Structural characterization of myosin from bovine brain

Myosins isolated from bovine brain, rabbit skeletal muscle, and chicken gizzard smooth muscle and their heavy meromyosin and light meromyosin fractions were studied in the electron microscope by negative staining with uranyl acetate. Under similar conditions of preparation and polymerization, the three myosins formed paracrystals of different structures. The light meromyosin portion of the skeletal muscle myosin also assembled in a different fashion than the brain or smooth muscle light meromyosins; the latter two assembled similarly. The heavy meromyosin portion from each of the three myosins was shown to interact with the actins isolated from each of the three tissue sources by the formation of the characteristic arrowhead patterns with similar periodicities. The brain heavy meromyosin attachment to both skeletal and brain actins was dissociated by ATP. It is suggested that differences in the light meromyosin portions of the three myosins may account in part for their differences in assembly in vivo.     

Studies on the amino groups of myosin ATPase. Trinitrophenylation of reactive lysyl residue in ventricular and atrial myosins

Myosin and heavy meromyosin from ventricular, atrial, and skeletal muscle were purified and trinitrophenylated by 2,4,6-trinitrobenzene sulfonate. The trinitrophenylation reaction followed a complex kinetics consisting of a fast and slow reaction in all preparations studied. Reactive lysine residues were trinitrophenylated during the fast reaction with a concomitant decrease in K⁺ (EDTA)-activated ATPase and an increase in Mg²⁺-stimulated ATPase activities of myosin. The extent of increase in Mg²⁺-mediated ATPase was the highest with skeletal and the lowest with atrial myosin. The trinitrophenylation of the less reactive lysyl residues continued during the slow reaction. The rate constants of the reactions and the number of reactive lysine residues were evaluated by computer analyses of the trinitrophenylation curves. Two reactive lysine residues were found in skeletal and ventricular myosins while their number in atrial myosin was somewhat lower. The rate of trinitrophenylation in skeletal muscle myosin or heavy meromyosin was always higher than in the two cardiac myosin isozymes. Addition of KCl increased the trinitrophenylation of both highly reactive and slowly reactive lysyl residues in all of the three heavy meromyosins, however, the effect was more profound with cardiac heavy meromyosins. Addition of MgADP induced spectral changes in trinitrophenylated skeletal but not in cardiac myosins. Similar changes occurred in skeletal and to a lesser degree in ventricular heavy meromyosin, but no definite spectral changes were observed in atrial heavy meromyosin. The findings suggest that structural differences exist around the reactive lysyl residue in the head portion of the three myosins.     

Purification and characterization of a smooth muscle myosin phosphatase from turkey gizzards

A phosphoprotein phosphatase that dephosphorylates smooth muscle myosin has been purified to apparent homogeneity from turkey gizzards. Smooth muscle phosphatase (SMP) IV has a molecular weight of 150,000 as determined by gel filtration on a Sephadex G-200 column and is composed of two subunits (Mr = 58,000 and 40,000). Although it is active toward a number of proteins, its activities toward the contractile proteins, intact myosin, heavy meromyosin, and isolated myosin light chains are higher than its activities toward phosphorylase alpha, histone IIA, and phosphorylase kinase. SMP-IV preferentially dephosphorylates the beta-subunit of phosphorylase kinase. The properties of the enzyme have been studied using heavy meromyosin, a soluble chymotryptic fragment of myosin, and isolated myosin light chains as substrates. SMP-IV has high affinity for both substrates and is optimally active at neutral pH. Divalent cations, Ca2+ and Mg2+, activate the dephosphorylation of heavy meromyosin but inhibit the activity toward myosin light chains. Low concentrations of ATP (1-5 mM) activate SMP-IV but concentrations higher than 5 mM are inhibitory. Inhibition of 50% of the activity of the enzyme by NaF and PPi requires concentrations higher than 10 mM. Rabbit skeletal muscle heat stable inhibitor-2 has no effect on the activity of SMP-IV toward heavy meromyosin, myosin light chains, and phosphorylase alpha.     

Immunochemical evidence for the species-specificity of mammalian cardiac myosin and heavy meromyosin.

