Novel parathyroid hormone (PTH) antagonists that bind to the juxtamembrane portion of the PTH/PTH-related protein receptor

Current antagonists for the parathyroid hormone (PTH)/PTH-related protein (PTHrP) receptor (PTHR) are N-terminally truncated or N-terminally modified analogs of PTH(1-34) or PTHrP(1-34) and are thought to bind predominantly to the N-terminal extracellular (N) domain of the receptor. We hypothesized that ligands that bind only to PTHR region comprised of the extracellular loops and seven transmembrane helices (the juxtamembrane or J domain) could also antagonize the PTHR. To test this, we started with the J domain-selective agonists [Gln(10),Ala(12),Har(11),Trp(14),Arg(19) (M)]PTH(1-21), [M]PTH(1-15), and [M]PTH(1-14), and introduced substitutions at positions 1-3 that were predicted to dissociate PTHR binding and cAMP signaling activities. Strong dissociation was observed with the tri-residue sequence diethylglycine (Deg)(1)-para-benzoyl-l-phenylalanine (Bpa)(2)-Deg(3). In HKRK-B7 cells, which express the cloned human PTHR, [Deg(1,3),Bpa(2),M]PTH(1-21), [Deg(1,3),Bpa(2),M]PTH(1-15), and [Deg(1,3),Bpa(2),M]PTH(1-14) fully inhibited (IC(50)s = 100-700 nm) the binding of (125)I-[alpha-aminoisobutyric acid(1,3),M]PTH(1-15) and were severely defective for stimulating cAMP accumulation. In ROS 17/2.8 cells, which express the native rat PTHR, [Deg(1,3),Bpa(2),M]PTH(1-21) and [Deg(1,3),Bpa(2),M]PTH(1-15) antagonized the cAMP-agonist action of PTH(1-34), as did PTHrP(5-36) (IC(50)s = 0.7 microm, 2.6 microm, and 36 nm, respectively). In COS-7 cells expressing PTHR-delNt, which lacks the N domain of the receptor, [Deg(1,3),Bpa(2), M]PTH(1-21) and [Deg(1,3),Bpa(2),M]PTH(1-15) inhibited the agonist actions of [alpha-aminoisobutyric acid(1,3)]PTH(1-34) and [M]PTH(1-14) (IC(50)s approximately 1 microm), whereas PTHrP(5-36) failed to inhibit. [Deg(1,3),Bpa(2),M]PTH(1-14) inhibited the constitutive cAMP-signaling activity of PTHR-tether-PTH(1-9), in which the PTH(1-9) sequence is covalently linked to the PTHR J domain, as well as that of PTHR(cam)H223R. Thus, the J-domain-selective N-terminal PTH fragment analogs can function as antagonists as well as inverse agonists for the PTHR. The new ligands described should be useful for further studies of the ligand binding and activation mechanisms that operate in the critical PTHR J domain.     

[13c]-Constant-Time [15n,1h]-Trosy-Hnca for Sequential Assignments of Large Proteins

The greatly improved sensitivity resulting from the use of TROSY during 15N evolution and amide proton acquisition enables the recording of HNCA spectra of large proteins with constant-time 13C evolution. In [13C]-ct-[15N,1H]-TROSY-HNCA experiments with a 2H/13C/15N-labeled 110 kDa protein, 7,8-dihydroneopterin aldolase from Staphylococcus aureus, nearly all correlation peaks seen in the [15N,1H]-TROSY-HNCA spectrum were also detected. The improved resolution in the 13C dimension then enabled a significant number of sequential assignments that could not be obtained with [15N,1H]-TROSY-HNCA without [13C]-constant-time period.     

