Enrichment of integral membrane proteins from small amounts of brain tissue

Subcellular fractionation represents an essential technique for functional proteome analysis. Recently, we provided a subcellular fractionation protocol for minute amounts of tissue that yielded a nuclear fraction, a membrane and organelle fraction, and a cytosolic fraction. In the current study, we attempted to improve the protocol for the isolation of integral membrane proteins, as these are particularly important for brain function. In the membrane and organelle fraction, we increased the yield of membranes and organelles by about 50% by introducing a single re-extraction step. We then tested two protocols towards their capacity to enrich membrane proteins present in the membrane and organelle fraction. One protocol is based on sequential solubilization using subsequent increases of chaotropic conditions such as urea, thereby partitioning hydrophobic proteins from hydrophilic ones. The alternative protocol applies high-salt and high-pH washes to remove non-membrane proteins. The enrichment of membrane proteins by these procedures, as compared to the original membrane and organelle fraction, was evaluated by 16-BAC-SDS-PAGE followed by mass spectrometry of randomly selected spots. In the original membrane and organelle fraction, 7 of 50 (14%) identified proteins represented integral membrane proteins, and 15 (30%) were peripheral membrane proteins. In the urea-soluble fraction, 4 of 33 (12%) identified proteins represented integral membrane proteins, and 10 (30%) were peripheral membrane proteins. In the high-salt/high-pH resistant sediment, 12 of 45 (27%) identified proteins were integral membrane proteins and 13 (29%) represented peripheral membrane proteins. During the analysis, several proteins involved in neuroexocytosis were detected, including syntaxin, NSF, and Rab3-interaction protein 2. Taken together, differential centrifugation in combination with high-salt and high-pH washes resulted in the highest enrichment of integral membrane proteins and, therefore, represents an adequate technique for region-specific profiling of membrane proteins in the brain.

Functional hemizygosity of PAFAH1B3 due to a PAFAH1B3-CLK2 fusion gene in a female with mental retardation, ataxia and atrophy of the brain

We report on the molecular characterization of a translocation t(1;19)(q21.3;q13.2) in a female with mental retardation, ataxia and atrophy of the brain. Sequence analysis of the breakpoints revealed an ALU:-repeat-mediated mechanism of recombination that led to truncation of two genes: the kinase CLK2 and PAFAH1B3, the gene product of which interacts with LIS1 as part of a heterotrimeric G protein complex PAF-AH1B. In addition, two reciprocal fusion genes are present. One expressed fusion gene encodes the first 136 amino acids of PAFAH1B3 followed by the complete CLK2 protein. Truncated PAFAH1B3 protein lost its potential to interact with LIS1 whereas CLK2 activity was conserved within the fusion protein. These data emphasize the importance of PAF-AH1B in brain development and functioning and demonstrate the first fusion gene apparently not associated with cancer.

Molecular cloning of Xp11 breakpoints in two unrelated mentally retarded females with X;autosome translocations

Mental retardation is a very common and extremely heterogeneous disorder that affects about 3% of the human population. Its molecular basis is largely unknown, but many loci have been mapped to the X chromosome. We report on two mentally retarded females with X;autosome translocations and breakpoints in Xp11, viz., t(X;17)(p11;p13) and t(X;20)(p11;q13). (Fiber-) FISH analysis assigned the breakpoints to different subbands, Xp11.4 and Xp11.23, separated by approximately 8 Mb. High-resolution mapping of the X- chromosome breakpoints using Southern blot hybridization resulted in the isolation of breakpoint-spanning genomic subclones of 3 kb and 0. 5 kb. The Xp11.4 breakpoint is contained within a single copy sequence, whereas the Xp11.23 breakpoint sequence resembles an L1 repetitive element. Several expressed sequences map close to the breakpoints, but none was found to be inactivated. Therefore, mechanisms other than disruption of X-chromosome genes likely cause the phenotypes.

A translocation breakpoint cluster disrupts the newly defined 3' end of the SNURF-SNRPN transcription unit on chromosome 15

Balanced translocations affecting the paternal copy of 15q11--q13 are a rare cause of Prader-Willi syndrome (PWS) or PWS-like features. Here we report on the cytogenetic and molecular characterization of a de novo balanced reciprocal translocation t(X;15)(q28;q12) in a female patient with atypical PWS. The translocation breakpoints in this patient and two previously reported patients map 70-80 kb distal to the SNURF-SNRPN gene and define a breakpoint cluster region. The breakpoints disrupt one of several hitherto unknown 3' exons of this gene. Using RT--PCR we demonstrate that sequences distal to the breakpoint, including the recently identified C/D box small nucleolar RNA (snoRNA) gene cluster HBII-85 as well as IPW and PAR1, are not expressed in the patient. Our data suggest that lack of expression of these sequences contributes to the PWS phenotype.

