Effect of specific mutations in helix α7 of domain I on the stability and crystallization of Cry3A in Bacillus thuringiensis

Insecticidal crystal (Cry) proteins of Bacillus thuringiensis crystallize after synthesis forming large inclusions that stabilize these toxins in the environment after cell lysis until eaten by an insect. Despite the biological importance of crystallization, little is known about the structural elements of Cry molecules that facilitate this process. We identified subdomains that affect Cry3A structure possibly through improper folding by chimeric-scanning mutagenesis, substituting short peptides of a truncated 70-kDa Cry1C molecule that does not crystallize into Cry3A, a wild-type 70-kDa molecule that crystallizes readily. Cry3A consists of three domains that contain five different blocks of conserved amino acids. Domain substitution and mutagenesis within these blocks suggested that the specific structure of block 2, which spans the junction between domains I and II, was important to the relative stability of Cry3A and subsequent crystallization. Amino acid sequences of particular importance to stability in Cry3A block 2 were identified using three substitution mutants, each spanning about a third of this block. One that consisted of Cry1C helix α7 yielded no detectable protein, whereas the other two produced characteristic Cry3A crystals. Specific mutations in this region showed tyrosine 268 was critical to normal stability of Cry3A and subsequent crystallization in that a mutant, 268L, was less stable than wild-type Cry3A and failed to form a characteristic Cry3A crystal. Circular dichroism analysis showed a decrease in this mutant’s α-helicity, indicating the importance of tyrosine 268 to the specific conformation of helix α7 that facilitates stability and normal crystallization.     

The toxicity test and hypothetical model ofBacillus thuringiensis Cry1Aa helix4

Development of targeted biological agents against agricultural insect pests is of prime importance for the elaboration and implementation of integrated pest management strategies that are environment-friendly, respectful of bio-diversity and safer to human health through reduced use of chemical pesticides. A major goal to understand how Bt toxins work is to elucidate the functions of their three domains. Domains II and III are involved in binding specificity and structural integrity, but the function of Domain I remains poorly understood. Using aManduca sexta BBMV (brush border membrane vesicles) system, we analyzed its responses to Cry1Aa 15 single-point mutations with altered Domain I helix 4 residues. Light scattering assay showed that toxicity was almost lost in 3 mutants, and we observed significantly reduced toxicity in other 7 mutants. However, 5 mutants retained wild-type toxicity. Using computer software, we simulated the three-dimensional structures of helix 4. Both experimental and bioinformatic analysis showed that residues in Cry1Aa Domain I helix 4 were involved in the formation of ion channels that is critical for its insect toxicity.     

The toxicity test and hypothetical model ofBacillus thuringiensis Cry1Aa helix4

Development of targeted biological agents against agricultural insect pests is of prime importance for the elaboration and implementation of integrated pest management strategies that are environment-friendly, respectful of bio-diversity and safer to human health through reduced use of chemical pesticides. A major goal to understand how Bt toxins work is to elucidate the functions of their three domains. Domains II and III are involved in binding specificity and structural integrity, but the function of Domain I remains poorly understood. Using a Manduca sexta BBMV (brush border membrane vesicles) system, we analyzed its responses to Cry1Aa 15 single-point mutations with altered Domain I helix 4 residues. Light scattering assay showed that toxicity was almost lost in 3 mutants, and we observed significantly reduced toxicity in other 7 mutants. However, 5 mutants retained wild-type toxicity. Using computer software, we simulated the three-dimensional structures of helix 4. Both experimental and bioinformatic analysis showed that residues in Cry1Aa Domain I helix 4 were involved in the formation of ion channels that is critical for its insect toxicity.     

The toxicity test and hypothetical model of Bacillus thuringiensis Cry1Aa helix4

Development of targeted biological agents against agricultural insect pests is of prime importance for the elaboration and implementation of integrated pest management strategies that are environment-friendly, respectful of bio-diversity and safer to human health through reduced use of chemical pesticides. A major goal to understand how Bt toxins work is to elucidate the functions of their three domains. Domains II and III are involved in binding specificity and structural integrity, but the function of Domain I remains poorly understood. Using a Manduca sexta BBMV (brush border membrane vesicles) system, we analyzed its responses to Cry1Aa 15 single-point mutations with altered Domain I helix 4 residues. Light scattering assay showed that toxicity was almost lost in 3 mutants, and we observed significantly reduced toxicity in other 7 mutants. However, 5 mutants retained wild-type toxicity. Using computer software, we simulated the three-dimensional structures of helix 4. Both experimental and bioinform     

Interaction between functional domains of Bacillus thuringiensis insecticidal crystal proteins.

