Large-scale purification of human BACE expressed in mammalian cells and removal of the prosegment with HIV-1 protease to improve crystal diffraction.

BACE, or beta-secretase, is an attractive target in the treatment of Alzheimer's Disease because of its involvement in the generation of amyloid beta peptides. BACE is a type I transmembrane aspartyl protease composed of pre-, pro-, catalytic, transmembrane and cytoplasmic domains. For the present study, the coding sequence was truncated just before the transmembrane domain and the resulting construct was extended with the C-terminal addition of a (His)(6) and expressed in several mammalian host cells. The enzyme expressed in CHO cells had the best crystallographic behavior and was purified in large quantities in a three step procedure. The purified BACE was comprised of two forms, namely the full length proBACE construct beginning with Thr(1), and a derivative missing the first 24 amino acids beginning with E(25). These BACE precursors co-crystallized in the presence of inhibitors yielding structures to 3.2 A resolution. HIV-1 protease treatment of this mixture resulted in complete cleavage of the F(39)-V(40) bond, leaving the V(40)EM...ES(432) (His)(6) derivative that was purified yielding an enzyme that was no more active than untreated BACE but co-crystallized with inhibitors producing well shaped, bipyramidal co-crystals diffracting to 2.6 A resolution.

BACE knockout mice are healthy despite lacking the primary beta-secretase activity in brain: implications for Alzheimer's disease therapeutics.

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by accumulation of amyloid plaques and neurofibrillary tangles in the brain. The major components of plaque, beta-amyloid peptides (Abetas), are produced from amyloid precursor protein (APP) by the activity of beta- and gamma-secretases. beta-secretase activity cleaves APP to define the N-terminus of the Abeta1-x peptides and, therefore, has been a long- sought therapeutic target for treatment of AD. The gene encoding a beta-secretase for beta-site APP cleaving enzyme (BACE) was identified recently. However, it was not known whether BACE was the primary beta-secretase in mammalian brain nor whether inhibition of beta-secretase might have effects in mammals that would preclude its utility as a therapeutic target. In the work described herein, we generated two lines of BACE knockout mice and characterized them for pathology, beta-secretase activity and Abeta production. These mice appeared to develop normally and showed no consistent phenotypic differences from their wild-type littermates, including overall normal tissue morphology and brain histochemistry, normal blood and urine chemistries, normal blood-cell composition, and no overt behavioral and neuromuscular effects. Brain and primary cortical cultures from BACE knockout mice showed no detectable beta-secretase activity, and primary cortical cultures from BACE knockout mice produced much less Abeta from APP. The findings that BACE is the primary beta-secretase activity in brain and that loss of beta-secretase activity produces no profound phenotypic defects with a concomitant reduction in beta-amyloid peptide clearly indicate that BACE is an excellent therapeutic target for treatment of AD.

BACE2 functions as an alternative alpha-secretase in cells.

BACE1 and BACE2 define a new subfamily of membrane-anchored aspartyl proteases. Both endoproteases share similar structural organization including a prodomain, a catalytic domain formed via DTG and DSG active site motifs, a single transmembrane domain, and a short C-terminal tail. BACE1 has been identified as the Alzheimer's beta-secretase, whereas BACE2 was mapped to the Down's critical region of human chromosome 21. Herein we show that purified BACE2 can be autoactivated in vitro. Purified BACE2 cleaves human amyloid precursor protein (APP) sequences at the beta-secretase site, and near the alpha-secretase site, mainly at A beta-Phe(20)--Ala(21) and also at A beta-Phe(19)--Phe(20). Alternatively, in cells BACE2 has a limited effect on the beta-secretase site but efficiently cleaves the sequences near the alpha-secretase site. The in vitro specificity of APP processing by BACE2 is distinct from that observed in cells. BACE2 localizes in the endoplasmic reticulum, Golgi, trans-Golgi network, endosomes, and plasma membrane, and its cellular localization patterns depend on the presence of its transmembrane domain. BACE2 chimeras that increase localization of BACE2 in the trans-Golgi network do not change its APP processing patterns. Thus, BACE2 can be distinguished from BACE1 on the basis of autoprocessing of the prosegment, APP processing specificity, and subcellular localization patterns.

