The quaking gene product necessary in embryogenesis and myelination combines features of RNA binding and signal transduction proteins.

The mouse quaking gene, essential for nervous system myelination and survival of the early embryo has been positionally cloned. Its sequence implies that the locus encodes a multifunctional gene used in a specific set of developing tissues to unite signal transduction with some aspect of RNA metabolism. The quaking(viable) (qkv) mutation has one class of messages truncated by a deletion. An independent ENU-induced mutation has a nonconservative amino acid change in one of two newly identified domains that are conserved from the C. elegans gld-1 tumour suppressor gene to the human Src-associated protein Sam68. The size and conservation of the quaking gene family implies that the pathway defined by this mutation may have broad relevance for rapid conveyance of extracellular information directly to primary gene transcripts.

Remarkable sequence conservation of transcripts encoding amphibian and mammalian homologues of quaking, a KH domain RNA-binding protein.

Mutations in the mouse quaking locus can result in two different types of developmental phenotypes: (1) a deficiency of myelin in the central nervous system that is accompanied by a characteristic tremor, or (2) embryonic lethality around day 9 of gestation. A quaking candidate gene (qkI) that encodes a KH motif protein has recently been identified. We have isolated and characterized cDNAs encoding the Xenopus quaking homologue (Xqua) and also assembled an almost complete human quaking sequence from expressed sequence tags. Sequence comparisons show that the amphibian and mammalian quaking transcripts exhibit striking conservation, both within the coding region and, unexpectedly, in the 3' UTR. Two Xqua transcripts 5 kb and 5.5 kb in length are differentially expressed in the Xenopus embryo, with the 5 kb transcript being detected as early as the gastrula stage of development. Using an in vitro assay, we have demonstrated RNA-binding activity for quaking protein encoded by the 5 kb transcript. Overall, the high sequence conservation of quaking sequences suggests an important conserved function in vertebrate development, probably in the regulation of RNA metabolism.

Stable gene expression from a mammalian artificial chromosome.

We have investigated the potential of PAC-based vectors as a route to the incorporation of a gene in a mammalian artificial chromosome (MAC). Previously we demonstrated that a PAC (PAC7c5) containing alpha-satellite DNA generated mitotically stable MACs in human cells. To determine whether a functional HPRT gene could be assembled in a MAC, PAC7c5 was co-transfected with a second PAC containing a 140 kb human HPRT gene into HPRT-deficient HT1080 cells. Lines were isolated containing a MAC hybridizing with both alpha-satellite and HPRT probes. The MACs segregated efficiently, associated with kinetochore proteins and stably expressed HPRT message after 60 days without selection. Complementation of the parental HPRT deficiency was confirmed phenotypically by growth on HAT selection. These results suggest that MACs could be further developed for delivering a range of genomic copies of genes into cells and that stable transgene expression can be achieved.

Mammalian artificial chromosome formation from circular alphoid input DNA does not require telomere repeats.

Mammalian artificial chromosomes (MACs) form in HT1080 cells after transfecting linear yeast artificial chromosome constructs minimally containing competent alphoid arrays, a selectable marker and terminal human telomere repeats. Restrictions on the structure of input DNA in MAC formation were investigated by transfecting circular or linear alphoid constructs with or without human telomere arrays and by varying the position and orientation of the telomere arrays on input linear constructs. Circular input DNA efficiently produced MACs. Absence of telomere arrays from circular input molecules did not significantly alter MAC formation rates. Linear constructs capped with telomere arrays generated MACs effectively, but a severe reduction in MAC formation was observed from linear constructs lacking telomere arrays. Human telomere arrays positioned 1-5 kb from linear construct ends and in either orientation were able to promote MAC formation with similar efficiencies. Both circular and linear input constructs generated artificial chromosomes that efficiently segregated in the absence of selection. Telomeres were not detected on the MACs, regardless of the inclusion of telomere arrays on input DNA, suggesting that circular chromosomes were formed. We found no evidence for acquisition of host cell DNA, which is consistent with de novo chromosome assembly.

Neural cell type-specific expression of QKI proteins is altered in quakingviable mutant mice.

qkI, a newly cloned gene lying immediately proximal to the deletion in the quakingviable mutation, is transcribed into three messages of 5, 6, and 7 kb. Antibodies raised to the unique carboxy peptides of the resulting QKI proteins reveal that, in the nervous system, all three QKI proteins are expressed strongly in myelin-forming cells and also in astrocytes. Interestingly, individual isoforms show distinct intracellular distributions: QKI-6 and QKI-7 are localized to perikaryal cytoplasm, whereas QKI-5 invariably is restricted to the nucleus, consistent with the predicted role of QKI as an RNA-binding protein. In quakingviable mutants, which display severe dysmyelination, QKI-6 and QKI-7 are absent exclusively from myelin-forming cells. By contrast, QKI-5 is absent only in oligodendrocytes of severely affected tracts. These observations implicate QKI proteins as regulators of myelination and reveal key insights into the mechanisms of dysmyelination in the quakingviable mutant.

Genomic organization and expression analysis of the mouse qkI locus.

qkI, encoding a KH domain-containing RNA binding protein, has been isolated as a candidate gene for the mouse neurological mutation quaking. Here, we describe detailed studies on its genomic structure and expression pattern. We isolated approximately 1 Mb of genomic region containing the quaking locus and determined its genomic organization. The qkI locus contains at least 9 exons spanning approximately 65 kb of DNA. It gives rise to six distinct transcripts encoding, theoretically, five different protein isoforms. Exons 1 through 4 are shared by all the transcripts, whereas coding exons and two distinct 3'-UTRs downstream to the exon 4 are differentially utilized. One isoform has a truncated KH domain and may act as an antagonist to the others. These findings and identification of a single transcription initiation site suggest that differential expression of each transcript is regulated by alternative splicing. Expression of each alternative transcript and protein product was also examined. Two types of transcripts, 5 kb-A and B, are most abundant in the brain of newborn mice and are gradually downregulated thereafter. In contrast, the other three messages, 6 kb, 7 kb-A and B, increase as myelination proceeds and peak at 2 weeks of age, corresponding to the most active stage of myelination. Although the qkI messages and their products are abundant in brain and heart, a lower level of expression was found in various other tissues tested. Alternative transcripts that share the same 3'-UTR showed very similar expression patterns, suggesting a regulatory role of the 3'-UTRs in qkI gene expression.