FosDH1 from module 1 of the fostriecin polyketide synthase (PKS) catalyzes the dehydration of a 3-hydroxybutyryl-SACP to the (E)-3-butenoyl-SACP. The steady-state kinetic parameters, kcat and kcat/Km, were determined over the pH range 3.0 to 9.2 for the FosDH1-catalyzed dehydration of the N-acetycsteamine thioester, 3-hydroxybutyryl-SNAC (3), to (E)-3-butenoyl-SNAC (4). The pH rate profiles for both log(kcat) and log(kcat/Km) each corresponded to a single pH-dependent ionization to give an active site general base, with a calculated pKa 6.1 ± 0.2 for kcat and pKa 5.7 ± 0.1 for kcat/Km. These results are inconsistent with the commonly suggested "two-base" (base-acid) mechanism for the dehydratases of PKS and fatty acid biosynthesis and support a simple one-base mechanism in which the universally conserved active site His residue acts as the base to deprotonate C-2 of the substrate, then redonates the proton to the C-3 hydroxyl group to promote C-O bond-cleavage and elimination of water. The carboxylate of the paired Asp or Glu residue is thought to bind and orient the hydroxyl group of the substrate in the stereoelectonically favored conformation.

Incubation of (±)-2-methyl-3-ketobutyryl-SNAC (3) and (±)-2-methyl-3-ketopentanoyl-SNAC (4) with BonKR2 or OxaKR5, ketoreductase domains from the bongkrekic acid (1) and oxazolomycin (2) polyketide synthases, in the presence of NADPH gave in each case the corresponding (2 R,3 S)-2-methyl-3-hydroxybutyryl-SNAC (5) or (2 R,3 S)-2-methyl-3-hydroxypentanoyl-SNAC (6) products, as established by chiral gas chromatography-mass spectrometry analysis of the derived methyl esters. Identical results were obtained by BonKR2- and OxaKR5-catalyzed reduction of chemoenzymatically prepared (2 R)-2-methyl-3-ketopentanoyl-EryACP6, (2 R)-2-methyl-3-ketobutyryl-BonACP2 (12), and (2 R)-2-methyl-3-ketopentanoyl-BonACP2 (13). The paired dehydratase domains, BonDH2 and OxaDH5, were then shown to catalyze the reversible syn dehydration of (2 R,3 S)-2-methyl-3-hydroxybutyryl-BonACP2 (14) to give the corresponding trisubstituted ( Z)-2-methylbutenoyl-BonACP2 (16).

Catalytic modules of assembly-line polyketide synthases (PKSs) have previously been observed in two very different conformations-an "extended" architecture and an "arch-shaped" architecture-although the catalytic relevance of neither has been directly established. By the use of a fully human naïve antigen-binding fragment (Fab) library, a high-affinity antibody was identified that bound to the extended conformation of a PKS module, as verified by X-ray crystallography and tandem size-exclusion chromatography-small-angle X-ray scattering (SEC-SAXS). Kinetic analysis proved that this antibody-stabilized module conformation was fully competent for catalysis of intermodular polyketide chain translocation as well as intramodular polyketide chain elongation and functional group modification of a growing polyketide chain. Thus, the extended conformation of a PKS module is fully competent for all of its essential catalytic functions.

S-Adenosyl methionine (SAM)-dependent C-methyltransferases are responsible for the C2-methylation of 3-ketoacyl-acyl carrier protein (ACP) intermediates to give the corresponding 2-methy-3-ketoacyl-ACP products during bacterial polyketide biosynthesis mediated by trans-AT polyketide synthases that lack integrated acyl transferase (AT) domains. A coupled ketoreductase (KR) assay was used to assign the stereochemistry of the C-methyltransferase-catalyzed reaction. Samples of chemoenzymatically generated 3-ketopentanoyl-ACP (9) were incubated with SAM and BonMT2 from module 2 of the bongkrekic acid polyketide synthase. The resulting 2-methyl-3-ketopentanoyl-ACP (10) was incubated separately with five (2R)- or (2S)-methyl specific KR domains. Analysis of the derived 2-methyl-3-hydroxypentanoate methyl esters (8) by chiral GC-MS established that the BonMT2-catalyzed methylation generated exclusively (2R)-2-methyl-3-ketopentanoyl-ACP ((2R)-10). Identical results were also obtained with three additional C-methyltransferases-BaeMT9, DifMT1, and MupMT1-from the bacillaene, difficidin, and mupirocin trans-AT polyketide synthases.

