Leishmania donovani bisubunit topoisomerase I gene fusion leads to an active enzyme with conserved type IB enzyme function

All eukaryotic topoisomerase I enzymes are monomeric enzymes, whereas the kinetoplastid family (Trypanosoma and Leishmania) possess an unusual bisubunit topoisomerase I. To determine what happens to the enzyme architecture and catalytic property if the two subunits are fused, and to explore the functional relationship between the two subunits, we describe here in vitro gene fusion of Leishmania bisubunit topoisomerase I into a single ORF encoding a new monomeric topoisomerase I (LdTOPIL-fus-S). It was found that LdTOPIL-fus-S is active. Gene fusion leads to a significant modulation of in vitro topoisomerase I activity compared to the wild-type heterodimeric enzyme (LdTOPILS). Interestingly, an N-terminal truncation mutant (1-210 amino acids) of the small subunit, when fused to the intact large subunit [LdTOPIL-fus-Delta(1-210)S], showed reduced topoisomerase I activity and camptothecin sensitivity in comparison to LdTOPIL-fus-S. Investigation of the reduction in enzyme activity indicated that the nonconserved 1-210 residues of LdTOPIS probably act as a 'pseudolinker' domain between the core and catalytic domain of the fused Leishmania enzyme, whereas mutational analysis of conserved His453 in the core DNA-binding domain (LdTOPIL) strongly suggested that its role is to stabilize the enzyme-DNA transition state through hydrogen bonding to one of the nonbridging oxygens. Taken together, our findings provide an insight into the details of the unusual structure of bisubunit topoisomerase I of Leishmania donovani.     

N-terminal region of the large subunit of Leishmania donovani bisubunit topoisomerase I is involved in DNA relaxation and interaction with the smaller subunit

Leishmania donovani topoisomerase I is an unusual bisubunit enzyme. We have demonstrated earlier that the large and small subunit could be reconstituted in vitro to show topoisomerase I activity. We extend our biochemical study to evaluate the role of the large subunit in topoisomerase activity. The large subunit (LdTOP1L) shows a substantial degree of homology with the core DNA binding domain of the topoisomerase IB family. Two N-terminal truncation constructs, LdTOP1Delta39L (lacking amino acids 1-39) and LdTOP1Delta99L (lacking amino acids 1-99) of the large subunit were generated and mixed with intact small subunit (LdTOP1S). Our observations reveal that residues within amino acids 1-39 of the large subunit have significant roles in modulating topoisomerase I activity (i.e. in vitro DNA relaxation, camptothecin sensitivity, cleavage activity, and DNA binding affinity). Interestingly, the mutant LdTOP1Delta99LS was unable to show topoisomerase I activity. Investigation of the loss of activity indicates that LdTOP1Delta99L was unable to pull down glutathione S-transferase-LdTOP1S in an Ni(2+)-nitrilotriacetic acid co-immobilization experiment. For further analysis, we co-expressed LdTOP1L and LdTOP1S in Escherichia coli BL21(DE3)pLysS cells. The lysate shows topoisomerase I activity. Immunoprecipitation revealed that LdTOP1L could interact with LdTOP1S, indicating the subunit interaction in bacterial cells, whereas immunoprecipitation of bacterial lysate co-expressing LdTOP1Delta99L and LdTOP1S reveals that LdTOP1Delta99L was significantly deficient at interacting with LdTOP1S to reconstitute topoisomerase I activity. This study demonstrates that heterodimerization between the large and small subunits of the bisubunit enzyme appears to be an absolute requirement for topoisomerase activity. The residue within amino acids 1-39 from the N-terminal end of the large subunit regulates DNA topology during relaxation by controlling noncovalent DNA binding or by coordinating DNA contacts by other parts of the enzyme.     

The large subunit of Leishmania topoisomerase I functions as the 'molecular steer' in type IB topoisomerase

Kinetoplastid topoisomerase IB is an unusual bisubunit enzyme where reconstitution of the large (LdTOPIL or L) and small (LdTOPIS or S) subunits shows functional activity. It is yet to be deciphered whether one subunit or both navigate the heterodimer to its cellular DNA targets. Tethering a specific DNA-binding protein to topoisomerase I alters its site specificity. The chimeric constructs UMSBP-LdTOPIL/S or U-L/S (fusion of UMSBP to the N-terminus of L and reconstituted with S) and LdTOPIL/UMSBP-LdTOPIS or L/U-S (fusion of UMSBP to the N-terminus of S and reconstituted with L) exhibit relaxation activity. Only U-L/S shows altered site specificity and enhanced DNA-binding affinity for the universal minicircle sequence (UMS) containing substrate. This proves that L alone serves as the 'molecular steer' for this heterodimer. Reconstituted U-L/S also induces cleavage close to UMS and causes minicircle linearization. The differential properties of the reconstituted chimeras U-L/S and L/U-S reveal the structural and functional asymmetry between the heterodimer. Therefore this study helps in a better understanding of the mechanistic details underlying topoisomerization by this bi-subunit enzyme.     