Mutation of CDH23, encoding a new member of the cadherin gene family, causes Usher syndrome type 1D

Usher syndrome type I (USH1) is an autosomal recessive disorder characterized by congenital sensorineural hearing loss, vestibular dysfunction and visual impairment due to early onset retinitis pigmentosa (RP). So far, six loci (USH1A-USH1F) have been mapped, but only two USH1 genes have been identified: MYO7A for USH1B and the gene encoding harmonin for USH1C. We identified a Cuban pedigree linked to the locus for Usher syndrome type 1D (MIM 601067) within the q2 region of chromosome 10). Affected individuals present with congenital deafness and a highly variable degree of retinal degeneration. Using a positional candidate approach, we identified a new member of the cadherin gene superfamily, CDH23. It encodes a protein of 3,354 amino acids with a single transmembrane domain and 27 cadherin repeats. In the Cuban family, we detected two different mutations: a severe course of the retinal disease was observed in individuals homozygous for what is probably a truncating splice-site mutation (c.4488G-->C), whereas mild RP is present in individuals carrying the homozygous missense mutation R1746Q. A variable expression of the retinal phenotype was seen in patients with a combination of both mutations. In addition, we identified two mutations, Delta M1281 and IVS51+5G-->A, in a German USH1 patient. Our data show that different mutations in CDH23 result in USH1D with a variable retinal phenotype. In an accompanying paper, it is shown that mutations in the mouse ortholog cause disorganization of inner ear stereocilia and deafness in the waltzer mouse.

Mutations in ARHGEF6, encoding a guanine nucleotide exchange factor for Rho GTPases, in patients with X-linked mental retardation

X-linked forms of mental retardation (XLMR) include a variety of different disorders and may account for up to 25% of all inherited cases of mental retardation. So far, seven X-chromosomal genes mutated in nonspecific mental retardation (MRX) have been identified: FMR2, GDI1, RPS6KA3, IL1RAPL, TM4SF2, OPHN1 and PAK3 (refs 2-9). The products of the latter two have been implicated in regulation of neural plasticity by controlling the activity of small GTPases of the Rho family. Here we report the identification of a new MRX gene, ARHGEF6 (also known as alphaPIX or Cool-2), encoding a protein with homology to guanine nucleotide exchange factors for Rho GTPases (Rho GEF). Molecular analysis of a reciprocal X/21 translocation in a male with mental retardation showed that this gene in Xq26 was disrupted by the rearrangement. Mutation screening of 119 patients with nonspecific mental retardation revealed a mutation in the first intron of ARHGEF6 (IVS1-11T-->C) in all affected males in a large Dutch family. The mutation resulted in preferential skipping of exon 2, predicting a protein lacking 28 amino acids. ARHGEF6 is the eighth MRX gene identified so far and the third such gene to encode a protein that interacts with Rho GTPases.

Molecular cloning of the critical region for glomerulopathy with fibronectin deposits (GFND) and evaluation of candidate genes

Glomerulopathy with fibronectin deposits (GFND, MIM 601894) is an autosomal dominant kidney disease that leads to terminal renal failure at a median age of 47 years. It represents a distinct entity of membranoproliferative glomerulonephritis (MPGN) type III and is characterized by the unique feature of massive glomerular deposits of fibronectin. We have recently localized a gene locus for GFND to human chromosome 1q32 by total genome linkage analysis in a large kindred, within a 4.1-cM critical interval between markers D1S2872 and D1S2891. This interval contains a cluster of genes for "regulators of complement activation" (RCA), which represent strong candidates for GFND. To identify positional candidate genes for GFND within the critical genetic interval, we here report the cloning of the entire critical GFND region in a complete YAC and partial PAC contig. We constructed a high-resolution transcriptional map, thereby defining positional and functional candidate genes for the disease. To evaluate their role in GFND, we performed functional studies on RCA proteins in GFND patients from the large kindred, as well as mutational analysis of the genes for complement receptor-2 (CR2), membrane cofactor protein (MCP), and decay accelerating factor (DAF). Although no loss-of-function mutation has been identified as yet, these data provide a basis for the examination of candidate genes for GFND and other genes for MPGN, which localize to the vicinity of the GFND region.