Structural differences between various myosins were investigated by means of antibodies to heavy meromyosin, a tryptic subfragment of myosin. Heavy meromyosin was purified from rabbit white skeletal and from pig and human cardiac muscles by gel filtration, and antisera were produced in guinea pigs. Analyses, carried out with the quantitative micro-complement fixation technique, indicated that the antibodies were specific to heavy meromyosin and myosin and not to other contractile proteins. For each muscle type, the corresponding intact myosin reacted, and the degree of dixation was always lower than with heavy meromyosin (50 and 70% fixation respectively). This vertical shift was the same for the three muscle types, indicating that the heavy meromyosin represent corresponding fragments of the myosin molecule from one muscle to the other. Antisera to pig or human cardiac heavy meromyosin clearly distinguished antigens (heavy meromyosins, myosins, or crude extracts) from the ventricles of various heterologous species. Relative to pig, the immunological distances were 50 for the rabbit, 73 for the rat and greater than 100 for human and mice. Relative to human, these values were 20 for the rat, 60 for the rabbit, 72 for the pig. These data provide direct evidence that mammalian cardiac myosin is species-specific.     

Two smooth muscle myosin heavy chains differ in their light meromyosin fragment

Smooth muscle myosin heavy chains [SM1, approximately 205 kilodaltons (kDa), and SM2, approximately 200 kDa] were separated on sodium dodecyl sulfate (SDS)-polyacrylamide gels. Peptide maps of the two heavy chains showed unique patterns. Limited proteolytic cleavage of purified swine stomach myosin was performed by using a variety of proteases to produce the major myosin fragments which were resolved on SDS gels. A single band was obtained for heavy meromyosin in the soluble fraction following chymotrypsin digestion. However, a variable number of bands were observed for light meromyosin fragments in the insoluble fraction after chymotrypsin digestion. Peptide mapping indicated that the two bands observed after short digestion times with chymotrypsin had relative mobility and solubility properties consistent with approximately 100- and 95-kDa light meromyosin (LMM) fragments. These results indicate that the region of difference between SM1 and SM2 lies in the LMM fragment.     

Effect of skeletal muscle myosin light chain 2 on the Ca2+-sensitive interaction of myosin and heavy meromyosin with regulated actin

A low-speed centrifugation assay has been used to examine the binding of myosin filaments to F-action and to regulated actin in the presence of MgATP. While the cross-linking of F-actin by myosin was Ca2+ insensitive, much less regulated actin was cross-linked by myosin in the absence of Ca2+ than in its presence. Removal of the 19000-dalton, phosphorylatable light chain from myosin resulted in the loss of this Ca2+ sensitivity. Readdition of this light chain partially restored the Ca2+-sensitive cross-linking of regulated actin by myosin. Urea gel electrophoresis has been used to distinguish that fraction of heavy meromyosin which contains intact phosphorylatable light chain from that which contains a 17000-dalton fragment of this light chain. In the absence of Ca2+, heavy meromyosin which contained digested light chain bound to regulated actin in MgATP about 10-fold more tightly than did heavy meromyosin which contained intact light chain. The regulated actin-activated ATPases of heavy meromyosin also showed that cleavage of this light chain causes a substantial increase in the affinity of heavy meromyosin for regulated actin in the absence of Ca2+. Thus, the binding of both myosin and heavy meromyosin to regulated actin is Ca2+ sensitive, and this sensitivity is dependent on the phosphorylatable light chain.     

Effects of actin and calcium ion on chymotryptic digestion of skeletal myosin and their implications to the function of light chains

Experiments have been carried out to assess the involvement of the myosin light chains [obtained by treatment of myosin with 5,5'-dithiobis(2-nitrobenzoic acid) (Nbs2)] in the control of cross-bridge movement and actomyosin interactions. Chymotryptic digestions of myosin, actomyosin, and myofibrils do not detect any Ca2+-induced change in the subfragment 2 region of myosin. Actin, like Ca2+, protects the in situ Nbs2 light chains from proteolysis and causes a partial switch in the digestion product of myosin from subfragment 1 to heavy meromyosin. This effect is independent of the state of aggregation of myosin, and it persists in acto heavy meromyosin and in actinomyosin in 0.6 M NaCl. Digestions and sedimentation studies indicate that there is no direct acto light chain interaction. Proteolysis of myosin shows a gradual transition from production of heavy meromyosin to subfragment 1 with lowering of the salt level. In the presence of Ca2+ heavy meromyosin is generated both in digestions of polymeric and of monomeric myosin. These results are explained in terms of localized changes within the Nbs2 light chains and subfragment 1. Subunit interactions in the myosin head lead to a Ca2+-induced reduction in the affinity of heavy meromyosin for actin in the presence of MgATP. The resulting Ca2+ inhibition of the actin-activated ATPase of myosin can be detected at high salt concentrations(75 mM KCl).     