The dehydratase domain FosDH1 from module 1 of the fostriecin polyketide synthase (PKS) catalyzed the stereospecific interconversion of (3R)-3-hydroxybutyryl-FosACP1 (5) and (E)-2-butenoyl-FosACP1 (11), as established by a combination of direct LC-MS/MS and chiral GC-MS. FosDH1 did not act on either (3S)-3-hydroxybutyryl-FosACP1 (6) or (Z)-2-butenoyl-FosACP1 (12). FosKR2, the ketoreductase from module 2 of the fostriecin PKS that normally provides the natural substrate for FosDH2, was shown to catalyze the NADPH-dependent stereospecific reduction of 3-ketobutyryl-FosACP2 (23) to (3S)-3-hydroxybutyryl-FosACP2 (8). Consistent with this finding, FosDH2 catalyzed the interconversion of the corresponding triketide substrates (3R,4E)-3-hydroxy-4-hexenoyl-FosACP2 (18) and (2Z,4E)-2,4-hexadienoyl-FosACP2 (21). FosDH2 also catalyzed the stereospecific hydration of (Z)-2-butenoyl-FosACP2 (14) to (3S)-3-hydroxybutyryl-FosACP2 (8). Although incubation of FosDH2 with (3S)-3-hydroxybutyryl-FosACP2 (8) did not result in detectable accumulation of (Z)-2-butenoyl-FosACP2 (14), FosDH2 catalyzed the slow exchange of the 3-hydroxy group of 8 with [18O]-water. FosDH2 unexpectedly could also support the stereospecific interconversion of (3R)-3-hydroxybutyryl-FosACP2 (7) and (E)-2-butenoyl-FosACP2 (13).

The sesquiterpene cyclase epi-isozizaene synthase (EIZS) catalyzes the cyclization of farnesyl diphosphate to form the tricyclic hydrocarbon precursor of the antibiotic albaflavenone. The hydrophobic active site pocket of EIZS serves as a template as it binds and chaperones the flexible substrate and carbocation intermediates through the conformations required for a multistep reaction sequence. We previously demonstrated that the substitution of hydrophobic residues with other hydrophobic residues remolds the template and expands product chemodiversity [Li, R., Chou, W. K. W., Himmelberger, J. A., Litwin, K. M., Harris, G. G., Cane, D. E., and Christianson, D. W. (2014) Biochemistry 53, 1155-1168]. Here, we show that the substitution of hydrophobic residues-specifically, Y69, F95, F96, and W203-with polar side chains also yields functional enzyme catalysts that expand product chemodiversity. Fourteen new EIZS mutants are reported that generate product arrays in which eight new sesquiterpene products have been identified. Of note, some mutants generate acyclic and cyclic hydroxylated products, suggesting that the introduction of polarity in the hydrophobic pocket facilitates the binding of water capable of quenching carbocation intermediates. Furthermore, the substitution of polar residues for F96 yields high-fidelity sesquisabinene synthases. Crystal structures of selected mutants reveal that residues defining the three-dimensional contour of the hydrophobic pocket can be substituted without triggering significant structural changes elsewhere in the active site. Thus, more radical nonpolar-polar amino acid substitutions should be considered when terpenoid cyclase active sites are remolded by mutagenesis with the goal of exploring and expanding product chemodiversity.

