Posttranscriptional Gene Silencing in Neurospora by a RecQ DNA Helicase

See allHide authors and affiliations

Science  17 Dec 1999:
Vol. 286, Issue 5448, pp. 2342-2344
DOI: 10.1126/science.286.5448.2342


The phenomenon of posttranscriptional gene silencing (PTGS), which occurs when a transgene is introduced into a cell, is poorly understood. Here, the qde-3 gene, which is required for the activation and maintenance of gene silencing in the fungusNeurospora crassa, was isolated. Sequence analysis revealed that the qde-3 gene belongs to the RecQ DNA helicase family. The QDE3 protein may function in the DNA-DNA interaction between introduced transgenes or with an endogenous gene required for gene-silencing activation. In animals, genes that are homologous to RecQ protein, such as the human genes for Bloom's syndrome and Werner's syndrome, may also function in PTGS.

Posttranscriptional gene silencing as a consequence of transgene introduction is a broadly diffused phenomenon in plants and fungi (1, 2). Introduction of double-stranded RNA (dsRNA) induces a similar phenomenon in animals (3). The wide occurrence of gene silencing among different organisms indicates that these phenomena may have evolved from an ancestral mechanism involved in genome protection from invading DNA (4) and viruses (5). Several models have been proposed to explain PTGS on the basis of the notion that the introduced transgenes result in the production of aberrant RNAs (aRNAs) (2) that are recognized as a template by host RNA-dependent RNA polymerase (RdRP). The RdRP enzyme may synthesize antisense RNA that can bind to mRNA and form dsRNAs that are targets for sequence-specific RNA degradation (6). These models have received experimental support. The qde-1 gene, which encodes a cellular component of PTGS in the fungus Neurospora crassa, is homologous to RdRP (7). Moreover, the accumulation of small antisense RNA molecules correlates with the occurrence of gene silencing in plants (8). It is unclear why PTGS is activated in some transgenic lines, whereas it is not activated in other lines. Gene silencing could be triggered by DNA pairing between homologous transgenes or with homologous resident genes (9). Such pairing, which could interfere with normal transcription, producing aRNA molecules, may occur only in some transgenic lines.

In gene silencing, also called “quelling” in N. crassa (10), three classes of quelling-defective mutants (qde-1, qde-2, and qde-3) have been isolated (11). To clone the qde-3 gene, we used random insertional mutagenesis of an al-1(albino-1) transgenic strain showing an albino (white) phenotype as a consequence of posttranscriptional silencing of the endogenous al-1 gene, which is involved in the biosynthesis of carotenoids (12). Mutation of qde genes releases al-1 gene silencing, resulting in the recovery of a wild-type (orange) phenotype that can be easily selected by visual inspection. A strain (627) showing the recovery of al-1 gene expression was isolated. By using a heterokaryon complementation analysis, we found that strain 627 belongs to one of the three previously identified qde complementation groups,qde-3. To isolate the qde-3 gene, we obtained (by plasmid rescue) genomic DNA from strain 627 flanking the insertion site (13). Two genomic cosmids were isolated by using the flanking sequences as a probe and were found to complement theqde-3 mutants, resulting in restoration of al-1gene silencing that was visible as the appearance of a white phenotype. Furthermore, a 9-kb Sph I fragment derived from the cosmids complemented qde-3 mutants. This DNA fragment was sequenced, revealing a long open reading frame of ∼6 kb that contains two putative introns identified by splicing consensus sequences and mapped by reverse transcriptase–polymerase chain reaction (RT-PCR) (14). To demonstrate that the putative 6-kb open reading frame is coincident with the qde-3 gene, we mapped the insertion site of the tagging plasmid in the qde-3mutant strain 627. The tagging plasmid was inserted immediately downstream from the second intron of the qde-3 gene, within the 3′-terminal acceptor site.

The putative QDE3 protein deduced from the qde-3nucleotide sequence contains 1955 amino acids. The encoded QDE3 polypeptide has a calculated molecular weight of 216,612 daltons. Using the predicted QDE3 peptide in a BLASTP search of amino acid sequence databases (15), we identified homologies with several peptides belonging to the family of RecQ DNA helicases. Homology is restricted to a 350–amino acid domain located in the center region of the polypeptide (residues 875 through 1228). This domain is coincident with the seven helicase domains that are strongly conserved among the RecQ helicases in organisms ranging from Escherichia coli to humans (Fig. 1). The qde-3 helicase domain shows the highest similarity with the Sgs1 protein ofSaccharomyces cerevisiae (54% identity) and theRqh1 protein of Schizosaccharomyces pombe (55% identity). Among RecQ proteins, however, QDE3 appears to belong to a subfamily of proteins that are considerably larger than the E. coli prototype (Fig. 1A). Related proteins belonging to this subfamily include three human genes [BML (Bloom's syndrome gene) (16), WRN (Werner's syndrome gene) (17), and RecQ4 (18)] and yeast genes Sgs1 (19) from S. cerevisiae and Rqh1 (20) from Sch. pombe. Other regions of the QDE3 protein, like the proteins of the human and yeast subfamilies, are rich in charged and polar amino acids, and the NH2-terminal region contains acidic domains (Fig. 1A).