Enzymatic mechanochemistry. I. The interaction of heavy meromyosin with “immobilized adenosine triphosphate”

ATP was covalently bound to an agarose gel. The insolubilized ATP was found to be capable of specifically binding heavy meromyosin. The adsorbed heavy meromyosin could be eluted by ATP in solution. Both binding and elution by ATP of heavy meromyosin were not much effected by Ca²⁺, Mg²⁺ or EDTA.While the water-soluble polyalanine-myosin was also found to be adsorbed, myosin in 0.5 M KCl did not seem to be adsorbed by agarose-ATP.Both Mg²⁺ and Ca²⁺ appear to activate the splitting of bound ATP by heavy meromyosin to practically the same extent.We prepared water-soluble derivatives of ATP in which ATP underwent the same chemical modification required for its coupling to agarose but in which the agarose component was absent. Their splitting by heavy meromyosin was also activated by Mg²⁺ though to a lesser extent but actin did not influence this reaction.Possible relations between our findings and the various stages of the reaction between myosin and ATP, as well as the potential use of columns filled with insolubilized NTPs for the separation and purification of myosin and of its subfragments, are discussed.     

Decoration of actin filaments with skeletal muscle heavy meromyosin containing either phosphorylated or dephosphorylated regulatory light chains

Heavy meromyosin containing almost intact regulatory light chains (LC2) was obtained from monomeric phosphorylated and dephosphorylated rabbit fast skeletal muscle myosin by brief chymotryptic digestion in the presence of CaCl2. Actin filaments, complexed with heavy meromyosin, display two different forms of arrowhead, depending on the form of LC2.     

Myosin and actomyosin from human skeletal muscle.

Human skeletal natural actomyosin contained actin, tropomyosin, troponin and myosin components as judged by polyacrylamide gel electrophoresis in sodium dodecyl sulfate. Purified human myosin contained at least three light chains having molecular weights (+/-2000) of 25 000, 18 000 and 15 000. Inhibitory and calcium binding components of troponin were identified in an actin-tropomyosin-troponin complex extracted from acetone-dried muscle powder at 37 degrees C. Activation of the Mg-ATPase activity of Ca2+-sensitive human natural or reconstituted actomyosin was half maximal at approximately 3.4 muM Ca2+ concentration (CaEGTA binding constant equals 4.4 - 10(5) at pH 6.8). Subfragment 1, isolated from the human heavy meromyosin by digestion with papain, appeared as a single peak after DEAE-cellulose chromatography. In the pH 6-9 range, the Ca2+-ATPase activity of the subfragment 1 was 1.8- and 4-fold higher that the original heavy meromyosin and myosin, respectively. The ATPase activities of human myosin and its fragments were 6-10 fold lower than those of corresponding proteins from rabbit fast skeletal muscle. Human myosin lost approximately 60% of the Ca2+-ATPase activity at pH 9 without a concomitant change in the number of distribution of its light chains. These findings indicate that human skeletal muscle myosin resembles other slow and fast mammalian muscles. Regulation of human skeletal actomyosin by Ca2+ is similar to that of rabbit fast or slow muscle.     

Calcium ions modulate regulation of smooth muscle contraction mediated by phosphorylation of myosin regulatory light chains

The effect of calcium ions on conformational changes of F-actin initiated by decoration of thin filaments with phosphorylated and dephosphorylated heavy meromyosin from smooth muscles was studied by fluorescence polarization spectroscopy. It is shown that heavy meromyosin with phosphorylated regulatory light chains (pHMM) promotes structural changes of F-actin which are typical for the "strong" binding of actin to the myosin heads. Heavy meromyosin with dephosphorylated regulatory light chains (dpHMM) causes conformational changes of F-actin which are typical for the "weak" binding of actin to the myosin heads. The presence of calcium enhances the pHMM effect and attenuates the dpHMM effect. We propose that a Ca2+-dependent mechanism exists in smooth muscles which modulates the regulation of actin--myosin interaction occurring via phosphorylation of myosin regulatory light chains.     