The N-terminal domain of the regulatory subunit is sufficient for complete activation of acetohydroxyacid synthase III from Escherichia coli

We have previously proposed a model for the fold of the N-terminal domain of the small, regulatory subunit (SSU) of acetohydroxyacid synthase isozyme III. The fold is an alpha-beta sandwich with betaalphabetabetaalphabeta topology, structurally homologous to the C-terminal regulatory domain of 3-phosphoglycerate dehydrogenase. We suggested that the N-terminal domains of a pair of SSUs interact in the holoenzyme to form two binding sites for the feedback inhibitor valine in the interface between them. The model was supported by mutational analysis and other evidence. We have now examined the role of the C-terminal portion of the SSU by construction of truncated polypeptides (lacking 35, 48, 80, 95, or 112 amino acid residues from the C terminus) and examining the properties of holoenzymes reconstituted using these constructs. The Delta35, Delta48, and Delta80 constructs all lead to essentially complete activation of the catalytic subunits. The Delta80 construct, corresponding to the putative N-terminal domain, has the highest level of affinity for the catalytic subunits and leads to a reconstituted enzyme with k(cat)/K(M) about twice that of the wild-type enzyme. On the other hand, none of these constructs binds valine or leads to a valine-sensitive enzyme on reconstitution. The enzyme reconstituted with the Delta80 construct does not bind valine, either. The N-terminal portion (about 80 amino acid residues) of the SSU is thus necessary and sufficient for recognition and activation of the catalytic subunits, but the C-terminal half of the SSU is required for valine binding and response. We suggest that the C-terminal region of the SSU contributes to monomer-monomer interactions, and provide additional experimental evidence for this suggestion.     

Chain length determination of prenyltransferases: both heteromeric subunits of medium-chain (E)-prenyl diphosphate synthase are involved in the product chain length determination

Among prenyltransferases, medium-chain (E)-prenyl diphosphate synthases are unusual because of their heterodimeric structures. The larger subunit has highly conserved regions typical of (E)-prenyltransferases. The smaller one has recently been shown to be involved in the binding of allylic substrate as well as determining the chain length of the reaction product [Zhang, Y.-W., et al. (1999) Biochemistry 38, 14638-14643]. To better understand the product chain length determination mechanism of these enzymes, several amino acid residues in the larger subunits of Micrococcus luteus B-P 26 hexaprenyl diphosphate synthase and Bacillus subtilis heptaprenyl diphosphate synthase were selected for substitutions by site-directed mutagenesis and examined by combination with the corresponding wild-type or mutated smaller subunits. Replacement of the Ala at the fifth position upstream to the first Asp-rich motif with bulky amino acids in both larger subunits resulted in shortening the chain lengths of the major products, and a double combination of mutant subunits of the heptaprenyl diphosphate synthase, I-D97A/II-A79F, yielded exclusively geranylgeranyl diphosphate. However, the combination of a mutant subunit and the wild-type, I-Y103S/II-WT or I-WT/II-I76G, produced a C(40) prenyl diphosphate, and the double combination of the mutants, I-Y103S/II-I76G, gave a reaction product with longer prenyl chain up to C(50). These results suggest that medium-chain (E)-prenyl diphosphate synthases take a novel mode for the product chain length determination, in which both subunits cooperatively participate in maintaining and determining the product specificity of each enzyme.     

