The Nuclear Cap-Binding Complex Mediates Meiotic Silencing by Unpaired DNA

In the filamentous fungus Neurospora crassa, cross walls between individual cells are normally incomplete, making the entire fungal network vulnerable to attack by viruses and selfish DNAs. Accordingly, several genome surveillance mechanisms are maintained to help the fungus combat these repetitive elements. One of these defense mechanisms is called meiotic silencing by unpaired DNA (MSUD), which identifies and silences unpaired genes during meiosis. Utilizing common RNA interference (RNAi) proteins, such as Dicer and Argonaute, MSUD targets mRNAs homologous to the unpaired sequence to achieve silencing. In this study, we have identified an additional silencing component, namely the cap-binding complex (CBC). Made up of cap-binding proteins CBP20 and CBP80, CBC associates with the 5′ cap of mRNA transcripts in eukaryotes. The loss of CBC leads to a deficiency in MSUD activity, suggesting its role in mediating silencing. As confirmed in this study, CBC is predominantly nuclear, although it is known to travel in and out of the nucleus to facilitate RNA transport. As seen in animals but not in plants, CBP20’s robust nuclear import depends on CBP80 in Neurospora. CBC interacts with a component (Argonaute) of the perinuclear meiotic silencing complex (MSC), directly linking the two cellular factors.

cap-binding proteins (CBPs) meiosis meiotic silencing by unpaired DNA (MSUD) Neurospora crassa RNA interference (RNAi) In eukaryotes, RNA polymerase II transcripts undergo multiple processing events to become mature mRNAs that can be translated into polypeptides. An important step is the addition of a 7-methylguanosine to the 59 end of nascent transcripts. This 59 cap structure plays a role in multiple aspects of mRNA metabolism, such as protection from exonucleolytic degradation, 39 end formation, pre-mRNA splicing, export from the nucleus, and translation (Topisirovic et al. 2011). These processes are often mediated by proteins that bind to the 59 cap, also known as cap-binding proteins (CBPs). Two of these, called CBP20 and CBP80, make up the heterodimeric cap-binding complex (CBC), which binds to the 59 cap and a spectrum of proteins necessary for various RNA processing events. CBC is known to be a key factor for transcription, mRNA stability, pre-mRNA 39 end processing, splicing, RNA export, the pioneer round of translation, and nonsensemediated mRNA decay (NMD) (Gonatopoulos-Pournatzis and Cowling 2014). Besides the aforementioned functions, CBC is also implicated in small RNA-mediated silencing (Gregory et al. 2008;Kim et al. 2008;Laubinger et al. 2008;Gruber et al. 2009;Sabin et al. 2009). In Neurospora crassa, an RNA interference (RNAi) system known as meiotic silencing by unpaired DNA (MSUD) exists to suppress expression from unpaired genes during sexual development (Shiu et al. 2001). A working model for MSUD begins with the detection of an unpaired DNA region during meiotic prophase I (with the help of suppressor of ascus dominance-6 or SAD-6, a presumptive homology search protein; Samarajeewa et al. 2014). A single-stranded aberrant RNA (aRNA) is transcribed from the unpaired DNA and subsequently exported to the perinuclear region. There, the aRNA encounters the meiotic silencing complex (MSC) (Decker et al. 2015), which contains several RNAi-related proteins. One of these is SAD-1, the RNA-directed RNA polymerase responsible for converting the aRNA into double strands (Shiu and Metzenberg 2002). Working alongside SAD-1, a helicase known as SAD-3 may assist in the RNA synthesis (Hammond et al. 2011a). The resulting double-stranded RNA (dsRNA) is then cut into small interfering RNAs (siRNAs) by the DCL-1 Dicer-like protein . The QIP (QDE-2-interacting protein) exonuclease converts the siRNA duplexes into single strands, which subsequently guide the SMS-2 (suppressor of meiotic silencing-2) Argonaute to target complementary mRNAs for silencing (Lee et al. 2003;Xiao et al. 2010). SAD-2 is thought to act as a scaffold protein and tether the aforementioned MSC components to the perinuclear region (Shiu et al. 2006;Decker et al. 2015). Two other proteins, SAD-4 and SAD-5, are essential for the production of siRNAs; however, their precise functions in MSUD remain unknown (Hammond et al. 2013a, b). In this study, we have shown that cap-binding proteins CBP20 and CBP80 also play a role in MSUD.

