Purification and characterization of transcription factors.

Transcription factors (TFs) are essential for the expression of all proteins, including those involved in human health and disease. However, TFs are resistant to proteomic characterization because they are frequently masked by more abundant proteins due to the limited dynamic range of capillary liquid chromatography-tandem mass spectrometry and protein database searching. Purification methods, particularly strategies that exploit the high affinity of TFs for DNA response elements (REs) on gene promoters, can enrich TFs prior to proteomic analysis to improve dynamic range and penetrance of the TF proteome. For example, trapping of TF complexes specific for particular REs has been achieved by recovering the element DNA-protein complex on solid supports. Additional methods for improving dynamic range include two- and three-dimensional gel electrophoresis incorporating electrophoretic mobility shift assays and Southwestern blotting for detection. Here we review methods for TF purification and characterization. We fully expect that future investigations will apply these and other methods to illuminate this important but challenging proteome.

DNA affinity chromatography.

DNA-affinity chromatography has been used for the purification of DNA-binding proteins that control various cellular processes. There have been improvements in coupling methods and choice of supports over the years. The procedure for coupling 5'-aminoethyl-(dT)18 to silica activated with N-hydroxysuccinimide and a carbodiimide has been described. Also, the cyanogen bromide mediated coupling of aminoethyl-(dT)18 to Sepharose is described. Determination of (dT)18-coupling to silica and Sepharose is by 5' end-labeling an oligonucleotide containing a (dA)18 stretch of sequence and determining how much hybridizes with the (dT)18 support. Enzymatic synthesis of a double-stranded DNA-silica or Sepharose prevents modification of nucleotide bases. We have explained the use of DNA and RNA templates for template-directed enzymatic synthesis of affinity columns. DNA-affinity chromatography is a powerful method with broad applicability and we are currently extending this technology for purifying transcription factors, polymerases, and nucleases.

Affinity purification of DNA-binding proteins.

The focus of this review is on DNA affinity chromatography, which is the most powerful tool for purification of DNA binding proteins. The use of nonspecific-, sequence specific- and single stranded-DNA affinity columns in purification of various DNA binding proteins is discussed. The purification strategies for transcription factors, restriction enzymes, telomerases, DNA and RNA polymerase and DNA binding antibodies are described. Different applications of DNA affinity chromatography are presented.

Mouse alpha1-syntrophin binding to Grb2: further evidence of a role for syntrophin in cell signaling.

Syntrophins have been proposed to serve as adapter proteins. Syntrophins are found in the dystrophin glycoprotein complex (DGC); defects in the constituents of this complex are linked to various muscular dystrophies. Blot overlay experiments demonstrate that alpha-dystroglycan, beta-dystroglycan, and syntrophins all bind Grb2, the growth factor receptor bound adapter protein. Mouse alpha1-syntrophin sequences were produced as chimeric fusion proteins in bacteria and found to also bind Grb2 in a Ca2+-independent manner. This binding was localized to the proline rich sequences adjacent to and overlapping with the N-terminal pleckstrin homology domain (PH1). Grb2 bound syntrophin with an apparent KD of 563 +/- 15 nM. Grb2-C-SH3 domain bound syntrophin with slightly higher affinity than Grb2-N-SH3 domain. Crk-L, an SH2/SH3 protein of similar domain structure but different specificity, does not bind these syntrophin sequences.

Comparative studies on discrete and concatemeric DNA-sepharose columns for purification of transcription factors.

Concatemers, tandem copies of DNA elements ligated together, are widely used for the DNA affinity chromatography of transcription factors. Purification of different transcription factors using discrete, concatemeric and T18:A18 tailed DNA affinity columns was studied. Columns having a discrete DNA sequence bound by cytidylic-adenylic-adenylic-thymidylic oligonucleotide (CAAT) enhancer binding protein (C/EBP) yields significantly more green fluorescent protein-C/EBP (GFP-C/EBP) fusion protein than a concatemeric DNA column made from five tandem repeats of the same DNA sequence. For lac repressor protein, the concatemeric and T18:A18 tailed columns show greater retention times than a discrete, untailed DNA affinity column. It was observed that the T18:A18 tailed column gives better resolution than either the discrete or concatemeric columns, of mixtures containing both lac repressor and GFP-C/EBP. Discrete concatemeric and T18:A18 tail columns all bound the Sp1 transcription factor and showed similar retention. The T18:A18 tailed column gives higher yield for Sp1 than the other columns. Our study shows concatemeric columns do not have any distinct advantage for the three different transcription factors we studied including Sp1, the original justification for the concatemeric approach.

Bi-column method for purification of transcription factors.

