Protection of actin against proteolysis by complex formation with deoxyribonuclease I

G-actin bound to deoxyribonuclease I (DNase I) is resistant to digestion by trypsin and chymotrypsin. In the absence of DNase I, G-actin is cleaved by these proteases to yield a 33 500 molecular weight core protein which is not degraded further. The major sites of proteolytic action in the amino acid sequence of actin have been identified as being adjacent to residues arginine-62 and lysine-68 for trypsin and leucine-57 for chymotrypsin. These residues are rendered inaccessible to proteases in the buffer by complex formation with DNase I. Digestion of G-actin with pronase from Streptomyces griseus yields fragmentation patterns that are similar to those observed with trypsin and chymotrypsin. This is likely to be because the specificities of the major constituents of pronase resemble those of trypsin and chymotrypsin. Again, complex formation with DNase I protects the otherwise vulnerable bonds in actin against proteolysis. Incubation with subtilisin Carlsberg leads to complete digestion of G-actin. No subtilisin-resistant core protein accumulates during the incubation. Protection of G-actin when complexed to DNase I is less than complete in this case but still is significant. This is interpreted in terms of the broad specificity of subtilisin and the observed fragmentation pattern of free G-actin when treated with subtilisin.     

Myosin subfragment 1 and structural elements of G-actin: effects of S-1(A2) on sequences 39-52 and 61-69 in subdomain 2 of G-actin.

The effect of myosin on the structure of two sequences on G-actin, a loop between residues 39 and 52 and a segment between residues 61 and 69 from the NH2-terminus, was probed by limited proteolytic digestions of G-actin in the presence of the myosin subfragment 1 isozyme S-1(A2). Under the experimental conditions of this work, no polymerization of actin was induced by S-1(A2) [Chen & Reisler (1991) Biochemistry 30, 4546-4552]. S-1(A2) did not change the rates of subtilisin and chymotryptic digestion of G-actin at loop 39-52. In contrast to this, the second protease-sensitive region on G-actin, segment 61-69, was protected strongly by S-1(A2) from tryptic cleavage. The minor if any involvement of loop 39-52 in S-1 binding was confirmed by determining the binding constants of S-1(A2) for pyrene-labeled G-actin (1.2 x 10(6) M-1), subtilisin-cleaved pyrenyl G-actin (0.3 x 10(6) M-1), and DNase I-pyrenyl G-actin complexes (0.3 x 10(6) M-1). Consistent with this, the activity of DNase I, which binds to actin loop 39-52 [Kabsch et al. (1990) Nature 347, 37-44], was inhibited almost equally well by actin in the presence and absence of S-1(A2). These results confirm the observation that DNase I and S-1(A2) bind to distinct sites on actin [Bettache et al. (1990) Biochemistry 29, 9085-9091] and demonstrate myosin-induced changes in segment 61-69 of G-actin.     

Interactions of myosin subfragment 1 isozymes with G-actin

The polymerization of G-actin by myosin subfragment 1 (S-1) isozymes, S-1(A1) and S-1(A2), and their proteolytically cleaved forms was studied by light-scattering, fluorescence, and analytical ultracentrifugation techniques. As reported previously, S-1(A1) polymerized G-actin rapidly while S-1(A2) could hardly promote the assembly reaction (Chaussepied & Kasprzak, 1989a; Chen and Reisler, 1990). This difference between the isozymes of S-1 was traced to the very poor, if any, ability of G-actin-S-1(A2) complexes to nucleate the assembly of actin filaments. The formation of G-actin-S-1(A2) complexes was verified in sedimentation velocity experiments and by fluorescence measurements using pyrene-labeled actin. The G-actin-S-1(A2) complexes supported the growth of actin filaments and accelerated the polymerization of actin in solutions seeded with MgCl2-, KCl-, and S-1(A1)-generated nuclei. The growth rates of actin-S-1(A2) filaments were markedly slower than those for actin-S-1(A1) filaments. Proteolytic cleavage of S-1 isozymes at the 50/20-kDa junction of the heavy chain greatly decreased their binding to G-actin and thus inhibited the polymerization of actin by S-1(A1). These results are discussed in the context of G-actin-S-1 interactions.     