N-terminal deletion of the gamma subunit affects the stabilization and activity of chloroplast ATP synthase

Five truncation mutants of chloroplast ATP synthase gamma subunit from spinach (Spinacia oleracea) lacking 8, 12, 16, 20 or 60 N-terminal amino acids were generated by PCR by a mutagenesis method. The recombinant gamma genes were overexpressed in Escherichia coli and assembled with alphabeta subunits into a native complex. The wild-type (WT) alphabetagamma assembly i.e. alphabetagammaWT exhibited high (Mg2+)-dependent and (Ca2+)-dependent ATP hydrolytic activity. Deletions of eight residues of the gamma subunit N-terminus caused a decrease in rates of ATP hydrolysis to 30% of that of the alphabetaWT assembly. Furthermore, only approximately 6% of ATP hydrolytic activity was retained with the sequential deletions of gamma subunit up to 20 residues compared with the activity of the alphabetaWT assembly. The inhibitory effect of the epsilon subunit on ATP hydrolysis of these alphabetagamma assemblies varied to a large extent. These observations indicate that the N-terminus of the gamma subunit is very important, together with other regions of the gamma subunit, in stabilization of the enzyme complex or during cooperative catalysis. In addition, the in vitro binding assay showed that the gamma subunit N-terminus is not a crucial region in binding of the epsilon subunit.     

Dissociation of m-calpain subunits occurs after autolysis of the N-terminus of the catalytic subunit, and is not required for activation.

Calpain is a heterodimeric, intracellular Ca(2+)-dependent, "bio-modulator" that alters the properties of substrates through site-specific proteolysis. It has been proposed that calpains are activated by autolysis of the N-terminus of the large subunit and/or its dissociation into the subunits. It is, however, unclear whether the dissociation into subunits is required for the expression of protease activity and/or for in vivo function. Recently, the crystal structure of m-calpain in the absence of Ca(2+) has been resolved. The 3D structure clearly shows that the N-terminus of the m-calpain large subunit (mCL) makes contact with the 30K subunit, suggesting that autolysis of the N-terminus of mCL changes the interaction of both subunits. To examine the relationship between autolysis, dissociation, and activation, we made and analysed a series of N-terminal mutants of mCL that mimic the autolysed forms or have substituted amino acid residue(s) interacting with 30K. As a result, the mutant m-calpains, which are incapable of autolysis, did not dissociate into subunits, whereas those lacking the N-terminal 19 residues (Delta 19), but not those lacking only nine residues (Delta 9), dissociated into subunits even in the absence of Ca(2+). Moreover, both Delta 9 and Delta 19 mutants showed an equivalent reduced Ca(2+) requirement for protease activity. These results indicate that autolysis is necessary for the dissociation of the m-calpain subunits, and that the dissociation occurs after, but is not necessary for, activation.     

A functional domain of the alpha1 subunit of soluble guanylyl cyclase is necessary for activation of the enzyme by nitric oxide and YC-1 but is not involved in heme binding

Soluble guanylyl cyclase is a heterodimeric enzyme consisting of an alpha(1) and a beta(1) subunit and is an important target for endogenous nitric oxide and the guanylyl cyclase modulator YC-1. The activation of the enzyme by both substances is dependent on the presence of a prosthetic heme group. It has been unclear whether this prosthetic heme group is sandwiched between the alpha(1) and beta(1) subunits or whether it exclusively binds to the beta(1) subunit. Here we analyze progressive amino-terminal deletion mutants of the human alpha(1) subunit after co-expression with the human beta(1) subunit in the baculovirus/Sf9 system. Spectral, biochemical, and pharmacological analysis shows that the first 259 amino acids of the alpha(1) subunit can be deleted without loss of sensitivity to nitric oxide (NO) or YC-1 or loss of heme binding of the respective enzyme complex with the beta(1) subunit. This is in contrast to previous data indicating that NO sensitivity and a functional heme binding site requires full-length amino termini of bovine alpha(1) and beta(1) subunits. Further deletion of the first 364 amino acids of the alpha(1) subunit leads to an enzyme complex with preserved heme binding but loss of sensitivity to NO or YC-1 despite induction of the typical spectral shift by NO binding to the prosthetic heme group. We conclude that 1) the amino-terminal part of the alpha(1) subunit is not involved in heme binding and 2) amino acids 259-364 of the alpha(1) subunit represent an important functional domain for the transduction of the NO activation signal and likely represent the target for NO-sensitizing substances like YC-1.     

