Biotechnology Journal - 2007 - Melissis - One Step Purification of Taq DNA Polymerase Using Nucleotide Mimetic Affinity

Biotechnol. J. 2007, 2, 121–132 DOI 10.1002/biot.200600191 www.biotechnology-journal.
com
Research Article
One-step purification of Taq DNA polymerase using

nucleotide-mimetic affinity chromatography
Sotirios Melissis, Nikolaos E. Labrou and Yannis D. Clonis

Laboratory of Enzyme Technology, Department of Agricultural Biotechnology, Agricultural University of Athens, Athens, Greece
The thermostable Thermus aquaticus DNA polymerase (Taq Pol) has been the key factor in trans-
Received 22 September 2006
forming the initial PCR method into one with huge impact in molecular biology and biotechnolo- Revised 10 November 2006
gy. Therefore, the development of effective affinity adsorbents for the purification of of Taq Pol, as Accepted 12 November 2006
well as other DNA polymerases, attracts the attention of the enzyme manufacturers and the re-
search laboratories. In this report we describe a simple protocol for the purification of Taq Pol
from E. coli lysates, leading to enzymes of high specific activity and purity. The protocol is based
on a single affinity chromatography step, featuring an immobilized ligand selected from a struc-
ture-biased combinatorial library of dNTP-mimetic synthetic ligands. The ligand library was
screened for its ability to bind and purify Taq Pol from E. coli lysates. One immobilized ligand
(mABSGu) of the general formula X–Trz–Y, bearing 9-aminoethylguanine (AEGu) and aniline-2-
sulfonic acid (mABS) linked on the triazine scaffold (Trz), displayed the highest purifying ability.
Adsorption equilibrium studies with this affinity ligand and Taq Pol determined a dissociation con-
stant (KD) of 0.12 mM for the respective complex, whereas ATP prevented the formation of the
mABSGu-Taq Pol complex. The mABSGu affinity adsorbent was exploited in the development of a
facile Taq Pol purification protocol, affording homogeneous enzyme (>99% purity, ~61 500 U/mg)
in a single chromatography step. Quality control tests showed that Taq Pol purified on the
mABSGu affinity adsorbent is free of nucleic acids and contaminating nuclease activities.
Keywords: Affinity chromatography · Biomimetic ligand · Combinatorial chemistry · Ligand design · Taq DNA polymerase
1 Introduction The design, synthesis and selection of affinity lig-

ands have progressed rapidly because of the accumulat-
The purification and isolation of biologically active pro- ed knowledge from combinatorial chemistry, structural
teins from complex mixtures remains one of the most biochemistry and computational chemistry [3, 4]. Re-
challenging tasks in biotechnology. Affinity chromatogra- cently, we reported a combinatorial library of novel nu-
phy [1–3] exploits the ability of biologically active macro- cleotide-mimetic synthetic ligands for DNA-recognizing
molecules to form specific and reversible complexes with enzymes, using Pyrococcus furiosus DNA polymerase
appropriate affinity ligands, and undoubtedly is the most (Pfu Pol) as a proof-of-principle example [5]. The nu-
specific and effective enzyme purification technique. cleotide-mimetic synthetic ligands were generated fol-
lowing the ‘structure-guided combinatorial method’ [3],
according to which, the ‘winner’ ligand is selected from
Correspondence: Professor Yannis D. Clonis, Laboratory of Enzyme Tech- an intentionally biased ligand library based on a ration-
nology, Department of Agricultural Biotechnology, Agricultural University ally designed ‘lead’ ligand. In the present study we in-
of Athens, 75 Iera Odos Street, 11855 Athens, Greece
vestigate the applicability of the nucleotide-mimetic lig-
E-mail: clonis@aua.gr
Fax: +30-210-5294307
and library to the selection of ligands for the affinity
chromatography purification of thermostable Thermus
Abbreviations: Taq Pol, Thermus aquaticus DNA polymerase; Pfu Pol, aquaticus DNA polymerase (Taq Pol) from recombinant
Pyrococcus furiosus DNA polymerase Escherichia coli cells.
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 121