Functional, chymotryptically split actin and its interaction with myosin subfragment 1

We have prepared chymotryptically split actin that retains the characteristic properties of intact actin. Chymotryptic digestion of G-actin produces an intermediate 35-kilodalton (kDa) fragment and from this a final product of 33 kDa known as the C-terminal "core". These fragments remain attached to an N-terminal 10-kDa fragment. The 35-kDa-10-kDa complex is able to polymerize upon addition of KCl and MgCl2, like intact actin, whereas the 33-kDa-10-kDa complex is not. The 35-kDa-10-kDa complex is here termed "split actin". In the rigor state, split actin binds to myosin subfragment 1 (S-1) strongly, with the same stoichiometry as intact actin. In the rigor state, split actin forms a carbodiimide-induced cross-linked product with S-1; the cross-linking sites on the split actin and on S-1 were proved to be the N-terminal 10-kDa fragment of split actin and the 20-kDa domain of S-1. There was no cross-linking between the 50-kDa domain of S-1 and the 10 kDa of actin. Therefore, the structure of the split actin-S-1 complex differs somewhat from that of the complex with intact actin. The cross-linking of split actin to S-1 causes superactivation of S-1 ATPase to approximately the same extent as does cross-linking of intact actin, whereas non-cross-linked split actin activates S-1 ATPase to a lesser extent. The N-terminus of the 35-kDa fragment was found to be residue 45 (Val-45) by amino acid sequence analysis; so there is no residue missing in split actin.     

Cross-linking of actin to myosin subfragment 1: course of reaction and stoichiometry of products

The cross-linking of actin to myosin subfragment 1 (S-1) with 1-ethyl-3-[3-(dimethyl-amino)propyl]carbodiimide was reexamined by using two cross-linking procedures [Mornet, D., Bertrand, R., Pantel, P., Audemard, E., & Kassab, R. (1981) Nature (London) 292, 301-306; Sutoh, K. (1983) Biochemistry 22, 1579-1585] and two independent methods for quantitating the reaction products. In the first approach, the cross-linked acto-S-1 complexes were cleaved with elastase at the 25K/50K and 50K/22K junctions in S-1. This enabled direct measurements of the cross-linked and un-cross-linked fractions of the 50K and 22K fragments of S-1. We found that in all cases actin was preferentially cross-linked to the 22K fragment and that the overall stoichiometry of the main cross-linked products was that of a 1:1 complex of actin and S-1. In the second approach, actin was cross-linked to tryptically cleaved S-1, and the course of these reactions was monitored by measuring the decay of the free 50K and 20K fragments and the formation of cross-linked products. After selecting the optimal cross-linking procedure and conditions, we determined that the rate of actin cross-linking to the 20K fragment of S-1 was 3-fold faster than the reaction with the 50K peptide. The overall rate of cross-linking actin to S-1 corresponded to the sum of the individual reactions of the 50K and 20K fragments, indicating their mutually exclusive cross-linking to actin. Thus, the reactions with tryptically cleaved S-1 were consistent with the 1:1 stoichiometry of actin and S-1 in the main cross-linked products and verified the preferential cross-linking of actin to the 20K fragment of S-1. These results are discussed in the context of the binding of actin to S-1.     

Cofilin induced conformational changes in F-actin expose subdomain 2 to proteolysis