Figure 1

qde-3 belongs to the RecQ DNA helicase family. (A) Schematic representation of the members of the RecQ DNA helicase family. The names of the gene products and the organisms are shown. The solid areas indicate the conserved helicase domains. Acidic domains are shown as shaded boxes. (B) Amino acid sequence alignment of the helicase domains of the members of the RecQ DNA helicase family (27). The gene product names and the positions of amino acid residues are shown. Identical residues are shown in black; dashes indicate spaces introduced to maintain sequence alignment. Boxes above the sequences indicate the positions of seven helicase domains (I, Ia, II, III, IV, V, and VI).

The yeast Sgs1p and the murine WRN interact with DNA topoisomerases (19, 21). To test for an interaction between QDE3 and topoisomerases in Neurospora, we assayed the sensitivity of several qde-3 mutants to the type II topoisomerase inhibitor, etoposide, and to the type I topoisomerase inhibitor, camptothecin (22). Etoposide, used at high concentration (10-fold higher than that generally used), did not show an inhibitory effect on either mutant or wild-type strains, indicating that Neurospora cells have low sensitivity to this drug. In camptothecin sensitivity assays, three qde-3mutant strains [627 and two ultraviolet (UV) irradiation–induced mutants, M17 and M18] showed a dramatic increase of sensitivity to the inhibitor (Fig. 2). By contrast, the control strain (6XW), which has the same genetic background as the qde-3 mutants, and the qde-1 (M20) and qde-2 (M10) mutant strains did not show increased sensitivity to camptothecin. Thus, the reason for increased sensitivity of qde-3 strains to the type I topoisomerase inhibitor camptothecin is probably a consequence of mutations within the qde-3 gene.

Figure 2

Sensitivity of qde-3mutants to type I topoisomerase inhibitor, camptothecin. Threeqde-3 mutants (strain 627, M17, and M18), a wild-type strain (WT), an al-1 silenced strain (6XW), a qde-1 mutant strain (M20), and a qde-2 mutant strain (M10) were assayed for sensitivity to camptothecin. Strains were grown in liquid cultures in the presence of different concentrations of camptothecin as indicated. For each mutant strain tested, the mass (in grams) of dried mycelia after 48 hours of growth is shown. Error bars indicate SD.

The fact that the qde-3 gene encodes a putative DNA helicase suggests a role for this gene in the activation step of gene silencing. A model for qde-3 function shows that the QDE3 DNA helicase could unwind double-stranded DNA, which may be required for DNA-DNA interactions between transgenic repeats. In addition, the DNA-pairing model proposes that DNA interaction between transgenes may induce changes in methylation or chromatin structure (or both), producing an “altered state” that could result in aRNA production (9). It has been proposed that DNA helicase or topoisomerase complexes may be involved in chromatin remodeling (23). The fact that QDE3 probably interacts with topoisomerases in vivo may suggest that QDE3 may have also a role in chromatin changes required for aberrant transcription. Alternatively, it has been proposed (3) that aRNAs could be dsRNAs produced from transgenic inverted repeats (IRs). The ability of RecQ helicases to process cruciform DNA structures (24) may indicate that QDE3 could be involved in resolving transgenic IR cruciforms to allow transcription of dsRNAs.

Eukaryotic RecQ DNA helicases have been generally implicated in DNA repair and in regulating recombination (16,17, 20). Our findings suggest that a specific RecQ helicase could be involved in a function other than DNA recombination and repair. In fact, in Neurospora, QDE3 seems to be specialized in gene silencing, because we found that the mutation in the qde-3 gene is sufficient to impair quelling, although at least another recQ homologous gene is present inNeurospora (25). Moreover, qde-3mutant strains and a wild-type strain showed the same ability to repair DNA damage induced by several mutagens (26).

The fact that RecQ-like protein is involved in gene silencing inNeurospora has begun to help us to understand the PTGS phenomenon. It also presents the obvious opportunity to test whether homologous recQ genes may be implicated in gene silencing in other organisms, especially in plants. This new function of a RecQ protein may also contribute to a deeper understanding of the biology of the recQ gene family and its function in higher eukaryotes, including humans.

  • * To whom correspondence should be addressed. E-mail: carlo{at}


View Abstract

Navigate This Article