Biotechnology
18607314, 2007, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/biot.200600191 by Cochrane Canada Provision, Wiley Online Library on [14/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Biotechnol. J. 2007, 2, 121–132
Journal
The introduction of Taq Pol into PCR has caused this at 72°C), and ending with incubation at 72°C for 10 min.
method to evolve into a widely used technique for genet- The resulting amplification product was digested with
ic analysis [6–8]. Taq Pol belongs to the family of DNA SphI and BqlII and cloned into pQE-70 digested with SphI
polymerases I (DNA Pol I), as does the E. coli DNA poly- and BqlII. Recombinant plasmids were transformed in
merase I [9]. Because of its high turnover number [10], lack E. coli XL-1 Blue strain and plated on LB media with ampi-
of proofreading activity (3’-5’ exonuclease activity) [11], cillin. Positive clones carrying the Taq Pol coding region
high temperature optimum (72–74°C), and ability to in- were verified by PCR screening and designated as pQE-
corporate 7-deaza-3-deoxyquanosine, Taq Pol has been Taq. To test for protein expression, pQETaq was trans-
used extensively in DNA sequencing and other molecular formed into E. coli M15 (pREP4). One colony was picked
biology techniques [12, 13]. This enzyme holds academic up and was grown on LB media with 100 µg/mL ampicillin
interest and commercial importance and was, therefore, and 25 µg/mL kanamycin at 37°C. An overnight culture
chosen for further evaluation of the ligand library. A suc- (50 mL) was used to inoculate 1 L (500 mL × 2) LB with
cessful affinity ligand selected from the library should be ampicillin and kanamycin. The cultures were grown at
able to provide, in a single chromatography step and good 37°C until the OD600 was 0.6, induced with isopropyl β-D-
yield, pure Taq Pol suitable for molecular biology applica- thiogalactopyranoside (IPTG, 1 mM or 0.5 mM) and al-
tions. lowed to grow for further 12 h. Cells were harvested
(10 000 × g, 10min) and suspended in Tris-HCl buffer
(50 mM, pH 7.9, 1 mM EDTA, 2 mM PMSF). The cell sus-
2 Materials and methods pension was disrupted by sonication on ice (5 min, 10-s
pulse with 10-s pause intervals) and centrifuged (13 000 ×
2.1 Materials g, 15 min, 4°C). The supernatant was incubated at 75°C
for 1 h, cooled on ice for 20 min, centrifuged (13 000 × g,
The pQE-70 vector and E. coli M15 (pREP4) strain was 15 min, 4°C) to remove denatured proteins, and analyzed
purchased from Qiagen (Germany). E. coli XL-1 Blue by SDS-PAGE, along with cell debris previously solubi-
strain was purchased from Stratagene (USA). Taq Pol, Pfu lized in 8 M urea.
Pol and deoxynucleotide triphosphates (dNTPs) were pur-
chased from Promega (UK). The pTacTaq vector, carrying 2.4 Preparation of pre-treated cell extract
the Taq Pol coding region was a kind gift of Dr. J. F. David-
son (Department of Pathology, University of Washington, Recombinant Taq Pol was expressed in E. coli M15
Seattle, USA). Protamine sulfate was obtained from Sig- (pREP4) as described above. Cells were harvested (10 000
ma-Aldrich (USA). × g, 10 min) and stored at –20°C. Cell paste (15 g) was sus-
pended in 15 mL Tris-HCl buffer (50 mM, pH 7.9, 1 mM
2.2 Construction of the ligand library EDTA, 2 mM PMSF). The cell suspension was disrupted
by sonication on ice (5 min, 10 s pulse with 10-s pause in-
The rational design and the chemical synthesis of the lig- tervals) and centrifuged (13 000 × g, 15 min, 4°C). The su-
and library were described before [5]. The chemical for- pernatant was incubated at 75°C for 1 h, cooled on ice for
mulae of the ligands are shown in Table 1. 20 min and centrifuged (13 000 × g, 15 min, 4°C) to remove
denatured proteins. Nucleic acids were removed by
2.3 Construction of the expression plasmid pQETaq adding a streptomycin sulfate solution (30%) to a final
concentration of 3 mg/mL. The mixture was gently stirred
Based on the DNA sequence of Taq Pol (GenBank acces- for 1 h at 4°C and centrifuged (13 000 × g, 15 min, 4°C).
sion no. J04639), two primers were synthesized: (a) the N- The supernatant was kept at –20°C for several months.
terminal sense primer 5´-TTTTGCATGCGGGGGATGCT-
GCCCCTCTTT-3´, which carries a unique SphI site 2.5 Screening of the library of immobilized ligands
(underlined), which includes an ATG starting site of trans-
lation, followed by the sequence coding the first 7 amino Chromatographic procedures were performed at 4°C us-
acids, and (b) the antisense primer 5´-TTTT- ing dialyzed pre-treated cell extract made in either Tris-
AGATCTTCACTCCTTGGCGGAGAGCCAGTC-3´, HCl buffer (20 mM, pH 7.5, 3 mM MgCl2) or potassium
which carries a unique underlined BqlII restriction site, phosphate buffer (20 mM, pH 7.5, 3 mM MgCl2). The ad-
followed by the stop codon. DNA amplification was per- sorbents bearing the formulae of Table 1 (1 mL, 0.9 g moist
formed using 2.5 U Pfu Pol in 100-µL reaction mixture of weight gel) were equilibrated in the same buffer and
PCR reaction buffer (supplied by vendor), 10 pmol of each loaded with pre-treated extract (0.6 mL, 1 mg protein), be-
primer, 0.2 mM of each dNTP and 50 ng of pTaq plasmid fore being washed with equilibration buffer until effluent
DNA. A standard PCR protocol [14] was applied to ampli- absorbance (A280) was less than 0.01. Bound proteins
fy the approximately 2500-bp fragment of the Taq Pol were eluted with 1 M KCl (4 mL) in Tris-HCl buffer (20 mM,
gene, using 35 amplification cycles (1 min at 94°C, 3.5 min pH 7.5, 3 mM MgCl2) or potassium phosphate buffer
122 © 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

123
www.biotechnology-journal.com
R2
–
Table 1. The structures of the ligands of the combinatorial library [5]
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

R1
–
Ligand designation
Biotechnol. J. 2007, 2, 121–132
AMSAd
AESAd
mABS
AEGu
AEAd
pABS
oABS
AMS
ABP
APP
AES
Entry
10
11
1
3
4
9
Biotechnol. J. 2007, 2, 121–132
R2