Cofilin/ADF affects strongly the structure of actin filaments and especially the intermolecular contacts of the DNase I binding loop (D-loop) in subdomain 2. In G-actin, the D-loop is cleaved by subtilisin between Met47 and Gly48, while in F-actin this cleavage is inhibited. Here, we report that yeast cofilin, which is resistant to both subtilisin and trypsin, accelerates greatly the rate of subtilisin cleavage of this loop in F-actin at pH 6.8 and at pH 8.0. Similarly, cofilin accelerates strongly the tryptic cleavage in F-actin of loop 60-69 in subdomain 2, at Arg62 and Lys68. The acceleration of the loops' proteolysis cannot be attributed to an increased treadmilling of F-actin for the following reasons: (i) the rate of subtilisin cleavage is independent of pH between pH 6.8 and 8.0, unlike F-actin depolymerization, which is pH-dependent; (ii) at high concentrations of protease the cleavage rate of F-actin in the presence of cofilin is faster than the rate of monomer dissociation from the pointed end of TRC-labeled F-actin, which limits the rate of treadmilling; and (iii) cofilin also accelerates the rate of subtilisin cleavage of F-actin in which the treadmilling is blocked by interprotomer cross-linking of the D-loop to the C terminus on an adjacent protomer. This suggests a substantial flexibility of the D-loop in the cross-linked F-actin. The increased cleavage rates of the D-loop and loop 60-69 reveal extensive exposure of subdomain 2 in F-actin to proteolytic enzymes by cofilin.     

Functional gamma-secretase inhibitors reduce beta-amyloid peptide levels in brain

To elucidate a role for the cytoskeletal protein actin in post-traumatic apoptotic cell death, the ability of actin-containing tissue extracts to inhibit exogenous DNase I was evaluated. In addition, cortical, hippocampal and thalamic extracts were examined for caspase-mediated actin cleavage and changes in actin polymerization state. Rats were anesthetized, subjected to lateral fluid percussion brain injury of moderate severity, and euthanized at 1 h, 6 h, 24 h, 1 week or 3 weeks post-injury (n = 3 per time-point). Tissue extracts from all brain regions of sham (uninjured) animals inhibited exogenous DNase I activity to a significant extent. However, inhibition of DNase I was significantly reduced at 1 h and 6 h in the injured hippocampus, and at 1 h, 6 h and 3 weeks in the thalamus. DNase I in cortical extracts of all injured animals was inhibited to a similar extent as that in uninjured animals. Actin fragments consistent with caspase-mediated proteolysis were observed in immunoblots of the injured hippocampus and thalamus at 1 h and 24 h, respectively, and were present up to 3 weeks post-injury. Transient actin hyperpolymerization was observed at 1 and 6 h post-injury in the thalamus and hippocampus, while actin depolymerization was observed at 1 and 3 weeks in the cortex and thalamus. Collectively our data suggest that DNase I disinhibition following brain trauma is associated predominantly with actin hyperpolymerization but also with actin depolymerization and concomitant caspase-mediated actin proteolysis.     

DNase I cleavage of adenoviral nucleoprotein.

Cleavage products resulting from DNase I treatment of adenoviral nucleoprotein were examined by gel electrophoresis, Southern blotting and hybridization to cloned restriction fragments derived from various regions of the viral genome. DNase I produced specific double-stranded cleavages in DNA of purified adenoviral cores and in DNA of intranuclear viral chromatin at early and late times of infection. At least some of these sites were also cleaved by DNase I in purified viral DNA, showing that sequence specificity of DNase I cleavage may contribute to the observation of specific double-stranded DNase I cleavage sites in adenoviral nucleoprotein. In addition, sites were observed which were specific either for cores or for intranuclear chromatin. In contrast to many cellular genes which have been characterized, there was no obvious relationship between DNase I cleavage sites and other features of the viral genome such as promoters or polyadenylation sites.     

Structural state of active and inactive genes during chromatin decondensation

Chromatin structure of globin and ovalbumin genes in chicken erythrocyte nuclei has been investigated by means of the "nuclease criterion" (described earlier). In intact nuclei (i.e. in the presence of 3 mM MgCl2) DNase I cleaves chromatin of both genes generating fragments multiple of a double-nucleosome repeat (2N-periodicity). However, in the case of the globin gene, apart from the 2N-periodicity, fragments were observed that are multiple of 100 b.p. and are characteristic for partially unfolded chromatin. This distinction in nuclease cleavage patterns correlates with a higher sensitivity of the globin gene as compared with the inactive ovalbumin gene. At 0.5-0.7 mM MgCl2 the transition from dinucleosomal fragmentation with DNase I and DNase II to fragmentation via a 100 b.p. interval occurs and the difference in digestibility of both genes is dramatically increased. If chromatin has been decondensed by incubation of nuclei in 10 mM Tris-buffer DNase Il generates an usual nucleosomal repeat, and in this ionic conditions one may not observe any difference in nuclease sensitivity of the analyzed genes. The data allow to suggest that the high nuclease sensitivity of potentially active genes can be conditioned by more relaxed arrangement of nucleosomes in higher order chromatin structure.     