The action of thiol reagents on the adenosine-triphosphatase activities of heavy meromyosin and L-myosin

1. The Ca²⁺-activated adenosine triphosphatase of heavy meromyosin is maximally stimulated by lower relative molar concentrations of phenylmercuric acetate than are required with myosin. 2. Stimulation of the Ca²⁺-activated adenosine triphosphatase of both heavy meromyosin and myosin by thiol reagents is markedly affected by ionic strength, the effects being greater with the former than with the latter. In particular, N-ethylmaleimide strongly inhibits the Ca²⁺-activated adenosine triphosphatase of heavy meromyosin at ionic strength below about 0·2. 3. The precise behaviour of the thiol reagents at low ionic strength is slightly modified by the age of the heavy meromyosin and myosin preparations. 4. Stimulation of the Mg²⁺-activated adenosine triphosphatase of heavy meromyosin by thiol reagents is relatively insensitive to ionic strength. 5. The adenosine triphosphatases of heavy meromyosin and myosin activated by potassium chloride in the absence of bivalent activators are inhibited by thiol reagents over the range of ionic strength at which stimulation occurs in the presence of calcium chloride as activator. 6. The modifying effects of potassium chloride and sodium chloride are qualitatively different when heavy-meromyosin adenosine triphosphatase is stimulated with phenylmercuric acetate. No such difference is observed when the enzyme is stimulated with N-ethylmaleimide.     

Degradation of smooth-muscle myosin by trypsin-like serine proteinases.

1. Hydrolysis of the myosins from smooth and from skeletal muscle by a rat trypsin-like serine proteinase and by bovine trypsin at pH 7 is compared. 2. Proteolysis of the heavy chains of both myosins by the rat enzyme proceeds at rates approx. 20 times faster than those obtained with bovine trypsin. Whereas cleavage of skeletal-muscle myosin heavy chain by both enzymes results in the generation of conventional products i.e. heavy meromyosin and light meromyosin, the heavy chain of smooth-muscle myosin is degraded into a fragment of mol. wt. 150000. This is dissimilar from heavy meromyosin and cannot be converted into heavy meromyosin. It is shown that proteolysis of the heavy chain takes place in the head region. 3. The 'regulatory' light chain (20kDa) of smooth-muscle myosin is degraded very rapidly by the rat proteinase. 4. The ability of smooth-muscle myosin to have its ATPase activity activated by actin in the presence of a crude tropomyosin fraction on introduction of Ca2+ is diminished progressively during exposure to the rat proteinase. The rate of loss of the Ca2+-activated actomyosin ATPase activity is very similar to the rate observed for proteolysis of the heavy chain and 3-4 times slower than the rate of removal of the so-called 'regulatory' light chain. 5. The significance of these findings in terms of the functional organization of the smooth muscle myosin molecule is discussed. 6. Since the degraded myosin obtained after exposure to very small amounts of the rat proteinase is no longer able to respond to Ca2+, i.e. the functional activity of the molecule has been removed, the implications of a similar type of proteolysis operating in vivo are considered for myofibrillar protein turnover in general, but particularly with regard to the initiation of myosin degradation, which is known to take place outside the lysosome (i.e. at neutral pH).     

Interaction of AMP-aminohydrolase with myosin and its subfragments.

We have shown that purified rabbit skeletal muscle AMP-aminohydrolase binds to rabbit muscle myosin, heavy meromyosin, and Subfragment 2 but does not bind to light meromyosin nor to Subfragment 1. The dissociation constant for binding to myosin was determined to be 0.14 muM. A new sedimentation boundary, presumably reflecting formation of a complex between AMP-aminohydrolase and heavy meromyosin or Subfragment 2, can be observed using the analytical ultracentrifuge. Binding of AMP-aminohydrolase to myosin, heavy meromyosin, or Subfragment 2 is abolished by phosphate (less than 10 mM), an inhibitor of AMP-aminohydrolase. No other rabbit muscle enzyme tested showed any interaction with myosin under the same conditions and there was no indication of complex formation between AMP-aminohydrolase and phosphofructokinase or phosphocreatine kinase in the analytical ultracentrifuge.     

The mechanism of regulatory light chain dissociation from scallop myosin.