A novel bi-column method for purifying transcription factors, using two different columns and two different elution strategies is described. Lac repressor elutes at lower heparin concentrations from a lower affinity lactose operatorl (Op1)-Sepharose column than from a higher affinity column containing the same sequence with a T18:A18 tail (Op1T18). A bi-column method was developed in which lac repressor fusion protein is eluted from the Op1-Sepharose with a low heparin concentration and trapped on a Op1T18-Sepharose column because of its higher affinity for the lac repressor protein. Elution of the latter column with buffer containing a high salt concentration gives significantly purer transcription factor than the conventionally used single column methods and removes residual heparin. Highly pure CAAT enhancer binding protein and the B3 transcription factor are also obtained by using variants of this bi-column method.

Oligomerization of mouse alpha 1-syntrophin and self-association of its pleckstrin homology domain 1 containing sequences.

Syntrophins are known to self-associate to form oligomers. Mouse alpha 1-syntrophin sequences were produced as chimeric fusion proteins in bacteria and were found to also oligomerize and in a micromolar Ca(2+)-dependent manner. The oligomerization was localized to the N-terminal pleckstrin homology domain (PH1) or adjacent sequences; the second, C-terminal PH2 domain did not show oligomerization. PH1 was found to self-associate, and calmodulin or Ca(2+)-chelating agents such as ethylene glycol bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA) could effectively prevent this oligomerization. A single calmodulin bound per syntrophin to cause inhibition of the precipitation. Since calmodulin inhibited syntrophin oligomerization in the presence or absence of Ca(2+), Ca(2+) binding to syntrophin is responsible for the inhibition by EGTA of syntrophin oligomerization.

Catalytic chromatography.

Catalytic chromatography exploits both specific biological affinity and catalytic specificity to selectively purify enzymes. Two different applications are presented. Purification of EcoRI restriction endonuclease to apparent homogeneity was accomplished in a single step with significantly greater yield and purification than was obtained with affinity chromatography. An attempt to purify the multiple DNA polymerase activities of Escherichia coli was also developed. Five well-resolved peaks of DNA polymerase activity were fractionated. In this new chromatographic mode, the enzyme binds immobilized substrate coupled to a column in the absence of some required cofactor. When the missing cofactor is added, the enzyme converts substrate to product and selectively elutes from the column.

Temperature dependence of DNA affinity chromatography of transcription factors.

Oligonucleotides bound by the CAAT enhancer binding protein (C/EBP), the lactose repressor, and Gal4 were chemically coupled to cyanogen bromide-activated Sepharose and the temperature dependence of transcription factor chromatography was characterized. Each transcription factor was applied to the appropriate column and eluted using a salt gradient at several temperatures. Each transcription factor showed a unique behavior. As temperature was increases, less salt was required to elute C/EBP, more salt was required to elute lac repressor, while Gal4 showed a biphasic dependency with the amount of salt first decreasing between 4 and 19 degrees C and then increasing above 19 degrees C. This temperature dependence is not due to protein or DNA unfolding but rather is a property of complex formation. By loading a column, washing it at a permissive temperature, and then rapidly changing the column temperature, highly selective elution can be obtained. The thermodynamics of this temperature effect are different for the binding of specific and nonspecific DNA sequences, making chromatography at different temperatures a potentially important way of purifying transcription factors.

Phosphorylation of dystrophin and alpha-syntrophin by Ca(2+)-calmodulin dependent protein kinase II.

A Ca(2+)-calmodulin dependent protein kinase activity (DGC-PK) was previously shown to associate with skeletal muscle dystrophin glycoprotein complex (DGC) preparations, and phosphorylate dystrophin and a protein with the same electrophoretic mobility as alpha-syntrophin (R. Madhavan, H.W. Jarrett, Biochemistry 33 (1994) 5797-5804). Here, we show that DGC-PK and Ca(2+)-calmodulin dependent protein kinase II (CaM kinase II) phosphorylate a common site (RSDS(3616)) within the dystrophin C terminal domain that fits the consensus CaM kinase II phosphorylation motif (R/KXXS/T). Furthermore, both kinase activities phosphorylate exactly the same three fusion proteins (dystrophin fusions DysS7 and DysS9, and the syntrophin fusion) out of a panel of eight fusion proteins (representing nearly 100% of syntrophin and 80% of dystrophin protein sequences), demonstrating that DGC-PK and CaM kinase II have the same substrate specificity. Complementing these results, anti-CaM kinase II antibodies specifically stained purified DGC immobilized on nitrocellulose membranes. Renaturation of electrophoretically resolved DGC proteins revealed a single protein kinase band (M(r) approximately 60,000) that, like CaM kinase II, underwent Ca(2+)-calmodulin dependent autophosphorylation. Based on these observations, we conclude DGC-PK represents a dystrophin-/syntrophin-phosphorylating skeletal muscle isoform of CaM kinase II. We also show that phosphorylation of the dystrophin C terminal domain sequences inhibits their syntrophin binding in vitro, suggesting a regulatory role for phosphorylation.