Single Transition-to-single Transition Polarization Transfer (ST2-PT) in [15N,1H]-TROSY

This paper describes the use of single transition-to-single transition polarization transfer (ST2-PT) in transverse relaxation-optimized spectroscopy (TROSY), where it affords a sensitivity enhancement for kinetically stable amide 15N-1H groups in proteins. Additional, conventional improvements of [15N,1H]-TROSY include that signal loss for kinetically labile 15N-1H groups due to saturation transfer from the solvent water is suppressed with the water flip back technique and that the number of phase steps is reduced to two, which is attractive for the use of [15N,1H]-TROSY as an element in more complex NMR schemes. Finally, we show that the impact of the inclusion of the 15N steady-state magnetization (Pervushin et al., 1998) on the signal-to-noise ratio achieved with [15N,1H]-TROSY exceeds by up to two-fold the gain expected from the gyromagnetic ratios of 1H and 15N.     

Genetic study of interactions between the cytoskeletal assembly protein sla1 and prion-forming domain of the release factor Sup35 (eRF3) in Saccharomyces cerevisiae.

Striking similarities between cytoskeletal assembly and the "nucleated polymerization" model of prion propagation suggest that similar or overlapping sets of proteins may assist in both processes. We show that the C-terminal domain of the yeast cytoskeletal assembly protein Sla1 (Sla1C) specifically interacts with the N-terminal prion-forming domain (Sup35N) of the yeast release factor Sup35 (eRF3) in the two-hybrid system. Sla1C and several other Sup35N-interacting proteins also exhibit two-hybrid interactions with the poly-Gln-expanded N-proximal fragment of human huntingtin, which promotes Huntington disease-associated aggregation. The Sup35N-Sla1C interaction is inhibited by Sup35N alterations that make Sup35 unable to propagate the [PSI(+)] state and by the absence of the chaperone protein Hsp104, which is essential for [PSI] propagation. In a Sla1(-) background, [PSI] curing by dimethylsulfoxide or excess Hsp104 is increased, while translational readthrough and de novo [PSI] formation induced by excess Sup35 or Sup35N are decreased. These data show that, in agreement with the proposed function of Sla1 during cytoskeletal formation, Sla1 assists in [PSI] formation and propagation, but is not required for these processes. Sla1(-) strains are sensitive to some translational inhibitors, and some sup35 mutants, obtained in a Sla1(-) background, are sensitive to Sla1, suggesting that the interaction between Sla1 and Sup35 proteins may play a role in the normal function of the translational apparatus. We hypothesize that Sup35N is involved in regulatory interactions with intracellular structural networks, and [PSI] prion may be formed as a by-product of this process.     

Transition-state analysis of a Vmax mutant of AMP nucleosidase by the application of heavy-atom kinetic isotope effects

The transition state of the Vmax mutant of AMP nucleosidase from Azotobacter vinelandii [Leung, H. B., & Schramm, V. L. (1981) J. Biol. Chem. 256, 12823-12829] has been characterized by heavy-atom kinetic isotope effects in the presence and absence of MgATP, the allosteric activator. The enzyme catalyzes hydrolysis of the N-glycosidic bond of AMP at approximately 2% of the rate of the normal enzyme with only minor changes in the Km for substrate, the activation constant for MgATP, and the Ki for formycin 5'-phosphate, a tight-binding competitive inhibitor. Isotope effects were measured as a function of the allosteric activator concentration that increases the turnover number of the enzyme from 0.006 s-1 to 1.2 s-1. The kinetic isotope effects were measured with the substrates [1'-3H]AMP, [2'-2H]AMP, [2'-2H]AMP, [9-15N]AMP, and [1',9-14C, 15N]AMP. All substrates gave significant kinetic isotope effects in a pattern that establishes that the reaction expresses intrinsic kinetic isotope effects in the presence or absence of MgATP. The kinetic isotope effect with [9-15N]AMP decreased from 1.034 +/- 0.002 to 1.021 +/- 0.002 in response to MgATP. The [1'-3H]AMP isotope effect increased from 1.086 +/- 0.003 to 1.094 +/- 0.002, while the kinetic isotope effect for [1',9-14C, 15N]AMP decreased from 1.085 +/- 0.003 to 1.070 +/- 0.004 in response to allosteric activation with MgATP. Kinetic isotope effects with [1'-14C]AMP and [2'-2H]AMP were 1.041 +/- 0.006 and 1.089 +/- 0.002 and were not changed by addition of MgATP.(ABSTRACT TRUNCATED AT 250 WORDS)     

Opioid receptor selectivity reversal in deltorphin tetrapeptide analogues.

