Inr. J Pepiide Prorein Ref. 46, 1995. 149-154 Prinred in Belgium - d l rights reserved

Copyrigh! 0 Munksgaard 1995

INTERNATIONAL JOURNAL OF PEPTIDE & PROTEIN RESEARCH

ISSN 0367-8377

Synthesis of the basic-helix-loophelix region of the immunoglobulin enhancer binding protein E47 and evaluation of its structural and

DNA binding properties

PATRICIA BISHOP, CORY JONES, INDRANEEL GHOSH and JEAN CHMIELEWSKI

Department of Chemistry, Purdue University, West Lafuyette, Indiana, USA

Received 23 January, accepted for publication 26 March 1995

The basic-helix-loop-helix (bHLH) region of the immunoglobulin enhancer binding protein E47 (IEB E47) was prepared in high yield by a solid-phase peptide synthesis methodology. Size-exclusion chromatography, sedimentation equilibrium and cross-linking data showed that the synthetic bHLH protein, 1, was dimeric, and higher-order aggregates of trimer and tetramer were also observed. The circular dichroism spectrum of 1 showed a high helical content, which increased upon addition of DNA containing the K E ~ sequence. Gel mobility shift experiments showed that protein 1 bound sequence specifically to the lcE2 sequence with a binding constant of 10 - lo M ~ , and had an affinity for other E box sequences as well. Comparisons between the co-crystal structure of IEB E47 with DNA and structural studies in solution showed lower helical con- tents in solution as would have been predicted from the crystal structure. 0 Munksgaard 1995.

Key words: helix-loop-helix; DNA binding; solid-phase synthesis

Immunoglobulin enhancer binding proteins El2 and E47 (IEB El2 and E47) were the first to be classified as basic-helix-loop-helix (bH1,H) proteins by Balti- more et al. (1) IEB E l2 and E47 play an important role in activating expression of the immunoglobulin light chain gene through binding at the icE2 enhancer site (2). IEB E47 also binds thepE2 and pE5 enhancer sites in a pre-T-cell line, resulting in immunoglobulin heavy chain transcription (3), and regulates tissue specific gene

Abbreviations: All amino acids exccpt glycine are of the L- configuration. Standard abbreviations for amino acids, peptides and protecting groups follow the recommendations of the IUPAC-UIB Commision on Biochemical Nomenclature. Other abbreviations used in the text are as follows: A, alanine; C, cysteine; D, aspartic acid; E, glutamic acid; F, phenylalanine; G, glycine; H, histidine; I, iso- leucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, pro- line; Q, glutamine; R, arginine; S, serine; T, threonine; V, valine; w , tryptophan; Y , tyrosine; bHLH, ba5ic-helix-loop-helix; HLH, helix-loop-helix; CD,circulardichroism, DIPCDI,N,N’-diisopropyl- carbodiimide; DIEA, N,N’-diisopropylethylamine; DMF, N,N’- dimethylformamide; DTT, dithiothreitol; Fmoc, 9-fluorenylmethyl- oxycarbonyl; HPLC, high-performance liquid chromatography; HOBt, 1-hydroxybenzotriazole; Mtr, 4-methoxy-2,3,6-trimethyl- benzenesulfonyl; NMP, N-methyl-2-pyrrolidinone; SDS-PAGE, sodium dodecylsulfate-polyacrylamide electrophoresis; BS3, bis- sulfosuccinimidyl-suherate; Sulfo-EGS, ethyleneglyco-bis-sulfo- succinimidylsuccinate; TFA, trifluoroacetic acid.

transcription by forming heterodimeric complexes with other bHLH proteins (4). The gene for IEB E47 lies at a t( 1 : 19) chromosomal breakpoint, which has been found in 30% of pediatric pre-B cell acute lymphoblas- tic leukemia patients (9, which suggests a role for the resulting fusion protein in abnormal cellular differen- tiation.