Poxvirus mRNA cap methyltransferase. Bypass of the requirement for the stimulatory subunit by mutations in the catalytic subunit and evidence for intersubunit allostery

The guanine-N7 methyltransferase domain of vaccinia virus mRNA capping enzyme is a heterodimer composed of a catalytic subunit vD1-(540-844) and a stimulatory subunit vD12. The poxvirus enzyme can function in vivo in Saccharomyces cerevisiae in lieu of the essential cellular cap methyltransferase Abd1. Coexpression of both poxvirus subunits is required to complement the growth of abd1delta cells. We performed a genetic screen for mutations in the catalytic subunit that bypassed the requirement for the stimulatory subunit in vivo. We thereby identified missense changes in vicinal residues Tyr-752 (to Ser, Cys, or His) and Asn-753 (to Ile), which are located in the cap guanine-binding pocket. Biochemical experiments illuminated a mechanism of intersubunit allostery, whereby the vD12 subunit enhances the affinity of the catalytic subunit for AdoMet and the cap guanine methyl acceptor by 6- and 14-fold, respectively, and increases kcat by a factor of 4. The bypass mutations elicited gains of function in both vD12-independent and vD12-dependent catalysis of cap methylation in vitro when compared with wild-type vD1-(540-844). These results highlight the power of yeast as a surrogate model for the genetic analysis of interacting poxvirus proteins and demonstrate that the activity of an RNA processing enzyme can be augmented through selection and protein engineering.     

The carboxyl terminus of the epsilon subunit of the chloroplast ATP synthase is exposed during illumination

The epsilon subunit of the chloroplast ATP synthase is an inhibitor of activity of the enzyme. Recombinant forms of the epsilon subunit from spinach chloroplasts lacking the last 10, 32, or 45 amino acids were immobilized onto activated Sepharose. A polyclonal antiserum raised against the epsilon subunit was passed over these immobilized protein columns, and the purified antibodies which were not bound recognized the portions of the epsilon subunit missing from the recombinant form present on the column. The full polyclonal antiserum can strip the epsilon subunit from the ATP synthase in illuminated thylakoid membranes [Richter, M. L., and McCarty, R. E. (1987) J. Biol. Chem. 262, 15037-15040]. Exposure of illuminated thylakoid membranes to antibodies recognizing the last 32 amino acids of the epsilon subunit collapses the proton gradient and hinders ATP synthesis with similar efficiency as the full polyclonal preparation. These results indicate that antibodies against the last 32 amino acids of the epsilon subunit are capable of stripping the subunit from the ATP synthase in illuminated membranes. Neither of these effects was seen when the membranes were exposed to the antibodies in the dark. This is direct evidence that the chloroplast ATP synthase undergoes a conformational shift during its activation by the electrochemical proton gradient which specifically alters the conformation of the carboxyl-terminal domain of the epsilon subunit from protected to solvent-exposed. The relation between this shift and activation of the enzyme by the electrochemical proton gradient is discussed.     

Studies on subunit structure and evidence that ligandin is a heterodimer

Several lines of evidence indicate that ligandin consists of two different subunits. The protein dissociates into two components that are detected by electrophoresis in a discontinuous sodium dodecyl sulfate system, or in acid-urea gels, and by isoelectric focusing in the presence of urea. The apparent molecular weights of the two polypeptides are 25,000 and 22,000. Alkylated or succinylated ligandins also exhibit subunit heterogeneity and resolved into two bands in these electrophoretic systems. Cross-linked ligandin showed only one band in sodium dodecyl sulfate-gel electrophoresis indicating that the two subunits are part of a heterodimeric protein rather than monomers of two different proteins. No dansylated terminal amino acids were detected suggesting that the NH2-terminal residues of both chains are blocked. One mole of arginine or phenylalanine was released per mole of ligandin after digestion with carboxypeptidase B or A, respectively. Tryptic maps of succinylated ligandin were consistent with identical disposition of arginine residues in both chains, but several additional tryptic peptides were obtained with native ligandin as compared to the predicted number if both subunits were identical. These observations are consistent with the possibility that both subunits contain common sequences and that a small peptide of about 25 to 30 amino acid residues is cleaved from the COOH-terminal of the larger subunit to produce the smaller subunit.     