Interactions among the three structural domains of Bacillus thuringiensis Cry1 toxins were investigated by functional analysis of chimeric proteins. Hybrid genes were prepared by exchanging the regions coding for either domain I or domain III among Cry1Ab, Cry1Ac, Cry1C, and Cry1E. The activity of the purified trypsin-activated chimeric toxins was evaluated by testing their effects on the viability and plasma membrane permeability of Sf9 cells. Among the parental toxins, only Cry1C was active against these cells and only chimeras possessing domain II from Cry1C were functional. Combination of domain I from Cry1E with domains II and III from Cry1C, however, resulted in an inactive toxin, indicating that domain II from an active toxin is necessary, but not sufficient, for activity. Pores formed by chimeric toxins in which domain I was from Cry1Ab or Cry1Ac were slightly smaller than those formed by toxins in which domain I was from Cry1C. The properties of the pores formed by the chimeras are therefore likely to result from an interaction between domain I and domain II or III. Domain III appears to modulate the activity of the chimeric toxins: combination of domain III from Cry1Ab with domains I and II of Cry1C gave a protein which was more strongly active than Cry1C.     

Cloning, characterization and expression of a new cry1Ab gene from Bacillus thuringiensisWB9

A new cry 1Ab-type gene, cry 1Ab17, was cloned from Bacillus thuringiensis WB9 by PCR. Nucleotide sequence indicated that the open reading frames (ORFs) consists of 3471 bases and encodes a protein of 1156 residues with a calculated molecular weight of 130.5 kDa and an pI value of 5.04. Homology comparison revealed that the deduced amino acid sequence of Cry1Ab17 had 95.4% to 99.7% identity with those of the known Cry1Ab proteins. The Cry1Ab17 was one residue longer than the known Cry1Ab (except for Cry1Ab2). Domain I (Tyr(33) to Arg(253)), II (Arg(265) to Phe(462)), III (Asn(464) to Thr(610)) of the Cry1Ab17 were 96.8%, 68.2% and 100% identical to the corresponding domains of Cry1Aa. Additionally, the cry 1Ab17 gene was expressed in Escherichia coli BL21 under the control of T7 promoter and the Cry1Ab17 isolated from the culture medium was toxic to 3rd instar Plutella xylostella larvae.     

Biochemical and preliminary crystallographic characterization of the vitamin D sterol- and actin-binding by human vitamin D-binding protein

Vitamin D-binding protein (DBP), a multi-functional serum glycoprotein, has a triple-domain modular structure. Mutation of Trp145 (in Domain I) to Ser decreased 25-OH-D(3)-binding by 80%. Furthermore, recombinant Domain I (1-203) and Domain I + II (1-330) showed specific and strong binding for 25-OH-D(3), but Domain III (375-427) did not, suggesting that only Domains I and II might be required for vitamin D sterol-binding. Past studies have suggested that Domain III is independently capable of binding G-actin. We exploited this apparently independent ligand-binding property of DBP to purify DBP-actin complex from human serum and rabbit muscle actin by 25-OH-D(3) affinity chromatography. Competitive (3)H-25-OH-D(3) binding curves for native DBP and DBP-actin complex were almost identical, further suggesting that vitamin D sterol- and actin-binding activities by DBP might be largely independent of each other. Trypsin treatment of DBP produced a prominent 25 kDa band (Domain I, minus 5 amino acids in N-terminus), while actin was completely fragmented by such treatment. In contrast, tryptic digestion of purified DBP-actin complex showed two prominent bands, 52 (DBP, minus 5 amino acids in the N-terminus) and 34 kDa (actin, starting with amino acid position 69) indicating that DBP, particularly its Domains II and III were protected from trypsin cleavage upon actin-binding. Similarly, actin, except its N-terminus, was also protected from tryptic digestion when complexed with DBP. These results provided the basis for our studies to crystallize DBP-actin complex, which produced a 2.5 A crystal, primitive orthorhombic with unit cell dimensions a=80.2A, b=87.3A, and c=159.6A, P2(1)2(1)2(1) space group, V(m)=2.9. Soaking of crystals of actin-DBP in crystallization buffer containing various concentrations of 25-OH-D(3) resulted in cracking of the crystal, which was probably a reflection of a ligand-induced conformational change in the complex, disrupting crystal contacts. In conclusion, we have provided data to suggest that although binding of 25-OH-D(3) to DBP might result in discrete conformational changes in the holo-protein to influence actin-binding, these binding processes are largely independent of each other in solution.     