Proteolytic cleavage of actin within the DNase-I-binding loop changes the conformation of F-actin and its sensitivity to myosin binding

Effects of subtilisin cleavage of actin between residues 47 and 48 on the conformation of F-actin and on its changes occurring upon binding of myosin subfragment-1 (S1) were investigated by measuring polarized fluorescence from rhodamine-phalloidin- or 1, 5-IAEDANS-labeled actin filaments reconstructed from intact or subtilisin-cleaved actin in myosin-free muscle fibers (ghost fibers). In separate experiments, polarized fluorescence from 1, 5-IAEDANS-labeled S1 bound to non-labeled actin filaments in ghost fibers was measured. The measurements revealed differences between the filaments of cleaved and intact actin in the orientation of rhodamine probe on the rhodamine-phalloidin-labeled filaments, orientation and mobility of the C-terminus of actin, filament flexibility, and orientation and mobility of the myosin heads bound to F-actin. The changes in the filament flexibility and orientation of the actin-bound fluorophores produced by S1 binding to actin in the absence of ATP were substantially diminished by subtilisin cleavage of actin. The results suggest that loop 38-52 plays an important role, not only in maintaining the F-actin structure, but also in the conformational transitions in actin accompanying the strong binding of the myosin heads that may be essential for the generation of force and movement during actin-myosin interaction.     

A structure of potentially active and inactive genes of chicken erythrocyte chromatin upon decondensation.

In the presence of 3 mM MgCl2 DNase I cleavage of bulk, globin and ovalbumin gene chromatin in chicken erythrocyte nuclei generates fragments which are multiples of a double-nucleosome repeat. However, in addition to the dinucleosomal periodicity beta-globin gene chromatin was fragmented into multiples of a 100 b.p. interval which is characteristic for partially unfolded chromatin. This distinction correlates with higher sensitivity of beta-globin domain to DNase I and DNase II as compared to the inactive ovalbumin gene. At 0.7 mM MgCl2 where these DNases fragment bulk chromatin into series of fragments with a 100 b.p. interval, the difference in digestibility of the investigated genes is dramatically decreased. When chromatin has been decondensed by incubation of nuclei in 10 mM Tris-buffer, DNase II generates a typical nucleosomal repeat, and the differential nuclease sensitivity of the analyzed genes is not observed. The data suggest that higher nuclease sensitivity of potentially active genes is due to irregularities in higher order chromatin structure.     

Nucleotide-induced changes in the proteolytically sensitive regions of myosin subfragment 1

Limited proteolytic digestions of myosin subfragment 1 (S-1) with elastase, subtilisin, papain, and thermolysin yield fragments that correspond within 1-2K daltons to the 25K, 50K, and 20K fragments produced by trypsin. While papain and thermolysin cut preferentially at the 26K/70K junction, elastase and subtilisin cleave both the 26K/70K and the 75K/22K junctions in S-1. Using the above proteases as conformational probes, we have previously demonstrated that the binding of actin is sensed at both the 26K/50K and the 50K/22K junctions [Applegate, D., & Reisler, E. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 7109-7112]. We report here that the binding of nucleotides at the active site is also sensed at both junctions. Both 2 mM MgADP and 5 mM MgATP slow the rate of elastase and subtilisin cleavage of the 95K heavy chain. With elastase, the 3-fold decrease in the rate of cleavage induced by nucleotides is evidenced at both the 26K/50K and the 50K/22K junctions. The analysis of subtilisin digestions is complicated by Mg nucleotide induced cleavage at a new site to produce a 91K fragment. Using N-methyl-6-anilinonaphthalene-2-sulfonyl chloride (MnsCl) to fluorescently label the 26K peptide, we demonstrate that the additional cleavage site is approximately 4K daltons from the N-terminal portion of the 95K heavy chain.(ABSTRACT TRUNCATED AT 250 WORDS)     

Carp liver actin: isolation, polymerization and interaction with deoxyribonuclease I (DNase I).