Comparative studies on chemically and enzymatically coupled DNA-Sepharose columns for purification of a lac repressor chimeric fusion protein.

The length of a DNA sequence attached to an affinity chromatography column affects column retention of transcription factors. Even when unrelated sequences such as a poly(A):poly(T) tail are included in a DNA sequence, transcription factors such as the lac repressor are bound more tightly by the column. The position of the additional sequences is also important. To compare coupling procedures, an identical DNA sequence was covalently attached to Sepharose by chemical coupling or produced enzymatically by template driven enzymatic primer extension. These two types of supports, containing the O1 operator sequence bound by lac repressor, were packed into identical columns and compared by purification of a lac repressor-beta-galactosidase fusion protein. We found that the purity and yield of proteins eluted from the two columns were similar. Overall, the results suggest that there is no significant advantage to either type of support for the purification of some proteins. The study revealed a potentially important effect of the length of DNA sequences on column selectivity.

Heparin elution of transcription factors from DNA-Sepharose columns.

A novel method using heparin for eluting transcription factors from DNA-Sepharose columns was characterized. CAAT enhancer binding protein (C/EBP) or lac repressor fusion proteins were both eluted with heparin from columns containing specific DNA sequences coupled to cyanogen bromide activated Sepharose. The amount of the lac repressor chimera which eluted from the column was shown to increase with increases in the mobile phase heparin concentration. The elution of the protein was also shown to be dependent on the amount of DNA coupled to the column and more protein eluted from columns containing lesser amounts of DNA. These data suggest that heparin and DNA compete for binding to the protein; this competition causes elution. Comparison of heparin- and salt-eluted protein demonstrated the heparin-eluted fraction was significantly purer and comparable to that obtained by elution with isopropyl beta-D-thiogalactopyranoside, a lactose analog. Heparin elution represents an important new tool in the purification of transcription factors and other DNA-binding proteins by DNA affinity chromatography.

Apocalmodulin.

Intracellular Ca2+ is normally maintained at submicromolar levels but increases during many forms of cellular stimulation. This increased Ca2+ binds to receptor proteins such as calmodulin (CaM) and alters the cell's metabolism and physiology. Calcium-CaM binds to target proteins and alters their function in such a way as to transduce the Ca2+ signal. Calcium-free or apocalmodulin (ApoCaM) binds to other proteins and has other specific effects. Apocalmodulin has roles in the cell that apparently do not require the ability to bind Ca2+ at all, and these roles appear to be essential for life. Apocalmodulin differs from Ca2+-CaM in its tertiary structure. It binds target proteins differently, utilizing different binding motifs such as the IQ motif and noncontiguous binding sites. Other kinds of binding potentially await discovery. The ApoCaM-binding proteins are a diverse group of at least 15 proteins including enzymes, actin-binding proteins, as well as cytoskeletal and other membrane proteins, including receptors and ion channels. Much of the cellular CaM is bound in a Ca2+-independent manner to membrane structures within the cell, and the proportion bound changes with cell growth and density, suggesting it may be a storage form. Apocalmodulin remains tightly bound to other proteins as subunits and probably hastens the response of these proteins to Ca2+. The overall picture that emerges is that CaM cycles between its Ca2+-bound and Ca2+-free states and in each state binds to different proteins and performs essential functions. Although much of the research focus has been on the roles of Ca2+-CaM, the roles of ApoCaM are equally vital but less well understood.

Pleckstrin homology domain 1 of mouse alpha 1-syntrophin binds phosphatidylinositol 4,5-bisphosphate.

Mouse alpha 1-syntrophin sequences were produced as chimeric fusion proteins in bacteria and found to bind phosphatidylinositol 4, 5-bisphosphate (PtdIns4,5P2). Half-maximal binding occurred at 1.9 microM PtdIns4,5P2 and when 1.2 PtdIns4,5P2 were added per syntrophin. Binding was specific for PtdIns4,5P2 and did not occur with six other tested lipids including the similar phosphatidylinositol 4-phosphate. Binding was localized to the N-terminal pleckstrin homology domain (PH1); the second, C-terminal PH2 domain did not bind lipids. Key residues in PtdIns4,5P2 binding to a PH domain were found to be conserved in alpha-syntrophins' PH1 domains and absent in PH2 domains, suggesting a molecular basis for binding.