Membrane-anchored aspartyl protease with Alzheimer's disease beta-secretase activity.

Mutations in the gene encoding the amyloid protein precursor (APP) cause autosomal dominant Alzheimer's disease. Cleavage of APP by unidentified proteases, referred to as beta- and gamma-secretases, generates the amyloid beta-peptide, the main component of the amyloid plaques found in Alzheimer's disease patients. The disease-causing mutations flank the protease cleavage sites in APP and facilitate its cleavage. Here we identify a new membrane-bound aspartyl protease (Asp2) with beta-secretase activity. The Asp2 gene is expressed widely in brain and other tissues. Decreasing the expression of Asp2 in cells reduces amyloid beta-peptide production and blocks the accumulation of the carboxy-terminal APP fragment that is created by beta-secretase cleavage. Solubilized Asp2 protein cleaves a synthetic APP peptide substrate at the beta-secretase site, and the rate of cleavage is increased tenfold by a mutation associated with early-onset Alzheimer's disease in Sweden. Thus, Asp2 is a new protein target for drugs that are designed to block the production of amyloid beta-peptide peptide and the consequent formation of amyloid plaque in Alzheimer's disease.

Nucleotide sequence analysis of the human KCNJ1 potassium channel locus.

Detailed analyses of transcripts encoding various isoforms of the human potassium (K+, inward rectifying) channel ROM-K (also referred to as K(ir)1.1) revealed the existence of at least five distinct transcripts [Shuck et al., J. Biol. Chem. 269 (1994) 24261-24270]. These five hROM-K transcripts appear to be the result of alternative splicing of five exons. The nucleotide sequence of the genomic DNA including and spanning these exons (the KCNJ1 locus) was obtained directly from lambda and P1 clones (a total of 40 kb). The organization of the hKCNJ1 gene was determined by combining this sequence information with data obtained from primer extension and RT-PCR experiments. It appears that the hKCNJ1 gene utilizes multiple promoters, with promoter-like elements found 5' of exons 1, 4, or 5. The promoter 5' of exon 5 was unexpected; thus, it appears that the hKCNJ1 gene is capable of producing six distinct hROM-K transcripts via the use of three promoters and alternative splicing of five exons. Comparisons of the rat and human ROM-K cDNA sequences find human homologs (orthologs) for two of the three distinct rROM-K transcripts. A search of the complete human KCNJ1 sequence with the exon sequence that defines the other rROM-K transcript located a region of shared nucleotides, a putative sixth exon, in the hKCNJ1 gene. This finding suggests that the rKCNJ1 gene may contain an exon that is no longer or infrequently used in transcripts derived from the hKCNJ1 gene.

Cloning and characterization of two K+ inward rectifier (Kir) 1.1 potassium channel homologs from human kidney (Kir1.2 and Kir1.3).

The DNA sequence encoding the rat brain inward rectifier-10 K+ channel was amplified from rat brain RNA using reverse transcription-polymerase chain reaction and used to clone the human homolog. Low stringency screening of a human kidney cDNA library and subsequent DNA sequence analysis identified two related K+ inward rectifier cDNAs, referred to as Kir1.2 and Kir1.3, which were derived from transcription of distinct human genes. Kir1.2 represents the human homolog of the rat BIRK-10 sequence, whereas Kir1.3 was unique compared with all available sequence data bases. The genes that encode Kir1.2 and Kir1.3 were mapped to human chromosomes 1 and 21, respectively. Both genes showed tissue-specific expression when analyzed by Northern blots. Kir1.2 was only detected in brain >> kidney and was detected at high levels in all brain regions examined. Kir1.3 was most readily detected in kidney and was also expressed in pancreas > lung. Comparative analysis of the predicted amino acid sequences for Kir1.2 and Kir1.3 revealed they were 62% identical. The most remarkable difference between the two polypeptides is that the Walker Type A consensus binding motif present in both Kir1.1 and Kir1.2 was not conserved in the Kir1.3 sequence. Expression of the Kir1.2 polypeptide in Xenopus oocytes resulted in the synthesis of a K+-selective channel that exhibited an inwardly rectifying current-voltage relationship and was inhibited by external Ba2+ and Cs+. Kir1.2 current amplitude was reduced by >85% when the pH was decreased from pH 7.4 to 5.9 using the membrane-permeant buffer acetate but was relatively unaffected when pH was similarly lowered using membrane-impermeant biphthalate. The inhibition by intracellular protons was voltage-independent with an IC50 of pH 6.2 and a Hill coefficient of 1.9, suggesting the cooperative binding of 2 protons to the intracellular face of the channel. In contrast, Kir1.3 expression in Xenopus oocytes was not detectable despite the fact that the cRNA efficiently directed the synthesis of a polypeptide of the expected Mr in an in vitro translation system. Co-expression of Kir1.3 with either Kir1.1 or Kir1.2 reduced currents resulting from expression of these inward-rectifier subunits alone, consistent with a dominant negative influence on Kir1.1 and Kir1.2 expression.