R1
Ligand designation
mABSAd
pABSAd
oABSAd
AMSGu
ABPGu
APPGu
AESGu
ABPAd
APPAd
Journal
Table 1. Continued
Biotechnology
Entry
124
12
13
14
15
16
17
18
19
20
Biotechnol. J. 2007, 2, 121–132 www.biotechnology-journal.com
Table 1. Continued
Entry Ligand designation R1 R2
21 oABSGu
22 mABSGu
23 pABSGu
24 AEAd-AEAd
25 AEGu-AEAd
26 AEGu-AEGu
(100 mM, pH 7.5), depending on the equilibration buffer with the equilibrating buffer until the effluent absorbance
used originally. The fractions collected (4 mL each) were (A280) was less than 0.01. Adsorbed protein was eluted
dialyzed against water, lyophilized and the purity of the with 100 mM potassium phosphate buffer, pH 7.0, 7.5 and
Taq Pol recovered was analyzed by SDS-PAGE [16]. 8.0. Collected fractions (4 mL) were dialyzed against wa-
ter, lyophilized and analyzed by SDS-PAGE [16].
2.6 Effect of pH on the purification of Taq Pol from the
adsorbent mABSGu 2.7 Purification of Taq Pol on the adsorbent mABSGu
Chromatographic procedures were performed at 4°C. A Chromatographic procedures were performed at 4°C. A
column containing adsorbent no. 22 (mABSGu) (1 mL, column containing adsorbent no. 22 (mABSGu) (1 mL,
0.9 g moist weight gel) was equilibrated, in three separate 0.9 g moist weight gel) was equilibrated with sodium
experiments, with 20 mM potassium phosphate buffer, phosphate buffer (Na2HPO4) (20 mM, pH 7.0, 2 mM
pH 7.0, 7.5 and 8.0, containing 3 mM MgCl2, respectively. MgCl2). A sample of pre-treated cell extract (1.5 mL,
A sample of pre-treated cell extract (0.6 mL, 1 mg protein) 2.5 mg protein) previously dialyzed against the same
previously dialyzed against the same equilibration buffer, equilibration buffer was applied to the adsorbent. The col-
was applied to the adsorbent. The column was washed umn was washed with the equilibration buffer (Na2HPO4;

Biotechnology
Biotechnol. J. 2007, 2, 121–132
Journal
20 mM, pH 7.0, 2 mM MgCl2) until effluent absorbance 2.12 Quality control assays of Taq Pol purified from the
(A280) was less than 0.01. Taq Pol was eluted with 80 mM mABSGu adsorbent
sodium phosphate buffer, pH 7.0 (4 mL). Collected fraction
(4 mL) was dialyzed against water, lyophilized and ana- The purified Taq Pol from the mABSGu adsorbent was as-
lyzed by SDS-PAGE [16]. sayed for the presence of endonuclease and exonuclease
activity using standard protocols (Promega). For endonu-
2.8 Adsorption equilibrium of Taq Pol with the adsorbent clease activity, 1 µg supercoiled plasmid DNA was incu-
mABSGu bated with 5 U purified Taq Pol for 8 h at 45°C followed by
8 h at 70°C in 1× Thermophilic DNA polymerase reaction
In a total volume of 1 mL sodium phosphate buffer buffer (50 mM KCl, 10 mM Tris-HCl, pH 9.0, 1.5 mM MgCl2
(Na2HPO4, 20 mM, pH 7.0, 2 mM MgCl2), varying amounts and 1%Triton X-100). Following incubation, the mixture
of purified Taq Pol (5–50 μg), previously dialyzed in the was analyzed by agarose electrophoresis to verify the ab-
same equilibration buffer, were mixed with 10 mg of ad- sence of visible cutting. For exonuclease activity, 1 µg
sorbent (mABSGu), in the presence or in the absence of lambda DNA/Hind III marker was incubated with 5 U pu-
ATP (15 mM). The suspensions were shaken for 120 min rified Taq Pol as described above. Following incubation,
for the system to reach equilibrium. The mixture was then the DNA products were analyzed by agarose elec-
centrifuged (5000 rpm, 2 min) and the amount of unbound trophoresis to verify the absence of visible smearing.
protein in the supernatant was determined by the method
of Bradford [17]. Bound protein was calculated by sub-
tracting the amount of unbound protein from the total 3 Results
amount of protein added. The data were analyzed ac-
cording to the method of Livingston and Chase [18]. 3.1 Directed combinatorial ligand design
2.9 Determination of the apparent capacity of adsorbent Comparison of the known polymerase structures shows
mABSGu for Taq Pol that the polymerase domain in each case contains a deep
cleft, whose overall shape has been compared to that of a
Chromatographic procedures were performed at 4°C. A half-open right hand [19–21], with the β-sheet that forms
column containing adsorbent no. 22 (mABSGu) (0.12 g the base of the cleft described as the ‘palm’, and the two
moist weight gel) was equilibrated with Tris-HCl buffer subdomains that form the walls of the cleft described as
(20 mM, pH 7.5, 3 mM MgCl2) and a solution of purified ‘fingers’ and ‘thumb’ (Fig. 1). Protein sequence align-
Taq Pol in the same buffer (0.05 mg/mL, 3 mL) was con- ments [19] and mutagenesis data (reviewed in [20]) indi-
tinuously applied on the column until effluent absorbance cate that the polymerase active site is located primarily on
(280 nm) was constant. Bound Taq Pol was eluted with the palm subdomain. The topology of the palm subdomain
1 M KCl in the same equilibration buffer (1 mL). is conserved among polymerase families, and contains
the three conserved aspartate residues involved in nu-
2.10 Determination of protein concentration cleotidyl transfer. The palm subdomain also contains the
The protein concentration was determined by the method

of Bradford [17].
2.11 Assay for relative activity of Taq Pol
The relative Taq Pol activity was determined by com-