The polyketide synthases responsible for the biosynthesis of the polyether antibiotics nanchangmycin (1) and salinomycin (4) harbor a number of redox-inactive ketoreductase (KR0) domains that are implicated in the generation of C2-epimerized (2S)-2-methyl-3-ketoacyl-ACP intermediates. Evidence that the natural substrate for the polyether KR0 domains is, as predicted, a (2R)-2-methyl-3-ketoacyl-ACP intermediate, came from a newly developed coupled ketosynthase (KS)-ketoreductase (KR) assay that established that the decarboxylative condensation of methylmalonyl-CoA with S-propionyl-N-acetylcysteamine catalyzed by the Nan[KS1][AT1] didomain from module 1 of the nanchangmycin synthase generates exclusively the corresponding (2R)-2-methyl-3-ketopentanoyl-ACP (7a) product. In tandem equilibrium isotope exchange experiments, incubation of [2-2H]-(2R,3S)-2-methyl-3-hydroxypentanoyl-ACP (6a) with redox-active, epimerase-inactive EryKR6 from module 6 of the 6-deoxyerythronolide B synthase and catalytic quantities of NADP+ in the presence of redox-inactive, recombinant NanKR10 or NanKR50, from modules 1 and 5 of the nanchangmycin synthase, or recombinant SalKR70 from module 7 of the salinomycin synthase, resulted in first-order, time-dependent washout of deuterium from 6a. Control experiments confirmed that this washout was due to KR0-catalyzed isotope exchange of the reversibly generated, transiently formed oxidation product [2-2H]-(2R)-2-methyl-3-ketopentanoyl-ACP (7a), consistent with the proposed epimerase activity of each of the KR0 domains. Although they belong to the superfamily of short chain dehydrogenase-reductases, the epimerase-active KR0 domains from polyether synthases lack one or both residues of the conserved Tyr-Ser dyad that has previously been implicated in KR-catalyzed epimerizations.

Terpenoid synthases catalyze isoprenoid cyclization reactions underlying the generation of more than 80,000 natural products. Such dramatic chemodiversity belies the fact that these enzymes generally consist of only three domain folds designated as α, β, and γ. Catalysis by class I terpenoid synthases occurs exclusively in the α domain, which is found with α, αα, αβ, and αβγ domain architectures. Here, we explore the influence of domain architecture on catalysis by taxadiene synthase from Taxus brevifolia (TbTS, αβγ), fusicoccadiene synthase from Phomopsis amygdali (PaFS, (αα)6), and ophiobolin F synthase from Aspergillus clavatus (AcOS, αα). We show that the cyclization fidelity and catalytic efficiency of the α domain of TbTS are severely compromised by deletion of the βγ domains; however, retention of the β domain preserves significant cyclization fidelity. In PaFS, we previously demonstrated that one α domain similarly influences catalysis by the other α domain [ Chen , M. , Chou , W. K. W. , Toyomasu , T. , Cane , D. E. , and Christianson , D. W. ( 2016 ) ACS Chem. Biol. 11 , 889 - 899 ]. Here, we show that the hexameric quaternary structure of PaFS enables cluster channeling. We also show that the α domains of PaFS and AcOS can be swapped so as to make functional chimeric αα synthases. Notably, both cyclization fidelity and catalytic efficiency are altered in all chimeric synthases. Twelve newly formed and uncharacterized C20 diterpene products and three C25 sesterterpene products are generated by these chimeras. Thus, engineered αβγ and αα terpenoid cyclases promise to generate chemodiversity in the greater family of terpenoid natural products.

Incubation of synthetic 2-methylneryl diphosphate (2-MeNPP, 10) with 2-methylisoborneol synthase (MIBS) gave a mixture of products that differed significantly from that derived from the natural substrate (E)-2-methylgeranyl diphosphate (3, 2-MeGPP). The proportion of (-)-2-methylisoborneol (1) decreased from 89 to 17% while that of 2-methylenebornane (4) increased from 10 to 26%, with the relative yields of the isomeric homo-monoterpenes 2-methyl-2-bornene (5) and 1-methylcamphene (6) remaining essentially unchanged (<1% each), as determined by chiral GC–MS analysis. The majority of the product mixture resulting from the MIBS-catalyzed cyclization of 2-MeNPP (10) consisted of the anomalous monocyclic homo-monoterpenes (±)-2-methylllimonene (15, 39%) and 2-methyl-α-terpineol (13, 10%), as well as the acylic derivatives 2-methylnerol (11, 7%) and 2-methyllinalool (14, <1%). The steady-state kinetic parameters of the MIBS-catalyzed reaction, determined using [1-3H]-2-methylneryl diphosphate (2-MeNPP), were kcat 0.0046±0.0003 s−1, Km 18±6 μm and kcat/Km 2.55 × 102 M−1 s−1. In comparison, the natural substrate 2-MeGPP had a kcat 0.105±0.007 s−1, Km 95±49 μm and kcat/Km 1.11 × 103 M−1 s−1. Taken together with earlier X-ray crystallographic studies of MIBS, as well as previous investigations of the mechanistically related plant monoterpene cyclase, bornyl diphosphate synthase, these results provide important insights into the binding and cyclization of both native substrates and intermediates and their analogs.