The key role of IEB E47 in gene activation has fueled efforts to determine the nature of its sequence-specific interactions with DNA. Early mutagenesis studies with IEB E47 delineated a basic region which was essential for DNA binding, and two regions with the potential to form amphiphilic helices which mediated dimerization (6). More recently the crystal structures of a number of bHLH proteins have been solved [Max (7), USF (8), IEB E47 (9)] which confirmed the initial prediction of the bHLH structure. Within the co-crystal structure of IEB E47 with DNA, the basic region bound in the major groove of DNA in a helical conformation, with the two amphiphilic helices forming the dimeric, four- helix bundle interface.

In an effort to extend the structure obtained in the solid state to that which exists in solution, we have chemically synthesized the 59 amino acid bHLH region of IEB E47, 1, by solid-phase peptide synthesis, and studied its conformation, self-assembling properties, and specific affinity for a number of DNA sequences.

149

P. Bishop et al.

basic region helix 1 loop helix 2

(1) ERRMANNARERVRVRDINEAFRELGRMCQMHLKSDKAQTKLLILQQAVQVILGLEQQVR

EXPERIMENTAL PROCEDURES

Protein synthesis Protein 1 was synthesized by a solid-phase peptide synthesis procedure on Wang's p-alkoxybenzylalcohol resin (lo), using a fluorenylmethyloxycarbonyl (Fmoc) -based strategy (1 1) on a Biosearch 9500 peptide syn- thesizer. The side-chain protecting groups were as follows: tert-butyl for Asp, Glu, Ser, Thr and Tyr; trityl for His and Cys; 4-methoxy-2,3,6-trimethylbenzene- sulfonyl (Mtr) for Arg; and tert-butyloxycarbonyl for Lys. The initial amino-acid substitution level for the synthesis was 0.6 mmol of Fmoc-Arg(Mtr) per gram of resin, and the synthesis was initiated on a 0.5 mmol scale. Prior to each coupling step, the N-terminal Fmoc group was removed by treatment with a 30% piperidine solution in DMF. Each amino acid was incorporated by two 1 h couplings using a 0.15 M solution of the amino acid and HOBT in NMP (4 eq. each), and an equimolar amount of a 0.3 M solution of DIPCDI in CH2C12. Following the coupling reactions, the resin was treated with a 0.3 M solution of acetic anhydride and DIEA in CH2C12 to block any unreacted amino- groups. Upon completion of the synthesis, the protein was cleaved from the solid support and the side-chain protecting groups were removed by suspending the resin-bound peptide (0.2 mmol) in a solution of TFA (9 ml), thioanisole (0.5 mL), ethanedithiol(O.3 mL) and anisole (0.2 mL) under nitrogen for 4 h at RT. The reaction mixture was filtered directly into dry ether, and the solution was stored at -20°C for 4 h. The protein was collected via centrifugation, washed twice with ether, and dried in vucuo.

The crude protein was purified to homogeneity by preparative reversed-phase HPLC using a Waters Delta-Pak CIS radial compression column ( 5 cm x 28 cm), a flow rate of 80 mL/min, and mobile phases of A (100% CH3CN/O.l% TFA) and B (100% HzO/ 0.1% TFA) with a linear gradient (15-60% A) over 60 min. The eluant was monitored at 214 nm. This pro- cedure provided an overall yield of 10% for the syn- thesis and purification of protein 1.

The protein was characterized by electrospray mass spectrometry: 6989, calc 6988.7 (Fig. 2); amino acid analysis: Ala ( 5 ) 5.01, Asx ( 5 ) 4.90, Glx (12) 12.53, Phe (1) 0.95, Gly (2) 2.07, His (1) 1.02, Ile (3) 2.62, Lys (3) 3.18, Leu (7) 7.57, Met (3) 2.79, Arg (9) 9.20, Ser (1) 0.83, Thr (1) 1.04, Val (5) 4.81; and amino-terminal sequencing through 58 amino acids which was consis- tent with the proposed primary sequence.