paring band intensities of PCR-amplified DNA obtained
using pre-treated E. coli extract, Taq Pol from the affini-
ty adsorbent mABSGu and commercial recombinant Taq
Pol (Promega). A standard PCR protocol (sense primer,
5´-ATGACCCTAAATATAGAAGATGAG-3´; antisense
primer, 5´-TTAATCAAGGCAGTTGTGTTGCAG-3´) was
applied to amplify a 1488-bp fragment of the pol gene of
Moloney Murine Leukemia virus, using 30 amplification
cycles (1.5 min at 94°C, 2.5 min at 52°C, 2.5 min at 72°C),
50 ng of sample DNA and 5 pmol of each primer. The PCR
products were visualized by agarose electrophoresis and Figure 1. Ribbon diagram of Taq DNA polymerase. Palm, thumb and fin-
analyzed using Kodak 1D Image Analysis Software. gers domains are labeled. The figure was generated from PDB coordinates
4KTQ using Pymol (DeLano Scientific LLC).

Figure 2. Structure of the Taq DNA polymerase

B-motif. The side-chains of Lys663 and Tyr671 are
labelled and shown as sticks. The figure was gener-
ated from PDB coordinates 4KTQ using Pymol
(DeLano Scientific LLC).
strictly conserved B-motif (RxxxKxxxFxxxYG, where x duction experiments with 1 mM and 0.5 mM IPTG
is any amino acid; Fig. 2). Amino acid residues from showed a similar expression level, therefore, 0.5 mM IPTG
motif-B contribute to dNTP recognition and binding [21, was used for enzyme expression.
22]. We have recently reported [5] the design and synthe-
sis of a directed (structure-guided) combinatorial library 3.3 Screening of the adsorbent library and determination of
of dNTPs-mimetic ligands for the purification of DNA binding capacity with Taq Pol
polymerases. dNTP-mimetic ligand design was based on
mimicking the natural interactions between dNTPs and All adsorbents (Table 1) were evaluated for their ability to
the B-motif. Considering the strictly conserved structural bind and purify Taq Pol from E. coli extract. Prior to affini-
features of B-motif and taking into account the spatial
proximity of the Tyr671 and Lys663 side chains of Taq
DNA polymerase, it is likely that putative bifunctional lig-
ands comprising, as suitable functionalities, a purine
base and a negatively charged aliphatic or aromatic
group, may function in a complementary fashion towards
the two B-motif conserved residues. These bifunctional
ligands may form aromatic stacking as well as electro-
static and hydrogen bond interactions with residues
Tyr671 and Lys663. Computer-aided molecular modeling
suggested that these triazinyl ligands bearing such func-
tionalities (Fig. 3) display a 2-D complementarity against
the target protein residues (Fig. 2). For example, the dis-
tance between Tyr671 Ca and Lys663 Ca is approximately
12.9 Å´, whereas that between the two substituents on the
triazine scaffold falls in the range 10–13 Å´.
3.2 Expression of Taq Pol DNA polymerase in E. coli M15

(pREP4)
Expression experiments of E. coli M15 (pREP4) trans-

formed with pQETaq revealed that Taq Pol was overex-
pressed in the supernatant, while almost no expression Figure 3. The structures of a B-motif-binding mimetic ligand, mABSGu
was observed in the form of inclusion bodies (Fig. 4). In- (Table 1, entry 22) (A), dGMP (B) and dGTP (C).

Biotechnology
Biotechnol. J. 2007, 2, 121–132
Journal
adsorbent AEGu-AEGu, Fig. 5E: adsorbent AEAd, Fig. 5G:

adsorbents AEAd-AEAd, AEGu-AEAd and AEGu-AEGu).
The remaining adsorbents (combinations of a purine and
an anionic substituents on the triazine scaffold) displayed
varying purification ability and enzyme recovery, depend-
ing on the elution system used (KCl in Figs. 5B and D, and
potassium phosphate in Figs. 5F and H). Adsorbent
mABSGu (Table 1, entry 22) achieved the highest purifi-
cation for Taq Pol in both elution systems tested (Figs. 5B
and F). The binding capacity of mABSGu adsorbent with
Taq Pol was 0.58 mg enzyme/g moist weight gel, and this
adsorbent was finally chosen for the purification protocol
of Taq Pol.
3.4 Adsorption equilibrium studies with adsorbent mABSGu

