Anomalous HIV-1 RNA, How Cap-Methylation Segregates Viral Transcripts by Form and Function

Abstract

The acquisition of m⁷G-cap-binding proteins is now recognized as a major variable driving the form and function of host RNAs. This manuscript compares the 5'-cap-RNA binding proteins that engage HIV-1 precursor RNAs, host mRNAs, small nuclear (sn)- and small nucleolar (sno) RNAs and sort into disparate RNA-fate pathways. Before completion of the transcription cycle, the transcription start site of nascent class II RNAs is appended to a non-templated guanosine that is methylated (m⁷G-cap) and bound by hetero-dimeric CBP80-CBP20 cap binding complex (CBC). The CBC is a nexus for the co-transcriptional processing of precursor RNAs to mRNAs and the snRNA and snoRNA of spliceosomal and ribosomal ribonucleoproteins (RNPs). Just as sn/sno-RNAs experience hyper-methylation of m⁷G-cap to trimethylguanosine (TMG)-cap, so do select HIV RNAs and an emerging cohort of mRNAs. TMG-cap is blocked from Watson:Crick base pairing and disqualified from participating in secondary structure. The HIV TMG-cap has been shown to license select viral transcripts for specialized cap-dependent translation initiation without eIF4E that is dependent upon CBP80/NCBP3. The exceptional activity of HIV precursor RNAs secures their access to maturation pathways of sn/snoRNAs, canonical and non-canonical host mRNAs in proper stoichiometry to execute the retroviral replication cycle.

Keywords: NCBP3; RNA virus; epigenetic modification; internal ribosome entry; junD; ribosome scanning; specialized translation; trimethylguanosine (TMG) cap.

Figure 1

Motifs within the HIV 5′-untranslated region (UTR). (A) Linear arrangement of TAR, Tat trans-activation responsive element, magenta; PolyA, blue; U5, Unique 5′ region, green; PBS-segment containing the primer binding site (PBS), tan; CES, core encapsidation signal, black, Numbers indicate position in the 5′-UTR. Lower line: TSS, transcription start site; DIS, dimerization initiation site inclusive of SL1; SD, splice donor; AUG, gag start codon. (B) Secondary structure model. Select nucleotides pairings coordinate TAR-PolyA stem loops and PolyA-U5 stem.

Figure 2

Precursor RNA fate is determined through the co-transcriptional recruitment of RNA binding proteins and subsequent rearrangement of RNP (ribonucleoprotein) components. Nascent RNA shown with paused RNA polymerase II, Blue. CE, capping enzyme. * CRM1 chaperone activity. Blue shape with black wavy line, RNA polymerase II paused on chromatin. NELF, negative elongation factor; CE, capping enzyme; pTEFb, positive transcription elongation factor b; NPC, nuclear pore complex. (A) Precursors of mRNA templates are completely spliced and licensed for nuclear export and CBP80/CBP20 nuclear cap-binding complex (CBC) exchange to eIF4E for canonical cap-dependent steady state translation. (B,C) snRNA and snoRNA gene products engage distinct RNA fate pathways that culminate in nucleolar trafficking and assembly into spliceosomes and ribosome. (D) Early HIV precursors become completely spliced mRNA templates by removal of alternative introns (dashed lines) that are licensed for the mRNA transport receptor NXF1/NXT1 and CBC exchange to eIF4E. These mRNAs form polysomes that translated Tat, Rev, Nef. (E) Late HIV precursors experience binding of Tat to TAR that trans-activates co-transcriptional capping and pTEFb activity. Rev binding to RRE activity results in 5′-cap hypermethylation, CRM1 nuclear export and specialized translation unaffected by mTOR. These mRNAs form polysomes that translated viral accessory and structural proteins. (F) The dashed line indicates a minority of unspliced RNA is bound by myristoylated (Myr, mauve bars) Gag polyprotein (mauve circles) and experienced dimerization at the plasma membrane. The dimeric RNA serves as genomic RNP that packaged into progeny virions.