Dehydratase (DH) domains of cryptic function are often found in polyketide synthase (PKS) modules that produce epimerized (2S)-2-methyl-3-ketoacyl-ACP (acyl carrier protein) intermediates. A combination of tandem equilibrium isotope exchange (EIX) and a newly developed Tandem Modules Epimerase assay revealed the intrinsic epimerase activity of NanDH1 and NanDH5, from modules 1 and 5, respectively, of the nanchangmycin (1) PKS as well of NigDH1, from module 1 of the nigericin (3) PKS. Unexpectedly, all three epimerase-active DH domains were also found to possess intrinsic dehydratase activity, whereas the conventional DH domains, EryDH4, from module 4 of the erythromycin synthase, and NanDH2 from module 2 of the nanchangmycin synthase, were shown to have cryptic epimerase activity.

Ketosynthase (KS) domains of assembly line polyketide synthases (PKSs) catalyze intermodular translocation of the growing polyketide chain as well as chain elongation via decarboxylative Claisen condensation. The mechanistic roles of ten conserved residues in the KS domain of Module 1 of the 6-deoxyerythronolide B synthase were interrogated via site-directed mutagenesis and extensive biochemical analysis. Although the C211A mutant at the KS active site exhibited no turnover activity, it was still a competent methylmalonyl-ACP decarboxylase. The H346A mutant exhibited reduced rates of both chain translocation and chain elongation, with a greater effect on the latter half-reaction. H384 contributed to methylmalonyl-ACP decarboxylation, whereas K379 promoted C-C bond formation. S315 played a role in coupling decarboxylation to C-C bond formation. These findings support a mechanism for the translocation and elongation half-reactions that provides a well-defined starting point for further analysis of the key chain-building domain in assembly line PKSs.

The role of the conserved active site tyrosine and serine residues in epimerization catalyzed by polyketide synthase ketoreductase (PKS KR) domains has been investigated. Both mutant and wild-type forms of epimerase-active KR domains, including the intrinsically redox-inactive EryKR3 degrees and PicKR3 degrees as well as redox-inactive mutants of EryKR1, were incubated with [2-H-2]-(2R,3S)-2-methyl-3-hydroxypentanoyl-SACP ([2-H-2]-2) and 0.05 equiv of NADP in the presence of the redox-active, epimerase-inactive EryKR6 domain. The residual epimerase activity of each mutant was determined by tandem equilibrium isotope exchange, in which the first-order, time-dependent washout of isotope from 2 was monitored by liquid chromatography tandem mass spectrometry with quantitation of the deuterium content of the diagnostic pantetheinate ejection fragment (4). Replacement of the active site Tyr or Ser residues, alone or together, significantly reduced the observed epimerase activity of each KR domain with minimal effect on substrate binding. Our results demonstrate that the epimerase and reductase activities of PKS KR domains share a common active site, with both reactions utilizing the same pair of Tyr and Ser residues.

In the biosynthesis of pentalenolactone (1), PenE and PntE, orthologous proteins from Streptomyces exfoliatus and S. arenae, respectively, catalyze the flavin-dependent Baeyer-Villiger oxidation of 1-deoxy-11-oxopentalenic acid (4) to the lactone pentalenolactone D (5), in which the less-substituted methylene carbon has migrated. By contrast, the paralogous PtlE enzyme from S. avermitilis catalyzes the oxidation of 4 to neopentalenolactone D (6), in which the more substituted methane substitution has undergone migration. We report the design and analysis of 13 single and multiple mutants of PntE mutants to identify the key amino acids that contribute to the regiospecificity of these two classes of Baeyer-Villiger monooxygenases. The L185S mutation in PntE reversed the observed regiospecificity of PntE such that all recombinant PntE mutants harboring this L185S mutation acquired the characteristic regiospecificity of PtlE, catalyzing the conversion of 4 to 6 as the major product. The recombinant PntE mutant harboring R484L exhibited reduced regiospecificity, generating a mixture of lactones containing more than 17% of 6. These in vitro results were corroborated by analysis of the complementation of the S. avermitilis ΔptlED double deletion mutant with pntE mutants, such that pntE mutants harboring L185S produced 6 as the major product, whereas complemention of the ΔptlED deletion mutant with pntE mutants carrying the R484L mutation gave 6 as more than 33% of the total lactone product mixture.

