TECHNICAL COMMENT◥

TRANSCRIPTION

Comment on “Mechanism ofeukaryotic RNA polymerase IIItranscription termination”Aneeshkumar G. Arimbasseri,1 George A. Kassavetis,2 Richard J. Maraia*1,3

Nielsen et al. (Reports, 28 June 2013, p. 1577) characterized their RNA polymerase III(Pol III) preparation and concluded that it requires an RNA hairpin/duplex structure forterminating transcription. We could not corroborate their findings using bona fide Pol IIIfrom two laboratory sources.We show that Pol III efficiently terminates transcription in theabsence of a hairpin/duplex in vitro and in vivo.

Abiochemical analysis of Saccharomycescerevisiae–derived RNA polymerase III(Pol III) by Nielsen et al. (1) showed thatalthough a fraction of the enzyme pausedat the known termination signal, oligo(T),

it failed to release the transcript in the absenceof a terminator-proximal RNA hairpin or otherRNA duplex structure. This work contrasts withthat of others that concluded that no hairpin ordyad structure is required for termination by Pol

III (2, 3). Soren Nielsen and Nikolay Zenkin gra-ciously provided us with a sample of their Pol III.Our analyses reveal that the Nielsen Pol IIIfails to terminate in the absence of a terminator-proximal RNA hairpin or other RNA duplexstructure, as they reported, but this polymerasepreparation differs substantially from Pol IIIpurified independently by A. Arimbasseri andG. Kassavetis using different methods in differ-ent laboratories.Using the template specified by Nielsen et al.

to produce a transcript lacking secondary struc-ture, we found that our S. cerevisiae Pol III effi-ciently released transcripts at oligo(T). Furtheranalyses indicated that the disparity must lie inthe polymerases rather than variables of the as-says. This was confirmed by direct comparisonsof the two polymerases, the Nielsen Pol III (NP)and ours (AP) (4), using two different templatesand two immobilization methods: (i) through the

RESEARCH

524-c 1 AUGUST 2014 • VOL 345 ISSUE 6196 sciencemag.org SCIENCE

1Intramural Research Program of the Eunice Kennedy ShriverNational Institute of Child Health and Human Development,National Institutes of Health, Bethesda, MD, USA. 2Division ofBiological Sciences, University of California, San Diego, CA,USA. 3Commissioned Corps, U.S. Public Health Service.*Corresponding author. E-mail: maraiar@mail.nih.gov

Fig. 1. Direct comparisons of NP polymeraseand AP Pol III for termination and transcriptionfactor TFIIIB+TFIIIC–dependent initiation. (A)Elongation complexes (EC) were assembled on tem-plates A and B and immobilized, through the RNApolymerase His6 tag, on Ni-NTA (nitrilotriacetic acid)agarose. NTPswere added, incubated for 10min, andthe released (R) andbound (B) transcripts separated,as indicated above each lane. (B) Comparison ofelongation complex formation and transcription byAPandNPon immobilized template B.The quantitiesof AP and NP used are the same as in (A); reactionswere for 1 min. Lanes 1 to 4 and 9 to 12 were done asdescribed (4),whereas lanes 5 to 8 and 13 to 16 weredone in the absence of EDTA and 40°C incubation.(C) Comparison of transcript release by three dif-ferent preparations of Pol III. AP was purified onNiNTA agarose (4) (used in other figures), whereasuntagged KP was highly purified by sequential ion-exchange chromatography (6).The quantities of APand NP used are the same as in (A) and (B). Re-leased, bound, and total (T) transcripts are indicated.(D) Pol III–specific transcription assay. Pre-initiationcomplexes were formed by incubating a SUP4 tRNAgene plasmid with S. cerevisiae S100 extract de-ficient in endogenous Pol III activity, followed by theaddition of AP or NP, as indicated above the lanes.Reactions were started by the addition of NTPs(with a-[32P] GTP) and MgCl2 and incubated for20 min.

