Carlo Morandi

Organization of the human gene encoding heterogeneous nuclear ribonucleoprotein type I (hnRNP I) and characterization of hnRNP I related pseudogene

The human gene hnRNPI encoding the heterogeneous nuclear ribonucleoprotein type I, an alternative splicing modulator of tissue-specific transcripts, also known as PTB (polypyrimidine tract-binding protein), was recently mapped on chromosome 14, as well as on chromosome 19, suggesting that two closely related copies of the same gene might exist in the human genome. We report here that the gene localized on chromosome 14 corresponds to a highly homologous processed pseudogene related to hnRNPI gene (psihnRNPI). Analysis by RT-PCR and by EST database comparison indicates that psihnRNPI is not expressed. In this report we have also analyzed the organization of the actual hnRNPI gene localized on chromosome 19. The DNA sequence at the intron-exon boundaries unveiled the possible mechanism by which three isoforms of the protein (namely hnRNPI, PTB2 and PTB3) are generated by means of alternative splicing of the same hnRNPI gene transcript.

Secondary structure prediction for RNA binding domain in RNP proteins identifies βαβ as the main structural motif

In eukaryotic cells transcript processing is strictly dependent upon binding of specific proteins. Nuclear RNA binding proteins share a common domain, which is involved in RNA binding. In order to characterize RNP‐RNA interactions we have performed a secondary structure prediction based both on statistical algorithms and comparative analysis of different proteins. A high conservation for secondary structure propensity between different RNPs was observed.

Modeling by homology of RNA binding domain in A1 hnRNP protein

Eukaryotic nuclear RNA binding proteins share a common sequence motif thought to be implicated in RNA binding. One of the two domains present in A1 hnRNP protein, has been modelled by homology in order to make a prediction of the main features of the RNA binding site. Acyl‐phosphatase (EC 3.6.1.7) was selected as template for the modeling experiment. The predicted RNA binding site is a β‐sheet containing the two RNP consensus sequences as well as lysines and arginines conserved among the family.

Transcriptional regulation of the human Raver2 ribonucleoprotein gene.

Raver2 is a putative modulator of the activity of the polypyrimidine-tract binding protein (PTB), one of the most intensively studied splicing repressors. Little is known about Raver2 expression, and all current data is from mice where it shows tissue specificity. In the present study, by comparing Raver2 transcript expression in human and mouse tissues, we found that human Raver2 is ubiquitously expressed in adult tissues. In order to investigate human Raver2 transcription regulation, we identified and characterized a putative promoter region in a 1000bp region upstream of the transcription starting site of the gene. Dual luciferase reporter assays demonstrated that this region had promoter activity conferred by the first 160bp. By mutagenic analyses of putative cis-acting regulatory sequences, we identified an individual site that decreased the promoter activity by up to 40% when mutated. Together, our results suggest that regulation of human Raver2 expression involves TATA-less transcriptional activity.

Analysis of the yeast NSR1 gene and protein domain comparison between Nsr1 and human hnRNP type A1

The yeast nucleolar protein-encoding gene NSR1 was isolated by low-stringency screening of a yeast genomic library with the human heterogeneous nuclear ribonucleoprotein type A1 (hnRNP A1) cDNA probe, and was mapped to chromosome VII. RNA abundance was determined and the transcription start point and polyadenylation site were mapped. A comparison between the Nsr1 and hnRNP A1 proteins, based on homopolymer RNA binding to their structural domains in vitro, revealed a striking biochemical similarity. When the N-terminal, lysine- and arginine-rich domain of Nsr1 was removed, the truncated protein behaved similarly to hnRNP A1; furthermore, the two RRM (RNA recognition motif) domains of Nsr1 behaved in the same manner as the two RRM domains of hnRNP A1. The biochemical data, therefore, would support the hypothesis that the two RRM domains in hnRNP A1 and Nsr1 interact with RNA in a similar manner in both mammalian and yeast cells, respectively.

Aging: a portrait from gene expression profile in blood cells

The availability of reliable biomarkers of aging is important not only to monitor the effect of interventions and predict the timing of pathologies associated with aging but also to understand the mechanisms and devise appropriate countermeasures. Blood cells provide an easily available tissue and gene expression profiles from whole blood samples appear to mirror disease states and some aspects of the aging process itself. We report here a microarray analysis of whole blood samples from two cohorts of healthy adult and elderly subjects, aged 43±3 and 68±4 years, respectively, to monitor gene expression changes in the initial phase of the senescence process. A number of significant changes were found in the elderly compared to the adult group, including decreased levels of transcripts coding for components of the mitochondrial respiratory chain, which correlate with a parallel decline in the maximum rate of oxygen consumption (VO2max), as monitored in the same subjects. In addition, blood cells show age-related changes in the expression of several markers of immunosenescence, inflammation and oxidative stress. These findings support the notion that the immune system has a major role in tissue homeostasis and repair, which appears to be impaired since early stages of the aging process.

Modelling by homology of RNA binding domain

In the last few years several evidences have been obtained about the specificity of RNP binding to RNA [for review see Bandziulis et al., 1989]. Since the three dimensional structure of RNA binding domain in RNP has not been resolved yet we know very little about the pattern of interaction between this class of proteins and RNA. B. Merril and collegues have shown that the highly conserved aromatics of the consensus sequences RNP1 and RNP2 are components of the binding site [Merril et al., 1988], but no data are available at the moment about the topology of the binding site or concerning residues eventually involved in RNA sequence recognition. For all these reasons, we have used a molecular modelling approach in order to obtain informations about protein folding of the 92 residues RNA binding domain. It is known that protein tertiary structure is often more highly conserved than primary structure, and starting from this observation several proteins have been modelled by homology [ for review, see Blundell et al., 1987 ]. We have analized a protein sequence data bank searching for proteins which three dimensional structure were already resolved, having sequence similarity with 22 different RNA binding domains. Horse muscle acylphosphatase (HMA)(EC 3.6.1.7) resulted to have similarity with RNPs, on average 25%, and its structure has been resolved by NMR techniques [Saudek et al., 1989]. Furthermore it shows a secondary structure pattern comparable to that recently hypothesized for RNA binding domain [Ghetti et al., 1989]. The RNA binding domain we have modelled is the second from the N terminal contained in human A1 hnRNP ( from aa. 92-184)(fig.1).

Identification of A 55 Kdal nuclear protein sharing homology with hnRNP type L

In our laboratory several clones were isolated by immunological screening of cDNA expression libraries from human liver mRNA, by means of rabbit antiserum against a subset of mammalian hnRNP (i). Among them, the cDNA of the hnRNP type A1 protein previously described (i) and a second cDNA clone hybridizing to a + poly A RNA species of 3300 nucleotides, the sequence determination of which, allowed us to derive the actual primary structure of the encoded protein.