Identification of selective inhibitors of cyclin dependent kinase 4

A new structural type of kinase inhibitor, containing a benzocarbazole nucleus, has been identified. Members of the series are selective for inhibition of the cyclin dependent kinase family of enzymes. Although the cdks are highly homologous, representatives of the series showed intra-cdk selectivities, especially for cdk4. SAR studies elucidated the important features of the molecules for inhibition.

Human mdm2 mediates multiple mono-ubiquitination of p53 by a mechanism requiring enzyme isomerization

The mdm2 gene product is an important regulator of p53 function and stability. mdm2 is an E3 ubiquitin ligase for p53 and the RING finger domain of mdm2 is critical for ligase activity. Ubiquitin (Ub) conjugation is a general targeting modification and poly-ubiquitin chains specifically target proteins to the proteasome for degradation. In this report, we show that the multistep cascade of mdm2-mediated p53 ubiquitination can be reduced to three purified recombinant proteins: ubiquitin-conjugated E2, mdm2, and p53. This simplification allows enzymatic analysis of the isolated ligase reaction. The simplified reaction recapitulates the ubiquitination of p53 observed with individual components and the p53-Ub((n)) is qualitatively similar to p53-Ub((n)) detected in lactacystin-treated cells. Surprisingly, we find that p53 is modified with multiple mono-ubiquitin moieties as opposed to a poly-ubiquitin chain. Finally, kinetic analysis indicates the transfer reaction proceeds either through a modified Ping Pong mechanism involving requisite enzyme isomerization steps, or through a Rapid Equilibrium Random Bi Bi mechanism involving very large anti-cooperative interactions between the two substrate binding pockets on the enzyme, mediated through allosteric changes in enzyme structure.

Quinazolines as cyclin dependent kinase inhibitors

Quinazolines have been identified as inhibitors of CDK4/D1 and CDK2/E. Aspects of the SAR were investigated using solution-phase, parallel synthesis. An X-ray crystal structure was obtained of quinazoline 51 bound in CDK2 and key interactions within the ATP binding pocket are defined.

Human carbon catabolite repressor protein (CCR4)-associative factor 1: cloning, expression and characterization of its interaction with the B-cell translocation protein BTG1

The human BTG1 protein is thought to be a potential tumour suppressor because its overexpression inhibits NIH 3T3 cell proliferation. However, little is known about how BTG1 exerts its anti-proliferative activity. In this study, we used the yeast 'two-hybrid' system to screen for interacting protein partners and identified human carbon catabolite repressor protein (CCR4)-associative factor 1 (hCAF-1), a homologue of mouse CAF-1 (mCAF-1) and Saccharomyces cerevisiae yCAF-1/POP2. In vitro the hCAF-1/BTG1 complex formation was dependent on the phosphorylation of a putative p34cdc2 kinase site on BTG1 (Ser-159). In yeast, the Ala-159 mutant did not interact with hCAF-1. In addition, phosphorylation of Ser-159 in vitro showed specificity for the cell cycle kinases p34CDK2/cyclin E and p34CDK2/cyclin A, but not for p34CDK4/cyclin D1 or p34cdc2/cyclin B. Cell synchrony experiments with primary cultures of rat aortic smooth-muscle cells (RSMCs) demonstrated that message and protein levels of rat CAF-1 (rCAF-1) were up-regulated under conditions of cell contact, as previously reported for BTG1 [Wilcox, Scott, Subramanian, Ross, Adams-Burton, Stoltenborg and Corjay (1995) Circulation 92, I34-I35]. Western blot and immunohistochemical analysis showed that rCAF-1 localizes to the nucleus of contact-inhibited RSMCs, where it was physically associated with BTG1, as determined by co-immunoprecipitation with anti-hCAF-1 antisera. Overexpression of hCAF-1 in NIH 3T3 and osteosarcoma (U-2-OS) cells was itself anti-proliferative with colony formation reduced by 67% and 90% respectively. Taken together, these results indicate that formation of the hCAF-1/BTG1 complex is driven by phosphorylation at BTG1 (Ser-159) and implicates this complex in the signalling events of cell division that lead to changes in cellular proliferation associated with cell-cell contact.