Intracellular IL-1 receptor antagonist promoter: cell type-specific and inducible regulatory regions.

The objective of these studies was to examine the molecular mechanisms involved in transcriptional regulation of the gene for the intracellular structural variant of the IL-1 receptor antagonist (icIL-1Ra) molecule. By reverse transcription-PCR analysis, constitutive expression of endogenous icIL-1Ra mRNA was observed in the epithelial cell lines A431 and HT-29, but not in the macrophage cell lines RAW 264.7 and U937, or in the lymphocyte cell lines Raji and Jurkat. However, icIL-1Ra mRNA expression was observed in response to stimulation with LPS in RAW 264.7 cells and to PMA and LPS in U937 cells. To examine the mechanisms of transcriptional regulation, 4.5 kb of the 5' flanking sequence was isolated from the human icIL-1Ra gene, sequenced, cloned into a luciferase expression vector (pIC4525.Luc), and examined in transfection studies. The pIC4525.Luc construct exhibited a pattern of expression in epithelial and macrophage cell lines similar to that of the endogenous icIL-1Ra gene. To obtain a generalized map of cell type-specific and inducible cis-acting DNA elements, nested 5' deletional mutants of the icIL-1Ra promoter were constructed. Results from transfection studies with these icIL-1Ra promoter/luciferase fusion constructs indicated that constitutive expression in epithelial cells was under the control of three positively acting regions located between bases -4525 to -1438, -288 to -156, and -156 to -49. In contrast, basal expression of pIC4525.Luc in transfected but unstimulated RAW 264.7 cells was under the control of a weak inhibitory region located between bases -4525 to -1438 and a strong positive element between -156 and -49. LPS induction of icIL-1Ra transcription in RAW 264.7 cells was regulated by strong positively acting DNA regions between bases -1438 to -909 and -156 to -49. In summary, the proximal region of the icIL-1Ra promoter, between bases -156 to -49, contains positive cis-acting elements that are needed for expression in both epithelial and monocyte cell lines. However, our results indicate that the ability of this proximal promoter region to control expression is strongly influenced, both positively and negatively, by other upstream cis-acting elements in a cell type-specific manner.

ROMK inwardly rectifying ATP-sensitive K+ channel. II. Cloning and distribution of alternative forms.