Fungal manipulation and genotypic information
Standard techniques from the Neurospora protocol guide were used throughout this work (http://www.fgsc.net/Neurospora/ NeurosporaProtocolGuide.htm). Strain names and genotypes are listed in Table 1. Knockouts and other markers were obtained from the Fungal Genetics Stock Center (FGSC) (McCluskey et al. 2010) and the Neurospora Functional Genomics Group (Colot et al. 2006). Fungal isolates were grown on Vogel's medium (Vogel 1956). Crosses were performed on synthetic crossing medium of Westergaard and Mitchell (1947).

Transformation and strain construction
Transformation by electroporation of conidia (asexual spores) was conducted using the technique of Margolin et al. (1997). All fluorescently tagged strains described in this study were constructed using the double-joint polymerase chain reaction (DJ-PCR) method (Hammond et al. 2011b;Samarajeewa et al. 2014).
Genotype screening and strain confirmation Genomic DNA was isolated from conidia (Henderson et al. 2005)  Bimolecular fluorescence complementation (BiFC) BiFC is an in vivo protein-protein interaction assay, and it relies on the reconstitution of the yellow fluorescent protein (YFP) when its nonfluorescing halves are brought together by two interacting proteins (Hu et al. 2002;Bardiya et al. 2008).
Generation of BiFC constructs was as described by Hammond et al. (2011b).
Photography and microscopy Z-stack pictures of protoperithecia (female structures) were taken using a Leica M205 FA stereomicroscope with a Leica DFC345 FX camera. Representative images of asci (spore sacs) and perithecia (fruiting bodies) were obtained using an Apple iPhone 5 with a Magnifi photoadapter (Arcturus Labs, Palo Alto, CA) on a Vanguard 1231CM microscope. An Olympus BX61 was used for fluorescent microscopy. Preparation and visualization of asci ($10/cross) were performed as described Xiao et al. 2010), with similar exposure times used across samples (50-150 ms for DAPI and 500-800 ms for fluorescent proteins).

Data availability
Strains are available upon request. The authors state that all data necessary for confirming the conclusions presented in the article are represented fully within the article.

Identification of two CBPs as MSUD players
To identify novel MSUD components, we have developed a highthroughput screen of the Neurospora knockout library for silencing mutants (Hammond et al. 2011a). Preliminary data from this screen identified two additional strains [FGSC 22440 (A) and 22441 (a)] that appear to be MSUD-deficient. These strains correspond to deletion mutants of ncu04187, also known as cbp80.
PCR amplification was used to verify the cbp80 deletion in these strains before its effect on silencing was evaluated in a quantitative MSUD assay. A Neurospora cross typically produces American footballshaped ascospores (sexual spores). However, if the Round spore gene n  were unpaired (and hence silenced), the cross (r + · r Δ ) would produce predominantly round ascospores (Shiu et al. 2001). This aberrant phenotype can be alleviated by the deletion of an MSUD gene. As seen in Figure 1, the removal of cbp80 increases the percentage of normal ascospores fivefold. This suggests that efficient silencing of unpaired genes requires CBP80. Because CBP80 is the large subunit of CBC, it is possible that its known binding partner, CBP20, also plays a role in silencing. Indeed, deletion of cbp20 (ncu00210) also affects MSUD, albeit to a lesser extent (Figure 1). When both CBPs are absent in a cross, the percentage of normal ascospores increases eightfold. These data suggest that CBC, a multifaceted eukaryotic factor, mediates MSUD.

Expression of CBC genes during both vegetative and sexual cycles
To determine the expression patterns of cbp20 and cbp80, we examined their transcripts via RNA-Seq datasets (see Materials and Methods). While the MSUD-exclusive genes examined here have relatively low vegetative expression (as compared to their sexual expression), cbp20 and cbp80 have robust expression levels in both lifecycles (Table 2). These data are consistent with the notion that CBC plays a role in both sexual and asexual stages. Indeed, it has been shown that CBC is important for NMD during vegetative growth in Neurospora (Zhang and Sachs 2015).