The structure of the exopolyphosphatase (PPX) from Escherichia coli O157:H7 suggests a binding mode for long polyphosphate chains

Polyphosphate (polyP) is a linear polymer consisting of tens to hundreds of phosphate molecules joined together by high-energy anhydride bonds. These polymers are found in virtually all prokaryotic and eukaryotic cells and perform many functions; prominent among them are the responses to many stresses. Polyphosphate is synthesized by polyP kinase (PPK), using the terminal phosphate of ATP as the substrate, and degraded to inorganic phosphate by both endo- and exopolyphosphatases. Here we report the crystal structure and analysis of the polyphosphate phosphatase PPX from Escherichia coli O157:H7 refined at 2.2 Angstroms resolution. PPX is made of four domains. Domains I and II display structural similarity with one another and share the ribonuclease-H-like fold. Domain III bears structural similarity to the N-terminal, HD domain of SpoT. Domain IV, the smallest domain, has structural counterparts in cold-shock associated RNA-binding proteins but is of unknown function in PPX. The putative PPX active site is located at the interface between domains I and II. In the crystal structure of PPX these two domains are close together and represent the "closed" state. Comparison with the crystal structure of PPX/GPPA from Aquifex aeolicus reveals close structural similarity between domains I and II of the two enzymes, with the PPX/GPPA representing an "open" state. A striking feature of the dimer is a deep S-shaped canyon extending along the dimer interface and lined with positively charged residues. The active site region opens to this canyon. We postulate that this is a likely site of polyP binding.     

Intramolecular proteolytic nicking and binding of Bacillus thuringiensis Cry8Da toxin in BBMVs of Japanese beetle

Bacillus thuringiensis (Bt) Cry8D insecticidal proteins are unique among Cry8 family proteins in terms of its insecticidal activity against adult Scarab beetles, such as Japanese beetle (Popillia japonica Newman). From the sequence homology with other Bt Cry proteins especially those active against beetles, such as Cry3Aa whose 3D structure is available, the structure of the Cry8D protein has been predicted to be a typical three-domain Cry protein type. In addition, the activation process of Cry8D in gut juice of susceptible insects is presumed to be similar to that of Cry3A (Yamaguchi et al., 2008). In this study, the activation process of Cry8Da in insect gut juice was closely examined. Japanese beetle gut juice proteases digested the 130 kDa Cry8Da protein to produce a 64 kDa protein. This 64 kDa protein was active against both adult and larval Japanese beetle and considered to be an activated toxin. N-terminal sequencing of this 64 kDa protein revealed that the Cry8Da leader sequence consisting of 63 amino acid residues from M¹ to F⁶³ was removed. As in the case of Cry3Aa, the proteases further digested the 64 kDa protein to two 8 kDa and 54 kDa fragments. N-terminal amino acid analysis of these smaller fragments indicated that the proteases digested the loop between Alpha Helix (Alpha for short) 3 and Alpha 4. This means that the 8 kDa fragment consists of Alpha 1-3 of Domain I and that the 54 kDa fragment contains the remaining Domain I and full Domain II and Domain III. Size exclusion chromatography and anion exchange chromatography could not separate these 64, 54 and 8 kDa proteins suggesting that the 54 kDa and 8 kDa fragments are still forming the toxin complex equivalent to the 64 kDa protein by size and ionic charge. The sequencing and chromatography results suggest that the gut juice proteases merely nicked the loop between Alpha 3 and Alpha 4. This nicking process appeared to be essential for receptor binding of the Cry8Da toxin. BBMV binding assay revealed that the Cry8Da toxin bound to BBMV preparations from both adult and larval Japanese beetle only after the loop was nicked. Only the 54 kDa fragment bound to the BBMV preparations but not the 64 kDa protein. Ligand blot showed that the protease activated Cry8Da toxin, presumably the 54 kDa fragment, bound to specific BBMV proteins, one or more of those would be receptor(s). The sizes and binding affinities of these Cry8Da-bound proteins of Japanese beetle BBMV differed between larvae and adults.     

Contribution of the 65-kilodalton protein encoded by the cloned gene cry19A to the mosquitocidal activity of Bacillus thuringiensis subsp. jegathesan.

