RNA-Dependent Epigenetic Silencing Directs Transcriptional Downregulation Caused by Intronic Repeat Expansions.

Transcriptional downregulation caused by intronic triplet repeat expansions underlies diseases such as Friedreich's ataxia. This downregulation of gene expression is coupled with epigenetic changes, but the underlying mechanisms are unknown. Here, we show that an intronic GAA/TTC triplet expansion within the IIL1 gene of Arabidopsis thaliana results in accumulation of 24-nt short interfering RNAs (siRNAs) and repressive histone marks at the IIL1 locus, which in turn causes its transcriptional downregulation and an associated phenotype. Knocking down DICER LIKE-3 (DCL3), which produces 24-nt siRNAs, suppressed transcriptional downregulation of IIL1 and the triplet expansion-associated phenotype. Furthermore, knocking down additional components of the RNA-dependent DNA methylation (RdDM) pathway also suppressed both transcriptional downregulation of IIL1 and the repeat expansion-associated phenotype. Thus, our results show that triplet repeat expansions can lead to local siRNA biogenesis, which in turn downregulates transcription through an RdDM-dependent epigenetic modification.

XRCC1 prevents toxic PARP1 trapping during DNA base excision repair.

Mammalian DNA base excision repair (BER) is accelerated by poly(ADP-ribose) polymerases (PARPs) and the scaffold protein XRCC1. PARPs are sensors that detect single-strand break intermediates, but the critical role of XRCC1 during BER is unknown. Here, we show that protein complexes containing DNA polymerase ß and DNA ligase III that are assembled by XRCC1 prevent excessive engagement and activity of PARP1 during BER. As a result, PARP1 becomes "trapped" on BER intermediates in XRCC1-deficient cells in a manner similar to that induced by PARP inhibitors, including in patient fibroblasts from XRCC1-mutated disease. This excessive PARP1 engagement and trapping renders BER intermediates inaccessible to enzymes such as DNA polymerase ß and impedes their repair. Consequently, PARP1 deletion rescues BER and resistance to base damage in XRCC1-/- cells. These data reveal excessive PARP1 engagement during BER as a threat to genome integrity and identify XRCC1 as an "anti-trapper" that prevents toxic PARP1 activity.

Observing a DNA polymerase choose right from wrong.

DNA polymerase (pol) ß is a model polymerase involved in gap-filling DNA synthesis utilizing two metals to facilitate nucleotidyl transfer. Previous structural studies have trapped catalytic intermediates by utilizing substrate analogs (dideoxy-terminated primer or nonhydrolysable incoming nucleotide). To identify additional intermediates during catalysis, we now employ natural substrates (correct and incorrect nucleotides) and follow product formation in real time with 15 different crystal structures. We are able to observe molecular adjustments at the active site that hasten correct nucleotide insertion and deter incorrect insertion not appreciated previously. A third metal binding site is transiently formed during correct, but not incorrect, nucleotide insertion. Additionally, long incubations indicate that pyrophosphate more easily dissociates after incorrect, compared to correct, nucleotide insertion. This appears to be coupled to subdomain repositioning that is required for catalytic activation/deactivation. The structures provide insights into a fundamental chemical reaction that impacts polymerase fidelity and genome stability.

Biochemical and structural characterization of Fapy•dG replication by Human DNA polymerase ß.

N6-(2-deoxy-α,ß-d-erythro-pentofuranosyl)-2,6-diamino-4-hydroxy-5-formamido-pyrimidine (Fapy•dG) is formed from a common intermediate and in comparable amounts to the well-studied mutagenic DNA lesion 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-OxodGuo). Fapy•dG preferentially gives rise to G → T transversions and G → A transitions. However, the molecular basis by which Fapy•dG is processed by DNA polymerases during this mutagenic process remains poorly understood. To address this we investigated how DNA polymerase ß (Pol ß), a model mammalian polymerase, bypasses a templating Fapy•dG, inserts Fapy•dGTP, and extends from Fapy•dG at the primer terminus. When Fapy•dG is present in the template, Pol ß incorporates TMP less efficiently than either dCMP or dAMP. Kinetic analysis revealed that Fapy•dGTP is a poor substrate but is incorporated ∼3-times more efficiently opposite dA than dC. Extension from Fapy•dG at the 3'-terminus of a nascent primer is inefficient due to the primer terminus being poorly positioned for catalysis. Together these data indicate that mutagenic bypass of Fapy•dG is likely to be the source of the mutagenic effects of the lesion and not Fapy•dGTP. These experiments increase our understanding of the promutagenic effects of Fapy•dG.

