Spt4/Spt5 Transcription Elongation Factor

• Spt4/Spt5 (Spt4/5) is a heterodimeric transcription elongation factor.
• Spt4/5 tightly associates with RNAPII.
• Spt4/5 regulates both transcriptional elongation and co-transcriptional pre-mRNA processing.
• Mechanisms by which Spt4/5 functions are not well understood.
• Human and Drosophila Spt4/5 complexes can bind nucleic acids in vitro.
• Yeast Spt4/5 can bind in a sequence-specific manner to single-stranded RNA containing AAN repeats.
• RNA-binding determinants are Spt4 together with the NGN domain of Spt5.
• KOW domains are not required for RNA recognition.
• A new function is attributed to a Spt4/5 domain that associates directly with RNAPII.
• Spt4/5 plays a role in transcriptional control.

Introduction

• Transcription is a highly dynamic and regulated process in eukaryotes.
• Gene transcription is carried out by three multi-subunit RNA polymerases (RNAPI, II, and III).
• These RNA polymerases produce distinct classes of RNA.
• RNAPs are related through common evolutionary histories, structures, and mechanisms.
• RNAPs transcribe RNA in a DNA-template-dependent manner.
• The catalytic cores of multi-subunit polymerases are found in all living organisms.
• Catalytic cores are deeply conserved.
• General regulatory proteins assist and direct the activities of these RNAPs.
• Regulatory proteins generally exhibit significant variation in structure and function across the three kingdoms of life and between functional classes of polymerase.
• A family of transcriptional regulators displays the same degree of conservation as seen in RNAP: the Spt5/NusG family.
• These proteins are essential for life and regulate transcription elongation in eukaryotes, archaea, and bacteria.
• Strict conservation across all domains of life suggests that proteins carry out an ancient, core function in transcription.
• The details of that function, are largely unknown.
• Eukaryotic Spt5 is a large multi-domain protein consisting of an N-terminal acidic domain, a NusG N-terminal (NGN) domain, several Kyprides, Ouzounis, Woese (KOW) domains and a set of C-terminal repeats (CTRs).
• CTR sequence varies across species.
• The conserved core of Spt5 consists of the NGN domain and a single KOW domain, found in archaeal Spt5 and the bacterial homolog NusG.
• In eukaryotes and archaea (but not bacteria), Spt5 forms a noncovalent complex with a small zinc finger protein, Spt4 (RpoE" in archaea), via its NGN domain.
• The mechanism by which the Spt4/5 heterodimer regulates elongation is unknown.
• Recent structural studies suggest that the NGN domains of NusG and Spt4/5 bind directly to RNAP, bridging its central cleft.
• This arrangement effectively seals the DNA into the elongation complex and may prevent the disengagement of the DNA template from the transcribing RNAP, enhancing the processivity of the elongating RNAP.
• This model places Spt4/5/NusG in a location suitable for allosterically modulating the RNAP active site and interacting with nucleic acids in the transcription elongation complex (TEC).
• Archaeal Spt4/5 interacts with both ds- and ssDNA; NusG interacts with T-rich nontemplate DNA in the transcription bubble inducing pausing of the elongation complex; Spt5 interacts with nontemplate DNA in the transcription bubble and upstream of the elongating RNAP II, appearing critical for the ability of Spt5 to modulate transcription pausing or arrest.
• Spt4/5 may couple the activities of the TEC to pre-mRNA processing, in addition to regulating elongation,.
• The CTR region of Spt5 is required for normal elongation control and genetically interacts with the CTD of RNAPII.
• Through its regulated phosphorylation, the CTR of Spt5 may serve as a scaffold for the cooperative assembly of transcription and RNA processing factors.
• Spt5 associates with a variety of 5' and 3' RNA processing factors including RNA capping enzymes, polyadenylation factors and RNA cleavage factors.
• Mutations affect splicing, polyadenylation and nuclear export of mRNA; Spt5 also independently facilitates splicing of transcription.
• Spt4/5 may exert its effects on elongation or processing through direct contacts with the nascent transcript.
• NusG and Spt4/5 can interact with nucleic acids.
• Spt4/5 associates with nascent transcripts soon after they emerge from the elongating polymerase, may depend on transcript binding.
• The KOW domains of Spt5 may mediate RNA binding.
• KOW domains are found in RNA helicases and ribosomal proteins, and in rRNA processing factor Mtr4 they have been observed to directly contact RNA.

Spt4/5 binds to ssRNA in vitro

• The RNA-binding properties of Spt4/5 was tested.
• A recombinant yeast Spt4/5 complex containing the Spt5 NGN and all five KOW domains was used.
• This complex was tested for the ability to bind double-stranded DNA (dsDNA), single-stranded DNA (ssDNA), and single-stranded RNA (ssRNA) pentaprobes using electrophoretic mobility shift assays (EMSAs).
• Pentaprobes are overlapping 100-nt oligonucleotides encompassing all possible 5-nt sequences.
• Spt4/5 gave rise to a clear concentration-dependent shift in the ssRNA probe.
• No binding to the dsDNA probe was observed.
• Only small amounts of ssDNA shifting could be seen at higher protein concentrations.
• The Spt4/5 complex binds preferentially to ssRNA.