The dissociation of the regulatory light chains from scallop myosin subfragments, on addition of EDTA, was investigated by using the fluorophore 8-anilinonaphthalene-1-sulphonate as a probe. The rate of this process (0.014 s-1) was partially limited by the rate of Mg2+ dissociation (0.058 s-1) from the non-specific high-affinity site. The dissociation of the regulatory light chain subfragment 1 was less extensive than from heavy meromyosin. Reassociation of the scallop regulatory light chain was induced on addition of Mg2+, but it appeared to be limited by a first-order step. The nature of this step was revealed by the kinetics of Mercenaria regulatory light chain association. Scallop heavy meromyosin, denuded of its regulatory light chains, exists in a refractory state, whose reversal to the nascent state limits the rate of light chain association (0.006 s-1). The formation of the refractory state is the driving force for the net dissociation of regulatory light chains from scallop heavy meromyosin. This mechanism is discussed with reference to existing structural information on light-chain-denuded myosin.     

Isolation of Messenger Ribonucleic Acid for Myosin Heavy Chain

A chicken embryonic polysome fraction that contains 50–60 monoribosomes and synthesizes the heavy chains of myosin is separated from other polysomes of smaller sizes by centrifugation through two cycles of discontinuous and continuous sucrose gradients. The unique properties of the polyadenylic acid segment present at the 3′-end of eukaryotic messenger RNA (mRNA) were used to purify the mRNA for myosin heavy chain from the phenol-extracted total RNA obtained from this polysome fraction. The total RNA was filtered thro ugh millipore filters resulting in partition of the riboscmal RNA (rRNA) and mRNA species. This millipore-bound RNA fraction, which consists of the mRNA and some ribosomal RNAs, was eluted from the filters with sodium dodecyl sulfate (SDS). Subsequent chromatography of this fraction on a cellulose column gave two well-separated peaks: an unadsorbed peak of ribosomal RNAs which was eluted with buffers of high ionic strength and an adsorbed peak of mRNA which was eluted only with a buffer of low ionic strength. Polyacrylamide gel electrophoresis of the mRNA peak fraction showed a single band with no detectable amounts of other RNAs, the mRNA migrating slower than 28S rRNA. The product of in vitro translation of the purified mRNA using a homologous cell-free system was identified as the myosin heavy chain by the following criteria: coprecipitation with carrier myosin at low ionic strength; elution properties on DEAE-cellulose column; and comigration with the heavy chain in polyacrylamide gel electrophoresis. In order to demonstrate the fidelity of translation of the mRNA, ¹⁴C-labeled products of the in vitro translation were copurified with unlabeled myosin heavy chains added as a carrier. The mixture of polypeptides was then cleaved with CNBr and the resulting peptides were separated by molecular sieving. The correlation between the radioactivity and the UV absorbance in the separated peptides indicates that total synthesis of the myosin heavy chain was achieved.     

Incorporation of the Ca2+ -binding component (g2) into heavy meromyosin and subfragment-1 of pig cardiac myosin.

A new protein component was found in heavy meromyosin and in subfragment-1 (S-1) prepared by chymotrypsin digestion of pig cardiac myosin in the presence of Ca2+. The molecular weight of this protein was estimated as 15,000 dalton. It was able to bind Ca2+ and showed a similar UV absorption spectrum to that of the g2 light chain. Heavy meromyosin and subfragment-1 which contained the 15,000 dalton component incorporated exogenous g2 and the 15,000 dalton component disappeared after such treatment. We concluded that the 15,000 dalton component was produced from g2 by limitted proteolysis. The subfragment-1 was separated into two protein fractions in equal yield by recycling the gel filtration. One contained the 15,000 dalton component and was able to bind Ca2+ while the other did not contain the component and was unable to bind Ca2+. According to analysis by SDS gel electrophoresis, the large polypeptide chain (the f component) of the first S-1 was approximately 5,000 dalton larger than the f component of the second S-1. The polypeptide corresponding to 5,000 dalton was designated polypeptide-C, because it was released from the C terminal of the f component. It seems to be essential for the attachment of the Ca2+-binding light chain g2. The location of g2 in myosin may thus be at the polypeptide-C which links the head to the tail of myosin.     