Unfilled gaps by polß lead to aberrant ligation by LIG1 at the downstream steps of base excision repair pathway.

Base excision repair (BER) involves the tightly coordinated function of DNA polymerase ß (polß) and DNA ligase I (LIG1) at the downstream steps. Our previous studies emphasize that defective substrate-product channeling, from gap filling by polß to nick sealing by LIG1, can lead to interruptions in repair pathway coordination. Yet, the molecular determinants that dictate accurate BER remains largely unknown. Here, we demonstrate that a lack of gap filling by polß leads to faulty repair events and the formation of deleterious DNA intermediates. We dissect how ribonucleotide challenge and cancer-associated mutations could adversely impact the ability of polß to efficiently fill the one nucleotide gap repair intermediate which subsequently results in gap ligation by LIG1, leading to the formation of single-nucleotide deletion products. Moreover, we demonstrate that LIG1 is not capable of discriminating against nick DNA containing a 3'-ribonucleotide, regardless of base-pairing potential or damage. Finally, AP-Endonuclease 1 (APE1) shows distinct substrate specificity for the exonuclease removal of 3'-mismatched bases and ribonucleotides from nick repair intermediate. Overall, our results reveal that unfilled gaps result in impaired coordination between polß and LIG1, defining a possible type of mutagenic event at the downstream steps where APE1 could provide a proofreading role to maintain BER efficiency.

Impact of DNA ligase 1 and IIIα interactions with APE1 and polß on the efficiency of base excision repair pathway at the downstream steps.

Base excision repair (BER) requires a tight coordination between the repair enzymes through protein-protein interactions and involves gap filling by DNA polymerase (pol) ß and subsequent nick sealing by DNA ligase (LIG) 1 or LIGIIIα at the downstream steps. Apurinic/apyrimidinic-endonuclease 1 (APE1), by its exonuclease activity, proofreads 3' mismatches incorporated by polß during BER. We previously reported that the interruptions in the functional interplay between polß and the BER ligases result in faulty repair events. Yet, how the protein interactions of LIG1 and LIGIIIα could affect the repair pathway coordination during nick sealing at the final steps remains unknown. Here, we demonstrate that LIGIIIα interacts more tightly with polß and APE1 than LIG1, and the N-terminal noncatalytic region of LIG1 as well as the catalytic core and BRCT domain of LIGIIIα mediate interactions with both proteins. Our results demonstrated less efficient nick sealing of polß nucleotide insertion products in the absence of LIGIIIα zinc-finger domain and LIG1 N-terminal region. Furthermore, we showed a coordination between APE1 and LIG1/LIGIIIα during the removal of 3' mismatches from the nick repair intermediate on which both BER ligases can seal noncanonical ends or gap repair intermediate leading to products of single deletion mutagenesis. Overall results demonstrate the importance of functional coordination from gap filling by polß coupled to nick sealing by LIG1/LIGIIIα in the presence of proofreading by APE1, which is mainly governed by protein-protein interactions and protein-DNA intermediate communications, to maintain repair efficiency at the downstream steps of the BER pathway.

Thumb-domain dynamics modulate the functional repertoire of DNA-Polymerase IV (DinB).

In order to cope with the risk of stress-induced mutagenesis, cells in all kingdoms of life employ Y-family DNA polymerases to resolve resulting DNA lesions and thus maintaining the integrity of the genome. In Escherichia coli, the DNA polymerase IV, or DinB, plays this crucial role in coping with these type of mutations via the so-called translesion DNA synthesis. Despite the availability of several high-resolution crystal structures, important aspects of the functional repertoire of DinB remain elusive. In this study, we use advanced solution NMR spectroscopy methods in combination with biophysical characterization to elucidate the crucial role of the Thumb domain within DinB's functional cycle. We find that the inherent dynamics of this domain guide the recognition of double-stranded (ds) DNA buried within the interior of the DinB domain arrangement and trigger allosteric signals through the DinB protein. Subsequently, we characterized the RNA polymerase interaction with DinB, revealing an extended outside surface of DinB and thus not mutually excluding the DNA interaction. Altogether the obtained results lead to a refined model of the functional repertoire of DinB within the translesion DNA synthesis pathway.