Deltorphin N-terminal tetrapeptides [DEL A: H-Tyr-D-Met-Phe-His-R, where R = -NH2, -NH-NH2, -OCH3, -OH, -NH-NH-CO-R' (R' = -CH3 or adamantane); DEL C: H-Tyr-D-Ala-Asp-R (R = -OH, -NHCH3)], were used in a receptor binding assay with [3H]DADLE and [3H]DPDPE for delta sites, and [3H]DAGO for mu sites; tetrapeptide Ki delta values were similar with either [3H]-delta ligand. DEL A tetrapeptides C-terminally substituted with -NH2, -NH-NH2, -OCH3, and -OH had 10 to greater than 1,000-fold decreased Ki delta values, while Ki mu increased 5 to 100-fold to yield mu selectivity. C-Terminal substitution with -NH-NH2 and -OCH3 conferred highest mu selectivities; adamantyl and acetyl hydrazide derivatives were non-selective. DEL-(1-4)-OH peptides had decreased delta and mu affinities: DEL A-[Asp4]-(1-4)-OH and DEL C-(1-4)-OH had low affinities (greater than 1 microM), however, the Ki delta of the former was 5-fold greater than the latter, and the Ki mu was less by 15-fold. The data suggest that the "message" domain of DEL exhibits receptor selectivity different from that of the heptapeptide.     

Bis(sulfo-N-succinimidyl) [15N,2H16]doxyl-2-spiro-4'-pimelate, a stable isotope-substituted, membrane-impermeant bifunctional spin label for studies of the dynamics of membrane proteins: application to the anion-exchange channel in intact human erythrocytes

We have synthesized and characterized an isotopically substituted homologue of the membrane-impermeant bifunctional spin label bis(sulfo-N-succinimidyl) doxyl-2-spiro-4'-pimelate (BSSDP) [Beth et al. (1986) Biochemistry 25, 3824-3832] in which the nitroxide N is substituted with 15N and all of the protons in the doxylpimelate moiety are replaced by deuterons ([15N,2H16]BSSDP). Like its normal isotope homologue, [15N,2H16]BSSDP reacts with the anion-exchange channel in intact human erythrocytes at a site that spans the single extracytoplasmic chymotryptic cleavage site and that overlaps the stilbenedisulfonate site. The narrower line widths in the EPR spectrum of [15N,2H16]BSDP-labeled anion channels allow calculation of a minimum separation of 16 A between spin labels bound at the functionally important stilbenedisulfonate sites on adjacent subunits of an anion channel dimer. The 15N and 2H isotopic substitutions also provide substantial improvement in signal to noise of motionally sensitive regions of the ST-EPR spectrum of [15N,2H16]BSSDP-labeled anion channels in intact erythrocytes. [15N,2H16]BSSDP-labeled anion channels in intact erythrocytes were cross-linked to covalent dimers in the extracytoplasmic domain with the membrane-impermeant cross-linking reagent bis(sulfo-N-succinimidyl) suberate [Staros (1982) Biochemistry 21, 3950-3955], and the saturation-transfer EPR spectrum of these cells was compared with that of cells treated with [15N,2H16]BSSDP but not subsequently cross-linked. The spectra were essentially identical, supporting the hypothesis that anion channel subunits form stable dimers in the membranes of intact erythrocytes.     