A novel, cell-type-specific mechanism for estrogen receptor-mediated gene activation in the absence of an estrogen-responsive element

The estrogen receptor (ER) typically activates gene transcription by binding to estrogen-responsive elements (EREs). The brain creatine kinase (BCK) promoter is responsive to estrogen but contains no ERE-related sequence. To investigate the mechanism of estrogen induction, we have introduced the estrogen receptor into HeLa cells and primary rat cardiomyocytes and fibroblasts along with 195 bp of BCK promoter linked to a chloramphenicol acetyltransferase (CAT) reporter gene. A 10-fold stimulation of CAT activity was observed in the presence of beta-estradiol in both HeLa and rat primary fibroblasts, but no induction was observed in primary rat cardiomyocytes. In contrast, a control vitellogenin gene construct which contains a typical ERE was induced in an ER-dependent manner in all cell types studied. Estrogen induction in HeLa was not sensitive to cycloheximide and was blocked by the ER antagonists tamoxifen and ICI 164,384. Analysis of 5' deletion and linker-scanning mutations indicates sequences between bp -45 and -75 including a TA-rich sequence and a CCAAT sequence to be crucial for stimulation of the BCK promoter by the ER. BCK estrogen induction is dependent on the DNA-binding domain and transactivation domain TAF2 of the ER. However, direct DNA binding is probably not required. Taken together, these results suggest a novel mechanism for ER-mediated gene activation. This mechanism is consensus ERE independent and cell type specific and requires interactions between the ER and molecules capable of interacting with the BCK promoter TA-rich region.

Multiple oligomeric states regulate the DNA binding of helix-loop-helix peptides

To study the protein-protein interactions that allow Id, a negative regulator of cell differentiation, to inhibit the DNA-binding activities of MyoD and E47, we have synthesized peptides corresponding to the helix-loop-helix domains of MyoD, E47, and Id. We show that Id preferentially inhibits the sequence-specific DNA-binding activity of MyoD, a muscle-specific protein, as compared to E47, a more ubiquitous protein. The Id helix-loop-helix domain itself forms stable tetramers, and its inhibitory activity arises from the formation of a heterotetrameric structure with MyoD. The formation of this higher order complex provides a general mechanism by which inhibitory proteins can generate sufficient interaction free energy to overcome the large DNA-binding free energy of dimeric DNA-binding proteins.

TATA box-mediated in vitro transcription by RNA polymerase III. Evidence for TATA-binding protein in a polymerase III type complex

We have shown previously that under specific conditions, a TATA box will mediate efficient in vitro transcription by RNA polymerase (pol) III in the absence of a PSE or other promoter elements. The reaction requires a HeLa cell phosphocellulose protein fraction, fraction B, which must be preincubated with the template DNA. Fraction B does not contain any detectable pol II type transcription factor IID (TFIID) activity. In this report, the relationship between fraction B and TFIID was further examined. Purified human TATA-binding protein (TBP) can substitute for fraction B to mediate TATA-dependent pol III transcription. Both TBP and fraction B prefer a reverse TATA box for pol III transcription, yet TBP bound to a reverse TATA box functions poorly for pol II transcription. Like TFIID, fraction B forms a template-committed complex with TATA-containing promoters. TBP, however, will not template commit for pol III transcription unless premixed with phosphocellulose fraction C. TBP-mediated pol III transcription is also more sensitive to the detergent Sarkosyl (N-lauroylsarcosine, Sigma) than is the fraction B reaction unless it is premixed with fraction C. Together, the data suggest that TBP can complex with a component of fraction C, and this complex is then functionally equivalent to fraction B. We propose that fraction B contains TBP in a complex with some other component(s) of the pol III transcription machinery and that this B complex TBP may be specific for pol III transcription.