Isolation of two novel human RhoGEFs, ARHGEF3 and ARHGEF4, in 3p13-21 and 2q22

RhoGEFs play an important role in various signaling cascades and are implicated in human conditions like cancer and mental retardation. A database search combined with screening of a human neuronal teratocarcinoma library identified two novel RhoGEFs, ARHGEF3 and ARHGEF4 (HGMW-approved symbols). The widely expressed ARHGEF3 transcript of 3561 nucleotides encodes a polypeptide of 526 amino acids with homology to NET1. The ARHGEF4 gene generates two transcripts of 3665 and 4000 nucleotides that translate into 720 amino acid residues. Expression of ARHGEF4 is restricted to brain and the encoded protein shows homology to collybistin. FISH analysis of genomic clones mapped ARHGEF3 to 3p13-21 and ARHGEF4 to 2q22.

DXS6673E encodes a predominantly nuclear protein, and its mouse ortholog DXHXS6673E is alternatively spliced in a developmental- and tissue-specific manner

DXS6673E is a candidate gene for nonspecific X-linked mental retardation and encodes a novel Zn-finger protein. The ortholog murine gene DXHXS6673E in XC-D was isolated and characterized. It is ubiquitously expressed in all embryonic stages and adult tissues. Two different transcription start sites exist that result in two major transcripts of 6055 and 5352 nucleotides, each composed of 25 exons. Exon 1A is tissue specific, whereas exon 1B is transcribed constitutively. Both variants are translated into the same 1370-amino-acid protein. Transcripts are subject to alternative splicing at the 5'-end. Some of the isoforms are developmental stage and tissue specific. Among them, one was present only in embryos and adult brain. Sequence analysis demonstrated evolutionary conservation down to the arthropods and defined several conserved protein motifs. Subcellular localization studies with green fluorescent protein as a reporter showed that DXS6673E is predominantly located in the nucleus due to several functional nuclear localization signals. Three distinct protein distribution patterns in COS-7 cells could be identified.

Systematic characterisation of disease associated balanced chromosome rearrangements by FISH: cytogenetically and genetically anchored YACs identify microdeletions and candidate regions for mental retardation genes

Disease associated balanced chromosome rearrangements (DBCRs) have been instrumental in the isolation of many disease genes. To facilitate the molecular cytogenetic characterisation of DBCRs, we have generated a set of >1200 non-chimeric, cytogenetically and genetically anchored CEPH YACs, on average one per 3 cM, spaced over the entire human genome. By fluorescence in situ hybridisation (FISH), we have performed a systematic search for YACs spanning translocation breakpoints. Patients with DBCRs and either syndromic or non-syndromic mental retardation (MR) were ascertained through the Mendelian Cytogenetics Network (MCN), a collaborative effort of, at present, 270 cytogenetic laboratories throughout the world. In this pilot study, we have characterised 10 different MR associated chromosome regions delineating candidate regions for MR. Five of these regions are narrowed to breakpoint spanning YACs, three of which are located on chromosomes 13q21, 13q22, and 13q32, respectively, one on chromosome 4p14, and one on 6q25. In two out of six DBCRs, we found cytogenetically cryptic deletions of 3-5 Mb on one or both translocation chromosomes. Thus, cryptic deletions may be an important cause of disease in seemingly balanced chromosome rearrangements that are associated with a disease phenotype. Our region specific FISH probes, which are available to MCN members, can be a powerful tool in clinical cytogenetics and positional cloning.

Mouse Dac, a novel nuclear factor with homology to Drosophila dachshund shows a dynamic expression in the neural crest, the eye, the neocortex, and the limb bud

Dac is a novel nuclear factor in mouse and humans that shares homology with Drosophila dachshund (dac). Alignment with available sequences defines a conserved box of 117 amino acids that shares weak homology with the proto-oncogene Ski and Sno. Dac expression is found in various neuroectodermal and mesenchymal tissues. At early developmental stages Dac is expressed in lateral mesoderm and in neural crest cells. In the neural plate/tube Dac expression is initially seen in the prosencephalon and gets gradually restricted to the presumptive neocortex and the distal portion of the outgrowing optic vesicle. Furthermore, Dac transcripts are detected in the mesenchyme underlying the Apical Ectodermal Ridge (AER) of the extending limb bud, the dorsal root ganglia and chain ganglia, and the mesenchyme of the growing genitalia. Dac expression in the Gli 3 mutant extra toes (Xt/Xt) shows little difference compared to the expression in wild-type limb buds. In contrast, a significant expansion of Dac expression are observed in the anterior mesenchyme of the limb buds of hemimelic extra toes (Hx/+) mice. FISH analysis reveals that human DAC maps to chromosome 13q22.3-23 and further fine-mapping defined a position of the DAC gene at 54cM or 13q21.1, a locus that associates with mental retardation and skeletal abnormalities.

