Sex differences in the functional organization of the brain for language

A much debated question is whether sex differences exist in the functional organization of the brain for language. A long-held hypothesis posits that language functions are more likely to be highly lateralized in males and to be represented in both cerebral hemispheres in females, but attempts to demonstrate this have been inconclusive. Here we use echo-planar functional magnetic resonance imaging to study 38 right-handed subjects (19 males and 19 females) during orthographic (letter recognition), phonological (rhyme) and semantic (semantic category) tasks. During phonological tasks, brain activation in males is lateralized to the left inferior frontal gyrus regions; in females the pattern of activation is very different, engaging more diffuse neural systems that involve both the left and right inferior frontal gyrus. Our data provide clear evidence for a sex difference in the functional organization of the brain for language and indicate that these variations exist at the level of phonological processing.

Modular organization of genes required for complex polyketide biosynthesis

In Saccharopolyspora erythraea, the genes that govern synthesis of the polyketide portion of the macrolide antibiotic erythromycin are organized in six repeated units that encode fatty acid synthase (FAS)-like activities. Each repeated unit is designated a module, and two modules are contained in a single open reading frame. A model for the synthesis of this complex polyketide is proposed, where each module encodes a functional synthase unit and each synthase unit participates specifically in one of the six FAS-like elongation steps required for formation of the polyketide. In addition, genetic organization and biochemical order of events appear to be colinear. Evidence for the model is provided by construction of a selected mutant and by isolation of a polyketide of predicted structure.

Tumors of the brain and nervous system after radiotherapy in childhood

We investigated the relation between radiotherapy in childhood for tinea capitis and the later development of tumors of the brain and nervous system among 10,834 patients treated between 1948 and 1960 in Israel. Benign and malignant tumors were identified from the pathology records of all Israeli hospitals and from Israeli national cancer and death registries. Doses of radiation to the neural tissue were retrospectively estimated for each patient (mean, 1.5 Gy). Sixty neural tumors developed in the patients exposed as children, and the 30-year cumulative risk (+/- SE) was 0.8 +/- 0.2 percent. The incidence of tumors was 1.8 per 10,000 persons per year. The estimated relative risk as compared with that for 10,834 matched general-population controls and 5392 siblings who had not been irradiated was 6.9 (95 percent confidence interval, 4.1 to 11.6) for all tumors and 8.4 (confidence interval, 4.8 to 14.8) when the analysis was restricted to neural tumors of the head and neck. Increased risks were apparent for meningiomas (relative risk, 9.5; n = 19), gliomas (relative risk, 2.6; n = 7), nerve-sheath tumors (relative risk, 18.8; n = 25), and other neural tumors (relative risk, 3.4; n = 9). A strong dose--response relation was found, with the relative risk approaching 20 after estimated doses of approximately 2.5 Gy. Our study confirms that radiation doses on the order of 1 to 2 Gy can significantly increase the risk of neural tumors.

Functional disruption in the organization of the brain for reading in dyslexia

Learning to read requires an awareness that spoken words can be decomposed into the phonologic constituents that the alphabetic characters represent. Such phonologic awareness is characteristically lacking in dyslexic readers who, therefore, have difficulty mapping the alphabetic characters onto the spoken word. To find the location and extent of the functional disruption in neural systems that underlies this impairment, we used functional magnetic resonance imaging to compare brain activation patterns in dyslexic and nonimpaired subjects as they performed tasks that made progressively greater demands on phonologic analysis. Brain activation patterns differed significantly between the groups with dyslexic readers showing relative underactivation in posterior regions (Wernicke's area, the angular gyrus, and striate cortex) and relative overactivation in an anterior region (inferior frontal gyrus). These results support a conclusion that the impairment in dyslexia is phonologic in nature and that these brain activation patterns may provide a neural signature for this impairment.