Vectorial polyketide biosynthesis on an assembly line polyketide synthase is the most distinctive property of this family of biological machines, while providing the key conceptual tool for the bioinformatic decoding of new antibiotic pathways. We now show that the action of the entire assembly line is synchronized by a previously unrecognized turnstile mechanism that prevents the ketosynthase domain of each module from being acylated by a new polyketide chain until the product of the prior catalytic cycle has been passed to the downstream module from the corresponding acyl carrier protein domain. The turnstile is closed by virtue of tight coupling to the signature decarboxylative condensation reaction catalyzed by the ketosynthase domain of each polyketide synthase module. Reopening of the turnstile is coupled to the eventual chain translocation step that vacates the module. At the maximal rate of substrate turnover, one would expect the chain release step to initiate a cascade of chain translocation events that sequentially migrate back upstream, thereby repriming each module and setting up the assembly line for the next round of polyketide chain elongation.

By the time I entered Harvard College in the fall of 1962, I knew, or thought I knew, that I wanted to become a scientist. In fact, I had very little idea what that actually meant, nor did I even know what branch of science to choose. My interest in chemistry was quickly kindled, however, by Prof Leonard K Nash’s introductory course in general chemistry. I was not only smitten by the subject itself but also I was inspired by Prof Nash’s informative and entertaining lectures, and his warm and scintillating personality. Indeed for many years thereafter, Prof Nash continued to be a source of encouragement and advice on both science and the art of teaching.

The final step in the biosynthesis of the sesquiterpenoid antibiotic pentalenolactone (1) is the highly unusual cytochrome P450-catalyzed, oxidative rearrangement of pentalenolactone F (2), involving the transient generation and rearrangement of a neopentyl cation. In Streptomyces arenae this reaction is catalyzed by CYP161C2 (PntM), with highly conserved orthologs being present in at least 10 other Actinomycetes. Crystal structures of substrate-free PntM, as well as PntM with bound substrate 2, product 1, and substrate analogue 6,7-dihydropentalenolactone F (7) revealed interactions of bound ligand with three residues, F232, M77, and M81 that are unique to PntM and its orthologs and absent from essentially all other P450s. Site-directed mutagenesis, ligand-binding measurements, steady-state kinetics, and reaction product profiles established there is no special stabilization of reactive cationic intermediates by these side chains. Reduced substrate analogue 7 did not undergo either oxidative rearrangement or simple hydroxylation, suggesting that the C1 carbocation is not anchimerically stabilized by the 6,7-double bond of 2. The crystal structures also revealed plausible proton relay networks likely involved in the generation of the key characteristic P450 oxidizing species, Compound I, and in mediating stereospecific deprotonation of H-3re of the substrate. We conclude that the unusual carbocation intermediate results from outer shell electron transfer from the transiently generated C1 radical to the tightly paired heme-•Fe(3+)-OH radical species. The oxidative electron transfer is kinetically dominant as a result of the unusually strong steric barrier to oxygen rebound to the neopentyl center C-1si, which is flanked on each neighboring carbon by syn-axial substituents.

Aristolochene synthase (ATAS) is a high-fidelity terpenoid cyclase that converts farnesyl diphosphate exclusively into the bicyclic hydrocarbon aristolochene. Previously determined crystal structures of ATAS complexes revealed trapped active site water molecules that could potentially interact with catalytic intermediates: water "w" hydrogen bonds with S303 and N299, water molecules "w1" and "w2" hydrogen bond with Q151, and a fourth water molecule coordinates to the Mg(2+)C ion. There is no obvious role for water in the ATAS mechanism because the enzyme exclusively generates a hydrocarbon product. Thus, these water molecules are tightly controlled so that they cannot react with carbocation intermediates. Steady-state kinetics and product distribution analyses of eight ATAS mutants designed to perturb interactions with active site water molecules (S303A, S303H, S303D, N299A, N299L, N299A/S303A, Q151H, and Q151E) indicate relatively modest effects on catalysis but significant effects on sesquiterpene product distributions. X-ray crystal structures of S303A, N299A, N299A/S303A, and Q151H mutants reveal minimal perturbation of active site solvent structure. Seven of the eight mutants generate farnesol and nerolidol, possibly resulting from addition of the Mg(2+)C-bound water molecule to the initially formed farnesyl cation, but no products are generated that would suggest enhanced reactivity of other active site water molecules. However, intermediate germacrene A tends to accumulate in these mutants. Thus, apart from the possible reactivity of Mg(2+)C-bound water, active site water molecules in ATAS are not directly involved in the chemistry of catalysis but instead contribute to the template that governs the conformation of the flexible substrate and carbocation intermediates.