Spt4/5 is a sequence-specific ssRNA-binding protein

• It was determined whether Spt4/5 recognizes ssRNA with any sequence or structural specificity.
• Systematic evolution of ligands by exponential enrichment (SELEX) experiment used.
• It involved using a library containing a random 24-nt region.
• Selection of the library was performed against an Spt4/5 heterodimer in which Spt5 comprised the NGN domain and the first KOW domain, a version containing the NGN domain and the first 2 KOW domains and Spt4/55K, in order to identify any differences in binding specificity between the constructs.
• Ubiquitin was used as a negative control.
• Monitoring SELEX pool enrichment was performed through selection following every second round using RNA EMSA (REMSA) with the enrichment defined by the complete protein-dependent shift of the library for any given round.
• Enrichment of the Spt4/5 constructs was evident after two to seven rounds of selection in comparison to the negative control.
• The rate of library enrichment was related to the number of KOW domains in the Spt4/5 complexes.
• For each multi-KOW construct, the most significantly enriched motif (5_AA) is 14 nt long and consists of the sequence AANAANAANAANAA, where N denotes any nucleotide.
• Structure predictions carried out for sequences containing this motif revealed a distinct lack of secondary structure.
• REMSAs and microscale thermophoresis (MST) selectivity for the AA repeat sequence was observed; 5_AA bound to Spt4 and a truncated form of Spt5 containing just the NGN domain (Spt4/5NGN) with a dissociation constant of 0.65 µM, whereas there was no measurable binding to sequences lacking the AA repeats.
• The specificity of Spt4/5 for ssRNA is sequence-based.

The Spt4/5 NGN domain is sufficient for RNA binding

• The enriched SELEX motif had no difference between the three Spt4/5 constructs.
• It was considered that some of the KOW domains may be dispensable for RNA binding.
• To determine Spt4/5's essential RNA-binding core, REMSAs were carried out using a 5_AA-containing sequence obtained from the SELEX experiment (AArich, Supporting Information Table S1) as the target probe.
• Testing Spt4/5 complexes binding included: (i) Spt4 with just the NGN domain of Spt5 (Spt4/5NGN) or (ii) Spt4/51k, Spt4/52k, and Spt4/55K.
• The complex composed only of Spt4 and the isolated NGN domain (Spt4/5NGN) bound RNA.

The role of KOW domains in the RNA binding of Spt4/5

• Spt4/5NGN is the minimal RNA-binding region of the complex.
• This contradicts that the KOW domains are responsible for the nucleic acid-binding capabilities of Spt4/5.
• There appeared to be very weak interactions between RNA and the tandem KOW1+2, and KOW4+5 domains.
• These low-affinity interactions are most likely around 100mM
• The KOW domains do not make a significant contribution to the high-affinity and sequence-specific binding observed for the Spt4/5NGN complex.
• It is notable that Spt4/5 complexes with more than one KOW domain enriched for AA-bearing RNA sequences in the SELEX experiment in fewer rounds of selection.
• MST experiments on Spt4/5 heterodimers binding to 5_AA RNA show microscopic Kd (~1 µM) along with a clear increase in the calculated Hill coefficient as more KOW domains are added.

A single full-length Spt4/5 heterodimer binds AA-rich RNA with micromolar affinity.

• Protein on the RNA increases the affinity of a subsequent heterodimer for the same RNA molecule.
• single REMSA band observed for Spt4/55k binding to RNA would represent a single multimeric complex (i.e. 2:1 or more) that is formed cooperatively.
• Multiple bands observed for Spt4/5NGN represent a mixture of multiple protein-RNA assemblies (i.e. 1:1 and 2:1 complexes) binding without cooperativity.

Spt4/5 heterodimer formation is associated with conformational change

• Since neither Spt4 nor Spt5 bound RNA strongly in isolation, the possibility that a conformational change of one or both subunits is required for high-affinity binding was considered.
• Compared far-UV circular dichroism (CD) spectra.

The NGN domain of Spt4/5 contains a novel heterodimeric RRM

• In light of the RNA-binding activity seen for Spt4/5NGN, the topology of the NGN domain of Spt5 was compared to other RNA binding modules, revealing a similarity to the well-established RNA recognition motif (RRM).
• The NGN domain of yeast Spt5 contains the typical RRM arrangement of four antiparallel ẞ-strands packed against two a-helices although its al-helix is extended and it has an additional C-terminal a-helix (3) that partially obscures what would be the RNA-binding surface of a canonical RRM.
• Spt4 interacts with the NGN domain of Spt5 through the alignment of their respective ẞ-sheets.
• Most canonical RRMs interact with ssRNA through three highly conserved aromatic rings located within two RNP motifs on the ẞ-sheet surface.
• Spt5 NGN lacks these canonical RNP sequences and conserved surface aromatics suggesting that Spt4/5 binds RNA in a noncanonical manner.