and Taq Pol
Figure 4. Expression of Taq Pol in E. coli M15 (pREP4). Induction was car-
ried out with 1 mM IPTG for 12 h. SDS-PAGE was performed on a 0.75-
mm-thick vertical gel containing 12.5% polyacrylamide (running gel) and Equilibrium adsorption studies were employed to charac-
2.5% stacking gel. Protein bands were stained with CBB R-250. Lane 1, terize the interaction of Taq DNA Pol with the adsorbent
heat-treated cell extract of E. coli M15 transformed with pQE-70; lane 2, mABSGu. This approach provides a relationship between
heat-treated cell extract of E. coli M15 transformed with pQETaq; lane 3, the concentration of the protein in solution and the
cell debris of E. coli M15 transformed with pQE-70 dialyzed in 8 M urea; amount of protein adsorbed to the solid phase when the
lane 4, cell debris of E. coli M15 transformed with pQETaq dialyzed in 8 M
two phases are at equilibrium [18, 23]. The model most of-
urea. The position of Taq Pol is indicated with the arrow.
ten employed for affinity systems is based on a second-or-
der reversible interaction, where the protein-ligand inter-
ty chromatography screening, the cell lysate was heat action has a characteristic binding energy [18] and pro-
treated (75°C) to precipitate contaminating cellular pro- ceeds in a monovalent fashion. At equilibrium, a familiar
teins. Streptomycin sulfate treatment was also examined Langmuir isotherm model, described by Eq. (1), can be
and adopted for removing contaminating nucleic acids. obtained [18]:
Assessment of the purifying effectiveness of the adsor-
bents was based on SDS-PAGE analysis, using all the q = qmaxc/(KD + c) (1)
amount of protein eluted from each adsorbent, thus ensur-
ing direct comparability of the adsorbents. Two screening where q is the bound adsorbate concentration at equilib-
experiments were performed, eluting bound proteins with rium (μg/mL adsorbent), c is the equilibrium liquid phase
1 M KCl (in 20 mM Tris-HCl buffer, pH 7.5, containing concentration (μg/mL), qmax is the Langmuir isotherm
3 mM MgCl2) in one case, and with 100 mM potassium maximum capacity constant (μg/mL adsorbent), and KD is
phosphate buffer pH 7.5 in the other case. Visual exami- the apparent dissociation constant. The batch adsorption
nation of the gels (Fig. 5) revealed that all adsorbents of Taq Pol on the mABSGu adsorbent is shown in Fig. 5,
bound the target enzyme, however, with significant varia- in the absence and in the presence of ATP. The calculat-
tion in their enzyme purifying ability and recovery (elu- ed dissociation constant was determined to be 0.12 mM
tion). Specifically, adsorbents having no adenine or gua- in the absence of ATP [24]. It appears that ATP is able to
nine moiety (Table 1, entries 1–7) bound the enzyme but perturb the enzyme-mABSGu complex, since the immo-
presented either low purifying ability (Figs. 5A and E, ad- bilized ligand has failed to adsorb Taq Pol in the presence
sorbents oABS, mABS, pABS) or low recovery (Figs. 5C of ATP (15 mM).
and G, adsorbents AMS, AES, APP, ABP). Furthermore,
adsorbents bearing only an adenine or guanine moiety 3.5 Purification of recombinant Taq Pol from E. coli extract
(Table 1, entries 8 and 9), or two purines (Table 1, entries on the mABSGu adsorbent
24–26) also bound the enzyme but presented either low
purifying ability (Fig. 5A: adsorbent AEGu, Fig. 5C: adsor- Prior to developing the purification protocol, the influence
bents AEAd-AEAd and AEGu-AEAd, Fig. 5E: adsorbent of pH on the binding process and desorption conditions
AEGu) or low recovery (Fig. 5A: adsorbent AEAd, Fig. 5C: were investigated. At pH 7.0, affinity adsorbent mABSGu
Figure 5. Adsorbent library screening with E. coli pre-treated extract containing Taq Pol activity. SDS-PAGE was performed on a 0.75-mm-thick vertical gel 䊳
containing 12.5% polyacrylamide (running gel) and 2.5% stacking gel. Protein bands were stained with CBB R-250. The chromatography was performed as
described under Section 2. All the protein recovered (eluted) from the adsorbent indicated was applied on each lane. A–D, library screening with KCl as
elution agent; E–H library screening with potassium phosphate buffer as elution agent. The position of Taq Pol is indicated with the arrow.