Identification of positional candidates for neurological disorders on chromsome 13q14-->q22

In the course of a research project aimed at the molecular characterization of balanced chromosome rearrangements associated with mental retardation (MR), several YACs spanning MR-associated chromosomal rearrangements in the 13q14-->q22 region were identified. To facilitate the search for relevant candidate genes, we have analyzed a total of 102 EST clones from this region. Sequence comparisons revealed that these 102 clones represent up to 72 distinct transcripts. When no physical mapping data were available, a minimal YAC contig was screened for each unique transcript by the polymerase chain reaction (PCR) or hybridization. Fifty-eight independent ESTs could be localized to YAC clones between the markers D13S1248 and D13S1201. Several ESTs are located on YAC clones detecting chromosomal rearrangements in MR patients. One EST was mapped within the critical region for Rieger syndrome type 2, and three transcripts were identified in the region for the nocturnal enuresis type 1. Some ESTs showed homologies to known genes, including the cadherin-related tumor suppressor gene from Drosophila, the yeast mitotic control protein DIS3, and the human alpha-2-macroglobulin receptor associated protein.

Identification of a novel Ran binding protein 2 related gene (RANBP2L1) and detection of a gene cluster on human chromosome 2q11-q12

The giant 358-kDa protein Ran binding protein 2 (RanBP2/Nup358) is localized at the cytoplasmic side of the nuclear pore complex and likely constitutes the Ran-GTP binding site at the cytoplasmic face of the complex. RanBP2/Nup358 furthermore acts as a chaperone for red/green opsin molecules. Here, we report on the physical mapping of human RanBP2 between markers D2S340 and D2S1893. A duplication of the 5'-end sequence of RanBP2 occurs within 3 Mb distal to RanBP2. Detailed sequence analysis resulted in primers specific for this distal duplication. Polymerase chain reaction-based screening of cDNA libraries indicates that this transcript, called RanBP2alpha (HGMW-approved symbol RANBP2L1), is expressed in several tissues. Screening of a fetal brain cDNA library yielded a 4057-bp partial cDNA clone for RanBP2alpha. Its 5'-end is almost identical to RanBP2, whereas its 3'-part is distinct from RanBP2. Northern blot analysis using a probe of the 3'-untranslated sequence of RanBP2alpha detected in several tissues an 8-kb transcript representing the full length of the transcript. In pancreas and placenta, an additional transcript of 14 kb was detected. PAC clones containing the bona fide RanBP2 sequences were localized to 2q11-q12 by FISH analysis, and a region of high similarity was detected on 2p11-p12. In summary, we have identified a RanBP2 gene cluster on 2q11-q12 together with a novel gene termed RanBP2alpha, with high sequence similarity to RanBP2.

Construction of a gene map of the nephronophthisis type 1 (NPHP1) region on human chromosome 2q12-q13

A gene for the autosomal recessive kidney disorder juvenile nephronophthisis (NPH) is located on chromosome 2q between markers D2S1893 and D2S1888. Recently, the presence of large homozygous deletions was described in the majority of NPH patients. We constructed an integrated YAC/PAC contig of 54 markers and 30 PAC clones that encompasses this deletion and the flanking inverted duplication. Thirty-six novel sequence-tagged site markers were generated for this region of 2-3 Mb, 22 of which represent PAC ends. Ten of 18 multiplex NPH families show a homozygous deletion for 8 consecutive markers. BlastN database search and expressed sequence tag (EST) mapping led to the localization of 18 EST clones to the integrated contig, representing 11 putative transcribed sequences. Seven EST clones were localized to the NPHP1 region between D2S1893 and D2S1888. Two EST clones, zc07a11 from a human parathyroid tumor library and yy63e10 from a multiple sclerosis lesion library, are located in the deletion region. PCR amplification experiments indicate that zc07a11 represents a chimeric cDNA. Through FISH analysis the NPHP1 deletion region was localized to 2q12-q13. In summary, our study provides a high-resolution physical map of the NPHP1 region with 7 precisely localized expressed sequences, 2 of which have recently been shown to be part of a gene for NPH. These data will alleviate the identification of further genes of a homozygous gene deletion syndrome in patients with NPH and oculomotor apraxia and will be instrumental in the characterization of the molecular mechanism leading to the large homozygous deletion in this region. The data furthermore provide an important step toward the construction of a sequence-ready PAC contig of this region.