Molecular characterization of helix-loop-helix peptides

A class of regulators of eukaryotic gene expression contains a conserved amino acid sequence responsible for protein oligomerization and binding to DNA. This structure consists of an arginine- and lysine-rich basic region followed by a helix-loop-helix motif, which together mediate specific binding to DNA. Peptides were prepared that span this motif in the MyoD protein; in solution, they formed alpha-helical dimers and tetramers. They bound to DNA as dimers and their alpha-helical content increased on binding. Parallel and antiparallel four-helix models of the DNA-bound dimer were constructed. Peptides containing disulfide bonds were engineered to test the correctness of the two models. A disulfide that is compatible with the parallel model promotes specific interaction with DNA, whereas a disulfide compatible with the antiparallel model abolishes specific binding. Electron paramagnetic resonance (EPR) measurements of nitroxide-labeled peptides provided intersubunit distance measurements that also supported the parallel model.

TATA box-mediated polymerase III transcription in vitro

We have defined conditions whereby a functional TATA box can mediate efficient in vitro transcription by RNA polymerase III. A TATA box is absolutely required for this reaction as a single-point mutation in this sequence completely abolishes transcription. Two protein components are also required: a HeLa cell phosphocellulose fraction (fraction B) and at least one other factor that can be supplied by various crude nuclear extracts or by HeLa cell phosphocellulose fractions C and D. The order of addition is critical; fraction B must be preincubated with the template DNA for TATA box-dependent polymerase III transcription to occur. Various TATA sequences are quite similar in their ability to mediate transcription by polymerases II and III. Despite the similarity in sequence requirements, fraction B does not appear to contain any detectable transcription factor (TF) IID activity, and TATA box-mediated polymerase III transcription does not appear to require TFIID in the form contained in phosphocellulose fraction D. It was recently reported that TFIID is required TFIID in the form contained in phosphocellulose fraction D. It was recently reported that TFIID is required for polymerase III transcription of the yeast and human U6 genes (Margottin, F., Dujardin, G., Gerard, M., Egly, J.-M., Huet, J., and Sentenac, A. (1991) Science 251, 424-426; Simmen, K. A., Bernues, J., Parry, H. D., Stunnenberg, H. G., Berkenstam, A., Cavallini, B., Egly, J.-M., and Mattaj, I. W. (1991) EMBO J. 10, 1853-1862). We propose that fraction B may contain TFIID in a modified form that is not functional for polymerase II transcription.

Human cytomegalovirus US3 and UL36-38 immediate-early proteins regulate gene expression

We have established the ability of the human cytomegalovirus (HCMV) UL36-38 and US3 immediate-early (IE) gene products to alter gene expression in human cells by using transient transfection assays. The cellular heat shock protein 70 (hsp70) promoter was transactivated following cotransfection with the HCMV IE regions in nonpermissive HeLa cells by UL36-38, US3, or IE1 and in permissive human diploid fibroblasts (HFF) by IE1 or IE2. Moreover, hsp70 expression was synergistically increased in HeLa cells cotransfected with US3 and UL36, with US3 and UL37, or with US3 and UL37x1. The synergistic transactivation of hsp70 expression by US3 and UL36-38 was not observed in HFF cells. Synergy was also not observed in HeLa cells between US3 and UL38, an early gene product encoded by the UL36-38 IE locus. Synergistic transactivation of hsp70 expression in HeLa cells required the syntheses of UL36-38 and US3 IE proteins, since nonsense mutants were not functional. hsp70 expression increased with increasing amounts of transfected US3 and UL37 DNA and occurred at the level of stable hsp70-promoted RNA. In contrast to the broad hsp70 response, promoters from the HCMV UL112 early gene and another cellular gene, brain creatine kinase, both responded strongly only to singly transfected IE2 in HeLa cells. Nevertheless, IE2 transactivation of the UL112 promoter was further stimulated by cotransfection of IE1 or of UL36-38 in both HeLa and HFF cells. Thus, different patterns of promoter transactivation and interactions between HCMV IE gene products in transactivation were found in HFF cells and in HeLa cells. These results establish the ability of the HCMV US3 and UL36-38 proteins to alter cellular and viral gene expression and are consistent with involvement of cellular transcription factors in HCMV IE regulation of gene expression.