129
www.biotechnology-journal.com

Biotechnol. J. 2007, 2, 121–132
Biotechnology
Biotechnol. J. 2007, 2, 121–132
Journal
(Fig. 6) displayed the highest purifying ability and enzyme the base functions (A and G) of the enzyme’s natural sub-
recovery. At higher pH values (7.5 and 8.0) the recovery of strate (dNTP). An aminoethyl group at the N-9 position of
Taq Pol declined. Consequently, pH 7.0 was chosen for the both bases serves as a functional chemical site for at-
enzyme purification protocol. With regard to the desorp- tachment to the chlorotriazine scaffold of the activated
tion conditions, 30% glycerol, 15 mM adenosine triphos- AH gel. Furthermore, the aminoethyl group acts as a
phate, 8 mM MnCl2, 100 mM sodium pyrophosphate and ‘spacer’ that protrudes the recognition moiety of the lig-
100 mM sodium phosphate buffer were tested. Sodium and to the solvent, thus facilitating the interaction with
phosphate buffer (Na2HPO4) led to the highest purifica- the target residue; and (ii) substitution of the last chlorine
tion and recovery and, therefore, was further investigat- atom of the 1,3,5-trichlorotriazine-activated AH gel by a
ed. All other agents failed either to desorb the enzyme or negatively charged phosphonate or sulfonate group (R1),
lead to an insufficient purification (results not shown). Ap- for mimicking the phosphate group of the natural sub-
plication of stepwise desorption, using increasing con- strate (dNTP) of the target enzyme.
centration of sodium phosphate buffer (20–100 mM) re- The screening of the ligand library with pre-treated
vealed that 80 mM buffer pH 7.0 was the most effective Taq Pol cell-extract verified, as in the case with Pfu Pol [5],
and, therefore, adopted for the purification of Taq Pol the rationale of ligand design. To better evaluate the chro-
(Table 2). SDS-PAGE analysis of the purified enzyme matographic behavior of the adsorbents, two elution sys-
showed a single protein band corresponding to 94 kDa tems (KCl and potassium phosphate) were examined. As
(Fig. 7, CBB R-250 staining). Furthermore, quality control already discussed in Section 3.3, adsorbents having no
tests revealed the absence of endonuclease and exonu- purine moiety (Table 1, entries 1–7) or only purine moi-
clease activities (Promega’s standard protocols). eties (Table 1, entries 8, 9, 24–26) showed an unsatisfac-
tory purifying ability and/or enzyme recovery (yield). All
other immobilized ligands displayed variable chromato-
4 Discussion graphic behavior depending on the elution system used
(Figs. 5B, D, F and H). Employing KCl as the eluting agent
We have recently reported [5] the design, synthesis and led to a sufficient recovery and low (Fig. 5B) to modest
application of a directed (structure-guided) combinatori- (Fig. 5D, adsorbent ABPAd) purification, or low recovery
al library of dNTPs-mimetic ligands for the purification of and purification (Fig. 5D, all adsorbents except ABPAd).
P. furiosus DNA polymerase (Pfu Pol), as a proof-of-princi- On the other hand, employing potassium phosphate
ple example. buffer as eluting agent led to low recovery and purification
In the present report we explore the dNTPs-mimetic (Fig. 5F, adsorbents mABSAd and pABSAd; Fig. 5H, all
ligand library for selecting ligands and respective affinity adsorbents except ABPAd) or sufficient recovery and low
adsorbents suitable for the purification of Taq Pol. All DNA purification (Fig. 5F, adsorbent oABSAd) or sufficient re-
polymerases seem to share a common mechanism of covery and purification (Fig. 5F, adsorbents oABSGu,
action for the synthesis of deoxyribonucleic acids in a mABSGu, pABSGu; Fig. 5H, adsorbent ABPAd). A closer
template-dependent fashion. For the Taq Pol to function inspection of electrophoretic pattern obtained for the
catalytically and synthesize a new DNA strand comple- most efficient adsorbents mABSGu and pABSGu (Fig. 5F)
mentary to a single-stranded DNA template (original DNA revealed that adsorbent mABSGu offered a comparative-
molecule), the presence of a short strand of poly-dNTP ly higher recovery (from every adsorbent, all eluted protein
(complementary DNA) as a primer is required. The ther- is analyzed by SDS-PAGE and presented). Therefore, this
mostable Taq Pol has been the key factor in transforming finding, combined with the good Taq Pol purifying ability,
the initial PCR method of Kary Mullis [6] into one with an led to the choice of adsorbent mABSGu for the purifica-
vast impact in molecular biology and biotechnology. tion protocol. The encouraging behavior of the mABSGu
Available crystal structures for most known polymerase adsorbent is further supported by adsorption equilibrium
families, including the Pol-I (of which Taq Pol is a mem- studies which led to a Langmuir isotherm and revealed
ber) and Pol-II families of DNA polymerases, have con- some degree of specificity of the interaction between the
firmed that such enzymes exhibit striking similarities in dNTP-mimetic ligand and the poly-dNTP binding area of
their overall architecture, the catalytic site, and the mech- the enzyme (Fig. 8). Such a pattern indicates an interac-
anism of nucleotidyl transfer [15, 25–29]. Therefore, a tion at a fixed number of enzyme sites each of which can
structure-biased combinatorial library of dNTP-mimetic only hold one ligand, and an interaction between the en-
ligands could serve as a useful pool for selecting ligands zyme and presumably the purine moiety of mABSGu that
suitable for a targeted DNA polymerase. is perturbed by ATP.
The rationale of constructing the combinatorial ligand Taq Pol has been purified previously by methods that
library (Table 1) is, briefly, as follows: (i) substitution of one mainly depended on heat treatment of cell lysate. A clas-
of the two remaining chlorine atoms of the 1,3,5-trichloro- sical purification protocol is that proposed by Engelke et
triazine-activated AH gel by base analogues, 9-amino- al. [14], combining heat pre-treatment and precipitation
ethyladenine or 9-aminoethylguanine (R2), for mimicking with polyethyleneimine, followed by ion exchange chro-

Figure 7. Taq Pol purified from affinity adsorbent mABSGu. SDS-PAGE

was performed on a 0.75-mm-thick vertical gel containing 12.5% poly-
Figure 6. Effect of pH on the purification of Taq Pol from adsorbent mAB-
acrylamide (running gel) and 2.5% stacking gel. Protein bands were
SGu. SDS-PAGE was performed on a 0.75-mm-thick vertical gel containing
stained with CBB R-250. The chromatography was performed as described
12.5% polyacrylamide (running gel) and 2.5% stacking gel. Protein bands
under Section 2. Lane 1, pre-treated cell extract; lane 2, flow through from
were stained with CBB R-250. The chromatography was performed as de-
the column after loading; lane 3, eluted fraction with 80 mM sodium
scribed under Section 2. Lane 1, pre-treated cell extract of E. coli M15
phosphate buffer (Na2HPO4), pH 7.0. The position of Taq Pol is indicated
transformed with pQETaq; lane 2, protein elution performed with 100 mM
with the arrow.
potassium phosphate buffer, pH 7.0; lane 3 protein elution performed
with 100 mM potassium phosphate buffer, pH 7.5; lane 4 protein elution
performed with 100 mM potassium phosphate buffer, pH 8.0; The posi-
tion of Taq Pol is indicated with the arrow.
matography, leading to Taq Pol with a specific activity of