Saccharomyces cerevisiae DNA polymerase IV overcomes Rad51 inhibition of DNA polymerase δ in Rad52-mediated direct-repeat recombination.

Saccharomyces cerevisiae DNA polymerase IV (Pol4) like its homolog, human DNA polymerase lambda (Polλ), is involved in Non-Homologous End-Joining and Microhomology-Mediated Repair. Using genetic analysis, we identified an additional role of Pol4 also in homology-directed DNA repair, specifically in Rad52-dependent/Rad51-independent direct-repeat recombination. Our results reveal that the requirement for Pol4 in repeat recombination was suppressed by the absence of Rad51, suggesting that Pol4 counteracts the Rad51 inhibition of Rad52-mediated repeat recombination events. Using purified proteins and model substrates, we reconstituted in vitro reactions emulating DNA synthesis during direct-repeat recombination and show that Rad51 directly inhibits Polδ DNA synthesis. Interestingly, although Pol4 was not capable of performing extensive DNA synthesis by itself, it aided Polδ in overcoming the DNA synthesis inhibition by Rad51. In addition, Pol4 dependency and stimulation of Polδ DNA synthesis in the presence of Rad51 occurred in reactions containing Rad52 and RPA where DNA strand-annealing was necessary. Mechanistically, yeast Pol4 displaces Rad51 from ssDNA independent of DNA synthesis. Together our in vitro and in vivo data suggest that Rad51 suppresses Rad52-dependent/Rad51-independent direct-repeat recombination by binding to the primer-template and that Rad51 removal by Pol4 is critical for strand-annealing dependent DNA synthesis.

Abasic site-peptide cross-links are blocking lesions repaired by AP endonucleases.

Apurinic/apyrimidinic (AP) sites are abundant DNA lesions arising from spontaneous hydrolysis of the N-glycosidic bond and as base excision repair (BER) intermediates. AP sites and their derivatives readily trap DNA-bound proteins, resulting in DNA-protein cross-links. Those are subject to proteolysis but the fate of the resulting AP-peptide cross-links (APPXLs) is unclear. Here, we report two in vitro models of APPXLs synthesized by cross-linking of DNA glycosylases Fpg and OGG1 to DNA followed by trypsinolysis. The reaction with Fpg produces a 10-mer peptide cross-linked through its N-terminus, while OGG1 yields a 23-mer peptide attached through an internal lysine. Both adducts strongly blocked Klenow fragment, phage RB69 polymerase, Saccharolobus solfataricus Dpo4, and African swine fever virus PolX. In the residual lesion bypass, mostly dAMP and dGMP were incorporated by Klenow and RB69 polymerases, while Dpo4 and PolX used primer/template misalignment. Of AP endonucleases involved in BER, Escherichia coli endonuclease IV and its yeast homolog Apn1p efficiently hydrolyzed both adducts. In contrast, E. coli exonuclease III and human APE1 showed little activity on APPXL substrates. Our data suggest that APPXLs produced by proteolysis of AP site-trapped proteins may be removed by the BER pathway, at least in bacterial and yeast cells.

Interlocking activities of DNA polymerase ß in the base excision repair pathway.

SignificanceBase excision repair (BER) is one of the major DNA repair pathways used to fix a myriad of cellular DNA lesions. The enzymes involved in BER, including DNA polymerase ß (Polß), have been identified and characterized, but how they act together to efficiently perform BER has not been fully understood. Through gel electrophoresis, mass spectrometry, and kinetic analysis, we discovered that the two enzymatic activities of Polß can be interlocked, rather than functioning independently from each other, when processing DNA intermediates formed in BER. The finding prompted us to hypothesize a modified BER pathway. Through conventional and time-resolved X-ray crystallography, we solved 11 high-resolution crystal structures of cross-linked Polß complexes and proposed a detailed chemical mechanism for Polß's 5'-deoxyribose-5-phosphate lyase activity.

Direct visualization of translesion DNA synthesis polymerase IV at the replisome.