Identification of the bovine gamma-aminobutyric acid type A receptor alpha subunit residues photolabeled by the imidazobenzodiazepine [3H]Ro15-4513

Ligands binding to the benzodiazepine-binding site in gamma-aminobutyric acid type A (GABA(A)) receptors may allosterically modulate function. Depending upon the ligand, the coupling can either be positive (flunitrazepam), negative (Ro15-4513), or neutral (flumazenil). Specific amino acid determinants of benzodiazepine binding affinity and/or allosteric coupling have been identified within GABA(A) receptor alpha and gamma subunits that localize the binding site at the subunit interface. Previous photolabeling studies with [(3)H]flunitrazepam identified a primary site of incorporation at alpha(1)His-102, whereas studies with [(3)H]Ro15-4513 suggested incorporation into the alpha(1) subunit at unidentified amino acids C-terminal to alpha(1)His-102. To determine the site(s) of photoincorporation by Ro15-4513, we affinity-purified ( approximately 200-fold) GABA(A) receptor from detergent extracts of bovine cortex, photolabeled it with [(3)H]Ro15-4513, and identified (3)H-labeled amino acids by N-terminal sequence analysis of subunit fragments generated by sequential digestions with a panel of proteases. The patterns of (3)H release seen after each digestion of the labeled fragments determined the number of amino acids between the cleavage site and labeled residue, and the use of sequential proteolytic fragmentation identified patterns of cleavage sites unique to the different alpha subunits. Based upon this radiochemical sequence analysis, [(3)H]Ro15-4513 was found to selectively label the homologous tyrosines alpha(1)Tyr-210, alpha(2)Tyr-209, and alpha(3)Tyr-234, in GABA(A) receptors containing those subunits. These results are discussed in terms of a homology model of the benzodiazepine-binding site based on the molluscan acetylcholine-binding protein structure.     

H(+)-ATPase gamma subunit of Escherichia coli. Role of the conserved carboxyl-terminal region

Cloned uncG genes (wild-type or in vitro mutagenized) for the Escherichia coli gamma subunit were introduced into the uncG mutant Gln-14----end), and the functions of the mutant subunits were studied. The F1's with Ala-283----end and Thr-277----end mutant gamma subunits had 63 and 14% of the wild-type ATPase activity, respectively, and mutants with these subunits showed reduced growth by oxidative phosphorylation, indicating that the 10 residues at the carboxyl terminus (286th residue) are important, but dispensable, for catalysis. On the other hand, F1 with a Gln-269----end gamma subunit was inactive. Replacement of conserved residues (Gln-269, Thr-273, or Glu-275) between Gln-269 and Leu-276 gave enzymes with significantly reduced ATPase activity (2-41% of that of the wild-type) and lower ATP-driven proton conduction. Thus these residues are required for the normal catalytic activity of F1, although they are not absolutely essential. Membranes with amino acid replacements (Thr-277----end, Gln-269----Leu, or Glu-275----Lys) and the frameshift mutation (downstream of Thr-277) had about 15% of the wild-type ATPase activity, but showed different degrees of ATP-dependent H+ translocation and growth yield by oxidative phosphorylation, suggesting that the gamma subunit, especially its carboxyl-terminal region, functions in coupling between catalysis and H+ translocation.     

Functionality of the seventh and eighth transmembrane domains of acyl-coenzyme A:cholesterol acyltransferase 1

Acyl-coenzyme A:cholesterol acyltransferase 1 (ACAT1) is a resident enzyme in the endoplasmic reticulum. ACAT1 is a homotetrameric protein and contains nine transmembrane domains (TMDs). His460 is a key active residue and is located within TMD7. Human ACAT1 has seven free Cys, but the recombinant ACAT1 devoid of free Cys retains full enzyme activity. To further probe the functionality of TMD7 (amino acids 446-460) and TMD8 (amino acids 466-481), we used a parental ACAT1 devoid of free Cys as the template to perform Cys-scanning mutagenesis within these regions. Each of the single Cys mutants was expressed in Chinese hamster ovary (CHO) cell line AC29 lacking endogenous ACAT1. We measured the effect of single Cys substitution on enzyme activity and used the Cu(1,10-phenanthroline)2SO4-mediated disulfide cross-linking method to probe possible interactions of engineered Cys between the two identical subunits. The results show that several residues in one subunit closely interact with the same residues in the other subunit; mutating these residues to Cys does not lead to large loss in enzyme activity. Helical wheel analysis suggests that these residues are located at one side of the coil. In contrast, mutating residues F453, A457, or H460 to Cys causes large loss in enzyme activity; the latter residues are located at the opposite side of the coil. A similar arrangement is found for residues in TMD8. Thus, helical coils in TMD7 and TMD8 have two distinct functional sides: one side is involved in substrate-binding/catalysis, while the other side is involved in subunit interaction.     