Stimulation by cyclic adenosine monophosphate of plasmid deoxyribonucleic acid replication and catabolite repression of the plasmid deoxyribonucleic acid-protein relaxation complex

Colicinogenic factors ColE1 and ColE2 are bacterial plasmids that exist in Escherichia coli as supercoiled deoxyribonucleic acid (DNA) and as strand-specific, relaxation complexes of supercoiled DNA and protein. Newly replicated ColE1 DNA becomes complexed with protein after the replication event. This association of DNA and protein can take place under conditions in which DNA or protein synthesis is arrested. The addition of cyclic adenosine monophosphate (c-AMP) to normal cells growing in glucose medium results in a six- to tenfold stimulation in the rate of synthesis of the protein component(s) of the complex and a three- to fivefold stimulation in the rate of ColE1 DNA replication. Employing mutants deficient in catabolite gene activator protein or adenylate cyclase, it was shown that synthesis of both the plasmid-determined protein colicin E1 and the protein component(s) of the ColE1 relaxation complex is mediated through the c-AMP-catabolite gene activator protein system. Addition of c-AMP to ColE2-containing cells results in the stimulation of synthesis of ColE2 DNA and relaxation protein(s) as well as in the production of a protein component of the ColE2 relaxation complex that renders it sensitive to induced relaxation by heat treatment. In the case of ColE2, synthesis of the relaxation protein(s) is not dependent upon catabolite gene activator protein.

Cerebral organization of component processes in reading

The cerebral organization of word identification processes in reading was examined using functional magnetic resonance imaging (fMRI). Changes in fMRI signal intensities were measured in 38 subjects (19 males and 19 females) during visual (line judgement), orthographic (letter case judgement), phonological (nonword rhyme judgement) and semantic (semantic category judgement) tasks. A strategy of multiple subtractions was employed in order to validate relationships between structure and function. Orthographic processing made maximum demands on extrastriate sites, phonological processing on a number of frontal and temporal sites, and lexical-semantic processing was most strongly associated with middle and superior temporal sites. Significant sex differences in the cerebral organization of reading-related processes were also observed.

Functional neuroimaging studies of reading and reading disability (developmental dyslexia)

Converging evidence from a number of neuroimaging studies, including our own, suggest that fluent word identification in reading is related to the functional integrity of two consolidated left hemisphere (LH) posterior systems: a dorsal (temporo-parietal) circuit and a ventral (occipito-temporal) circuit. This posterior system is functionally disrupted in developmental dyslexia. Reading disabled readers, relative to nonimpaired readers, demonstrate heightened reliance on both inferior frontal and right hemisphere posterior regions, presumably in compensation for the LH posterior difficulties. We propose a neurobiological account suggesting that for normally developing readers the dorsal circuit predominates at first, and is associated with analytic processing necessary for learning to integrate orthographic features with phonological and lexical-semantic features of printed words. The ventral circuit constitutes a fast, late-developing, word identification system which underlies fluent word recognition in skilled readers.

Effect of estrogen on brain activation patterns in postmenopausal women during working memory tasks

CONTEXT

Preclinical studies suggest that estrogen affects neural structure and function in mature animals; clinical studies are less conclusive with many, but not all, studies showing a positive influence of estrogen on verbal memory in postmenopausal women.

OBJECTIVE

To investigate the effects of estrogen on brain activation patterns in postmenopausal women as they performed verbal and nonverbal working memory tasks.

DESIGN

Randomized, double-blind, placebo-controlled, crossover trial from 1996 through 1998.

SETTING

Community volunteers tested in a hospital setting.

PATIENTS

Forty-six postmenopausal women aged 33 to 61 years (mean [SD] age, 50.8 [4.7] years).

INTERVENTION

Twenty-one-day treatment with conjugated equine estrogens, 1.25 mg/d, randomly crossed over with identical placebo and a 14-day washout between treatments.

MAIN OUTCOME MEASURES

Brain activation patterns measured using functional magnetic resonance imaging during tasks involving verbal and nonverbal working memory.

RESULTS

Treatment with estrogen increased activation in the inferior parietal lobule during storage of verbal material and decreased activation in the inferior parietal lobule during storage of nonverbal material. Estrogen also increased activation in the right superior frontal gyrus during retrieval tasks, accompanied by greater left-hemisphere activation during encoding. The latter pattern represents a sharpening of the hemisphere encoding/retrieval asymmetry (HERA) effect. Estrogen did not affect actual performance of the verbal and nonverbal memory tasks.

CONCLUSIONS

Estrogen in a therapeutic dosage alters brain activation patterns in postmenopausal women in specific brain regions during the performance of the sorts of memory function that are called upon frequently during any given day. These results suggest that estrogen affects brain organization for memory in postmenopausal women.