Lack of large, homozygous deletions of the nephronophthisis 1 region in Joubert syndrome type B. APN Study Group. Arbeitsgemeinschaft für Pädiatrische Nephrologie

Joubert syndrome type B (JSB) is a developmental disorder of the nephronophthisis (NPH) complex with multiple organ involvement, including NPH, coloboma of the eye, aplasia of the cerebellar vermis, and the facultative symptoms of psychomotor retardation, polydactyly, and neonatal tachypnea. In isolated autosomal recessive NPH type 1 (NPH1), homozygous deletions have been described as causative in more than 80% of patients. Since different combinations of the extrarenal symptoms with NPH occur in JSB, a contiguous gene deletion syndrome in the NPH1 genetic region would seem a highly likely cause for JSB. We therefore examined 11 families with JSB for the presence of extended deletions at the NPH1 locus. Genomic DNA was examined using four consecutive polymerase chain reaction (PCR) markers that are deleted in NPH1 and three PCR makers flanking the NPH1 deletion. In all seven markers examined, there was no homozygous deletion detected in any of the 11 JSB families studied. Since these markers saturate the NPH1 deletion region at high density, this finding excludes the presence of large homozygous deletions of the NPH1 region in these JSB families, making it unlikely that deletions of the NPH1 region are a primary cause for JSB.

A novel gene encoding an SH3 domain protein is mutated in nephronophthisis type 1

Juvenile nephronophthisis (NPH), an autosomal recessive cystic kidney disease, is the primary genetic cause of chronic renal failure in children. About two thirds of patients with NPH carry a large homozygous deletion at the gene locus NPH1 on 2q13. We here identify a novel gene. NPHP1, which extends over most of this common deletion. The 4.5-kb transcript encodes a protein with an SH3 domain, which is highly conserved throughout evolution. The 11-kb interval between the 3' end of NPHP1 and an inverted repeat containing the distal deletion breakpoint was found to contain the first exon of a second gene, MALL. In patients with a hemizygous deletion of the NPH1 region, additional point mutations were found in NPHP1 but not in MALL.

Molecular cloning of the interleukin-1 gene cluster: construction of an integrated YAC/PAC contig and a partial transcriptional map in the region of chromosome 2q13

Genes of the interleukin-1 (IL-1) gene cluster localized on chromosome 2q13 are implicated in many physiological and pathophysiological processes. We present here a high-resolution physical map of this region between markers D2S2008 and D2S4/PAX8. An integrated YAC/PAC contig and a partial transcriptional map were constructed by STS-constent mapping using the CEPH YAC library and three PAC libraries. A total of 3 YACs, 34 PACs, and 56 STSs were integrated: 33 newly generated probes to PAC end sequences, 9 polymorphic and 4 nonpolymorphic markers, 5 known genes, 4 expressed sequence tags, and 1 pseudogene. Within the map, a complete PAC contig of > 1 Mb encompasses the IL-1 gene cluster and PAX8, a paired-box-containing gene. This allowed us to define the transcriptional orientation of GLVR1, IL1B, and IL1RN and to show that PAX8 is localized outside the IL-1 gene cluster. FISH analysis localized PAC clones containing the IL-1 gene cluster to 2q12-q13. The data provide the basis for further characterization of the IL-1 gene cluster and for the construction of a sequence-ready PAC contig of this region.

Molecular genetic identification of families with juvenile nephronophthisis type 1: rate of progression to renal failure. APN Study Group. Arbeitsgemeinschaft für Pädiatrische Nephrologie