The rat ROMK gene encodes inwardly rectifying, ATP-regulated K+ channels [K. Ho, C. G. Nichols, W. J. Lederer, J. Lytton, P. M. Vassilev, M. V. Kanazirska, and S. C. Hebert. Nature Lond. 362: 31-38, 1993; H. Zhou, S. S. Tate, and L. G. Palmer. Am. J. Physiol. 266 (Cell Physiol. 35): C809-C824, 1994], and mRNA encoding these channels is widely expressed in distal cortical and outer medullary nephron segments [see companion study; W.-S. Lee and S. C. Hebert. Am. J. Physiol. 268 (Renal Fluid Electrolyte Physiol. 37): F1124-F1131, 1995]. Using approaches based on homology to ROMK1, we have identified two additional ROMK isoforms, ROMK2b and ROMK3. Analysis of the nucleotide sequences of the ROMK isoforms indicates that molecular diversity of ROMK transcripts is due to alternative splicing at both the 5'-coding and 3'-noncoding regions. The splicing at the 5' end of ROMK gives rise to channel proteins with variable-length NH2 termini containing different initial amino acid sequences. Functional expression of these isoforms in Xenopus oocytes showed that they form functional Ba(2+)-sensitive K+ channels. The nephron distribution of mRNAs encoding alternatively spliced isoforms of ROMK (ROMK1-ROMK3) was investigated by reverse transcription-polymerase chain reaction (RT-PCR) of nephron segments dissected from rat kidney. Nondegenerate PCR primer pairs were designed to span at least one intron and to amplify specific alternatively spliced forms of ROMK.(ABSTRACT TRUNCATED AT 250 WORDS)

Intracellular H+ inhibits a cloned rat kidney outer medulla K+ channel expressed in Xenopus oocytes.

The pH sensitivity of a cloned rat kidney K+ channel, ROMK1, was examined after expression in Xenopus oocytes. Membrane currents and intracellular pH (pHi) were concomitantly monitored by the two-microelectrode voltage-clamp technique and a pH-sensitive microelectrode. Oocytes injected with ROMK1 cRNA developed a hyperpolarized resting potential of -98.7 +/- 0.98 mV and a slightly inwardly rectifying Ba(2+)-sensitive K+ current. Lowering external pH from 7.4 to 6.7 using membrane-permeable acetate buffer reduced measured pHi from 7.2 to 6.6 and reduced the ROMK1 current by 80%. The H+ blockade of ROMK1 currents was voltage independent. The relationship between ROMK1 slope conductance and pHi fitted to a titration curve suggested binding of four H+ to a site with a pK of 6.79. Extracellular acidification from pH 7.4 to 6.0 using membrane-impermeable biphthalate buffer had no effect on the ROMK1 current. The pH sensitivity of the ROMK1 channel is similar to that reported for a small-conductance native kidney K+ channel.

Characterization of U-92016A as a selective, orally active, high intrinsic activity 5-hydroxytryptamine1A agonist.

The purpose of the present study was to characterize U-92016A [(+)-R)-2-cyano-N,N-dipropyl-8-amino-6,7,8,9-tetrahydro-3H-benz[e] indole] as a 5-hydroxytryptamine (5-HT)1A receptor agonist and to compare its activity with that of standard 5-HT1A receptor agonists. U-92016A binds with high affinity to human 5-HT1A receptors expressed in Chinese hamster ovary cells (Ki = 0.2 nM). Radioligand binding studies also indicate that U-92016A is selective for the 5-HT1A receptor over other biogenic amine receptors. In Chinese hamster ovary cells expressing the human 5HT1A receptor, U-92016A decreased the forskolin-induced increase in cyclic AMP synthesis and had an intrinsic activity of 0.82 relative to 5-HT. U-92016A potently decreased rectal temperature in mice. The maximum temperature decrease was significantly greater than that observed for 8-hydroxy-di-n-propyl aminotetralin, buspirone, gepirone, ipsapirone or flesinoxan. U-92016A also elicited the 5-HT-mediated syndrome in rats and resulted in a dose-related decrease in 5-hydroxytryptophan accumulation. The compound also decreased arterial blood pressure in spontaneously hypertensive rats and inhibited sympathetic nerve activity in cats. In these assays U-92016A displayed excellent potency and a long duration of action. U-92016A also inhibited the firing of dorsal raphe 5-HT neurons and was active in two social interaction assays. The p.o. bioavailability of U-92016A was calculated to be 45%. Taken together, these data indicate that U-92016A is a metabolically stable, p.o. active 5-HT1A receptor agonist with an exceptionally high degree of intrinsic activity.

Cloning and characterization of multiple forms of the human kidney ROM-K potassium channel.

