,26 (i.e. a human or insect cell expression system) with the strength of our protocol (i.e., catalytic-based purification) to further improve upon this method.
Although this protocol is currently specific for G4R1, it could easily be adapted to any ATP-dependent, G4-resolving proteins, including, but not limited to, BLM, WRN, FANCJ, hnRNP proteins, hPif1, and/or ChlR1/DDX1. By altering the sequence of the nucleic acid bound to the streptavidin beads, this protocol could be adapted to purify other ATP-dependent nucleic acid enzymes for which the product of the enzymatic reaction is no longer a substrate for the enzyme, including nucleic acid helicases and nucleases.
The authors have nothing to disclose.
We would like to thank our funding sources, including a generous gift from the Ware Foundation (to J.P.V.), The National Institutes of HealthGrant T32-CA079448 (to P.J.S.), and Ball State University startup funds (to P. J. S.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
TriEx4-DHX36 plasmid | Addgene | 68368 | |
Rosetta2(DE3)plysS competent cells | Novagen | 71403-4 | |
S.O.C medium | Thermo Fisher Scientific | 15544034 | |
Difco Terrific Broth | Becton Dickinson | 243820 | |
Glycerol | Sigma-Aldrich | G5516 | |
Chloramphenicol | Sigma-Aldrich | C1919 | 35 µg/ml in bacterial plates/large cultures |
Carbenicillin (plant cell culture tested) | Sigma-Aldrich | C3416 | 50 µg/ml in bacterial plates/large cultures |
Isopropyl β-D-1-thiogalactopyranoside (IPTG) | Sigma-Aldrich | I6758 | |
Lysozyme (from chicken egg white) | Sigma-Aldrich | L6876 | |
1 M Tris-HCl pH=8 | Universal Scientific Supply Co. | 1963-B | or From standard source |
1 M Tris-HCl pH=7 | Universal Scientific Supply Co. | 1966 | or From standard source |
1.5 M Tris-HCl, pH=8.8 | For casting resolving gel (for protein quantitation gel); From standard source | ||
1 M Tris-HCl, pH=6.8 | For casting stacking gel (for protein quantitation gel); From standard source | ||
1 M Tris-Acetate, pH=7.8 | Universal Scientific Supply Co. | 1981 | or From standard source |
70% Ethanol | From standard source | ||
Magnesium chloride (1 M solution) | Life Technologies | AM9530G | |
Sodium chloride | Sigma-Aldrich | S7653 | |
Sodium acetate | Sigma-Aldrich | S8750 | |
20x SSC | Universal Scientific Supply Co. | 1665 | or From standard source |
β-mercaptoethanol (2-BME) | Sigma-Aldrich | 63689 | |
Protease inhibitor cocktail | Sigma-Aldrich | P8849 | |
Leupeptin hemisulfate | Sigma-Aldrich | L8511 | |
Streptavidin paramagnetic beads | Promega | Z5482 | |
0.5 M EDTA, pH=8 | Universal Scientific Supply Co. | 0718 | or From standard source |
0.2 M EDTA, pH=6 | Universal Scientific Supply Co. | From standard source; initially adjust pH with NaOH, then adjust pH back down with HCl. | |
A-lactalbumin (Type 1 from bovine milk) | Sigma-Aldrich | L5385 | |
Cobalt metal affinity beads | Clonetech | 635502 | |
L-Histidine | Sigma-Aldrich | H8000 | |
Acetic acid, glacial | Fisher Scientific | A38-500 | |
Adenosine 5'-Triphosphate (from bacterial source) | Sigma | A7699 | |
40% acrylamide/Bis solution (37.5:1) | Biorad | 161-0148 | |
Glycine | Sigma-Aldrich | 50046 | to make protein gel running buffer (192 mM glycine, 25 mM Tris Base, 0.1% SDS) |
10 % Sodium dodecyl sulfate | Universal Scientific Supply Co. | 1667 | to make protein gel running buffer (192 mM glycine, 25 mM Tris Base, 0.1% SDS); or From standard source |
10x TBE | Sigma-Aldrich | 11666703001 | or From standard source |
Tris base | Fisher Scientific | BP152-1 | to make protein gel running buffer (192 mM glycine, 25 mM Tris Base, 0.1% SDS); From standard source |
TEMED | Sigma-Aldrich | 411019 | |
Ammonium persulfate | Sigma-Aldrich | A3678 | |
Broad Range Protein MW markers | Promega | V8491 | |
Biotinylated Z33 oligo ("Z33-Bio") | Oligos Etc | 5’ AAA GTG ATG GTG GTG GGG GAA GGA TTC GGA CCT-biotin 3’ | |
TAMRA-Z33 oligo ("Z33-TAM") | Oligos Etc | 5’ TAMRA-AAA GTG ATG GTG GTG GGG GAA GGA TTC GGA CCT 3’ | |
Fluor-coated TLC plate | Life Technologies | AM10110 | |
Ficoll | Sigma-Aldrich | F2637 | 30% in H2O |
Coomassie R-250 | Sigma-Aldrich | 27816 | |
Methanol | Fisher Scientific | A412 | |
Multiband UV lamp | Capable of emitting UV light at 365 nm | ||
Table-top centrifuge (with swinging bucket rotor) | Capable of being cooled to 4 °C | ||
Microcentrifuge | Capable of being cooled to 4 °C | ||
Digital Sonfier | Branson | Or equivalent capable of delivering sonication pulses (30% amplitude, 2s ON 2s OFF) | |
50 °C water bath | For formation of Z33 into quadruplex | ||
37 °C incubator for bacteria | For bacterial transformations and initial overnight growth of large cultures of Rosetta2 E. coli transformed with TriEx4-DHX36 | ||
37 °C/14 °C shaking incubator for bacteria | For growth and protein induction of large cultures of Rosetta2 E. coli transformed with TriEx4-DHX36 | ||
Spectrophotometer | capable of reading OD600; capable of reading oligomer concentrations based on base sequence (such as Biorad SmartSpec 3000) | ||
Thermometer | From standard source | ||
PCR strip tubes | From standard source | ||
15- and 50-ml centrifuge tubes (polypropylene) | From standard source | ||
Microcentrifuge tubes (2.0 ml) | From standard source | ||
500 ml centrifuge bottles (polypropylene) | Thermo Scientific | 3141-0500 | |
Standard array of pipet tips and serological pipettes | From standard source | ||
Gel-loading tips | From standard source | ||
Automatic repeating pipette | For quick aliquoting of rG4R1; From standard source | ||
Thermal cycler | From standard source | ||
Liquid Nitrogen | From standard source | ||
Dry ice | From standard source | ||
Laemlli sample buffer | Biorad | 161-0737 | |
Apparatus for running large slab gels | Biorad | We have used the Protean II xi cell apparatus from Biorad | |
Magnet | Life Technologies | 12301D | We use a magnet from One Lambda (Now a Thermo Fisher Scientific brand); and Life is also a subsidiary of Thermo, and thus the magnet listed here should be a suitable replacement |
Razor blades | From standard source | ||
Filter paper and funnel | From standard source | ||
Glass casserole dish | From standard source | ||
Orbital shaker | From standard source | ||
Kimwipes | From standard source | ||
Clear sheet protectors | From standard source | ||
Scanner and associated TWAIN software | From standard source | ||
Image analysis software | Such as Fuji Multiguage, or equivalent | ||
Microsoft Excel | Or equivalent |