Identification of cis-acting regulatory elements in the promoter region of the rat brain creatine kinase gene

The functional organization of the rat brain creatine kinase (ckb) promoter was analyzed by deletion, linker scanning, and substitution mutagenesis. Mutations were introduced into the ckb promoter of hybrid ckb/neo (neomycin resistance gene) genes, and the mutant genes were expressed transiently in HeLa cells. Expression was assayed by primer extension analysis of neo RNA, which allowed the transcription start sites and the amount of transcription to be determined. Transfections and primer extension reactions were internally controlled by simultaneous analysis of transcription from the adenovirus VA gene located on the same plasmid as the hybrid ckb/neo gene. We demonstrate that 195 bp of the ckb promoter is sufficient for efficient in vivo expression in HeLa cells. A nonconsensus TTAA element at -28 bp appears to provide the TATA box function for the ckb promoter in vivo. Two CCAAT elements, one at -84 bp and the other at -54 bp, and a TATAAA TA element (a consensus TATA box sequence) at -66 bp are required for efficient transcription from the TTAA element. In addition, we present evidence that the consensus beta-globin TATA box responds to the TATAAATA element in the same way as the ckb nonconsensus TTAA element.

Brain and muscle creatine kinase genes contain common TA-rich recognition protein-binding regulatory elements

We have previously reported that the rat brain creatine kinase (ckb) gene promoter contains an AT-rich sequence that is a binding site for a protein called TARP (TA-rich recognition protein). This AT-rich segment is a positively acting regulatory element for the ckb promoter. A similar AT-rich DNA segment is found at the 3' end of the 5' muscle-specific enhancer of the rat muscle creatine kinase (ckm) gene and has been shown to be necessary for full muscle-specific enhancer activity. In this report, we show that TARP binds not only to the ckb promoter but also to the AT-rich segment at the 3' end of the muscle-specific ckm enhancer. A second, weaker TARP-binding site was identified in the ckm enhancer and lies at the 5' end of the minimal enhancer segment. TARP was found in both muscle cells (C2 and L6 myotubes) and nonmuscle (HeLa) cells and appeared to be indistinguishable from both sources, as judged by gel retardation and footprinting assays. The TARP-binding sites in the ckm enhancer and the ckb promoter were found to be functionally interchangeable. We propose that TARP is active in both muscle and nonmuscle cells and that it is one of many potential activators that may interact with muscle-specific regulators to determine the myogenic phenotype.

Two different RNA polymerase II initiation complexes can assemble on the rat brain creatine kinase promoter

The rat brain creatine kinase (CKB) gene has a structurally complex promoter. Although it contains a consensus RNA polymerase II TATA box, this is located at -60 relative to the main transcription start site, an unprecedented distance for mammalian TATA boxes. Steady-state messenger RNA for CKB is extremely abundant in brain, yet barely detectable in liver. Using a nuclear run-on analysis, we show that this difference is due, at least in part, to a difference in the rate at which CKB is transcribed in these tissues. We also demonstrate a tissue-specific hypersensitive site in the promoter region of the CKB gene. In vitro transcription experiments using nuclear extract isolated from liver, brain, and HeLa cells show a marked difference in the way these extracts recognize the CKB promoter. Two different preinitiation complexes can assemble on the CKB promoter, one which initiated at the same downstream start-site used in normal adult rat brain and one which initiated 35 base pairs further upstream. Extracts from brain form exclusively the downstream initiating complex. However, liver extracts favor formation of the upstream initiating complex, but formation of this complex can be blocked if brain or HeLa extract is added first. We demonstrate that the upstream consensus TATA box can function as a TATA box to mediate upstream initiation, but is not required for downstream initiation. We have shown previously that brain and HeLa cell nuclear extracts contain an abundant factor, TARP, which binds to the upstream consensus TATA box and have proposed that this factor is distinct from the classical TATA box binding factor TFIID. We propose that binding of TARP to the upstream consensus TATA box may block formation of the upstream initiating complex.