Familial juvenile nephronophthisis (NPH), an autosomal recessive cystic disease of the kidney, is the most common genetic cause of end-stage renal disease (ESRD) in the first two decades of life. A gene locus for nephronophthisis type 1 (NPH1) has been mapped by linkage analysis to chromosome 2q13. We performed a haplotype analysis in 16 NPH families with at least two affected patients with the typical history, clinical signs and histology of NPH using microsatellite markers of the NPH1 genetic region. By demonstration of a recombinant event marker D2S1893 was identified as a novel centromeric flanking marker to the NPH1 critical genetic region. Absence of linkage to the NPH1 locus in six NPH families confirmed the existence of at least one additional gene locus for NPH. Linkage to the NPH1 locus was demonstrated in 10 families. In 8 of these families a homozygous deletion was identified. These data permit for the first time the study of the development of renal failure in a subset of NPH1 families, which is most likely homogeneous with regard to the responsible gene locus. We present a statistical description of serial serum creatinine measurements in NPH1. Analysis of renal death revealed a median of 13.1 years. Age-dependent quartiles were generated for serum creatinine. In summary, the new marker provides a diagnostic tool to aid in the diagnosis of NPH, while the progression charts offer a standard for an assessment of the rate of progression to ESRD for patients with NPH1 to be used in future therapeutic trials and for a prediction of the individual course of the disease.

Physical mapping of the gene for juvenile nephronophthisis (NPH1) by construction of a complete YAC contig of 7 Mb on chromosome 2q13

Familial juvenile nephronophthisis (NPH) is an autosomal recessive cystic disease of the kidney that leads to end-stage renal failure in adolescence. NPH is the most common genetic cause of end-stage renal disease in children. A gene locus for nephronophthisis (NPH1) has been mapped by linkage analysis to chromosome 2q13. We report here the construction of a complete YAC contig in the minimum genetic region for NPH1 by STS content mapping using clones of the CEPH YAC libraries. A physical map of maximum distances between 32 STS markers was constructed, thereby defining the order of a total of 27 STS markers. Since D2S340 and D2S121 have previously been identified as flanking markers to the NPH1 gene, the new contig defines on a physical map the NPH1 minimum genetic region to a 6.4-Mb interval. As a novel assignment, expressed genes, some of which may be candidates for the disease, were localized to the NPH1 region. In addition, the known interstitial telomeric repeat on chromosome 2 was physically mapped to this region. This contig assembly provides the basis for closer definition of the NPH1 critical region through identification of more narrow flanking markers and for the construction of a transcriptional map of the region towards isolation of the NPH1 gene.

Phylogenic relationships of the amino acid sequences of prosome (proteasome, MCP) subunits

Prosomes [or proteasomes, Multi-Catalytic Proteinase (MCP) are multisubunit protein complexes, found from archaebacteria to man, the structure of which (a 4-layer cylinder) is remarkable conserved. They were first observed as subcomplexes of untranslated mRNP, and then as a multicatalytic proteinase with several proteolytic activities. A number of sequences from subunits of these complexes are now available. Analysis of the sequences shows that these subunits are evolutionarily related, and reveals three highly conserved amino acid stretches. Based on a phylogenic approach, we propose to classify the sequenced subunits into 14 families, which fall into two superfamilies, of the alpha- and beta-type. These data, together with several recently published observations, suggest that some subunits may be interchangeable within the complexes, which would thus constitute a population of heterogenous particles.

Isolation and characterization of alpha-type HC3 and beta-type HC5 subunit genes of human proteasomes

Eukaryotic proteasomes from an evolutionarily conserved multi-gene family and are thought to have originated from a common ancestral gene and diverged into alpha-type and beta-type subgroups. To understand the molecular basis of the proteasome genes, we isolated and characterized two human proteasome genes econding the alpha-type HC3 and beta-type HC5 subunit. The functional genes for HC3 and HC5 are similar in being approximately 15 kb in length, but differ in having exon numbers of 9 and 6, respectively. Analyses of about 2.5 to 3.0 kb of the 5'-flanking regions of these two genes revealed the absence of TATA and CAAT promoter elements. However, two or three GC boxes were found. By analysis of the transcriptional regulatory activities in the 5'-flanking regions of the two genes, these GC boxes were found to function coordinately as promoters of the two genes. Interestingly, the HC3 gene possesses an additional silencer element in the 5'-upstream region near the first exon. This element is also able to repress the promoter activities of other genes, such as the HC5 and the type 1 glucose transporter genes, irrespective of whether it has a sense or antisense orientation, indicating that it acts as a general transcriptional silencer. The HC5 gene does not have this silence element, and its promoter activity is five to ten times that of HC3. These results show that the human proteasomal HC3 and HC5 genes differ not only in their genomic structures, such as their numbers of exons and their exon-intron organizations, but also in the mechanisms regulating their transcription, suggesting that they diverged at an early stage of evolution.