Assignment of the 1H and 15N NMR spectra of Rhodobacter capsulatus ferrocytochrome c2

The peptide resonances of the 1H and 15N nuclear magnetic resonance spectra of ferrocytochrome c2 from Rhodobacter capsulatus are sequentially assigned by a combination of 2D 1H-1H and 1H-15N spectroscopy, the latter performed on 15N-enriched protein. Short-range nuclear Overhauser effect (NOE) data show alpha-helices from residues 3-17, 55-65, 69-88, and 103-115. Within the latter two alpha-helices, there are three single 3(10) turns, 70-72, 76-78, and 107-109. In addition alpha H-NHi+1 and alpha H-NHi+2 NOEs indicate that the N-terminal helix (3-17) is distorted. Compared to horse or tuna cytochrome c and cytochrome c2 of Rhodospirillium rubrum, there is a 6-residue insertion at residues 23-29 in R. capsulatus cytochrome c2. The NOE data show that this insertion forms a loop, probably an omega loop. 1H-15N heteronuclear multiple quantum correlation experiments are used to follow NH exchange over a period of 40 h. As the 2D spectra are acquired in short time periods (30 min), rates for intermediate exchanging protons can be measured. Comparison of the NH exchange data for the N-terminal helix of cytochrome c2 of R. capsulatus with the highly homologous horse heart cytochrome c [Wand, A. J., Roder, H., & Englander, S. W. (1986) Biochemistry 25, 1107-1114] shows that this helix is less stable in cytochrome c2.     

Structural similarities between the N-terminal domain of Clostridium pasteurianum hydrogenase and plant-type ferredoxins

An N-terminal domain of Clostridium pasteurianum hydrogenase I, encompassing 76 residues out of the 574 composing the full-size enzyme, had previously been overproduced in Escherichia coli and shown to form a stable fold around a [2Fe-2S] cluster. This domain displays only marginal sequence similarity with [2Fe-2S] proteins of known structure, and therefore, two-dimensional 1H NMR has been implemented to elucidate features of the polypeptide fold. Despite the perturbing presence of the paramagnetic [2Fe-2S] cluster, 57 spin systems were detected in the TOCSY spectra, 52 of which were sequentially assigned through NOE connectivities. Several secondary structure elements were identified. The N terminus of the protein consists of two antiparallel beta strands followed by an alpha helix contacting both strands. Two additional antiparallel beta strands, one of them at the C terminus of the sequence, form a four-stranded beta sheet together with the two N-terminal strands. The proton resonances that can be attributed to this beta2alphabeta2 structural motif undergo no paramagnetic perturbations, suggesting that it is distant from the [2Fe-2S] cluster. In plant- and mammalian-type ferredoxins, a very similar structural pattern is found in the part of the protein farthest from the [2Fe-2S] cluster. This indicates that the N-terminal domain of C. pasteurianum hydrogenase folds in a manner very similar to those of plant- and mammalian-type ferredoxins over a significant part (ca. 50%) of its structure. Even in the vicinity of the metal site, where 1H NMR data are blurred by paramagnetic interactions, the N-terminal domains of hydrogenase and mammalian- and plant-type ferredoxins most likely display significant structural similarity, as inferred from local sequence alignments and from previously reported circular dichroism and resonance Raman spectra. These data afford structural information on a kind of [2Fe-2S] cluster-containing domain that occurs in a number of redox enzymes and complexes. In addition, together with previously published sequence alignments, they highlight the widespread distribution of the plant-type ferredoxin fold in bioenergetic systems encompassing anaerobic metabolism, photosynthesis, and aerobic respiratory chains.     

The H29D Mutation Does Not Enhance Cytosolic Ca2+ Activation of the Cardiac Ryanodine Receptor