Mutations in two distinct regions of acetolactate synthase regulatory subunit from Streptomyces cinnamonensis result in the lack of sensitivity to end-product inhibition

Acetolactate synthase small subunit encoding ilvN genes from the parental Streptomyces cinnamonensis strain and mutants resistant either to valine analogues or to 2-ketobutyrate were cloned and sequenced. The wild-type IlvN from S. cinnamonensis is composed of 175 amino acid residues and shows a high degree of similarity with the small subunits of other valine-sensitive bacterial acetolactate synthases. Changes in the sequence of ilvN conferring the insensitivity to valine in mutant strains were found in two distinct regions. Certain point mutations were located in the conserved domain near the N terminus, while others resulting in the same phenotype shortened the protein at V(104) or V(107). To confirm whether the described mutations were responsible for the changed biochemical properties of the native enzyme, the wild-type large subunit and the wild-type and mutant forms of the small one were expressed separately in E. coli and combined in vitro to reconstitute the active enzyme.     

Deletion study of DNA topoisomerase IB from Leishmania donovani: searching for a minimal functional heterodimer

The substantial differences between trypanosomal and leishmanial DNA topoisomerase IB concerning to their homologues in mammals have provided a new lead in the study of the structural determinants that can be effectively targeted. Leishmania donovani, the causative agent of visceral leishmaniasis, contains an unusual heterodimeric DNA topoisomerase IB. The catalytically active enzyme consists of a large subunit (LdTopIL), which contains the non-conserved N-terminal end and the phylogenetically conserved "core" domain, and of a small subunit (LdTopIS) which harbors the C-terminal region with the characteristic tyrosine residue in the active site. Heterologous co-expression of LdTopIL and LdTopIS genes in a topoisomerase I deficient yeast strain, reconstitutes a fully functional enzyme LdTopIL/S which can be used for structural studies. An approach by combinatorial cloning of deleted genes encoding for truncated versions of both subunits was used in order to find out structural insights involved in enzyme activity or protein-protein interaction. The role played by the non-conserved N-terminal extension of LdTopIL in both relaxation activity and CPT sensitivity has been examined co-expressing the full-length LdTopIS and a fully active LdTopIDeltaS deletion with several deletions of LdTopIL lacking growing sequences of the N-terminal end. The sequential deletion study shows that the first 26 amino acids placed at the N-terminal end and a variable region comprised between Ala548 to end of the C-terminal extension of LdTopIL were enzymatically dispensable. Altogether this combinatorial approach provides important structural insights of the regions involved in relaxation activity and for understanding the atypical structure of this heterodimeric enzyme.     

Carboxyl-terminal sequences critical for inositol 1,4,5-trisphosphate receptor subunit assembly

The inositol 1,4,5-trisphosphate receptor (InsP(3)R) is a tetrameric assembly of conserved subunits that each contains six transmembrane regions (TMRs) localized near the carboxyl terminus. Receptor subunit assembly into a tetramer appears to be a multideterminant process involving an additive contribution of membrane spanning helices and the short cytosolic carboxyl terminus (residues 2590-2749). Previous studies have shown that of the six membrane-spanning regions in each subunit, the 5th and 6th transmembrane regions, and the carboxyl terminus are strong determinants for assembly. The fifth and sixth TMRs contain numerous beta-branched amino acids that may participate in coiled/coil formation via putative leucine zipper motifs. InsP(3)R truncation mutants were expressed in COS-1 cells and analyzed by sucrose density gradient sedimentation and gel filtration for their ability to assemble. Chemical cross-linking with the homobifunctional reagents sDST or DMS of mammalian and bacterially expressed carboxyl-terminal containing receptor fragments reveals that sequences within the carboxyl terminus confer the formation of subunit dimers. A series of InsP(3) receptor carboxyl-terminal fragments and glutathione S-transferase (GST)/InsP(3)R chimeras were expressed in Escherichia coli and used in an in vitro assay to elucidate the minimal sequence responsible for association of the carboxyl termini into dimers. The results presented here indicate that this minimal sequence is approximately 30 residues in length and is localized between residues 2629 and 2654. These residues are highly conserved between the three InsP(3)R isoforms ( approximately 80% identity) as well as the ryanodine receptor ( approximately 40% identity) and suggest that a conserved assembly motif may exist between the two intracellular receptor families. We propose that assembly of the InsP(3) receptor to a tetramer involves intersubunit interactions mediated through both the membrane-spanning regions and residues 2629-2654 of the carboxyl terminus possibly through the formation of a dimer of dimers.