Dimerizution studies

Size-exclusion chromatography. A solution of 1 (0.5 mL, 1 5 0 ~ ~ ) was passed through Sephadex G-50 (90x 1.6 cm column) with a flow rate of 0.3 mL/min at 4' C, and equilibrated in a 10 mM phosphate, 150 mM NaC1, 1 mM DTT buffer at pH 6.0 (Buffer A). The eluant was monitored at 214 nm. A standard molecular-weight curve was generated using bovine serum albumin, car- bonic anhydrase, cytochrome C and aprotinin, and the apparent molecular weight of 1 was determined by interpolation.

Equilibrium sedimentation. A solution of 1 (100 pL, 1 5 0 p ~ , in a buffer containing 2 0 m ~ phosphate, 160 mM KCl, 5 mg/mL Dextran T 40, pH 7.4) was cen- trifuged at lo5 rpm at 4°C for 48 h using a TLA 100 rotor in a Beckman Optima TLX Ultracentrifuge according to the method of Pollet (12). Ten 10 pL aliquots were removed and assayed for total protein concentration using the fluoraldehyde peptide/protein assay reagent (Pierce Chemical Company).

Cross-linking experiments. Protein 1 was allowed to react with 10 eq. ofthe BS3 (Pierce Chemical C.) cross- linking reagent in Buffer A at RT for 12 h. All reactions were diluted to 100 PM and then analyzed by 17% SDS-PAGE. Coomassie Brilliant Blue R-250 was used to stain the protein bands.

Circular dichroism spectroscopy Spectra of 1 were recorded on a Jasco 5600 spectro- polarimeter at 4" C using Buffer A. The concentration of the stock protein solution was quantitated using the Bio-Rad protein assay. The spectra were an average of three scans from 200 to 260 nm, with a resolution of 0.2 nm and a scan speed of 10 nm/min. The helical content was calculated from the value of the mean resi- due ellipticity at 222 nm: % helicity = ([ 191222 - [ O ] o ) / ([ &,,, - [ 010 (13). In the CD experiments with equimo- lar amounts of added DNA, a C D spectrum of the DNA was first obtained and then subtracted from the spectrum of 1 plus the DNA.

DNA binding A 29 base-pair fragment of DNA containing the lcE2 E box was cloned into the EcoRIIBamHI site of pUC19 (New England Biolabs). After amplification, this plas- mid was digested with NarI and HindIII to yield a 21 1 bp fragment containing three E box sequences: CAGGTG ( K E ~ ) , CAGCTG and CAACTG. Alterna- tively, the plasmid was cut with: EcoRI and HindIII to

150

The bHLH region of IEB E47

data correlate well with those obtained for longer se- quences containing the bHLH of IEB E47 (6, 14).

Covalent cross-linking of the protein aggregates, such that they could be observed under denaturing condi- tions, was used as an additional confirmation of dimer formation. Cross-linked complexes of 1 were seen at 1 0 0 ~ ~ as well as at 7 5 0 ~ ~ protein concentration (Fig. 3). The approximate molecular weight of the lower band was 6925, which corresponds to monomer, and the upper band had an approximate molecular weight of 11 005, which corresponds to a cross-linked dimer. Higher-ordered aggregates were also evident at the higher concentration of 1; bands at molecular weight 21 925 and 29 510 were seen corresponding to trimer and tetramer, respectively. Tetrameric structures have been observed with other HLH proteins such as USF (8), MyoD (14, 15), myogenin (16), Id (17) and TFEB (18), and this tetramerization event may be an addi- tional mechanism for transcriptional regulation.

yield a 60 bp fragment containing only the rcE2 E box (CAGGTG); PvuI and EcoRI to yield a 121 bp frag- ment containing the CAGCTG site; and NurI and PvuII to yield a 72 bp fragment containing the CAACTG site. These fragments were radiolabellcd with either the Kle- now fragment of DNA Polymerase or T4 Polymerase and dCTP, dTTP, dGTP, and ~ ' - I x - ~ ~ P - ~ A T P (Amer- sham). DNA binding was assayed under half-maximal binding conditions. Protein and DNA were incubated at RT in Buffer A containing 57,, glycerol, and 1 pg of poly-dIdC for 30 min followed by analysis on non- denaturing PAGE. The concentration of DNA was 10 - M in all experiments. The dissociation constant was determined by densitometric analysis of the result- ing autoradiograph.