Figure 3

Structural basis for 7-methylguanosine (m⁷G) base recognition by cap-binding proteins. (A) (Transcription start site of nascent RNAPII transcripts experience 5′-5′ linkage to guanosine that is methylated at the N⁷ position to form the m⁷G-cap. (B) Space-filling models present the conserved m⁷G-cap binding pockets of: CBP20 of the CBP20/CBP80 heterodimer; cytoplasmic cap binding protein, eIF4E; trimethylguanosine synthase, TGS1; poly(A)-specific ribonuclease, PARN.

Figure 4

Comparison of the cap-dependent translation pathways of HIV early and late mRNAs that are licensed by m⁷G-cap or m^2,2,7-trimethylguanosine cap(TMG), respectively. (Left) m⁷G-capped early mRNAs assemble nuclear cap-binding complex (CBC) composed of CBP80/CBP20. After nuclear export, CBC exchanges to eIF4E. eIF4E recruits eIF4G/eIF4A for eIF4E-dependent translation initiation that is modulated by mTOR (mechanistic target of rapamycin). Activated mTOR (+) hyper-phosphorylates 4E-binding protein (BP), blocking its interaction with eIF4E. mTOR inhibition (−) upregulates hypo-phosphorylated (hypo) 4E-BP for allosteric binding with eIF4E that blocks eIF4G interaction and promptly halts eIF4E-dependent translation. (Right) m⁷G-capped late mRNAs contain the Rev-responsive element (RRE) (gray RNA element) and require Rev binding (gray shape). The RRE-dependent mRNAs tether DHX9/RNA helicase A (RHA) through the post-transcriptional control element (PCE) and experience non-canonical CBC exchange to trimethylguanosine synthase 1 (TGS1). TGS1 hypermethylates m⁷G-cap to TMG-cap and exchanges to CBP80-NCBP3 for pecialized translation that is unaffected by mTOR.

Figure 5

The hypermethylation of the m⁷G-cap eliminates capacity for Watson-Crick base pairing and for CBC/eIF4E/TGS1 binding. (A) The ^2,2,7-trimethylguanosine (TMG-cap) appended to the 5′-terminus of RNA. (B) Space-filling model of the TMG-binding pocket of Snurportin1 and NCBP3 modeled on the structure of PARN. In each, the TMG-cap exhibits base stacking with the aromatic ring of tryptophan (W) in the polypeptide and guanine in the cognate RNA.

Figure 6

The HIV early and late mRNAs engage independent translation pathways that interdependently control the expression of viral regulatory proteins and structural/accessory proteins. (Left) HIV early mRNAs are fully processed mRNAs that undergo canonical eIF4E-dependent translation to Tat, Rev and Nef. Hypo-phosphorylated (hypo) 4E-BP allosterically binds eIF4E and blocks eIF4G interaction and halts the translation initiation. (Right) Trimethylguanosine (TMG)-cap of the Rev/Rev-responsive element (RRE)-dependent mRNAs licenses specialized translation independently of eIF4E activity. HIV unspliced and singly spliced (US and SS) mRNAs engage Rev at the RRE (gray protein binding gray RNA element) and subvert canonical CBC exchange to eIF4E and engages CBP80/NCBP3 Vpr upregulates hypo-4E-BP [75] and inhibits eIF4E-dependent translation of host proteins and Tat, Rev, Nef. The specialized translation of the Rev/RRE-dependent TMG-capped mRNAs endures until lack of Tat/Rev activity curtails biosynthesis of the US/SS RNA [20]. The downregulation of the TMG-capped mRNAs encoding viral accessory/structural protein and attenuates virus proliferation [20].

Figure 7

Predicted tertiary structure context of the HIV Capof the HIV ^NL4−3 5′-UTR given transcription start site heterogeneity of one guanosine (G) or GGG. Zoom-in on the junction of TAR (magenta)-PolyA (blue)-U5 (green) helices. Cap, magenta atom; G104, blue atom and U105, green atom. (A) Cap-G at the 5′ terminus. (B) Cap-GGG at the 5′ terminus repositioned Cap ~26 Angstroms from the junction of PolyA-U5. Rotation of these structures is provided in Video S1.