The upstream muscle-specific enhancer of the rat muscle creatine kinase gene is composed of multiple elements

A series of constructs that links the rat muscle creatine kinase promoter to the bacterial chloramphenicol acetyltransferase gene was generated. These constructs were introduced into differentiating mouse C2C12 myogenic cells to localize sequences that are important for up-regulation of the creatine kinase gene during myogenic differentiation. A muscle-specific enhancer element responsible for induction of chloramphenicol acetyltransferase expression during myogenesis was localized to a 159-base-pair region from 1,031 to 1,190 base pairs upstream of the transcription start site. Analysis of transient expression experiments using promoters mutated by deletion indicated the presence of multiple functional domains within this muscle-specific regulatory element. A DNA fragment spanning this region was used in DNase I protection experiments. Nuclear extracts derived from C2 myotubes protected three regions (designated E1, E2, and E3) on this fragment from digestion, which indicated there may be three or more trans-acting factors that interact with the creatine kinase muscle enhancer. Gel retardation assays revealed that factors able to bind specifically to E1, E2, and E3 are present in a wide variety of tissues and cell types. Transient expression assays demonstrated that elements in regions E1 and E3, but not necessarily E2, are required for full enhancer activity.

Identification of a novel TA-rich DNA binding protein that recognizes a TATA sequence within the brain creatine kinase promoter

The rat brain creatine kinase gene possesses a structurally complex promoter with multiple potential regulatory elements. Two CCAAT sequences, a TATAAATA sequence and a TTAA sequence are found within the first one hundred base pairs. We present evidence that favors the allocation of the downstream TTAA sequence as the potential TATA box. We show that the CCAAT sequences and the upstream TATAAATA sequence are binding sites for potential regulatory factors and that sequences in this region are capable of regulating expression from the downstream TTAA sequence. We suggest that the protein that binds to the upstream TATAAATA sequence is not a classical TFIID factor but rather may serve to block the binding of TFIID and/or to promote transcription from the downstream start site. We have been able to define conditions in vitro under which binding to this upstream TATAAATA sequence does not occur. Under these conditions we are able to detect transcription from both potential TATA sequences, a situation which we have been unable to detect in vivo. Our experiments suggest the existence in HeLa and brain nuclei of a protein that recognizes the concensus TATAAATA sequence, that is distinct from TFIID, and that may function in part to deny access of TFIID to this potential promoter element.

Isolation of four rat creatine kinase genes and identification of multiple potential promoter sequences within the rat brain creatine kinase promoter region

Rat genes encoding muscle creatine kinase (gene ckm) and brain creatine kinase (gene ckb) have been cloned. The two genes have an identical intron-exon distribution, although introns of the ckb gene are much smaller than those for ckm. The complete nucleotide sequence has been determined for the ckb gene. 5'-Flanking sequences have been compared for both ckm and ckb. The ckb gene has a structurally complex 5'-flanking region with multiple TATA and CAAT sequences suggesting that the gene may contain overlapping promoter elements. Examination of ckb RNA from brain tissue provides no evidence for the existence of more than one set of transcriptional start points. The ckm and ckb genes show little homology in the 5'-flanking regions, although we observed some sequence conservation in the regions immediately surrounding the CAAT and TATA sequences. The rat genome also contains two other creatine kinase-like sequences which may represent processed pseudogenes or mitochondrial creatine kinase.

Expression of a rat brain creatine kinase-beta-galactosidase fusion protein in Escherichia coli and derivation of the complete amino acid sequence of rat brain creatine kinase

Expression of a rat brain creatine kinase (CKB)/beta-galactosidase fusion protein in Escherichia coli has allowed isolation of a rat CKB cDNA clone by direct antibody screening of a rat brain gamma gt11 expression library. This clone is 1416 bp long and includes 202 bp of 3'-untranslated region and 29 bp of 5'-untranslated region. The coding sequence of this clone has enabled us to deduce the complete amino acid (aa) sequence of rat CKB protein.