In bacterial cells, DNA damage tolerance is manifested by the action of translesion DNA polymerases that can synthesize DNA across template lesions that typically block the replicative DNA polymerase III. It has been suggested that one of these translesion DNA synthesis DNA polymerases, DNA polymerase IV, can either act in concert with the replisome, switching places on the ß sliding clamp with DNA polymerase III to bypass the template damage, or act subsequent to the replisome skipping over the template lesion in the gap in nascent DNA left behind as the replisome continues downstream. Evidence exists in support of both mechanisms. Using single-molecule analyses, we show that DNA polymerase IV associates with the replisome in a concentration-dependent manner and remains associated over long stretches of replication fork progression under unstressed conditions. This association slows the replisome, requires DNA polymerase IV binding to the ß clamp but not its catalytic activity, and is reinforced by the presence of the γ subunit of the ß clamp-loading DnaX complex in the DNA polymerase III holoenzyme. Thus, DNA damage is not required for association of DNA polymerase IV with the replisome. We suggest that under stress conditions such as induction of the SOS response, the association of DNA polymerase IV with the replisome provides both a surveillance/bypass mechanism and a means to slow replication fork progression, thereby reducing the frequency of collisions with template damage and the overall mutagenic potential.

Modifying the Basicity of the dNTP Leaving Group Modulates Precatalytic Conformational Changes of DNA Polymerase ß.

The catalytic function of DNA polymerase ß (pol ß) fulfills the gap-filling requirement of the base excision DNA repair pathway by incorporating a single nucleotide into a gapped DNA substrate resulting from the removal of damaged DNA bases. Most importantly, pol ß can select the correct nucleotide from a pool of similarly structured nucleotides to incorporate into DNA in order to prevent the accumulation of mutations in the genome. Pol ß is likely to employ various mechanisms for substrate selection. Here, we use dCTP analogues that have been modified at the ß,γ-bridging group of the triphosphate moiety to monitor the effect of leaving group basicity of the incoming nucleotide on precatalytic conformational changes, which are important for catalysis and selectivity. It has been previously shown that there is a linear free energy relationship between leaving group pKa and the chemical transition state. Our results indicate that there is a similar relationship with the rate of a precatalytic conformational change, specifically, the closing of the fingers subdomain of pol ß. In addition, by utilizing analogue ß,γ-CHX stereoisomers, we identified that the orientation of the ß,γ-bridging group relative to R183 is important for the rate of fingers closing, which directly influences chemistry.

Visualizing the coordination of apurinic/apyrimidinic endonuclease (APE1) and DNA polymerase ß during base excision repair.

Base excision repair (BER) is carried out by a series of proteins that function in a step-by-step process to identify, remove, and replace DNA damage. During BER, the DNA transitions through various intermediate states as it is processed by each DNA repair enzyme. Left unrepaired, these BER intermediates can transition into double-stranded DNA breaks and promote genome instability. Previous studies have proposed a short-lived complex consisting of the BER intermediate, the incoming enzyme, and the outgoing enzyme at each step of the BER pathway to protect the BER intermediate. The transfer of BER intermediates between enzymes, known as BER coordination or substrate channeling, remains poorly understood. Here, we utilize single-molecule total internal reflection fluorescence microscopy to investigate the mechanism of BER coordination between apurinic/apyrimidinic endonuclease 1 (APE1) and DNA polymerase ß (Pol ß). When preformed complexes of APE1 and the incised abasic site product (APE1 product and Pol ß substrate) were subsequently bound by Pol ß, the Pol ß enzyme dissociated shortly after binding in most of the observations. In the events where Pol ß binding was followed by APE1 dissociation during substrate channeling, Pol ß remained bound for a longer period of time to allow disassociation of APE1. Our results indicate that transfer of the BER intermediate from APE1 to Pol ß during BER is dependent on the dissociation kinetics of APE1 and the duration of the ternary complex on the incised abasic site.

The enzymatic properties of Arabidopsis thaliana DNA polymerase λ suggest a role in base excision repair.

Base excision repair (BER) generates gapped DNA intermediates containing a 5'-terminal 2-deoxyribose-5-phosphate (5'-dRP) group. In mammalian cells, gap filling and dRP removal are catalyzed by Pol ß, which belongs to the X family of DNA polymerases. In higher plants, the only member of the X family of DNA polymerases is Pol λ. Although it is generally believed that plant Pol λ participates in BER, there is limited experimental evidence for this hypothesis. Here we have characterized the biochemical properties of Arabidopsis thaliana Pol λ (AtPol λ) in a BER context, using a variety of DNA repair intermediates. We have found that AtPol λ performs gap filling inserting the correct nucleotide, and that the rate of nucleotide incorporation is higher in substrates containing a C in the template strand. Gap filling catalyzed by AtPol λ is most efficient with a phosphate at the 5'-end of the gap and is not inhibited by the presence of a 5'-dRP mimic. We also show that AtPol λ possesses an intrinsic dRP lyase activity that is reduced by mutations at two lysine residues in its 8-kDa domain, one of which is present in Pol λ exclusively and not in any Pol ß homolog. Importantly, we also found that the dRP lyase activity of AtPol λ allows efficient completion of uracil repair in a reconstituted short-patch BER reaction. These results suggest that AtPol λ plays an important role in plant BER.