Neurobiological studies of reading and reading disability

Evidence from neuroimaging studies, including our own, suggest that skilled word identification in reading is related to the functional integrity of two consolidated left hemisphere (LH) posterior systems: a dorsal (temporo-parietal) circuit and a ventral (occipito-temporal) circuit. This posterior system appears to be functionally disrupted in developmental dyslexia. Relative to nonimpaired readers, reading-disabled individuals demonstrate heightened reliance on both inferior frontal and right hemisphere posterior regions, presumably in compensation for the LH posterior difficulties. We propose a neurobiological account suggesting that for normally developing readers, the dorsal circuit predominates at first, and in conjunction with premotor systems, is associated with analytic processing necessary for learning to integrate orthographic with phonological and lexical semantic features of printed words. The ventral circuit constitutes a fast, late-developing, word form system, which underlies fluency in word recognition.

LEARNING OUTCOMES

As a result of this activity, (1) the participant will learn about a model of lexical processing involving specific cortical regions. (2) The participant will learn about evidence which supports the theory that two dorsal LH systems may be disrupted in developmental dyslexia. (3) The participant will learn that individuals with reading impairment may rely on other regions of the brain to compensate for the disruption of posterior function.

Polyketide synthesis: prospects for hybrid antibiotics

Polyketides fall into two structural classes: aromatic and complex. The former are built mainly from acetate units through a reiterative process wherein the beta-carbonyl groups formed after each condensation cycle are left largely unreduced. Complex polyketides are composed of acetates, propionates, or butyrates, and the extent of beta-carbonyl reduction varies from one cycle to the next. Two themes for polyketide synthases are emerging. Aromatic PKSs are determined by four to six genes encoding mono- or bifunctional enzymes; one PKS complex is used for all synthesis steps. Complex PKSs are composed of several multifunctional polypeptides that contain enzymatic domains for the condensation and reduction steps; each domain is used at a unique step in the pathways, and the extent of beta-carbonyl processing depends on the functional domains operating at that cycle. Mutations rendering certain domains nonfunctional have been introduced into genes for complex polyketides, resulting in the production of novel molecules.

Cloning and heterologous expression of the epothilone gene cluster

The polyketide epothilone is a potential anticancer agent that stabilizes microtubules in a similar manner to Taxol. The gene cluster responsible for epothilone biosynthesis in the myxobacterium Sorangium cellulosum was cloned and completely sequenced. It encodes six multifunctional proteins composed of a loading module, one nonribosomal peptide synthetase module, eight polyketide synthase modules, and a P450 epoxidase that converts desoxyepothilone into epothilone. Concomitant expression of these genes in the actinomycete Streptomyces coelicolor produced epothilones A and B. Streptomyces coelicolor is more amenable to strain improvement and grows about 10-fold as rapidly as the natural producer, so this heterologous expression system portends a plentiful supply of this important agent.

Efficacy and safety of sitagliptin when added to insulin therapy in patients with type 2 diabetes

OBJECTIVE

To evaluate the efficacy and tolerability of sitagliptin when added to insulin therapy alone or in combination with metformin in patients with type 2 diabetes.

METHODS

After a 2 week placebo run-in period, eligible patients inadequately controlled on long-acting, intermediate-acting or premixed insulin (HbA1c > or = 7.5% and < or = 11%), were randomised 1:1 to the addition of once-daily sitagliptin 100 mg or matching placebo over a 24-week study period. The study capped the proportion of randomised patients on insulin plus metformin at 75%. Further, the study capped the proportion of randomised patients on premixed insulin at 25%. The metformin dose and the insulin dose were to remain stable throughout the study. The primary endpoint was HbA1c change from baseline at week 24.

RESULTS

Mean baseline characteristics were similar between the sitagliptin (n = 322) and placebo (n = 319) groups, including HbA1c (8.7 vs. 8.6%), diabetes duration (13 vs. 12 years), body mass index (31.4 vs. 31.4 kg/m(2)), and total daily insulin dose (51 vs. 52 IU), respectively. At 24 weeks, the addition of sitagliptin significantly (p < 0.001) reduced HbA1c by 0.6% compared with placebo (0.0%). A greater proportion of patients achieved an HbA1c level < 7% while randomised to sitagliptin as compared with placebo (13 vs. 5% respectively; p < 0.001). Similar HbA1c reductions were observed in the patient strata defined by insulin type (long-acting and intermediate-acting insulins or premixed insulins) and by baseline metformin treatment. The addition of sitagliptin significantly (p < 0.001) reduced fasting plasma glucose by 15.0 mg/dl (0.8 mmol/l) and 2-h postmeal glucose by 36.1 mg/dl (2.0 mmol/l) relative to placebo. A higher incidence of adverse experiences was reported with sitagliptin (52%) compared with placebo (43%), due mainly to the increased incidence of hypoglycaemia (sitagliptin, 16% vs. placebo, 8%). The number of hypoglycaemic events meeting the protocol-specified criteria for severity was low with sitagliptin (n = 2) and placebo (n = 1). No significant change from baseline in body weight was observed in either group.