Sequence analyses and inter-species comparisons of three novel human proteasomal subunits, HsN3, HsC7-I and HsC10-II, confine potential proteolytic active-site residues

Proteasomes play a major role in non-lysosomal pathways of protein turnover mediated by distinct multiple proteolytic activities. Identification of their active-site residues is important for elucidating their catalytic mechanisms. Here we report the nucleotide sequences of three human proteasomal subunits, HsN3, HsC7-I and HsC10-II, coding for proteins with 264, 201 and 205 amino acid residues with calculated molecular weights of 29,192, 22,836 and 22,931, respectively. Sequence comparison showed that all three proteins belong to the beta-type superfamily and that they are the human counterparts of subunits reported to participate in the peptidyl-glutamyl-peptide hydrolyzing, chymotrypsin-like and trypsin-like activity of this complex. Alignments of the putative catalytically active subunits of various species revealed several family-specifically conserved serinyl residues within highly conserved amino acid stretches. Based on localization and hydrophobicity, the roles of these amino acid residues as active site and substrate binding site candidates are discussed.

cDNA cloning and interferon gamma down-regulation of proteasomal subunits X and Y

Proteasomes are the proteolytic complex responsible for major histocompatibility complex (MHC) class I-restricted antigen presentation. Interferon gamma treatment increases expression MHC-encoded LMP2 and LMP7 subunits of the proteasome and decreases expression of two proteasome subunits, named X and Y, which alters the proteolytic specificity of proteasomes. Molecular cloning of complementary DNAs encoding X and Y showed that their proteins are proteasomal subunits with high amino acid similarity to LMP7 and LMP2, respectively. Thus, interferon gamma may induce subunit replacements of X and Y by LMP7 and LMP2, respectively, producing proteasomes perhaps more appropriate for the immunological processing of endogenous antigens.

The prosomal RNA-binding protein p27K is a member of the alpha-type human prosomal gene family

Monoclonal antibodies demonstrated high conservation during evolution of a prosomal protein of M(r) 27,000 and differentiation--specific expression of the epitope. More than 90% of the reacting antigen was found as a p27K protein in the free messenger ribonucleoprotein (mRNP) fraction but another protein of M(r) 38,000, which shared protease fingerprint patterns with the p27K polypeptide, was also labelled in the nuclear and polyribosomal fractions. Sequencing of cDNA recombinant clones encoding the p27/38K protein and comparison with another prosomal protein, p30-33K, demonstrated the existence of a common characteristic sequence pattern containing three highly conserved segments. The genes Hs PROS-27 and Hs PROS-30 were mapped to chromosomes 14 (14q13) and 11 (11p15.1), respectively. The structure of the p27K protein shows multiple potential phosphorylation sites, an NTP-binding fold and an RNA-binding consensus sequence. The Hs PROS-27/beta-galactosidase fusion protein binds a single RNA of about 120 nucleotides from total HeLa cell RNA. Sequence comparisons show that the Hs PROS-27 and Hs PROS-30 genes belong to the gene family that encodes the prosome--MCP (multicatalytic proteinase)--proteasome proteins. Comparison with other members of the family from various species allowed us to show that the tripartite consensus sequence characteristic of the alpha-type sub-family is conserved from archeobacteria to man. The members of this gene family are characterised by very high evolutionary conservation of amino acid sequences of homologous genes and 20%-35% sequence similarity, between different family member within the same species and are clearly distinct from the beta-type family.

Disruption of prosomes by some bivalent metal ions results in the loss of their multicatalytic proteinase activity and cancels the nuclease resistance of prosomal RNA

Prosomes are ribonucleoprotein particles constituted by a variable set of about 20 proteins found associated with untranslated mRNA. In addition, they contain a small RNA, the presence of which has been an issue of controversy for a long time. The intact particles have a multicatalytic proteinase (MCP) activity and are very stable; we have never observed autodigestion of the particle by its intrinsic proteinase activity. Surprisingly it was found that Zn2+ and Cu2+ ions at concentrations of 0.1-1 mM disrupt the prosome particles isolated from HeLa cells and duck erythroblasts and abolish instantaneously its MCP activity, without altering the two-dimensional electrophoretic pattern of the constituent proteins. Fe2+, however, seems to induce autodegradation rather than dissociation of the prosome constituents. Most interestingly, protein or oligopeptide substrates protect the particle and its proteinase activity from disruption by Zn2+ or Cu2+. Nuclease-digestion assays reveal that the prosomal RNA, which is largely resistant in the intact particle, becomes digestible after dissociation of prosomes by Zn2+. These data give, for the first time, unambiguous proof of the presence of an RNA in the particle. Furthermore, they demonstrate a structure-function relationship between the complex and its enzyme activity, which seems to be based on the particle as an entity and not on the single constituent proteins.