Loss of CBC impairs sexual development
None of the known components of MSUD are essential for cell viability (Samarajeewa et al. 2014), and CBP20 and CBP80 are no exception (Figure 2). On the other hand, the first few reported MSUD proteins are all necessary for the sexual cycle. For example, the absence of either qip or dcl-1 results in the production of perithecia that contain no asci, suggesting that they are important for early sexual development Xiao et al. 2010). A less severe phenotype can be seen in crosses lacking sad-1, sad-2, or sad-3, which are able to form asci that abort before ascospore production (Shiu et al. 2001(Shiu et al. , 2006Hammond et al. 2011a). Hence, for a long period of time, it was thought that MSUD is required for sexual development. However, the discovery of sad-4, sad-5, and sad-6 revealed this to be untrue, since crosses lacking any one of these genes are able to produce a decent number of mature ascospores (Hammond et al. 2013b;Samarajeewa et al. 2014).
To determine if CBC plays a role in sexual development, ascospore production was examined in a series of crosses. Although cbp20 and cbp80 are not absolutely required for sexual reproduction, their absence correlates with an appreciable decrease in the progeny count ( Figure 3A). In fact, homozygous crosses of double knockouts (cbp20 Δ /80 Δ · cbp20 Δ /80 Δ ) yield only one-sixth of the normal number of ascospores, while those of a single knockout are somewhat more fertile. This CBC-dependent sporulation defect is reminiscent of that observed in yeast (Qiu et al. 2012). CBC-deficient crosses appear to have fewer mature perithecia and asci ( Figure 3B).

Localization of CBC components
The first few described MSUD proteins localize in the perinuclear region, where they form a complex (MSC) to process RNAs exiting the nucleus (Shiu et al. 2006;Alexander et al. 2008;Xiao et al. 2010;Hammond et al. 2011a, b;Decker et al. 2015). The discovery of SAD-5 and SAD-6 demonstrated that some parts of the MSUD machinery are located inside the nucleus (Hammond et al. 2013b;Samarajeewa et al. 2014). Given CBC's documented role in various nuclear processes in other organisms, it is conceivable that Neurospora CBP20 and CBP80 are also found mainly in the nucleus. To test this hypothesis, CBP20 and CBP80 were tagged with mCherry and green fluorescent protein (GFP), respectively. As expected, microscopic examination revealed that they localize predominantly in the nucleus (Figure 4, A-H). This result mirrors observations seen in other organisms, such as Saccharomyces cerevisiae, Chironomus tentans, Homo sapiens, and Arabidopsis thaliana (Görlich et al. 1996;Visa et al. 1996;Izaurralde et al. 1994;Kierzkowski et al. 2009). Not surprisingly, CBP20 colocalizes with CBP80, confirming that the two proteins occupy the same subcellular compartment ( Figure 4D).
CBP20 and CBP80 interact with each other and not with SAD-5 Since most eukaryotic CBP20 and CBP80 proteins are known to form a heterodimer, we set out to investigate whether the two Neurospora homologs have physical association during sexual development. The loss of CBP20 and/or CBP80 is associated with a considerable reduction in ascospore production. (B) Although protoperithecial development appears proficient in a given single or double cbp mutant, there only seem to be roughly half the normal number of mature perithecia in the corresponding homozygous cross. Dissection of mutant perithecia showed that they contain frequent ascus abortions. Deletion of sad-5, in comparison, does not significantly affect ascospore production (Hammond et al. 2013b). 1, F2-01 · P3-08. 2, F8-02 · P20-47. 3, F6-19 · P20-39. 4, F8-04 · P20-68.
Expectedly, we have shown that CBP20 and CBP80 have close interaction using BiFC ( Figure 5A). SAD-5 is another MSUD protein known to localize predominantly in the nucleus, although its precise function has yet to be established (Hammond et al. 2013b). We asked if SAD-5 is directly linked to CBC in the nuclear part of the MSUD mechanism. As shown in Figure 5 (B and C), SAD-5 does not interact with either CBP20 or CBP80, suggesting that it may be spatially separated from the CBC-related part of the silencing pathway. CBP20 requires CBP80 for efficient nuclear localization Given the fact that CBP80 mediates the active nuclear import of CBP20 in animals (Izaurralde et al. 1995), it is possible for Neurospora to show a similar pattern. Our result indicates that CBP20 can be found in both nuclear and perinuclear regions when CBP80 is absent in a cross (Figure 4, I-L). This suggests that although CBP20 can independently enter the nucleus, its robust nuclear entry (and/or reentry) requires its heterodimeric partner.