Two new crystal protein genes, cry19A and orf2, isolated from Bacillus thuringiensis subsp. jegathesan were cloned and characterized. The cry19A gene encodes a 74.7-kDa protein, and the orf2 gene encodes a 60-kDa protein. Cry19A contains the five conserved blocks present in most B. thuringiensis delta-endotoxins. The ORF2 amino acid sequence is similar to that of the carboxy terminus of Cry4 proteins. The cry 19A gene was expressed independently or in combination with orf2 in a crystal-negative B. thuringiensis host. The proteins accumulated as inclusions. Purified inclusions containing either Cry19A alone or Cry19A and ORF2 together were toxic to Anopheles stephensi and Culex pipiens mosquito larvae. They were more toxic to C. pipiens than to A. stephensi. However, inclusions containing Cry19A and ORF2 together were more toxic than inclusions of Cry19A alone but less toxic than the wild-type inclusions of B. thuringiensis subsp. jegathesan.     

Three-dimensional structure of the ribosomal translocase: elongation factor G from Thermus thermophilus.

The crystal structure of Thermus thermophilus elongation factor G without guanine nucleotide was determined to 2.85 A. This GTPase has five domains with overall dimensions of 50 x 60 x 118 A. The GTP binding domain has a core common to other GTPases with a unique subdomain which probably functions as an intrinsic nucleotide exchange factor. Domains I and II are homologous to elongation factor Tu and their arrangement, both with and without GDP, is more similar to elongation factor Tu in complex with a GTP analogue than with GDP. Domains III and V show structural similarities to ribosomal proteins. Domain IV protrudes from the main body of the protein and has an extraordinary topology with a left-handed cross-over connection between two parallel beta-strands.     

Role of Bacillus thuringiensis Toxin Domains in Toxicity and Receptor Binding in the Diamondback Moth

The toxic fragment of Bacillus thuringiensis crystal proteins consists of three distinct structural domains. There is evidence that domain I is involved in pore formation and that domain II is involved in receptor binding and specificity. It has been found that, in some cases, domain III is also important in determining specificity. Furthermore, involvement of domain III in binding has also been reported recently. To investigate the role of toxin domains in the diamondback moth (Plutella xylostella), we used hybrid toxins with domain III substitutions among Cry1C, Cry1E, and Cry1Ab. Neither Cry1E nor G27 (a hybrid with domains I and II from Cry1E and domain III from Cry1C) was toxic, whereas Cry1C and F26 (the reciprocal hybrid) were equally toxic. H04 (a hybrid with domains I and II from Cry1Ab and domain III from Cry1C) showed toxicity that was of a similar level as that of Cry1Ab and significantly higher than that of Cry1C. Binding assays with ¹²⁵I-Cry1C showed that Cry1C and F26 competed for the same binding sites on midgut membrane vesicles, whereas Cry1E, G27, and H04 did not bind to these sites. Our results show that, in contrast to findings in other insects for the toxins and hybrids used here, toxin specificity as well as specificity of binding to membrane vesicles in the diamondback moth is mediated by domain II (and/or I) and not by domain III.     

Full-length sequence of the cDNA for human erythroid beta-spectrin

Spectrin is the major molecular consituent of the red cell membrane skeleton. We have isolated overlapping human erythroid beta-spectrin cDNA clones and determined 6773 base pairs of contiguous nucleotide sequence. This includes the entire coding sequence of beta-spectrin. The sequence translates into a 2137 amino acid, 246-kDa peptide. beta-Spectrin is found to consist of three distinct domains. Domain I, at the N terminus, is a 272-amino acid region lacking resemblance to the spectrin repetitive motif. Sequences in this region exhibit striking sequence homology, at both nucleotide and amino acid levels, to the N-terminal "actin-binding" domains of alpha-actinin and dystrophin. Between residues 51 and 270 there is 55% amino acid identity to human dystrophin, with only four single amino acid gaps in alignment. Domain II consists of 17 spectrin repeats. Several sequence variations are observed in typical repeat structure. Homology to alpha-actinin extends beyond domain I into the N-terminal portion of domain II. Domain III, 52 amino acid residues at the C terminus, does not adhere to the spectrin repeat motif. Combining knowledge of spectrin primary structure with previously reported functional studies, it is possible to make several inferences regarding structure/function relationships within the beta-spectrin molecule.     