The protein of M(r) 21,000 constituting the prosome-like particle of duck erythroblasts is homologous to apoferritin

In duck erythroblasts, two major populations of untranslated messenger (m) RNP can be separated by sucrose gradient centrifugation in low ionic strength. One of these contains globin mRNA associated to protein factors, among them the prosomes. The other, sedimenting in the 35S zone, contains non-globin mRNA. From this '35S' mRNP, a new RNP particle called the prosome-like particle was isolated and characterized [Akhayat, O., Infante, A. A., Infante, D., Martins de Sa, C., Grossi de Sa, M.-F. & Scherrer, K. (1987) Eur. J. Biochem. 170, 23-33]. The PLP is a multimer of a protein of M(r) 21,000, and contains small RNA species. The particle is tightly associated with repressed mRNA and inhibits in vitro protein synthesis. We show here that the protein of M(r) 21,000, constituting the prosome-like particle, is apoferritin. Different approaches confirm the RNP character of this particle and provide evidence that some of its RNA species are tRNA. The hypothesis is discussed as to whether (apo-)ferritin might serve other functions in addition to iron storage.

Prosomes and their multicatalytic proteinase activity

Prosomes were first described as being mRNA-associated RNP (ribonucleoprotein) particles and subcomponents of repressed mRNPs (messenger ribonucleoprotein). We show here that prosomes isolated from translationally inactive mRNP have a protease activity identical to that described by others for the multicatalytic proteinase complex (MCP, 'proteasome'). By RNase or non-ionic detergent treatment, the MCP activity associated with repressed non-globin mRNP from avian erythroblasts, sedimenting at 35 S, could be quantitatively shifted on sucrose gradients to the 19-S sedimentation zone characteristic of prosomes, which were identified by monoclonal antibodies. The presence of small RNA in the enzymatic complex was shown by immunoprecipitation of the protease activity out of dissociated mRNP using a mixture of anti-prosome monoclonal antibodies; a set of small RNAs 80-120 nucleotides long was isolated from the immunoprecipitate. Furthermore, on CsCl gradients, colocalisation of the MCP activity with prosomal proteins and prosomal RNA was found, and no difference in the prosomal RNA pattern was observed whether the particles were fixed or not prior to centrifugation. These data indicate that the MCP activity is a property of prosomes, shown to be in part RNP and subcomplexes of in vivo untranslated mRNP. A hypothesis for the role of the prosome-MCP particles in maintaining homeostasis of specific protein levels is proposed.

The major RNA in prosomes of HeLa cells and duck erythroblasts is tRNA(Lys,3)

Two-dimensional gel electrophoresis of HeLa cell prosomal RNAs, 3'-end labeled by RNA ligase, revealed one prominent spot. Determination of a partial sequence at the 3'-end indicated full homology to the 18 nucleotides at the 3'-end of tRNA(Lys,3) from rabbit, the bovine and the human species. An oligonucleotide complementary to the 3'-end of tRNA(Lys,3) hybridized on Northern blots with prosomal RNA from both HeLa cells and duck erythroblasts. In two-dimensional PAGE, the major pRNA of HeLa cells co-migrated with bovine tRNA(Lys,3). Reconstitution of the CCA 3'-end of RNA from both human and duck prosomes, by tRNA-nucleotidyl-transferase, confirmed the tRNA character of this type of RNA. Furthermore, it revealed at least one additional tRNA band about 85 nt long among the prosomal RNA from both species. Finally, confirming an original property of prosomal RNA, we show that in vitro synthesized tRNA(Lys,3) hybridizes stably to duck globin mRNA, and to poly(A)(+)- and poly(A)(-)-RNA from HeLa cells.

Structure and RNA content of the prosomes

Duck erythroblasts prosomes were analysed by small angle neutron scattering (SANS), dynamic light scattering and (cryo-)electron microscopy. A molecular weight of approximately 720,000 +/- 50,000, a radius of gyration of 64 +/- 2 A and a hydrodynamic radius of approximately 86 A were obtained. Electron micrographs show a hollow cylinder-like particle with a diameter of 120 A, a height of 170 A and a diameter of 40 A for the cavity, built of four discs, the two outer ones being more pronounced than those in the center. Results from SANS indicate less then 5% of RNA in the purified prosomes, but nuclease protection assays confirm its presence.