CBC interacts with a component of MSC
Since CBC travels in and out of the nucleus (Gonatopoulos-Pournatzis and Cowling 2014), we asked if it interacts with the perinuclear MSC. In eukaryotes, Argonaute is known as the core effector of the RNA-induced silencing complex, which uses small RNAs to target complementary transcripts. In a BiFC assay, we have shown that both CBP20 and CBP80 interact with the SMS-2 Argonaute, a component of MSC (Figure 5,D and E). This suggests that CBC mediates silencing through its interaction with MSC.
As expected, CBP20 and CBP80 directly interact to form CBC in Neurospora. This heterodimeric complex is predominantly nuclear. In eukaryotes, CBC is known to travel in and out of the nucleus to facilitate RNA export. As seen in animals but not in plants (Kierzkowski et al. 2009), robust nuclear import of CBP20 depends on CBP80 in Neurospora.
Similar to other reported MSUD components, CBC is not essential for somatic growth. However, it is one of the few MSUD players (along with quelling factors DCL-1 and QIP) that have a clear role during the vegetative cycle (i.e., NMD). CBC is important for sexual development, as its absence is associated with a drastic reduction in ascospore production. Thus far, we have observed a range of sexual phenotypes in various silencing Figure 4 CBC localizes predominantly in the nucleus. (A-D) CBP20 and CBP80 colocalize in the nucleus, excluding the nucleolus (arrow). P20-44 · P20-45. (E-H) CBP80 does not appear to linger outside the nuclear envelope (labeled by nucleoporin NUP120). P20-34 · P20-35. (I-L) CBP20 accumulates outside the nuclear envelope in the absence of CBP80 (arrow), suggesting that its robust nuclear entry (and/or reentry) requires the latter. P23-28 · P23-29. The chromatin was stained with DAPI. Bar, 5 mm. mutants (from complete ascus abortion to normal), and the results here further expand the complexity of how MSUD proteins may affect sexual development.
Our study shows that CBC has physical interaction with a component of MSC, thereby connecting the two formerly unrelated RNA-binding complexes. CBC is the first well-characterized RNA hub to be linked to MSUD, and it will be interesting to see if others follow suit. Although CBC and SAD-5 (a poorly defined MSUD protein) are both predominantly nuclear, they do not have direct interaction, leaving the role of SAD-5 still unclear.
Our current model holds that MSUD begins in the nucleus, where aRNAs are made from any gene lacking a pairing partner. These aRNAs are then exported to the perinuclear region, where they are converted into siRNAs that are capable of guiding MSC to target complementary mRNAs. An mRNA is normally bound at the 59 cap by CBC, which aids its nuclear export. Since MSC resides just outside of the nuclear envelope and at least one of its components (SMS-2) interacts with CBC, one possibility is that MSC may efficiently detect exiting mRNAs by interacting with CBC. The absence of CBC may make it harder for MSC to recognize and capture target mRNAs, allowing some to reach the translational machinery and thus leading to a lower silencing efficiency. Alternatively, CBC may have a stimulatory effect on the silencing pathway directly or indirectly ). Future studies of CBC and related proteins should allow us to better define their connections to MSUD.

ACKNOWLEDGMENTS
We thank James Birchler and Patrice Albert for their equipment sharing and advice. We are indebted to the FGSC, the Neurospora Functional Genomics Group, University of Missouri (MU) Core Facilities, colleagues from our community, and members of the Shiu Laboratory for their materials and help. We are pleased to acknowledge use of materials generated by P01 GM068087 "Functional Analysis of a Model Filamentous Fungus." L.M.D. was supported by a GK-12 Fellowship from the National Science Foundation (DGE1045322). E.C.B. was supported by a Graduate Assistance in Areas of National Need Fellowship from the U.S. Department of Education. B.S.S. and S.L.K. were supported by the MU Life Sciences Undergraduate Research Opportunity Program. S.F.B. and J.B.H. were supported by the Federal Work-Study Program and the MU Exposure to Research for Science Students (EXPRESS) Program, respectively. This work was supported by the MU Research Board, the MU Research Council, and the National Science Foundation (MCB1157942). Figure 5 Interactions among CBP20, CBP80, and SMS-2. Similar to their eukaryotic counterparts, Neurospora CBP20 and CBP80 form a complex (A). The two cap-binding proteins interact with the SMS-2 Argonaute (D and E) and not SAD-5 (B and C). In a BiFC assay, each protein is tagged with either the N-or C-terminal fragment of YFP. A positive interaction (F) reconstitutes the yellow fluorophore, while a negative interaction (G) does not. Micrographs illustrate prophase asci expressing (A) yfpn-cbp20 and yfpc-cbp80 (P21-10 · P21-11), (B) yfpn-cbp20 and yfpc-sad-5 (P21-LITERATURE CITED