RESULTS AND DISCUSSION

Protein synthesis The bHLH portion of IEB E47,1, was synthesized and purified in high yield (10% overall) to provide a homo- geneous material as judged by reversed-phase HPLC (Fig. 1A) and SDS PAGE (Fig. IB), and with the cor- rect mass as judged by electrospray mass spectrometry (Fig. 2). As one of the first reported syntheses of a bHLH protein (14), it bodes well for the chemical syn- thesis of proteins in this class, including mutant and fusion proteins, using an Fmoc-based methodology.

Dimerizatioiz studies Direct evidence for dimerization was provided from both size-exclusion chromatography and sedimentation equilibrium. Size exclusion Chromatography with 1 gave an apparent molecular weight of' 15 300 for 1 by inter- polation of the protein standards curve, which corre- sponds to an aggregation state of 2.2. Sedimentation equilibrium was also performed to determine the ag- gregation state of 1. A plot of aliquot concentration vs. r2, the square of the distance of the aliquot from the center of the rotor, yielded a straight line whose slope was proportional to a molecular weight for 1 of 11 740, which corresponds to an aggregation state of 1.7. These

Circular dichroism studies The CD spectrum for protein 1 exhibited a character- istic helical spectrum with a double minima at 208 and 222 nm (Fig. 4) (19). The mean residue ellipticity, [ 01, was calculated per residue (n = 59), and the value at 222 nm was used to calculate a maximum helical con- tent of 44% for 1. The concentration dependence of the CD spectra was also investigated to determine whether the protein was self-associating in solution resulting in helix stabilization. In the range 1 - 1 5 0 ~ ~ the helical content exhibited a concentration dependent behavior for 1. By analogy with the co-crystal structure of IEB E47 (9), a helical content of 66% was predicted, which is significantly higher than the helical content of 44% obtained for 1 by CD. It may be possible that in solution the ends of the amphiphilic helices in 1 are unwinding due to limited intra- and intermolecular contacts.

A CD spectra in the presence of DNA containing the rcE2 sequence was obtained to investigate the effect of DNA binding on the conformation of 1 (Fig. 4). When this DNA was added to 1 at a concentration near maxi- mal helicity (50 PM), 1 gained an additional 8% helical

FIGURE 1 Analysis of the purity of protein 1 by: (A) analytical reversed-phase HPLC on Vydac C8 column (0.46 cm x 25 cm) using a linear gradient (30 min) of 25-80% CH,CN/H20 (0.1 ''(, TFA) with detection at 214 nm and (B) 17% SDS PAGE with Coornassie Blue stain: lane 1, pure protein; lane 2, molecular weight markers.

151

P. Bishop et a/.

B) 100- 1165.7 1 6988.E

0

FIGURE 2 Electrospray mass spectrometry of purified 1: (A) spectrum showing the multiply charged ions and (B) the deconvoluted spectrum.

5000

FIGURE 3 SDS PAGE of crosslinked 1: lane 1,750 j t ~ 1 + BS’; lane 2, 100 j t ~

1 + BS’; lane 3, 1 alone; lane 4, molecular weight markers.

content as seen by the increase in the value of the mean residue ellipticity at 222 nm. In the control experiment, when DNA lacking the K E ~ sequence was added to 1 the helical content decreased by 9%. In the co-crystal structure of IEB E47 the bask region is bound in the major groove of the DNA in a helical conformation (9), whereas in the absence of DNA this region is believed to exist as a random coil (14). If the entire basic region of 1 adopted a helical conformation an increase of 22 % should have been observed. This discrepancy may be due to an equilibrium state in which some amount of protein exists in an unbound state, which is reflected in the lower helicity of 1. Alternatively, the basic region,