Interaction between Functional Domains of Bacillus thuringiensis Insecticidal Crystal Proteins

Interactions among the three structural domains of Bacillus thuringiensis Cry1 toxins were investigated by functional analysis of chimeric proteins. Hybrid genes were prepared by exchanging the regions coding for either domain I or domain III among Cry1Ab, Cry1Ac, Cry1C, and Cry1E. The activity of the purified trypsin-activated chimeric toxins was evaluated by testing their effects on the viability and plasma membrane permeability of Sf9 cells. Among the parental toxins, only Cry1C was active against these cells and only chimeras possessing domain II from Cry1C were functional. Combination of domain I from Cry1E with domains II and III from Cry1C, however, resulted in an inactive toxin, indicating that domain II from an active toxin is necessary, but not sufficient, for activity. Pores formed by chimeric toxins in which domain I was from Cry1Ab or Cry1Ac were slightly smaller than those formed by toxins in which domain I was from Cry1C. The properties of the pores formed by the chimeras are therefore likely to result from an interaction between domain I and domain II or III. Domain III appears to modulate the activity of the chimeric toxins: combination of domain III from Cry1Ab with domains I and II of Cry1C gave a protein which was more strongly active than Cry1C.     

Binding characteristics to mosquito-larval midgut proteins of the cloned domain II-III fragment from the Bacillus thuringiensis Cry4Ba toxin

Receptor binding plays an important role in determining host specificity of the Bacillus thuringiensis Cry delta-endotoxins. Mutations in domains II and III have suggested the participation of certain residues in receptor recognition and insect specificity. In the present study, we expressed the cloned domain II-III fragment of Cry4Ba and examined its binding characteristics to mosquito-larval midgut proteins. The 43-kDa Cry4Ba-domain II-III protein over-expressed in Escherichia coli as inclusion bodies was only soluble when carbonate buffer, pH 10.0 was supplemented with 4 M urea. After renaturation via stepwise dialysis and subsequent purification, the refolded domain II-III protein, which specifically reacts with anti Cry4Ba-domain III monoclonal antibody, predominantly exists as a beta-sheet structure determined by circular dichroism spectroscopy. In vitro binding analysis to both histological midgut tissue sections and brush border membrane proteins prepared from susceptible Aedes aegypti mosquito-larvae revealed that the isolated Cry4Ba-domain II-III protein showed binding functionality comparable to the 65-kDa full-length active toxin. Altogether, the data present the 43-kDa Cry4Ba fragment comprising domains II and III that was produced in isolation was able to retain its receptor-binding characteristics to the target larval midgut proteins.     

Bacillus thuringiensis Cry4A and Cry4B mosquito-larvicidal proteins: homology-based 3D model and implications for toxin activity

Three-dimensional (3D) models for the 65-kDa activated Cry4A and Cry4B delta-endotoxins from Bacillus thuringiensis subsp. israelensis that are specifically toxic to mosquito-larvae were constructed by homology modeling, based on atomic coordinates of the Cry1Aa and Cry3Aa crystal structures. They were structurally similar to the known structures, both derived 3D models displayed a three-domain organization: the N-terminal domain (I) is a seven-helix bundle, while the middle and C-terminal domains are primarily comprise of anti-parallel beta-sheets. Circular dichroism spectroscopy confirmed the secondary structural contents of the two homology-based Cry4 structures. A structural analysis of both Cry4 models revealed the following: (a) Residues Arg-235 and Arg-203 are located in the interhelical 5/6 loop within the domain I of Cry4A and Cry4B, respectively. Both are solvent exposed. This suggests that they are susceptible to tryptic cleavage. (b) The unique disulphide bond, together with a proline-rich region within the long loop connecting alpha4 and alpha5 of Cry4A, were identified. This implies their functional significance for membrane insertion. (c) Significant structural differences between both models were found within domain II that may reflect their different activity spectra. Structural insights from this molecular modeling study would therefore increase our understanding of the mechanic aspects of these two closely related mosquito-larvicidal proteins.     

