DSIF, a novel transcription elongation factor that regulates RNA polymerase II processivity, is composed of human Spt4 and Spt5 homologs

Abstract

We report the identification of a transcription elongation factor from HeLa cell nuclear extracts that causes pausing of RNA polymerase II (Pol II) in conjunction with the transcription inhibitor 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole (DRB). This factor, termed DRB sensitivity-inducing factor (DSIF), is also required for transcription inhibition by H8. DSIF has been purified and is composed of 160-kD (p160) and 14-kD (p14) subunits. Isolation of a cDNA encoding DSIF p160 shows it to be a homolog of the Saccharomyces cerevisiae transcription factor Spt5. Recombinant Supt4h protein, the human homolog of yeast Spt4, is functionally equivalent to DSIF p14, indicating that DSIF is composed of the human homologs of Spt4 and Spt5. In addition to its negative role in elongation, DSIF is able to stimulate the rate of elongation by RNA Pol II in a reaction containing limiting concentrations of ribonucleoside triphosphates. A role for DSIF in transcription elongation is further supported by the fact that p160 has a region homologous to the bacterial elongation factor NusG. The combination of biochemical studies on DSIF and genetic analysis of Spt4 and Spt5 in yeast, also in this issue, indicates that DSIF associates with RNA Pol II and regulates its processivity in vitro and in vivo.

Figure 1

Effect of DRB or H8 on transcription elongation. (A) Kinetically synchronized reactions proceeded as described in Materials and Methods except reaction products from the pTF3-6C₂AT-100 template were analyzed by a 20% sequencing gel. Numbers at the right indicate the positions of markers (nucleotides) and a solid bar corresponds to positions of RNA transcripts accumulated by addition of DRB. DRB was added to the reaction after the 45-min preincubation to various concentrations indicated above the gel. (B) Reactions proceeded as described in A except reactions were further incubated for 30 min after the 10-min initiation/elongation step in the absence [lanes 3,4, chase (−)] or presence [lanes 5,6, chase (+)] of cold ATP, CTP, and UTP (final concentration, 1 mm each). Reaction products from the pTF3-6C₂AT-100 template in the absence (lanes 1,3,5) or presence (lanes 2,4,6) of 50 μm DRB, which was added to the reaction after the 45-min preincubation, were analyzed by a 10% sequencing gel. (C) Reactions proceeded as described in A. Reaction products from the pTF3-6C₂AT-100 template in the presence of 50 μm DRB (lane 2) or 50 μm H8 (lane 3), which was added to the reaction after the 45-min preincubation, were analyzed by a 10% sequencing gel. Numbers at the right side indicate the positions of markers (nucleotides).

Figure 2

Purification of DSIF. (A) The purification scheme for DSIF is illustrated and is described in Materials and Methods. (B) Kinetically-synchronized transcription reaction was carried out with 2 μl of the concentrated P.5 and P.85 fractions from the phosphocellulose column (except lanes 1 and 2, which contained 4 μl of HeLa cell nuclear extracts). Even numbered lanes contained DRB to 50 μm. P.3 (6 μl) was added to a transcription reaction (lanes 5,6). P.3 (6 μl) was incubated at 68°C for 10 min prior to its addition to the reaction (lanes 7,8). (C) Fractions (1 μl) containing DSIF activity from a Superose 12 column were assayed for their ability to confer DRB sensitivity on a partially purified transcription system. Reactions were as described in Materials and Methods. Reaction products were resolved by 8% urea–PAGE. (In) Column input fraction (0.5 μl). (D) Aliquots (1 μl) of the Superose 12 column fractions were analyzed in a 12.5% SDS–polyacrylamide gel and the proteins were visualized by silver. Numbers to the right of gel indicate the position of protein molecular size standards. (E) Purified DSIF was assayed for its ability to confer DRB or H8 sensitivity on a partially purified transcription system. Reactions proceeded as described in B. DRB to 50 μm (lanes 2,4) or H8 to 50 μm (lanes 6,8) was added to the reaction after the 45-min preincubation. Lanes 3, 4, 7, and 8 contained 10 ng of purified DSIF.