Isolation and sequence analysis of cDNA clones coding for rat skeletal muscle creatine kinase

A series of cDNA clones corresponding to 1494 bases of rat muscle creatine kinase mRNA has been isolated and characterized. The identity of these clones has been confirmed by DNA sequence analysis and by comparison of the predicted amino acid sequence with that determined for the purified protein. The cDNA sequence accounts for the entire coding sequence of the creatine kinase protein in addition to the complete 3' untranslated region and 68 bases of 5' noncoding region. Sequences corresponding to the active site region of the protein, the initiation codon, the termination codon, and poly(A) addition signal have been identified.

Myosin isozymes in avian skeletal muscles. II. Fractionation of myosin isozymes from adult and embryonic chicken pectoralis muscle by immuno-affinity chromatography

Chicken pectoralis consists primarily of large white fibres, which react exclusively with antibodies prepared against adult fast myosin. There is, however, a small region of uniformly red fibres which responds to antibodies against adult slow myosin as well as adult fast myosin. The myosin extracted from this red region is also heterogeneous as shown by the presence of both slow and fast light chains. By means of immunoadsorbents, it has been possible to separate the 'red myosin' into a 'fast' component and a 'slow' component. These two fractions have been characterized with respect to their light and heavy chain content by one-dimensional and two-dimensional gel electrophoresis. The myosin heavy chain was reduced to the smaller fragments required for electrophoresis by proteolytic degradation. We conclude from the electrophoretic patterns that the 'fast' and 'slow' myosin components from the pectoralis red region closely resemble the myosin from the white region of the pectoralis and the myosin from the slow anterior latissimus dorsi (ALD) muscle. The demonstration of a 'slow myosin' in adult pectoralis muscle raises the possibility that the crossreactivity of embryonic pectoralis myosin with anti-slow (ALD) myosin antibodies might be due to the presence of such slow components in embryonic chicken muscle. Direct isolation of a slow component from embryonic pectoralis was achieved by immunoadsorption, as described for adult mixed muscle myosin. Analysis of the subunit composition by gel electrophoresis shows an enrichment in adult-type slow light chains, but the heavy chain pattern is quite distinct from that of adult slow heavy chain. These studies suggest that several myosin isozymes exist in embryonic chicken pectoralis, but that none is identical to those myosins found in the different fibres of the adult pectoralis muscle.

Myosin isozymes in avian skeletal muscles. I. Sequential expression of myosin isozymes in developing chicken pectoralis muscles

Myosin has been purified from chicken pectoralis muscle at various stages of development, from 10 days' incubation to approximately 10 months after hatching. Embryonic myosin from the earliest stage showed a high level of ATPase activity, similar to that obtained for adult pectoralis myosin. Two-dimensional peptide mapping of partial chymotryptic digests showed, however, that is heavy chain is quite different from that of adult fast myosin. The immunological crossreactivity observed between embryonic myosin and adult fast (pectoralis) myosin is therefore due to shared antigenic determinants rather than the presence of any adult isoforms. In an accompanying paper we will show that embryonic myosin at 10 days' incubation is not a single species, but consists of at least two heavy chain isozymes. The minor fraction binds slow light chains preferentially, and appears to be largely responsible for the observed crossreactivity with slow (ALD) myosin. None of the embryonic myosins is equivalent to the adult forms. Prior to hatching, LC3f is present only in very small amounts (less than 5%), and the adult light chain pattern, containing LC1f and LC3f in equimolar amounts, is not generated until after one week post-hatching. At about that time a new heavy chain population is detected, different from either the embryonic heavy chain or the adult heavy chain. The adult heavy chain peptide pattern appears from about three weeks' post-hatching, but a map indistinguishable from that of adult myosin is not observed until about 26 weeks. None of the observed differences in peptide maps can be related to different strains of chicken; pectoralis myosin from adult White Rock gave an identical map to that from White Leghorn. Unexpectedly, posterior latissimus dorsi (PLD) myosin from White Leghorn appears to be different from pectoralis myosin from the same strain, despite the histochemical and immunocytochemical similarity of the two muscles. We conclude that myosin polymorphism is widespread in muscle tissue, and that the expression of myosin isozymes and their subunits is under developmental regulation.