The N-terminal domain of the cardiac ryanodine receptor (RyR2) harbors a large number of naturally occurring mutations that are associated with stress-induced ventricular tachyarrhythmia and sudden death. Nearly all these disease-associated N-terminal mutations are located at domain interfaces or buried within domains. Mutations at these locations would alter domain-domain interactions or the stability/folding of domains. Recently, a novel RyR2 mutation H29D associated with ventricular arrhythmia at rest was found to enhance the activation of single RyR2 channels by diastolic levels of cytosolic Ca²⁺. Unlike other N-terminal disease-associated mutations, the H29D mutation is located on the surface of the N-terminal domain. It is unclear how this surface-exposed H29D mutation that does not appear to interact with other parts of the RyR2 structure could alter the intrinsic properties of the channel. Here we carried out detailed functional characterization of the RyR2-H29D mutant at the molecular and cellular levels. We found that the H29D mutation has no effect on the basal level or the Ca²⁺ dependent activation of [³H]ryanodine binding to RyR2, the cytosolic Ca²⁺ activation of single RyR2 channels, or the cytosolic Ca²⁺- or caffeine-induced Ca²⁺ release in HEK293 cells. In addition, the H29D mutation does not alter the propensity for spontaneous Ca²⁺ release or the thresholds for Ca²⁺ release activation or termination. Furthermore, the H29D mutation does not have significant impact on the thermal stability of the N-terminal region (residues 1–547) of RyR2. Collectively, our data show that the H29D mutation exerts little or no effect on the function of RyR2 or on the folding stability of the N-terminal region. Thus, our results provide no evidence that the H29D mutation enhances the cytosolic Ca²⁺ activation of RyR2.     

Vasoactive intestinal polypeptide VPAC1 and VPAC2 receptor chimeras identify domains responsible for the specificity of ligand binding and activation.

In order to identify the receptor domains responsible for the VPAC1 selectivity of the VIP1 agonist, [Lys15, Arg16, Leu27] VIP (1-7)/GRF (8-27) and VIP1 antagonist, Ac His1 [D-Phe2, Lys15, Arg16, Leu27] VIP (3-7)/GRF (8-27), we evaluated their binding and functional properties on chimeric VPAC1/VPAC2 receptors. Our results suggest that the N-terminal extracellular domain is responsible for the selectivity of the VIP1 antagonist. Selective recognition of the VIP1 agonist was supported by a larger receptor area: in addition to the N-terminal domain, the first extracellular loop, as well as additional determinants in the distal part of the VPAC1 receptor were involved. Furthermore, these additional domains were critical for an efficient receptor activation, as replacement of EC1 in VPAC1 by its counter part in the VPAC2 receptor markedly reduced the maximal response.     

1H, 15N, and 13C assignments of the N-terminal activation domain of Dictyostelium discoideum Formin C

Dictyostelium discoideum Formin C (ForC) plays an important role in the fruiting body formation during the multicellular stages of the slime mold. Formins are multidomain proteins that are known to regulate the actin cytoskeleton. Here, we report the assignments of the ¹H, ¹⁵N, and ¹³C nuclei of the N-terminal activation domain (residues 1–100) of ForC. Chemical shifts have been deposited at the BioMagResBank under the BMRB accession number 17,029. The N-terminal region of the 131 kDa ForC protein is supposed to form a GTPase-binding domain required for activation of the formin.     

Backbone dynamics of the N-terminal domain in E. coli DnaJ determined by 15N- and 13CO-relaxation measurements.

The backbone dynamics of the N-terminal domain of the chaperone protein Escherichia coli DnaJ have been investigated using steady-state 1H-15N NOEs, 15N T1, T2, and T1 rho relaxation times, steady-state 13C alpha-13CO NOEs, and 13CO T1 relaxation times. Two recombinant constructs of the N-terminal domain of DnaJ have been studied. One, DnaJ(1-78), contains the most conserved "J-domain" of DnaJ, and the other, DnaJ(1-104), includes a glycine/phenylalanine rich region ("G/F" region) in addition to the "J-domain". DnaJ(1-78) is not capable of stimulating ATP hydrolysis by DnaK, despite the fact that all currently identified sites responsible for DnaJ-DnaK interaction are located in this region. DnaJ(1-104), on the other hand, retains nearly the full ATPase stimulatory activity of full length DnaJ. Recently, a structural analysis of these two molecules was presented in an effort to elucidate the origin of their functional differences [Huang, K., Flanagan, J. M., and Prestegard, J. H. (1999) Protein Science 8, 203-214]. Herein, an analysis of dynamic properties is presented in a similar effort. A generalized model-free approach with a full treatment of the anisotropic overall rotation of the proteins is used in the analysis of measured relaxation parameters. Our results show that internal motions on pico- to nanosecond time scales in the backbone of DnaJ(1-78) are reduced on the inclusion of the "G/F" region, while conformational exchange on micro- to millisecond time scales increases. We speculate that the enhanced flexibility of residues on the slow time scale upon the inclusion of the "G/F" region could be relevant to the ATPase stimulatory activity of DnaJ if an "induced-fit" mechanism applies to DnaJ-DnaK interactions.     