The aim of this study was to isolate and to characterize actin from the carp liver cytosol and to examine its ability to polymerize and interact with bovine pancreatic DNase I. Carp liver actin was isolated by ion-exchange chromatography, followed by gel filtration and a polymerization/depolymerization cycle or by affinity chromatography using DNase I immobilized to agarose. The purified carp liver actin was a cytoplasmic beta-actin isoform as verified by immunoblotting using isotype specific antibodies. Its isoelectric point (pI) was slightly higher than the pI of rabbit skeletal muscle alpha-actin. Polymerization of purified carp liver actin by 2 mM MgCl(2) or CaCl(2) was only obtained after addition of phalloidin or in the presence of 1 M potassium phosphate. Carp liver actin interacted with DNase I leading to the formation of a stable complex with concomitant inhibition of the DNA degrading activity of DNase I and its ability to polymerize. The estimated binding constant (K(b)) of carp liver actin to DNase I was calculated to be 1.85x10(8) M(-1) which is about 5-fold lower than the affinity of rabbit skeletal muscle alpha-actin to DNase I.     

Proteolysis of Paracoccus denitrificans cytochrome oxidase by trypsin and chymotrypsin

Paracoccus oxidase containing only two subunits was subjected to proteolysis by trypsin and chymotrypsin. Both subunits of the purified enzyme were cleaved at only a few sites and enzymatic activity was not inhibited. The cleavage sites were identified by protein sequencing. Subunit I was cleaved near the amino-terminus and subunit II in the loop connecting the two predicted trans-membrane helices. In native membrane fragments, but not in intact spheroplasts, this loop was accessible to both proteases. These results provide experimental evidence for the folding of subunit II in the membrane.     

LC20 and kinetics of gizzard myosin subfragment-1: digestion with papain vs. S. aureus protease.

Previous reports have shown that papain-digested gizzard subfragment-1 (PAP-S1) has a cleaved regulatory light chain (LC20), and Vmax similar to phosphorylated heavy meromyosin (HMM) (Greene et al., Biochemistry 22:530-535, 1983; Sellers et al., J. Biol. Chem. 257:13880-13883, 1982; Umemoto et al., J. Biol. Chem. 264:1431-1436, 1989], while S. aureus protease-digested S-1 (SAP-S1) has intact LC20, but Vmax closer to that of unphosphorylated HMM [Ikebe and Hartshorne, 1985]. To determine whether intact LC20 inhibits ATPase activity for subfragment-1 (S1), we compared the kinetic properties and structures of unphosphorylated PAP-S1 and SAP-S1. SDS-PAGE showed that SAP-S1 had 68 and 24 KDa heavy chain and 20 and 17 KDa light chain components. PAP-S1 (15 minutes digestion at 20 degrees C) also had 68 and 17 KDa bands, but the single 24 KDa band (24HC) was replaced by a group of 22-24 KDa fragments and LC20 was cleaved to a 16 KDa fragment. At 13 mM ionic strength, both PAP-S1 and SAP-S1 had Vmax similar to phosphorylated HMM (1.1-1.5 s-1). SAP-S1 had the same KATPase as phosphorylated HMM (38 microM actin), but KATPase for PAP-S1 was 3-fold stronger (11 microM actin). Subsequent digestion of SAP-S1 with papain did not significantly change Vmax, but as LC20 and 24HC were cleaved, both KATPase and Kbinding strengthened 3- to 5-fold. Thus, intact LC20 did not inhibit, and cleavage of LC20 did not increase Vmax for S1. Rather, papain cleavage of LC20 and 24HC was associated with strengthened actin binding.     