Figure 3

Renaturation of DSIF activity. Purified DSIF (fraction 5 from Superose 12, 40 μl) was separated by a 12.5% SDS–polyacrylamide gel. Bands corresponding to 160- and 14-kD polypeptides were cut out of gel. The renaturation method was described in Materials and Methods. Aliquots (1 μl) of the fraction 5 were subjected to 12.5% SDS-PAGE and proteins were visualized by silver staining (lane 1). Renatured proteins were tested for their ability to confer DRB sensitivity on a partially purified reconstituted transcription system (lanes 2–9), and reaction products were resolved by 8% urea–PAGE. DRB was added at the final concentration of 50 μm to lanes 3, 5, 7, and 9. Transcription reactions were performed with purified DSIF (lanes 2,3, 10 ng). p160 and p14 were renatured together and added lanes 4 and 5 (7.5 ng of each protein), and renatured p14 (15 ng) alone was added to lanes 6 and 7, and renatured p160 (15 ng) alone was added to lanes 8 and 9.

Figure 4

Characterization of DSIF normal function in vitro. (A) Purified DSIF was added to a partially reconstituted transcription system (lanes 1–5) or a minimal transcription system (lanes 6–10) that was reconstituted by 10 ng of purified recombinant TBP, 30 ng of purified recombinant TFIIB, 16 ng of purified recombinant TFIIF, and 0.5 μl of highly purified RNA Pol II (Usuda et al. 1991). Kinetically synchronized reactions proceeded as described in Materials and Methods except reactions contained the pTF3-6C₂AT-100 template. Reaction products were analyzed by a 10% sequencing gel. Numbers to the right of gel indicate the position of markers (nucleotides) and solid bars correspond to positions of RNA transcripts accumulated by the addition of DSIF. (Lanes 2,7) 8 ng of DSIF; (lanes 3,8) 16 ng of DSIF; (lanes 4,9) 32 ng of DSIF; (lanes 5,10) 64 ng of DSIF. (B) Effect of DRB on transcription was tested in the absence or presence of DSIF. Reactions proceeded as described in A. Even numbered lanes contained DRB to 50 μm, which was added to the reaction after the 45-min preincubation. P.3 (6 μl) was added to a partially reconstituted transcription system (lanes 1,2) and DSIF (24 ng) was added to the system (lanes 5,6). Numbers to the right of gel indicate the position of markers (nucleotides) and solid bars correspond to positions of RNA transcripts induced by the addition of DSIF.

Figure 5

DSIF is composed of human Spt4 and Spt5 homologs. (A) Schematic comparison of DSIF p160 and yeast Spt5 and C. elegans Spt5 homologs. The DSIF p160 is compared with the C. elegans Spt5 homolog and yeast Spt5 proteins. The positions of the acidic domain at the amino terminus, highly conserved regions, and the carboxy-terminal repeats are marked by closed boxes. DSIF p160, C. elegans Spt5 homolog, and yeast Spt5 are highly related over their entire lengths. Numbers in boxes indicate percentages of identical amino acid within the region. (B) Recombinant Supt4h protein was tested for its ability to confer DRB sensitivity on a partially purified reconstituted transcription system. The DSIF p160 and rSupt4h were renatured independently as described in Materials and Methods. DRB was added to a final concentration of 50 μm lanes 2, 4, 6, and 8. Transcription reactions were performed with 10 ng of purified DSIF (lanes 1,2), 12.5 ng of p160 (lanes 3–6), 12.5 ng of rSupt4h (lanes 5–8). Reaction products were resolved by 8% urea–PAGE. (In) DSIF input fraction for SDS-PAGE before separation of p160 and p14 on the gel.

Figure 5

DSIF is composed of human Spt4 and Spt5 homologs. (A) Schematic comparison of DSIF p160 and yeast Spt5 and C. elegans Spt5 homologs. The DSIF p160 is compared with the C. elegans Spt5 homolog and yeast Spt5 proteins. The positions of the acidic domain at the amino terminus, highly conserved regions, and the carboxy-terminal repeats are marked by closed boxes. DSIF p160, C. elegans Spt5 homolog, and yeast Spt5 are highly related over their entire lengths. Numbers in boxes indicate percentages of identical amino acid within the region. (B) Recombinant Supt4h protein was tested for its ability to confer DRB sensitivity on a partially purified reconstituted transcription system. The DSIF p160 and rSupt4h were renatured independently as described in Materials and Methods. DRB was added to a final concentration of 50 μm lanes 2, 4, 6, and 8. Transcription reactions were performed with 10 ng of purified DSIF (lanes 1,2), 12.5 ng of p160 (lanes 3–6), 12.5 ng of rSupt4h (lanes 5–8). Reaction products were resolved by 8% urea–PAGE. (In) DSIF input fraction for SDS-PAGE before separation of p160 and p14 on the gel.