CONCLUSION

In this 24-week study, the addition of sitagliptin to ongoing, stable-dose insulin therapy with or without concomitant metformin improved glycaemic control and was generally well tolerated in patients with type 2 diabetes.

The angular gyrus in developmental dyslexia: task-specific differences in functional connectivity within posterior cortex

Converging evidence from neuroimaging studies of developmental dyslexia reveals dysfunction at posterior brain regions centered in and around the angular gyrus in the left hemisphere. We examined functional connectivity (covariance) between the angular gyrus and related occipital and temporal lobe sites, across a series of print tasks that systematically varied demands on phonological assembly. Results indicate that for dyslexic readers a disruption in functional connectivity in the language-dominant left hemisphere is confined to those tasks that make explicit demands on assembly. In contrast, on print tasks that do not require phonological assembly, functional connectivity is strong for both dyslexic and nonimpaired readers. The findings support the view that neurobiological anomalies in developmental dyslexia are largely confined to the phonological-processing domain. In addition, the findings suggest that right-hemisphere posterior regions serve a compensatory role in mediating phonological performance in dyslexic readers.

The FK520 gene cluster of Streptomyces hygroscopicus var. ascomyceticus (ATCC 14891) contains genes for biosynthesis of unusual polyketide extender units

FK520 (ascomycin) is a macrolide produced by Streptomyces hygroscopicus var. ascomyceticus (ATCC 14891) that has immunosuppressive, neurotrophic and antifungal activities. To further elucidate the biosynthesis of this and related macrolides, we cloned and sequenced an 80kb region encompassing the FK520 gene cluster. Genes encoding the three polyketide synthase (PKS) subunits (fkbB, fkbC and fkbA), the peptide synthetase (fkbP), the 31-O-methyltransferase (fkbM), the C-9 hydroxylase (fkbD) and the 9-hydroxyl oxidase (fkbO) had the same organization as the genes reported in the FK506 gene cluster of Streptomyces sp. MA6548 (Motamedi, H., Shafiee, A., 1998. The biosynthetic gene cluster for the macrolactone ring of the immunosuppressant FK506. Eur. J. Biochem. 256, 528-534). Disruption of a PKS gene in the cluster using the &phi;C31 phage vector, KC515, led to antibiotic non-producing strains, proving the identity of the cluster. Previous labeling data have indicated that FK520 biosynthesis uses novel polyketide extender units (Byrne, K.M., Shafiee, A., Nielson, J., Arison, B., Monaghan, R.L., Kaplan, L., 1993. The biosynthesis and enzymology of an immunosuppressant, immunomycin, produced by Streptomyces hygroscopicus var, ascomyceticus. Dev. Ind. Microbiol. 32, 29-45). Genes in the flanking regions of the FK520 cluster were identified that appear to be involved in synthesis of these extender units. All but two of these genes were homologous to genes with known function. In addition to a crotonyl-CoA reductase gene (fkbS), at least two other genes are proposed to be involved in biosynthesis of the atypical PKS extender unit ethylmalonyl-CoA, which accounts for the ethyl side chain on C-21 of FK520. A set of five contiguous genes (fkbGHIJK) is proposed to be involved in biosynthesis of an unusual PKS extender unit bearing an oxygen on the alpha-carbon, and leading to the 13- and 15-methoxy side chains. These putative precursor synthesis genes in the flanking regions of the FK520 cluster are not found in the flanking regions of the rapamycin cluster (Molnár, I., Aparicio, J.F., Haydock, S.F., Khaw, L.E., Schwecke, T., König, A., Staunton, J., Leadlay, P.F., 1996. Organisation of the biosynthetic gene cluster for rapamycin in Streptomyces hygroscopicus: analysis of genes flanking the polyketide synthase. Gene 169, 1-7), consistent with labeling data showing that rapamycin biosynthesis uses only malonyl and methylmalonyl extender units.