on April 25, 2020

http://science.sciencem

ag.org/D

ownloaded from

polymerase His6 tag (4) (Fig. 1A) or (ii) throughthe 5′-biotin–tagged nontemplate strand DNA(Fig. 1B) (1). For the experiment shown in Fig. 1A,templates A and B contain differently positioned9T and 10T terminators, respectively, producing5′-32P–labeled RNA primer-directed transcriptspredicted to lack secondary structure; AP ef-ficiently released transcripts at the T tracts ofboth templates (lanes 1 and 3). In contrast, NPproduced more read-through transcripts (RT),and most of the transcripts with 3′ ends at the Ttract were in the bound fractions (lanes 6 and 8).For the experiment shown in Fig. 1B, we usedimmobilized DNA, unlabeled RNA primer anda-[32P]GTP (guanosine triphosphate) incorpora-tion (1), and the same amounts of polymerases asin Fig. 1A. In this assay, NP was far more activethan AP and also produced more RT. Transcriptspaused at oligo(T) were not released by NP butwere efficiently released by AP (Fig. 1B). It is note-worthy that substantial RT was also producedby NP for transcripts with a hairpin and on evenlonger T tracts [Fig. 2C in (1)], reflecting unex-pectedly low termination efficiency relative towhat would result with other preparations ofPol III [with comparable nucleoside triphosphate(NTP) concentrations; see (3–5)]. We also com-pared NP and AP with an untagged preparationof Pol III, KP (6), using immobilized DNA (Fig.1C). KP was nearly indistinguishable from AP,whereas NP again produced mostly bound tran-scripts, arrested at oligo(T). It is likely importantthat NP was substantially more active than AP inthe immobilized DNA assay relative to the immo-bilized polymerase assay. NP also differed in

read-through beyond its oligo(T) elongation-arrest site compared with read-through beyondthe oligo(dT) pause/release site by AP and KP.Thus, NP is very different from AP and KP. Its ac-tivity also differs from Pol IIID, a form that lacksits C53/C37 subcomplex, because Pol IIID efficientlyreleases transcripts at long, 9T terminators (4).We also compared AP and NP for factor-

dependent initiation of transcription of the SUP4tRNA gene by assembling initiation complexeswith an S. cerevisiae S100 extract deficient inPol III activity (from Pol IIID–expressing cells(4)). Addition of AP produced the expected 104-nucleotide (nt) transcript (6) and smaller productsconsistent with pre-tRNA processing in cellularextracts (7), whereas NP produced substantiallyless (Fig. 1D, lanes 4 and 5). Mixing AP and NPyielded levels similar to AP alone (lane 6). BecauseNP terminates efficiently on the SUP4 gene (1),the low ratio of specific to nonspecific transcrip-tion observed with NP indicates an additionaldifference between the Pol III preparations.Examining the 280 Pol III transcribed genes

in the S. cerevisiae genome for potential RNA sec-ondary structure within 12 nt upstream of theirterminators, Nielsen et al. proposed that their RNAstructure model is relevant to Pol III terminationin vivo (1). We employed an Schizosaccharomycespombe suppressor-tRNA reporter system to exam-ine this issue directly (8, 9). Figure 2A depictspSer7T, which has a short 3′ trailer betweenthe end of the tRNA module and the 7T termi-nator. The nearest RNA duplex structure forpSer7T has its 3′ end 9 nt upstream from theterminator. Construct pSer7T-US contains a 26-nt

CA-repeat unstructured spacer (US) in the 3′trailer, and pSer7T-HP contains a 26-nt spacercontaining a G+C–rich dyad sequence formingan 8-bp hairpin (HP). Structure prediction of thefull-length sequences (10) confirmed that the USwould be single-stranded, whereas HP wouldform an independent hairpin. The control con-structs pSer3T-US and pSer3T-HP have only 3Ts in the terminator region and should produceRT transcripts only; the downstream com-plementary read-through (cRT) region is com-plementary to, and can base pair with, thetRNA sequence, thereby blocking normal pro-cessing (8). The three 7T constructs producedhigh levels of suppression, reflected by whitecolony/patch color (Fig. 2B, top row). The 3T con-structs demonstrated that suppression is notachieved in the absence of termination. Northernblot analysis shown in Fig. 2C using an intron-based probe of the suppressor-tRNA pSer con-structs in Fig. 2B displayed the expected primarynascent RNAs of different lengths reflected inFig. 2A (Fig. 2C). Previous characterizations in-dicate that the major—i.e., lowest—band designatedby the arrow in Fig. 2C is an intron-containingintermediate substrate for mature tRNA produc-tion (11). Comparable levels of this band in lanes1 to 3 reflect the comparable suppression activitiesin Fig. 2B. These findings provide strong evidencethat a terminator-proximal hairpin is largely in-consequential to suppressor tRNA production byPol III in vivo.A mechanistic link between termination and

reinitiation is widely believed to underlie thehigh productivity of Pol III, which is important