Ketoreductases (KRs) are the most widespread tailoring domains found in individual modules of assembly line polyketide synthases (PKSs), and are responsible for controlling the configurations of both the α-methyl and β-hydroxyl stereogenic centers in the growing polyketide chain. Because they recognize substrates that are covalently bound to acyl carrier proteins (ACPs) within the same PKS module, we sought to quantify the extent to which protein–protein recognition contributes to the turnover of these oxidoreductive enzymes using stand-alone domains from the 6-deoxyerythronolide B synthase (DEBS). Reduced 2-methyl-3-hydroxyacyl-ACP substrates derived from two enantiomeric acyl chains and four distinct ACP domains were synthesized and presented to four distinct KR domains. Two KRs, from DEBS modules 2 and 5, displayed little preference for oxidation of substrates tethered to their cognate ACP domains over those attached to the other ACP domains tested. In contrast, the KR from DEBS module 1 showed an ~10–50-fold preference for substrate attached to its native ACP domain, whereas the KR from DEBS module 6 actually displayed an ~10-fold preference for the ACP from DEBS module 5. Our findings suggest that recognition of the ACP by a KR domain is unlikely to affect the rate of native assembly line polyketide biosynthesis. In some cases, however, unfavorable KR–ACP interactions may suppress the rate of substrate processing when KR domains are swapped to construct hybrid PKS modules.

The potential for recombining intact polyketide synthase (PKS) modules has been extensively explored. Both enzyme-substrate and protein-protein interactions influence chimeric PKS activity, but their relative contributions are unclear. We now address this issue by studying a library of 11 bimodular and 8 trimodular chimeric PKSs harboring modules from the erythromycin, rifamycin, and rapamycin synthases. Although many chimeras yielded detectable products, nearly all had specific activities below 10% of the reference natural PKSs. Analysis of selected bimodular chimeras, each with the same upstream module, revealed that turnover correlated with the efficiency of intermodular chain translocation. Mutation of the acyl carrier protein (ACP) domain of the upstream module in one chimera at a residue predicted to influence ketosynthase-ACP recognition led to improved turnover. In contrast, replacement of the ketoreductase domain of the upstream module by a paralog that produced the enantiomeric ACP-bound diketide caused no changes in processing rates for each of six heterologous downstream modules compared with those of the native diketide. Taken together, these results demonstrate that protein-protein interactions play a larger role than enzyme-substrate recognition in the evolution or design of catalytically efficient chimeric PKSs.

Fusicoccin A is a diterpene glucoside phytotoxin generated by the fungal pathogen Phomopsis amygdali that causes the plant disease constriction canker, first discovered in New Jersey peach orchards in the 1930s. Fusicoccin A is also an emerging new lead in cancer chemotherapy. The hydrocarbon precursor of fusicoccin A is the tricyclic diterpene fusicoccadiene, which is generated by a bifunctional terpenoid synthase. Here, we report X-ray crystal structures of the individual catalytic domains of fusicoccadiene synthase: the C-terminal domain is a chain elongation enzyme that generates geranylgeranyl diphosphate, and the N-terminal domain catalyzes the cyclization of geranylgeranyl diphosphate to form fusicoccadiene. Crystal structures of each domain complexed with bisphosphonate substrate analogues suggest that three metal ions and three positively charged amino acid side chains trigger substrate ionization in each active site. While in vitro incubations reveal that the cyclase domain can utilize farnesyl diphosphate and geranyl diphosphate as surrogate substrates, these shorter isoprenoid diphosphates are mainly converted into acyclic alcohol or hydrocarbon products. Gel filtration chromatography and analytical ultracentrifugation experiments indicate that full-length fusicoccadiene synthase adopts hexameric quaternary structure, and small-angle X-ray scattering data yield a well-defined molecular envelope illustrating a plausible model for hexamer assembly.