Outcome model of asphyxial cardiac arrest in rats

An outcome model with asphyxial cardiac arrest in rats has been developed for quantifying brain damage. Twenty-two rats were randomized into three groups. Control group I was normal, was conscious, and had no asphyxia (n = 6). Sham group II had anesthesia and surgery but no asphyxia (n = 6). All 12 rats in groups I and II survived to 72 h and were functionally and histologically normal. Arrest group III (the model; n = 10) had light anesthesia and apneic asphyxia of 8 min, which led to cessation of circulation at 3-4 min of apnea, resulting in cardiac arrest (no flow) of 4-5 min. All 10 rats had spontaneous circulation restored by standard external cardiopulmonary resuscitation. Nine rats survived controlled ventilation for 1 h and observation to 72 h, while one rat died before extubation. All nine survivors were conscious at 72 h, with neurologic deficit scores (0% = best; 100% = worst) of 7 +/- 6% (2-16%). All brain regions at five coronal levels were examined for ischemic neurons. The prevalence of ischemic neurons in five regions was categorically scored. The average total brain histopathologic damage score in group III (n = 9) was 2.1 (p < 0.05 vs. group I or II). A reproducible outcome model of cardiac arrest in rats was documented. It provides a tool for investigating pathophysiological mechanisms of neuronal death caused by a transient global hypoxic-ischemic brain insult.

Organization of the enzymatic domains in the multifunctional polyketide synthase involved in erythromycin formation in Saccharopolyspora erythraea

Localization of the enzymatic domains in the three multifunctional polypeptides from Saccharopolyspora erythraea involved in the formation of the polyketide portion of the macrolide antibiotic erythromycin was determined by computer-assisted analysis. Comparison of the six synthase units (SU) from the eryA genes with each other and with mono- and multifunctional fatty acid and polyketide synthases established the extent of each beta-ketoacyl acyl-carrier protein (ACP) synthase, acyltransferase, beta-ketoreductase, ACP, and thioesterase domain. The extent of the enoyl reductase (ER) domain was established by detecting similarity to other sequences in the database. A segment containing the putative dehydratase (DH) domain in EryAII, with a potential active-site histidine residue, was also found. The finding of conservation of a portion of the DH-ER interdomain region in the other five SU, which lack these two functions, suggests a possible evolutionary path for the generation of the six SU.

Auditory selective attention: an fMRI investigation

In the present experiment, 25 adult subjects discriminated speech tokens ([ba]/[da]) or made pitch judgments on tone stimuli (rising/falling) under both binaural and dichotic listening conditions. We observed that when listeners performed tasks under the dichotic conditions, during which greater demands are made on auditory selective attention, activation within the posterior (parietal) attention system and at primary processing sites in the superior temporal and inferior frontal regions was increased. The cingulate gyrus within the anterior attention system was not influenced by this manipulation. Hemispheric differences between speech and nonspeech tasks were also observed, both at Broca's Area within the inferior frontal gyrus and in the middle temporal gyrus.

Strategies for visual word recognition and orthographical depth: A multilingual comparison

We investigated the psychological reality of the concept of orthographical depth and its influence on visual word recognition by examining naming performance in Hebrew, English, and Serbo-Croatian. We ran three sets of experiments in which we used native speakers and identical experimental methods in each language. Experiment 1 revealed that the lexical status of the stimulus (high-frequency words, low-frequency words, and nonwords) significantly affected naming in Hebrew (the deepest of the three orthographies). This effect was only moderate in English and nonsignificant in Serbo-Croatian (the shallowest of the three orthographies). Moreover, only in Hebrew did lexical status have similar effects on naming and lexical decision performance. Experiment 2 revealed that semantic priming effects in naming were larger in Hebrew than in English and completely absent in Serbo-Croatian. Experiment 3 revealed that a large proportion of nonlexical tokens (nonwords) in the stimulus list affects naming words in Hebrew and in English, but not in Serbo-Croatian. These results were interpreted as strong support for the orthographical depth hypothesis and suggest, in general, that in shallow orthographies phonology is generated directly from print, whereas in deep orthographies phonology is derived from the internal lexicon.