Proteolysis of smooth muscle myosin by Staphylococcus aureus protease: preparation of heavy meromyosin and subfragment 1 with intact 20 000-dalton light chains

The proteolysis of gizzard myosin by Staphylococcus aureus protease produces both heavy meromyosin and subfragment 1 in which the 20 000-dalton light chains are intact, and conditions are suggested for the preparation of each. Cleavage of the myosin heavy chain to produce subfragment 1 is dependent on the myosin conformation. Proteolysis of myosin in the 10S conformation yields predominantly heavy meromyosin, and myosin in the 6S conformation yields mostly subfragment 1 and some heavy meromyosin. Two sites are influenced by myosin conformation, and these are located at approximately 68 000 and 94 000 daltons from the N-terminus of the myosin heavy chain. The latter site is thought to be located at the subfragment 1-subfragment 2 junction, and cleavage at this site results in the production of subfragment 1. The time courses of phosphorylation of both heavy meromyosin and subfragment 1 can be fit by a single exponential. The actin-activated Mg2+-ATPase activity of heavy meromyosin is markedly activated by phosphorylation of the 20 000-dalton light chains. From the actin dependence of Mg2+-ATPase activity the following values are obtained: for phosphorylated heavy meromyosin, Vmax approximately 5.6 s-1 and Ka (the apparent dissociation constant for actin) approximately 2 mg/mL; for dephosphorylated heavy meromyosin, Vmax approximately 0.2 s-1 and Ka approximately 7 mg/mL. The actin-activated ATPase activity of subfragment 1 is not influenced by phosphorylation, and Vmax and Ka for both the phosphorylated and dephosphorylated forms are 0.4 s-1 and 5 mg/mL, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)     

The disappearance of the dependence of actin-myosin interaction on the phosphorylation of myosin light chains in the "freezing" of the structure of heavy meromyosin by a bifunctional reagent

Using glycerinated muscle fibers, free of myosin, tropomyosin and troponin, a study was made of the structural state of F-actin modified by N-(iodoacetyl)-N'-(1-naphthyl-5-sulfo)-ethylendiamine (1.5-IAEDANS) and by rhodaminyl--phalloin at decoration of thin filaments with a proteolytic fragment of myosin--heavy meromyosin containing phosphorylated and dephosphorylated myosin light chains. The heavy meromyosin used has three SH-groups of heavy chain SH1, SH2 and SH chi modified by bifunctional reagent N,N'-n-phenylmaleimide (SH1-SH2, SH2-SH chi). At decoration of thin filaments with heavy meromyosin, some changes in polarized fluorescence of rhodaminyl--phalloin and 1.5-IAEDANS independent of phosphorylation of myosin light chains were found. Fluorescence anisotropy of the fiber was found to depend primarily on the character of heavy chain of SH-group modification. The ability of heavy chains to change their conformations is supposed to play an important role in the mechanism of myosin system modulation of muscle contraction.     

Phosphorylation-dependent structural changes in the regulatory light chain domain of smooth muscle heavy meromyosin.

Smooth muscle heavy meromyosin, a double-headed proteolytic fragment of myosin lacking the COOH-terminal two-thirds of the tail, has been shown previously to be regulated by phosphorylation. To examine phosphorylation-dependent structural changes near the head-tail junction, we prepared five well regulated heavy meromyosins containing single-cysteine mutants of the human smooth muscle regulatory light chain labeled with the photocross-linking reagent, benzophenone-iodoacetamide. For those mutants that generated cross-links, only one type of cross-linked species was observed, a regulatory light chain dimer. Irradiated mutants fell into two classes. First, for Q15C, A23C, and wild type (Cys-108), a regulatory light chain dimer was formed for dephosphorylated but not thiophosphorylated heavy meromyosin. These data provide direct chemical evidence that in the dephosphorylated state, Gln-15, Ala-23, and Cys-108 on one head are positioned near (within 8.9 A) the regulatory light chain of the partner head and that thiophosphorylation abolishes proximity. This behavior was also observed for the Q15C mutant on a truncated heavy meromyosin lacking both catalytic domains. For the actin-heavy meromyosin complex, cross-links were formed in both de- and thiophosphorylated states. S59C and T134C mutants were in a second mutant class, where regulatory light chain dimers were not detected in dephosphorylated or thiophosphorylated heavy meromyosin, suggesting positions outside the region of interaction of the regulatory light chains.