Correct Nucleotide Selection Is Confined at the Binding Site of Polymerase Enzymes.

DNA polymerases (Pols) add incoming nucleotides (deoxyribonucleoside triphosphate (dNTPs)) to growing DNA strands, a crucial step for DNA synthesis. The insertion of correct (vs incorrect) nucleotides relates to Pols' fidelity, which defines Pols' ability to faithfully replicate DNA strands in a template-dependent manner. We and others have demonstrated that reactant alignment and correct base pairing at the Pols catalytic site are crucial structural features to fidelity. Here, we first used equilibrium molecular simulations to demonstrate that the local dynamics at the protein-DNA interface in the proximity of the catalytic site is different when correct vs incorrect dNTPs are bound to polymerase ß (Pol ß). Formation and dynamic stability of specific interatomic interactions around the incoming nucleotide influence the overall binding site architecture. This explains why certain Pols' mutants can affect the local catalytic environment and influence the selection of correct vs incorrect nucleotides. In particular, this is here demonstrated by analyzing the interaction network formed by the residue R283, whose mutant R283A has an experimentally measured lower capacity of differentiating correct (G:dCTP) vs incorrect (G:dATP) base pairing in Pol ß. We also used alchemical free-energy calculations to quantify the G:dCTP →G:dATP transformation in Pol ß wild-type and mutant R283A. These results correlate well with the experimental trend, thus corroborating our mechanistic insights. Sequence and structural comparisons with other Pols from the same family suggest that these findings may also be valid in similar enzymes.

Dna2 nuclease deficiency results in large and complex DNA insertions at chromosomal breaks.

Insertions of mobile elements1-4, mitochondrial DNA5 and fragments of nuclear chromosomes6 at DNA double-strand breaks (DSBs) threaten genome integrity and are common in cancer7-9. Insertions of chromosome fragments at V(D)J recombination loci can stimulate antibody diversification10. The origin of insertions of chromosomal fragments and the mechanisms that prevent such insertions remain unknown. Here we reveal a yeast mutant, lacking evolutionarily conserved Dna2 nuclease, that shows frequent insertions of sequences between approximately 0.1 and 1.5 kb in length into DSBs, with many insertions involving multiple joined DNA fragments. Sequencing of around 500 DNA inserts reveals that they originate from Ty retrotransposons (8%), ribosomal DNA (rDNA) (15%) and from throughout the genome, with preference for fragile regions such as origins of replication, R-loops, centromeres, telomeres or replication fork barriers. Inserted fragments are not lost from their original loci and therefore represent duplications. These duplications depend on nonhomologous end-joining (NHEJ) and Pol4. We propose a model in which alternative processing of DNA structures arising in Dna2-deficient cells can result in the release of DNA fragments and their capture at DSBs. Similar DNA insertions at DSBs are expected to occur in any cells with linear extrachromosomal DNA fragments.

Three-metal ion mechanism of cross-linked and uncross-linked DNA polymerase ß: A theoretical study.

In our recent publication, we have proposed a revised base excision repair pathway in which DNA polymerase ß (Polß) catalyzes Schiff base formation prior to the gap-filling DNA synthesis followed by ß-elimination. In addition, the polymerase activity of Polß employs the "three-metal ion mechanism" instead of the long-standing "two-metal ion mechanism" to catalyze phosphodiester bond formation based on the fact derived from time-resolved x-ray crystallography that a third Mg2+ was captured in the polymerase active site after the chemical reaction was initiated. In this study, we develop the models of the uncross-linked and cross-linked Polß complexes and investigate the "three-metal ion mechanism" vs the "two-metal ion mechanism" by using the quantum mechanics/molecular mechanics molecular dynamics simulations. Our results suggest that the presence of the third Mg2+ ion stabilizes the reaction-state structures, strengthens correct nucleotide binding, and accelerates phosphodiester bond formation. The improved understanding of Polß's catalytic mechanism provides valuable insights into DNA replication and damage repair.