5263 U/mg. Alternative purification protocols employ
freezing and thawing of expression cell cultures (–70°C
and 75°C or room temperature) [30], or ammonium sulfate
precipitation [31]. Heat-treatment of the cell lysate, as the
only means of effecting Taq Pol purification, has been re-
ported to lead to Taq Pol with a specific activity of 342 612
U/mg [32]. Finally, another possibility is to combine
heat-treatment and ammonium sulfate-Polymin P precip-
itation, followed by phenyl-Sepharose and heparin-
Figure 8. Equilibrium adsorption of Taq Pol with affinity adsorbent mABS-
Sepharose chromatography [11]. The present strategy Gu. The plot describes the equilibrium in solid-phase Taq Pol concentra-
(Table 2) leads, in a single chromatography step, to Taq tion (Taq Pol bound) vs. the equilibrium in liquid phase Taq Pol concen-
Pol with a specific activity of 61 538 U/mg, employing a tration (Taq Pol unbound), for the Lanqmuir isotherm in the absence (o)
low-cost elution agent (80 mM sodium phosphate buffer, or in the presence of ATP (15 mM) (•).
pH 7.0). Although the 10.2-fold purification achieved nu-
merically is not impressive, nevertheless, it is sufficient
for an overexpressed enzyme. Purified Taq Pol (Figs. 7 and used in-house for routine applications, e.g., PCR screen-
9) was kept at –20°C for more than 2 years in 50 mM Tris- ing and cloning.
HCl, pH 8.0, containing 100 mM NaCl, 0.1 mM EDTA, In conclusion, in this report we proposed a new purifi-
1 mM DTT, 1% Triton X-100 and 50% glycerol, without ap- cation protocol for Taq Pol, featuring heat-treatment of E.
preciable loss of its activity. The purified Taq Pol has been coli lysate and streptomycin sulfate precipitation (pre-
Table 2. Purification protocol of recombinant Taq Pol from E. colia)
Step Volume Activity Protein Specific activity Purification Yield

(mL) (U) (mg)b) (U/mg) (fold) (%)
Pre-treated cell extract 1.5 15000 2.5 6.000 1.0 100
mABSGu affinity chromatographyc) 4.0 4000 0.065 61.538 10.2 26.6
a) Procedures were performed at 4°C. For details see text.
b) The protein concentration of the pre-treated extract was determined by the Bradford method.
c) Desorption with 80 mM sodium phosphate buffer, pH 7.0.