Identification of the site important for the actin-activated MgATPase activity of myosin subfragment-1.

The lysine-rich sequence (-KKGGKKK-) located at the 50,000/20,000 Mr junction of myosin subfragment-1 (S-1) was cleaved by endoprotease Arg-C or by trypsin in the presence of ATP and an equimolar amount of actin. Under these conditions, cleavage by Arg-C was between the first and second lysine residues, whereas cleavage by trypsin was between the third and fourth lysine residues. The actin-activated MgATPase activity of the S-1 cleaved by Arg-C was almost the same as native S-1, but S-1 cleaved by trypsin showed markedly reduced ATPase activity.     

Conformation changes of actin during formation of filaments and paracrystals and upon interaction with DNase I, cytochalasin B, and phalloidin

Spin labels attached to rabbit muscle actin became more immobilized upon conversion of actin from the G state to the F state with 50 mM KCl. Titration of G-actin with MgCl2 produced F-actin-like EPR spectra between 2 and 5 mM-actin filaments by electron microscopy. Higher concentrations of MgCl2 produced bundles of actin and eventually paracrystals, accompanied by further immobilization of spin labels. The effects of MgCl2 and KCl were competitive: addition of MgCl2 to 50 mM could convert F-actin (50 mM KCl) to paracrystalline (P) actin; the reverse titration (0 to 200 mM KCl in the presence of 20 mM MgCl2) was less complete. Addition of DNase I to G- or F-actin gave the expected amorphous electron micrographic pattern, and the actin was not sedimentable at (400,000 x g x h). EPR showed that the actin was in the G conformation. Addition of DNase I to paracrystalline actin gave the F conformation (EPR) but the actin was "G" by electron microscopy. Phalloidin converted G-actin to F-actin, had no effect on F-actin, and converted P-actin to the F state by electron microscopy but maintained the P conformation by EPR. Cytochalasin B produced no effects observable by EPR or centrifugation but "untwisted" paracrystals into nets. Since actin retained its P conformation by EPR in two states which were morphologically not P, we conclude that the P state is a distinct conformation of the actin molecule and that actin filaments aggregate to form bundles (and eventually paracrystals) when actin monomers are able to enter the P conformation.     

Expression and characterization of a DNase I-Fc fusion enzyme

Recombinant human deoxyribonuclease I (DNase I) is an important clinical agent that is inhaled into the airways where it degrades DNA to lower molecular weight fragments, thus reducing the viscoelasticity of sputum and improving the lung function of cystic fibrosis patients. To investigate DNases with potentially improved properties, we constructed a molecular fusion of human DNase I with the hinge and Fc region of human IgG1 heavy chain, creating a DNase I-Fc fusion protein. Infection of Sf9 insect cells with recombinant baculovirus resulted in the expression and secretion of the DNase I-Fc fusion protein. The fusion protein was purified from the culture medium using protein A affinity chromatography followed by desalting by gel filtration and was characterized by amino-terminal sequence, amino acid composition, and a variety of enzyme-linked immunosorbent assays (ELISA) and activity assays. The purified fusion contains DNase I, as determined by a DNase I ELISA and an actin-binding ELISA, and an intact antibody Fc region, which was quantified by an Fc ELISA, in a 2:1 stoichiometric ratio, respectively. The dimeric DNase I-Fc fusion was functionally active in enzymatic DNA digestion assays, albeit about 10-fold less than monomeric DNase I. Cleavage of the DNase I-Fc fusion by papain resulted in a specific activity comparable to the monomeric enzyme. Salt was inhibitory for wild type monomeric DNase I but actually enhanced the activity of the dimeric DNase I-Fc fusion. The DNase I-Fc fusion protein was also less Ca2+-dependent than DNase I itself. These results are consistent with a higher affinity of the dimeric fusion protein to DNA than monomeric DNase I. The engineered DNase I-Fc fusion protein described herein has properties that may have clinical benefits.