NMDA Receptor-Mediated Regulation of AMPA Receptor Properties in Organotypic Hippocampal Slice Cultures

Abstract: Activation of the calcium-dependent protease calpain has been proposed to be a necessary step in the formation of long-term potentiation (LTP) in the hippocampus, and stimulation of N-methyl-d -aspartate (NMDA) receptors leads to an increase in intracellular calcium concentration, calpain activation, proteolysis of cytoskeletal elements, and modification of α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor properties. In the present study, we evaluated the effects of NMDA treatment of cultured hippocampal slices on the properties of AMPA receptors. Cultured hippocampal slices were treated with NMDA (100 µM) for 15 min and [³H]AMPA binding to membrane fractions was measured. NMDA-treated slices exhibited an increase in both “high-affinity” and “low-affinity” [³H]-AMPA binding, with smaller changes in 6-cyano-7-nitro[³H]quinoxaline-2,3-dione binding. The increase in [³H]AMPA binding was significantly reduced by preincubation of cultures with calpain inhibitor I or calpeptin (100 µM). Furthermore, NMDA exposure decreased the number of GluR1 subunits of AMPA receptors detected by an antibody against the C-terminal domain of the subunit in western blots and resulted in the formation of a lower molecular weight species detected by an antibody against the N-terminal domain. Both effects were completely prevented by calpain inhibitors. These results indicate that NMDA receptor activation produces calpain activation and complex modifications of AMPA receptor properties, which could be involved in NMDA receptor-mediated changes in synaptic efficacy.     

Evidence for a signal sequence at the N terminus of the common precursor to adrenocorticothrophin and beta-lipotropin in mouse pituitary cells

The precursor to corticotropin and beta-endorphin was synthesized in a reticulocyte cell-free system under the direction of mRNA from mouse AtT-20 pituitary tumor cells in the presence of [3H]proline, [3H]phenylalanine, [3H]leucine, [3H]valine, [3H]isoleucine or [35S]methionine. Automatic Edman degradation of the radioactive cell-free product showed the following N-terminal sequence: Pro-1, Met-2, Leu-11, Leu-12, Leu-13, Leu-15, Leu-16, Leu-17, Ile-21 and Val-23. The corticotropin-endorphin precursor was also labeled in AtT-20 cells with [3H]valine, [3H]leucine, [3H]tryptophan, [3H]serine, [35S]methionine or [35S]cysteine. Automatic Edman degradation of the radioactive intact cell form gave the following N-terminal sequence: Trp-1, Cys-2, Leu-3, Ser-5, Ser-6, Val-7, Cys-8, Leu-11, Leu-17, Leu-18 and tentatively Met-27. The sequence of the intact cell form from AtT-20 cells matches the sequence of the cell-free form of bovine pituitary precursor beginning at Trp-27, as determined by recombinant DNA technology [Nakanishi, S., Inoue, A., Kita, T., Nakamura, M., Chang, A. C. Y., Cohen, S. N., and Numa, S. (1979) Nature (Lond.) 278, 423-427]. The sequence of the mouse pituitary mRNA-directed cell-free translation product also matches the bovine precursor beginning at Pro-2. The results suggest that both the mouse and bovine precursors possess a signal sequence of 26 amino acids which is cleaved in intact cells. CNBr cleavage of [35S]cysteine-labelled intact cell precursor gave rise to an N-terminal fragment of a size compatible with the presence of a methionyl residue at or near position 27.     