Engineered biosynthesis of a complete macrolactone in a heterologous host

Macrocyclic polyketides have been subjects of great interest in synthetic and biosynthetic chemistry because of their structural complexity and medicinal activities. With expression of the entire 6-deoxyerythronolide B synthase (DEBS) (10,283 amino acids) in a heterologous host, substantial quantities of 6-deoxyerythronolide B (6dEB), the aglycone of the macrolide antibiotic erythromycin, and 8,8a-deoxyoleandolide, a 14-membered lactone ring identical to 6dEB except for a methyl group side chain in place of an ethyl unit, were synthesized in Streptomyces coelicolor. The biosynthetic strategy utilizes a genetic approach that facilitates rapid structural manipulation of DEBS or other modular polyketide synthases (PKSs), including those found in actinomycetes with poorly developed genetic methods. From a technological viewpoint, this approach should allow the rational design of biosynthetic products and may eventually lead to the generation of diverse polyketide libraries by means of combinatorial cloning of naturally occurring and mutant PKS modules.

An erythromycin analog produced by reprogramming of polyketide synthesis

The polyketide-derived macrolactone of the antibiotic erythromycin is made through successive condensation and processing of seven three-carbon units. The fourth cycle involves complete processing of the newly formed beta-keto group (beta-keto reduction, dehydration, and enoyl reduction) to yield the methylene that will appear at C-7 of the lactone ring. Synthesis of this molecule in Saccharopolyspora erythraea is determined by the three large eryA genes, organized in six modules, each governing one condensation cycle. Two amino acid substitutions were introduced in the putative NAD(P)H binding motif in the proposed enoyl reductase domain encoded by eryAII. The metabolite produced by the resulting strain was identified as delta 6,7-anhydroerythromycin C resulting from failure of enoyl reduction during the fourth cycle of synthesis of the macrolactone. This result demonstrates the involvement of at least the enoyl reductase from the fourth module in the fourth cycle and indicates that a virtually complete macrolide can be produced through reprogramming of polyketide synthesis.

Patients with type 2 diabetes mellitus have higher risk for acute pancreatitis compared with those without diabetes

AIM

The aetiology of acute pancreatitis (AP) is complex, and many risk factors for AP are shared by patients with type 2 diabetes mellitus (T2DM). However, few have assessed risk factors for AP specifically in T2DM patients.

METHODS

Patients in the General Practice Research Database (2 984 755, 5.0% with T2DM) were used to estimate incidence of AP for T2DM relative to non-diabetes, adjusting for prior pancreatitis, gallbladder disease, obesity, smoking and alcohol use. Multivariate Cox regression analysis adjusting for risk factors and Charlson comorbidity index (CCI) was used to estimate hazard ratios (HR) with 95% confidence intervals (CI).

RESULTS

Between 2003 and 2007, 301 of 148 903 patients with T2DM and 2434 of almost 3 million patients without diabetes developed AP. Patients with T2DM had higher risk for AP compared with patients without diabetes (crude HR: 2.89, 95% CI: 2.56-3.27). Patients with T2DM had significantly higher rates of prior alcohol and tobacco exposure (44.2 and 61.9% vs. 34.1 and 35.9%, p < 0.001) and of comorbid conditions (14.7% with CCI > or =1 vs. 4.3%, p < 0.001). Histories of obesity, pancreatitis, gallbladder disease, smoking or alcohol use were significant predictors of AP. After adjusting for these factors, age, gender and comorbidities, the risk of developing AP remained elevated in patients with T2DM (adjusted HR: 1.49, 95% CI: 1.31-1.70).

CONCLUSION

After adjusting for risk factors, patients with T2DM had an elevated risk of AP compared with patients without diabetes. Physicians should be aware of the increased risk in patients with T2DM, particularly in those with prior pancreatitis.

Isolation and characterization of the epothilone biosynthetic gene cluster from Sorangium cellulosum

The epothilone biosynthetic gene cluster was isolated from Sorangium cellulosum strain SMP44. The gene cluster contains seven genes and spans approx. 56kb. The genes encoding the PKS, epoA, epoC, epoD, epoE, and epoF, are divided into nine modules. The EpoB protein is a non-ribosomal peptide synthetase (NRPS) that catalyzes formation of the thiazole found in the epothilones. EpoK is a P450 enzyme responsible for the epoxidation of epothilones C and D to epothilones A and B, respectively. EpoK was expressed in Escherichia coli, and the purified protein was shown to convert epothilone D to epothilone B in vitro.