The Impact of SNP-Induced Amino Acid Substitutions L19P and G66R in the dRP-Lyase Domain of Human DNA Polymerase ß on Enzyme Activities.

Base excision repair (BER), which involves the sequential activity of DNA glycosylases, apurinic/apyrimidinic endonucleases, DNA polymerases, and DNA ligases, is one of the enzymatic systems that preserve the integrity of the genome. Normal BER is effective, but due to single-nucleotide polymorphisms (SNPs), the enzymes themselves-whose main function is to identify and eliminate damaged bases-can undergo amino acid changes. One of the enzymes in BER is DNA polymerase ß (Polß), whose function is to fill gaps in DNA. SNPs can significantly affect the catalytic activity of an enzyme by causing an amino acid substitution. In this work, pre-steady-state kinetic analyses and molecular dynamics simulations were used to examine the activity of naturally occurring variants of Polß that have the substitutions L19P and G66R in the dRP-lyase domain. Despite the substantial distance between the dRP-lyase domain and the nucleotidyltransferase active site, it was found that the capacity to form a complex with DNA and with an incoming dNTP is significantly altered by these substitutions. Therefore, the lower activity of the tested polymorphic variants may be associated with a greater number of unrepaired DNA lesions.

Probing the Conformational Restraints of DNA Damage Recognition with ß-L-Nucleotides.

The DNA building blocks 2'-deoxynucleotides are enantiomeric, with their natural ß-D-configuration dictated by the sugar moiety. Their synthetic ß-L-enantiomers (ßLdNs) can be used to obtain L-DNA, which, when fully substituted, is resistant to nucleases and is finding use in many biosensing and nanotechnology applications. However, much less is known about the enzymatic recognition and processing of individual ßLdNs embedded in D-DNA. Here, we address the template properties of ßLdNs for several DNA polymerases and the ability of base excision repair enzymes to remove these modifications from DNA. The Klenow fragment was fully blocked by ßLdNs, whereas DNA polymerase κ bypassed them in an error-free manner. Phage RB69 DNA polymerase and DNA polymerase ß treated ßLdNs as non-instructive but the latter enzyme shifted towards error-free incorporation on a gapped DNA substrate. DNA glycosylases and AP endonucleases did not process ßLdNs. DNA glycosylases sensitive to the base opposite their cognate lesions also did not recognize ßLdNs as a correct pairing partner. Nevertheless, when placed in a reporter plasmid, pyrimidine ßLdNs were resistant to repair in human cells, whereas purine ßLdNs appear to be partly repaired. Overall, ßLdNs are unique modifications that are mostly non-instructive but have dual non-instructive/instructive properties in special cases.

Polymerase IV Plays a Crucial Role in Pollen Development in Capsella.

In Arabidopsis (Arabidopsis thaliana), DNA-dependent RNA polymerase IV (Pol IV) is required for the formation of transposable element (TE)-derived small RNA transcripts. These transcripts are processed by DICER-LIKE3 into 24-nucleotide small interfering RNAs (siRNAs) that guide RNA-directed DNA methylation. In the pollen grain, Pol IV is also required for the accumulation of 21/22-nucleotide epigenetically activated siRNAs, which likely silence TEs via post-transcriptional mechanisms. Despite this proposed role of Pol IV, its loss of function in Arabidopsis does not cause a discernible pollen defect. Here, we show that the knockout of NRPD1, encoding the largest subunit of Pol IV, in the Brassicaceae species Capsella (Capsella rubella), caused postmeiotic arrest of pollen development at the microspore stage. As in Arabidopsis, all TE-derived siRNAs were depleted in Capsella nrpd1 microspores. In the wild-type background, the same TEs produced 21/22-nucleotide and 24-nucleotide siRNAs; these processes required Pol IV activity. Arrest of Capsella nrpd1 microspores was accompanied by the deregulation of genes targeted by Pol IV-dependent siRNAs. TEs were much closer to genes in Capsella compared with Arabidopsis, perhaps explaining the essential role of Pol IV in pollen development in Capsella. Our discovery that Pol IV is functionally required in Capsella microspores emphasizes the relevance of investigating different plant models.