152

-20000 ’ I 200 260

Wavelength (nm)

FIGURE 4 Circular dichroism spectra of 1 at 128 pM (solid line); 1 plus either 1.2 eq. of xE2 containing DNA (dotted line) or 1.2 eq. of poly-dIdC (dashed line). The spectra were obtained by subtracting the spectrum of the DNA from the spectrum of 1 plus DNA.

which is at the extreme N-terminus of 1, may lack ad- ditional protein contacts needed for increased stability of the helix.

DNA binding experiments Gel mobility shift assays were used to determine the affinity of 1 for a DNA probe containing three differ- ent E box sequences: CAGGTG, CAGCTG and

The bHLH region of IEB E47

FIGURE 5 DNA binding of 1 to E box sequences. 211 bp fragments containing three E boxes: CAGGTG (kE2 site), CAGCTG, CAACTG. Lane 1, DNA alone; lane 2, DNA + 0.5 p~ 1; lane 3, DNA + 0.75 pM 1; lane 4, DNA + 1.0 PM 1; lane 5, DNA + 2.5 p~ 1; lane 6, DNA + 5.0 p~ I ; lane 7, DNA + 7.5 pM 1; lane 8, DNA + 10 pM 1; lane 9, DNA + 25 pM 1; lane 10, DNA + 50 pM 1; lane 11, DNA + 75 p~ 1; lane 12, DNA + 100 /AM 1.

CAACTG (Fig. 5) . As the protein concentration was increased to 1 p~ a shifted band was obtained indicat- ing that 1 had affinity for a site within the DNA probe, and as the concentration was further increased, a second shifted band at 5 y ~ was obtained. At even higher concentrations (25 p ~ ) 1 displayed affinity for a third site in the DNA probe resulting in a third shifted band. At 75 y~ non-specific aggregation between the protein and DNA occurred, resulting in a high molecu- lar weight complex observed as a band at the top of the gel. To determine which of the three E box sequences was bound, three additional fragments of DNA each containing one of the above E boxes were used in gel mobility shift experiments (Fig. 6). As expected, 1 had the highest affinity for the xE2 site as demonstrated by the presence of the shifted bands in lanes 2 and 3. The shifted band in lane 6 indicates that 1 has higher affinity for the CAGCTG sequence than for the CAACTG sequence.

From the gel mobility shift experiments the first quan- titation of a binding constant of 1 for the rcE2 site was

FIGURE 6 Affinity of 1 for three separate E box aequences. Lanes 1-3, 51 bp fragment containing CAGGTG (kE2 \ite); lanes 4-6, 121 bp frag- ment containing CAGCTG; lanes 7-9. 72 bp fragment containing CAACTG. Lanes 1, 4 and 7 contain DNA alone; lanes 2, 5 and 8 contain 2.5 pM 1; lanes 3, 6 and 9 conlain 7.5 p M 1.

calculated to be 10- lo M ~ , as compared to 10- M~ obtained with IEB E47 (20). This decrease in binding of the smaller protein may be due to the absence of additional protein-DNA contacts found in regions out- side of the bHLH region which may exist in the full- length protein. The dimeric complex of 1 may also not be as stable as the dimer of IEB E47, perhaps due to loss of intermolecular protein-protein interactions which may also lower the DNA binding constant. Full- length IEB E47 has been shown to bind to other E box sequences (21). The 59 amino acid protein 1 also bound to one other E box, CAGCTG. However, as with IEB E47, lower-affinity binding was observed as compared to the xE2 site.

CONCLUSION

These results indicate that studying individual bHLH domains can provide valuable structural and DNA binding information for the HLH class of proteins, including those whose structure and function are unknown. Because of the reduced complexity and ease of synthesis of individual domains, fusion proteins, mutant protein and deletion proteins can be generated which may provide useful insight into the behavior of this very important class of DNA binding proteins.