Biotechnology
Biotechnol. J. 2007, 2, 121–132
Journal
[11] Suzuki, M., Avicola, A. K., Hood, L., Loeb, L. A., Low fidelity mutants
in the O-helix of Thermus aquaticus DNA polymerase I. J. Biol.
Chem. 1997, 272, 11228–11235.
[12] Li, Y., Mitaxov, V., Waksman, G., Structure-based design of Taq DNA
polymerases with improved properties of dideoxynucleotide incor-
poration. Proc. Natl. Acad. Sci. USA 1999, 96, 9491–9496.
[13] Innis, M. A., Myambo, K. B,, Gelfand, D. H., Brow, M. A., DNA se-
Figure 9. An ethidium bromide-
quencing with Thermus aquaticus DNA polymerase and direct se-
stained agarose gel (1%), showing the
quencing of polymerase chain reaction-amplified DNA. Proc. Natl.
presence or absence of contaminating
Acad. Sci. USA 1988, 85, 9436–9440.
nucleic acids before and after strepto-
[14] Engelke, D. R., Krikos, A., Bruck, M. E., Ginsburg, D., Purification of
mycin sulfate treatment of heat-treated
Thermus aquaticus DNA polymerase expressed in Escherichia coli.
cell extract of E. coli M15 transformed
Anal. Biochem. 1990, 191, 396–400.
with pQETaq. Procedures were per-
[15] Hingorani, M. M., O’Donnel, M., DNA polymerase structure and
formed as described in Section 2. Lane
mechanisms. Curr. Org. Chem. 2000, 4, 887–913.
1, heat-treated cell extract of E. coli
[16] Laemmli, U. K., Cleavage of structural proteins during the assembly
M15 transformed with pQETaq before
of the head of bacteriophage T4. Nature 1970, 227, 680–685.
streptomycin sulfate treatment; lane 2,
[17] Bradford, M. M., A rapid and sensitive method for the quantitation
heat-treated cell extract of E. coli M15
of microgram quantities of protein utilizing the principle of protein-
transformed with pQETaq after strep-
dye binding. Anal. Biochem. 1976, 72, 248–254.
tomycin sulfate treatment.
[18] Livingston, A. G., Chase, H.A., Preparation and characterization of
adsorbents for use in high performance liquid affinity chromatogra-
phy. J. Chromatogr. 1989, 481, 159–174.
treated extract), followed by affinity chromatography on [19] Delarue, M., Poch, O., Tordo, N., Moras, D. et al., An attempt to uni-
the dNTP-mimetic ligand mABSGu, selected from a fy the structure of polymerases. Protein Eng. 1990, 3, 461–467.
[20] Joyce, C. M., Steitz, T. A., Function and structure relationships in
structure-biased combinatorial ligand library. The pro-
DNA polymerases. Annu. Rev. Biochem. 1994, 63, 777–822.
posed purification protocol is simple and effective, yield- [21] Steitz, T. A., DNA Polymerases: Structural Diversity and Common
ing recombinant Taq Pol of the highest specific activity so Mechanisms. J. Biol. Chem. 1999, 274, 17395–17398.
far reported. This ligand library is now being studied with [22] Li, Y., Korolev, S., Waksman, G., Crystal structures of open and closed
respect to other DNA-recognizing enzymes. forms of binary and ternary complexes of the large fragment of Ther-
mus aquaticus DNA polymerase I: structural basis for nucleotide in-
corporation. EMBO J. 1998, 17, 7514–7525.
[23] Katsos, N. E., Labrou, N. E., Clonis, Y. D., Interaction of L-glutamate
5 References oxidase with triazine dyes: selection of ligands for affinity chro-
matography J. Chromatogr. B. 2004, 807, 277–285.
[1] Bailon, P., Ehrlich, G. K., Fung, W.-J., Berthold, W. (Eds.), Affinity [24] Li, R., Dowd, V., Stewart, D. J., Burton, S. J. et al., Design, synthesis,
Chromatography: Methods and Protocols, Humana Press, Clifton and application of a protein A. Nat. Biotechnol. 1998, 16, 190–195.
2000. [25] Iyer, L. M., Koonin, E. V., Leipe, D. D., Aravind, L., Origin and evolu-
[2] Vijayalakshmi, M. A. (Ed.), Biochromatography: Theory and Prac- tion of the archaeo-eukaryotic primase superfamily and related
tice, Taylor & Francis, London 2002. palm-domain proteins: structural insights and new members. Nu-
[3] Clonis, Y. D., Affinity chromatography matures as bioinformatic and cleic Acids Res. 2005, 33, 3875–3896.
combinatorial tools develop. J. Chromatogr. A. 2006, 1101, 1–24. [26] Evans, S. J., Fogg, M. J., Mamone, A., Davis, M. et al., Improving
[4] Mondal, K., Gupta, M. N., The affinity concept in bioseparation: dideoxynucleotide-triphosphate utilisation by the hyper-ther-
evolving paradigms and expanding range of applications. Biomol. mophilic DNA polymerase from the archaeon Pyrococcus furiosus.
Eng. 2006, 23, 59–76. Nucleic Acids Res. 2000, 28, 1059–1066.
[5] Melissis, S., Labrou, N. E., Clonis, Y. D., Nucleotide-mimetic syn- [27] Hogg, M., Wallace, S. S., Doublie, S., Crystallographic snapshots of a
thetic ligands for DNA-recognizing enzymes One-step purification replicative DNA polymerase encountering an abasic site. EMBO J.
of Pfu DNA polymerase. J. Chromatog. A. 2006, 1122, 63–75. 2004, 23, 1483–1493.
[6] Mullis, K. B., Faloona, F. A., Specific synthesis of DNA in vitro via a [28] Doublie, S., Tabor, S., Long, A. M., Richardson, C. C. et al., Crystal
polymerase-catalyzed chain reaction. Methods Enzymol. 1987, 55, structure of a bacteriophage T7 DNA replication complex at 2.2 A
335–350. resolution. Nature 1998, 391, 251–258.
[7] Saiki, R. K., Scharf, S., Faloona, F., Mullis, K .B. et al., Enzymatic am- [29] Biles, B. D., Connolly, B. A., Low-fidelity Pyrococcus furiosus DNA
plification of beta-globin genomic sequences and restriction site polymerase mutants useful in error-prone PCR. Nucleic Acids Res.
analysis for diagnosis of sickle cell anemia. Science 1985, 230, 2004, 15, 176–182.
1350–1354. [30] Grimm, E., Arbuthnot, P., Rapid purification of recombinant Taq
[8] Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J. et al., Primer-di- DNA polymerase by freezing and high temperature thawing of bac-
rected enzymatic amplification of DNA with a thermostable DNA terial expression cultures. Nucleic Acids Res. 1995, 23, 4518–4519.
polymerase. Science 1988, 239, 487–491. [31] Pluthero, F. G., Rapid purification of high-activity Taq DNA poly-
[9] Datta, K., LiCata, V. J., Thermodynamics of the binding of Thermus merase. Nucleic Acids Res. 1993, 21, 4850–4851.
aquaticus DNA polymerase to primed-template DNA. Nucleic Acids [32] Desai, U. J., Pfaffle, P. K., Single-step purification of a thermostable
Res. 2003, 31, 5590–5597. DNA polymerase expressed in Escherichia coli. Biotechniques 1995,
[10] Barnes, W. M., The fidelity of Taq polymerase catalyzing PCR is im- 19, 780–784.
proved by an N-terminal deletion. Gene1992, 112, 29–35.

Biotechnology Journal - 2007 - Melissis - One Step Purification of Taq DNA Polymerase Using Nucleotide Mimetic Affinity

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Biotechnology Journal - 2007 - Melissis - One Step Purification of Taq DNA Polymerase Using Nucleotide Mimetic Affinity

Uploaded by

Copyright:

Available Formats

Biotechnol. J. 2007, 2, 121–132 DOI 10.1002/biot.200600191 www.biotechnology-journal.

One-step purification of Taq DNA polymerase using

Sotirios Melissis, Nikolaos E. Labrou and Yannis D. Clonis

1 Introduction The design, synthesis and selection of affinity lig-

© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 121

122 © 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Entry Ligand designation R1 R2

© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 125

The protein concentration was determined by the method

2.11 Assay for relative activity of Taq Pol

The relative Taq Pol activity was determined by com-

126 © 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 2. Structure of the Taq DNA polymerase

3.2 Expression of Taq Pol DNA polymerase in E. coli M15

Expression experiments of E. coli M15 (pREP4) trans-

© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 127

adsorbent AEGu-AEGu, Fig. 5E: adsorbent AEAd, Fig. 5G:

3.4 Adsorption equilibrium studies with adsorbent mABSGu

128 © 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

130 © 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 7. Taq Pol purified from affinity adsorbent mABSGu. SDS-PAGE

matography, leading to Taq Pol with a specific activity of

Table 2. Purification protocol of recombinant Taq Pol from E. colia)

Step Volume Activity Protein Specific activity Purification Yield

© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 131

132 © 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

You might also like