Distribution and properties of myosin isozymes in developing avian and mammalian skeletal muscle fibers

Isozymes of myosin have been localized with respect to individual fibers in differentiating skeletal muscles of the rat and chicken using immunocytochemistry. The myosin light chain pattern has been analyzed in the same muscles by two-dimensional PAGE. In the muscles of both species, the response to antibodies against fast and slow adult myosin is consistent with the speed of contraction of the muscle. During early development, when speed of contraction is slow in future fast and slow muscles, all the fibers react strongly with anti-slow as well as with anti-fast myosin. As adult contractile properties are acquired, the fibers react with antibodies specific for either fast or slow myosin, but few fibers react with both antibodies. The myosin light chain pattern slow shows a change with development: the initial light chains (LC) are principally of the fast type, LC1(f), and LC2(f), independent of whether the embryonic muscle is destined to become a fast or a slow muscle in the adult. The LC3(f), light chain does not appear in significant amounts until after birth, in agreement with earlier reports. The predominance of fast light chains during early stages of development is especially evident in the rat soleus and chicken ALD, both slow muscles, in which LC1(f), is gradually replaced by the slow light chain, LC1(s), as development proceeds. Other features of the light chain pattern include an "embryonic" light chain in fetal and neonatal muscles of the rat, as originally demonstrated by R.G. Whalen, G.S. Butler- Browne, and F. Gros. (1978. J. Mol. Biol. 126:415-431.); and the presence of approximately 10 percent slow light chains in embryonic pectoralis, a fast white muscle in the adult chicken. The response of differentiating muscle fibers to anti-slow myosin antibody cannot, however, be ascribed solely to the presence of slow light chains, since antibody specific for the slow heavy chain continues to react with all the fibers. We conclude that during early development, the myosin consists of a population of molecules in which the heavy chain can be associated with a fast, slow, or embryonic light chain. Biochemical analysis has shown that this embryonic heavy chain (or chains) is distinct from adult fast or slow myosin (R.G. Whalen, K. Schwartz, P. Bouveret, S.M. Sell, and F. Gros. 1979. Proc. Natl. Acad. Sci. U.S.A. 76:5197-5201. J.I. Rushbrook, and A. Stracher. 1979. Proc Natl. Acad. Sci. U.S.A. 76:4331-4334. P.A. Benfield, S. Lowey, and D.D. LeBlanc. 1981. Biophys. J. 33(2, Pt. 2):243a[Abstr.]). Embryonic myosin, therefore, constitutes a unique class of molecules, whose synthesis ceases before the muscle differentiates into an adult pattern of fiber types.

Extended amino acid sequences around the active-site lysine residue of class-I fructose 1,6-bisphosphate aldolases from rabbit muscle, sturgeon muscle, trout muscle and ox liver

1. Amino acid sequences covering the region between residues 173 and 248 [adopting the numbering system proposed by Lai, Nakai & Chang (1974) Science 183, 1204-1206] were derived for trout (Salmo trutta) muscle aldolase and for ox liver aldolase. A comparable sequence was derived for residues 180-248 of sturgeon (Acipenser transmontanus) muscle aldolase. The close homology with the rabbit muscle enzyme was used to align the peptides of the other aldolases from which the sequences were derived. The results also allowed a partial sequence for the N-terminal 39 residues for the ox liver enzyme to be deduced. 2. In the light of the strong homology evinced for these enzymes, a re-investigation of the amino acid sequence of rabbit muscle aldolase between residues 181 and 185 was undertaken. This indicated the presence of a hitherto unsuspected -Ile-Val-sequence between residues 181 and 182 and the need to invert the sequence -Glu-Val- to -Val-Glx- at positions 184 and 185. 3. Comparison of the available amino acid sequences of these enzymes suggested an early evolutionary divergence of the genes for muscle and liver aldolases. It was also consistent with other evidence that the central region of the primary structure of these enzymes (which includes the active-site lysine-227) forms part of a conserved folding domain in the protein subunit. 4. Detailed evidence for the amino acid sequences proposed has been deposited as Suy Lending Division, Boston Spa, Wetherby, West Yorkshire LS23 7BQ, U.K., from whom copies can be obtained on the terms indicated in Biochem. J. (1978) 169, 5.