Transcription factor-green fluorescent protein chimeric fusion proteins and their use in studies of DNA affinity chromatography.

A new plasmid, pJ22, was produced by introducing the enhanced green fluorescent protein (GFP) coding sequence into the pET28 plasmid while retaining much of the multiple cloning site. This new plasmid was then used to produce a chimeric fusion protein containing the DNA-binding region of the rat liver CAAT enhancer binding protein (C/EBP) fused to the COOH-terminus of GFP. This new GFP-C/EBP fusion protein also contains (His)6 to facilitate purification by Ni(2+)-agarose and several other useful features. The plasmid and protein were developed to allow us to more rapidly investigate the DNA-Sepharose affinity chromatography of transcription factors. The GFP-C/EBP protein is virtually identical in its DNA-binding properties to a well-characterized, bacterially expressed protein called C/EBP 62 which has been shown to mimic rat wild-type C/EBP DNA-binding. GFP-C/EBP also binds to DNA-Sepharose which contains the CAAT element and is eluted by a salt gradient. Salt-dependent elution was highly temperature-dependent over the range of 4-19 degrees C. Since temperature-dependent DNA-binding has also been reported for other DNA-binding proteins, this may also occur with other transcription factors. DNA-affinity chromatography gave higher purity than that obtained by Ni(2+)-agarose chromatography and chromatography on the same DNA-Sepharose column at two different temperatures resulted in the greatest purification, to near homogeneity. This temperature-dependent affinity chromatography provides an important new approach to transcription factor purification.

Ca2+-calmodulin binding to mouse alpha1 syntrophin: syntrophin is also a Ca2+-binding protein.

Syntrophins are peripheral membrane proteins which have been found associated with dystrophin, the protein product of the Duchenne muscular dystrophy gene locus. Mouse alpha1 syntrophin binds the COOH-terminal domain of dystrophin, and calmodulin inhibits this interaction in a Ca2+-dependent fashion. Where calmodulin binds to syntrophin was investigated by constructing fusion proteins containing different regions of syntrophin's sequence. Syntrophin contains at least two regions which bind calmodulin in different ways. The COOH-terminal 24 residues contain a Ca2+-calmodulin binding site, named CBS-C, which binds calmodulin with an apparent affinity of 18 nM and which is highly conserved in all syntrophins. The amino-terminal 174 residue section of syntrophin contains other calmodulin binding, and binding occurs in either the presence or absence of Ca2+ with an apparent affinity of 100 nM. Syntrophin was shown to bind Ca2+ at two or more sites residing in the amino-terminal 274 residues, and Ca2+ binding to syntrophin affects calmodulin binding at high concentrations of syntrophin. Syntrophin A (residues 4-274) is predominantly a dimer in EGTA. A model of syntrophin's complex interactions with itself (i.e., oligomerization), calmodulin, and Ca2+ is presented.

Hybrid selection with cDNA-silica.

A partial length ovalbumin cDNA-silica was produced using primer extension of (dT)18-silica with annealed partial ovalbumin RNA and reverse transcriptase. This cDNA-silica was used to test whether full-length ovalbumin RNA could be selectively purified in the presence of a large excess of other (mouse muscle) RNA. The cDNA-silica synthesized had minimally 60 pmol cDNA per gram silica and had a capacity for full-length ovalbumin RNA of minimally 38 micrograms/g. Even when other RNA was present in greater than 1000-fold excess, ovalbumin RNA was selectively retained by the cDNA-silica and was eluted in yields of 43% with an enrichment which varied over the range of 29-162-fold in various experiments. These results show that even rare RNAs can be selectively purified in high yield using cDNA-silica. The importance of these results to hybrid selection and subtractive library preparation is discussed.

Ca2+-calmodulin binds to the carboxyl-terminal domain of dystrophin.

The unique COOH-terminal domain of dystrophin (mouse dystrophin protein sequences 3266-3678) was expressed as a chimeric fusion protein (with the maltose-binding protein), and its binding to calmodulin was assessed. This fusion protein, called DysS9, bound to calmodulin-Sepharose, bound biotinylated calmodulin, caused characteristic changes in the fluorescence emission spectrum of dansyl-calmodulin, and had an apparent affinity for dansyl-calmodulin of 54 nM. Binding in each case was Ca2+-dependent. The maltose-binding protein does not bind calmodulin, and thus binding resides in the dystrophin-derived sequences. Deletion mutation experiments further localize the high affinity calmodulin binding to mouse dystrophin protein sequences 3293-3349, and this domain contains regions with chemical characteristics found in the calmodulin-binding sequences in other proteins. The COOH-terminal domain provides sites of attachment of dystrophin to membrane proteins, and calmodulin binding may modulate these interactions.