SCIENCE sciencemag.org 1 AUGUST 2014 • VOL 345 ISSUE 6196 524-c

Fig. 2. Pol III termination is independent of tran-script terminator–proximal RNA structure in vivo.(A) Schematic of the different suppressor tRNAconstructs used. pSer7T is the parent construct(8). Promoter elements and other features areindicated (arrow, start site; box A; intron; box B;7T or 3T terminators; T2, read-through termina-tor). The dashed vertical lines delineate the 5′and 3′ matured/processed ends of the tRNA. Anunstructured spacer (US) or a spacer that formsa stable hairpin (HP) was introduced within the 3′trailer, immediately upstream of the terminator.On the right side is a schematic of the resultantpredicted pre-tRNA products. (B) Patch suppres-sion assay analysis of S. pombe colonies carryingthe different suppressor constructs. Red color indi-cates lack of suppression, and white color indi-cates efficient suppression. (C) Northern blotanalysis of the nascent tRNAs produced by thedifferent suppressor constructs using an intron-specific probe. 7T- and 7T-US/HP designations tothe left side indicate the position of the nascent 7T-terminated pre-tRNAs of pSer7Tand pSer7T-US/HPconstructs, respectively. “Endo” indicates the nas-cent transcript of the endogenous dimeric tRNAser-tRNAi-met gene. 3T-T2 indicates the 3′ extended transcript generated by read-through transcription of the constructswiththe 3T terminator.The arrow indicates a partially processed pre-tRNA intermediate before nuclear export.The differencesin the profiles in lanes 2 and 3 likely reflect alternative 3′ processing pathways (15).The lower panel shows the same blotprobed for the pol II-transcribed U5 snRNA as a sample recovery control.

RESEARCH | TECHNICAL COMMENTon A

pril 25, 2020

http://science.sciencemag.org/

Dow

nloaded from

for proliferation, growth, and development (12, 13).High productivity and additional features alsomake Pol III a compelling choice for therapeuticexpression of small interfering RNAs for RNAinterference (RNAi) and other purposes (14). Inthis regard, our data indicate that a terminator-proximal hairpin is not essential for termina-tion, either in vivo or in vitro.

REFERENCES AND NOTES

1. S. Nielsen, Y. Yuzenkova, N. Zenkin, Science 340, 1577–1580 (2013).2. D. F. Bogenhagen, D. D. Brown, Cell 24, 261–270 (1981).3. X. Wang, W. R. Folk, J. Biol. Chem. 269, 4993–5004 (1994).

4. A. G. Arimbasseri, R. J. Maraia, Mol. Cell. Biol. 33, 1571–1581(2013).

5. H. Matsuzaki, G. A. Kassavetis, E. P. Geiduschek, J. Mol. Biol.235, 1173–1192 (1994).

6. G. A. Kassavetis, B. R. Braun, L. H. Nguyen, E. P. Geiduschek,Cell 60, 235–245 (1990).

7. D. R. Engelke, P. Gegenheimer, J. Abelson, J. Biol. Chem. 260,1271–1279 (1985).

8. M. Hamada, A. L. Sakulich, S. B. Koduru, R. J. Maraia, J. Biol.Chem. 275, 29076–29081 (2000).

9. Y. Huang, R. V. Intine, A. Mozlin, S. Hasson, R. J. Maraia, Mol.Cell. Biol. 25, 621–636 (2005).

10. M. Zuker, Science 244, 48–52 (1989).11. V. Cherkasova, J. Bahler, D. Bacikova, K. Pridham, R. J. Maraia,

Mol. Biol. Cell 23, 480 (2011).

12. A. G. Arimbasseri, K. Rijal, R. J. Maraia, Transcription 4, (2013).13. L. Marshall, S. J. Goodfellow, R. J. White, PLOS Biol. 5, e286

(2007).14. M. C. Yen et al., Cancer Gene Ther. 20, 351–357 (2013).15. R. J. Maraia, T. N. Lamichhane, WIRES RNA 2, 362–375 (2011).

ACKNOWLEDGMENTS

We thank N. Zenkin and S. Nielsen for NP3, and E. P. Geiduschek,M. Kashlev, R. Landick, and A. Vannini for discussion andcomments. This work was supported by direct funding from theIntramural Research Program of the Eunice Kennedy ShriverNational Institute of Child Health and Human Development.

24 March 2014; accepted 25 June 201410.1126/science.1253783

524-c 1 AUGUST 2014 • VOL 345 ISSUE 6196 sciencemag.org SCIENCE

RESEARCH | TECHNICAL COMMENTon A

pril 25, 2020

http://science.sciencemag.org/

Dow

nloaded from

Comment on ''Mechanism of eukaryotic RNA polymerase III transcription termination''Aneeshkumar G. Arimbasseri, George A. Kassavetis and Richard J. Maraia

DOI: 10.1126/science.1253783 (6196), 524.345Science 

ARTICLE TOOLS http://science.sciencemag.org/content/345/6196/524.3

CONTENTRELATED

http://science.sciencemag.org/content/sci/340/6140/1577.fullhttp://science.sciencemag.org/content/sci/345/6196/524.4.full

REFERENCES

http://science.sciencemag.org/content/345/6196/524.3#BIBLThis article cites 15 articles, 8 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2014, American Association for the Advancement of Science

on April 25, 2020

http://science.sciencem

ag.org/D

ownloaded from