ADP-ribosylation factor 1 of Arabidopsis plays a critical role in intracellular trafficking and maintenance of endoplasmic reticulum morphology in Arabidopsis

ADP-ribosylation factors (Arf), a family of small GTP-binding proteins, play important roles in intracellular trafficking in animal and yeast cells. Here, we investigated the roles of two Arf homologs, Arf1 and Arf3 of Arabidopsis, in intracellular trafficking in plant cells. We generated dominant negative mutant forms of Arf 1 and Arf3 and examined their effect on trafficking of reporter proteins in protoplasts. Arf1[T31N] inhibited trafficking of H(+)-ATPase:green fluorescent protein (GFP) and sialyltransferase (ST):GFP to the plasma membrane and the Golgi apparatus. In addition, Arf1[T31N] caused relocalization of the Golgi reporter protein ST:GFP to the endoplasmic reticulum (ER). In protoplasts expressing Arf1[T31N], ST:red fluorescent protein remained in the ER, whereas H(+)-ATPase:GFP was mistargeted to another organelle. Also, expression of Arf1[T31N] in protoplasts resulted in profound changes in the morphology of the ER. The treatment of protoplasts with brefeldin A had exactly the same effect as Arf1[T31N] on various intracellular trafficking pathways. In contrast, Arf3[T31N] did not affect trafficking of any of these reporter proteins. Inhibition experiments using mutants with various domains swapped between Arf1 and Arf3 revealed that the N-terminal domain is interchangeable for trafficking inhibition. However, in addition to the T31N mutation, motifs in domains II, III, and IV of Arf1 were necessary for inhibition of trafficking of H(+)-ATPase:GFP. Together, these results strongly suggest that Arf1 plays a role in the intracellular trafficking of cargo proteins in Arabidopsis, and that Arf1 functions through a brefeldin A-sensitive factor.     

Truncation of motifs III and IV in human lens betaA3-crystallin destabilizes the structure

The purpose of our study was to determine the effects of specific truncations on the structural properties of human betaA3-crystallin. The following eight deletion mutants of betaA3-crystallin were generated: (i) N-terminal extension (NTE) 21 amino acids (betaA3[21] mutant), (ii) NTE 22 amino acids (betaA3[22] mutant), (iii) NTE (betaA3[N] mutant), (iv) NTE plus motif I (betaA3[N+I] mutant), (v) NTE plus motifs I and II (betaA3[N+I+II] mutant), (vi) NTE plus motifs I and II and connecting peptide (betaA3[N+I+II+CP] mutant), (vii) motifs III and IV (betaA3[III+IV] mutant), and (viii) motif IV (betaA3 [IV] mutant). The DNA sequencing and MALDI-TOF mass spectrometric methods confirmed desired specific deletions, and the purified mutant proteins exhibited a single band during SDS-PAGE analysis. When ANS bound, all the mutant proteins exhibited fluorescence quenching and a red shift, suggesting that the truncations caused changes in the exposed hydrophobic patches. The CD spectra showed that deletion of either NTE or the N-terminal domain (motifs I and II) had a relatively weaker effect on the structural stability than deletion of the C-terminal domain (motifs III and IV). Intrinsic Trp fluorescence spectral studies suggested changes in the microenvironment of the mutant proteins following truncations. HPLC multiangle light scattering analyses showed that truncation led to higher-order aggregation compared to that in the wild-type protein. Equilibrium unfolding and refolding of WT betaA3 with urea were best fit to a three-state model with transition midpoints at 2.2 and 3.1 M urea. However, the two transition midpoints of betaA3[21] and betaA3[22] and betaA3[N] mutants were similar to those of the wild type, suggesting that these truncations had a minimal effect on structural stabilization. Further, the mutant proteins containing the N-terminal domain (i.e., betaA3[III+IV] and betaA3[IV] mutants) exhibited higher transition midpoints compared to the transition midpoints of the mutant protein with the C-terminal domain (i.e., betaA3[N+I+II+CP] mutant). The results suggested that the N-terminal domain is relatively more stable than the C-terminal domain in betaA3-crystallin.