Figure 6

DSIF confers DRB sensitivity on transcription through its interaction with RNA Pol II. (A) Western blotting with α-p160. HeLa cell nuclear extracts, purified DSIF and DSIF p160 were analyzed by Western blotting with α-p160. (Lanes 1–3) 0.0625, 0.25, and 1 μl of nuclear extracts, respectively; (lane 4) 4 ng of p160 in DSIF; (lane 5) 4 ng of recombinant p160. (B) Immunodepletion of DSIF. Nuclear extracts were treated twice by the protein G resin [Ab(-) and the α-p160 affinity resin Ab(+)] as described in Materials and Methods. A kinetically synchronized transcription reaction with the pTF3-6C₂AT template was carried out with 4 μl of nuclear extracts treated by the resin. DRB was added to 50 μm (lanes 2,4,6). Purified DSIF (10 ng) was added in lanes 5 and 6. Reaction products were resolved by 8% urea–PAGE. (C,D) Western blotting of materials in immunodepletion experiments with α-p160, α-CTD, and specific antibodies against general transcription factors indicated in D. (IN) Nuclear extracts (1 μl); (UB) second unbound fraction (1 μl); (EL) eluate from the resin (2.5 μl). Ab(−) and Ab(+) were described in B. (E) Immunoprecipitation with α-CTD and Western blotting with α-p160 and α-CTD. Reactions were performed as described in Materials and Methods. (IC) Immunoadsorbed complexes (lanes 3,6) 0.3 μl, (lanes 4,7) 1 μl.

Figure 7

Characterization of the transcription elongation activity of DSIF in vitro. (A) Multiple sequence alignment of the KOW motif as found in bacterial NusG, ribosomal protein RL24, human DSIF p160, C. elegans Spt5 homolog and yeast Spt5. The amino acids of each protein are numbered next to the sequences. A BLAST search of the GenBank database with the amino-terminal 800 amino acids residues of p160 revealed that the central region of p160 shares significant sequence similarity with the KOW motif in Drosophila ambivalens NusG (P = 0.022), Sulfolobus solfataricus RL24 (P = 0.022), S. solfataricus NusG (P = 0.0012), and Methanococcus jannaschii NusG (P = 0.00019). The alignment and shading of multiple sequences was performed with Multiple sequence alignment with hierarchical clustering (Multalin, Corpet 1988) and BOXSHADE 3.21 software under default conditions. (B) Kinetically synchronized transcription reaction containing various concentrations of ribonucleoside triphosphates. Reactions were carried out with 2 μl of the concentrated P.5 and P.85 fractions from phosphocellulose column fractionation as described in Materials and Methods except that a 20 min incubation proceeded after addition of 60 μm ATP, indicated concentrations of CTP and UTP containing 5 μCi of [α-³²P]UTP (800 Ci/mmole). Numbers on the left indicate the position of markers (nucleotides). (C) Transcriptional activity of native and recombinant DSIF. A kineticallysynchronized transcription reaction was carried out as described inB except that an incubation after addition of 60 μm ATP, 10 μm CTP, and 1 μm UTP containing 5 μCi of [α-³²P]UTP (800 Ci/mmole) proceeded for the indicated times (top, in minutes) in the absence (lanes 1–3) or presence of 16 ng of purified DSIF (lanes 4–6), or 2 ng of p14 and 16 ng of p160 (lanes 7–9). The pTF3-6C₂AT template was used. Numbers at left indicate the position of markers (nucleotides). (D) Comparison of specific activities of p14 and a mixture of p14 and p160. Reactions were carried out as described in C except an incubation after addition of ribonucleoside triphosphates proceeded for the indicated times (top, in minutes) in the presence of 2 ng of p14 and 16 ng of p160 (lanes 2,4,6,8). Numbers at left indicate the position of markers (nucleotides). (E) DRB-sensitive transcription under limiting concentrations of ribonucleoside triphosphates. Reactions were carried out as described in D except a 20 min incubation proceeded after addition of ribonucleoside triphosphates in the presence of 2 ng of p14 and 16 ng of p160 (lanes 3,4). Lanes 2 and 4 contained 50 μm of DRB. Numbers on the left indicate the position of markers (nucleotides).