ACKNOWLEDGEMENTS

We gratefully acknowledge the financial support of the NIH ( 5 R29 G M 46936-02) and the Monsanto Chemical Co., and mass spectral analysis from the Nebraska Center for Mass Spectrometry.

REFERENCES

1. Murre, C., McCaw, P.S. & Baltimore, D. (1989) Cell 56, 777-

2. Lenardo, M., Pierce, J.W. & Baltimore, D. (1987) Science 236,

3. Schlissel, M., Voronova, A. & Baltimore, D. (1991) Genes and

783

1573-1577

Dev. 5, 1367-1376

153

P. Bishop et al.

4. Lassar, A.B., Davis, R.L., Wright, W.E., Kadesch, T., Murre, C., Voronova, A., Baltimore, D. & Weintraub, H. (1991) Cell 66,

5. Kamps, M.P., Murre, C., Sun, X.H. & Baltimore, D. (1990) Cell

6. Voronova, A. & Baltimore, D. (1990) Proc. Narl. Acad. Sci. USA

7. Ferre-D’Amare, A.R., Prendergast, G.C., Ziff, E.B. & Burley,

8. Ferre-DAmare, A.R., Pognonec, P., Roeder, R.G. & Burley,

9. Ellenherger, T., Fass, D., Amaud, M. & Harrison, S.C. (1994)

305-3 15

60,547-555

87, 4722-4726

S.K. (1993) Narure (London) 363, 38-45

S.K. (1994) EMBO J. 13, 180-189

Genes & Dev. 8, 970-980 10. Wang, S.S. (1973)J. Am. Chem. SOC. 95, 1328-1333. 11. Atherton, E. & Sheppard, R.C. (1987) in Gross, E. & Meien-

hofer, J., eds, The Peprides, Academic Press, New York, Vol. 9,

12. Pollet, R. J. (1985) Methods Enzymol. 117, 3-27, Pollet, R.J., Haase, B.A. & Standaert, M.L. (1979)J. Biol. Chern. 254,30-33

13. Lyu, P.C., Sherman, J.C., Chen, A. & Kallenbach, N.R. (1991) Proc. Natl. Acad. Sci. USA 88, 5317-5320

14. Anthony-Cahill, S.J., Benfield, P.A., Fairman, R., Wasserman, Z.R., Brenner, S.L., Stafford 111, W.F., Altenbach, C., Hubbell, W.L. & DeGrado, W.F. (1992) Science 255, 979-983

pp. 1-38

15. Starovasnik, M.A., Blackwell, T.K., Laue, T.M., Weintraub, H.

16. Farmer, K., Catala, F. & Wright, W.E. (1992)JBiol. Chem. 267,

17. Fairman, R., Beran-Steed, R.K., Anthony-Cahill, S.J., Lear, J.D., Stafford 111, W.F., DeGrado, W.F., Benfield, P.A. & Bren- ner, S.L. (1993) Proc. Narl. Acad. Sci. USA 90, 10429-10433

18. Fisher, D.E., Carr, C.S., Parent, L.A. & Sharp, P.A. (1991) Genes Dev. 5, 2342-2352

19. Greenfield, N. & Fasman, G.D. (1969) Biochemistry 8, 4108- 41 16

20. Sun, X.-H. & Baltimore, D. (1991) Cell 64, 459-470 21. Mum, C., McCaw, P.S., Vaessin, H., Caudy, M., Jan, L.Y., Jan,

Y.N.,Cabrera,C.V., Buskin, J.N.,Hauschka, S.D.,Lassar,A.B., Weintraub. H. & Baltimore, D. (1989) Cell 58, 537-544

& Klevit, R.C. (1990) Biochemistry 31, 9891-9903

5631-5636

Address:

Jean Chmielewski Department of Chemistry Purdue University West Lafayette, IN 47907 USA