The rat kidney ROM-K1 potassium channel cDNA was used to clone the homolog from human kidney using a combination of cDNA cloning, reverse transcriptase-polymerase chain reaction (RT-PCR), and primer extension cloning methods. In addition to the human species homolog of ROM-K1, four additional transcripts that are formed by alternative splicing of a single human gene were also characterized (hROM-K2 to hROM-K5). All five transcripts share a common 3' exon that encodes the majority of the channel protein and in three of the isoforms translation is initiated at a start codon contained within this exon (hROM-K2, hROM-K4, and hROM-K5). The two other transcripts contain additional exons that potentially extend the open reading frame by either 19 amino acid residues (hROM-K1) or by 17 amino acid residues (hROM-K3). Comparison of the translation products from the three representative transcripts (hROM-K1, hROM-K2, and hROM-K3) confirmed that hROM-K1 gave the largest product (41.6 kDa) and was translated more efficiently than either hROM-K2 or hROM-K3. Also, despite the presence of several additional canonical acceptor sites for Asn-linked glycosylation relative to rat ROM-K1, all three channel polypeptides were glycosylated to a similar extent in the in vitro translation reactions when canine pancreatic microsomes were included. A survey of the tissue distribution of expression of the various forms in selected human tissues showed that the core-exon linked to all four possible 5' exons are detected almost exclusively in kidney. The core-exon was also detected in human kidney and lower amounts were detected in skeletal muscle > pancreas > spleen > brain = heart > liver RNAs by RT-PCR. Alternatively, Northern blot analysis of poly(A)+ RNAs from these same tissues revealed a 2.8-kilobase transcript only in kidney. Heterologous expression of either the hROM-K1, hROM-K2, or hROM-K3 channel transcripts in Xenopus oocytes led to the expression of K(+)-selective, Ba(2+)-sensitive inwardly rectifying channels as measured by whole cell currents. At this level of analysis, the channel properties of the individual forms could not be distinguished.

Cloning, sequencing and phylogenetic analysis of a human 5-hydroxytryptamine 1D receptor pseudogene.

A third member of the human 5HT1D gene family has been identified using a combination of homology cloning and DNA sequence analysis. This human gene is most related to the 5HT1D alpha subtype (77% shared identity) and is a pseudogene, based on the lack of an open reading frame (ORF) caused by multiple in-frame stop codons and nucleotide (nt) deletions relative to the functional 5HT1D alpha gene (encoding the 5-hydroxytryptamine 1D alpha receptor). The 5HT1D pseudogene also contained an insertion that shares 87% identity to the Alu consensus sequence. Phylogenetic analysis of the three human genes in this family reveals that although the two functional genes, 5HT1D alpha and 5HT1D beta, are detected in all mammalian species examined, the 5HT1D pseudogene is only detected in a subset of primates (catarrhines) that evolved approximately 35-45 million years (Myr) ago. Alternatively, based on the 23% divergence between the functional 5HT1D alpha gene and the 5HT1D pseudogene, we estimate that these two genes began to diverge approximately 50 Myr ago.

Cloning, heterologous expression and characterization of murine interleukin 1 receptor antagonist protein.

A cDNA coding for the human interleukin 1 receptor antagonist protein (IL 1Ra) was used to clone the corresponding murine cDNA. The nucleotide sequence of the open reading frame coding for the processed form of mIL 1Ra predicted a 152-residue protein that was 77% identical to human IL 1Ra. The cellular and tissue distribution of murine IL 1Ra (mIL 1Ra) transcripts showed high levels in macrophages and skin while lower levels were detected in tissues that contain significant numbers of resident macrophages. The portion of the mIL 1Ra cDNA that codes for the mature form of the protein was placed under the control of a Trp promoter and expressed in E. coli at a level of 37% of total cell protein. The expressed protein was secreted into the periplasm and was purified to homogeneity in a single step by cation-exchange chromatography. Recombinant mIL 1Ra competitively inhibited 125I-labeled IL 1 alpha binding to murine type I IL 1R present on EL4 6.1 cells (Ki value of 0.21 nM) and antagonized IL 1-stimulated co-mitogenesis in murine thymocytes (0.7 x 10(6)-1.1 x 10(6) units/mg).