Shared Binding Sites for the Bacillus thuringiensis Proteins Cry3Bb,Cry3Ca,and Cry7Aa in the African Sweet Potato Pest Cylas puncticollis (Brentidae)

Bacillus thuringiensis Cry3Bb, Cry3Ca, and Cry7Aa have been reported to be toxic against larvae of the genus Cylas, which are important pests of sweet potato worldwide and particularly in sub-Saharan Africa. However, relatively little is known about the processing and binding interactions of these coleopteran-specific Cry proteins. The aim of the present study was to determine whether Cry3Bb, Cry3Ca, and Cry7Aa proteins have shared binding sites in Cylas puncticollis to orient the pest resistance strategy by genetic transformation. Interestingly, processing of the 129-kDa Cry7Aa protoxin using commercial trypsin or chymotrypsin rendered two fragments of about 70 kDa and 65 kDa. N-terminal sequencing of the trypsin-activated Cry7Aa fragments revealed that processing occurs at Glu⁴⁷ for the 70-kDa form or Ile⁸⁸ for the 65-kDa form. Homologous binding assays showed specific binding of the two Cry3 proteins and the 65-kDa Cry7Aa fragment to brush border membrane vesicles (BBMV) from C. puncticollis larvae. The 70-kDa fragment did not bind to BBMV. Heterologous-competition assays showed that Cry3Bb, Cry3Ca, and Cry7Aa (65-kDa fragment) competed for the same binding sites. Hence, our results suggest that pest resistance mediated by the alteration of a shared Cry receptor binding site might render all three Cry toxins ineffective.     

Trix, a novel Rac guanine-nucleotide exchange factor from Dictyostelium discoideum is an actin-binding protein and accumulates at endosomes

Small Rho family GTPases are involved in regulation of actin cytoskeleton dynamics. These molecular switches are themselves mainly controlled by specific GTPase-activating proteins (GAPs) and guanine-nucleotide exchange factors (GEFs). We have cloned and initially characterized a novel putative RhoGEF from Dictyostelium discoideum. The predicted 135-kDa protein displays a unique domain organization in its N-terminus by harboring two type3 calponin homology (CH) domains followed by a single type1 CH domain. The C-terminal region encompasses a diffuse B-cell lymphoma homology/pleckstrin homology tandem domain that is typically found in RhoGEFs. We therefore refer to this protein as Trix (triple CH-domain array exchange factor). A recombinant N-terminal region of Trix carrying all three CH domains binds to F-actin and bundles actin filaments. Trix-null mutants are viable and display only subtle defects when compared to wild-type cells with the exception of a substantial decrease in exocytosis of a fluid-phase marker. GFP fusions with the full-length protein or the N-terminal part containing all three CH domains revealed that Trix localizes to the cortical region and strongly accumulates on late endosomes. Our results suggest that Trix is specifically involved in a Rho GTPase-signaling pathway that is required for regulation of the actin cytoskeleton during exocytosis.     

The architecture of the animal fatty acid synthetase complex. IV. Mapping of active centers and model for the mechanism of action

The fatty acid synthetase of animal tissue consists of two subunits, each containing seven catalytic centers and an acyl carrier site. Proteolytic cleavage patterns indicate that the subunit is arranged into three major domains, I, II, and III. Domain I contains the NH2-terminal end of the polypeptide and the catalytic sites of beta-ketoacyl synthetase (condensing enzyme) and the acetyl-and malonyl-transacylases. This domain, therefore, functions as a site for acetyl and malonyl substrate entry into the process of fatty acid synthesis and acts in part as the site of carbon-carbon condensation, resulting in chain elongation. Domain II is the medial domain and contains the beta-ketoacyl and enoyl reductases, probably the dehydratase, and the 4'-phosphopantetheine prosthetic group of the acyl carrier protein site. Domain II, therefore, is designated as the reduction domain where the keto carbon is reduced to methylene carbon by sequential processes of reduction, dehydration, and reduction again. Throughout these processes, the acyl group is attached to the pantetheine-SH of the acyl carrier protein. The latter site is distal to the cysteine-SH of the beta-ketoacyl synthetase, constitutes the 15000-dalton polypeptide at the COOH-terminal end of Domain II, and connects to Domain III. When the growing chain reaches C16 carbon length, the fatty acyl group is released by the thioesterase activity, which is contained in Domain III. A functional model is proposed based on the aforementioned results and the recent evidence that the synthetase subunits are arranged in a head-to-tail fashion, such that the pantetheine-SH of the acyl carrier protein of one subunit and the cysteine-SH of the beta-ketoacyl synthetase of the second subunit are juxtaposed. In this model, a palmitate synthesizing site contains Domain I of one subunit and Domains II and III of the second subunit. Therefore, even though each subunit contains all of the partial activities of the reaction sequence, the actual palmitate synthesizing unit consists of one-half of a subunit interacting with the complementary half of the other subunit.